Skip to main content

Full text of "Year book - Carnegie Institution of Washington"

See other formats

-/? •3>-rSrV(? £& 

^~sy *&V J>y*.* c*»*'<- fZ?'^ 


Year Book 61 

Digitized by the Internet Archive 

in 2012 with funding from 

LYRASIS Members and Sloan Foundation 


Year Book 

July 1, 1961 - June 30, 1962 


'"**' ^^"^gaJOOawoooo rnm . K ^^ rf tSX'i^'^SSK^K^Z^^SSX 

Sixtieth Anniversary 

Library of Congress Catalog Card Number 3-16716 
Garamond Press, Baltimore, Maryland 



Officers and Staff v 

Report of the President 1 

Reports of Departments and Special Studies 

Mount Wilson and Palomar Observatories 3 

Geophysical Laboratory 51 

Department of Terrestrial Magnetism 209 

Committee on Image Tubes for Telescopes 295 

Department of Plant Biology 303 

Department of Embryology 367 

Department of Genetics 435 

Bibliography 477 

Administrative Reports 479 

Report of the Executive Committee 481 

Report of Auditors 483 

Abstract of Minutes of the Sixty-Fourth Meeting of the 

Board of Trustees 497 

Articles of Incorporation 499 

By-Laws of the Institution 503 

Index 507 


President and Trustees 


Caryl P. Haskins 


Barklie McKee Henry 

Henry S. Morgan 
V ice-Chairman 

Garrison Norton 

Robert Woods Bliss 1 
Amory H. Bradford 
Omar N. Bradley 
Vannevar Bush 
Walter S. Gifford 
Carl J. Gilbert 
Crawford H. Greenewalt 
Caryl P. Haskins 
Barklie McKee Henry 
Alfred L. Loomis 
Robert A. Lovett 
Keith S. McHugh 
Margaret Carnegie Miller 
Henry S. Morgan 
Seeley G. Mudd 
William I. Myers 
Garrison Norton 
Richard S. Perkins 
Elihu Root, Jr. 
William W. Rubey 
Henry R. Shepley 
Charles P. Taft 
Juan T. Trippe 
James N. White 
Robert E. Wilson 

i Died April 19, 1962. 

Trustees continued 


Keith S. McHugh, Chairman 
Alfred L. Loomis 
Juan T. Trippe 



Henry S. Morgan, Chairman 
Amory H. Bradford 
Walter S. Gifford 
Caryl P. Haskins 
Barklie McKee Henry 
Robert A. Lovett 
Garrison Norton 
James N. White 
Robert E. Wilson 

Omar N. Bradley, Chairman 
Henry S. Morgan 
Garrison Norton 
James N. White 



James N. White, Chairman 
Walter S. Gifford 
Alfred L. Loomis 
Henry S. Morgan 
Richard S. Perkins 
Elihu Root, Jr. 

Seeley G. Mudd, Chairman 
Amory H. Bradford 
Crawford H. Greene wait 
Elihu Root, Jr. 


Alfred L. Loomis, Chairman 
Margaret Carnegie Miller 
William I. Myers 
Charles P. Taft 


Amory H. Bradford, Chairman 
Barklie McKee Henry 
Richard S. Perkins 
Charles P. Taft 


Juan T. Trippe, Chairman 
Barklie McKee Henry 
Richard S. Perkins 
Robert E. Wilson 


Former Presidents and Trustees 


Daniel Coit Gilman, 1902-1904 Robert Simpson Woodward, 1904-1920 

John Campbell Merriam, President 1921-1938; President Emeritus 1939-1945 
Vannevar Bush, 1939-1955 


Alexander Agassiz 


Henry Cabot Lodge 


George J. Baldwin 


Seth Low 


Thomas Barbour 


Wayne MacVeagh 


James F. Bell 


Andrew W. Mellon 


John S. Billings 


Roswell Miller 


Robert Woods Bliss 


Darius O. Mills 


Lindsay Bradford 


S. Weir Mitchell 


Robert S. Brookings 


Andrew J. Montague 


John L. Cadwalader 


William W. Morrow 


William W. Campbell 


William Church Osborn 


John J. Carty 


James Parmelee 


Whitefoord R. Cole 


Wm. Barclay Parsons 


Frederic A. Delano 


Stewart Paton 


Cleveland H. Dodge 


George W. Pepper 


William E. Dodge 


John J. Pershing 


Charles P. Fenner 


Henning W. Prentis, Jr. 


Homer L. Ferguson 


Henry S. Pritchett 


Simon Flexner 


Gordon S. Rentschler 


W. Cameron Forbes 


David Rockefeller 


James Forrestal 


Elihu Root 


William N. Frew 


Julius Rosenwald 


Lyman J. Gage 


Martin A. Ryerson 


Cass Gilbert 


Theobald Smith 


Frederick H. Gillett 


John C. Spooner 


Daniel C. Gilman 


William Benson Storey 


John Hay 


Richard P. Strong 


Myron T. Herrick 


William H. Taft 


Abram S. Hewitt 


William S. Thayer 


Henry L. Higginson 


James W. Wadsworth 


Ethan A. Hitchcock 


Charles D. Walcott 


Henry Hitchcock 


Frederic C. Walcott 


Herbert Hoover 


Henry P. Walcott 


William Wirt Howe 


Lewis H. Weed 


Charles L. Hutchinson 


William H. Welch 


Walter A. Jessup 


Andrew D. White 


Frank B. Jewett 


Edward D. White 


Samuel P. Langley 


Henry White 


Ernest 0. Lawrence 


George W. Wickersham 


Charles A. Lindbergh 


Robert S. Woodward 


William Lindsay 


Carroll D. Wright 


Under the original charter, from the date of organization until April 28, 1904, the following were 
ex officio members of the Board of Trustees : the President of the United States, the President of the 
Senate, the Speaker of the House of Representatives, the Secretary of the Smithsonian Institution, 
and the President of the National Academy of Sciences . 




813 Santa Barbara Street 
Pasadena, California 

Ira S. Bowen, Director 

Horace W. Babcock, Asst. Director 

Halton C. Arp 

William A. Baum 

Armin J. Deutsch 

Olin J. Eggen 

Jesse L. Greenstein 

Robert F. Howard 

Robert P. Kraft 

Guido Munch 

J. Beverley Oke 

Allan R. Sandage 

Maarten Schmidt 

Otto Struve 

Olin C. Wilson 

Fritz Zwicky 



2801 Upton Street, N.W. 
Washington 8, D. C. 

Philip H. Abelson, Director 
Francis R. Boyd, Jr. 
Felix Chayes 
Sydney P. Clark, Jr. 1 
Gordon L. Davis 
Gabrielle Donnay 
Joseph L. England 
Hugh J. Greenwood 
Thomas C. Hoering 
Gunnar Kullerud 
Patrick L. Parker 2 
J. Frank Schairer 
George R. Tilton 
Hatten S. Yoder, Jr. 

52 U Broad Branch Road, N.W. 
Washington 15, D. C. 

Merle A. Tuve, Director 
L. Thomas Aldrich 
Ellis T. Bolton 
Roy J. Britten 
Bernard F. Burke 
Dean B. Cowie 
John W. Firor 3 
Scott E. Forbush 
W. Kent Ford, Jr. 
Stanley R. Hart 4 
Norman P. Heydenburg 
Brian J. McCarthy 
Richard B. Roberts 
T. Jefferson Smith 5 
John S. Steinhart 
Georges M. Temmer 
Harry W. Wells 6 

1 Resigned June 30, 1962. 

2 Appointed September 1, 1961. 
s Through September 15, 1961. 

4 From September 1, 1961. 

5 From June 1, 1962. 

6 On leave of absence to serve as State 
Department Scientific Attache" sta- 
tioned at Rio de Janeiro, Brazil, 
through April 30, 1962. 


Staff continued 


Stanford, California 

C. Stacy French, Director 
Jeanette S. Brown 
David C. Fork 
William M. Hiesey 
Harold W. Milner 
Malcolm A. Nobs 


115 West University Parkway 
Baltimore 10, Maryland 

James D. Ebert, Director 
David W. Bishop 
Bent G. Boving 
Robert K. Burns 1 
Robert L. DeHaan 
Irwin R. Konigsberg 
Elizabeth M. Ramsey 
Mary E. Rawles 


Cold Spring Harbor 
Long Island, New York 

Berwind P. Kaufmann, Director 1 
Elizabeth Burgi 
Helen Gay 
Alfred D. Hershey 
Barbara McClintock 
Margaret R. McDonald 

i Retired June 30, 1962. 


Staff continued 


1530 P Street, N.W., Washington 5, D. C. 

Caryl P. Haskins President 

Edward A. Ackerman Executive Officer 

Ruth L. McCollum Assistant to the President 1 

Marjorie H. Walburn Acting Assistant to the President 2 

Ailene J. Bauer Director of Publications 

Lucile B. Stryker Editor 

James W. Boise Bursar; Secretary-Treasurer Retirement Trust 

Kenneth R. Henard Assistant Bursar; Assistant Treasurer Retirement Trust 

Donald J. Patton Administrative Associate 

James F. Sullivan Assistant to the Bursar 

Richard F. F. Nichols Executive Secretary to the Finance Committee 

Marshall Hornblower Counsel 

Staff Members in Special Subject Areas 

Tatiana Proskouriakoff 
Anna 0. Shepard 

1 Retired June 30, 1962. 

2 Effective from May 21, 1962. 

Staff continued 


Carnegie Research Associates 

William A. Arnold 

Oak Ridge National Laboratory 

J. D. McGee 

Imperial College of Science and Technology, University of London 

Jan H. Oort 

Leiden Observatory, The Netherlands 

Paul Ramdohr 

Heidelberg University 

C. E. Tilley 

Cambridge University 

Evelyn M. Witkin 

State University of New York 

Research Associates of Carnegie Institution of Washington 

Louis B. Flexner 

University of Pennsylvania 

John H. Holland 

Logic of Computers Group, University of Michigan 

Peter Milner 

The Department of Psychology, McGill University, Montreal 

Harry E. D. Pollock 

Carnegie Institution of Washington 

Donald L. Richards 

The Cooley Electronics Laboratory, University of Michigan 


The Report of the President 

I look upon the Carnegie Institution as the most interesting effort the world 
has known for the development of a national interest in research. 

Henry S. Pritchett 

in a letter to Major Henry L. Higginson, May 1904 

Without the degree of liberty which culture demands even a perfect society will 
be no better than a jungle. For this reason all authentic creation is a gift to 
the future. 

Albert Camus 

"Y Artiste et son temps' ' 

Actuelles II, chroniques 1948-1953 

The difference is infinitely small between a system of labour which leads men 
to discover the beauty of the world and one which hides it from them. But this 
infinitely small difference is real, and no effort of the imagination can bridge it. 

Simone Weil 

"Cette guerre est une guerre de religions' 7 

Ecrits de Londres et dernieres lettres 

Institution of Washington. Sixty years ago, in 1902, Andrew Carnegie 
transmitted to a newly elected Board of Trustees a deed of trust conveying 
the sum of ten million dollars "to found, in the city of Washington, an 
Institution which with the cooperation of institutions now or hereafter 
established, there or elsewhere, shall in the broadest and most liberal 
manner encourage investigation, research, and discovery. . . ." At the end 
of January in that year, the Trustees elected Daniel Coit Gilman, fresh 
from the career for which he was already noted as president of the Johns 
Hopkins University, as first president of the Carnegie Institution, and 
resolved "to promote original research by systematically sustaining projects 
of broad scope that may lead to the discovery and utilization of new forces 
for the benefit of man . . . projects of minor scope that may fill in gaps 
of knowledge of particular things or restricted fields of research . . . admin- 
istration of a definite or stated research under a single direction by compe- 
tent individuals." 

It was not the first of Andrew Carnegie's great philanthropic gifts. Far 
from it indeed. In the last decade of the closing century in Pittsburgh he 
had established the Carnegie Institute with its natural history museum, its 
music hall, and its department of fine arts, and had made possible the 
Carnegie Institute of Technology, grown now to front rank among the 



scientific and technical universities of the nation. In the opening years of 
the new century he had established the Carnegie Trust for the Universities 
of Scotland, and the Carnegie Dunfermline Trust in benefit of his native 
town. Nor was it, by many removes, to be the last. There were to follow 
the Carnegie Foundation for the Advancement of Teaching, the Carnegie 
Endowment for International Peace, Carnegie Hero Funds in no less than 
eleven countries, and finally, in culmination, the Carnegie Corporation of 
New York. And long before all of them — indeed well before the publication 
of his pioneering " Gospel of Wealth" in the North American Review in 
1889 — he had initiated that career of benefactions which was to be so 
profoundly influential in all the subsequent shaping of American philan- 
thropic tradition with the gift of a library to his native Dunfermline. 

But the establishment of the Carnegie Institution of Washington marked 
a new direction in the kinds of institutions made possible by Mr. Carnegie's 
gift. In fact, it established a new kind of institution for America — the first 
to be devoted wholly and completely, in intent and in philosophy, to the 
ideal of research scholarship over wide fronts of science in its broadest, 
most unfettered, most completely uncommitted aspect. This was a novel 
concept, and quite obviously, from some of the records of the time, one 
neither everywhere comprehensible nor even everywhere palatable in a 
youthful nation with a strongly established pragmatic tradition. It repre- 
sented, indeed, a notably original idea, which six following decades have 
shown to be both great and enduring. 

Four years after the establishment of the Institution, it had been granted 
a new Charter by special Act of Congress and had been organized into no 
less than fourteen departments, representing as many subjects. Over the 
next five years, definitive judgments were made as to where and how the 
Institution could work most effectively. One of them made during these 
years of experiment and trial was to prove crucial. It involved the decision 
to concentrate the resources of the Institution primarily on the research of 
its various departments ; to make of it, in essence, an operating rather than 
a granting scientific organization. By 1911, its endowment more than 
doubled by subsequent additions by Mr. Carnegie, its departments firmly 
established but now reduced to ten, the Institution was molded to the 
purpose, and had taken on essentially the form of organization, that 
characterize it to this day. Through the following years new departments 
have arisen, departments have been consolidated, and some departments 
have been closed, as the needs and the research frontiers of each decade 
have dictated. Whole fields that were represented in the Institution in 1911, 
like economics and sociology, historical research, meridian astrometry, 
nutrition in the medical sense, no longer are included in its program as the 
resources of the nation in those areas have strengthened and enlarged. 


Other fields not represented then but now on the frontiers of research, like 
modern embryology, molecular and cellular biology, the study of the 
mechanisms of photosynthesis, have been included in its purview in more 
recent years. Today there are five instead of ten departments in the Insti- 
tution. Most originated in planning going back to the very beginning, 
though the work they conduct today, under the same general titles with 
which they began, has expanded far beyond the original concepts embodied 
in those rubrics, and may have wandered far afield from them as well. The 
Department of Terrestrial Magnetism was founded in 1904, the Geophysical 
Laboratory in 1906, and a Desert Laboratory, later to become the Division 
and then the Department of Plant Biology, appeared in 1903. A Solar 
Observatory for Mount Wilson was planned as early as 1902. Studies of the 
sun remain at the pioneering fringes of investigation in that part of the 
Institution to this day. But now the Solar Observatories have metamor- 
phosed to the complex of giant telescopes included in the Mount Wilson 
and Palomar Observatories, operated jointly with the California Institute 
of Technology. To the intensive program of solar investigations of which 
George Ellery Hale dreamed and which he initiated with his striking 
discoveries of magnetic fields in the sun have been added a goodly share of 
the world's most important findings about the farthest reaches of the 
celestial universe. 

But through all the years the major philosophies of the Institution and 
one major feature of its organizational pattern have stood constant, tested 
and retested in situation after situation and proved as fresh and relevant 
today as when they were conceived. The decision made at the outset that 
flexibility and effectiveness in the kind of research to which the Institution 
is dedicated can best be achieved through a series of rather small unit 
laboratories, each mobile and relatively independent, each able to seize the 
initiative in new and appropriate fields as they appear, yet all sufficiently 
connected so that they may be of mutual assistance as the needs arise, was 
a remarkable one, both for its uniqueness at the time and for the subtlety 
of the vision that dictated it. Over the decades, as research has burgeoned 
in the nation and groups devoted to research have multiplied, many other 
experiments in organizational form have been tried. But it is especially 
interesting that some of the most modern thinking and experimenting in 
organization for research, in this country as well as abroad, has returned to 
precisely this pattern as one of the most effective in exploring the dynamic 
frontiers of scientific knowledge. 

Organization, however, is only a framework, vital but at last only 
supporting. Most significant — and most truly enduring — have been the 
elements of philosophy and purpose which inaugurated the Institution and 
which have remained unchanged through all the years : the philosophy that 


all its resources, all its deepest purposes, are centered in the creative 
individual, whatever be his field, that in the truest sense he is the uncom- 
mitted investigator, suitably endowed and suitably protected, whose time, 
quite literally, is bought by the Institution and then returned as uncon- 
strained endowment. And with this goes the philosophy, equally deep-seated 
and equally important, that this freedom from fixed commitment applies 
to fields of endeavor as well as to men: that high mobility within specific 
fields, that the unfettered crossing of fields, that the fashioning of uncon- 
ventionally wide-ranging programs, are subject only to the limitations 
imposed by Nature and by the judgment of gifted and discriminating 
investigators, and that making this mobility and this flexibility possible is 
a principal objective of the Institution. 

Over the years that philosophy, and the programs that have followed 
from it, have led to many pioneering practical discoveries within the 
Institution. The elucidation of the genetic principles underlying the 
development of hybrid corn, first accomplished by Shull in the Department 
of Genetics at Cold Spring Harbor working with East at Harvard, provided 
the fountainhead for an agricultural innovation which by 1952 was esti- 
mated to have brought an economic gain for the United States of almost 
forty billion dollars. For many of the predominantly agricultural countries 
of the world, moreover, the technique of hybrid corn has provided one of 
those basic resources which, as Galbraith has recently pointed out, is in 
the truest sense a fundamental contribution to their economic strength — an 
advance of really general application. At the same Department, during the 
second world war, studies of mutations occurring under X-ray bombard- 
ment in the famous mold Penicillium resulted in the development of a 
strain of that fungus which produced three to five times as much of the 
vitally needed penicillin as the highest-yielding strains then known. 

In 1925, fully fifteen years before the intensive research on radar for 
combat in the second world war, Breit and Tuve at the Department of 
Terrestrial Magnetism, experimenting with a modified Navy transmitter, 
produced radio pulses and for the first time observed their echoes from the 
ionosphere. In the course of those experiments, moreover, they detected a 
curious interference of normal echoes by passing planes — prophecy of the 
field of radar. At the Geophysical Laboratory, Day and Shepherd early 
undertook studies in the field of low-expansion quartz glasses that proved 
basic to the evolution of Pyrex — a program which during the first world 
war supplied the United States with ninety-seven per cent of its require- 
ments for optical glass. In 1935 a modified formula for annealing that same 
Pyrex glass proved fundamental to the manufacture of the mirror for the 
two-hundred-inch telescope on Palomar Mountain. Later, in the same 
laboratory, studies by Morey on lanthanum and borate glasses of high 


refractive index led to a whole new family of glasses of great importance in 
the manufacture of photographic lenses — a development having important 
implications for the second world war. In the Geophysical Laboratory, 
again, Rankin and Wright as early as 1915 were able to solve the age-old 
riddle of cement, and their classic work has served ever since as a guide in 
the chemical aspects of the cement industry. From the same laboratory in 
later years have come new refractories for the steel industry, studies of 
natural geothermometers and geochronometers of fundamental concern to 
practical mining and oil prospecting as much as to fundamental geology, 
and, as recently as 1959, synthetic diamonds produced with new substrates 
and under new conditions of pressure. 

Such practical innovation within the Institution has not been confined 
to the substantive aspects of its concerns. In both world wars the Institution 
played a major role in initiating forms of research organization for armed 
conflict. In the first war, the scientific and technical role of the Institution 
overshadowed its organizational one. But in World War II, through its 
President, the Institution served as a core of thinking and effort from which, 
in the following war years, the Office of Scientific Research and Develop- 
ment was to grow and to assume the lead in civilian scientific and technical 
military development in the nation. Through its activity and its influence, 
a preponderant share of all the major scientific and technical advances in 
the military art were achieved, from radar to modern submarine detection 
to proximity fuzes to nuclear weapons to new and improved prosthetic aids 
for the war wounded and the war blinded. 

But as critical as the technical findings developed from its activities, and 
in the final analysis perhaps more enduring, was the dramatic and conclusive 
demonstration of the crucial role that science as a whole must play in our 
national life in the years to come, in formal peace as in formal war. Experi- 
ments in the organization of science were initiated in the O.S.R.D. which 
were ultimately to find fruition in such government instruments for the 
furtherance of scientific development throughout the nation as the Office 
of Naval Research and later the National Science Foundation, and in such 
bodies as the Atomic Energy Commission, whose present organizational 
patterns, first tested in the Manhattan Project of the Army Corps of 
Engineers, were likewise pioneered in the O.S.R.D. They were reflected, too, 
in such special resources of military thinking and planning as the Rand 
Corporation, founded shortly after the close of the war. Before those 
wartime demonstrations of the crucial role of science and technology in the 
very web of our national life had been made, the greater part of the scientific 
activity of the nation was prosecuted outside the sphere of government and 
of public funds. Today, probably sixty-five per cent of the total research of 
the nation is supported by federal funds, and the proportion is continuing 


to grow. It is a dramatic demonstration of how deeply, in the public view, 
the scientific and technical development of the nation has, in fact, become 
the whole nation's concern. This situation has brought its own problems, of 
a wholly new order of scope and depth. They, too, must be important 
concerns in the future for the Carnegie Institution. 

To have initiated such practical contributions to the public welfare on 
the scientific and technical fronts, to have participated actively and sig- 
nificantly in the initiation of major currents of scientific history whose 
sweep has now carried us to realms far beyond what was remotely imagined 
even twenty years ago, to have pioneered forms of organization that are 
today in the furnace of national trial and test, sum to considerable useful 
achievement, and might be thought, in and of themselves, to justify the 
vision upon which the Institution was founded and through which it lives 
today. Yet, in one sense, they represent mere by-products, mere projecting 
iceberg tips, as it were, of that vision, indicators only of the submerged 
seven-eighths. That seven-eighths lies in the kingdom of the mind. It lies 
in that devotion to deeper patterns, the symmetries, the lights and shades 
of Nature, wherever the search may lead, to which the Institution was 
originally dedicated, and which, undeviatingly, it pursues today. 

That seven-eighths too has been productive of striking innovations in its 
own realm, and these, possibly in a truer sense than the practical "firsts," 
stand as proper signatures of the Institution. They range over many fields. 
While the thinking which underlay the famous Michelson-Morley experi- 
ment on "ether drift" was yet fresh, Professor Michelson, holder of the first 
Nobel prize in the natural sciences to be awarded in America, within the 
Institution repeated the experiment with an accuracy hitherto unattained, 
giving strong support to the theory of relativity, itself still at the stage of 
question and of doubt. Within the Institution, too, Michelson repeated 
with greater refinement that classic work that he had first undertaken as 
Ensign A. A. Michelson of the United States Navy, determining the 
velocity of light with a new precision, first across a path between the peaks 
of Mount Wilson and Mount San Antonio, then in one defined by a mile- 
long line of evacuated pipe at the Irvine Ranch in southern California. At 
the Mount Wilson and Palomar Observatories Hale's pioneering discovery 
that sunspots mark strong magnetic fields has been followed in more recent 
years by studies of solar magnetism of unprecedented refinement, and by 
the discovery, among the stars, of the most intense magnetic fields ever 
observed in any astronomical body. Hubble's studies of the phenomenon of 
the redshift in stellar spectra led to the theory of the expanding universe, 
culminating dramatically a year ago in the measurement of the redshift of 


by far the most distant celestial object yet recorded. At the Observatories, 
too, Baade's studies of the structure and stellar composition of galaxies, 
with those of others, have suggested concepts of stellar evolution, of growth 
and decay, undreamed of as little as a quarter century ago. 

At the Department of Terrestrial Magnetism a series of conferences on 
theoretical physics, held shortly before the second world war in cooperation 
with the George Washington University, among other things stimulated 
the suggestion that the source of energy in the sun and the stars is a nuclear 
reaction involving carbon — a notion leading within the next year to a 
classical model of the hydrogen-helium reaction now familiar as one of the 
accepted sources of stellar energy, ancillary to the hydrogen-deuterium- 
helium reaction recognized in recent years as more important. In the 
Geophysical Laboratory studies of the biochemistry of ancient sediments 
have given new dimensions to our concept of the age of terrestrial life, while 
studies of the artificial synthesis of amino acids from inorganic components 
under a variety of physical and chemical conditions, besides shedding new 
light on the probable modes of the origin of life on earth and the nature of 
its chemical environments, have also carried important theoretical impli- 
cations for our notions about the existence of life on other planets. 

At the Department of Plant Biology, work on photosynthesis has pro- 
duced suggestive insights about that critical step which, with all the research 
that has been brought to bear for the last half-century, still eludes our 
understanding — the initial process by which the energy of light is used in 
the fixation of carbon dioxide. It has brought suggestions, too, about that 
further mystery, still elusive, of what it is about the chloroplast that enables 
it alone, when intact, to bring this about, whereas extracted chlorophyll 
itself will not. And in that Department, too, investigations of many years' 
duration have illuminated the detailed bases of plant evolution — of the 
roles of mutation and selection, of the development of ecological races and 
of speciation — and have revealed the often enormously complex and 
exquisitely coordinated detail of the evolutionary patterns they compose, 
at the levels both of form and of physiological function. 

In three laboratories of the Institution — the Department of Terrestrial 
Magnetism, the Department of Embryology, and the Department of 
Genetics at Cold Spring Harbor — investigations of cellular metabolism and 
development, of cellular differentiation, and of the mechanisms of heredity 
at the molecular level have brought striking new knowledge of the detailed 
ways in which the materials of heredity and of development interact at the 
level of the cell nucleus and of its cytoplasm, at the level of the germinal 
cell and of the body cell of the plant or animal, at the level of differentiation 
and development of the individual organism, and at the level of its heredity. 

Such discoveries and results are but scattered samples taken from a rich 


matrix of sixty years of Institution work. But they are fair examples of its 
most typical fruit — the truest product of the philosophy in which it was 
founded and through which it lives. It may well be said that all else is in 
one sense by-product. 

In the seventh decade of the twentieth century, it is hard to recast the 
scientific and technical America in which the Carnegie Institution was 
founded in 1902. In the America of 1902, few if any corporations in the 
United States could boast over sixty thousand stockholders. The American 
Telephone and Telegraph Company, as example, admitted to less than 
eight thousand. A third of all the manufactured products of the country 
were produced by partnerships or by individual proprietors. Speech had 
been transmitted by wireless, but the Fleming valve was still to be produced, 
and the first audion was not to be developed for eight more years. The first 
aerial flight, the twelve-second achievement of Orville Wright at Kitty 
Hawk, was not to occur until the following year. The earliest motion picture 
to tell a connected story, The Great Train Robbery, was yet to be produced. 
A large proportion of such great technical industries of today as the movie 
and the aircraft industries had not been born, and even the technical prin- 
ciples underlying the television industry were not yet remotely conceived. 

The independent industrial laboratory had been pioneered some years 
earlier by the Arthur D. Little Company, but the concept of such a labora- 
tory within an industry had just been formulated and put into practice 
with the establishing of the General Electric research laboratory in 1901 
and of that of the du Pont Company in the same year the Institution was 
founded. Of all the great complex of industrial laboratories that were to 
transform the nature of American industrial science and technology in the 
twentieth century, not one other had yet appeared. 

For the scope of science in that day, it is worth noting that in genetics 
it was only two years before that the work of Gregor Mendel had been 
rediscovered and its significance truly appreciated by Hugo de Vries and 
Correns and von Tschermak-Seysenegg. The very notion that some genetic 
characteristics can be dealt with in crosses in numerical ratios was still 
unfamiliar, while ideas of genetic linkage and dominance, or the notion of 
the linear array of genes, was still almost a decade away. Indeed, there was 
no proper science of genetics at all, and the word gene itself had yet to be 
coined. In astronomy, it is probably fair to say that the entire known 
universe was thought to lie within our own Galaxy. By contrast, within the 
range of the two-hundred-inch Hale telescope today lie perhaps a thousand 
million such galaxies. 


Only seven years before the Institution was established, Wilhelm 
Roentgen had given the first demonstration of the X rays that bear his 
name, and the first Nobel award in science had gone to him for that dis- 
covery only a year before the founding of the Institution. The electron had 
been discovered by J. J. Thomson but five years earlier, and radium and 
thorium had been isolated by the Curies only four years before. Max Planck 
had advanced the quantum theory in the year preceding the founding of 
the Institution. And the special theory of relativity was not to appear for 
three more years. The Institution was five years old when the first Nobel 
award in science to be made in the United States came to Albert Michelson. 

In the world of technology, plastics, synthetic fibers, vitamins, anti- 
biotics, all were unknown. And in practical medicine, it is striking that the 
national death rate from influenza and pneumonia was reckoned at one 
hundred and eighty-two per one hundred thousand of the population — a 
figure to be reduced to thirty-nine forty-eight years later. In the same 
period deaths from scarlet fever fell from more than eleven per thousand to 
a total of sixty-eight for the entire country. It is worth recalling that, when 
Lord Lister, scientific disciple of Pasteur to whom the whole concept of 
antisepsis and sterilization in medical practice may be said to have been 
due, died in 1912, the Institution was already completing its first decade. 
Such was the world scene of science and technology within which the 
Institution took its place. 

In 1902 science and technology were already familiar concerns within the 
federal government. They were indeed concerns as old as the nation itself. 
It was Thomas Jefferson who as Secretary of State in 1790 submitted a 
"Report ... on the Subject of Establishing the Uniformity of the Weights, 
Measures, and Coins of the United States," and who, upon recommendation 
of the American Philosophical Society, transmitted to the Congress a 
proposal for the establishment of a United States Coast Survey, which was 
set up within the Treasury Department seventeen years later. And it was 
John Quincy Adams, when he was Secretary of State, who personally 
prepared for the Congress a similar report upon weights and measures. It 
was Adams, too, who led the fight to accept the bequest from James 
Smithson, who had died in 1829, to found the organization that was to grow 
to the Smithsonian Institution of today. The establishment of the Depart- 
ment of Agriculture dated from Civil War days, contemporary with the 
passage of the Morrill Act. So also did the National Academy of Sciences, 
from whose recommendations, somewhat later, were to follow the Geo- 
logical Survey and the Weather Bureau. 


These early involvements of the federal government in science and 
technology, however, gave little hint of the massive and commanding role 
it would play on the national scene in little more than half a century. Even 
at the end of the fourth decade of the twentieth century the total federal 
research program is estimated to have cost annually only about one hundred 
million dollars — less than the annual budget for the National Science 
Foundation alone in 1962. Twenty years later, however, yearly federal 
expenditures for research and development had grown to over a billion 
dollars out of a total estimated national commitment of about three billion. 
By 1960 the national total had climbed to fourteen billion dollars or more, 
of which the federal government supplied some nine billion. Today it may 
have reached sixteen to eighteen billion. The budget of the National Science 
Foundation for scientific research and related activities as submitted to the 
Congress for 1963 will total one hundred and sixty-five million dollars, while 
the Department of Defense is expected to spend about seven billion dollars 
on research and development, the National Aeronautics and Space Admin- 
istration about two and one-half billion, the Atomic Energy Commission 
approximately another one and one-half billion. The total government funds 
spent in research and development in 1963 are expected to reach almost 
twelve and one-half billion dollars, of which expenditures for research alone 
may attain to one and one-half billion dollars, as compared with approxi- 
mately one billion for the present year. 

It has been calculated that the total funds expended for research and 
development in the United States over the past decade have increased at 
approximately fifteen per cent per year, leading to a doubling of volume 
every five years. If the present rate of increase of our expenditures in the 
field were to continue, indeed, our projected monetary support of research 
and development in their current definition could formally exceed our total 
governmental budget before 1975, and could exceed our gross national 
product before the end of the century — a reflection, however hypothetical, 
that vividly illuminates the scientific and technical dynamism and the 
scientific and technical problems with which we live. How different is this 
scene from that upon which the Institution entered ! 

The implications of this astonishing vista are many. One is the degree to 
which, with almost explosive suddenness since World War II, science and 
technology have been universally recognized as of major national concern. 
Another, of course, reflects the depth and intensity of technological compe- 
tition in the world and our own needs in national defense. A third mirrors 
both the rate of population growth and, most pointedly, the growth of 
wealth in the United States. And the climates in which these expenditures 
on both the private and the public fronts have taken place and the govern- 
mental patterns through which they are effected in the public sector — 


patterns at present in perhaps their most active phases of evolution and of 
adjustment — make a compelling chapter in the history of development both 
of American scientific enterprise and awareness and of American political 
institutions, and reveal much about their nature. 

All these factors — the vast increase in the volume of our scientific and 
technical resources, in human and in monetary terms and in terms of 
scientific and technical facilities, the pressing demands of overriding national 
objectives, economic and military, the consequent larger and larger partici- 
pation of federal resources in the total funding of the national research and 
more especially of the national technical effort — have, not unnaturally, had 
profound impacts on our thinking about science generally. Bit by bit they 
may have led to some subtle changes, perhaps well-nigh unconscious ones, 
in our conception of the ways in which, typically, the frontiers of truly new 
scientific knowledge are pushed back. This evolution could carry implica- 
tions grave enough to warrant serious thought. 

In all the years of American scientific research, from the times of Josiah 
Willard Gibbs to those of the second world war, we were accustomed to 
think of the great advances in scientific thought, of the initiation of its 
great new directions, as being predominantly the product of individual 
genius, working in environments which, however modest, and in part 
perhaps because of that very modesty, were especially adapted for flexi- 
bility, for absence of constraint, for a maximum of freedom in concept and 
in execution. We thought of the outstanding scientific conquest as typically 
an achievement of extraordinary brilliance, originality, and insight in 
individual innovation, giving significant new dimensions to its time, and 
ideally climaxing a career of unfettered scholarship. We did not particularly 
conceive research in this sense as the composite product of large numbers 
of men working in numerous and highly organized groups. 

Since the second world war, however, following the spectacular demon- 
strations of technical conquest wrought by great organizations, of which the 
Manhattan Project was but the forerunner, we have sometimes been 
inclined by analogy to conceive of pioneering research for basically new 
ideas in rather similar terms — inclined, perhaps, to more than half believe 
that in the contemporary world it too may require such teams. It is then 
only logical to reason that if, at this stage of the world's scientific develop- 
ment, pioneering scientific research critically depends upon the large-scale 
efforts of highly organized and massively implemented teams, its effective- 
ness may be roughly proportionate to the material resources bestowed upon 
it — and that cost and magnitude themselves may provide an important 
index of scientific significance. We have even been tempted at times to 
imagine that the speed and effectiveness with which new scientific frontiers 
are breached may be a simple function of numbers of men and rates of 


expenditure, and to expect that the attainment of new scientific vision in 
an area of basic research may be accelerated in direct proportion to the 
size of teams and the amounts of money committed to the search. 

This philosophy, so directly derived from the demonstrated course of 
practical achievement, appeals especially to that keen pragmatic instinct 
that has run like a golden thread through all the fabric of our development 
as a nation, and to the genius for organization which has so long been one of 
our most pronounced national characteristics. Nor is there lack of evidence 
that at first sight seems to confirm the idea. It is patent today that the 
physical equipment required on the frontiers of research in many of the 
sciences, especially those of the greatest conceptual maturity, is massive, 
complex, and expensive, and requires the collaboration of sizable teams in 
designing it, in manipulating it, and in gathering data with it if truly new 
information is to be obtained. The productiveness of research in many such 
fields since men and money have made possible the design of powerful new 
tools and massive teams have been assembled to operate them gives vivid 
testimony to how powerful, and indeed how indispensable, resources of this 
kind may be in some of the most highly developed fields of science. 

Yet in a deeper sense this judgment may harbor a considerable, and 
sometimes a positively dangerous, misconception, especially when it is 
assumed that great teams and high costs are prerequisites for the setting of 
new directions in scientific thought. A part of that misconception doubtless 
stems from a failure to demark sufficiently two general approaches in 
research, which, though they are complementary and often intergrade, yet 
have certain characteristics and pose certain requirements that are quite 
distinct. In one the basic ends of the investigation are generally evident, if 
not wholly clear in detail, at or near its beginning. The preeminent challenge 
to the investigator is to chart the road toward his goal — mapping it, 
projecting it, building it, all that it may approach a citadel already at least 
dimly visible on the horizon. The other general kind of research may begin 
without specific ends or, indeed, without consciously conceived objectives 
of any kind. Its driving motive is likely to be pure curiosity, the winning 
from Nature of deeply new knowledge, of knowledge won wholly for its 
own sake. The talents and the training demanded by these two kinds of 
research, and the difficulty of the scientific challenges posed by each, are 
often much the same. At one end of a spectrum of research they intergrade, 
and any distinction attempted between them becomes formal and unreal. 
At their extremes, however, the challenges they present are undoubtedly 
quite different, often to be met in widely divergent ways. Above all, whereas 
research programs of the first kind can frequently be visualized in a general 
way ahead of time, and so planned intelligently, the same is rarely true in 
the second type of research. A very large share of the concerns of such a 


great team effort as was involved in the program of the Manhattan Project, 
for instance, fell into the former category. The deeply underlying theoretical 
knowledge, the unexpected and radically new ideas about Nature, on which 
the whole program of the Project was based and on which it turned, had 
been achieved by investigators like Meitner and Hahn and Strassmann in 
Europe in 1938, by such individuals as Rutherford and his colleagues at 
Cambridge in 1914. They had been won through research of the second 
kind, conducted by a very few gifted scientists working in the settings we 
have traditionally visualized as consonant with the finest of individual 
creative effort. 

It is no accident that today we sometimes make these distinctions less 
clearly than we might. At a very deep level it may be a consequence of our 
peculiar history and circumstances. Throughout our earlier years as a 
technically developing nation we were able to rely on the older countries of 
Europe for basic ideas on which to build our applications as implicitly, and 
often as unconsciously, as we relied upon the British navy for the protection 
of our seas. It was both natural and adaptive that the kind of scientific and 
technical contributions at which we early became most adept and developed 
most highly, and to which perhaps we initially attached greatest attention 
and attributed greatest value, should have involved the brilliantly organ- 
ized, the meticulously careful development, often undertaken on the boldest 
and most breathtaking scale, of basic ideas that had been conceived abroad. 
Today such ideas are much more often drawn from our own resources. But 
historically our first attachment was to their execution rather than to their 
generation. And so it is not surprising that we sometimes fail to distinguish 
innovation from execution, and have not always recognized the limitations 
within which we can extrapolate experience from one kind of activity to 
the other. 

But there is more to the matter than this. For it is demonstrably true 
that gains in our knowledge of Nature as new and fundamental and unex- 
pected as any in the world can come, unbidden, from the investigations of 
great teams for research and development in many areas. As our resources 
for team research grow in the coming years, we can properly expect the 
rate at which such new knowledge is revealed to increase also — if not 
proportionately, at least very substantially. And so we should not fail to 
ask an implied question of great importance. The philosophy that envisaged 
the environment of brilliant, original, unfettered individual research as the 
milieu in which the great new directions of scientific thought were born and 
nourished, the philosophy which has had such confirmation in recent 
scientific history, was itself developed in the days of scarcity in science — 


scarcity not only of material wealth, but especially scarcity of scientific 
workers. Now we live and work in a nation committed to an unparalleled 
rate of growth in the material resources for research, and in a world in 
which perhaps eighty per cent of all the scientists who have ever lived are 
our contemporaries. Is it possible that the philosophy itself was adjusted to 
the needs of other times ; that it is not relevant to an era of plenty? May it 
actually be true today not only that major advances in new knowledge, the 
setting of radically new scientific directions, can be achieved in the environ- 
ment of great and highly organized research teams, but also that, in practice, 
such environments are indeed essential, or, at any rate, the most favorable, 
to the process? Is it possible that we are witness to a profound revolution 
in the very character of research itself? Is it possible that the small and 
mobile groups to which we earlier looked for some of the most significant 
scientific innovations, the groups which in the past characteristically had 
an influence on scientific progress out of all proportion to their numbers or 
their social cost, can no longer in our day provide such significant approaches 
to the unknown? 

Such a radical query, of course, bears profoundly on the whole philosophy 
of research. It is far more than a practical question. It touches some of the 
deepest wellsprings of scientific faith. It touches belief in the very nature 
and effectiveness of the individual search for truth in our time. In subtle 
ways it touches on the nature of scientific truth itself. It is an important 
question for the Carnegie Institution, deeply committed to the faith that 
the distinguished, unfettered individual can bring unique gifts to his 
society, and deeply committed, too, to belief in the uniqueness and the im- 
portance of the influence which a community of independent scholars can 
exercise on scientific progress. 

For a question of such magnitude and gravity, abstract analysis will not 
suffice. Contemporary evidence alone can give convincing answers. Have 
the recent great advances in our knowledge of the universe and of our own 
more immediate environment, the original ideas of scientific stature 
achieved in the last few years which promise to open truly novel avenues of 
thought for the future — have these been necessarily, or even primarily, 
associated with the massive programs of great teams? Or do the basic 
contributions of small and mobile research groups continue in our day to 
have their old significance? 

Such an abundance of evidence springs to mind, provided by striking 
advances no more than a half-dozen years old, in so many regions of scientific 
inquiry, that its very selection poses a problem and must necessarily be 
arbitrary. But three outstanding areas of recent investigation are particu- 


larly interesting to consider from this standpoint, because their environ- 
ments and circumstances span such an extraordinary range of magnitude 
and character and form. 

The first example may comprehend that immense complex of research and 
development dedicated to the placing of man in outer space and ultimately 
on the moon or on neighboring planets, its present great achievements in our 
country vividly symbolized by the voyages of Shepard and of Grissom, of 
Carpenter, Glenn, and Schirra. The second is of quite a different kind. It 
involves an achievement in astronomy of the year just past which in the stag- 
gering distances with which it deals emphasizes anew what a thin terrestrial 
shell is the outer space so far entered by man. It is the identification of what 
has proved to be by far the most remote celestial object ever discovered in the 
heavens — an object certainly billions of light years distant from us — and 
the measurement of the redshift of its spectrum. The third selected area of 
advance may in some ways be the most profound of all, though it is far from 
the best known. It includes the experimental evidence so brilliantly obtained 
in the last few years, and the reasoning directing the search for it, indicating 
beyond reasonable doubt that the information governing the inheritance of 
all the qualities of living things is structurally graven on the chromosomes 
within their germ cells in the form of a genuine code. It includes, as a climax, 
the demonstration of the general nature of that code, which the year just 
past has witnessed. These findings may well mark the greatest single ad- 
vance in genetics since the demonstration five decades ago that the genes of 
heredity lie in the chromosomes in a linear array. 

These three advances in natural knowledge bear much resemblance in 
certain fundamental qualities. All have won important and striking new 
knowledge. In all of them, the research for that knowledge has included a 
variety of scientific disciplines apparently far removed from the main 
concern — in the case of the third as far removed as crystallography seems to 
be from conventional genetics. Profoundly new directions of thought have 
resulted from all three. Possibly the third has produced the most thoroughly 
revolutionary new insights. The first has brought a sense of liberating con- 
quest and a wealth of first-hand information about regions known hitherto 
only palely and at second hand. 

But in many features of the modes and environments of research char- 
acterizing them, the three examples diverge about as much as scientific 
activities can differ. The contrast is particularly vivid when cast in terms of 
the parameters under special consideration: the relative size of the efforts, 
the sheer volume of human and material sources brought to bear, the kinds 
and degrees of organization. The enormous magnitude of the space program 
and the tremendous cooperative efforts currently involved in its prosecution 
and planned for the future need little emphasis. In this respect, indeed, 


Project Apollo is much in the tradition of a Manhattan Project, though 
yet bolder in both variety and scale. It is estimated that by the close of the 
budget for 1963 the National Aeronautics and Space Administration will 
have spent more than four thousand millions of dollars for the conduct of 
research and development. For research facilities alone it will have expended 
more than eight hundred and twenty millions. Behind the great individuals 
who have manned the space vehicles, and have recorded and analyzed the 
data of research, and who will do so in the future, lie the years of develop- 
ment on a scale of unprecedented magnitude and the immense organizations 
required for its successful prosecution. Behind the fashioning of the tools 
the final explorers command lie combinations of highly specialized dis- 
ciplines and intricate techniques of the most varied kind — chemical, elec- 
tronic, mechanical — ranging from the arts of propulsion engineering to 
those of miniaturization. It is interesting to notice in this connection that 
the cast of the effort at present is, as it perforce must be, importantly 
oriented about the design and use of tools. In considerable measure it is 
basically an engineering effort — perhaps the most exciting and compelling 
engineering effort of this century. 

Shortly after the second world war, when instruments of radio detection 
were being put to a new use in the service of astronomy, several surveys of 
the skies were undertaken to detect and locate the positions of celestial 
bodies that were emitters of radio waves. The equipment then available, 
however, was relatively poor in both resolution and accuracy. It could not 
effectively complement the far more precise tools of optical astronomy. 
Resolution and precision were often too low to permit a reliable identifica- 
tion of radio sources with corresponding objects observed optically, though 
sometimes they were suspected to be the same. As the techniques of radio 
astronomy sharpened, however, as larger dishes were built and manned and 
put into use, both penetration and resolving power improved greatly. At the 
radio observatory of the Cavendish Laboratory in England and at the 
observatory of the California Institute of Technology at Bishop in the 
Owens Valley, instruments of outstanding capacity were built. During 1959 
and 1960 two fresh surveys of the skies were undertaken with them: in 
Cambridge at 169 and 189 centimeters, in California at about a sixth that 
wavelength (31.2 cm). In the course of these surveys the celestial positions 
of certain emitters of radio waves were determined with a new precision. So 
precise was the location of one of these objects, indeed, that the two- 
hundred-inch Hale telescope could be brought to bear upon it. The peculiar 
color characteristics of the object suggested that it might include a pair of 
galaxies in collision, and so might be expected to have one or more emission 


lines in its spectrum. And so it happened that a prescient astronomer of the 
Mount Wilson and Palomar Observatories was able to obtain two spectra of 
the visible light from this source and to measure the degree of redshift in 
them. At the same time another observer, obtaining multicolor photometric 
observations of two of the fainter galaxies of the same cluster and construct- 
ing their curves of continuous emission, confirmed this measurement of red- 
shift. It corresponded to a recession velocity of nearly half the speed of 
light. This heavenly body defines a new boundary for the universe compre- 
hended within human ken. It marks by far the most searching probe into 
unplumbed reaches of space that the mind and hand of man have yet 
accomplished, ranging certainly to the order of several billion light years. 
When it is recalled that a single light year amounts to almost six million 
million miles — about sixty-three thousand times the distance of our own 
world from the sun — it makes the orbits of earth satellites, spectacular as 
they are, yet appear as comparatively near-neighborhood adventures. 

Perhaps the greatest ultimate significance of this achievement will lie in 
the contribution it can make to our ideas about the basic nature of the 
universe. Indeed, this newly determined point of distance, so far beyond any 
other yet obtained, has already offered suggestive evidence on the great 
question of whether our universe is a continuously expanding one, or a 
universe in which the continuous creation and destruction of matter stand 
in equilibrium, or whether the universe in fact may experience alternate 
expansion and contraction extending over astronomic periods of time. 

In sharp contrast to the first example, the planning of these observations, 
their confirmation, and the deductions from them were not the work of great 
teams of highly coordinated technical workers. These were the fruits of 
observations and calculations made by a few individuals laboring in relative 
solitude, the fruits of work of a relative handful of gifted astronomers. 
Perhaps never in science has the work of individuals been more clearly 
identifiable. The contrast with the first example is sharp. 

Yet behind this classical achievement of gifted individuals lay many 
decades of research and engineering focused on the design of the powerful 
modern tools of optical and radio astronomy. Without them the achieve- 
ment itself would have been quite impossible. These tools, like those in- 
volved in the space effort, were the products of hands and minds and toil in 
literally hundreds of specialized skills. And it was not skill and art that alone 
were brought to bear, but with them the magnificent resources of intellect 
and materials and time and research that gave them scope and effectiveness. 
The achievement itself dramatically underlines how significant and how 
essential the gifted and untrammeled individual investigator is today on 
some of the most advanced frontiers of the physical sciences. It was pri- 
marily focused on the gathering and the interpretation of information about 


nature, not on the design of tools. Yet its success depended in turn on a 
panoply of instruments brought to perfection in other times and other 
places, the development of which had required a structure of science and 
technology of whose cumulative magnitude and scope no scientist of an 
earlier generation could have had the faintest dream. 

The third example embodies yet a different pattern. It would be hard to 
imagine a more fundamental or more sweeping discovery than one elucidat- 
ing, at a deeper level than had hitherto been imagined, the manner in which 
the information governing all the qualities of inheritance may be recorded 
and stored in the chromosomes of plants and animals and men — stored with 
such extraordinary effectiveness and such enduring stability that there are 
organisms living today whose hereditary characteristics have been main- 
tained more durably than the very rocks within whose strata the fossils of 
their remote ancestors are preserved. Yet in terms of magnitude the human 
and the material resources committed to that search, by comparison with 
the preceding illustrations, have been positively minuscule. 

In 1953 Linus Pauling and Robert Brainard Corey at the California 
Institute of Technology suggested that the molecular structure of the unit 
of heredity, the "molecule" of deoxyribonucleic acid, might consist of chains 
of polynucleotides intertwined in the form of a helix, with four characteristic 
bases, the purines adenine and guanine and the pyrimidines thymine and 
cytosine, attached to them and projecting outward, while phosphate groups 
were oriented to the center. There were features of this model which con- 
flicted with experimental evidence, notably that it was hard to reconcile the 
fact that DNA is an acid with the existence of bases lying, as it were, on the 
outside of the molecule. But the model involved one very great idea which, 
though it was not widely credible in terms of that particular construction, 
yet was to prove fundamental to all further thinking on the matter. It was 
the idea that the biological specificity of the unit of DNA, on which its 
power of determining inheritance must rest, must inhere in the sequence of 
occurrence of these bases along the molecular chain and the suggestion that 
the periodic distances at which these bases occur might be of the right order 
to permit them to order the sequence of amino acids in the construction of a 
protein. This was a most important foundation upon which to rear what 
would prove a truly extraordinary arch of reasoning. But for long even the 
idea that the nucleic acid structure could be locally specific was resisted. Until 
that idea had been widely accepted, its more detailed consequence could 
hardly gain effective credence. Both these developments were made possible 
by a second great idea, which might be likened to a keystone of the arch. 

This critical idea was provided by J. D. Watson when, in a flash of insight 


reminiscent of Kekule's vision of the structure of the benzene molecule that 
came to him in a London bus almost a hundred years ago, he imagined 
the consequences of, in effect, turning the model inside out, pointing the 
bases inward, and pairing the purine molecules with the smaller pyrimidines. 
Highly significant correspondences with nature were achieved by this 
remarkable insight. The first and fundamental rule of the composition of 
deoxyribonucleic acid, namely that it incorporates purines and pyrimidines 
in equal ratio, was given a rational basis. And the contradiction between 
the acidic nature of DNA and its presumed outwardly pointing bases, which 
had plagued the model of Pauling and Corey, was resolved. But there were 
impressive difficulties to be met also. The idea that the bases were outward- 
pointing had not resulted simply from neglecting the alternative that they 
might point inward. That possibility, indeed, had been carefully examined in 
formulating the earlier model. But it had been concluded that such a 
structure was not possible. For the new model to be convincing, the physical 
possibility of such an arrangement had to be demonstrated, and the details 
of the linkages between the purines and pyrimidines had to be worked out — 
formidable tasks requiring concepts and techniques familiar to those dealing 
with the structure of crystals. 

And so it was that, also in 1953, Watson and F. H. C. Crick, working in 
the Molecular Biology Unit of the British Medical Research Council 
adjacent to the Cavendish Laboratory at Cambridge, announced their 
brilliant hypothesis of the structure of the unit of heredity, of the "molecule" 
of deoxyribonucleic acid, as a pair of "ribbons" wound in the form of a 
double helix around a common axis and linked by the four bases, the purines 
adenine and guanine and the pyrimidines thymine and cytosine, paired in a 
highly specific fashion. The model of Pauling and Corey had suggested that 
the bases could not be packed in the center of the molecule. The new model 
proved that indeed they could, and from that demonstration came perhaps 
the most significant idea in the whole chain — the concept of base pairing 
itself, and with it the associated and important notion that a maximum of 
four kinds of base pairs could be involved. The beauty and credibility of the 
model gave firmness and emphasis to the earlier idea that the biological 
specificity of the unit of heredity must derive in large measure from the 
ordering of the pairs of bases along the chain of the deoxyribonucleic acid. 

All together, three biological consequences stemmed directly from the 
model, which must rank among the most important advances of our age in 
the understanding of the fundamental nature of earthly life. First, the model 
allowed the extraordinary phenomenon of the replication of the genetic 
pattern which occurs at every division of every living cell — the mechanism 
fundamental to the very process of the growth and multiplication of life on 
earth — to be understood consistently for the first time. Second, the nature 


of the phenomenon of the sudden changes in inheritance which we call 
mutation, intensively studied since the days of de Vries but never under- 
stood in their fundamental molecular mechanisms, now for the first time 
became comprehensible at that level, in terms of known changes in bases 
which could result in alterations of their sequence to produce such changes. 
Third, and greatest of all, perhaps, was the full rationalization of the key 
concept that biological specificity in inheritance must in large part derive 
from the sequential ordering of the bases in the nucleic acids. 

This third great consequence was to lead to a scientific vision of new and 
unexpected dimensions. That vista was provided by the idea that genetic 
information might in fact be coded in the DNA molecule in the form of a 
linear message for which the four permissible combinations of bases might 
serve as alphabet, in a manner, indeed, reminiscent of the coding of a 
message on the punched tape of a computer. This radical concept was first 
examined in detail by the astrophysicist Gamow in 1954. Although the 
precise form of the code suggested at that time has since proved incorrect, 
the basic idea has become established as one of the great theoretical ad- 
vances in our view of the nature of the living world. And so was posed the 
pointed question: if such a code exists, what is its specific nature? 

It is that question which theoretical and experimental work of the past 
two years has done much to answer. An important share of the answer, like 
the original question, has come once again from the laboratory of the Unit 
for Molecular Biology at Cambridge ; other critical parts have followed from 
several American university laboratories, from the National Institutes of 
Health, from the Carnegie Institution of Washington. Suffice it to say that 
preponderant evidence suggests that the code employs words containing 
very few "letters," probably not more than three. 

A virus may include within its single chromosome something of the order 
of a hundred thousand base pairs. A billion pairs of bases may be included 
within the total store of information of our own chromosomes. It is a 
startling concept that if the DNA strands from all the cells in a single human 
body were uncoiled their total length might well span the solar system. 
There is ample opportunity for diversity in the ways that the elements of 
the code can be combined. 

With this conceptual advance, carrying the implication that one of the 
basic challenges offered by the problem of heredity might lie, in effect, in the 
decoding of a script, progress in meeting that challenge has come with 
remarkable speed. What may well prove to be a Rosetta stone has been 
provided by the development of methods of accomplishing protein synthesis 
in cell-free systems under the influence of artificial ribonucleic acids com- 
posed of only two bases in known ratios and therefore containing specified 
code words in known frequencies. The composition of the resulting protein 


should yield the key to code " letters" in terms of the ratios of specific amino 
acids corresponding to them. Another highly promising approach involves 
techniques for investigating the coupling between the base-pair patterns of 
the deoxyribonucleic acid of an organism and the "messenger RNA" of 
related forms, which may differ in their coding only in relatively minor, but 
specific and determinable, particulars. The current year sees work of this 
kind at a peak of activity. With wing-swift speed, a whole new area in our 
understanding of the basic mechanisms of heredity at the molecular level is 
being exploited. 

Here, then, are three genuinely great advances marking the technical and 
scientific progress of the last three years. In a profound sense all three are 
typical of their age, and, for a variety of reasons, could not have occurred at 
any earlier time. Obviously neither space exploration nor the astronomical 
investigations of the new "edge of the universe" now within our ken could 
have been achieved with the tools of any other era. The peculiar modernity 
of the third example involves especially a yet different circumstance. For the 
very idea that the information of inheritance may be recorded as a code is 
peculiarly consonant with our age — perhaps so characteristic that it should 
be treated with a caution doubled by this very fact. In the nascence of 
primitive biological thought fire was a living thing, dangerous and bright, 
and the expression "vital fires within us" remains to remind us how much we 
once thought of life as the "inhabiting property" of something that was 
obviously dynamically alive. In an age when the frontiers of engineering 
exploration concerned pumps and hydraulics the mechanism of the circula- 
tion of the blood was a fascinating and fertile subject of physiological 
speculation and of physiological research. For the age of Descartes, strings 
and pulleys provided compelling images for the mechanisms of life, and 
images of clockwork for the mind. In the early nineteenth century, domi- 
nated by the vision of steam power engineering, energy transformations 
seemed among the most important aspects of life, and the rise of large-scale 
electrical power engineering in the latter part of the nineteenth and the 
early twentieth century reinforced the vision. Then, in our own era, with its 
emphasis on small-current engineering and the modulated control of gigantic 
mechanical and electrical processes, the aspects of living processes included 
under the rubric of Cybernetics have occupied a center of the stage. Studies 
of those fascinating properties of living systems involving, in all their varied 
and exquisitely elaborate mechanisms, the maintenance of homeostasis, the 
preservation of balance in dynamic systems, have held a special attraction 
for our time. And in our immediate day, when communication of new orders 
of content and of speed, and with it the massive processing of information, 


so dominates our lives, when we are inevitably so much concerned with the 
coding of information and the unraveling of such codes, it is scarcely 
surprising that a natural process operating upon those principles, which has 
evidently been central to the evolution of all life, as no doubt it was also in 
its origin, should only now have so powerfully focused our attention as to be 
on the threshold of solution. It follows, too, that, just as each of the earlier 
interpretations of living processes subsequently gave central place to its 
successor but left the residue of its own unalterable truth to contribute 
permanently to our basic understanding, we must be prepared to accept — 
and indeed to welcome — the same fate for the concept of genetic coding. 

The likenesses uniting these three examples, then, lie deep. It would be 
hard to select the most significant among them, though in the achievement 
of particular new insights the second and especially the third may pre- 
dominate. What now of the parameters of scale, of magnitude of the re- 
sources committed, of the extent of organization of the work, as criteria of 
its significance? Here it would be difficult to imagine wider contrasts. 

At every point in the extraordinary conceptual development that marks 
the third example, the commitment to it in terms of numbers of workers, in 
terms of material resources, was extraordinarily modest. The Unit for 
Molecular Biology of the Medical Research Council at Cambridge began 
with two crystallographers. Ten years later, when its revolutionary dis- 
coveries were well launched, it numbered perhaps a dozen workers and was 
housed in a temporary building behind the Cavendish Laboratory and in 
various University rooms — a very minimum of space. It was, indeed, 
superbly instrumented for its task. But such instrumentation was in- 
credibly modest in both mass and cost compared with that required in 
either of the other fields. In that free and flexible atmosphere, built about 
the largely unfettered efforts of a few gifted individuals working within a 
minimum of formal organization, have been made some of the most im- 
portant advances in man's concept of his world and of himself possible to 
the twentieth century. It is striking to compare this situation with that in 
which the exploration of space must go forward. 

This, then, is the character of the contemporary evidence. Such contrasts 
of size and structure and organization in the modes of some of the most 
significant assaults on the frontiers of natural knowledge in this decade 
strongly suggest that these parameters, broadly considered, bear little 
direct relation to their scientific significance. They inspire compelling re- 
flections about the continuing effectiveness, in our own day, of the scale and 


the pattern and the philosophy of research to which the Carnegie Institution 
is so deeply committed. It seems abundantly clear that the essential qualities 
and requirements of inquiry at the very frontiers of man's knowledge of his 
universe do not now, and in all probability will not in the foreseeable future, 
differ significantly from those of our classical scientific past. Such inquiry 
will surely continue to bear the unmistakable stamp of the gifted and un- 
trammeled individual, whatever may be the scale of resources, in knowledge, 
in tools, in human and material support, which he may require. 

Bronowski has pointed out that perhaps the most fundamental discovery 
of the scientific age was that Nature was to be approached and won, not by 
attempting to outwit her by magic, as many a medieval alchemist had 
imagined reflecting a prevailing climate of his time, but rather by discover- 
ing the true quality of natural laws and taking care to work within them. It 
is easy to forget how tremendous was that change of view, how much of trial 
and vision was comprehended within Newton's simple admonition that 
"science must be kept free from occult influences." The atmosphere of true 
research is still as it was when that great advance of philosophy was made, 
still the atmosphere in which, as Lionel Trilling has recalled, Faraday re- 
fused to be called physicist, holding the term too narrowly imprisoning a 
chamber for his life's commitment. These are the dimensions, whatever be 
the nature of the structures in which they are embedded, which still evoke 
the great advances of today. 

In the central context of discovery, it seems clear that the magnitude and 
organization of a research effort may be the least meaningful of parameters 
in any fundamental or enduring sense. One may indeed think of the large 
and the small research enterprises in our society as essentially symbiotic, 
each fulfilling its specific role — one more example of the rich diversity by 
which we live. 

The relation, however, is actually more subtle. The responsibility that 
devolves upon small and mobile groups dedicated to the exploration of new 
frontiers is clearly greater in our own clay than merely that of one component 
in a many-hued panoply of research. At least one aspect of the relation is far 
more serious, and wears a significance which must inevitably sharpen further 
in the coming years. It is not only important that the small and mobile re- 
search group be maintained and strengthened to ensure continuing 
advance along those remote boundaries of natural knowledge so vital to our 
spiritual as well as to our material well-being. It is not only important be- 
cause, in such a massive and highly advanced technical and engineering 
society as our own is today and must even more become tomorrow, the 
scientific "leverage" of such pioneering groups must inevitably increase. It 


is a further and a significant truth that, while climates that foster innovation 
can be maintained in the midst of complex and highly organized technical 
undertakings, preserving them intact is no common or easy achievement. It 
requires a particular determination, an extraordinary persistence of vision 
and pertinacity of will, an unusual sensitivity and skill, to sustain conditions 
favorable to original, exploratory research on remote and far-flung frontiers 
of the mind in massive working environments over considerable periods of 
time, undeflected by all the immediate demands that architecting to known 
ends in those environments inevitably imposes, in some multiple proportion 
of intensity to scale. Without the sustaining view that small and mobile 
groups attaining great discoveries can offer, without their inspiration, the 
task must become doubly difficult. These circumstances may define for the 
small and mobile group the most demanding and important of all its 
functions — the heavy responsibility of the keeper of a vision — the vision of 
the creating individual. 

In the future that responsibility may well become not only wider but yet 
more challenging. For it is abundantly evident that science and technology, 
in the world as a whole as well as in our own nation, have entered phases of 
development in our day so different in scale and complexity from their 
beginnings — or from what, incidentally, the newly developing nations of the 
world may confront or may require in their own immediate futures — as to 
differ essentially in kind. As Pierre Teilhard de Chardin has written with 
sensitive perception, "The Earth is covering itself not merely by myriads of 
thinking units, but by a single continuum of thought, and finally forming a 
functionally single Unit of Thought of planetary dimensions." An important 
aspect of the qualitative growth of contemporary science, of course, inheres 
in its essentially additive nature, in the formidable integration of knowledge 
and of thought characteristic of a pursuit where discoveries in one field may 
in the span of a few months alter the entire basis against which thinking in 
very different areas must be projected. Another concerns almost the op- 
posite situation. The significance of great research is largely measured by 
the impact of its results over a wide range of frontiers of inquiry, demanding 
the widest and swiftest communication possible and challenging human 
intellectual capacities for assimilation and generalization to their limits. 
But the processes of research bring heavy demands on quite opposite 
qualities — on extraordinarily detailed knowledge of a single field, on that 
supreme mastery of all its coordinates down to the most minute, developed 
over long periods of years, which so often is prerequisite to significant and 
sustained advance. In the past, science has been able to reconcile these two 
quite opposite requirements in tolerable fashion. With increase of scale the 
problem takes on new dimensions. 

Science in the last decades has responded to the challenge with enormously 
increased sophistication, with vastly expanded organization and integration 


of knowledge, with, indeed, quite a new development of recent years, the 
field of research on research itself. But as science has matured in its modes of 
cultivating the whole vast field of its thought, as its power has grown to enter 
and occupy new areas of research in force so soon as the first hint of them 
appears, these very qualities have brought novel and troubling consequences 
for the gifted individual, particularly for the gifted young research student 
just entering upon his life's work, upon whom so much of the future de- 
pends. As A. B. Pippard, among others, has pointed out dramatically, the 
legions of investigators can now be mobilized with such speed and effective- 
ness at a new and attractive breach in the frontier of knowledge that, 
particularly if the area offers a promise of practical benefit, a green and 
fertile intellectual valley can be reduced to aridity for the innovator within 
less than the working life of a generation of young scientists. The conse- 
quences incident to such swift and locustlike invasions, however effective 
and profitable they may be for a technical society in the large, can be dis- 
couraging to vulnerable individuals, and they bear at precisely the points of 
talent and dedication most precious to us. There can be no more urgent 
imperative than the creation of opportunity for individuals faced with this 
dilemma to address themselves once again to wholly new fields of inquiry. 
This too lies peculiarly in the domain of small and mobile and basically 
highly uncommitted research groups. 

What, in final essence, is the deepest meaning of the scientific way? In the 
profoundest sense, what is the meaning of the individual human life dedi- 
cated to it? Within the scientific context, as well as outside it, what, at last, 
are people for? A generation, perhaps even a decade, ago such a question 
was all but unasked by most Americans. Certainly it was all but unasked in 
1902. Even if put, in that day, it would have appeared to many not only 
irrelevant but quite possibly sinister. But in a world with a population 
estimated at nearly three billion and predicted by conservative demog- 
raphers to reach almost four billion by- 1980 and to attain nearly seven 
billion by the turn of the century, the question wears quite a different 
aspect. In our own nation, with a population now over one hundred and 
seventy million and destined perhaps to reach two hundred and twenty 
million by 1975, the revolutionary consequences of this flood tide upon 
every facet of the world we know demand no emphasis. It must profoundly 
affect every circumstance of our society, of its organization and its function. 
It must affect the individual's inner view of himself and his conception of his 
relation to his universe, his understanding and his reach in his own physical 
world, and much else besides. 

The rate of growth of the scientific effort today considerably exceeds that 
of the population as a whole. Inevitably, it would seem, it must change after 


two or three more periods of doubling. But in absolute terms it would seem 
beyond reasonable doubt that the legions of technically trained people in the 
future will vastly exceed in numbers those now active, even as these in turn 
so vastly exceed the numbers of only a few decades ago. Great technical and 
engineering efforts will be ready and available to confer rich meaning on the 
lives of many. In massive and compelling developmental undertakings op- 
portunities will continue to be provided to great numbers of active minds to 
labor for ends not only dramatic, not only economically and socially adap- 
tive, but as creative and as meaningful in our times as the tasks of the 
builders of Chartres or of the Parthenon must have been in theirs. Pippard 
has presciently pointed out that, if the field of technology is to prove 
sufficiently magnetic to attract first-class intellects to it, opportunities for 
the dramatic and the spectacular, outlets for the moral impulse to share in 
socially significant undertakings, the sheer intellectual quality of the under- 
takings themselves, must provide the motivations. Among the great and 
challenging technical and engineering undertakings of our time, all three 
motivations are presented on a scale the world may never have experienced 

But there will be other scientific workers, too, of other and less specially 
identifiable tastes and talents, hostages to a more distant future. For them 
the requirements will be quite different. Perhaps the deepest question the 
times can pose for them, and as well the most poignant for all man's spirit- 
ual welfare, will be this. In a society as densely packed, as intricately 
organized, as highly urbanized, as our own must inevitably become in 
future years, can small and mobile enclaves of thoughtful and imaginative 
men and women continue to maintain integrity and distinctive freedom 
within the greater society? On their ability to do so in the broadest context 
will depend in no small measure the fate of the individual and of those goals 
and motivations through which in the past we have lived and taken our 
national being. In a very real sense their persistence alone can effectively 
preserve the priceless jewel of the opportunity for quietness and temporary 
solitude which in our past has been so vital a nursery for individual American 
greatness as well as for that of our society as a whole. For it is the gifted, 
unorthodox individual in the laboratory or the study or the walk by the 
river at twilight who has always brought to us, and must continue to bring 
to us, all the basic resources by which we live. His position must be guarded 
and honored and implemented with every resource that we can muster, now 
and in the future, for he is irreplaceable. This matter too, and all the circum- 
stances attendant upon it, must be a central and abiding concern through 
all the coming years for the Carnegie Institution of Washington. As Chaucer 
said six hundred years ago, so may we today: "Out of the old fields cometh 
the new corn." 

The Year in Review 

It is fascinating to compare the Institution of approximately sixty years 
ago with that of today. There was, of course, very little to report from the 
first year or two of the Institution's existence, which was spent in a search 
for profitable lines of endeavor and experiments with organization toward 
that end. As early as 1904, however, the lines the Institution was to follow 
for some years were discernible, and the report for the year 1905 (Year Book 
4) describes the nature of the Institution's work in nearly all of the broad 
fields in which it was active during 1961-1962. Some glimpses of these early 
activities, set alongside typical activities in our several fields for 1961-1962, 
give a most illuminating view of the progress of the Institution, and indeed 
of science in the United States. 

In 1905 the resources and objectives of the Institution were much more 
widely dispersed than they are today. The total budget for that year was 
$586,000, a little more than half of which was allotted to ten " Departments 
of Investigations" which included the forerunners of all the Institution's 
present fields except embryology. Among the Departments were several that 
have since been terminated (Marine Biology, Economics and Sociology, 
History, Nutrition, and Horticulture). Half of the total budget for the 
Departments ($302,700) went to the Solar Observatory on Mount Wilson, 
which was under construction in that year. In addition, 43 individuals or 
organizations outside the Institution received grants to the sum of $130,625 
in the fields of anthropology, archaeology, astronomy, bibliography, botany, 
chemistry, geology, history, paleontology, philology, phonetics and linguis- 
tics, physics, and zoology. The Institution also had in 1905 a program of 
subsidizing outside publications of "meritorious works which would not 
otherwise be readily printed." Nearly $30,000 was expended in 1905 for this 
purpose and for the publication of works written within the Institution 

By contrast the Institution's budget for 1961-1962 was $2,848,480, all of 
which was spent upon the six operating Departments that have been 
maintained in recent years. Except for departmental fellowships the 
Institution made no outside grants and did not subsidize publication for 
works written outside the Institution. While a great variety of subjects was 



under investigation within the Institution in 1961-1962, research was under- 
taken in a better organized and more purposeful manner. 

Four of the more promising lines of research, as viewed by the President 
and Trustees of the Institution in 1905, lay in the work of its Solar Ob- 
servatory, in its Department of Terrestrial Magnetism, in geophysical re- 
search, and in biological investigations. With rather remarkable perception 
the importance of fundamental research in the physical and biological 
sciences is commented upon in the 1905 report. The Solar Observatory is 
described as ranking among Institution projects "first in order of cost for 
initial construction and equipment. This cost, however, is no more than 
commensurate with the magnitude of the problem attacked. . . ." Of the 
biological investigations, including those of the Station for Experimental 
Evolution and the Desert Botanical Laboratory, which was the predecessor 
of the Department of Plant Biology, the report noted that fundamental 
research in plant and animal biology "for a series of years can hardly fail to 
yield results of signal practical and theoretical value." 

The Department of Genetics 

In our series of "then and now" snapshots it is appropriate 

to begin with the Department of Genetics, whose prede- 
1905 cessor in 1905 was the Station for Experimental Evolution, 

one of the most active parts of the Institution in that 

Even though the Station for Experimental Evolution at Cold Spring 
Harbor had been in existence for only a little more than a year, a year of 
very full activity was reported. Following the inspiration of Hugo de Vries, 
who had given the dedication lecture at the Station the year before, C. B. 
Davenport described the long-range objectives of the Station's work. "The 
factors of evolution are three — variation, inheritance, and adjustment. 
Studies may be made on any one of these factors or on all three together; 
as a matter of fact, they can hardly be studied wholly independently. 
. . . Since studies in inheritance have been relatively neglected. . . our first 
efforts have been directed primarily toward such studies." 1 Already five 
principal investigators and the Director, Dr. Davenport, had commenced 
their programs of research. 

From a modern point of view the range of the work undertaken was 
astonishing. It was described as "investigations into inheritance and 
variability" of plants, insects, and other invertebrates; "investigations upon 

1 Year Book 4, p. 87. 


aquatic vertebrates"; "studies on inheritance in domesticated animals"; 
and "investigations into the cytological basis of heredity." Experiments 
were in progress on eight beetle species, three species of moth, flies, aphids, 
crickets, bees, and snails (Helix nemoralis). The brown trout and several 
killifishes (Fundulus sp.) were studied, and the Station experimented with 
goats, sheep, and cats. During the year George H. Shull became well started 
on the research which led to his later valuable knowledge of maize re- 
production. But in 1905 he was searching for suitable material for experi- 
ment, and had a garden of 81 different species of biennials, perennials, and 
annuals. Along with this search he conducted a variety of experiments, 
which included investigation of the inheritance of seed weights in beans 
(repeating W. Johannsen's experiments) and the vegetative habits of 
Russian sunflowers (Helianthus annuus) and other species. He had also 
begun his observation of the characteristics of maize. The particular 
character chosen for study in 1905 was the number of rows on the maize ear. 
Although the importance of cytological research was recognized, the 
year's effort failed to devise even a suitable experiment. The report ob- 
served, "The results of the last three years confirm the belief in the im- 
portance of the chromatic material in inheritance. This chromatic material 
exhibits a bewildering complexity and diversity scarcely less than that of 
adult organisms." 2 

It is interesting to find in 1961-1962 two lines of investi- 
gation which were at a germinal stage in 1905. Experiments 
1961-1962 with maize are still productive of fundamental results, and 
cytological research using flies (now the familiar Drosophila) 
formed an important part of the departmental program. 
Thus in one way or another these lines have held some of the departmental 
attention for more than 56 years. 

The approach of the Department in 1961-1962, however, was a vastly 
different enterprise. In a sense Barbara McClintock's methods of working 
with maize genes are lineal descendants of the variation and inheritance 
techniques that Shull was commencing to pioneer by counting rows of 
kernels on ears. But in Dr. McClintock's hands these methods have 
become highly sensitive and one of the sharpest tools in modern genetics. 
She has made them a match for other sharp new tools heavily dependent on 
chemistry and physics. For more than a dozen years she has been interested 
in the elements associated with genes that activate, control, suppress, or 
regulate genie action. Her work during these years has revealed the presence 
in maize of two controlling systems, an Activator (Ac) system, whose pres- 
ence or absence is associated with the appearance or nonappearance of 
mutations of a particular gene, and the Suppressor-mutator system (Spra), 

2 Year Book 4, P- 94. 


which causes a varied expression of the action of a single gene as observed 
in somatic cells. In her research this observation has been associated 
especially with the appearance of the reddish-blue pigment anthocyanin. 
Depending on its phase, the Suppressor-mutator element may either inhibit 
or activate the gene expression which results in the formation of antho- 
cyanin in maize leaves or kernels. 

A second theme of Dr. McClintock's work through these years has been a 
search for evidence that even the fine structure of inheritance is basically 
similar for all forms of life. In a much more general way Davenport and 
others started with the same hypothesis at the Station for Experimental 
Evolution, attempting to observe genetic expression in many forms of life. 
Dr. McClintock's first experimental evidence on the similarity of operation 
of genie control elements in different forms of life was reported in 1950. 3 
In that year she observed, " Because the same types of mutability as those 
observed in maize have been described for a wide variety of organisms, it is 
probable that the same events, involving the same chromosome materials, 
may occur in all organisms. 4 

During the year 1961-1962 Dr. McClintock continued to examine the 
parallels between the gene-control systems in maize and bacteria. She 
observes in her report that both organisms have gene-control systems 
composed of an " operator" element directly controlling genie activity ad- 
jacent to the structural gene and a "regulator" element acting upon the 
operator element. Other investigators have shown that the position of the 
regulator element on the bacterial chromosome may differ for individual 
systems. 5 It may be near to or removed from the locus of the operator 
element. Dr. McClintock's work during the year confirmed her hypothesis 
that there is a high probability that genie control systems in maize and 
bacteria act in similar fashion. She concludes her report by stating that 
her findings "are sufficiently extensive to leave no doubt that a two-element 
system of control of gene action, composed of an operator element at the 
locus of the gene and a regulator element located elsewhere, may arise 
at a gene locus that initially carried the regulator of the system." It 
would appear that one more link has thus been added to the gradually 
extending chain of evidence on basic similarities for many forms of life at 
the cellular level. 

A second field of departmental interest in 1905 survived to 1961-1962. 
This was the application of cytology to genetics, which was considered, but 
only futilely explored, in 1905. Indeed, successful development of this field 

3 Proceedings of the National Academy of Sciences, 36, 344-355, 1950. 

4 Year Book 49, p. 165, 1950. 

6 F. Jacob and J. Monod, On the regulation of gene activity, Cold Spring Harbor Symposia on 
Quantitative Biology, 26, 193-209, 394-395, 1961, presented completely for the first time evidence 
on the operator and regulator elements in bacteria. 


actually was postponed for more than 15 years after 1905, when in the 
1920's the work of John Belling finally laid the foundations for modern 
cytogenetics. This work was continued in 1961-1962 in the research of 
Berwind P. Kaufmann, Helen Gay, Margaret McDonald, and their associ- 
ates. The general objectives of the group bore some resemblance to the 
crudely stated convictions about the importance of cytology in the 1905 
report. The group continued its work of nearly two decades, charting the 
changes occurring in the organization of chromosomes and cytoplasmic 
organelles as cells in higher organisms grow and differentiate. Their methods, 
however, were a world apart from those of 1905, including as they did 
electron microscopy, fluorescent microscopy, enzyme chemistry, and bio- 
chemically specific stains. In addition, they had at their disposal the vast 
knowledge that has accumulated over 40 years on the genetic characteristics 
of Drosophila flies, which continued to be one of the objects of their ob- 
servations. A second material for study has been the plant Tradescantia 
(spiderwort family), which offers a very favorable opportunity for cyto- 
plasmic study during microsporogenesis. 6 Of particular interest has been 
the effort of this group to approach the problems of charting the submicro- 
scopic organization of chromosomes by means of "enzymatic dissection." 

All these techniques were employed during the year, adding to the results 
obtained in other years. Experiments were conducted on the mutagenic 
properties of deoxyribonuclease when introduced into Drosophila. An 
enzyme analogue, 5-bromodeoxyuridine, was added to the list of mutagenic 
agents employed on both Drosophila and Tradescantia. Perhaps the most 
interesting results from this group's program during the year were two 
discoveries: (1) The finding that direct chromosomal breakage occurs in 
Tradescantia root tips in the presence of 5-bromodeoxyuridine. This enzyme 
analogue acts by modifying the base sequences in nucleic acid rather than 
the phosphate-sugar helices attacked by deoxyribonuclease. (2) The 
observation that Golgi bodies, one of the types of cytoplasmic organelle, 
exhibit different forms in the progression of microsporogenesis in Trades- 

A third activity important to the 1961-1962 Department was not even 
dreamed of in 1905. It is represented in the work of Alfred D. Hershey and 
his associates, who are gradually charting the molecular structure of the 
viral chromosome. Dr. Hershey's work illustrates, more than anything else 
in the Department, the observation made by M. Demerec as early as 1942 
that "From the purely biological science of early days, genetics has de- 
veloped into a science where cooperation with physics, chemistry, and 
mathematics is essential." 7 Hershey and his associates observe in their 

6 Microspore = pollen. 

7 Year Book 41, p. 171. 


report of this year that methods have been devised in recent years to 
characterize and differentiate among different types of deoxyribonucleic 
acid (DNA) molecules. Among these methods are optical analysis of 
thermal denaturation, chromatographic analysis, measurement of fragility 
and buoyant density, and specific enzymatic tests. But these tests do not 
give information about molecular structure, which remains a more or less 
"plausible inference." Hershey's objective is to remove genetics' dependence 
on inference for its concepts of molecular structure of genetic material. To 
this end, he and his associates are experimenting with the DNA of several 
types of bacteriophage. 8 He considers these DNA's to be favorable material 
for experiment because: (1) they can be isolated in a molecularly homo- 
geneous state, permitting correlation between structure and biological 
function ; (2) their synthesis can be studied in infected cells that have been 
proved suitable for metabolic study in the past; and (3) present intensive 
study of the genetics of a few bacteriophage species gives valuable refer- 
ence points for physical and chemical findings. He considers his current 
work at least in part "exploratory." 

Several interesting results ensued from Dr. Hershey's exploration of 
physical techniques in measuring molecular weight during the year. In one 
he established the molecular weight of the DNA of a bacteriophage known 
as T5 by first establishing an ingenious pair of "scales" by analyzing DNA 
fragments of another phage (T2). One scale is established by determining 
sedimentation constants 9 of fragments of labeled T2 DNA as separated by 
column chromatography. 10 The other was obtained from fragility tests that 
measured the rate of breakage of T2 DNA fragments of a given sedimenta- 
tion coefficient when stirred in a mixer at a given speed. The sedimentation 
coefficient 9 and the fragility index of T5 DNA were then determined. By 
comparison with the T2 "scales" a molecular weight of 84 million was 
determined. The T5 DNA matched very closely fragments of T2 DNA in 
one sedimentation coefficient range (48.5-49.5). 

By similar techniques Dr. Hershey also brought to light during the year 
some interesting molecular characteristics of the DNA of phage lambda, 
which was found to have astonishingly different molecular properties from 
other well known DNA's. Of particular interest was a broad range of 
denaturation temperatures, like that of bacterial DNA's and contrasting 
with an exceedingly narrow range typical of other phage DNA's. On one 
hand these and other properties suggest a marked tendency of the molecules 
to interact with each other, and on the other, a remarkable differentiation 
in structure along their lengths. These exceptional properties may be 

8 Bacteriophage — any of a number of intracellular virus parasites of bacteria. 

9 Measure of the rate of precipitation of particles in suspension in a solution when centrifuged. 

10 Chromatography — a method of separating and analyzing chemical substances by inducing 
differential migration and adsorption from solution in a porous, insoluble, sorptive medium. 


related to each other and to some of the well known biological peculiarities 
of phage lambda. 

By infecting bacteria with isotopically labeled phage particles and by 
labeling DNA synthesized in the bacteria after infection, Dr. Hershey and 
Dr. F. R. Frankel have determined that cells subjected to such infection 
always contain a considerable fraction of their total DNA in a form indis- 
tinguishable from that found in finished phage particles. They note that 
this points to a mechanism for the preservation and determination of 
molecular length that operates continuously during DNA replication, not 
only at some terminal stage in the formation of the phage particle. This 
conclusion is considered significant evidence bearing upon several hypoth- 
eses about genetic mechanisms. 

The Department of Plant Biology 

The Department of Plant Biology also has developed from 
an operation under way in 1905. The Desert Botanical 
1905 Laboratory was active that year, located at Tucson, Arizona. 
The program in 1905 was not as varied as that of the Station 
for Experimental Evolution. Twelve investigators were 
associated with the Laboratory in that year, most of them as recipients of 
grants. As might be expected, their investigations were heavily weighted 
toward the characteristics of arid-region plants, especially transpiration 11 
and water-conducting mechanisms. A substantial amount of attention was 
paid to the character of plant environment, as in D. T. MacDougal's 
observations of soil temperature and B. E. Livingston's study of the 
relations of desert plants to soil moisture and evaporation. More typical, 
however, was F. E. Lloyd's study of correlation between stomatal 12 action 
and transpiration in certain types of desert plants. (No positive correlation 
was observed.) But along with these was displayed at least a secondary 
interest in what later became biochemistry and biophysics. For example, 
A. L. Dean conducted an " Investigation of the proteolytic enzymes of 
plants" and W. T. Swingle received a grant for an "Investigation of electro- 
magnetic and electrostatic effects on lines of force found in living plant 
cells." No conclusive results were reported from the latter study, but Dean 
reported finding an ereptic enzyme 13 in all tissues of a species of bean 
(Phaseolus vulgaris). 

11 Transpiration — the escape of water vapor from living plants. 

12 Stomata — minute pores in the epidermis of plants, through which gases and water enter or 
escape from the plant. 

13 A type of enzyme that breaks down proteoses and peptones, as in the intestinal tract of 


Most interesting about the program of the Desert Botanical Laboratory 
in 1905 was the complete absence of any attention to the problems of photo- 
synthesis, which have since become a major preoccupation of the Depart- 
ment of Plant Biology. Although the basic physical-chemical relations of 
photosynthesis 14 had been suggested sixty years before, there was no hint 
of the importance of these problems in the 1905 program. Th. W. Engelmann 
in 1887 discovered that light absorbed by pigments other than chlorophyll 
also produced photosynthesis, more than fifteen years before the establish- 
ment of the Laboratory. Even during the year of the 1905 report, the 
English plant physiologist, F. F. Blackman, demonstrated that photo- 
synthesis includes at least one "dark" reaction not initiated by light. 

The interest of the Institution in photosynthesis actually began six 
years later, in 1911, when H. A. Spoehr came to the Department of Botan- 
ical Research at Tucson, which succeeded the Desert Botanical Laboratory. 
Spoehr first came to the Institution to study the "chemical physiology" of 
plants but very soon became immersed in the problems of photosynthesis, 
an interest he maintained actively until his retirement in 1950. Just as 
intensively as in Spoehr's time the Department of Plant Biology today 
applies its research efforts to the great problem of unraveling the complex- 
ities of photosynthesis. 

The work of the Department in 1961-1962 on photosynthesis 
still centers on a problem the general outlines of which 
1961-1962 emerged in Engelmann's time : the exact function of the two 
sets of pigments, chlorophyll and the accessory pigments, 
both of which induce photosynthesis. It is now supposed 
that photosynthesis comprises at least two photochemical events, one driven 
by chlorophyll a, the other by the accessory pigments. Two discoveries 
made about 1955 provided some evidence for this hypothesis. One discovery 
was Blinks' chromatic transient effect, a momentary change in photo- 
synthetic rate observed when light absorbed by chlorophyll is changed to a 
color absorbed by accessory pigments. The other was Emerson's enhance- 
ment effect. In this effect photosynthesis resulting from wavelengths 
absorbed by chlorophyll a alone, when augmented by wavelengths absorbed 
through accessory pigments, is increased more than would be predicted 
from the simple sum of the effects from both radiations presented separately. 
A major effort is now being made in the world of research to define the 

14 Joseph Priestley demonstrated the production of "good air" (oxygen) by plants in 1772; 
Jan Ingenhousz in 1778 showed that the effect noted by Priestley resulted from the influence of 
sunlight; Jean Senebier noted in 1782 that "bad air" (carbon dioxide) was a necessary input; 
Lavoisier determined the composition of carbon dioxide in 1784; Nicolas de Saussure showed 
precisely in 1804 that water, light, and carbon dioxide were inputs, and oxygen plus organic 
matter outputs; Julius Mayer, through his concepts of the conservation of energy, in 1845 sug- 
gested the place of sunlight and vegetative organisms in chemical action taking place on a global 
basis at the earth's surface. 


nature of these two essential photochemical reactions and relate them to the 
chain of events in photosynthesis that results in oxygen evolution and 
carbon dioxide reduction. As throughout the long history of research in 
photosynthesis, ingenious theories currently exist to explain in detail most 
of the known effects. Generally considered, each investigator has his own 
favored concept of the process, and the different hypotheses are not entirely 
compatible with one another. Further experiments and more comprehensive 
concepts are still needed for an adequate understanding of photosynthesis. 

At the Department of Plant Biology, C. Stacy French and his associates 
continued their efforts to provide experimental evidence on the exact 
functions of the different plant pigments. 

A year ago they found in a red alga (Porphyridium cruentum) that 
chlorophyll a but not the accessory pigment, phycoerythrin, produces a 
chemically unidentified substance that rapidly consumes oxygen. Some of 
it is left over after a light exposure, as is demonstrated by the temporarily 
accelerated rate of oxygen uptake after an exposure to light absorbed by 
chlorophyll a. This material is also believed to be an intermediate in the 
process of photosynthesis. 

This year the persistence of the chemically unidentified material previ- 
ously formed by illumination of chlorophyll a was measured by French and 
Jeanette Brown. This was done by observing the increased oxygen pro- 
duction of the algae upon exposure to individual flashes of light at the 
wavelength absorbed by phycoerythrin. The presence of the material 
enhances the oxygen evolution by a light flash that activates phycoerythrin. 
The half-life of the material measured in this way was found to be about 
18 seconds under certain conditions. By contrast, preillumination by 
phycoerythrin-absorbed light did not enhance oxygen production when 
chlorophyll a was subsequently activated. 

Another series of experiments, made this year, shows even more complex 
relations between the effects of different pigments of green leaves. The story 
began about eighty years ago, when Engelmann found traces of oxygen 
evolution from isolated chloroplasts. This effect was further investigated by 
Molish early in this century, but since then the reaction has had very little 
attention until recently, no doubt owing to R. Hill's discovery in 1937 that 
the addition of oxidants such as ferricyanide greatly increases the amount 
of oxygen produced. An avalanche of papers on the Hill reaction followed, 
and experiments with the evolution of oxygen from within chloroplasts 
without added substances have been all but abandoned. 

In the past year, however, Y. de Kouchkovsky of the Centre National de 
la Recherche Scientifique, Gif-sur-Yvette, France, and David C. Fork of 
the Department of Plant Biology, have reexamined this effect with greatly 
improved methods. The work, started independently at the two laboratories, 


was continued as a collaborative effort during Dr. Fork's visit to Gif-sur- 
Yvette in March 1962. 

By measuring oxygen exchange of Swiss chard chloroplasts Fork showed 
that it is possible to distinguish four separate effects of light, each with its 
characteristic action spectrum. They are: 

1. The evolution of oxygen from chloroplasts without added oxidants is 
driven most effectively by light having a wavelength of 650 millimicrons 
(red). 15 This corresponds to the absorption peak of chlorophyll b in chloro- 
plasts, thereby showing that chlorophyll b is more effective than chlorophyll 
a in this reaction. A shoulder on the curve of the action spectrum, however, 
shows that at least one of the three forms of chlorophyll a is also active. 
This oxygen production within the chloroplast goes rapidly for only a few 
seconds, then its rate drops to a very low value. Storage in the dark revives 
the ability to evolve oxygen. Apparently light consumes some material 
found in chloroplasts which is restored in darkness. 

2. Dr. Fork found the recovery process to be strongly accelerated by 
exposure to far-red light. A wavelength of about 730 millimicrons was most 
effective for this purpose. This wavelength suggests identity with phy to- 
chrome, a substance which, though present in very small amounts, controls 
many plant responses. In addition to the 730-millimicron peak, however, 
the action spectrum for the regeneration of the chloroplasts' ability to 
evolve oxygen also has a peak in the blue wavelengths which does not 
activate phytochrome. 

3. Ferricyanide [K 3 Fe(CN)e], when added to chloroplasts, substitutes for 
the natural oxidant substance responsible for photoproduction of oxygen. 
The rate of oxygen evolution remains for long light exposures, and the action 
spectrum, which peaks at 678 millimicrons, shows that chlorophyll a is 
more effective than chlorophyll b when ferricyanide is present. 

4. A very specific inhibitor for oxygen production by chloroplasts is the 
herbicide DCMU. 16 When this poison is added to chloroplasts the photo- 
consumption of oxygen can be measured without interference by oxygen 
evolution and shows a maximum efficiency at wavelength 690 millimicrons 

Four different action spectra have thus been measured for oxygen 
exchange in isolated chloroplasts. French raises the question of the exact 
function of each pigment in these various photoprocesses. He says that the 
answer is clear for chlorophylls a and b (678- and 650-millimicron peaks) : 
they are concerned with oxygen evolution. But it is not yet known why 
chlorophyll b is more effective than chlorophyll a for the reaction within the 

15 One millimicron = 10 -6 millimeter. 

16 3-(3,4-Dichlorophenyl)-l,l-dimethylurea; manufactured by E. I. du Pont de Nemours and 


natural chloroplast whereas the reverse is true when ferrieyanide is added. 

The two action spectra with peaks at 730 and 690 millimicrons are more 
obscure. They do not necessarily indicate that there are active pigments 
with absorption maxima at either wavelength. Instead, spectra may result 
from the activation of two pigments whose reactions either reinforce or 
counteract each other. In both cases the action spectra maxima may differ 
greatly from the absorption maxima of the reacting pigments. These are 
interesting subjects for further investigation. 

Ellen C. Weaver started an attack on the problems of photosynthesis 
with an intriguing and promising new technique, that of electron para- 
magnetic 17 resonance (EPR) spectroscopy. She notes in her report the well 
established fact that illuminated chlorophyll-containing material has a 
higher level of unpaired electrons than material in the dark, suggesting that 
some phase of photosynthesis proceeds by single-electron transfers. Even 
though several research groups outside the Institution had employed this 
new technique (about six years old) in studying photosynthesis, no rigorous 
demonstration had yet been made that electron resonance 18 had a direct 
connection with photosynthesis. 

Dr. Weaver set out during the year first to determine whether or not the 
established resonance was associated with chlorophyll. She observed two 
distinctly different light-induced resonances. One is the R (rapid-decaying) 
signal, seen only when cells are illuminated. The other may persist for hours 
in the absence of light, and it is designated the S (slow-decaying) signal. 
Using a yellow mutant (no chlorophyll) of the fresh-water alga Chlamy- 
domonas reinhardi, Dr. Weaver obtained no R signals in EPR observation, 
suggesting that the R signal is ascribable to chlorophyll. She also discovered 
by using dilute cell suspensions that 680-millimicron light (near the absorp- 
tion peak for chlorophyll a) was the most effective for producing the R 
signals. Another interesting result is her discovery that the amplitude of 
the R signal has a strictly linear (proportional) relation to light intensity 
for the wavelengths least absorbed by chlorophyll, whereas wavelengths 
most strongly absorbed by chlorophyll have no linear relation to light 
intensity (assuming low light levels in both cases). Dr. Weaver's tentative 
conclusion from these observations is that the R signal is associated with 
chlorophyll and arises from the "primary" act of photosynthesis. 

Dr. Weaver also discovered that any inhibition of oxygen evolution, as 
by DCMU or by limiting the manganese-ion concentration in the growing 
medium, will produce an enhanced R signal. This suggests that if the 

17 Paramagnetic — atoms having spin systems with magnetic moment (or materials containing 
those atoms) are paramagnetic. 

18 Electron resonance — a property of unpaired electrons, whereby precession of the spinning 
electron may be inferred when it is subjected to an electromagnetic field at a specific frequency, 
as in EPR spectroscopy. 


pathway of the electrons is in any way obstructed the net level of unpaired 
spins rises. The result indicates further that the alteration of photosynthetic 
processes other than oxygen evolution may provide a fruitful field for 
experiment using the EPR spectroscopic technique. Interestingly, the S 
(slow-decaying) signal is not altered by blocking the oxygen evolution 
pathway with DCMU, but manganese starvation reduces that signal to an 
extremely low level. It is thought that this result may be correlated with a 
lack of plastoquinone, 19 previously determined elsewhere to be a necessary 
and apparently universal factor in the oxygen evolution of green plants. 

Dr. Weaver has thus presented evidence that chlorophyll is the source of 
one type of free electrons in an intact photosynthetic organism and that 
plastoquinone is the site of another type. She has also demonstrated the 
correlation of the two types of signals with the evolution of photosynthetic 
oxygen. The method and her results are of more than usual interest, because 
photosynthesis is essentially a photoreduction process when viewed in a 
highly general way, that is, the transfer of electrons from one substance 
to another. 

Although photosynthesis still presents an awesome complexity to those 
investigating it, studies like those of French, Fork, Brown, and Weaver 
examining the effects of light on metabolic reactions are continually 
changing concepts of how synthesis takes place and, step by step, are build- 
ing a more complete understanding of this vastly important phenomenon. 

Another field in plant biology, experimental taxonomy, can trace its 
origin to the activities of the 1905 Desert Laboratory. Again, however, the 
diffuse approach of 1905 is gone. William M. Hiesey and his associates note 
in the 1961-1962 report that current developments in precise techniques 
have greatly extended the horizon of this field. Instead of the compart- 
mentalizing of botanical study, which was commencing in 1905, they see 
"a truly integrated plant science whereby contributions from the various 
specialized fields, including taxonomy, ecology, cytology, genetics, physi- 
ology, developmental morphology, and biochemistry, can be incorporated 
in a panoramic view of plant relationships and evolution." Their goal is an 
integrated understanding of the chain of mechanisms that determine plant 
evolution, including the genetic and the biochemical. For a number of years 
plants of the genus Mimulus 20 had been used for comparative growth 
studies of altitudinal effects at the Stanford, Mather, and Timberline 
stations. More recently the races of one species, Mimulus cardinalis, have 
been subjected to controlled growth chamber experiments. 

During the year Harold W. Milner made some particularly interesting 

19 Plastoquinone — quinone found in chloroplasts ; the structure of this compound is given in 
figure 31 of the report of the Department of Plant Biology. 

20 The garden "monkey flower" belongs to the Mimulus genus. 


studies of the photosynthetic rates of six races of the species originating in 
diverse climates and altitudes. Among the variations observed were a 60 
per cent difference among the races in the light intensity required to saturate 
photosynthesis at high temperature, and a 100 per cent difference at a very 
low temperature (0°C). Significant variance in photosynthetic rate at 
extreme temperatures also was observed, as well as disparate abilities to 
maintain a high rate of photosynthesis over a long period. From these and 
other results one may conclude that climatic races within the same species 
may show differential patterns of response undoubtedly linked with vari- 
ations in internal physiology. 

During the year an important step was taken toward establishing tissue 
cultures from Mimulus plants, so as to make quantitative measurements of 
growth and photosynthetic rates in tissue cultures similar to those for whole 
plants. By examining the physiological requirements of tissue from various 
plant organs, it should be possible to localize the site of physiological 
differences within the plant. 

In addition, the group extended its work during the year to species of 
Solidago (goldenrod), particularly in the collaborative work of Malcolm 
Nobs of the Department working at the Institute of Plant Systematics and 
Genetics at Uppsala, Sweden. The same type of difference in response to 
light intensity was observed between two races of Solidago virgaurea: one a 
shade-loving race from Sweden and the other an alpine race from Norway. 
The alpine race has a much higher requirement for light saturation than 
the shade race, and its chloroplasts remain normal at light intensities that 
cause the disintegration of those from the shade race. 

The Department of Terrestrial Magnetism 

The Department of Terrestrial Magnetism was also among 
the active Departments of the Institution in the year 1905. 
1905 The work of the Department in 1905 faithfully followed its 
name, although a wide range of projects was reported, with 
activity on an almost worldwide basis. A major preoccupa- 
tion of the Department during that year was an effort to start a systematic 
series of magnetic observations on most parts of the globe, which at that 
time were informational blanks. L. A. Bauer, Director of the Department 
in that year, stated, "our progress with regard to the great and principal 
facts of the earth's magnetism will be at a standstill unless a magnetic 
survey of the whole globe be undertaken immediately." Toward that end a 
wooden sailing vessel, the brig Galilee, had been manned and outfitted, and 
had undertaken trial runs. This was the beginning of a program that 


continued for almost 25 years thereafter, in which the sailing vessels Galilee 
and Carnegie logged more than 400,000 miles to undertake magnetic and 
other scientific observations in every ocean area of the globe. It ended only 
with the accidental destruction by fire of the Carnegie at Samoa in Novem- 
ber 1929. 

A very extensive land survey program also was being initiated for 
magnetic observations in 1905. Many of the islands of the West Indies were 
covered in that year, and arrangements were being completed for observa- 
tions on the South Pacific Islands and in Canada, Mexico, Central America, 
South America, and China. Cooperative arrangements for observations and 
research were maintained with several German scientific institutions and 
with the St. Petersburg Academy of Sciences in Russia. 

Besides its primary program on the study of and basic data collection for 
terrestrial magnetism, the Department in 1905 organized and participated 
in the program of observing the solar eclipse of that year, and it began 
cooperating with the Institution's Solar Observatory in the study of several 
solar phenomena. 

The Institution in 1905 also expressed a substantial interest in physics 
research, but entirely through a program of grants to fourteen American 
physicists. Among the grants were several for studies of emission spectra 
and a study of the theory of light. 

Although the emphasis so prominent in the 1905 program 
of the Department of Terrestrial Magnetism was continued 
1961-1962 until the early 1930's with relatively slight changes, the 
program of 1961-1962 in the Department was a much differ- 
ent one. The principal activities reminiscent of the earlier 
days of the Department came in the research of Scott E. Forbush, but again 
in an environment strikingly different from that of the first twenty-five 
years of the Department. Forbush's principal investigations during the 
report year were devoted to the intensity of the charged particles in the 
Van Allen trapped-radiation belt adjacent to the earth, as recorded during 
the transits of the satellite Explorer VII through the belt between 1959 and 
1960. He also had under way studies examining the southward shift of the 
auroral-zone current system during magnetic storms in its probable associ- 
ation with particles coming from the outer Van Allen belt. 

The bulk of the Department's varied and imaginative research in 1961- 
1962, however, derived from applying the techniques of physics to a wide 
variety of geophysical and biological problems. They ranged from the ex- 
amination of the interior of living cells to charting the hydrogen clouds of 
our Galaxy. 

Perhaps the most significant results to emerge from the year's work were 
from a quarter that could hardly have been envisioned as associated with 


the Department even twenty years ago. They came from the work of the 
Biophysics Section (E. T. Bolton, R. J. Britten, D. B. Cowie, B. J. Mc- 
Carthy, J. E. Midgley, and R. B. Roberts) on the fine structure of, and 
biochemical processes taking place within, bacterial and other cells. 

As the end of the report year approached, the Section was engrossed in 
some striking experiments involving "messenger" ribonucleic acid (RNA). 
This type of RNA contains nucleotide 21 sequences complementary to those 
in the appropriate DNA which provides the genetic information. In follow- 
ing a lead provided by E. K. F. Bautz and B. D. Hall at the University of 
Illinois it was discovered that single-stranded DNA could be immobilized 
in agar and complementary RNA could be caused to hybridize with it 
through the formation of hydrogen bonds. By washing, the immobilized 
hybrid DNA-RNA combination was freed of other contaminating RNA. 
The hybridized RNA could then be reclaimed, in a state of high purity, by 
dissociation of the hydrogen bonds, and could be chemically analyzed. 

With this simple and effective new procedure it has been possible to 
demonstrate that the DNA-like RNA comprises about 1 per cent of the 
total RNA of bacterial cells and that it has a half-life during active syn- 
thesis of approximately 2 minutes. On the assumption that this RNA is in 
fact the active template for protein synthesis, the measurements of its 
quantity and half-life show that a single molecule acts catalytically for the 
synthesis of many polypeptide 22 chains. 

Further work has revealed that the method can be used to exploit the 
specificity inherent in the hybridization process, which depends upon long 
regions of complementary nucleotide sequences in molecules of RNA and 
DNA. Thus, RNA from bacteriophage T2 will hybridize well with DNA 
of the genetically closely related phage T4 but not with the apparently 
unrelated T7 DNA. Several species of bacteria have also been tested, and 
cross reactions have been found to occur to a greater or lesser degree in 
accord with accepted taxonomic relationships. Thus, the method has made 
feasible a quantitative chemical analysis of the amount of genetic informa- 
tion held in common among species. 

Since the method is a general one, applying to the DNA of all species and 
tissues, it can be used in studies of the transcription of genetic information 
and of differentiation, two of the key subjects of modern biology. 

During the year the Biophysics Section also contributed a new hypothesis 
about the code associated with the role of nucleic acid in specifying the 
order of amino acids in protein. The prevailing hypothesis interprets ex- 
perimental findings in terms of a "three-letter" or triplet code. The experi- 

21 Precursor of or decomposition product from nucleic acid, composed of a nitrogenous base, a 
ribose sugar, and phosphoric acid. 

22 Peptides are proteins linked by amide (RCO-NHR/), or "peptide," bond. 


ments of the Section lead its members to believe that a two-letter or doublet 
code eliminates the major failing of the triplet code, which implies an un- 
realistically high uridylic acid 23 content for the "template" material of 
protein synthesis. The doublet code apparently provides a good correlation 
between the amino acid composition of the bacterially synthesized protein 
and the nucleotide composition of the RNA templates on which it is formed. 

Other fields in which the techniques of physics are being applied by Staff 
Members of the Department are seismological exploration of the earth's 
crust, radioactive dating of rocks, radio astronomy, and the development 
of image tubes for use in astronomical studies. 

It is of particular interest that all these programs in one respect or another 
are cooperative, carrying on the tradition of joint investigations or joint 
enterprise which was started and even widely used in the earliest days of 
the Department. As Merle Tuve, the Director of the Department, observes 
in the introduction of his 1961-1962 report, " 'cooperation' . . . has many 
very different aspects in the current work of the Department, but in each 
case it represents a situation where there is special usefulness in our freedom 
of initiative and recognition of the infectious characteristic of personal 
enthusiasm." To some extent, the same thing might have been said for the 
programs in biophysics and geomagnetic studies. 

A good example of the Department's cooperative approach is shown in 
its radio astronomy program. With the support of the National Science 
Foundation a new Carnegie Radio Astronomy Station will soon be estab- 
lished in Argentina. Parts for a major instrument, a parabolic antenna 
nearly 100 feet (30 meters) in diameter, are now being manufactured in 
this country and will be shipped to Buenos Aires for assembly there during 
1962-1963. The Argentinian National Council for Scientific and Technical 
Investigations and the Research Council of the State of Buenos Aires have 
created a new National Institute of Radio Astronomy to participate in the 
construction and operation of the station. Later operation will be a cooper- 
ative venture among the Carnegie Institution, the University of Buenos 
Aires, and the University of La Plata. Invitations will be extended to 
astronomers in other institutions in South America to participate in the 
research program. Some fellowships are being offered by the Institution to 
bring students and professional research men interested in radio astronomy 
to this country for training in the use of parabolic antennas and for ac- 
quiring educational background in radio astronomy. 

The observational program in radio astronomy using the Department's 

23 Uridylic acid — a nucleotide; technically uracil (2,6-dioxypyrimidine) + D-ribose sugar -f- 
phosphoric acid. 


instruments also continued during the year. Observations of the hydrogen 
gas content at the center of our Galaxy confirmed previous observations at 
Leiden, the Netherlands, and Sydney, Australia, that the motions of hy- 
drogen close to the Galactic center are complex, and that the hydrogen gas 
not only is rotating about the center of the Galactic mass but also is ex- 
panding. Because of its latitudinal position, the Derwood, Maryland, 
Station of the Department was able to extend observations nearly 20° 
farther south along the Galactic plane than the Dutch station. 

The Department also decided during the year, after considerable experi- 
ment, to begin construction of an interferometer array from parabolic 
dish antennas, to be able to obtain precise positions of radio noise sources 
in the sky. A 30-meter dish closely following the design of the Argentinian 
radio telescope is now being constructed at Derwood and will be used with 
the existing 60-foot parabolic antenna as a two-element interferometer. 
These two antennas will be employed in experiment to evaluate the po- 
tentialities of such a system in determining precise radio-star positions. 

A second cooperative venture of the Department in the area of astro- 
nomical study has been the work of the Committee on Image Tubes for 
Telescopes, of which Merle Tuve is chairman. In this the Department has 
collaborated with the Mount Wilson and Palomar Observatories, the Lowell 
Observatory, the National Bureau of Standards, and the United States 
Naval Observatory to develop electronic image tubes for magnifying signals 
received on optical telescopes. This work has also been supported in large 
part by generous grants from the National Science Foundation. 

During the year the Committee continued the testing of tubes manu- 
factured experimentally upon its order by the International Telephone and 
Telegraph Corporation Laboratories and by the Radio Corporation of 
America. The tests conducted were largely undertaken by W. K. Ford, Jr., 
of the Department. The tubes proved to have better operating character- 
istics than the Committee had hoped for only three years ago. Telescope 
observations were made at the Lowell Observatory with the tubes to ex- 
amine their reliability and effectiveness, and laboratory investigations were 
conducted to distinguish among the relative merits of the several tubes. 
On the basis of the spectrographic tests from telescope observations and 
the laboratory tests, the Committee believes that the two types of tubes 
recently examined (mica-window and cascaded) will have wide application 
in astronomy because of their advantages over conventional photography. 
Development will be continued, again with the support of the grant from 
the National Science Foundation. 

A major project of the seismic studies group in the Earth's Crust Section 
of the Department (J. S. Steinhart, L. T. Aldrich, M. A. Tuve, and associ- 
ates) was an intensive study of the earth's crust in Maine, in which col- 


leagues from the University of Wisconsin, Princeton University, Penn- 
sylvania State University, the University of Michigan, and the Woods Hole 
Oceanographic Institution participated, and the United States Coast Guard 
assisted in detonating explosions in the Gulf of Maine in July 1961. 

The data obtained from the explosions have since been the subject of 
appraisal to determine the application of explosion seismology to designation 
of crustal structures. This is a very real geological problem, because the 
traditional conception of the earth's crust as one or more horizontal layers 
of constant seismic wave velocity has appeared inadequate for more than a 
decade. Efforts to find the proper reflections from the surfaces of the sup- 
posed layers have been unsuccessful; and laboratory measurements of 
seismic velocities in various rock types contradict the layer hypothesis. 
Field evidence suggests significant lateral as well as vertical differences in 
structure. Several models that might conform to the seismic results received 
from the explosions were therefore constructed. 

On the basis of these models it seems fairly certain that in Maine the 
Mohorovicic discontinuity 24 lies at 36 ± 3 kilometers below the surface. 
The most likely models suggest that the upper 3 kilometers of the crust is 
granitic and that below the granite the percentage of gabbro 25 increases at 
a rate that maintains a steady gradient in seismic wave velocity change to 
a depth of about 20 kilometers. These findings are of interest geologically 
in that they postulate appreciably less granitic material than is customarily 
thought to be in a continental crust. 

The radioactive dating group is not only interinstitutional but also 
interdepartmental (L. T. Aldrich and S. R. Hart of the Department of 
Terrestrial Magnetism, G. L. Davis, G. R. Tilton, and B. R. Doe of the 
Geophysical Laboratory and associates). During the year the Department 
of Terrestrial Magnetism members of the group participated in an exchange 
program with the Geological and Mineralogical Institute of the University 
of Kyoto. 

Dr. I. Hayase, of the University of Kyoto, spent part of the year at the 
Department becoming familiar with its techniques of measuring mineral 
ages. In the course of his visit he analyzed samples collected in Japan. The 
data were of interest as the first measurement of the kind from Japan. They 
showed no contradictions between the isotopically determined ages and 
ages implied by geological structure. They also showed discordances between 
rubidium-strontium and potassium-argon age determinations commonly 
enough to indicate a complex geological history for the Islands. 

As a second part of the exchange, L. T. Aldrich of the Department is now 

24 A phenomenon recorded in the changing speed of seismic waves at certain depths. 

25 A granitic rock formed of plagioclase (light-colored) feldspar and a monoclinic pyroxene like 
augite (dark-colored). 


in Kyoto as a visiting professor at the University. He is assisting in the 
establishment of a complete laboratory for the measurement of mineral 
ages. To facilitate this work the Department constructed and shipped to 
the University a mass spectrometer 26 which Dr. Aldrich now has in opera- 
tion at the Institute there. It is expected that the spectrometer will serve 
as a model for similar equipment to be built elsewhere in Japan. We hope 
that this particular interinstitutional collaboration will continue indefinitely. 

The Geophysical Laboratory members of the group also worked with a 
staff member of the Geological Survey of Finland, O. Kouvo, on the dating 
in two orogenic (mountain-building) belts in Finland: the Karelian belt 
extending from southeastern Finland northwesterly to Finnish Lapland, 
and the Svecofennian extending east-west in southern Finland. It is gener- 
ally believed by geologists that the Svecofennian belt is older than the 
Karelian. The radioactive dating work, however, gives strong evidence that 
the intrusion of igneous rocks occurred about 1.9 billion years ago in both 
orogenic belts, and the two orogenies therefore are approximately con- 

The radioactive dating group has also compiled a new map of age dis- 
tribution in crystalline basement rocks of North America. This shows one 
belt of rocks, ranging from 0.9 to 1.2 billion years old, extending from 
Labrador to Texas; another, 1.2 to 1.55 billion years old, occupying a large 
part of the central and southwestern part of the country; a third, 2.0 to 2.8 
billion years of age, from the Rocky Mountains northeastward over the 
Laurentian Shield to Quebec; and still another, 1.55 to 2 billion years old, 
in Alberta and northwestern Canada. A picture of the geographical differ- 
entiation of ancient rocks in North America is thus beginning to emerge. 

In cooperation with the University of Basel, Switzerland, the Department 
completed the installation of a polarized ion source in the departmental 
accelerator during the year. It consists of a discharge tube for the production 
of atomic hydrogen, diaphragms and pumps for defining the atomic beam, 
a quadrupole magnet for selecting and focusing the atoms having the de- 
sired orientation, an ionizer for the atomic beam, and a device for pre- 
accelerating and focusing the ionized atoms. The machine was operated 
successfully. It is planned to use the polarized deuteron beam in the study 
of a number of nuclear reactions, thus returning the Department more 
directly to the field of nuclear physics than at any time since the end of 
World War II. For more than fifteen years after the mid- 1 920 's the 
Department maintained a pioneering effort in nuclear physics, operating 
one of the first accelerators in this country. 

26 Mass spectrometer — an instrument for determining the masses of atoms or molecules in a 
gas, liquid, or solid. In it a beam of ions is directed through electric and magnetic fields so as to 
produce a mass spectrum identifiable by an electrical detector. 


The Geophysical Laboratory 

Other than the Terrestrial Magnetism program, geophys- 
ical research in 1905 was not carried on within the premises 
1905 of the Institution but nonetheless was considered an im- 
portant part of the total program. It was the type of project 
that President Woodward advocated continuing, in his 
" Suggestions Concerning Pending Problems of the Institution." 27 Indeed, 
a large part of the total geophysical program in that year was carried on 
in close collaboration with the United States Geological Survey in Wash- 
ington, thus commencing a friendly professional relation that has continued 
ever since. The two principal investigators of that year, Arthur L. Day and 
G. F. Becker, held appointments in the Survey even though a substantial 
proportion of Dr. Day's time was spent on Institution projects. Included 
was a three-month visit by Dr. Day to Europe for the purpose of studying 
laboratory equipment for geophysical research and making an inventory 
of European research. 

Becker's research was concerned entirely with an effort to determine 
experimentally the relation between stress and strain. The main part of his 
apparatus was a 3-inch tube 480 feet long erected in the Washington Monu- 
ment, within which steel tapes were suspended. He made some observations 
by means of this equipment during the year. In another project, F. D. 
Adams of McGill University conducted experiments on the cubic compressi- 
bility, the modulus of shear, and the flow of rocks, in which hundred-ton 
pressures were used. 

The heart of the 1905 program, however, lay in the work of Dr. Day. 
Much of his time was spent in setting up his newly designed laboratory 
equipment. It comprised, among other apparatus, a furnace capable of 
reaching 2100°C in oxidizing or reducing atmospheres, a large electric 
furnace in which pressures up to 500 pounds or a vacuum could be main- 
tained, and a water-pressure plant capable of reaching 2000 atmospheres. 
A similar plant capable of reaching 3000 atmospheres was under construc- 
tion. Dr. Day's research included the completion of a three-year investi- 
gation of the lime-soda feldspar 28 group of rocks. His results showed "that 
the lime-soda feldspars form a continuous series of mixed crystals capable 
of stable existence in any proportion of the two component minerals. " 
Experimental proof of this isomorphism was established by correlating 
melting points with change in the mixes of the two components. Experi- 
ments also were conducted on wollastonite (CaSi0 3 ), determining for the 
first time the exact temperature of crystallization of this mineral as found 
in nature. 

27 Year Book 4, pp. 28-29. 

28 Feldspar is one component of granite. 


Inspired by his thought on silicates, Dr. Day already was looking toward 
the future, as he mentioned two practical problems to which his laboratory 
later contributed most significantly. He notes that "the study of lime-silica 
mixtures is fundamental in the preparation of Portland cement. Questions 
of technical interest in glass manufacture reappear everywhere in handling 
silicate solutions." 29 He concluded in a satisfied vein, "grave doubts were 
entertained as to the feasibility of handling physical phenomena at high 
temperatures with anything like the certainty attained at ordinary tem- 
peratures, but the experience of this first year has justified the effort ... ." 
If Dr. Day could look in on the Geophysical Laboratory of 
today he should feel greatly gratified, both because his 
1961-1962 beginning work in 1905 accurately forecast a direction and 
method of research that continues to be highly productive 
after nearly sixty years and because of the enormously great 
range and resolving power of the methods now in use. 

The techniques upon which the Laboratory depend have become enor- 
mously more powerful and more sensitive than in Dr. Day's time. The 
3000-atmosphere pressures, which were tremendous to Dr. Day, have been 
succeeded in 1961-1962 by pressures of 100,000 atmospheres. Moreover, 
these elevated pressures can be employed in combination with almost any 
temperature needed in geophysical experiment. In Dr. Day's 1905 experi- 
ments, high temperatures could be accompanied by a pressure of only a 
few hundred pounds. The present-day Laboratory has firm grasp of these 
tools, and it applies them to the whole range of problems on the frontiers 
of modern geology. From the first explorations of the potentiality of these 
geophysical techniques it has arrived at the full power of applying them to 
revelation of the earth's interior and its history. Furthermore, the capacities 
of the Laboratory now include a wide variety of techniques — beyond those 
of high temperature and high pressure — taken from modern physics, 
chemistry, and mathematics. The Department of 1961-1962 included work 
in experimental petrology, statistical petrology, crystallography, ore 
minerals, meteorite analysis, geothermal calculations, the ages of rocks and 
minerals, and organic geochemistry. 

Among the numerous investigations carried on in these fields in 1961- 
1962, three will be described briefly to illustrate more in detail the charac- 
teristics of research at the Geophysical Laboratory. These are experimental 
petrology, in which much of the work this year was focused on pyroxene 
minerals, and emphasized the study of phase equilibria 30 at higher pressures; 

29 Later work of the Laboratory made fundamental contributions to the technology of both 

30 In chemical terms, any crystalline compound or liquid is a phase ; hence, a mineral separated 
from a rock is also a phase. Assemblages of phases (or minerals) which do not melt or react at a 
particular temperature and pressure are said to be at equilibrium. Study of these mineral equi- 
libria is a means of understanding the conditions of formation of rocks. 


the mineralogy of meteorites; and organic geochemistry, including analysis 
of Precambrian carbonaceous materials. 

The program of studying the mineralogical composition of meteorites, 
begun last year, continued to produce most interesting results. Particularly 
relevant as a preview of the solid matter to be found in the spatial environs 
of the earth, the meteorites studied are continuing to yield mineralogical 
surprises. P. Ramdohr and G. Kullerud examined more than a hundred 
stony meteorites during the year, rinding in them fourteen new minerals 
thought to be observed for the first time anywhere. Only one of them has 
been given a name, the others being referred to simply by letters of the 
alphabet for the time being. Because they occur in amounts too small to 
permit performance of standard chemical analyses or X-ray powder 
diffraction studies, the component elements in only two have been identified. 
These were a nickel-iron sulfide [(NiFe) 2 S] called the Henderson phase, and 
a colorless mineral of spinel 31 type (Mg 2 Ti0 4 ). Several of the remaining 
twelve minerals are thought to be sulfides, and one, having an hexagonal 
layered structure, seems to be a compound of iron, carbon, and sulfur. One 
is thought to contain arsenic. The electron probe is considered to have 
promising potentialities for assisting in the chemical identification of these 
minerals. Another method of identification of the new phases is synthesis, 
once the major constituents are surmised from deductions about the origin 
of the minerals. Ramdohr and Kullerud state that their efforts in this 
direction are increasingly successful. 

Ramdohr and Kullerud also made a number of observations on distinctive 
structural and textural phenomena in meteorites. They include evidences of 
mechanical distortion and crystallization in many meteorites, evidence of 
spontaneous melting in the interior of many, and the effects of terrestrial 
weathering, which may yield products that may be mistaken for primary 
components. Magnetite (Fe 3 4 ) frequently may be such a product. In 
another set of analyses on meteorites, S. P. Clark, Jr., has identified an 
unknown mineral in tektites (glassy bodies probably of meteoric origin) as 
schreibersite (Fe 3 P). He concludes that the content of minor elements in 
meteoric bodies, like sulfur, phosphorus, or carbon, should be helpful in 
identifying the number of meteoric falls in complex fields like those of 
southeast Asia or Australia. Presumably the minor elements would be the 
same in each fall but would differ in separate falls. 

Another development of 1961-1962 meriting special mention is the study 
of phase equilibria at high pressures. This study has extended over sev- 

31 Spinel is typically magnesium aluminate (MgO-Al 2 3 ), but it has a wide variety of forms 
containing ferrous iron, manganese, ferric iron, and chromium. It may be red, yellow, green, 
black, or some other color. A general formula is R"0 -R/'^Os, where R" may be one of the bivalent 
metals, magnesium, zinc, manganese, iron, nickel, cobalt, or cadmium, and R"' may be trivalent 
aluminum, cobalt, iron, chromium, or gallium. 


eral years and has drawn increasing effort by Laboratory Staff Members. 

Geochemical studies at pressures up to 100,000 atmospheres have per- 
mitted geologists to take a fresh approach to various problems that have 
been the subject of spirited theoretical discussion for decades. Is the 
Mohorovicic discontinuity a phase change from basalt to eclogite? 32 Is it 
the same under the continents as under the oceans? What is the mineralogy 
of the earth's mantle? 33 Can the various types of basaltic lava be related to 
variations in the melting of mantle rocks at different depths and pressures? 
What temperatures are present in the lower mantle and core? 34 As yet none 
of the questions can be fully answered, but the high-pressure studies of the 
last five years have contributed to an understanding of all and promise to 
contribute far more. 

The Mohorovicic discontinuity continues to be one of the more absorbing 
geological problems. High-pressure high-temperature experiment with the 
synthesis of rocks expected at the depths of the discontinuity has given 
some indication of the rocks to be found there. Under the continents they 
are principally basalt and eclogite. Basalt is transformed by high pressure 
to the denser eclogite. Eclogite consists essentially of jadeite-bearing 
pyroxene and pyrope-bearing garnet. 35 Both jadeite [NaAl(Si0 3 ) 2 ] and 
pyrope (Mg 3 Al 2 Si 3 0i2) are high-pressure phases, and their pressure- 
temperature fields of stability have been established in recent years at 
elevated temperatures. Significantly, both the reactions leading to the 
formation of jadeite and pyrope take place in a relatively narrow pressure- 
temperature range. The experimental results now indicate that in the depth 
range 50 to 100 kilometers in the mantle, where basaltic lava is believed to 
form, the mineral assemblage will be characteristic of eclogites. The experi- 
mental data for the transition fit reasonably well the hypothesis that the 
continental Mohorovicic discontinuity is a basalt-eclogite transition. The 
nature of the discontinuity under the oceans apparently is different from 
the continental, and is a challenging question for future thought and 

Additional experimental data for constructing concepts of the earth's 
mantle and crust are being obtained in quantity at the Laboratory from an 
examination of the melting relations of silicates at high pressure. As a result 
the present-day conceptions of reactions by which basalts form in the 
partial fusion of mantle rocks are wholly different from those of earlier 
workers. The system of petrology developed by N. L. Bowen and others 

32 A dense rock equivalent in composition to basalt, found in association with Russian and 
South African diamond pipes, and occurring in rocks, elsewhere on the earth's surface, thought 
to originate from deep in the earth's mantle. 

33 That part of the earth's interior between the Mohorovicic discontinuity and the core. 

34 The core is thought to commence at a depth of about 2900 kilometers. 

35 Garnet has the general formula R"R'"(Si04)3, where R" may be bivalent iron, magnesium, 
manganese, or calcium, and R'" may be trivalent iron, aluminum, or chromium. 


earlier at this Laboratory from experiments at atmospheric pressure 
successfully explained many characteristics of igneous rocks. It now is clear, 
however, that pressures as low as 10,000 to 20,000 atmospheres produce 
very pronounced changes in crystal-liquid equilibria in silicate rock systems. 
Even though the data on phase relations at high pressures still do not 
permit the construction of a system of petrology for the lower crust and 
upper mantle of the earth, answers to some important questions are being 

One of the intriguing questions concerned the formation of silica-saturated 
basalt rocks of the crust from silica-undersaturated mantle rocks. F. R. 
Boyd, Jr., and J. L. England experimented during the year with the melting 
of pyrope garnet, thought to be an important constituent of the mantle, at 
pressures prevailing where basalts are considered to be formed. They found 
that pyrope garnet melts incongruently at these pressures to spinel and 
liquid. This melting relationship could explain the formation of the silica- 
saturated basalts, such as are found in Hawaii, from the typical minerals 
assumed to be in the upper mantle. 

Continuing the experiments reported in Year Book 60, H. S. Yoder, Jr., 
and C. E. Tilley examined the possible origin of alkali basalt and tholeiitic 
basalt, two groups of rocks that are very important components of the 
earth's crust. Their previous experiments with natural rocks and synthetic 
mineral systems established that the same magma (liquid rock), depending 
on pressure, could yield both types of basalt. They now have suggested 
mechanisms whereby both alkali and tholeiitic basalt may be generated 
from an eclogitic liquid deep within the earth's mantle. 

Sydney P. Clark, Jr., J. F. Schairer, and John de Neufville have attacked 
the same problem with a different approach. They also believe that it is 
necessary to examine critically important systems of minerals in their 
entirety under pressure before inferences about melting and solidification 
within the mantle can be drawn with confidence. They chose to examine 
the important but complicated quaternary system 36 that includes among 
its phases the oxides spinel and corundum, forsterite (Mg 2 Si0 4 ), diopside 
(CaMgSi 2 06), pyrope garnet, various forms of silica, and still other minerals. 
Their experiments were conducted at atmospheric pressure and at a pressure 
of 20,000 atmospheres. Their observations showed a range of solid solution 
in pyroxene minerals (e.g., diopside and others) at high pressures that is 
far more extensive than in the same system at atmospheric pressure. 
Pressure therefore undoubtedly produces profound changes in the melting 
relations within at least this mineral system. For some of the compositions, 
the system at 20,000 atmospheres is not even qualitatively similar to the 

36 Quaternary system — a system of phase relations among minerals having four end members, 
schematically expressible in a tetrahedral diagram. 


system at atmospheric pressure, as in the appearance of quartz on the 
liquidus above 1000°C. The experiments showed that this system is well 
suited to the study of the complex chemical equilibria at high pressures. 
Further study should yield important contributions to the petrology of the 
earth's rocks at depth. 

Heat is the source of energy for most geological processes, and knowledge 
of the temperatures at the depth of the core-mantle boundary 37 in the earth 
is of fundamental importance. It is probable that the temperature at this 
depth is not far below the minimum melting temperature of rocks in the 
lower mantle and not far above the solidifying temperature of the iron- 
nickel alloy believed to comprise the outer core. Data obtained at low pres- 
sures showed that the slopes of silicate melting curves were two to five 
times greater than the slopes of the melting curves of most metals. However, 
recent results for diopside and a few other silicates at pressures up to 50,000 
atmospheres yield diagrams with slopes having a pronounced curvature. 
Extrapolation of the diopside data to a pressure at the core-mantle bound- 
ary 38 indicates that a temperature of 3750°C would be required to melt 
diopside there. Similar extrapolation of data on the melting of iron indicates 
a temperature of 5200° at the boundary. Although the uncertainties in these 
extrapolations are very numerous, it is interesting that the estimates are 
close. They are furthermore in rough agreement with estimates made by 
other, equally uncertain, methods. 

An elegant example of the application of the sensitive and powerful 
modern research techniques in geophysics to a problem that scarcely could 
have been touched even a decade ago is shown in the identification of com- 
pounds characteristic of life from ancient rocks. P. H. Abelson and P. L. 
Parker have isolated fatty acids from rocks as old as 500,000,000 years. This 
is the oldest known occurrence of these substances. Among the compounds 
identified were the saturated acids myristic [CH 3 (CH)i 2 C0 2 H], palmitic 
[CH 3 (CH) 14 C0 2 H], and stearic [CH 3 (CH) 16 C0 2 H], the last being the 
most abundant. Although the quantities found are minute (10 micrograms 39 
per gram of organic carbon), gas-liquid chromatography permits isolation, 
identification, and quantitative measurement of the individual acids, even 
when major amounts of impurities are present. 

The same fatty acids were isolated from recent sediments. Palmitic acid 
was the major component in the young rocks, being as much as ten times as 
abundant as stearic acid, the more abundant in the old rocks. Thus although 
the fatty acids in very young and in old rocks are qualitatively similar a 
puzzling quantitative difference has been noted. 

37 Postulated to be at a depth of about 2900 kilometers. 

38 1,400,000 atmospheres. 

39 A microgram is 0.000001 gram. 


T. C. Hoering has investigated two important aspects of the geochemical 
record of very early life on earth. He has studied some of the earth's oldest 
sedimentary rocks, which contain structures geologists consider related to 
algal activity. He has measured stable carbon isotope ratios, C 13 /C 12 , in 
coexisting carbonates and reduced carbon obtained from these specimens. 
He has found that the isotopes have been fractionated into a C 13 -enriched 
carbonate phase and a C 13 -depleted reduced carbon phase. The amount of 
this fractionation is nearly identical to that found in contemporaneous algal 
cells and their associated carbonates. The magnitude of the effect is also 
similar to that found between limestones and coals of all geological ages. 
Such isotope fractionation is caused by a slightly different rate of photo- 
synthesis for molecules of carbon dioxide containing C 12 as compared with 
those containing C 13 . 

The samples examined include a limestone from the Belt Series of Glacier 
Park, Montana, with a minimum age of 1.2 billion years, the Randville 
dolomite of Crystal Falls, Michigan, with a minimum age of 1.5 billion 
years, and the Bulawayan limestone of Southern Rhodesia with a minimum 
age of 2.7 billion years. In these rocks, which are among the oldest known 
sedimentary rocks, the isotopic evidence is consistent with the presence of 
photosynthetic algae in the very early Precambrian era. 

The second study by Hoering was on the reduced carbon of Precambrian 
sedimentary rocks. A successful effort was made to extract and partially 
identify organic molecules from them. By means of a number of chemical 
degradations he was able to liberate soluble fractions from the insoluble 
"fabric" of the reduced carbon. The fractions were analyzed with the aid of 
ultraviolet spectroscopy and chromatography. The results indicate that the 
insoluble reduced carbon may be related to the kerogen of more recent 
rocks. Kerogen is produced by interactions of organic products of cells when 
deposited in sediments deprived of oxygen. Thus additional evidence has 
been produced pointing to the existence of life in very early Precambrian 
times, more than two billion years ago. 

Mount Wilson and Palomar Observatories 

In 1905 the astronomical activities of the Institution were 

mainly those of the Solar Observatory, whose building and 

1905 equipment were under construction during the year, with 

view to completion in 1906. Mount Wilson was considered an 

especially favorable site because "The unusually favorable 

atmospheric conditions which prevail day and night at the site of the 

observatory have attracted the attention of astronomers and astrophysi- 


cists generally. " 40 "It has been been found that the average night-seeing is 
exceedingly good, while the low wind-velocity, coupled with the trans- 
parency of the atmosphere, afford. . . advantages which should render Mount 
Wilson an ideal site for the 5-foot reflector." George E. Hale, the Director, 
defined his purposes as: "(1) The investigation of the sun (a) as a typical 
star, in connection with the study of stellar evolution: (b) as the central 
body of the solar system, with special reference to possible changes in the 
intensity of its heat radiation, such as might influence the conditions of life 
upon the earth. (2) The choice of an effective mode of attack, involving (a) 
the application of new methods in solar research; (6) the investigation of 
stellar and nebular phenomena, especially such as are not within the reach of 
existing instruments ; and (c) the interpretation of these celestial phenomena 
by means of laboratory experiments." He was at this time already con- 
sidering the design of "a large reflecting telescope and of new types of 
instruments." He also looked forward to "The furtherance of international 
cooperation in astrophysical research through the invitation to Mount 
Wilson, from time to time, of investigators especially qualified to take 
advantage of the opportunities afforded. . . ." 

A large part of Dr. Hale's report in 1905 was necessarily devoted to a 
statement on the numerous construction projects that had absorbed his 
attention during the year. They included ten buildings on Mount Wilson, 
and the Pasadena office and shop, which were constructed on land given by 
citizens of Pasadena. Dr. Hale nonetheless found time not only for instru- 
ment testing but also for an observing program and planning a future 
research program. Daily direct photographs of the sun on a scale of 6.7 
inches to the solar diameter were taken on the Snow telescope. Observations 
were made to test an hypothesis of Dr. Hale's about the relation of calcium 
vapor to the faculae and plages 41 of the sun. Some experimental study of 
the spectra of sunspots, plages, and the chromosphere was undertaken for 
instrument design. Photographs were also taken of bright stars with a long- 
focus grating spectrograph. They included a photograph of the blue region 
of the first-order spectrum of Arcturus, which required an exposure of 14 
hours on three successive nights. 

Visiting investigators had already found their way to Mount Wilson. E. 
E. Barnard of the Yerkes Observatory photographed the southern part of 
the Milky Way, described by Dr. Hale as "a most important contribution 
to our knowledge of the structure of the Milky Way and of the remarkable 
nebulae within it." The Smithsonian Institution also sent an expedition 

40 Year Book 4, p. 25. 

41 Faculae — small irregular bright patches in the photosphere (visible disk) of the sun, sur- 
rounding sunspots. 

Plages — faculae of the chromosphere, which is the outer layer of the sun's "atmosphere," 
extending to a height of several thousand kilometers from the visible disk. 


to the mountain for observing solar radiation, directed by C. G. Abbot. 
Other astronomical work supported by the Institution in 1905 included 
the compilation, by Lewis Boss of the Dudley Observatory, Albany, New 
York, of a Preliminary General Catalogue of Stars for the 6000 stars visible 
to the naked eye. Also included were grants to Simon Newcomb of Washing- 
ton, D. C., for an " Investigation of the mean motion of the moon," and a 
rather enigmatical grant "To aid investigations in mathematical astronomy, 
statistical methods, and economic science." The economic science part was 
never reported upon, either in 1905 or in the four succeeding years when 
Dr. Newcomb held sequel grants. 

The program of 1961-1962 at the Observatory, as for each of 

the four preceding Departments, differed vastly from its 

1961-1962 ancestor of sixty years ago. The primary emphasis on solar 

studies gave way in 1918 to a more general astronomy 

program with the completion of the 100-inch telescope. 

Nonetheless, solar observation and solar study have continued to the present 

day, but with gradually decreasing emphasis. A major change came in 1958- 

1959, with a decision to drastically curtail routine observations. Since then 

solar studies have centered on the sun's magnetic fields, of which daily 

observations are made with the aid of the solar magnetograph originated 

and developed by H. D. and H. W. Babcock of the Observatories. Daily 

solar magnetograms have been made since 1957. 

During the 1961-1962 year R. F. Howard commenced an extensive study 
of the accumulated magnetograms to classify magnetic regions, and cor- 
related them with optical and radio phenomena. He has already obtained 
the very interesting finding that the unipolar magnetic (UM) regions of the 
sun correlate in their position with calcium absorption phenomena observed 
spectroscopically. It may thus now become possible to extend observation 
of UM regions backward for 50 years or more, using the Observatory's 
extensive collection of spectroheliograms showing the absorption lines of the 

Responding somewhat to the explosion of national interest in inter- 
planetary space, the year was also marked at the Observatories by renewed 
attention to the planets, which have been subject to recurring study at the 
Observatories in the past. It has seemed important to press ground-based 
observations like those that can be undertaken at the Observatories to the 
limits made possible with new photometric and infrared techniques, be- 
cause information about the planets can be acquired by these techniques at a 
cost of much less effort and money than by observations from rockets. G. 
Munch, with the collaboration of H. Spinrad of the Jet Propulsion Labora- 
tory of the California Institute of Technology, and R. Younkin of the same 
laboratory, began studies of the spectra of the major planets. Two 


lines of the hydrogen molecule were found in the spectrum of Saturn, pro- 
viding the first firm evidence of the presence of hydrogen in the atmosphere 
of that planet. Spinrad also analyzed high-dispersion spectra of Venus, 
finding evidence of large changes in the temperature of the atmosphere of 
Venus. B. Murray of the California Institute of Technology continued 
studies of the photoelectric colorimetry of the moon with the Mount 
Wilson facilities. 

A major part of the Observatories' program, however, has been devoted to 
a study of the masses, luminosities, surface temperatures, and chemical 
composition of stars, and the variation of luminosity and surface tempera- 
ture with age. During recent decades these have been among the major 
problems in astronomy. Even though such research has become increasingly 
important with time, steps toward the modern understanding of these 
phenomena date back to the early years of this century. The first important 
step was taken by E. Hertzsprung and H. N. Russell, when they plotted a 
diagram of the absolute magnitude of stars in the solar neighborhood against 
their surface temperatures as indicated by spectral class or color. They 
found that most stars fall in a narrow diagonal band on their diagram, the 
very hot giants being at one end and the cool dwarfs at the other. Later 
theoretical investigations based on nuclear physics showed that the fusion of 
hydrogen into helium was an important source of energy for most stars 42 
and that stars obtaining their energy from this reaction logically fall on the 
color-magnitude diagram in the narrow "main sequence" band noted by 
Hertzsprung and Russell. 

During the second world war, Walter Baade of the Observatories made a 
detailed investigation of the stellar content of the Andromeda galaxy. He 
found that the brightest stars in its spiral arms are very hot giants similar to 
those in the solar neighborhood, which also is on a spiral arm. In the nucleus, 
however, the brightest stars were cool red giants. To differentiate these, 
Baade introduced the concept of Population I (younger) stars typically on 
the spiral arms and Population II (older) stars typically at galactic centers. 

Theory then predicted that, as the hydrogen fuel approaches exhaustion 
in a stellar core, a star expands greatly but cools and thereby moves off the 
narrow main-sequence band in the color-magnitude diagram and becomes a 
red giant. Since the brightest stars use their fuel most rapidly this change 
starts at the upper end of the main sequence and moves down the sequence 
with time. Obviously, the hot giants in the solar neighborhood and in 
galactic spiral arms indicate a population of stars that have formed re- 
cently, whereas the red giants in a galactic nucleus represent a population 
of old stars. Theory permits one to go even further and fix the age of a group 

42 The hydrogen-helium reaction is now considered ancillary to the hydrogen-deuterium-helium 


of stars by observing the magnitude at which stars are just beginning to 
move off the main sequence. Color-magnitude diagrams have been con- 
structed for a large number of globular and galactic clusters 43 by A. R. 
Sandage, H. C. Arp, and W. A. Baum of the Observatories, and many 
others. Ages of a few million up to ten billion or more years have been found. 

With the aid of high-dispersion spectra it has become possible recently to 
make detailed quantitative chemical analyses of stellar atmospheres. The 
first studies of the sun and of bright nearby stars indicated a surprising uni- 
formity of chemical composition. When these measurements were extended 
to some of the distant older clusters, however, it was found that their stars 
were deficient in the heavier metallic elements, often by factors of 100 or 
more compared with the stars near the sun. Since most of the strong 
metallic lines fall in the ultraviolet ( U) region of the spectrum, a star of high 
metallic content exhibits a depressed U region compared with the blue (B) 
or green-yellow (V, " visible' ') spectral regions. Within the past two years 
astronomers at the Observatories have found it possible to fix the metallic 
content from a comparison of the magnitudes of a star measured in the 
U, B, and V regions. This makes feasible the extension of abundance 
determinations to stars far too faint for detailed spectrum analysis. 

In general, old stars such as those in the globular clusters, or high- velocity 
stars, 44 which presumably were formed at the same time as those in the 
galactic nucleus, are metal deficient compared with the younger stars. 
Theory suggests that metals are formed late in the evolution of a star, after 
the hydrogen fuel has been exhausted in the stellar core and the central 
temperature has increased to many times that of stars on the main sequence. 
Therefore the metal-containing material in recently formed stars has gone 
through one or more earlier generations of stars in which the metals are 
formed and then blown off into space either in a gradual flow 45 or explosively 
in a nova or supernova outburst. 

Obviously, a project to understand stars will require decades for comple- 
tion as well as investigation by many astronomers at a number of observa- 
tories. The Institution can take pride not only in the participation of the 
Observatories in the grand conception of such a project but also in their 
preeminent position as contributor of observational data leading to widen- 
ing views of the universe which these studies are providing. During the year 
1961-1962 the staff of the Observatories skillfully exploited the wonderful 
instruments at their disposal to give us further insights on this frontier of 

43 A globular cluster is a group of many thousands of stars arranged in a regular form showing 
spherical symmetry. Many are located outside the plane of the Milky Way. A smaller group of 
stars always found near the plane of the Milky Way is known as a galactic cluster. 

44 A high-velocity star is a star that is moving about the galactic nucleus with a velocity 
markedly different from that of the sun. 

45 As studied by A. Deutsch of the Observatories. See Year Book 59, p. 8, and other year books. 



logT 4.5 






O o ° 

G ft NTS 

















Generalized Hertzsprung-Russell diagram of star color-magnitude relation. Log T — logarithm of 
temperature, degrees Kelvin; Mboi = bolometric magnitude, as measured from calculated total 
energy emission. (Adapted from Cecilia Payne-Gaposchkin, Introduction to Astronomy, Prentice- 
Hall, New York, 1954.) 


the universe. Among the results were new knowledge about the differences 
in chemical composition among stars, the correlation of chemical composi- 
tion and star movement, a determination of the time of formation of the 
Galaxy in which we are located, and new evidence on the expansion of the 

The staff of the Observatories participated in detailed chemical investiga- 
tions of a number of stars. J. Greenstein and R. A. Parker of the Observa- 
tories have collaborated with G. Wallerstein of the University of California, 
H. L. Heifer of the University of Rochester, and L. Aller of the University of 
Michigan to study one group of three red giant stars. They found that the 
common metals were only 1/500 as abundant in this group as in the sun, and 
the heavy elements strontium, zirconium, barium, cerium, and europium 
were deficient by a factor of 25,000. Considering the deficiencies, they 
estimate that these stars, which are part of our Galaxy, probably condensed 
within a few hundred million years after the formation of the Galaxy. In 
investigating several dozen peculiar B and A stars, 46 J. Jugaku, W. L. W. 
Sargent, and L. T. Searle found that the abundances of individual elements 
vary erratically compared with neighboring elements in the periodic table, 
often fluctuating by factors of 100 or more. Elements found to have marked 
over- or underabundance in certain stars are beryllium, carbon, nitrogen, 
oxygen, silicon, phosphorus, and mercury. 

For some years a group of stars have been recognized and designated as 
subdwarfs because they lie appreciably below the main sequence on the 
color-magnitude diagram. Early studies showed that they were metal- 
deficient, therefore old, stars. They have high velocities considered in 
relation to the sun. To learn more about these rather rare stars, A. Sandage 
and C. T. Kowal have started a program for the photoelectric observation of 
the ultraviolet-blue-visible magnitudes of the high-velocity stars given in 
the Giclas Proper Motion Catalogue. More than 100 new metal-deficient 
subdwarfs have been discovered among the 700 stars observed thus far. 
Spectroscopic studies by Greenstein and by Sandage of an enlarged sample 
of these subdwarfs confirmed the high velocity of all. 

In a further effort to obtain information on the relation of the subdwarfs 
to other principal groups of stars, O. J. Eggen and Sandage studied the effect 
of "line blanketing" 47 on the position of a star in the color-magnitude dia- 
gram. The results were of special astronomical interest, for Eggen and 
Sandage found that if proper correction is made for line blanketing the sub- 
dwarf stars move into the same position as normal dwarf stars on the main 

46 The accepted spectral classification of stars designates them by arbitrary letters as O, B, A, 
F, G, K, and M. O and B stars have the highest temperatures, and M the lowest. 

47 Line blanketing refers to the situation in which the abundance in a star of the metallic 
elements having strong absorption bands in the ultraviolet is so great that it causes an appreciable 
deficiency in the spectral region compared with other parts of the spectrum of the star. 


sequence of the color-magnitude diagram. Thus another addition was made 
to our understanding of the wonderful order which astronomers have been 
slowly illuminating with the aid of modern instruments. 

Eggen, D. Lynden-Bell, and Sandage also studied the orbits around the 
nucleus of our Galaxy of a large number of dwarf stars, including both the 
normal and subdwarf types. They find a close correlation between metal 
deficiency and the eccentricity and angular momentum of the stellar orbit. 
They interpret this as indicating that metal-deficient stars were formed in 
an early period while our Galaxy was rapidly contracting. From the age of 
these stars they were able to fix the time of formation of our Galaxy out of 
the medium of the universe at about ten billion years ago. 

Studies of stellar properties are important not only for understanding the 
characteristics of the stars themselves but also to provide a firm basis for 
the measurements on which the conceptions of the structure and origin of 
the entire universe depend. For example, nearly all determinations of large 
astronomical distances depend on the comparison of the apparent brightness 
of a nearby object with that of an identical object in a distant cluster or 
galaxy. Thus cepheid variables 48 were used by E. P. Hubble to fix the dis- 
tance of the nearby Andromeda galaxy, and the galaxies themselves were 
used to estimate the distances of clusters of galaxies at the extreme range 
of telescopic penetration into space. However, the discovery of different 
stellar populations with major differences in age and chemical composition 
raised many doubts about the identity in absolute magnitude of a star in 
our own neighborhood with that of a star in a nearby galaxy, which might 
or might not have similar age or chemical composition. 

One of the uncertainties in these extrapolations toward a picture of the 
universe has been the effect of light absorption by dust clouds along the 
path between star and observer. This is especially troublesome, since the 
shorter wavelengths of the spectrum are absorbed more strongly than the 
longer, producing a reddening effect. During the year H. C. Arp reexamined 
this problem in color-magnitude studies of globular clusters of stars. He 
found that the correction for absorption should be appreciably larger than 
had been allowed formerly. A substantial revision downward of previously 
determined globular cluster ages therefore must be made. This also elimi- 
nates a discrepancy existing between age determined from position on the 
color-magnitude diagram and age determined from models of cosmological 
expansion. They now become consistent. 

Extrapolations to distant galaxies are also handicapped because most 
are too distant to permit observation of enough stars for the construction of 
a color-magnitude diagram. However, it is possible to analyze the integrated 

48 Stars whose light emission varies in a definite pattern over a relatively short period but 
longer than 24 hours. 


light received from a galaxy and from it obtain information about the 
distribution in temperature, magnitude, and chemical composition of the 
component stars. W. A. Baum has studied some nearby galaxies of different 
types by photoelectric scanning methods in order to obtain information 
about the evolution of galaxies. His evidence indicates that some definitely 
are composed of true Population II (older) stars whereas others (large 
ellipticals) have mainly Population I stars. Such observations are important 
in the construction and interpretation of cosmological models. Present 
interpretation of the observable universe conceives it as having a radius of 
five billion or so light years, 49 expanding at its limits of observation at 
nearly half the speed of light. This interpretation depends on assumptions 
made about the magnitude-redshift relation, that is the reddening of the 
observed spectrum caused by recession of the distant galaxies in relation to 
the solar system. Distant galaxies, of course, are seen at an earlier age 
than nearby ones — billions of years of difference for the most distant. Since 
individual stars undergo large changes in luminosity and temperature with 
age, the observable integrated light of a galaxy also changes with time. 
How, for example, has the extremely distant galaxy 3C295 (now redesig- 
nated 1410+5224), mentioned in Year Book 59, changed in the five billion 
years since the observed light that fell on the Palomar photographic plate 
left the galaxy? Answers to questions like these will be obtained only from 
studies such as those undertaken by Baum and other Staff Members of the 
Observatories on stellar properties and evolution. 

The Department of Embryology 

Although the Department of Embryology was not estab- 
lished until 1914, when it was organized by Franklin P. Mall, 
1905 even its subject was not ignored among the activities of the 
Institution in 1905. In that year L. B. Mendel of Yale 
University was given a grant for " Study of physiology of 
growth, especially in its chemical processes/ ' Professor Mendel's grant was 
renewed in each of several years thereafter. He reported that he was 
studying the " chemical composition of the developing animal body and 
the equipment of this organism for its nutrition, upon which growth 
essentially depends. Data are being collected at first hand regarding the 
composition of various embryonic tissues at different stages of embryonic 
growth. For the nervous system a correlation between morphological and 

49 Light year — the distance traveled by light in a vacuum during one year; about 5.88 X 10 12 


chemical development is already apparent. The chemistry of embryonic 
muscle is also already under investigation. "The purin content of the liver 
and muscles at various embryonic stages has been determined. ... It is 
hoped ... to ascertain whether the purin metabolism of the young is 
essentially different from that of the adult." 50 A grant to L. E. Griffin was 
also made in the same year "to secure material for a study of the embry- 
ology, histology, and physiology of the Nautilus." Studies supported at the 
Marine Biological Laboratory, Woods Hole, Massachusetts, included one 
on the "segmentation of certain fertilized eggs"; on "regenerative processes 
and structures"; and on "muscle-fibers of the fish heart." 

These studies, however, were not in any sense an organized group. Nor 
did they command a major interest on the part of the Institution's admin- 
istration, as was shown by the termination of this type of grant at the end 
of 1908. It remained for Dr. Mall to set in 1914 the lines on which the 
Department continued so long, an examination of the morphology and 
histology of the human embryo and the embryonic physiology of primates. 
Even at the beginning of the Department, however, other organisms were 
studied, as illustrated by the 1914 study of E. L. and E. R. Clark on the 
movements of the lymph heart in living chick embryos, and their report in 
that year "that the muscle of the lymph heart is derived from the 
myotomes." 51 

The year 1961-1962 was marked by the setting of an im- 
portant milestone in the history of the Department. After 
1961 1962 several years of preparation the new Department of Embry- 
ology building, adjacent to the Homewood Campus of the 
Johns Hopkins University in Baltimore, Maryland, was 
completed. This building, specially designed for embryological research, 
should free the staff of the Department from the inconveniences that 
attended work in their former cramped quarters at the Johns Hopkins 
Medical School near the center of the city. The Department started to 
move on August 1, 1961, and was able to assume full operation by early 
November, in spite of a long delay in equipping the building because of an 
electricians' strike. The new building appears to have met with staff ap- 
proval. Director J. D. Ebert describes it as having "an unusual combination 
of fine qualities, pleasing to both aesthetic and practical senses." 

The Department in 1961-1962 is described by Dr. Ebert in the introduc- 
tion of his report of this year as one holding to its "traditional organization 
of a group of independent investigators whose interests range widely from 
biochemistry and microbiology to anatomy and physiology, with sub- 
stantial overlapping in experience and approach. ... in developmental 

60 Year Book 4, pp. 259-260. 

51 Year Book 13, p. 112. A myotome is a muscle mass in a developing animal. 


biology today it appears to favor the generation and interchange of 
ideas. . . ." 

The multifaceted program Dr. Ebert describes included an interesting 
study of the physiologic aspect of frog-embryo growth from the stage of the 
fertilized egg onward by D. D. Brown and J. D. Caston, the nature of the 
testicular antigen in induced aspermatogenesis 52 by G. L. Carlson and D. 
W. Bishop, the role of deoxyribonuclease II during the metamorphosis of 
the tadpole by J. R. Coleman, a comprehensive study of the developing 
human eye by R. O'Rahilly, and still others. 

Of particular interest among these was the Brown-Caston study of the 
embryonic development of the frog Rana pipiens. They found that the 
early embryos contain a measurable but small population of ribosomes in 
their cells. The early ribosomal content changes little until a stage near the 
end of morphogenesis, 53 when there is a very rapid appearance of more 
particles. This coincides with the time when the embryo has been shown to 
require magnesium ions from outside. In addition, the iron storage molecule, 
ferritin, was definitely identified in the egg. Also, although ribosomal 
synthesis was shown to begin after much of morphogenesis is completed, 
high-molecular-weight RNA, with a base composition identical to ribosomal- 
ion RNA, was found to be present in all stages of the embryo. 

Perhaps the most striking progress to be reported from the Department 
during the year came in the studies of I. R. Konigsberg, who joined the 
Institution as a Staff Member on July 1, 1961. They will be described in 
some detail as an illustration of the methods and approach of the present- 
day Department. 

From its beginning the Department of Embryology has numbered, among 
its Staff Members, investigators dedicated to the study of the development, 
structure, chemistry, and physiology of muscles. W. H. and M. R. Lewis 
(Department of Embryology, 1915-1940) pioneered in analyzing the origin 
of muscle fibrils in tissue culture; and Arpad Csapo (1949-1954) was among 
the first students of muscle chemistry to characterize the contractile pro- 
teins of the uterus and to examine their regulation under different physio- 
logical conditions. More recently J. D. Ebert and R. L. DeHaan and their 
associates have focused attention on the biochemistry of developing con- 
tractile proteins and on morphogenetic movements and relations of 
contractile and conductile cells in the heart. D. W. Bishop has contributed 
importantly to our understanding of mechanisms in primitive contractile 
systems like sperm tails. 

To this roster the Department now adds Konigsberg's name. During the 
year he made substantial progress in a hitherto refractory subject, the 

62 Destruction of the power to produce sperm. 

63 The emergence of the specific structure of an animal during embryonic development. 


investigation of the cytodifferentiation of embryonic skeletal muscle cells in 
dispersed cell culture. His system of culture is designed to offer greater 
opportunity for rigorous control of both the quantitative aspects of the 
cellular population and the extracellular environment than can be achieved 
either in vivo or in organ culture. Many years' experience by numerous 
previous investigators suggested that such culturing techniques could be 
expected to promote the loss of differentiative character and would not 
favor a progressive increase in the effects of cell specialization on mor- 
phology. No generally satisfactory explanation for this previously observed 
incompatibility has ever been given. However, Konigsberg's results with 
monolayer cultures of embryonic skeletal muscle cells are in striking dis- 
agreement with expectations from earlier experience. 

Monolayer cultures prepared from suspensions of 11- to 12-day chick 
embryonic leg muscle pass through three recognizable phases. The period 
immediately following plating of the cells is marked by rapid proliferation 
with a mean generation time of 24 hours. During this period cultures consist 
exclusively of mononucleated cells and have the general appearance of 
cultures of "fibroblast-like" 54 cells such as might be derived from a great 
variety of tissues. The transition from the first to the second phase occurs in 
a matter of hours and is characterized by the formation of long multinuclear 
"ribbonlike" cells. Formation of these multinuclear cells coincides with the 
attainment of cell confluence in the culture. The effect of cell density is 
further suggested by experiments in which the inoculum size was varied. 
The smaller the inoculum, the greater the time of transition from phase one 
("fibroblast-like" cell) to phase two (multinuclear "ribbon"), and vice 
versa. The abrupt appearance of multinucleated myotubes 55 in this second 
phase is paralleled by an equally abrupt break downward in the rate of 
proliferation. Again, the time required for this development can be shifted 
by varying the inoculum size. 

Differentiation beyond the stage of the mononucleated myoblast 56 occurs 
in culture after cells have ceased rapid multiplication. This observation is 
consistent with Konigsberg's earlier findings, as well as with those from 
several laboratories, that myotube nuclei are postmitotic 57 and that they 
form by cellular fusion. The third phase of muscle differentiation in culture 
is characterized by the progressive development of the cross-striated 
myofibrillar pattern and the initiation of the spontaneous contraction char- 
acteristic of muscular tissue. 

All Konigsberg's studies before the past year had been restricted to mono- 

64 Fibroblasts — elongated mononuclear cells which develop into and are also part of connective 

65 Aggregated-cell constituent of muscle. 

66 Unassociated single "premuscle" cell. 

67 Mitosis is cell division. 


layer cultures established with inocula of 1 million to 2.5 million cells each. 
Such cultures reach confluence between the second and fourth day of 
culture, depending on the size of the inoculum. To probe for the lower limit 
of inoculum size that would still permit differentiation to occur he turned to 
the single-cell plating technique developed by T. T. Puck and his associates. 
In this procedure small numbers of cells are dispersed over a relatively large 
area. During appropriate periods of incubation the individual cells give rise 
to discrete colonies visible to the naked eye. The technique has been applied 
most successfully to permanently established cell strains. 

Using freshly isolated embryonic muscle cells Konigsberg observed a 
plating efficiency of approximately 10 per cent. In plates cultivated for 10 
to 13 days approximately 1 in 10 colonies exhibited unmistakable signs of 
skeletal muscle cell differentiation. The proportion of differentiated cells 
ranged from colonies containing several elongated myotubes in colonies of 
predominantly mononucleated cells to colonies in which virtually every 
nucleus was in syncytial 58 association. Under polarized light or bright-field 
illumination after staining, the myotubes showed the presence of longitu- 
dinal fibrils, which frequently exhibited the pattern of cross striation typical 
of mature skeletal muscle cells. It is apparent that some myoblasts, at least, 
through a sequence of rapid multiplications, can produce a large number of 
progeny that retain the capacity for differentiation. 

Two major questions emerged from these observations. First, what is the 
significance of the finding that only 1 in 10 colonies eventually differentiates? 
Second, what is the stimulus initiating myotube formation? Konigsberg is 
attacking the second problem by examining the relationship of cell density 
to myotube formation. Two general mechanisms by which cell density 
might affect myotube formation were considered. Since myotube formation 
is a result of cell fusion, high cell density might ensure that a sufficient 
number of effective cell-to-cell collisions occur. Another, and equally likely, 
possibility is that a high cell density may be either supplementing the 
culture medium with cell products or removing some components. 

Konigsberg designed experiments to test that possibility. His first tests 
showed that the medium is altered by the metabolic activity of cells 
cultured in it. In cultures grown on a medium preconditioned by the pres- 
ence of other cells, myotube formation commenced as much as 24 hours 
earlier than initial cultures of equal numbers of cells from the same cell 
suspension but cultured in fresh medium. Furthermore, the cells in condi- 
tioned medium attached to the glass more firmly, presenting a strikingly 
different appearance from the control cultures. 

These results are impressive in themselves, and indeed they represent 
something of a technical breakthrough in the difficult task of cell culture. 

58 Referring to a multinucleated aggregate of imperfectly separated cells, or a multinuclear cell. 


But as so often in science they are probably more important for the questions 
they raise than for the results they give. Already they have pointed the way 
to a number of additional experiments to probe the relation between condi- 
tioned media and cell differentiation. But in the hint given of a hitherto 
unsuspected closeness of relation between cell and environment we touch a 
problem of wide application and perhaps vast significance in understanding 
all higher forms of life. 

Although a report describing work like Konigsberg's can give something 
of the sense of high adventure experienced by scientists within the Institu- 
tion and elsewhere, there are dimensions to the scientific life of today that 
must always escape any progress report. Most of those who work within the 
Institution share a deep conviction about the humanity of their calling and 
about the community of fellowship that not only is vital to the progress of 
their work but also is a deeply felt reward in itself. Happily these convic- 
tions occasionally shine through more esoteric daily concerns. They are 
notable this year in the comments of James Ebert and Merle Tuve, each on a 
point of his philosophy. 

Ebert has written particularly of his own deep attachment to the essential 
unity of living science. He quotes Frank R. LilhVs memorable words that 
" Scientific discovery is a truly epigenetic process in which the germs of 
thought develop in the total environment of knowledge." The life of the 
laboratory, where one must be quick to acknowledge what has gone before, 
alert to the current actions of others having similar interests, and mindful of 
the needs for others to know, can be a social experience almost beyond 
comparison. Dr. Ebert notes his pleasure at having visiting investigators 
from other institutions: "They do contribute vitally to the Department . . . 
but of far greater moment is the question whether such a visit adds measur- 
ably to the man's ability as an investigator and teacher when he returns to 
his home laboratory. Has he found new direction or meaning for his re- 
search? Has the opportunity for reflection. . . led to a searching reexamina- 
tion of his program?" With pride, the Institution can record that its 
Departments provided literally hundreds of such opportunities during the 

Tuve's comments touch upon the aesthetic experience of being a scientist, 
and upon what is perhaps one of the deepest motivations in "exact" science. 
Contrasting it with the disorder and transience he sees in the life of men in 
the mass, he expresses his admiration at "the beautiful regularity and 
systematic relatedness. . . in every aspect of the natural phenomena. . . from 
distant stars to living bacteria." He considers this a cause of the sense of 
very deep satisfaction in scientific studies. Through science, man, bit by bit, 


is adding to his stature and to "his awe of the stupendous and beautifully 
intricate universe in which he finds himself." Tuve considers it a "great good 
fortune" for scientists to be able to devote their energy and talents to 
illuminating "the intricate and orderly patterns of the physical world around 
us." To him this is "a princely gift of our time and circumstances." 

Such motivations lead to a dedication which is the wonder of all who have 
not experienced their attractions. It is a dedication measured only in part by 
a voluntary 70-hour week, by long nights on a mountaintop in below-freezing 
weather, by a hundred frustrations with equipment design, or by a willing- 
ness to work at the modest salaries that fundamental science is able to 
provide. We can hope that Andrew Carnegie, after sixty years, might be 
approving of both the dedication and the insights of these men and their 
predecessors as they have striven "to secure, if possible, for the United 
States of America leadership in the domain of discovery. . . of new forces." 59 

Losses . . . 

I must report with great sorrow the loss of a devoted member of the 
Board of Trustees, the Honorable Robert Woods Bliss, and of a highly 
valued Staff Member of the Mount Wilson and Palomar Observatories, 
Don 0. Hendrix. 

Robert Woods Bliss, a Trustee of the Institution for twenty-six years, 
died in Washington, D. C, on April 19, 1962. Elected a Trustee in 1936, he 
became a member of the Executive Committee the following year. He was 
Secretary of the Board of Trustees from 1953 until his death. He also served 
continuously from 1939 on the Committee on Archaeology, from 1939 to 
1945 on the Auditing Committee, and from 1950 to 1953 and 1958 to 1961 
on the Nominating Committee. 

Before his association with the Institution he had already had a distin- 
guished career of 33 years in the diplomatic corps of the United States, 
where he held many important posts. He was especially concerned with 
efforts to bring about world security through arms control and international 
organization. In 1908 he was United States delegate to the International 
Conference to Consider Measures for the Revision of Arms and Ammunition 
Regulations in Brussels. As counselor to our embassy in Paris from 1916 to 
1919 he assisted in preparations for the Versailles Peace Conference and in 
its work. Again, in 1921, he was a member of the United States delegation 
to the Washington Conference on the Limitation of Armaments. His 
beautiful estate, Dumbarton Oaks, in Washington, was the scene of the 
conference that laid plans for the United Nations. 

69 Andrew Carnegie, Trust Deed Creating a Trust for the benefit of the Carnegie Institution 
of Washington, D. C, January 28, 1902. 


Just before his retirement in 1933 he had served for six years as am- 
bassador to Argentina; and during World War II he was called back from 
his technical retirement to serve as special consultant and then special 
assistant to the Secretary of State. 

Mr. Bliss will be remembered by the Washington community for his 
many philanthropic and cultural contributions. The Institution will 
remember his dedication to its welfare and his gentle but always penetrating 
counsel on every problem. 

Another loss that is especially felt is that of Don O. Hendrix of the Mount 
Wilson and Palomar Observatories, who died on December 26, 1961, at the 
age of 57. Joining the staff of the Observatories in 1913, Hendrix became 
Superintendent of its optical shop in 1947, where he carried out such im- 
portant projects as the optical design for the 48-inch schmidt telescope and 
the final figuring of the 200-inch mirror after it had been moved to Palomar. 
His extraordinary skill was largely responsible for the high efficiency of the 
present equipment of the Observatories. 

With keen regret I also record the loss to the Institution of four retiring 
members of the staff. Dr. Berwind P. Kaufmann, Director of the Depart- 
ment of Genetics, Dr. Robert K. Burns, Staff Member of the Department of 
Embryology, Mrs. Ruth L. McCollum, Assistant to the President, and 
Wilbur A. Pestell, Administrative Assistant at the Department of Plant 
Biology, all retired on June 30, 1962. 

Dr. Kaufmann came to the Department of Genetics in 1937 from the 
University of Alabama, where he had served for ten years as professor and 
department head. Since that time his professional interests have touched on 
many facets of the broad field of cytogenetics, with emphasis on the varying 
patterns of chromosome structure that influence gene action. These interests 
stemmed from experience in the area of descriptive cytology, gained in the 
early 1920's, when chromosomes were generally regarded as uniformly 
staining rod-shaped structures with no discernibly precise pattern of internal 
organization. By developing and applying ingenious techniques, Dr. 
Kaufmann demonstrated that chromosomes contain paired, helically dis- 
posed strands at all phases of somatic and meiotic mitoses. 

Upon joining the Institution's staff, Dr. Kaufmann undertook an analysis 
of the types and frequencies of chromosomal rearrangements induced by 
ionizing radiations, using the giant chromosomes in the salivary glands of 
Drosophila for diagnostic purposes. His discovery and evaluation of the 
effects of near-infrared radiation on the frequencies of X-ray-induced re- 
arrangements was an outstanding accomplishment of that period. 

In 1960 Dr. Kaufmann succeeded Dr. M. Demerec, first as Acting 
Director and in 1961 as Director of the Department. During his twenty-five 


years at Cold Spring Harbor he maintained a strong interest in science 
education and in the training of young biologists. He has now returned to a 
university environment, having been appointed Professor of Zoology and 
Senior Research Scientist at the University of Michigan, where his sincerity, 
dedication, and technical skill will be inspiring to those who have the good 
fortune to work with him. 

Dr. Burns joined the Department of Embryology in 1940 from the 
University of Rochester, where he had been a member of the Department of 
Anatomy, of which Dr. George W. Corner was the head before his own move 
to the Institution. When he went to Baltimore, Dr. Burns rejoined Dr. 
Corner and another long-time Rochester colleague, B. H. Willier, who had 
assumed the direction of the Johns Hopkins Department of Biology. Burns, 
who held the title of Honorary Professor of Biology at the University, 
served as an important link between the two departments, pointing the 
way to the close association that exists today. 

Dr. Burns has devoted his entire career to studying the mechanisms of sex 
differentiation. A student of Ross G. Harrison, he began by demonstrating 
sex reversal in amphibians, using the technique of embryonic parabiosis. His 
was the first convincing laboratory research following up Frank R. Lillie's 
analysis of the freemartin. 60 Later he turned his attention to mammals, and 
again produced the first convincing evidence of sex reversal by the use of 
purified sex hormones in his analysis of the effects of estradiol on the pro- 
spective male opossum. 

Dr. Burns has returned to Bridgewater College, where he received his 
first degree. He is teaching embryology and continuing his research on sex 

A loss most keenly felt by the President and the Office of Administration 
was the retirement of Mrs. Ruth McCollum, Administrative Assistant to 
the President. Mrs. McCollum joined the Institution staff in the administra- 
tive office of the Department of Terrestrial Magnetism in 1942, where she 
gave distinguished assistance during the difficult period of the war. In 1946 
she transferred to the Bursar's office in the Office of Administration, where 
she served for thirteen years, first as secretary to the Bursar and then as 
Accountant. During the latter part of this period Mrs. McCollum con- 
tributed part of her time and skill to general responsibilities of the Office of 
Administration. Early in 1959 she became Administrative Assistant to the 
President. Her management of arrangements for the Annual Meeting of the 
Board of Trustees was always a model of organization and good taste. Her 
artistic talent appeared in many ways in her work, much to the Institution's 
advantage, as in the annual departmental exhibits. No problem was too 
difficult to tax her good humor, and long hours only increased her devotion 

60 A modified female of bovine heterosexual twins. 


to the Institution. Her many talents and fine spirit are much missed by all 
who worked with her. 

Wilbur A. Pestell, Administrative Assistant at the Department of Plant 
Biology, also retired on June 30, 1962. He was actively associated with the 
Institution for 42 years, a period of dedicated service seldom equaled by past 
employees. He worked first in the Division of Publications in Washington, 
subsequently at the Desert Botanical Laboratory near Tucson, then at the 
Coastal Laboratory at Carmel, California, and finally as Secretary in the 
Department of Plant Biology at Stanford. His faithful work cleared routine 
tasks from the way of many others whose scientific results have been re- 
ported in these Year Books. 

. . . and Changes . . . 

The year 1962, in addition to signalizing the sixtieth anniversary of the 
Institution, also marked a significant change in its internal organization. 
Upon the retirement of Dr. Berwind P. Kaufmann, the fourth Director of 
the Department of Genetics, on June 30, 1962, the status of genetics re- 
search within the Institution was altered. As of July 1, the Department of 
Genetics became the Genetics Research Unit, with Alfred D. Hershey as 
Director. The work of the Unit will center on the research of Hershey, 
Barbara McClintock, and their associates at Cold Spring Harbor. In 
September 1962, Helen Gay, another Staff Member of the Unit, transferred 
her work to Ann Arbor, Michigan, where she will continue her association 
with Dr. Kaufmann. 

The Department of Genetics was formed in 1921 from a merger of the 
former Department for Experimental Evolution and the Eugenics Records 
Office. The Department for Experimental Evolution, which was formed in 
1906, had been preceded by the Station for Experimental Evolution, 
established at the present site in Cold Spring Harbor, New York, during 
1904. The dominant traits of the Department of Genetics were clearly those 
it inherited from the Department for Experimental Evolution. For fifty- 
eight years the research groups that successively made Cold Spring Harbor 
their scientific home maintained a research tradition which in many ways 
has been the story of genetics progress in the United States. Originally con- 
ceived by its first Director, C. B. Davenport, and inspired by the preceding 
work of Hugo de Vries in the Netherlands, the Cold Spring Harbor labora- 
tory has had an uncanny record of association with and stimulation of the 
main currents in genetic thought during the more than half century of its 
existence. Even twenty years ago Milislav Demerec, on the eve of his be- 
coming the third Director of the Department, could say that the "backing 


given to genetical research by the Institution undoubtedly accounts to a 
large degree for the fact that the United States now occupies the leading 
position in this branch of science." 61 

The Station's first work followed Dr. Davenport's lead. He had a con- 
suming ambition to prove experimentally the broad application of Mendel's 
law as rediscovered in 1900 by de Vries in the Netherlands, Correns in 
Germany, and von Tschermak-Seysenegg in Austria. Davenport's early 
work on poultry, birds, and mammals did actually furnish classic experi- 
mental confirmation of the broad application of Mendelian inheritance. 

Davenport's student and later colleague G. H. Shull provided one of the 
most unusual chapters in the Laboratory's history by laying the theoretical 
foundations of hybrid corn cultivation, described by Mangelsdorf as "the 
most far-reaching development in applied biology of this quarter century." 62 
Shull's recognition and exploitation of heterosis (hybrid vigor), which he 
named, gave the basic principle "which underlies almost the entire hybrid 
corn enterprise." 63 More recently the same plant in the hands of Barbara 
McClintock has been a highly successful medium for discovering the muta- 
tional behavior of genes. Both may certainly be counted among the most 
significant achievements in genetics. 

Illustrative of the range of interests to be found in the work of the 
Laboratory through its fifty-eight years of history are the pioneering experi- 
mental studies of C. C. Little on the inheritance of tumors, followed later by 
E. C. MacDowell's and J. S. Potter's studies of mouse leukemia; the 
foundation of cytogenetics by John Belling, followed by the productive 
cytogenetic research undertaken on Datura (of the potato family) by A. F. 
Blakeslee, the second Director of the Department, and his colleagues, and 
succeeded more recently by B. P. Kaufmann's cytogenetic studies on 
Drosophila; the painstaking studies of Milislav Demerec, mapping the gene 
loci of Escherichia coli and Salmonella; and the work of A. D. Hershey, 
exploring the molecular structure of the bacterial phage chromosome. 

Through the years the Department has been no less favored by geneticists 
who have been associated with it on a part-time basis. Among the Research 
Associates and Guest Investigators who were connected with the Depart- 
ment at one time or another in its history were: W. E. Castle, E. B. Wilson, 
C. B. Bridges, H. E. Crampton, E. B. Badcock, L. C. Dunn, Th. Dobzhan- 
sky, M. Delbrlick, A. Hollaender, D. G. Catcheside, M. Westergaard, C. 
Stern, and S. Brenner. 

The Institution will continue support of genetics research, although at a 

61 Carnegie Institution of Washington Year Book 4U P- 170. 

62 Paul C. Mangelsdorf, "Hybrid corn: its genetic basis and its significance in human affairs," 
in Genetics in the Twentieth Century, edited by L. C. Dunn, The Macmillan Company, New York, 
1951, p. 555. 

63 Ibid., p. 653. 


reduced scale by comparison with the Department's peak staff. The Genetics 
Research Unit will remain at Cold Spring Harbor. It is hoped that it will be 
joined by an interuniversity-sponsored research group investigating quanti- 
tative biology, the formation of which was being explored at the year's end. 

. . . and Gains 

Two new members were elected to the Board of Trustees of the Institution 
on May 11, 1962: William Walden Rubey and Carl Joyce Gilbert. 

Dr. Rubey is one of the country's most distinguished geologists. From 
1920 until 1960 he was associated with the United States Geological Survey, 
where his work received signal recognition in the Award of Excellence of the 
Department of the Interior in 1943 and the Distinguished Service award in 
1950. His contributions have added significantly to scientific understanding 
in several fields of geology, notably in knowledge of the original formation of 
the oceans, the transport of particles and sediments by running water, and 
the mechanics of very large overthrust faults. He is a graduate of the 
University of Missouri and holds honorary doctoral degrees both from that 
University and from Yale. At present he is serving as a member of the 
National Science Board (National Science Foundation) and of the board of 
directors of the American Association for the Advancement of Science. 

Mr. Gilbert is Chairman of the Board, Gillette Company, Boston. He is a 
graduate of the University of Virginia and the Harvard Law School. He is a 
member of the board of directors of several corporations, including the 
Raytheon Manufacturing Company, the Fiduciary Trust Company, and the 
Pepperell Manufacturing Company. Devoted to public service as well as to 
business, Mr. Gilbert is a member of the board of managers and past presi- 
dent of the Boston Dispensary, vice-chairman of the Massachusetts Port 
Authority, trustee of the New England Center Hospital, member of the 
administrative board of the New England Medical Center, vice-president 
of the New England Council, and trustee and member of the executive 
committee of Tufts College. Before he became chairman of the board of the 
Gillette Company in 1958 he had served as its president. 

It is always a special pleasure to record the honors that have come to 
members of the Institution. 

Presentation of the Kettering award for 1961 to Dr. Vannevar Bush, 
retired President of the Institution, was made at a conference in Washing- 
ton, D. C, of the Patent, Trademark, and Copyright Foundation of George 
Washington University for outstanding work in the field of patents, trade- 
marks, and related areas. 


At the Mount Wilson and Palomar Observatories, Ira S. Bowen, the 
Director, was elected a member of the Royal Society of Sciences of Uppsala, 
Sweden. The Newcomb Cleveland prize of $1000 was awarded to Halton C. 
Arp, Staff Member, on December 29, 1961, by the American Association 
for the Advancement of Science for "a noteworthy paper, representing an 
outstanding contribution in science.' ' Robert P. Kraft, Staff Member, re- 
ceived the Helen B. Warner prize of the American Astronomical Society for 
outstanding research by a young member of the Society. Guido Munch and 
Allan R. Sandage, Staff Members, were elected fellows of the American 
Academy of Arts and Sciences. Fritz Zwicky, Staff Member, was elected a 
member of the International Academy of Astronautics. This organization, 
which is only a year old, is the first international academy of scientists and 
engineers who have made contributions to space technology. It is limited to 
165 active members in the life sciences, basic sciences, and engineering. 

At the Geophysical Laboratory, Philip H. Abelson, the Director, received 
on April 6, 1962, the Washington State University Regents' Distinguished 
Alumnus award for the academic year 1961-1962. 

Scott E. Forbush, Staff Member of the Department of Terrestrial 
Magnetism, was elected to membership in the National Academy of 
Sciences, April 24, 1962, and on June 14, 1962, he received an honorary 
doctor of science degree from Case Institute of Technology, Cleveland, 
Ohio, for his contributions to our understanding of cosmic-ray phenomena. 

At the Department of Plant Biology, Jens Clausen, retired Staff Member, 
was made a Knight of the Order of Dannebrog by the King of Denmark in 
recognition of his contributions to botany and genetics. The Danish consul 
presented the decoration to Dr. Clausen on October 13, 1961, in San 

M. Demerec, retired Director of the Department of Genetics, was awarded 
the Kimber Genetics medal of the National Academy of Sciences on April 
24, 1962, "in recognition of his many contributions to the understanding of 
the genetics of various plants, Drosophila, bacteria, and viruses, and 
especially for his leadership in the investigation of unstable genes, the 
mutation process, genetics of micro-organisms and the genetic fine structure 
of the gene." 

J. E. S. Thompson, retired Staff Member of the Department of Archae- 
ology, received the honorary degree of doctor of humane letters and the 
Drexel Medal for Archaeology from the University of Pennsylvania in 
February 1962. 

Scientists and Scholars, 1902 - 1962 

In essence, the whole quality of the Institution, and its history, lie with those who 
have been associated with it over the years. Following are the names of the senior 
scientific staff members of all departments of the Institution over the last fifteen 
years (or over the last fifteen years of the existence of terminated departments). 
Below them in each department are listed the names of eminent and representative 
scientists and scholars who have been members of the staff or otherwise affiliated 
with the Institution since it was founded in 1902. A list is also given of all Fellows of 
the Carnegie Institution of Washington since the beginning of its Fellowship Program 
in 1947, and a list of grantees and others affiliated with the Institution but not with 
any particular department. 





Desert Laboratory, opened in 1903, became headquarters of Department of Botanical Research 
in 1905; name changed to Laboratory for Plant Physiology in 1923; reorganized in 1928 as Division 
of Plant Biology, including ecology; name changed to Department of Plant Biology in 1951. 


Daniel T. MacDougal, 1906-1927 

Herman A. Spoehr, 1927-September 1930, September 1931-1947 {Chairman)) 1947-1950 

{Chairman Emeritus) 

C. Stacy French, 1947— 

Staff Members 

Jeanette S. Brown, 1958 — 
Jens C. Clausen, 1931-1956 
David C. Fork, 1961 — 
Paul Grun, 1949-1954 
William M. Hiesey, 1926— 
David C. Keck, 1928-1951 

Donald W. Kupke, 1955-1956 
Harold W. Milner, 1927— 
Malcolm A. Nobs, 1939-1941, 1951- 
James H. C. Smith, 1925-1961 
Harold H. Strain, 1927-1962 
Ellen C. Weaver, 1961-1962 

Violet (Koski) Young, 1949-1953 

John Belling, 1921-1933 
William A. Cannon, 1903-1924 
Frederic E. Clements, 1917-1941 
Waldo S. Glock, 1931-1938 
Harvey M. Hall, 1918-1932 

Garrett J. Hardin, 1942-1946 
Burton E. Livingston, 1906-1909 
Francis E. Lloyd, 1906 
Winston M. Manning, 1941-1946 
Forrest Shreve, 1908-1945 

Godfrey G. Sykes, 1906-1929 

Other Scientists and Scholars Associated with the Department 

Leroy R. Abrams, 1932 

(Stanford University) 
Ernest Anderson, Research Associate 

1932-1936 (University of Arizona) 
William A. Arnold, Research Associate 


(Oak Ridge National Laboratory) 
Eric Ashby, 1930 

(Clare College, Cambridge University) 
Daniel I. Axelrod, 1937, 1939, 1944, 1950, 1959 

(University of California) 
Ernest B. Babcock, Research Associate 

1926-1945 (University of California) 
Irving W. Bailey, Research Associate 

1928-1930, 1932-1939 

(Harvard University) 
Charles E. Bessey, 1914 

(University of Nebraska) 

Nathaniel L. Britton, Research Associate 

1902, 1912-1916, 1918-1922 

(New York Botanical Garden) 
Ursula Brodfiihrer, 1956 

(University of Munich) 
Douglas H. Campbell, 1911 

(Stanford University) 
Ralph W. Chaney, Research Associate 


(University of California, Berkeley) 
William S. Cooper, 1919-1925 

(University of Minnesota) 
Frederick V. Coville, 1902-1905 

(U. S. Department of Agriculture; 

later, U. S. National Museum) 
Pierre Dansereau, 1949 

(University of Montreal; later, 

New York Botanical Garden) 



John P. Decker, 1957 

(U. S. Forest Service) 
Lee R. Dice, Research Associate 

1929-1930, 1932-1934-1938 

(University of Michigan) 
Erling Dorf, 1930, 1936, 1942 

(Princeton University) 
A. E. Douglass, Research Associate 

1924-1938 (University of Arizona) 
Newton B. Drury, Research Associate 


(California State Parks Commission) 
Benjamin M. Duggar, Research Associate 

1920-1921 (Missouri Botanical Garden; 

later, University of Wisconsin) 
Friedrich Ehrendorfer, 1951-1952 

(University of Vienna) 
Robert Emerson, Research Associate 


(California Institute of Technology) 
G. E. Erdtmann, 1930 

(University of Stockholm) 
William G. Farlow, 1905 

(Harvard University) 
Edward E. Free, Research Associate, 1920 

(U. S. Department of Agriculture) 
Martin Gibbs, 1962 (Cornell University) 
John W. E. Glattfeld, Research Associate 

1920-1921 (University of Chicago) 
Richard H. Goodwin, 1950 

(Connecticut College) 
Verne E. Grant, 1949-1950 

(Rancho Santa Ana Botanic Garden) 
Helen M. Habermann, 1959 

(Goucher College) 
Per Halldal, 1955-1957 (University of Oslo) 
Francis T. Haxo, 1957 

(Scripps Institution of Oceanography) 
Robert Hill, 1952 (Cambridge University) 
A. Stanley Holt, 1959 

(National Research Council of Canada) 
Ellsworth Huntington, Research Associate in 

Geology, 1903-1904, 1910-1912, 1915-1917, 

1922-1923 (Yale University) 
Donald A. Johansen, 1931-1932 

(private research) 
Ivan M. Johnston, 1942 (Harvard University) 
Erik G. J0rgensen, 1959 

(Royal Danish School of Pharmacy) 
Robert W. Krauss, 1951-1955 

(University of Maryland) 
Elias Landolt, 1953-1955 

(Swiss Federal Institute of Technology) 

Charlton M. Lewis, Research Associate 

1938-1941 (Patent Agent, 

Barkelow and Lewis, Pasadena) 
Harlan Lewis, 1954-1955 

(University of California, Los Angeles) 
Esmond R. Long, 1914-1915 

(University of Chicago; later, Henry Phipps 

Institute, University of Pennsylvania) 
John M. Macfarlane, 1902 

(University of Pennsylvania) 
Axel Madsen, 1962 (Royal Veterinary 

and Agricultural College, Copenhagen) 
Herbert L. Mason, 1925 

(University of California) 
Max Milner, 1957 (UN Children's Fund, 

Food Conservation Division) 
George T. Moore, 1914 

(Missouri Botanical Garden, St. Louis) 
Vladimir Moravek, Research Associate, 1926 

(University of Brno, Czechoslovakia) 
Jack E. Myers, 1950-1951, 1959 

(University of Texas) 
Hedda Nordenshiold, 1949 

(Royal Agricultural College, Uppsala) 
Axel Nygren, 1950 

(Royal Agricultural College, Uppsala) 
Winthrop J. V. Osterhout, Research Associate 

1922-1924 (Harvard University; later, 

Rockefeller Institute for Medical Research) 
James B. Overton, Research Associate 

1903, 1926-1927 (University of Wisconsin) 
George J. Peirce, 1910-1912 

(Stanford University) 
Gifford Pinchot, 1902 

(U. S. Department of Agriculture; later, 

Yale University and Governor of 

Thomas R. Pray, 1960-1961 

(University of Southern California) 
Joseph N. Rose, Research Associate 

1908, 1910-1923 (U. S. National Museum) 
Gilbert M. Smith, Research Associate 

1926-1927 (Stanford University) 
Roger Y. Stanier, 1959 

(University of California, Berkeley) 
G. Ledyard Stebbins, Jr., 1934-1936, 1945 

(University of California, Davis) 
Bernard Strehler, 1955 (National Heart 

Institute, Baltimore City Hospital) 
Walter T. Swingle, 1904 

(U. S. Department of Agriculture) 
Hiroshi Tamiya, 1952-1953 (Tokugawa 

Institute for Biological Research) 



Edwin W. Tisdale, 1959 (University of Idaho) 
Sam F. Trelease, 1914 (Columbia University) 
Vladimir tJlehla, Research Associate, 1924 

(University of Brno, Czechoslovakia) 
Cornelius B. Van Niel, 1931-1932 

(Stanford University) 
Chakrauarti S. Venkatesh, 1955-1956 

(Forest Research Institute, India) 
Wolf Vishniac, 1957 (Yale University; 

later, University of Rochester) 
Diter von Wettstein, 1959 

(University of Copenhagen) 

Heinrich Walter, 1929 

(University of Stuttgart) 
John E. Weaver, Research Associate 

1922-1930 (University of Nebraska) 
George R. Wieland, Research Associate 

1903-1934, 1941 (Yale University) 
Ira L. Wiggins, Research Associate 

1932-1933, 1936 (Stanford University) 
Paul C. Wilbur, 1926-1927 (Food Machinery 

and Chemical Corporation, San Jose) 
S. W. Williston, 1904 (University of Chicago) 
Frederick T. Wolf, 1960 

(Vanderbilt University) 


Mount Wilson Observatory organized in 1904; unified operation with the Palomar Observatory 
of the California Institute of Technology began in 1948. 


George E. Hale, 1904-1923; 1923-1936 (Honorary) 

Walter S. Adams, 1924-1945 

Ira S. Bowen, 1946— 

Staff Members 

Halton C. Arp, 1957— 
Walter Baade, 1931-1958 
Harold D. Babcock, 1909-1948 
Horace W. Babcock, 1946— 
William A. Baum, 1950— 
Arthur D. Code, 1956-1958 
Armin J. Deutsch, 1951 — 
Olin Eggen, 1961 — 
Jesse L. Greenstein, 1948 — 
Robert F. Howard, 1961— 
Fred Hoyle, 1957-1962 
Edwin P. Hubble, 1919-1953 
Milton L. Humason, 1917-1957 
Alfred H. Joy, 1915-1948 
Robert B. King, 1938-1948 
Robert P. Kraft, 1960— 
Paul W. Merrill, 1919-1952 

J. A. Anderson, 1916-1943 
Theodore Dunham, Jr., 1930-1947 
Arthur S. King, 1908-1943 

F. G. Pease, 1904-1938 

G. W. Ritchey, 1905-1919 
Charles E. St. John, 1908-1930 

Rudolph L. Minkowski, 1937-1960 
Guido Munch, 1951— 
Seth B. Nicholson, 1915-1957 
J. Beverley Oke, 1958— 
Donald E. Osterbrock, 1953-1958 
Edison Pettit, 1920-1955 
Alexander Pogo, 1950-1959 
Robert S. Richardson, 1931-1958 
Allan R. Sandage, 1952— 
Roscoe F. Sanford, 1918-1949 
Maarten Schmidt, 1959 — 
Otto Struve, 1962— 
Henrietta H. Swope, 1952 — 
Albert G. Wilson, 1948-1953 
Olin C. Wilson, 1931 — 
Ralph E. Wilson, 1938-1951 
Fritz Zwicky, 1925— 

Frederick H. Seares, 1909-1940 
Harlow Shapley, 1914-1921 
Sinclair Smith, 1923-1938 
Gustaf Stromberg, 1917-1946 
Adrian van Maanan, 1912-1946 



Other Scientists and Scholars Associated with the Department 

Charles G. Abbot, 1909-1948 

(Smithsonian Institution) 
Giorgio Abetti, 1909, 1930 

(Observatorio di Arcetri) 
Lawrence Aller, 1946-1961 

(University of Indiana; later, University of 

Michigan and University of California, 

Los Angeles) 
Edward E. Barnard, 1904-1905, 1912 

(Yerkes Observatory) 
W. Becker, 1962 (University of Basel) 
Dirk Brouwer, Research Associate 

1940-1944 (Yale University) 
John A. Carroll, 1924 (Cambridge University) 
William de Sitter, 1932 

(Observatory of Leiden) 
Albert Einstein, 1933 (Preussische 

Akademie der Wissenschaft, Berlin) 
E. Freundlich, 1926 

(Astrophysical Observatory, Potsdam) 
Henry G. Gale, Research Associate 

1909-1911 (University of Chicago) 
Leo Goldberg, 1940 

(Harvard College Observatory) 
Guillermo Haro, 1958 

(Tonantzintla Observatory) 
George Herbig, 1948, 1950, 1954 

(Lick Observatory) 
Ejnar Hertzsprung, 1912 

(Potsdam Observatory) 
Erik Holmberg, 1940-1941, 1947, 1951, 1955 

(Lund Observatory) 
Sir James Hop wood Jeans, Research Associate 

1922-1947 (Royal Society of London) 
Jacobus C. Kapteyn, Research Associate 

1908-1922 (University of Groningen) 
Philip C. Keenan, 1953-1962 

(Perkins Observatory) 
Gerard P. Kuiper, 1942, 1950, 1954 

(Yerkes Observatory; later, 

University of Arizona) 
Robert B. Leighton, 1951-1962 

(California Institute of Technology) 
Abbe Le Maitre, 1933 (University of Louvain) 
A. O. Leuschner, Research Associate 

1906-1907, 1924 (University of California) 
Bertil Lindblad, 1920-1921, 1950 

(Stockholm Observatory) 
Knut Lundmark, 1922-1923, 1930, 1933, 1938 

(University of Uppsala) 
W. J. Luyten, 1951, 1959 

(University of Minnesota) 

Robert R. McMath, 1950-1960 

(McMath-Hulbert Observatory) 
N. U. Mayall, 1951 (Lick Observatory; later, 

Kitt Peak National Observatory) 
A. A. Michelson, Research Associate 

1903-1904, 1919-1931 

(University of Chicago) 
Dayton C. Miller, 1921 

(Case School of Applied Science) 
S. A. Mitchell, Research Associate 

1924-1927, 1934-1944 

(University of Virginia) 
W. W. Morgan, 1957-1962 

(Yerkes Observatory) 
Earnest F. Nichols, Research Associate 

1908-1909 (Dartmouth College; later, 

Yale University and Massachusetts 

Institute of Technology) 
Y. Ohman, 1934 (University of Uppsala) 
Jan H. Oort, Research Associate 

1924, 1939, 1952, 1958-1959, 1961 

(Leiden Observatory) 
P. Th. Oosterhoff, 1934, 1960 

(Leiden Observatory) 
L. Perek, 1959 (Astronomical Institute of 

Czechoslovak Academy of Sciences) 
L. Plaut, 1956-1959 

(Kapteyn Astronomical Laboratory) 
Frank E. Ross, 1903-1909, 1927-1939 

(Yerkes Observatory) 
S. Rosseland, 192&-1927 

(International Research Fellow; later, 

University of Oslo) 
Henry N. Russell, Research Associate 

1903-1905, 1921-1947 

(Princeton University) 
Edwin E. Salpeter, Research Associate, 1959 

(Cornell University) 
Jan Schilt, 1925-1926 (International Research 

Fellow; later, Yale University and 

Rutherfurd Observatory, Columbia 

Martin Schwarzschild, 1946-1954 

(Princeton University) 
C. D. Shane, 1929-1930 

(University of California) 
Frederick Slocum, 1933-1934 

(Van Vleck Observatory) 
Lyman Spitzer, Jr., 1937-1940, 1948-1959 

(Princeton University) 
Joel Stebbins, Research Associate 

1930, 1932-1948 (University of Wisconsin) 



Carl Stormer, Research Associate, 1912-1915 

(University of Christiania) 
Bengt Stromgren, 1950, 1960 

(Institute for Advanced Study) 
Pol Swings, Research Associate 

1944-1946, 1958-1959 (University of Liege) 
A. D. Thackeray, 1935-1936 (Commonwealth 

Fellow; later, Radcliffe Observatory) 
Albrecht Unsold, 1929, 1957, 1961 

(University of Kiel) 

H. C. Van de Hulst, 1954 

(Leiden Observatory) 
Albert E. Whitford, 1933-1957 

(University of Wisconsin; later, 

Lick Observatory) 
Rupert Wildt, 1935-1936 (National Research 

Fellow; later, Yale University) 
R. v. d. R. Woolley, Research Associate 

1929-1931, 1958-1959, 1961 

(Royal Greenwich Observatory) 


Organized as the Department of International Research in Terrestrial Magnetism on April 1, 1904. 
Name changed to Department of Terrestrial Magnetism in 1905. 


Louis A. Bauer, 1904-1929 

John A. Fleming, 1929-1934 (Acting); 1935-1946 

Merle A. Tuve, 1946— 

Staff Members 

Philip H. Abelson, 1939-1953 
L. T. Aldrich, 1950— 
Lloyd V. Berkner, 1933-1951 
Ellis T. Bolton, 1951— 
Roy J. Britten, 1951 — 
Bernard F. Burke, 1953— 
Dean B. Cowie, 1944— 
John W. Firor, Jr., 1953-1961 
Scott E. Forbush, 1927— 
W. Kent Ford, Jr., 1957— 
Oliver H. Gish, 1922-1948 
John W. Graham, 1951-1958 
Stanley R. Hart, 1961 — 
Norman P. Heydenburg, 1935- 
Ellis A. Johnson, 1935-1956 

Brian J. McCarthy, 1960— 
W. C. Parkinson, 1913-1950 
Richard B. Roberts, 1937 — 
William J. Rooney, 1924-1949 
T. Jefferson Smith, 1962— 
John S. Steinhart, 1961 — 
Howard E. Tatel, 1947-1957 
Georges M. Temmer, 1953 — 
George R. Tilton, 1951-1956 
Oscar W. Torreson, 1923-1952 
Ernest H. Vestine, 1938-1957 
George R. Wait, 1920-1951 
Harry W. Wells, 1932— 
George W. Wetherill, 1954-1960 

J. P. Ault, 1905-1929 
S. J. Barnett, 1917-1926 
E. H. Bramhall, 1941-1944 
G. Breit, 1924-1929 
O. Dahl, 1926-1936 

F. T. Davies, 1929-1939 
H. M. W. Edmonds, 1906-1930 
H. W. Fisk, 1905-1932 
Lawrence R. Hafstad, 1928-1946 
James A. Van Allen, 1939-1941 



Other Scientists and Scholars Associated with the Department 

Samuel J. Mcintosh Allen, Research Associate 

1924-1928 (University of Cincinnati) 
E. Amaldi, 1936-1937 

(Royal University of Rome) 
J. Bartels, Research Associate, 1930-1940 

(Fortsliche Hochschule, Eberswalde, 

Germany; later, Geophysikalisches Institut, 

Gottingen, Germany) 
Carl Barus, Research Associate 

1902, 1904-1923, 1926 (Brown University) 
Jesse W. Beams, 1934-1935 

(University of Virginia) 
J. C. Beattie, Research Associate, 1908-1911 

(South Africa College, Cape Town) 
Ralph D. Bennett, Research Associate 

1933-1940 (Massachusetts Institute of 

Technology; later, Vallecitos Atomic 

Laboratory, Pleasanton, California) 
Hans A. Bethe, 1936-1941 

(Cornell University) 
Henry G. Booker, Research Associate 

1938-1940 (Cornell University) 
Edward L. Bowles, Research Associate 

1939-1945 (Massachusetts Institute of 

Joseph C. Boyce, Research Associate, 

1939-1950 (Massachusetts Institute of 

Technology; later, New York University 

and Illinois Institute of Technology) 
Robert B. Brode, Research Associate 

1939-1941 (University of California, 

Richard E. Byrd, 1931-1932 

(U. S. Navy, Arctic explorer) 
Sydney Chapman, Research Associate 

1934-1940, 1951-1953 (Trinity College, 

Cambridge; later, Imperial College, 

London; High Altitude Observatory, 

Boulder, Colorado, and Geophysical 

Institute, College, Alaska) 
Georges N. Cohen, Research Associate 

1956-1959 (Institut Pasteur, Paris) 
Arthur H. Compton, Research Associate 

1931-1945 (University of Chicago; later, 

Washington University) 
Karl T. Compton, Research Associate 

1928-1934 (Princeton University; later, 

Massachusetts Institute of Technology) 
T. G. Cowling, 1950-1951 

(Princeton L^niversity) 

Hugh H. Darby, Research Associate 

1948-1950 (consultant in biochemistry, 

Mt. Airy, Maryland) 
N. Ernest Dorsey, Research Associate 

1912-1913 (National Bureau of Standards) 
George Gamow, Research Associate, 1935-1944 

(George Washington University; later, 

University of Colorado) 
Enrique Gaviola, Research Associate 

1928-1929 (Comision de Astrofisica y 

Radioastronomia, Universidad de Buenos 

Ross Gunn, Research Associate, 1938-1944 

(Naval Research Laboratory; later, 

U. S. Weather Bureau and American 

Anton L. Hales, Research Associate, 1960 

(University of the Witwatersrand ; later, 

Graduate Research Center, Inc., Dallas, 

John S. Hall, 1954— 

(Lowell Observatory, Flagstaff, Arizona) 
Raymond G. Herb, 1935 

(University of Wisconsin) 
Victor F. Hess, Research Associate, 1940-1946 

(Fordham University) 
Thomas H. Johnson, Research Associate 

1933-1946 (Bartol Research Foundation; 

later, Brookhaven National Laboratory, 

Atomic Energy Commission, and Raytheon 

Manufacturing Company) 
Arthur E. Kennelly, Research Associate, 

1924-1935 (Harvard University and 

Massachusetts Institute of Technology) 
Serge A. Korff, Research Associate, 1936-1945 

(New York University) 
D. la Cour, 1931-1932 (Danish 

Meteorological Survey, Copenhagen) 
H. A. Lorentz, Research Associate, 1920 

(University of Leiden) 
Frank T. McClure, Research Associate 

1955-1960 (Applied Physics Laboratory, 

Johns Hopkins University) 
J. D. McGee, Research Associate, 1959-1960 

(Imperial College) 
Kenneth R. McQuillen, Research Associate 

1951-1960 (University of Cambridge) 
Robert A. Millikan, Research Associate 

1921-1945 (California Institute of 




B. Y. Mills, 1953-1954 

(Radiophysics Laboratory, Commonwealth 

Industrial and Research Organization, 

Sydney, Australia) 
S. K. Mitra, 1936-1937 

(University of Calcutta) 
T. Nagata, 1950-1951 

(Geophysical Institute, Tokyo) 
J. L. Pawsey, 1957-1958 (Commonwealth 

Industrial and Research Organization, 

Greenleaf W. Pickard, Research Associate 

1927-1935 (consultant in electrical 

engineering, Newton Centre, Massachusetts) 
Wilson M. Powell, Research Associate 

1942-1943 (Lawrence Radiation 

Laboratory, University of California) 
A. T. Price, 1952 

(The Royal Technical College, Glasgow) 
Norman Ramsey, 1938-1939 

(Harvard University) 
J. A. Ratcliffe, 1950-1951 

(Cavendish Laboratory, Cambridge) 
Bruno Rossi, 1932-1933 

(Massachusetts Institute of Technology) 
Sir Arthur Rucker, 1904-1915 

(Royal College of Science, 

South Kensington, London) 

M. N. Saha, 1936-1937 

(Allahabad University, India) 
Marcel Schein, 1939-1944 (University of 

Chicago; later, University of California) 

A. Schmidt, 1905-1907 
(Potsdam Magnetic Observatory) 

B. F. J. Schonland, 1952 
(University of the Witwatersrand) 

Frederick Slocum, Research Associate, 1920 

(Brown University; later, Wesleyan 

F. Graham Smith, 1952-1954 

(Cavendish Laboratory, Cambridge) 
J. C. Street, Research Associate, 1933-1934 

(Harvard University) 
H. U. Sverdrup, Research Associate, 1926-1939 

(Geophysical Institute, Bergen) 
W. F. G. Swann, 1916-1920 

(Bartol Research Foundation) 
John T. Tate, Research Associate, 1941-1945 

(University of Minnesota) 
Edward Teller, 1936-1937 (University of 

Chicago; later, Lawrence Radiation 

Laboratory, University of California) 
Manuel S. Vallarta, Research Associate 

1940-1941, 1948-1950 

(National University of Mexico) 
John von Neumann, 1948-1949, 1955-1957 

(Institute for Advanced Study, Princeton) 


Station for Experimental Evolution opened in 1904; name changed to Department of Experimental 
Evolution in 1906; combined with Eugenics Record Office in 1921 to form Department of Genetics. 


Charles B. Davenport, 1904-1934 

Albert F. Blakeslee, 1935-November 30, 1941 

Milislav Demerec, December 1, 1941-1942 (Acting); 1943- June 30, 1960 

Berwind P. Kaufmann, July 1, 1960- January 31, 1962 (Acting); February 1, 1962- June 30, 1962 

Staff Members 

Ernst Caspari, 1947-1949 Margaret R. McDonald, 1946- 

Helen Gay, 1960-1962 E. C. MacDowell, 1914-1952 

Alfred D. Hershey, 1950— George Streisinger, 1956-1960 

Barbara McClintock, 1942— Bruce Wallace, 1947-1949 

Evelyn M. Witkin, 1950-1955 

Arthur M. Banta, 1909-1930 
Robert W. Bates, 1931-1941 

Barbara S. Burks, 1936-1941 
John Belling, 1920-1929 



Ugo Fano, 1940-1946 
Ross A. Gortner, 1909-1914 
J. Arthur Harris, 1907-1924 
H. H. Laughlin, 1918-1940 
S. E. Luria, 1945-1946 
Frank E. Lutz, 1904-1909 

Charles W. Metz, 1914-1930 
Oscar Riddle, 1912-1945 
Sophie Satina, 1924-1942 
George H. Shull, 1904-1915 
Morris Steggerda, 1930-1944 
H. E. Warmke, 1938-1945 

Benjamin B. Wells, 1941-1942 

Other Scientists and Scholars Associated with the Department 

Edgar Anderson, 1941, 1945 

(Washington University; later, 

Missouri Botanical Garden) 
Ernest Ball, 1942-1943 

(University of North Carolina) 
Hans Bauer, 1936 (Kaiser- Wilhelm Institut 

fur Biologie, Berlin-Dahlem; later, 

Max-Planck Institut fur Meeresbiologie, 

Wilhelmshaven, Germany) 
George W. Beadle, 1935 

(California Institute of Technology; 

later, University of Chicago) 
John J. Biesele, Research Associate, 1944-1946 

(University of Texas) 
Dietrich Bodenstein, 1944 

(University of Virginia) 
Sydney Brenner, 1954 

(University of the Witwatersrand; later, 

Cavendish Laboratory) 
Vernon Bryson, Research Associate, 1942-1943 

(Rutgers University) 
John T. Buchholz, 1921-1941 (University of 

Arkansas; later, University of Illinois) 
Sir Macfarlane Burnet, 1950 

(The Walter and Eliza Hall Institute of 

Medical Research, Melbourne) 
J. Gordon Carlson, 1937-1940 

(University of Alabama; later, 

University of Tennessee) 
J. Lincoln Cartledge, 1921-1924 

(University of West Virginia) 
William E. Castle, Research Associate 

1904-1943 (Harvard University; later, 

University of California) 
David G. Catcheside, Research Associate 

1957-1959 (University of Birmingham, 

Donald R. Charles, 1929-1930 

(Sarah Lawrence College; later, 

University of Rochester) 
Albert Claude, 1946 

(Rockefeller Institute for Medical Research; 

later, The Free University of Brussels) 

Marie E. Conklin, 1937-1941 

(Adelphi College) 
J. N. Couch, 1925-1926 
M (University of North Carolina) 

ax Delbnick, 1937 

(California Institute of Technology) 
Hugo de Vries, Research Associate, 1904-1918 

(University of Amsterdam) 
Th. Dobzhansky, Research Associate 

1936-1949 (California Institute of 

Technology and Columbia University; 

later, Rockefeller Institute for Medical 

L. C. Dunn, 1929 (Columbia University) 
Boris Ephrussi, 1937 (Institut de Biologie 

Physico-Chimique, Paris; later, Centre 

National de la Recherche Scientifique, 

Harold D. Fish, Research Associate, 1919-1924 

(Denison University; later, University of 

Pierre Fredericq, 1958 (University of Liege) 
Gabriel Gasid, 1945-1946 (University of Chile) 
Norman H. Giles, Jr., 1941 

(Yale University) 
Joseph S. Gots, 1954, 1955, 1957 

(University of Pennsylvania) 
John W. Gowen, 1915, 1940 

(Iowa State College) 
Ludwig von Graf, 1906 (University of Graz) 
C. C. Guthrie, 1909 (University of Pittsburgh) 
Ike Gustafsson, 1937-1938 (Institute for 

Genetic Research, Svalof, Sweden; later, 

Forest Research Institute of Sweden, 

Olli Halkka, 1959 (University of Helsinki) 
Alexander Hollaender, Research Associate 

1942-1944 (National Institutes of Health; 

later, Oak Ridge National Laboratory) 
Sally Hughes-Schrader, 1934 

(Sarah Lawrence College; later, 

Duke University) 



C. Leonard Huskins, 1935, 1936 

(McGill University; later, 

University of Wisconsin) 
Fritz Kaudewitz, 1954 (Max-Planck Institut 

fur Virusforschung, Tubingen) 
Tage Kemp, 1932 (University of Copenhagen) 
P. C. Koller, 1938 (University of Edinburgh; 

later, Chester Beatty Research Institute, 

Jaroslav Krizenecky, 1928-1929 

(Zootechnical Research Institute, Brno, 

Victor K. LaMer, 1916-1917 

(Columbia University) 
Raymond Latarjet, 1945-1946 

(Institut Pasteur, Paris) 
Albert Levan, 1951 (University of Lund) 
Cyrus Levinthal, 1951 

(University of Michigan; later, 

Massachusetts Institute of Technology) 
C. C. Little, Research Associate, 1919-1925 

(Roscoe B. Jackson Memorial Laboratory, 

Bar Harbor, Maine) 
Edward L. Mark, Research Associate 

1904-1910 (Harvard University) 
Horace N. Marvin, 1941-1942 

(University of Arkansas) 
William J. Moenkhaus, 1904-1906 

(Indiana University) 
G. Montalenti, 1951 (University of Naples; 

later, University of Rome) 
H. J. Muller, 1921 (Indiana University) 
Robert K. Nabours, Research Associate 

1929-1930 (Kansas State College) 
James V. Neel, 1940 (Dartmouth College; 

later, University of Michigan) 
Howard B. Newcombe, Research Associate 

1938, 1945-1947 (Atomic Energy 

Commission of Canada, Ltd.) 
Theophilus S. Painter, 1923 

(University of Texas) 
Raymond Pearl, Research Associate 

1904-1906 (University of Michigan; later, 

Johns Hopkins University) 
Marcus M. Rhoades, 1941 (Columbia 

University; later, Indiana University) 
Maurice N. Richter, 1930-1952 

(Columbia University; later, 

New York University Medical Center) 
Franz Schrader, 1934 (Columbia University; 

later, Duke University) 
Edmund W. Sinnott, 1938 (Columbia 

University; later, Yale University) 
B. M. Slizynski, 1936-1937 (University of 

Cracow; later, University of Edinburgh) 

Evelyn E. B. Smith, Research Associate, 1958 

(University of Glasgow) 
Laurence H. Snyder, 1922 

(University of Hawaii) 
Arnold H. Sparrow, 1940, 1941 

(Brookhaven National Laboratory) 
Warren P. Spencer, 1935 (College of Wooster) 
S. G. Stephens, Research Associate, 1945-1947 

(North Carolina State College) 
Curt Stern, 1933, 1938, 1944, 1946 

(University of Rochester; later, 

University of California, Berkeley) 
A. Tavcar, 1951 (University of Zabreb) 
Howard J. Teas, 1942-1943 

(Nuclear Center, Mayagiiez, Puerto Rico) 
Rene Thomas, 1957-1958 

(University of Brussels) 
N. TimofeerT-Ressovsky, 1932 

(Kaiser- Wilhelm Institut fur Hirnforschung; 

later, Academy of Sciences, Novosibirsk, 

Jun-ichi Tomizawa, Research Associate 

1957-1959 (National Institute of Health, 

William L. Tower, Research Associate 

1904-1917 (University of Chicago) 
J. van Overbeek, 1940, 1941 

(California Institute of Technology; later, 

Agricultural Laboratory, Shell 

Development Company, Modesto, 

C. H. Waddington, 1938 (Cambridge 

University; later, University of Edinburgh) 
Mogens Westergaard, Research Associate 

1957-1959 (Universitetets Genetiske 

Institut, Copenhagen) 
Fritz von Wettstein, 1938 

(Kaiser- Wilhelm Institut fur Biologie, 

Berlin-Dahlem; later, University of Vienna) 
M. J. D. White, 1947, 1950-1952 

(University of London and University of 

Texas; later, Commonwealth Scientific and 

Industrial Research Organization, 

Canberra, Australia) 
P. W. Whiting, 1933-1935 

(University of Pennsylvania) 
Maurice Whittinghill, 1938 

(Bennington College; later, 

University of North Carolina) 
Edmund B. Wilson, Research Associate 

1904-1909, 1936-1938 

(Columbia University) 
Charles Yanofsky, 1956 (Western Reserve 

University; later, Stanford LTniversity) 



The following persons carried on genetic 
studies with Carnegie Institution support: 

Calvin B. Bridges, Research Associate 
1916-1938 (Columbia University; later, 
California Institute of Technology) 

Thomas Hunt Morgan, Research Associate 

1916-1945 (Columbia University; later, 
California Institute of Technology) 

Jack Schultz, Research Associate, 1929-1941 
(California Institute of Technology; later, 
Institute for Cancer Research, Philadelphia) 

A. H. Sturtevant, Research Associate 
1916-1931 (Columbia University; later, 
California Institute of Technology) 


Organized in 1906. Opened in 1907. 


Arthur L. Day, 1909-1936 
Leason H. Adams, 1936-1937 (Acting); 1938- July 31, 1952 
George W. Morey, August 1, 1952-August 31, 1953 (Acting) 
Philip H. Abelson, September 1, 1953 — 

Staff Members 

Norman L. Bowen, 1910-1937, 1947-1953 

Francis R. Boyd, Jr., 1953— 

John S. Burlew, 1936-1952 

Felix Chayes, 1947— 

Sydney P. Clark, Jr., 1957-1962 

Gordon L. Davis, 1941 — 

Gabrielle Donnay, 1955 — 

Joseph L. England, 1926 — ■ 

Hans P. Eugster, 1952-1958 

Roy W. Goranson, 1926-1951 

Hugh J. Greenwood, 1960 — 

Joseph W. Greig, 1922-1960 

Thomas C. Hoering, 1959 — 

Earl Ingerson, 1935-1947 

Eugene T. Allen, 1907-1933 
Olaf Andersen, 1912-1918 
Tom. F. W. Barth, 1929-1940 
F. Russell von Bichowsky, 1916-1919 
A. F. Buddington, 1919-1920 
J. K. Clement, 1904-1907 
Pentti Eskola, 1921-1922 
Clarence N. Fenner, 1910-1937 
Michael Fleischer, 1936-1938 
Ralph E. Gibson, 1924-1946 
Sterling B. Hendricks, 1926 
James H. Hibben, 1928-1939 
John C. Hostetter, 1912-1919 

Frank C. Kracek, 1923-1956 
Gunnar Kullerud, 1954 — 
William S. MacKenzie, 1951-1957 
Patrick L. Parker, 1961 — 
Charles S. Piggot, 1925-1947 
Eugene Posnjak, 1913-1947 
Howard S. Roberts, 1917-1947 
J. Frank Schairer, 1927— 
George R. Tilton, 1956— 
George Tunell, 1925-1947 
O. Frank Tuttle, 1947-1953 
William D. Urry, 1938-1949 
Hatten S. Yoder, Jr., 1948— 
Emanuel G. Zies, 1913-1949 

John Johnston, 1908-1916 
Esper S. Larsen, Jr., 1907-1909 
Robert H. Lombard, 1915-1927 
Herbert E. Merwin, 1909-1959 
Elbert F. Osborn, 1938-1945 
George A. Rankin, 1907-1916 
Earnest S. Shepherd, 1904-1946 
Robert B. Sosman, 1908-1928 
Henry S. Washington, 1912-1934 
Walter P. White, 1904-1935 
Erskine D. Williamson, 1914-1923 
Fred E. Wright, 1906-1944 
Ralph W. G. Wyckoff, 1919-1927 



Other Scientists and Scholars Associated with the Department 

Frank D. Adams, 1903-1911 

(McGill University) 
Samuel K. Allison, 1925-1926 

(University of Chicago) 
George F. Becker, 1903-1909 

(U. S. Geological Survey) 
J. C. Branner, 1906-1907 

(Arkansas Geological Survey) 
L. E. J. Brouwer, 1931-1932 

(Royal Dutch Petroleum) 
Thomas C. Chamberlin, Research Associate 

1903-1927 (University of Chicago) 
Hessel de Vries, Research Associate, 1958-1959 

(University of Groningen, the Netherlands) 
J. D. H. Donnay, 1953— 

(Johns Hopkins University) 
William H. Emmons, 1903 

(University of Chicago) 
Henry Faul, 1956-1957 

(U. S. Geological Survey) 
Grove Karl Gilbert, 1904, 1906-1907 

(U. S. Geological Survey) 
Harry H. Hess, Research Associate, 1940-1942 

(Princeton University) 
Joseph P. Iddings, 1905 

(University of Chicago) 
Emilie Jager, 1958-1959 

(University of Bern, Switzerland) 
Willard F. Libby, Research Associate 

1954-1959 (University of California at 

Los Angeles) 

George D. Louderback, 1903-1906 

(University of Nevada) 
Gordon J. F. MacDonald, 1955— 

(Institute of Geophysics and Planetary 

Physics, University of California at Los 

Forest Ray Moulton, Research Associate 

1903-1922 (University of Chicago) 
Paul Niggli, 1913-1914 (Zurich) 
C. C. Patterson, 1958 

(California Institute of Technology) 
Frank A. Perret, Research Associate 

1938-1943 (Volcanological Museum, 

St. Pierre, Martinique) 
Hans Ramberg, Research Associate, 1955-1958 

(University of Chicago) 
Paul Ramdohr, Research Associate, 1960 — 

(University of Heidelberg) 
C. S. Slichter, 1903, 1906 

(University of Wisconsin) 
David B. Stewart, 1954-1956 

(U. S. Geological Survey) 
C. E. Tilley, Research Associate, 1955 — 

(Cambridge University) 
Johan August Udden, 1925, 1928 

(University of Texas) 
C. R. Van Hise, 1902-1903 

(University of Wisconsin) 
C. E. Van Orstrand, 1904, 1906-1910 

(U. S. Geological Survey) 

Organized 1914. 



Franklin P. Mall, 1914-1917 

George L. Streeter, 1918-1940 

George W. Corner, 1941-1955 

James D. Ebert, 1956— 

Staff Members 

George W. Bartelmez, 1949-1960 
David W. Bishop, 1952— 
Bent G. Boving, 1951 — 
Donald D. Brown, 1962— 

Robert K. Burns, 1940-1962 

Arpad Csapo, 1951-1955 

Robert L. DeHaan, 1956 — 

James F. Didusch, 1913-1940, 1945-1955 



Louis B. Flexner, 1940-1951 
Osborne 0. Heard, 1913-1956 
Chester H. Heuser, 1921-1950 
I. R. Konigsberg, 1961 — 
Elizabeth M. Ramsey, 1949— 

Herbert M. Evans, 1913-1915 
Carl G. Hartman, 1925-1941 
Margaret R. Lewis, 1915-1946 

Mary E. Rawles, 1957— 
Samuel R. M. Reynolds, 1941-1955 
Royal F. Ruth, 1956-1961 
David B. Tyler, 1947-1950 
Walter S. Wilde, 1944-1947 

Warren H. Lewis, 1914-1940 
Charles W. Metz, 1930-1940 
Adolph H. Schultz, 1916-1925 

Other Scientists and Scholars Associated with the Department 

William E. Adams, 1957-1958 

(University of Otago, New Zealand) 
Ines de Allende, 1941-1943 

(University of Cordoba) 
Howard D. Andervont, 1923-1926 

(Johns Hopkins School of Hygiene and 

Public Health; later, National Cancer 

T. S. and B. F. Argyris, 1961-1962 

(Syracuse University) 
Alexander Barry, 1947 

(University of Michigan) 
T. H. Bast, 1929-1930 

(University of Wisconsin) 
J. D. Boyd, 1934-1935 

(Cambridge University) 
E. A. Boyden, 1939-1940 (University of 

Minnesota; later, University of Washington) 
Washington Buno, 1945-1946 

(University of Montevideo) 
Gerald L. Carlson, 1960-1962 

(Massachusetts Institute of Technology) 
Eliot R. Clark and Eleanor L. Clark, 

1907-1914 (Johns Hopkins School of 

Medicine; later, University of 

George W. Corner, Jr., 1943 — 

(Johns Hopkins School of Medicine) 

E. V. Cowdry, 1913-1916 

(Johns Hopkins School of Medicine; 
later, Washington University) 
Maria Victoria de la Cruz, 1949-1950 
(Institute of Cardiology, Mexico City) 

F. Cuajunco, 1927-1928 
(University of the Philippines) 

Harold Cummins, 1927-1928 

(Tulane University) 
Vera Danchakoff, 1924 

(Columbia University) 
Carl Lawrence Davis, 1920 

(University of Maryland) 
Vincent J. De Feo, 1955-1957 

(University of Illinois) 

Anatole S. Dekaban, 1959 — 

(National Institutes of Health) 
Charles A. Doan, 1923-1924 

(Johns Hopkins School of Medicine; 

later, Ohio State University) 
Jules Duesberg, Research Associate, 1915-1918 

(University of Liege) 
Robert K. Enders, 1930-1932 

(Swarthmore College) 
Thomas R. Forbes, 1937-1938 

(Johns Hopkins School of Medicine; 

later, Yale University) 
Fritz Fuchs, 1950-1951 

(University of Copenhagen) 
Ernest D. Gardner, 1955 

(Wayne State University) 
E. M. K. Geiling, 1935-1936 

(Johns Hopkins School of Medicine; 

later, University of Chicago and 

U. S. Food and Drug Administration) 
Isidore Gersh, 1934-1935 

(Johns Hopkins School of Medicine; 

later, University of Chicago) 
G. O. Gey, 1924-1930 

(Johns Hopkins School of Medicine) 
Joseph Gillman, Research Associate 

1941-1942, 1946-1948 

(University of the Witwatersrand) 
G. Gitlin, 1950 

(Hebrew University, Jerusalem) 
Timothy Glover, 1961 

(University of Liverpool) 
Charles M. Goss, 1948 (Louisiana State 

University School of Medicine) 
Donald J. Gray, 1955 (Stanford University) 
Gilbert S. Greenwald, 1954-1956 

(University of Washington; later, 

University of Kansas) 
Paul W. Gregory, 1928-1929 

(Harvard University; later, 

University of California, Davis) 
E. Grodzinski, 1928-1929 

(University of Cracow) 



Alan F. Guttmacher, 1921-1922 

(Johns Hopkins School of Medicine; later, 

Mt. Sinai Hospital, New York City) 
Manfred S. Guttmacher, 1921-1922 

(Johns Hopkins School of Medicine; 

later, private practice in psychiatry, 

John W. S. Harris, 1961 

(London Hospital Medical College) 
Arthur T. Hertig, Research Associate 

1933-1956 (Harvard Medical School) 
Marion Hines, 1925-1947 

(University of Chicago and Johns Hopkins 

University; later, Emory University) 
A. St. G. Huggett, 1952-1953 

(St. Mary's Hospital Medical School, 

Irwin H. Kaiser, 1946-1947 

(University of Minnesota) 
Seymour Katsh, 1955-1958 

(University of Colorado) 
Franz Keibel, Research Associate, 1914-1918 

(Anatomical Institute, Strassburg) 
Benjamin F. Kingsbury, 1917-1918 

(Cornell University) 
Abraham Kulangara, 1959-1961 

(University of California, Los Angeles; 

later, All India Institute of Medical 

Sciences, New Delhi) 
Orthello R. Langworthy, 1924-1930 

(Johns Hopkins University) 
Hans Laufer, 1957-1959 

(Johns Hopkins University) 
John McKenzie, 1959 

(University of Aberdeen, Scotland) 
Joseph E. Markee, 1935-1936 

(Stanford University; later, 

Duke University) 
Arthur Meyer, 1917-1918 

(Stanford University) 
Tom Mori, 1960-1961 (Tohoku University) 
Harland W. Mossman, 1934-1935 

(University of Wisconsin) 
William B. Muchmore, 1959 

(University of Rochester) 
Jacques Mulnard, 1957 

(University of Brussels) 
Alton M. Mun, 1959-1961 

(Washington State College; later, 

University of Maine) 
G. Muratori, 1934-1935 

(University of Padua, Italy) 
Roberto Narbaitz, 1959 

(University of Buenos Aires) 

Catherine Neill, 1952-1953 (London; later, 

Johns Hopkins Medical School) 
Martin Nordmann, 1929-1930 

(University of Tubingen) 
Ronan O'Rahilly, 1961-1962 

(St. Louis University) 
F. Orts Llorca, 1959 (University of Madrid) 
John Papaconstantinou, 1958-1960 

(Johns Hopkins School of Medicine; 

later, University of Connecticut) 
W. M. Paul, 1954 (University of Toronto) 
Donald F. Poulson, 1936-1937 

(Yale University) 
Curt P. Richter, 1927-1930 

(Johns Hopkins School of Medicine) 
Eduardo de Robertis, 1941-1942 

(University of Buenos Aires) 
John Rock, 1938-1945 

(Harvard Medical School) 
Edward Roosen-Runge, 1957 

(University of Washington) 
Florence Sabin, 1914-1924 

(Johns Hopkins University; later, 

Rockefeller Institute) 
Jorgen U. Schlegel, 1948-1949 

(University of Copenhagen) 
Harold D. Senior, 1917-1918 

(New York University) 
Ronald Singer, 1951-1952 

(University of Cape Town, South Africa; 

later, University of Chicago) 
William L. Straus, Jr., 1925-1930 

(Johns Hopkins School of Medicine) 
Fritz Strauss, 1950 (University of Bern) 
Somers H. Sturgis, 1942-1943 

(Massachusetts General Hospital) 
Francis H. Swett, 1924-1927 

(Johns Hopkins School of Medicine; 

later, Duke University) 
Pierre Tardent, 1959-1960 

(Zoological Station, Naples) 
M. H. Toosy, 1948-1949 

(Lahore Medical School) 
Theodore W. Torrey, 1952 

(Indiana University) 
U. U. Uotila, 1939-1940 

(Harvard Medical School) 
W. J. van Doorenmaalen, 1958-1959 

(Municipal University, Amsterdam) 
R. Walmsley, 1935-1936 

(University of Edinburgh) 
Lewis H. Weed, Research Associate 

1914-1919, 1921-1935 

(Johns Hopkins University) 



Karl M. Wilson, 1913-1924 

(Johns Hopkins School of Medicine; 

later, University of Rochester) 
Milton C. Winternitz, 1914-1916 

(Johns Hopkins School of Medicine; 

later, Yale University) 

George Wislocki, 1916-1931 

(Johns Hopkins School of Medicine; 

later, Harvard University ) 
Emil Witschi, 1941-1942 

(University of Iowa) 


Organized as a "bureau" in 1903; became a "department" in 1905; terminated as a department 
and incorporated as the Section of United States History in a new Division of Historical Research, 


Andrew C. McLaughlin, 1903-1905 

John F. Jameson, 1905-1928 

None, 1928-1930 

Staff Members 

Edmund C. Burnett, 1907-1932 
Frances G. Davenport, 1905-1927 
Elizabeth Donnan, 1911-1919 
Shirley Farr, 1921-1922 
Mary F. Griffin, 1922-1925 

Waldo G. Leland, 1903-1945 
Marguerite M. McKee, 1925-1929 
David W. Parker, 1909-1913, 1925-1928 
Charles O. Paullin, 1912-1936 
Leo F. Stock, 1910-1945 

Other Scientists and Scholars Associated with the Department 

Charles Francis Adams, 1902 

(Massachusetts Historical Society) 
Ephraim D. Adams, 1904 

(Stanford University) 
William H. Allison, 1906-1911 (Bryn Mawr 

College; later, Colgate University) 
Charles M. Andrews, 1904-1918 

(Yale University) 
James C. Ballagh, 1907-1908 

(Johns Hopkins University; later, 

University of Pennsylvania) 
Adolf F. A. Bandelier, 1911-1914 

(Columbia University) 
Eugene C. Barker, 1906 

(University of Texas) 
John S. Bassett, 1921-1928 (Smith College) 
Herbert C. F. Bell, 1916-1923 

(Bowdoin College; later, Wesleyan 

Samuel F. Bemis, 1923-1925 

(George Washington University; later, 

Yale University) 
Herbert E. Bolton, 1907-1913 

(Stanford University; later, 

University of California) 

Harold Martin Bowman, 1907-1908 

(Boston University School of Law) 
Julian P. Bretz, 1906 

(later, Cornell University) 
Helen T. Catterall, 1918-1933 (Boston Bar) 
Isaac Joslin Cox, 1906-1908 

(University of Cincinnati; later, 

Northwestern University) 
Walter F. Dodd, 1908 (Library of Congress; 

later, Yale University) 
Max Farrand, 1912-1913 (Yale University; 

later, Henry E. Huntington Library and 

Art Gallery) 
Albert B. Faust, 1912-1916 

(Cornell University) 
William S. Ferguson, 1906-1908 

(University of California; later, 

Harvard University) 
CarlR. Fish, 1908-1911 

(University of Wisconsin) 
Worthington C. Ford, 1903-1906 

(Library of Congress; later, 

Massachusetts Historical Society) 
Dixon R. Fox, 1919-1920 (Columbia 

University; later, Union College) 



Frank A. Golder, 1914-1922 

(State College of Washington; later, 

Stanford University) 
Evarts B. Greene, 1918 (University of 

Illinois; later, Columbia University) 
Charles W. Hackett, 1918-1929 

(University of Texas) 
Charles H. Haskins, 1905-1908 

(Harvard University) 
Roscoe R. Hill, 1910-1917 

(later, Nicaraguan High Commission; 

National Archives) 
Frank H. Hodder, 1912 (University of Kansas) 
William Wirt Howe, 1904 (Board of Trustees, 

Carnegie Institution of Washington) 
William I. Hull, 1914 (Swarthmore College) 
Herman G. James, 1923 (University of 

South Dakota; Ohio University) 
Marcus W. Jernegan, 1907-1926 

(University of Chicago) 
Louise P. Kellogg, 1922 

(Wisconsin State Historical Society) 
Benjamin B. Kendrick, 1910 (Women's 

College of University of North Carolina) 
Marion D. Learned, 1908-1912 

(University of Pennsylvania) 
Orin G. Libby, 1912 

(University of North Dakota) 
George W. Littlehales, 1915 

(George Washington University; formerly 

with Carnegie Institution of Washington's 

Department of Terrestrial Magnetism) 
Alfred T. Mahan, 1914-1915 

(U. S. Navy, retired) 
William R. Manning, 1908-1910 

(George Washington University; later, 

Department of State) 

John J. Meng, 1936-1954 

(Catholic University of America; later, 

Hunter College) 
Herbert L. Osgood, 1912-1918 

(Columbia University) 
Edwin W. Pahlow, 1926-1927 

(Ohio State University) 
Frederic L. Paxson, 1910-1914 

(University of Wisconsin; later, 

University of California) 
Francis S. Philbrick, 1914-1915 

(University of Pennsylvania) 
Jesse S. Reeves, 1912-1913 

(University of Michigan) 
James A. Robertson, 1909-1917, 1931-1932 

(Stetson University; later, Archives of 

Robert W. Rogers, 1924-1927 

(Drew Theological Seminary) 
Joseph Schafer, 1918-1919 

(University of Oregon) 
George W. Scott, 1903-1905 (Library of 

Congress; Columbia University) 
William R. Shepherd, 1905-1908 

(Columbia University) 
William A. Slade, 1904-1905 

(Library of Congress) 
Frederick J. Turner, 1916-1917 

(Harvard University) 
Arnold J. F. van Lear, 1919-1926 

(New York State Library; later, 

New York State Education Department) 
Claude H. Van Tyne, 1904-1908 

(University of Michigan) 
Ray H. Whitbeck, 1914-1915 

(University of Wisconsin) 
Irene A. Wright, 1925-1928 (Library of 

Congress; later, National Archives) 


Established 1930, superseding the Department of Historical Research, which became a section of 
United States History in the Division. The other two sections were the Section of Aboriginal American 
History, which continued the archaeological work already begun by Sylvanus G. Morley in Central 
America and by E. H. Morris in southwestern United States, and the Section of the History of 
Science. Became the Department of Archaeology, 1951. 

Alfred V. Kidder, Chairman, 1930-1950 
Harry E. D. Pollock, Director, 1951-1958 



Staff Members 

Eleanor B. Adams, 1934-1949 
Robert S. Chamberlain, 1937-1947 
Sylvanus G. Morley, 1914-1948 
Earl H. Morris, 1925-1955 
Alexander Pogo, 1929-1950 
Tatiana A. Proskouriakoff, 1939 — 
Ralph L. Roys, 1930-1953 
Karl Ruppert, 1925-1956 
George A. L. Sarton, 1918-1949 

France V. Scholes, 1931-1946 
Anna 0. Shepard, 1933— 
Edwin M. Shook, 1933-1958 
A. Ledyard Smith, 1929-1958 
Robert E. Smith, 1931-1960 
Gustav Stromsvik, 1926-1957 
Sol Tax, 1938-1947 
J. Eric S. Thompson, 1935-1959 
Alfonso Villa Rojas, 1932-1947 

Manuel J. Andrade, 1932, 1936-1940 J. Ignacio Rubio Man6, 1936-1942 

Abraham M. Halpern, 1941-1942 Oliver G. Ricketson, Jr., 1920-1940 

Henry B. Roberts, 1926-1939 

Other Scientists and Scholars Associated with the Division 

Sophie D. Aberle, Research Associate 

1933-1940 (United Pueblo Agency, 

Albuquerque, New Mexico; later, 

Chief Nutrition, Bernalillo County 

Indian Hospital) 
Robert M. Adams, Jr., 1951-1952 

(Oriental Institute, University of Chicago) 
Monroe Amsden, 1923-1924, 1927 

(southwestern archaeologist) 
E. Wyllys Andrews, 

1939-1940, 1941-1942, 1947-1948 

(Tulane University) 
Herman Beyer, 1937 (Tulane University) 
Franz Blom, 1924-1925 (Tulane University) 
Stephen F. de Borhegyi, 1949 

(University of Oklahoma; later, 

Milwaukee Public Museum) 
George W. Brainerd, 1939-1942, 1948-1949 

(University of California, Los Angeles) 
Kirk Bryan, 1945 (Harvard University) 
W. R. Bullard, 1951-1953 

(Peabody Museum, Harvard University) 
Alfonso Caso y Andrade, Research Associate 

1936-1939 (Instituto Nacional Indigenista, 

Kenneth M. Chapman, 1935 

(University of New Mexico) 
Jean Chariot, 1926-1931 (painter and teacher) 
Ann Chowning, 1954-1955 

(Bryn Mawr College; later, 

University of Pennsylvania) 
I. Bernard Cohen, 1938-1941 

(Harvard University) 
Fay-Cooper Cole, 1931 

(University of Chicago) 

G. W. Collins, 1936-1937 

(U. S. Department of Agriculture) 
Frank H. Connell, 1931-1932 

(Dartmouth College) 
Luther S. Cressman, Research Associate 

1936-1942 (University of Oregon) 
John H. Denison, Jr., 1937-1938 

(Big Horn, Wyoming) 
Rollins A. Emerson, 1934-1935 

(Cornell University) 
F. W. Gaige, 1930-1931 

(University of Michigan) 
Rutherford J. Gettens, 1955 

(Freer Gallery of Art, Washington, D. C.) 
John P. Gillin, 1941-1943, 1945-1946 

(Duke University; later, University of 

Antonio Goubaud, 1944-1945 

(Instituto de Antropologia e Historia, 

Guatemala City) 
Carl E. Guthe, Research Associate, 1921-1922 

(New York State Museum, Albany) 
Lewis U. Hanke, 1935-1939 

(Library of Congress; later, University of 

Texas and Columbia University) 
Mark R. Harrington, Research Associate 

1930-1936 (Southwest Museum, 

Los Angeles) 
William A. Heidel, 1928-1939 

(Wesleyan University) 
Edgar B. Howard, 1934-1942 

(University of Pennsylvania) 
William T. Howard, Jr., 1924 

(Johns Hopkins University) 
Jesse D. Jennings, 1936-1937 

(University of Utah) 



J. H. Kempton, 1934-1938 

(U. S. Department of Agriculture) 
J. Steward Lincoln, 1940-1941 

(Guatemala City) 
John M. Longyear, III, 1937-1939, 

1941-1942, 1945-1947, 1949-1950 

(Colgate University) 
Samuel K. Lothrop, 1922-1933 

(Peabody Museum, Harvard University) 
Cyrus L. Lundell, Research Associate 

1933-1941 (University of Michigan) 
Maud Worcester Makemson, 1943 

(Vassar College) 
Norman A. McQuown, 1937-1949 

(University of Chicago) 
Paul S. Martin, 1926-1928 

(Chicago Natural History Museum) 
Ann Axtell Morris, 1926-1931 

(Boulder, Colorado) 
Lila M. O'Neale, 1935-1936 

(University of California, Berkeley) 
Arthur S. Pearse, 1928-1936 

(Duke University) 
Wilson Popenoe, 1935-1936 

(United Fruit Company; later, Escuela 

Agricola Panamericana, Tegucigalpa, 

Robert Redfield, Research Associate 

1930-1949 (University of Chicago) 
Ruth Reeves, Research Associate, 1934-1935 

(New York City, New York) 
Juan de Dios Rosales, 1944-1946 

(Instituto Indigenista de Guatemala) 
George M. Saunders, 1930-1932 

(Harvard University) 

Adolph H. Schultz, Research Associate 

1916-1925, 1937-1938 (Johns Hopkins 

University; later, University of Zurich) 
George C. Shattuck, 1929-1939 

(Boston City Hospital; later, 

Massachusetts General Hospital) 
Joseph L. Smith, 1941 

(Boston Museum of Fine Arts) 
Philip E. Smith, 1953-1954 

(University of Toronto) 
R. Stadelman, 1936-1938 

(U. S. Department of Agriculture) 
L. C. Stuart, 1932-1933 

(University of Michigan) 
John Teeple, 1928-1931 

(consulting chemist, New York City) 
Antonio Tejeda F., 1938-1939, 1944-1947 

(Museo Nacional de Arqueologia y 

Etnologia, Guatemala City) 
Donald E. Thompson, 1954-1955 

(University of Wisconsin) 
Aubrey S. Trik, 1935-1938 (University 

Museum, University of Pennsylvania) 
Melvin Tumin, 1942-1944 

(University of North Carolina) 
George C. Vaillant, 1925-1940 

(American Museum of Natural History; 

later, University Museum, University of 

Robert Wauchope, 1933-1936 

(Tulane University) 
Howell Williams, 1949-1950 

(University of California, Berkeley) 
Clark Wissler, Research Associate, 1924-1933 

(American Museum of Natural History; 

later, Yale University) 


Organized 1904; terminated 1916. 

Board Members 

Carroll D. Wright, Director, 1904-1909 (Clark College) 

Henry W. Farnam, Chairman, 1909-1916 (Yale University) 

Kenyon L. Butterfield, Agriculture and Forestry, 1904-1915 (Rhode Island College of 

Agriculture and Mechanic Arts) 

Victor S. Clark, Manufactures, 1906-1916 (Census Bureau) 

John R. Commons, The Labor Movement, 1909-1915 (University of Wisconsin) 

Davis R. Dewey, Money and Banking, 1904-1914 (Institute of Technology, Boston) 

Henry B. Gardner, Federal and State Finance, 1904-1914 (Brown University) 

J. W. Jenks, Industrial Organization, 1904-1914 (Cornell University) 



Emory R. Johnson, Domestic and Foreign Commerce, 1904-1915 (University of Pennsylvania) 

B. H. Meyer, Transportation, 1904-1916 (University of Wisconsin) 

S. N. D. North, Manufactures, 1904 (Census Bureau) 

Edward W. Parker, Mining, 1904-1915 (U. S. Geological Survey) 

W. Z. Ripley, Transportation, 1904 (Newton Centre, Massachusetts) 

Alfred Holt Stone, The Negro in Slavery and Freedom, 1906-1914 (Dunleith, Mississippi) 

Walter F. Willcox, Population and Immigration, 1904-1914 (Cornell University) 

Other Scientists and Scholars Associated with the Department 

Edith Abbott, 1905-1910 

(University of Chicago) 
Henry C. Adams, 1904 

(University of Michigan) 
Charles H. Ambler, 1910 

(Randolph-Macon College; later, 

(University of West Virginia) 
John B. Andrews, 1913-1915 

(numerous activities in labor economics) 
Oliver Edwin Baker, 1912 

(U. S. Department of Agriculture) 
Emily Greene Balch, 1904-1907 

(Wellesley College) 
F. Spencer Baldwin, 1908-1909 

(Boston University) 
J. Lynn Barnard, 1905-1908 

(Philadelphia School of Pedagogy) 
Alvard Longley Bishop, 1907-1908 

(Yale University) 
Frank W. Blackmar, 1904-1914 

(University of Kansas) 
Ernest Ludlow Bogart, 1904-1912 

(Oberlin College; later, Princeton 

University and University of Illinois) 
Beverley Waugh Bond, 1908-1909 

(Purdue University; later, 

University of Cincinnati) 
William K. Boyd, 1910-1913 

(Duke University) 
James E. Boyle, 1905-1908 

(University of North Dakota) 
Solon J. Buck, 1906-1913 

(University of Indiana; later, 

Archivist of the United States) 
Thomas N. Carver, 1904-1912 

(Harvard University) 
Robert E. Chaddock, 1909 

(Columbia University) 
John B. Clark, 1902 (Columbia University) 
Frederick A. Cleveland, 1905-1913 

(New York University; later, in charge of 

President Taft's Commission on Economy 

and Efficiency) 

Thomas Conway, Jr., 1904-1913 

(University of Pennsylvania) 
Mary Roberts Coolidge, 1907-1910 

(Mills College) 
John Lee Coulter, 1908-1912 

(University of Minnesota; later, 

U. S. Tariff Commission) 
James Walter Crook, 1908-1911 

(Amherst College) 
Ira Brown Cross, 1909-1913 

(University of California) 
Stuart Daggett, 1904-1913 

(University of California) 
Edgar M. Dawson, 1908-1913 

(Princeton University; later, 

Hunter College) 
Clive Day, 1907-1909 (Yale University) 
David T. Day, 1908-1911 

(U. S. Geological Survey; later, 

U. S. Bureau of Mines) 
Robert N. Denham, Jr., 1908 

(University of Michigan; later, 

National Labor Relations Board) 
Carroll W. Doten, 1906-1908 

(Massachusetts Institute of Technology) 
W. E. B. Dubois, 1908 (Atlanta University) 
Edwin C. Eckel, 1904-1908 

(U. S. Geological Survey) 
Richard T. Ely, 1904-1909 

(University of Wisconsin) 
Fred Rogers Fairchild, 1904-1909 

(Yale University) 
Henry Pratt Fairchild, 1908-1909 

(Bowdoin College; later, 

New York University) 
John I. Falconer, 1912-1913 

(Ohio State University) 
Albert B. Faust, 1907-1910 

(Cornell University) 
Walter L. Fleming, 1908-1911 

(Louisiana State University and 

Vanderbilt University) 



Albert A. Giesecke, 1904-1910 

(University of Pennsylvania; later, 

University of Cuzco, Peru) 
Eugene A. Gilmore, 1910 

(University of Wisconsin; later, 

State University of Iowa) 
E. A. Goldenweiser, 1904-1908 

(various economic posts in U. S. 

L. C. Graton, 1908-1913 

(U. S. Geological Survey; later, 

Harvard University) 
Elmer C. Griffith, 1908-1911 

(Kalamazoo College) 
George Gorham Groat, 1904-1908 

(Ohio Wesleyan University; later, 

University of Vermont) 
James Edward Hagerty, 1905-1909 

(Ohio State University) 
Robert M. Haig, 1910-1913 

(Columbia University) 
Matthew Brown Hammond, 1904-1909 

(Ohio State University) 
Glover D. Hancock, 1908-1911 

(Amherst College; later, 

Washington and Lee University) 
Lewis Henry Haney, 1906-1910 

(New York University) 
Hugh Sisson Hanna, 1905-1908 

(U. S. Bureau of Labor Statistics) 
Adelaide R. Hasse, 1905-1917 

(New York Public Library; later, 

Brookings Institution) 
Frank I. Herriott, 1905-1911 

(Drake University) 
Benjamin H. Hibbard, 1908-1914 

(Iowa State College; later, 

University of Wisconsin) 
Henry E. Hoagland, 1911-1913 

(Ohio State University) 
Roy Jay Holden, 1908-1915 

(Virginia Polytechnic Institute) 
Jacob H. Hollander, 1904-1907 

(Johns Hopkins University) 
Solomon S. Huebner, 1904-1911 

(University of Pennsylvania) 
Walter Renton Ingalls, 1904-1908 

(construction engineer, New York City) 
Theodore H. Jack, 1909-1911 

(Emory University; later, 

Randolph-Macon College) 
Edward D. Jones, 1908-1914 

(University of Michigan) 
T. J. Jones, 1908 (Hampton Institute; 

later, Phelps Stokes Fund) 

Clyde L. King, 1911-1913 

(University of Pennsylvania) 
Julius Klein, 1909 

(U. S. Department of Commerce) 
Francis Baker Laney, 1904-1915 

(U. S. National Museum; later, 

U. S. Geological Survey) 
John Lapp, 1908 (Cornell University; 

later, Marquette University) 
Laurence M. Larson, 1905-1909 

(University of Illinois) 
C. K. Leith, 1904-1915 

(University of Wisconsin) 
Isaac P. Lippincott, 1909 

(Washington University) 
Oliver C. Lockhart, 1908-1913 

(Ohio State University) 
Isaac A. Loos, 1905-1909 

(State University of Iowa) 
Gerald Francis Loughlin, 1915 

(U. S. Geological Survey) 
David A. McCabe, 1912 

(Princeton University) 
Charles McCarthy, 1904 

(U. S. Commission on Industrial Relations; 

later, U. S. Food Administration) 
James Farley McClelland, 1904-1905 

(Columbia School of Mines; later, 

Yale University) 
George McCutchen, 1908-1912 

(University of South Carolina) 
S. J. McLean, 1906-1910 

(University of Toronto) 
F. L. McVey, 1908-1911 

(University of North Dakota; later, 

University of Kentucky) 
E. T. Miller, 1905-1915 (University of Texas) 
H. A. Millis, 1909-1912 

(Stanford University; later, 

University of Chicago) 
Wesley C. Mitchell, 1904-1908 

(University of California; later, 

New School for Social Research) 
Blaine F. Moore, 1908-1909 

(U. S. Commission on Industrial Relations; 

later, University of Kansas) 
Charles E. Munroe, 1904-1910 

(George Washington University) 
Henry R. Mussey, 1904 (Wellesley College) 
W. T. Nardin, 1905 (Pet Milk Company) 
Selig Perlman, 1911-1915 

(University of Wisconsin) 
Warren Milton Persons, 1908 

(Colorado College; later, Harvard 




John B. Phillips, 1908-1909 

(University of Colorado; later, 

University of Indiana) 
Ulrich B. Phillips, 1904-1910 

(Tulane University; later, University of 

Michigan and Yale University) 
Charles F. Pidgin, 1908 

(Massachusetts Bureau of Statistics of 

CarlC. Plehn, 1904-1911 

(University of California) 
Fred Wilbur Powell, 1909-1913 

(Brookings Institution) 
Joseph Hyde Pratt, 1904-1910 

(University of North Carolina) 
E. P. Puckett, 1908-1913 (Central College) 
Charles Lee Raper, 1905-1909 

(University of North Carolina; later, 

Syracuse University) 
William A. Rawles, 1904-1911 

(University of Indiana) 
Heinrich Ries, 1904-1909 

(Cornell University) 
Thomas James Riley, 1908-1909 

(University of Missouri; later, 

Washington University) 
Clyde Orval Ruggles, 1908-1911 

(Ohio State University; later, 

Harvard University) 
Aaron M. Sakolski, 1906 

(New York University) 
David J. Saposs, 1911-1915 

(various government posts in labor 

William O. Scroggs, 1905-1911 

(Louisiana State University) 
A. E. Sheldon, 1904-1905 

(Nebraska Historical Society) 
St. George L. Sioussat, 1904-1913 

(University of the South; later, 

University of Pennsylvania) 
J. Russell Smith, 1904-1908 

(University of Pennsylvania; later, 

Columbia University) 
Yates Snowden, 1910-1911 

(University of South Carolina) 

Don C. Sowers, 1910-1913 

(University of Colorado) 
Robert James Sprague, 1908-1909 

(University of Maine; later, 

Rollins College) 
Harry Harkness Stoek, 1904-1908 

(editor, Mining and Minerals; later, 

University of Illinois) 
Edgar M. Sydenstricker, 1908-1915 

(U. S. Public Health Service) 
Henry C. Taylor, 1908-1915 

(University of Wisconsin; later, 

Farm Foundation) 
D. Y. Thomas, 1907-1908 

(University of Florida; later, 

University of Arkansas) 
William H. Tolman, 1908 

(Pawtucket, Rhode Island) 
Walter Sheldon Tower, 1905-1908 

(Bethlehem Steel Corporation; later, 

Iron and Steel Institute) 
Robert James Usher, 1905 

(Howard-Tilton Memorial Library, 

Tulane University) 
Francis Walker, 1909-1910 

(Federal Trade Commission) 
Royal Brunson Way, 190&-1908 

(Northwestern University; later, 

Beloit College) 
Nathan Austin Weston, 1905-1910 

(University of Illinois) 
Horace L. Wilgus, 1905-1909 

(University of Michigan) 
C. C. Williamson, 1905-1908 

(New York Public Library; later, 

Columbia University) 
Calvin Dill Wilson, 1908-1912 

(clergyman and author) 
Edwin E. Witte, 1911-1912 

(University of Wisconsin) 
R. R. Wright, Jr., 1908-1910 

(Georgia State Industrial College) 
Allyn A. Young, 1905-1910 

(Stanford University; later, Cornell 

University and Harvard University) 
Frederic G. Young, 1905-1913 

(University of Oregon) 




Established in 1804. Name changed to Tortugas Laboratory in 1923. Activities terminated in 1939. 


Alfred G. Mayer, 1904-1922 

William Harding Longley, 1923-1927 (Administrative Officer); 1928-1937 (Executive Officer) 

David Hilt Tennent, 1938-1939 (Executive Officer) 

Staff Members 

Paul S. Conger, 1924-1929, 1937-1938 Albert Mann, 1919-1933 

John W. Mills, 1906-1939 

Other Scientists and Scholars Associated with the Department 

Percy L. Bailey, Jr., 1937 

(College of the City of New York) 
Stanley C. Ball, 1913-1914, 1917 

(Massachusetts Agricultural College; 

later, Bishop Museum, Honolulu, and 

Peabody Museum, Yale University) 
Paul Bartsch, 

1912-1917, 1919, 1921-1927, 1930-1932 

(U. S. National Museum; later, 

George Washington University) 
Norman J. Berrill, 1937 (McGill University) 
Lawrence R. Blinks, 1925-1928 

(Rockefeller Institute; later, Stanford 

University and Hopkins Marine Station) 
H. Boschma, 1924 (Rijksuniversiteit, Leiden) 
Howard H. M. Bowman, 1915-1916 

(University of Pennsylvania; later, 

University of Toledo) 
Alan A. Boyden, 1931, 1933, 1935 

(Rutgers University) 
Charles M. Breder, Jr., 1928 

(New York Aquarium and 

American Museum of Natural History) 
Floyd J. Brinley, 1936-1937 

(North Dakota Agricultural College; 

later, University of Toledo) 
William K. Brooks, 1905-1907, 1909 

(Johns Hopkins University) 
Dugald E. S. Brown, 1934 

(New York University Medical School; 

later, University of Michigan) 
Walter E. Bullington, 1929-1930, 1934 

(Randolph-Macon College) 
Martin Burkenroad, 1928 (Tulane 

University; later, Marine Biological Station, 

National Museum of Panama) 

Lewis R. Cary, Research Associate 

1910-1918, 1920, 1929-1933, 1935 

(Princeton University) 
Edward L. Chambers, 1936 

(Princeton University; later, 

University of Miami School of Medicine, 

Coral Gables) 
Robert Chambers, 1936 

(Washington Square College, New York 

University; later, Marine Biological 

Laboratory, Woods Hole) 
Frank M. Chapman, 1907-1909 

(American Museum of Natural History) 
Hubert Lyman Clark, Research Associate 

1912-1917, 1929-1930 (Museum of 

Comparative Zoology, Harvard University) 
Leonard B. Clark, 1936-1937 

(Union College) 
Frank W. Clarke, 1919 

(U. S. Geological Survey) 
Leon J. Cole, 1906-1914 (Yale University; 

later, University of Wisconsin) 
John Colman, 1930 (Cambridge University) 
Edwin G. Conklin, 1905, 1907, 1909, 1915 

(Princeton University) 
Benjamin R. Coonfield, 1937 

(Brooklyn College) 
Rheinart P. Cowles, 1905-1909, 1914 

(Johns Hopkins University) 
Paul R. Cutright, 1936 (Beaver College) 
Ulric Dahlgren, 1906, 1908, 1911-1922 

(Princeton University) 
Reginald A. Daly, 1919 (Harvard University) 
John H. Davis, Jr., 1936-1937 

(Southwestern College; later, 

University of Florida) 



May W. de Laubenfels, 1926-1927, 1931, 1935 

(Pasadena Junior College; later, 

Oregon State College) 
George S. de Renyi, 1933 

(University of Pennsylvania) 
Richard B. Dole, 1913 

(U. S. Geological Survey) 
Henry H. Donaldson, 1916 

(Wistar Institute of Anatomy) 
William L. Doyle, 1933-1934 

(Johns Hopkins University; later, 

University of Chicago) 
George Harold Drew, 1911-1913 

(Christ's College, Cambridge University) 
Gilman A. Drew, 1912 

(Marine Biological Laboratory, 

Woods Hole) 
Charles H. Edmondson, 1906-1907 

(Iowa Wesleyan; later, University of 

Hawaii and Bishop Museum, Honolulu) 
Richard M. Field, 1919 (Museum of 

Comparative Zoology, Harvard University ; 

later, Princeton University) 
A. Haldane Gee, 1929 

(Scripps Institution of Oceanography; 

later, Foster D. Snell, Inc., New York) 
John H. Gerould, 1915, 1921-1922 

(Dartmouth College) 
Isidore I. Gersh, 1934 

(Johns Hopkins University Medical School; 

later, University of Chicago School of 

Abraham J. Goldforb, 1912-1913, 1915-1916 

(College of the City of New York) 
Hubert B. Goodrich, 1934 

(Wesleyan University) 
Myron Gordon, 1927, 1932 

(Cornell University; later, 

American Museum of Natural History and 

New York Zoological Society) 
James N. Gowanlock, 1929 

(Dalhousie University) 
Caswell Grave, 

1924, 1926-1929, 1932, 1934-1935 

(Washington University) 
George M. Gray, 1912 

(Marine Biological Laboratory, 

Woods Hole) 
Eugene W. Gudger, 1908, 1912-1915 

(North Carolina College for Women; later, 

American Museum of Natural History) 
George T. Hargitt, 1905 

(Northwestern University; later, 

Syracuse University) 

John E. Harris, 1933-1934, 1936 

(Cambridge University) 
J. A. Harrison, 1936 (University of London) 
Robert Hartmeyer, 1907 

(Berlin Zoological Museum) 
E. Newton Harvey, Research Associate 

1909-1925, 1929 (Princeton University) 
Shinkishi Hatai, 1916-1917 

(Wistar Institute of Anatomy) 
Frederick R. Hayes, 1931 

(Institute of Oceanography, 

Dalhousie University, Halifax) 
Edwin R. Helwig, 1932 

(University of Pennsylvania; later, 

University of Colorado) 
Walter N. Hess, 1930, 1937 

(Hamilton College) 
Davenport Hooker, 1905, 1907-1909, 1914 

(Yale University) 
Dwight L. Hopkins, 1928-1930 

(Duke University; later, 

Mundelein College, Chicago) 
Robert Tracy Jackson, 1912 

(Museum of Comparative Zoology, 

Harvard University) 
Merkel H. Jacobs, 1911 

(University of Pennsylvania) 
Norris Jones, 1936-1937 

(Swarthmore College) 
Harvey E. Jordan, 1907, 1909, 1912-1914 

(University of Virginia) 
E. Jorgensen, 1910 (University of Bergen) 
Carl Kellner, 1905-1907, 1909 

(Yale University) 
Milton J. Kopac, 1932-1934, 1936 

(University of California; later, 

New York University) 
Beverly W. Kunkel, 1930 (Lafayette College) 
Karl S. Lashley, 1913-1915 

(Johns Hopkins University; later, 

Harvard University and Yerkes 

Laboratories of Primate Biology) 
Marius Le Compte, 1936 

(Royal Museum of Natural History, 

James L. Leitch, 1931, 1933, 1935 

(University of California; later, 

Armstrong College) 
Ivey F. Lewis, 1927 (University of Virginia) 
Frank R. Lillie, 1935-1936 

(University of Chicago) 
Edwin Linton, 1906-1909 

(Washington and Jefferson College; later, 

University of Pennsylvania) 



Charles B. Lipman, 1920, 1922-1923 

(University of California) 
Balduin Lucke, 1936-1937 

(University of Pennsylvania Medical 

Jesse F. McClendon, 

1908-1910, 1916-1917, 1919 

(University of Missouri; later, 

University of Minnesota and 

Einstein Medical Center, Philadelphia) 
Oliver McCoy, 1927-1928 

(Johns Hopkins University; later, 

University of Rochester and 

China Medical Board of New York) 
Harold W. Manter, 1929-1931, 1933 

(University of Nebraska) 
Gordon Marsh, 1929, 1934-1937 

(University of Iowa) 
James C. Martin, 1933 

(University of California) 
Cloyd Heck Marvin, 1932 

(George Washington University) 
Samuel 0. Mast, 1910 (Goucher College; 

later, Johns Hopkins University) 
George Matthai, 1915 (Emmanuel College, 

Cambridge University) 
Grace Medes, 1915 (Bryn Mawr College; 

later, Lankenau Hospital Research Center, 

Seth E. Meek, 1909 

(Field 'Museum of Natural History, 

Charles W. Merriam, 1932 

(University of California; later, 

Cornell University and U. S. Geological 

Harry M. Miller, Jr., 1924-1926, 1928 

(Washington University; later, 

Rockefeller Foundation, Paris) 
Sergius Morgulis, 1923-1924 

(Creighton University) 
Charles E. Moritz, 1935 

(University of California; later, Redlands 

College and Philip Morris and Company) 
Theodor Mortensen, 1916 

(University of Copenhagen) 
Paul A. Nicoll, 1932, 1934-1935, 1937 

(Washington University; later, 

Indiana School of Medicine) 
Raymond C. Osburn, 1908, 1914 

(New York Aquarium; later, 

Ohio State University) 
Fernandus Payne, 1932, 1937 

(University of Indiana) 

Arthur S. Pearse, 1927, 1930 

(Duke University) 
Henry F. Perkins, 1903-1905 

(University of Vermont) 
Alexander Hamilton Phillips, 1915 

(Princeton University) 
Robert F. Pitts, 1935 ' 

(New York University; later, 

Cornell University) 
Harold H. Plough, 1935-1937 

(Amherst College) 
Frank M. Potts, 1913-1915, 1920, 1922 

(Cambridge University) 
Philip B. A. Powers, 1932, 1936 

(University of Pennsylvania) 
Henry S. Pratt, 1909-1910, 1924 

(Haverford College) 
Jacob E. Reighard, 1905, 1907, 1909 

(University of Michigan) 
Edwin E. Reinke, Research Associate 

1911-1915 (Vanderbilt University) 
Oscar W. Richards, 1933, 1935 

(Yale University; later, American Optical 

Company, Southbridge, Massachusetts) 
Gordon A. Riley, 1937 (Yale University; 

later, Bingham Oceanographic Laboratory, 

Yale University) 
Asa A. Schaeffer, Research Associate 

1919, 1921-1927, 1929 

(University of Tennessee; later, 

Temple University) 
Waldo L. Schmitt, 1924, 1929-1931 

(U. S. National Museum) 
William A. Setchell, 1920, 1922-1923 

(University of California) 
Eugene W. Shaw, 1915 

(U. S. Geological Survey) 
Clarence R. Shoemaker, 1925 

(U. S. National Museum) 
Charles F. Silvester, 1915 

(Princeton University; later, 

Captain, U. S. Army) 
H. G. Smith, 1933 (University of Bristol) 
Frederick C. Steward, 1932-1934, 1936 

(University of London; later, 

Cornell University) 
Charles R. Stockard, 1907-1910 

(Cornell University Medical College) 
Raymond G. Stone, 1930-1931, 1934 

(University of Kansas City) 
Frank A. Stromsten, 1907-1910 

(University of Iowa) 
Geoffrey Tandy, 1930, 1932 

(British Museum of Natural History) 
Vance Tartar, 1937 (Yale University) 



Shiro Tashiro, 1914-1915 

(University of Chicago; later, 

University of Cincinnati) 
Charles V. Taylor, 1924-1925 

(University of California; later, 

Stanford University) 
William R. Taylor, 1924-1925 

(University of Pennsylvania) 
David M. Tennent, 1936 (Yale University; 

later, Merck Institute for Therapeutic 

Research and Hess and Clark Division of 

Richardson-Merrell, Inc.) 
Harry Beal Torrey, Research Associate 

1926-1927 (Cornell University Medical 

School; later, Stanford University) 
Aaron L. Treadwell, 

1904, 1909-1910, 1913-1916, 1918, 

1920-1921 (Vassar College) 
Joseph M. Valentine, 1925 

(Yale University; later, 

Alabama Museum of Natural History) 
Gilbert Van Ingen, 1915 

(Princeton University) 
T. Wayland Vaughan, Research Associate 

1908-1917, 1919, 1922-1923 

(U. S. Geological Survey; later, 

Scripps Institution of Oceanography) 
J. Paul Visscher, 1929-1930 

(Western Reserve University) 

W. Seward Wallace, 1908 

(University of Nevada) 
John C. Waller, 1915 

(King's College, Cambridge University) 
William B. Wartman, 1928 

(University of Pennsylvania Medical 

School; later, Northwestern University) 
John B. Watson, 1907, 1909-1915 

(University of Chicago; later, 

William Esty and Company, New York) 
John W. Wells, 1931 (Cornell University) 
Roger C. Wells, 1919 

(U. S. Geological Survey) 
E. I. Werber, 1915 (Yale University) 
Douglas M. Whitaker, 1925 

(Stanford University) 
J. L. Williams, 1931 

(University of California) 
Benjamin H. Willier, 1935 

(University of Rochester; later, 

Johns Hopkins University) 
Henry V. Wilson, 1924 

(University of North Carolina) 
J. M. Wilson, 1932-1933 

(Medical College of South Carolina) 
C. M. Yonge, 1933 (University of Bristol) 
Charles Zeleny, 1906-1909 

(University of Indiana; later, 

University of Illinois) 






Albany, New York 


Lewis Boss, 1905-October 5, 1912 
Benjamin Boss, 1912-1936; Chairman, Committee on Meridian Astrometry, 1936-1938 

Staff Members 

Sebastian Albrecht, 1913-1937 Harry Raymond, 1905-1940 

Heroy Jenkins, 1909-1937 Arthur J. Roy, 1903-1936 

William B. Varnum, 1903-1936 




Organized in 1907, opened in 1908. Activities terminated January 1, 1946. 


Francis G. Benedict, 1907-1937 
Thorne M. Carpenter, 1938-1942 (Acting); 1943-1945 

V. Ooropatchinsky, 1923-1946 

Staff Members 

Robert C. Lee, 1929-1944 

Harold L. Higgins, 1908-1915 Walter R. Miles, 1914-1922 

H. Monmouth Smith, 1913-1920 

Other Scientists and Scholars Associated with the Department 

Henry P. Armsby, 1919-1920 

(Pennsylvania State College) 
James E. Ash, 1915 

(Harvard University Medical School; 

later, Army Medical Museum) 
Cornelia Golay Benedict, 

1918-1920, 1923, 1925-1926, 1929 
Edward H. Bensley, 1935 

(Montreal General Hospital) 
C. C. Benson, 1912, 1928-1929 

(University of Toronto) 
Alice F. Blood, 1917-1918 

(Simmons College) 
Samuel Brody, 1927 (University of Missouri) 
Ernest W. Brown, 1911 

(U. S. Navy Medical Corps) 
John M. Bruhn, 1932-1934 

(Yale Anthropoid Experiment Station, 

Orange Park, Florida; later, 

University of Alabama School of Medicine) 
M. Lucien Bull, 1914 (Institut Marey, Paris) 
Walter G. Cady, 1912-1913 

(Wesleyan University) 
E. P. Cathcart, Research Associate 

1912-1914 (University of Glasgow) 
Elizabeth E. Crofts, 1924 

(Mount Holyoke College) 
G. H. de Paula Souza, 1920 (Sao Paulo, Brazil) 
David B. Dill, 1935 (Harvard University) 
Raymond Dodge, 1912-1913 

(Wesleyan University; later, 

Yale University) 
Eugene F. Du Bois, 

1915, 1921, 1925-1927, 1930 

(Russell Sage Institute of Pathology; later, 

Cornell University Medical College) 

David L. Edsall, 1912 

(Washington University Medical School; 

later, Harvard University) 
H. T. Edwards, 1935 (Harvard University) 
W. Falta, 1909 

(First Medical Clinic, Vienna) 
Gertrude A. Farr, 1925-1929 

(University of New Hampshire) 
John M. Fuller, 1925-1926 

(New Hampshire Agricultural Experiment 

James L. Gamble, 1913 

(Harvard University Medical School) 
H. S. D. Garven, 1927-1932 

(Moukden Medical College, Manchuria) 
Florence Gustafson, 1925-1927 

(Wellesley College) 
Tom S. Hamilton, 1925 (University of 

Illinois Agricultural Experiment Station) 
C. S. Hicks, 1927-1930 

(University of Adelaide, South Australia) 
Fred A. Hitchcock, 1932 

(Ohio State University) 
John Homans, 1910-1912 

(Harvard University Medical School) 
Roy G. Hoskins, 1933 

(Harvard University Medical School; 

later, Tufts College) 
Elliott P. Joslin, 1909-1925, 1930, 1941-1943 

(New England Deaconess Hospital, Boston; 

later, Harvard University Medical School) 
Howard T. Karsner, 1914-1915 

(Harvard University Medical School; later, 

Bureau of Medicine and Surgery, 

Navy Department) 



Leslie G. Kilborn, 1927-1932 

(West China Union University; later, 

University of Hong Kong) 
Zing Yang Kuo, 1938 (Hangchow, China) 
Walter Landauer, 1931 

(Storrs Agricultural Experiment Station) 
Milton 0. Lee, 1935-1936 

(Harvard University Medical School) 
Helge Lundholm, 1929 

(McLean Hospital, Waverley, 

Massachusetts; later, Duke University) 
Grace MacLeod, 1922-1927 

(Teachers College, Columbia University) 
Eleanor D. Mason, 1927-1933 

(Women's Christian College, Madras) 
James H. Means, 1913-1915 

(Massachusetts General Hospital, Boston; 

later, Massachusetts Institute of 

Mary Henderson Meyer, 1931 

(Massachusetts Home, Boston) 
Carey D. Miller, 1928-1935 

(University of Hawaii Experiment Station) 
Sergius Morgulis, 1913 (Creigh ton University; 

later, University of Nebraska College of 

John R. Murlin, 1909 

(Cornell University Medical College; later, 

University of Rochester College of Medicine) 
Hans Murschhauser, Research Associate 

1914 (Diisseldorf, Germany) 
Julius Nitzulescu, 1928 

(Faculty of Medicine, Jassy, Roumania) 
Francis W. Peabody, 1915 

(Peter Bent Brigham Hospital, Boston) 
Josef M. Petrik, 1929 

(Masaryk University, Brno, 

Joseph H. Pratt, 1911-1913 

(New England Medical Center, Boston) 

E. G. Ritzman, Research Associate, 1933-1939 
(University of New Hampshire) 

F. W. Rolph, 1918 (University of Toronto) 
Howard F. Root, 1921-1927, 1930, 

1933-1934, 1936, 1939-1943 

(New England Deaconess Hospital, Boston) 

Paul Roth, 1911-1914, 1917-1921, 1923 

(Battle Creek Sanitarium, Michigan) 
George C. Shattuck, 1929-1930 

(Harvard University Medical School) 
Henry C. Sherman, 1934-1936 

(Columbia University) 
Hazeltine L. Stedman-Parmenter, 1925-1927 

(Mount Holyoke College) 
Nils Stenstrom, 1920 (Stockholm, Sweden) 

F. Strieck, 1928 

(University of Wiirzburg, Germany) 
Fritz B. Talbot, 1911-1922, 1924 

(Harvard University Medical School) 
Carl Tigerstedt, Research Associate, 1913-1914 

(University of Helsingfors) 
Harry C. Trimble, 1939-1940 

(Harvard University Medical School) 
Abby H. Turner, 1924, 1927-1929 

(Mount Holyoke College) 
E. C. van Leersum, 1920 

(Institute for Human Nutrition, 

H. S. Halcro Wardlaw, 1931 (Australia) 
Laurence G. Wesson, 1938 

(Veader Leonard Laboratory of 

Experimental Therapeutics, Baltimore; 

later, Massachusetts Institute of 

Paul Dudley White, 1937 

(Massachusetts General Hospital, Boston) 
Priscilla White, 1936, 1939 

(New England Deaconess Hospital, Boston) 
John C. Whitehorn, 1929 

(McLean Hospital, Waverley, 

Massachusetts; later, Johns Hopkins 

Francis H. Williams, 1912 

(Boston City Hospital) 

G. D. Williams, 1926-1927 
(Washington University Medical School) 

Stanley D. Wilson, 1930-1935 

(Yenching University, Peiping) 
Robert M. Yerkes, 1932-1934 

(Yale Anthropoid Experiment Station, 

Orange Park, Florida) 


Office of Administration 
Horace B. Barlow, 1961 (King's College, Cambridge University) 



Department of Plant Biology 

Herbert G. Baker, 1948-1949 

(University of California) 
Shao-lin Chen, 1949-1950 

(Red Star Yeast Company) 
Edwin A. Davis, 1949-1950 

(U. S. Department of Agriculture) 
L. N. M. Duysens, 1952-1953 

(University of Leiden) 
Fulton J. F. Fisher, 1956-1957 

(University of Melbourne) 
Joop C. Goedheer, 1957-1958 

(University of Utrecht) 
Bessel Kok, 1951-1952 

(Research Institute for Advanced Studies, 


Paul H. Latimer, 1956-1957 

(Auburn University) 
Josef E. Loeffler, 1954-1955 

(Shell Development Company) 
Fergus D. H. Macdowall, 1947-1949 

(Canadian National Research Council) 
Guy C. McLeod, 1959-1960 

(SIAS Institute, Brooks Hospital, 

Brookline, Massachusetts) 
Ruth Sager, 1961 (Columbia University) 
Jerome A. Schiff, 1962 (Brandeis University) 
Kazuo Shibata, 1956 

(Tokugawa Institute for Biological 

Hemming I. Virgin, 1954 

(University of Gothenburg) 

Mount Wilson and Palomar Observatories 

M. K. Vainu Bappu, 1951-1952 

(Astrophysical Observatory, Kodaikanal, 

Geoffrey R. Burbidge, 1955-1957 

(University of California, La Jolla) 
William A. Buscombe, 1950-1952 

(Mount Stromlo Observatory, 

Australian National University, 

Canberra, Australia) 
Edward R. Dyer, Jr., 1949-1950 

(National Academy of Sciences) 
Roger F. Griffin, 1960-1961 

(St. John's College, Cambridge University) 
Colin S. Gum, 1959-1960 

(Radiophysics Laborato^, Commonwealth 

Scientific and Industrial Research 

Organization, Sydney, Australia) 
Karl G. Henize, 1955-1957 

(Dearborn Observatory, Northwestern 


Leo Houziaux, 1960-1962 

(Institut d'Astrophysique, 

University of Liege) 
Thomas A. Matthews, 1956-1958 

(California Institute of Technology) 
Charles Robert O'Dell, 1962— 

(Mount Wilson and Palomar Observatories) 
George W. Preston, III, 1959-1961 

(Lick Observatory, Mount Hamilton) 
Alexander W. Rodgers, 1959-1960 

(Mount Stromlo Observatory, Australian 

National University, Canberra, Australia) 
John B. Rogerson, Jr., 1954-1956 

(Princeton University Observatory) 
Stewart L. Sharpless, 1952-1953 

(U. S. Naval Observatory) 
Carlos M. Varsavsky, 1959 

(Comision de Astrofisica y Radioastronomia, 

Buenos Aires) 
Merle F. Walker, 1952-1954 

(Lick Observatory, Mount Hamilton) 

Department of Terrestrial Magnetism 

Arthur I. Aronson, 1959-1960 

(Purdue University) 
Toshi Asada, 1960-1962 

(Geophysical Institute, Tokyo) 
Manuel N. Bass, 1958-1959 

(Northwestern University) 
Prabhat K. Bhattacharya, 1948-1950 

(California Institute of Technology) 
Louis Brown, 1961 — (University of Basel) 

Mateo Casaverde, 1948 

(Instituto Geofisico del Peru) 
William Compston, 1958 

(Australian National University) 
E. H. Creaser, 1955-1956 

(University of Cambridge) 
J. D. Duerksen, 1959-1960 

(National Institute for Medical Research, 




William C. Erickson, 1956-1957 

(Leiden Observatory) 
Gonzalo Fernandez, 1948-1949 

(Instituto Geofisico del Peru) 
George B. Field, 1953 

(Princeton University Observatory) 
J. W. Findlay, 1952 

(National Radio Astronomy Observatory, 

Green Bank) 
Kenneth L. Franklin, 1954-1956 

(Hayden Planetarium) 
John W. Graham, 1947-1949 

(Woods Hole Oceanographic Institution) 
Ronald Green, 1961-1962 

(University of Tasmania, Hobart) 
Richard Hall, 1962 — (Indiana University) 
Pembroke Jones Hart, 1952-1954 

(National Science Foundation) 
H. Lawrence Heifer, 1953-1957 

(University of Rochester) 
Ellis S. Kempner, 1958 

(National Institutes of Health) 
John J. Leahy, 1956-1957 

(City of Hope Hospital, California) 
Howard M. Lenhoff, 1958 

(Howard Hughes Medical Institute, Miami) 
Soren Lovtrup, 1951-1952 

(Carlsberg Laboratories, Copenhagen) 
John E. Midgley, 1960-1962 

(Oxford University) 
Thomas Murphy, 1947-1948 

(National University of Ireland) 

Jatinder Nath Nanda, 1949-1951 

(Indian Naval Physical Laboratory, 

New Delhi) 
Leif Owren, 1953-1954 

(Geophysical Institute, College, Alaska) 
W. D. Parkinson, 1947-1948 

(Bureau of Mineral Resources, 

Melbourne, Australia) 
Gerald C. Phillips, 1950-1952 (Rice Institute) 
George F. Pieper, 1956-1957 

(Applied Physics Laboratory, 

Johns Hopkins University) 
Hector Rojas, 1961-1962 

(Pan American College Observatory, 

Edinburgh, Texas) 
Hermann Rudin, 1962 — (University of Basel) 
Jorma J. Ruhimas, 1959-1960 

(University of Helsinki) 
George C. Sponsler, 1950 

(U. S. Department of the Navy) 
M. Sugiura, 1955 

(Geophysical Institute, College, Alaska) 
Harold Weaver, 1956-1957 

(Lick Observatory, Mount Hamilton) 
James A. Weinman, 1958-1960 

(University of Wisconsin) 
Dexter Whitehead, 1947-1948 

(University of Virginia) 
Francis Waverly Wood, 1949-1951 

(Bureau of Mineral Resources, 

Melbourne, Australia) 

Department of Genetics 

Guiseppe Bertani, 1948-1949 

(Karolinska Institutet, Stockholm) 
Katherine S. Brehme (Warren), 1939-1941 

(National Institutes of Health) 
Hugh J. Cairns, 1960-1961 

(Australian National University, Canberra) 
H. Clark Dalton, 1948-1950 

(Washington Square College, 

New York University) 
Berthe Delaporte, 1948-1949 

(Ecole Pratique des Hautes Etudes, Paris) 
A. H. Doermann, 1947-1949 

(Vanderbilt University) 
Kazuo Hashimoto, 1957-1958 

(Keio University School of Medicine, 

Etta Kiifer (Boothroyd), 1956-1957 

(McGill University) 

Joseph D. Mandell, 1955-1957 

(Palo Alto Medical Research Foundation) 
Hermann Moser, 1953-1956 

(Frances Delafield Hospital, 

New York City) 
Frank H. Mukai, 1959 

(Biological Laboratory, 

Long Island Biological Association) 
Kenneth Paigen, 1950-1952 

(Roswell Park Memorial Institute, Buffalo) 
Catherine Roesel, 1950-1951 

(University of Georgia School of Medicine) 
Janine Sechaud, 1960 (University of Oregon) 
Atif Sengiin, 1957 

(University of Istanbul, Turkey) 
Robert C. von Borstel, 1952-1953 

(Oak Ridge National Laboratory) 



Geophysical Laboratory 


Ralph Arnold, 1956-1959 

(Princeton University; later, 

Saskatchewan Research Council, 

University of Saskatchewan) 
D. Kenneth Bailey, 1962— 

(Trinity College, Dublin, Ireland) 
Hubert L. Barnes, 1956-1959 

(Columbia University; later, 

Pennsylvania State University) 
Robin Brett, 1961 — (Department of 

Geological Sciences, Harvard University) 
Charles W. Burnham, 1961 — 

(Massachusetts Institute of Technology) 
Peter R. Buseck, 1961 — 

(Department of Geology, Columbia 

G. A. Chinner, 1958-1960 

(University of Cambridge) 
John de Neufville, 1961-1962 

(Yale University; later, 

Harvard University) 
Bruce R. Doe, 1960-1962 

(California Institute of Technology; later, 

U. S. Geological Survey) 
W. Gary Ernst, 1955-1958 

(Johns Hopkins University; later, 

University of California, Los Angeles) 
Jeff J. Fawcett, 1961 — 

(University of Manchester) 

B. Halferdahl, 1954-1958 

(Johns Hopkins University; later, 

Research Council of Alberta, 

Edmonton, Alberta, Canada) 
Kai Hytonen, 1959-1961 

(University of Helsinki; later, 

Geological Survey of Finland, Otaniemi) 
Mackenzie L. Keith, 1947-1950 

(Pennsylvania State University) 
Donald H. Lindsley, 1960-1962 

(Johns Hopkins University) 
Giinter Moh, 1962 (Heidelberg University) 
Nobuo Morimoto, 1957-1959, 1962 

(Mineralogical Institute, 

University of Tokyo) 
Kaarlo J. Neuvonen, 1948-1950 

(Geological Survey of Finland; later, 

University of Turku, Finland) 

Louis Otto Nicolaysen, 1951-1954 

(Massachusetts Institute of Technology; 

later, Bernard Price Institute of 

Geophysical Research, Johannesburg, 

South Africa) 
Philip M. Orville, 1957-1958 

(Yale University; later, Cornell University) 
Edwin W. Roedder, 1947-1948 

(Columbia University; later, 

U. S. Geological Survey) 
Eugene H. Roseboom, 1956-1959 

(Harvard University; later, 

U. S. Geological Survey) 
Bruno Sabels, 1962 (University of Nevada) 
Th. G. Sahama, 1947-1949 

(University of Helsinki) 
Werner F. Schreyer, 1958-1959, 1962— 

(University of Kiel) 
James R. Smith, 1954-1957 

(Princeton University; later, 

Saskatchewan Research Council, 

University of Saskatchewan) 
Joseph Victor Smith, 1951-1954 

(Cavendish Laboratory, University of 

Cambridge; later, Pennsylvania State 

University and University of Chicago) 
Yoshio Suzuki, 1960-1962 

(Hakkaido University, Japan; later, 

Geological Survey of Japan) 
Per-Fredrick Troften, 1960 

(Norwegian Geological Survey; later, 

Geof ysisk Malmleting, Trondheim, Norway) 
Allan C. Turnock, 1958-1960 

(University of Manitoba; later, 

Department of Mines and Technical 

Surveys, Ottawa) 
J. R. Vallentyne, 1956-1957 

(Queen's University, Ontario; later, 

Cornell University) 
Bruce Velde, 1962— 

(Montana State University) 
David R. Wones, 1957-1959 

(Massachusetts Institute of Technology; 

later, U. S. Geological Survey) 
Kenzo Yagi, 1950-1951, 1960-1961 

(Tohoko University, Japan) 
Richard A. Yund, 1959-1961 

(University of Illinois; later, 

Brown University) 



Department of Embryology 

Michael Abercrombie, 1962 

(University College, London) 
Vittorio Danesino, 1953-1954 

(University of Naples) 
L. E. DeLanney, 1957 (Wabash College) 
Christine Gilbert, 1950-1951 

(University of the Witwatersrand) 
Perry W. Gilbert, 1949-1950 

(Cornell University) 
E. Clark Gillespie, 1948 

(Johns Hopkins University; later, 

University of Arkansas) 
Richard J. Goss, 1960-1961 

(Brown University) 
Jerome S. Harris, 1948-1949 

(Johns Hopkins University; later, 

private practice in obstetrics in Denver) 
Beni Horvath, 1952-1953 

(Columbia University; later, 

National Institutes of Health) 
Yoshihiro Kato, 1959-1961 

(Tokyo University; later, 

University of Nagoya) 

Efstathios J. Kokrikos, 1953-1954 

(Red Cross Hospital, Athens) 
Ben C. Moffett, Jr., 1954 

(University of Alabama; later, 

Armed Forces Institute of Pathology) 
Brenda Schofield, 1953-1954 

(Oxford University) 
E. Carl Sensenig, 1945 — 

(University of Alabama) 
Peter H. S. Silver, 1961-1962 

(Middlesex Hospital Medical School, 

Malcolm S. Steinberg, 1956-1958 

(Johns Hopkins University) 
Ikuo Takeuchi, 1959-1961 

(Princeton University; later, 

University of Osaka) 
L. J. Wells, 1948 (University of Minnesota) 
Douglas R. Wilkie, 1955 

(University of London) 
Ian B. Wilson, 1961-1962 

(University College of North Wales) 
Fred H. Wilt, 1958-1960 (Purdue University) 

Department of Archaeology 

Robert H. Barlow, 1949-1950 

(Mexico City College) 
Heinrich Berlin, 1952-1955 

(Instituto de Antropologia e Historia de 


Joseph A. Hester, Jr., 1952-1954 

(University of California, Los Angeles) 

William T. Sanders, 1954-1955 
(Pennsylvania State University) 

Raymond H. Thompson, 1950-1952 
(University of Arizona) 





Solomon F. Acree, 1904-1913 

(Johns Hopkins University; later, 

National Bureau of Standards) 
Charles Baskerville, 1903-1905 

(College of the City of New York) 
Gregory P. Baxter, Research Associate 

1904-1914, 1924 (Harvard University) 
Gustavus E. Behr, 1906 (Llarvard University) 
Amos P. Brown, 1904-1908 

(University of Pennsylvania) 

Paul B. Davis, Research Associate, 1916-1917 
(Johns Hopkins University; later, 
Davison Chemical Corporation, Baltimore) 

Louis M. Dennis, 1903 (Cornell University) 

Howard W. Doughty, 1904 

(Johns Hopkins University; later, 
Amherst College) 

George S. Forbes, 1906 (Harvard University) 

Joseph C. W. Frazer, 1916-1918 
(Johns Hopkins University) 



Moses Gomberg, 1904-1905 

(University of Michigan) 
Harry C. Jones, 1903-1916 

(Johns Hopkins University) 
George B. Kistiakowsky, Research Associate 

1942 (Harvard University) 
Philip A. Leigh ton, Research Associate 

1934-1935 (Stanford University) 
Harmon N. Morse, 1902-1918 

(Johns Hopkins University) 
Arthur A. Noyes, Research Associate 

1903-1930 (California Institute of 

I. I. Rabi, Research Associate, 1934-1935 

(Columbia University) 
Ira Remsen, 1902, 1913, 1917 

(Johns Hopkins University) 

Theodore W. Richards, Research Associate 

1902-1928 (Harvard University) 
Edgar Fahs Smith, Research Associate 

1902, 1909, 1916-1918, 1920-1922 

(University of Pennsylvania) 
Julius Stieglitz, 1909 (University of Chicago) 
Wilfred N. Stull, 1903 (Harvard University) 
James B. Sumner, Research Associate in 

Biochemistry, 1931-1932 

(Cornell University) 
John Bishop Tingle, 1903-1905 

(Johns Hopkins University) 
Harold C. Urey, Research Associate, 1934-1935 

(Columbia University; later, 

University of Chicago) 
Hobart H. Willard, 1910 (Harvard University) 
Edgar B. Wilson, Research Associate 

1936-1937 (Harvard University) 


Joseph S. Ames, 1904-1905 

(Johns Hopkins University) 
Carl D. Anderson, 1942-1943 

(California Institute of Technology) 
G. F. Barker, 1904 (Washington, D. C.) 
Samuel J. Barnett, Research Associate 

1904-1905 (Stanford University; later, 

University of California, Los Angeles, and 

California Institute of Technology) 
Ralph D. Bennett, Research Associate 

1932-1933 (Massachusetts Institute 

of Technology; later, Naval Ordnance 

Charles F. Burgess, 1904-1908 

(University of Wisconsin) 
William Campbell, 1904-1905 

(Columbia University) 
Henry S. Carhart, 1904-1905 

(University of Michigan) 
Clement D. Child, 1903-1904 

(Colgate University) 
William W. Coblentz, 1903-1908, 1911 

(National Bureau of Standards) 
Henry Crew, 1902-1904 

(Northwestern University) 
Paul S. Epstein, 1937-1939 

(California Institute of Technology) 
J. A. Folse, 1926 

(Rosenwald Industrial Museum, Chicago) 
William S. Franklin, 1906 

(Lehigh University; later, 

Massachusetts Institute of Technology) 

L. A. Freudenberger, 1906 (Delaware College) 
Robert H. Goddard, 1929-1930 

(Clark University) 
John F. Hay ford, Research Associate 

1911-1913, 1915-1917, 1919-1925 

(Northwestern University) 
Henry M. Howe, 

1906-1911, 1913-1914, 1916-1920 

(Columbia University) 
H. Victor Neher, 1943 

(California Institute of Technology) 
Edward L. Nichols, Research Associate 

1905-1906, 1908-1918, 1920-1925 

(Cornell University) 
Francis E. Nipher, 1914 

(Washington University) 
Gennady W. Potapenko, 1937-1939 

(California Institute of Technology) 
Allen G. Shenstone, Research Associate 

1931-1933 (Princeton University) 
William W. Strong, 1908-1911 

(Johns Hopkins University; later, 

Scientific Instrument and Electrical 

Machine Company) 
Horace S. Uhler, 1905 (Johns Hopkins 

University; later, Yale University) 
John B. Whitehead, 1903-1905 

(Johns Hopkins University) 
Robert W. Wood, 1902-1904 

(Johns Hopkins University) 
Albert F. Zahm, 1905 

(Catholic University of America; later, 

Library of Congress) 




Arthur B. Coble, 1903-1904 

(University of Missouri; later, 

University of Illinois) 
Floyd F. Decker, 1910 (Syracuse University) 
Leonard E. Dickson, Research Associate 

1904, 1912, 1919, 1922, 1927-1928 

(University of Chicago) 
George W. Hill, 1905-1907 

(West Nyack, New York) 
John Holland, Research Associate, 1960-1961 

(University of Michigan) 
Derrick N. Lehmer, Research Associate 

1904-1909, 1911-1912, 1925-1928, 

1931-1932, 1936 (University of California) 
Arthur C. Lunn, 1909 (University of Chicago) 
William D. MacMillan, 1909 

(University of Chicago) 

Eliakim H. Moore, 1902 

(University of Wisconsin; later, 

University of Chicago) 
Frank Morley, Research Associate 

1902, 1908, 1910-1918, 1920-1921, 1923, 

1926, 1928, 1930-1931, 1933-1936 

(Johns Hopkins University) 
James B. Shaw, 1907 

(James Millikin University; later, 

University of Illinois) 
Henry W. Stager, 1911 (Fresno, California) 
Ormond Stone, 1902 (Leander McCormick 

Observatory, Charlottesville, Virginia) 
Ernest J. Wilczynski, Research Associate 

1903-1905 (University of California; later, 

University of Chicago) 


William H. Burr, 1902 
(Columbia University) 

William F. Durand, 1903-1906 
(Cornell University; later, 
Stanford University) 

George Gibbs, 1902 

(Baldwin Locomotive Works, Philadelphia; 
later, consulting engineer, Pennsylvania 

William F. M. Goss, 1904-1908 
(University of Illinois) 

George S. Morison, 1902 

(civil engineer, New York City) 
Harold Pender, 1902-1903 

(Syracuse University; later, 

University of Pennsylvania) 
Charles P. Steinmetz, 1902 

(General Electric Company) 
Robert H. Thurston, 1902 

(Cornell University) 
Leonard Waldo, 1903 (consulting engineer in 

metallurgy and electronics, 

Plainfield, New Jersey) 

Geography, Geology, and Geophysics 

Cleveland Abbe, 1902 (U. S. Weather Bureau; 

later, Johns Hopkins University) 
Adalbert E. Benfield, Research Associate 

1940-1941 (Williams College; later, 

Harvard University) 
Tor Bergeron, 1951-1957 

(University of California; later, 

Meteorological Institute, Uppsala, Sweden) 
J. Bjerknes, 1951-1957 

(University of California) 
V. Bjerknes, Research Associate in 

Meteorology, 1906-1948 

(University of Oslo) 
Eliot Blackwelder, 1903-1904 

(University of Wisconsin; later, 

Stanford University) 

Robert C. Bundgaard, 1951-1957 

(U. S. Air Force) 
Ian Campbell, Research Associate, 1933-1939 

(California Institute of Technology) 
Rollin T. Chamberlin, 1908 

(University of Chicago) 
George Davidson, 1906-1907 

(University of California) 
William Morris Davis, 1902, 1925-1926 

(Harvard University) 
C. L. Godske, 1951-1957 

(University of Bergen) 
Frank T. Gucker, Jr., Research Associate 

1940-1950 (Northwestern University; 

later, Indiana University) 
Norman E. A. Hinds, 1931, 1933-1935 

(University of California, Berkeley) 



William H. Hobbs, Research Associate, 

(University of Michigan) 
John H. Maxson, Research Associate 

1932-1939 (California Institute of 

Technology; later, Anderson-Pritchard 

Oil Corporation, Denver) 
Walter H. Newhouse, Research Associate 

1939-1945 (Massachusetts Institute of 

Technology; later, 

University of Chicago) 

1930 Sverre Petterssen, 1951-1957 

(University of Chicago) 
J. W. Sandstrom, 1906-1908 

(University of Stockholm) 
Alexander Silverman, Research Associate 

1939-1942 (University of Pittsburgh) 
H. Solberg, 1951-1957 (University of Oslo) 
William Van Roy en, Research Associate, 1934 

(University of Nebraska; later, 

Brooklyn College) 


Oscar S. Adams, 1925-1926 

(U. S. Coast and Geodetic Survey) 
F. B. Bassett, 1923 (U. S. Navy Department) 
George L. Bean, 1928-1931 

(U. S. Coast and Geodetic Survey) 
Hugo Benioff, 1932-1936 

(California Institute of Technology) 
William Bowie, 1925-1926 

(U. S. Coast and Geodetic Survey) 
Perry Byerly, 1925-1926 

(University of California, Berkeley) 
Charles Lewis Gazin, 1931-1932 

(U. S. Geological Survey; later, 

U. S. National Museum) 
Herbert E. Gregory, 1925-1926 

(Yale University; later, 

Bishop Museum, Honolulu) 
Beno Gutenberg, 1930-1931, 1933-1935 

(California Institute of Technology) 
William Stephen Webster Kew, 1922-1923 

(U. S. Geological Survey; later, 

Standard Oil Company of California) 
Andrew C. Lawson, 1906-1907 

(University of California, Berkeley) 
James B. Macelwane, S.J., 1924-1925 

(St. Louis University) 

Levi F. Noble, 1922-1923 

(U. S. Geological Survey) 
Harry Fielding Reid, 1906-1907 

(Johns Hopkins University) 
Charles F. Richter, 1927-1928, 1932-1937 

(California Institute of Technology) 
Arnold Romberg, 1921-1923 

(University of Hawaii; later, 

University of Texas) 
Maple D. Shappell, 1930-1934 

(California Institute of Technology) 
Frederick P. Vickery, 1922-1923 

(University of Southern California, 

Los Angeles; later, 

Sacramento Junior College) 
Frank Wenner, 1922-1923 

(National Bureau of Standards) 
Walter T. Whitney, 1913, 1917, 1922-1923 

(California Institute of Technology; later, 

Pomona College) 
Bailey Willis, Research Associate 

1903-1907, 1912, 1930, 1934 

(Stanford University) 
Harry O. Wood, Research Associate 

1920-1931, 1936-1940 

(California Institute of Technology) 

Physiological Chemistry 

John J. Abel, 1903-1905 

(Johns Hopkins University) 
Wilder D. Bancroft, 

1902, 1904-1906, 1908-1910 

(Cornell University) 
Russell H. Chittenden, 1904-1907 

(Yale University) 
Walter H. Eddy, Research Associate 

1927-1933 (Columbia University) 

Lafayette B. Mendel, 1905-1906, 1927-1930 

(Yale University) 
Thomas B. Osborne, Research Associate 

1904-1927 (Connecticut Agricultural 

Experiment Station) 
Hubert B. Vickery, Research Associate 

1922-1937 (Connecticut Agricultural 

Experiment Station) 
Robert R. Williams, Research Associate 

1927-1933 (Bell Telephone Laboratories) 




John W. Baird, 1903-1904 
(Cornell University; later, 
Clark University) 

James Mark Baldwin, 1902 
(Princeton University) 

Clarence B. Farrar, 1904-1906 

(Shepperd and Enoch Pratt Hospital, 

Baltimore; later, 

Toronto Psychiatric Hospital) 

Shephard I. Franz, 1903-1911, 

1913, 1915-1917 (St. Elizabeth's Hospital, 

Washington, D. C.) 
S. Stanley Hall, 1903-1904 (Clark University) 
Peter Milner, Research Associate, 1960-1961 

(McGill University) 
James P. Porter, 1907 

(Clark University; later, Ohio University) 
Henry A. Ruger, Research Associate 

1927-1929 (Columbia University) 


Wilbur O. Atwater, 1903-1905 

(Wesleyan University) 
Henry P. Bowditch, 1902 

(Harvard Medical School) 
Simon Flexner, 1902-1903 

(University of Pennsylvania; later, 

Rockefeller Institute for Medical Research) 
Alexander Forbes, 1906 (Harvard University) 
Robert H. Gault, Research Associate 

1927-1929 (Northwestern University) 
Charles C. Guthrie, 1908 

(Washington University; later, 

University of Pittsburgh) 
Frank A. Hartman, Research Associate 

1931-1934 (Ohio State University) 
William H. Howell, Research Associate 

1902, 1933-1934 (Johns Hopkins University) 

Leo Loeb, 1903-1905, 1907-1909 

(University of Pennsylvania; later, 

Washington University) 
S. Weir Mitchell, 1902 

(Philadelphia, Pennsylvania) 
Aubrey T. Mussen, Research Associate 

1929-1931 (Johns Hopkins University) 
Hideyo Noguchi, 1903-1908 

(University of Pennsylvania) 
Earle B. Phelps, Research Associate, 1931-1933 

(Columbia University) 
Edward T. Reichert, 1904, 1908-1914 

(University of Pennsylvania) 
George Oscar Russell, Research Associate 

(Ohio State University; later, 

Gallaudet College) 


Anton J. Carlson, 1903-1904 

(Stanford University; later, 

University of Chicago) 
A. B. Clawson, 1907 (University of Michigan) 
Henry E. Crampton, Research Associate 

1902, 1904, 1906, 1908, 1916, 1919, 

1923-1925, 1927-1928, 1930-1933, 1935, 

1939 (Columbia University) 
Joseph A. Cushman, 1912, 1919, 1939-1941 

(Cushman Laboratory for Foraminiferal 

Research, Sharon, Massachusetts) 
Bashford Dean, 1906 (Columbia University) 
Carl H. Eigenmann, 1903-1904 

(Indiana University) 
Ross G. Harrison, Research Associate 

1944-1948 (Yale University) 
Leland O. Howard, 1903-1904 

(U. S. Department of Agriculture) 

Herbert S. Jennings, 1902-1905 

(University of Michigan; later, 

Johns Hopkins University) 
Charles A. Kofoid, Research Associate 

1921-1925 (University of California) 
Ralph S. Lillie, 1904 (University of Nebraska; 

later, University of Chicago) 
Joseph A. Long, 1911 (Harvard University) 
Clarence E. McClung, 1903-1905 

(University of Kansas; later, 

University of Penns} 7 lvania) 
Hansford MacCurdy, 1907 

(Harvard University) 
Edward L. Mark, 1906-1910 

(Harvard University) 
C. Hart Merriam, 1902 

(U. S. Biological Survey) 
Albert P. Morse, 1903-1905 

(Wellesley, Massachusetts) 



Henry Fairfield Osborn, 1902 

(Columbia University) 
William Patten, 1904-1905 

(Dartmouth College) 
Raymond Pearl, 1904-1906 

(University of Michigan; later, 

Johns Hopkins University Medical School) 
John C. Phillips, 1911 

(Harvard University) 

Porter E. Sargent, 1904 
(Harvard University; later, 
Sargent School Service) 

Nettie M. Stevens, 1904-1905 
(Bryn Mawr College) 

Edmund B. Wilson, 1903 
(Columbia University) 

Naohide Yatsu, 1905-1906 
(Columbia University) 


Ira S. Allison, 1939-1944 

(Oregon State College) 
Earl H. Bell, Research Associate, 1934-1935 

(University of Nebraska; later, 

U. S. Department of Agriculture) 
John P. Buwalda, Research Associate 

1925-1938 (California Institute of 

Frank M. Carpenter, Research Associate 

1931-1932 (Harvard University) 
Ermine C. Case, Research Associate 

1903-1905, 1908-1909, 1911-1912, 

1914-1919, 1921-1922 

(University of Michigan) 
Carlton Condit, 1938, 1944 

(University of California; later, 

Illinois State Museum, Springfield) 
Lyman H. Daugherty, 1941 

(San Jose State College) 
Hellmut De Terra, Research Associate 

1934-1939 (Yale University; later, 

Columbia University) 
A. L. Du Toit, Research Associate, 1923 

(Pretoria, South Africa) 
Eustace L. Furlong, 

1921-1924, 1928, 1931-1933, 1938-1942 

(University of California) 
William K. Gregory, 1938 

(American Museum of Natural History; 

later, Columbia University) 
Oliver P. Hay, Research Associate 

1902-1907, 1911-1927 

(American Museum of Natural History) 
Norman E. A. Hinds, 1936 

(University of California) 
Edgar B. Howard, Research Associate 

1934-1942 (University of Pennsylvania) 
Hildegarde Howard, 

1932, 1938, 1939, 1942, 1946, 1949 

(Los Angeles County Museum) 
Hsen Hsu Hu, 1940 

(Fan Memorial Institute of Biology, 


Remington Kellogg, 1925-1942 

(U. S. National Museum) 
Robert Smith La Motte, 1935-1936 

(U. S. Forest Service; later, 

University of California) 
Harry D. MacGinitie, 1933, 1937, 1941, 1953 

(Humboldt State College) 
Edwin D. McKee, Research Associate 

1936-1942 (U. S. National Park Service; 

later, U. S. Geological Survey) 
Earl L. Packard, 

1926, 1928, 1931-1932, 1938-1939, 

1941-1943 (Oregon State College) 
Llewellyn I. Price, Research Associate 

1939-1940 (Harvard University) 
Malcolm J. Rogers, Research Associate, 1937 

(San Diego Museum) 
Paul B. Sears, 1936, 1938 

(University of Oklahoma; later, 

Yale University and Wake Forest College, 

Winston-Salem, North Carolina) 
Chester Stock, Research Associate, 1925-1943 

(California Institute of Technology) 
Alexander A. Stoyanow, Research Associate 

1928 (University of Arizona; later, 

University of California) 
G. H. R. von Koenigswald, Research Associate 

1936-1938, 1947 (Bandung, Java) 
David White, Research Associate, 1925-1932 

(National Academy of Sciences) 
Henry S. Williams, 1902 (Yale University) 
Howell Williams, 1943-1944 

(University of California) 
Samuel W. Williston, 1903 

(University of Chicago) 
Robert W. Wilson, 

1933-1934, 1936-1937, 1940, 1942, 1949 

(California Institute of Technology; later, 

University of Colorado and 

University of Kansas) 
Wendell P. Woodring, 1925, 1928, 1932 

(California Institute of Technology; later, 

U. S. Geological Survey) 



Archaeology and Anthropology 

Marion E. Blake, Research Associate 

1937-1938, 1940-1941, 1945 

(American Academy, Rome) 
Franz Boas, 1902 

(American Museum of Natural History; 

later, Columbia University) 
William T. Brigham, 1906-1912 

(Bernice Pauahi Bishop 

Museum, Honolulu) 
George A. Dorsey, 1902 

(Field Museum of Natural History, 

Arthur L. Frothingham, 1913 

(Princeton University) 
William H. Holmes, 1902-1904 

(U. S. National Museum) 
Walter W. Hyde, Research Associate 

1919-1920 (University of Pennsylvania) 
Waldemar Jochelson, 

1923-1924, 1926, 1929-1930, 1934-1935 

(American Museum of Natural History) 
Allan C. Johnson, 1910-1911 

(Princeton University) 

W. Max Muller, 1904-1907, 1910-1911 

(University of Pennsylvania) 
Raphael Pumpelly, 1903-1906 

(Newport, Rhode Island) 
Hubert Schmidt, 1903-1904 

(Museum fur Volkerkunde, Berlin) 
George W. Scott, 1904-1905, 1911-1914 

(Law Librarian of Congress and 

Supreme Court) 
Thomas D. Seymour, 1903 

(American School of Classical Studies, 

Esther Boise Van Deman, Research Associate 

1906-1925 (American School of Classical 

Studies, Rome; later, 

University of Michigan) 
William Hayes Ward, Research Associate 

1903-1908 (editor, The Independent, 

New York) 
Andrew F. West, 1905-1911 

(Princeton University) 
James R. Wheeler, 1905-1912 

(American School of Classical Studies, 

Athens; later, Columbia University) 


Cyrus Adler, 1902 (Smithsonian Institution) 
J. McKeen Cattell, 1902-1904 

(Columbia University; later, 

editor, American Men of Science) 
Wilberforce Eames, 1906-1908 

(librarian, Bibliographical Society of 


Fielding H. Harrison, Research Associate 

1903-1927 (Army Medical Museum; later, 

Surgeon General's Office) 
Herbert Putnam, 1902-1907 

(Library of Congress) 
J. David Thompson, 1903-1907 

(Library of Congress) 

Literature, Linguistics, and Philology 

Manuel J. Andrade, Research Associate 

1933-1940 (University of Chicago) 
John Pawley Bate, 1910-1917 

(Inns of Court, London) 
Henry Bergen, Research Associate 

1912-1927, 1933 (Brooklyn, New York) 
J. Leslie Brierly, 1910-1911 

(Lincoln's Inn, and Trinity College, Oxford) 
Morgan Callaway, 1913 

(Johns Hopkins University) 
William Churchill, Research Associate 

1911, 1915-1919, 1921 (Committee on 

Public Information, Washington, D. C.) 
Lane Cooper, 1916 (College of St. James) 
Albert G. de Lapradelle, 1910-1916 

(University of Paris) 

Charles G. Fenwick, 1910-1916 

(Carnegie Endowment for International 

Peace; later, Bryn Mawr College) 
Ewald Flugel, 1904-1908 

(Stanford University) 
George D. Gregory, 1916 

(Carnegie Endowment for International 

John W. Hebel, 1917 (Cornell University) 
George Hempl, 1904-1905, 1909 

(University of Michigan; later, 

Stanford University) 
Charles W. Hodell, Research Associate 

1907-1908 (Goucher College) 
Thomas Erskine Holland, 1910-1911 

(LTniversity of Oxford) 



Arthur G. Kennedy, 1916-1918, 1920, 1923 

(Stanford University) 
Henry C. Lancaster, Research Associate, 1912 

(Johns Hopkins University) 
Elias A. Lowe, Research Associate 

1910-1929, 1936-1940, 1947 

(Institute for Advanced Study, 

Princeton University) 
John D. Maguire, 1909-1913 

(Catholic University of America) 
Ernest Nys, 1917 (University of Brussels) 
Charles G. Osgood, 1915 

(Princeton University) 
James Brown Scott, Research Associate 

1910-1917 (Department of State; later, 

Carnegie Endowment for International 


E. W. Scripture, 1903-1906 (Yale University) 
H. Oskar Sommer, Research Associate 

1906-1907, 1909-1912 ("Astolat," 

Camberley, Surrey, England) 
Benjamin F. Stelter, 1916 

(University of Southern California; later, 

Occidental College) 
John S. P. Tatlock, Research Associate 

1916-1918, 1920, 1923 (Stanford University; 

later, Harvard University and University of 

Ludwig von Bar, 1910-1916 

(University of Gottingen) 
John Westlake, 1910-1912 

(University of Cambridge) 
Herbert Francis Wright, 1917 

(Catholic University of America; later, 

Georgetown University) 

Political Science 

N. Andrew N. Cleven, Research Associate 

1930-1931 (Duke University) 
Isaac J. Cox, Research Associate, 1925-1927 

(Northwestern University) 
Herman G. James, Research Associate 

1922-1923 (University of Texas; later, 

President, Ohio University) 

Percy A. Martin, Research Associate, 
1926-1928 (Stanford University) 

William W. Pierson, Jr., Research Associate 
1927-1928 (University of North Carolina) 

Leo S. Rowe, 1904, 1906-1910, 1917 
(University of Pennsylvania) 

Graham H. Stuart, 1924 (Stanford University) 

Reports of Departments 
and Special Studies 

Mount Wilson and Palomar Observatories 

Geophysical Laboratory 

Department of Terrestrial Magnetism 

Committee on Image Tubes for Telescopes 

Department of Plant Biology 

Department of Embryology 

Department of Genetics 

Mount Wilson and Palomar 


Operated by Carnegie Institution of Washington 
and California Institute of Technology 

Pasadena, California 

Ira S. Bowen 

Horace W. Babcock 
Assistant Director 


Ira S. Bowen, 

Carl D. Anderson 

Horace W. Babcock 

Jesse L. Greenstein 

Robert B. Leighton 

Allan R. Sandage 


Introduction 5 

Observing Conditions 6 

Solar Observations 6 

Solar magnetic fields 7 

Forbidden nitrogen lines in the 

solar spectrum 7 

Planets 7 

Comets 8 

Stellar Spectroscopy and Photometry . 8 
Chemical composition of stellar 

atmospheres 8 

Line blanketing ....... 11 

Color-magnitude and chemical- 
composition relationships . . .12 

Color-spectral-type relationships . . 13 

Photometry of stellar clusters and 

associations 14 

Photometry of double stars . . .15 
Photometry of variable stars . . .16 
Photometry of the Giclas proper 

motion catalogue . . . . . . 17 

Subdwarfs 17 

White dwarfs .18 

Faint blue stars 18 

Balmer lines in early-type stars . . 18 

RR Lyrae variables 19 

Supernovae 19 

U Geminorum stars (dwarf novae) . 20 

Old novae 20 

Shell stars 20 

Mass loss from stars with extended 

atmospheres 21 

Segregation of elements in magnetic 

stars 21 

Radial velocities of magnetic stars . 22 

Stellar polarization 23 

Gaseous Nebulae and Interstellar gas . 24 

Galaxies 25 

Structure and internal motions of the 

Galaxy 25 

Rotation and internal motions of 

galaxies 26 

Emission nebulae in galaxies ... 27 

Variable stars in galaxies .... 28 

Photometry and stellar content and 

evolution 30 

Catalogue of galaxies and clusters 

of galaxies 32 

Internal motions of clusters of galaxies 33 

Redshift-magnitude relations ... 33 

Radio Sources 34 

Theoretical Studies 35 

Stellar atmospheres 35 

Star formation 37 

Stellar dynamics 37 

Cosmology 37 

Miscellaneous 38 

Instrumentation 39 

Guest Investigators 39 

Staff and Organization 45 

Bibliography 47 

Carnegie Institution of Washington Year Book 61, 1961-1962 


In 1904 George E. Hale, acting under the space age astronomy, was sponsored by 

auspices of the National Academy of the Douglas Aircraft Company and was 

Sciences, invited the scientific academies held at the California Institute of Tech- 

and the astronomical and physical socie- nology on August 7, 8, and 9. It was 

ties of a number of countries to send attended by about 100 engineers and 

representatives to a meeting to be held in astronomers. 

connection with the International Con- Because of the interest in the large 

gress of Science at the St. Louis Exhi- telescopes at Mount Wilson and Palomar 

bition for the purpose of establishing Mountain, arrangements were made by 

"co-operation among individuals and the Observatories to provide transporta- 

institutions engaged in Solar Research." tion from Los Angeles and entertainment 

This meeting resulted in the formation of on the mountains for the foreign delegates 

the International Union for Co-operation to the Assembly of the Union. Trips to 

in Solar Research, which held later Mount Wilson were scheduled on the 

meetings at Oxford (1905), Meudon afternoons of August 11 and 25, and to 

(1907), Mount Wilson (1910), and Bonn Palomar on August 12 and 26. About 275 

(1913). The Mount Wilson meeting was delegates took advantage of this oppor- 

attended by about 80 members of the tunity to visit the facilities on the 

Union and invited guests. mountains. 

After World War I, the Union was Nearly all members of the staff of the 

reorganized on a broader basis to include Observatories attended the Assembly at 

all branches of astronomy and its name Berkeley and participated in the sessions 

was changed to the International Astro- of the various commissions of which they 

nomical Union. Assemblies of the Union were members. 

were held at Rome (1922), Cambridge, Throughout the history of the Observa- 

England (1925), Leiden (1928), Cam- tories the major emphasis has been placed 

bridge, Massachusetts (1932), Paris on observations of the sun, stars, nebulae, 

(1935), and Stockholm (1938). After a and galaxies. From time to time, however, 

ten-year intermission caused by the when the Observatories' equipment was 

second World War, meetings occurred at suitable, attention has been given to 

Zurich (1948), Rome (1952), Dublin observations of planets and satellites. 

(1955), and Moscow (1958). The next For example, satellites X, XI, and XII of 

General Assembly in 1961 was planned Jupiter and the very unusual asteroids 

for the United States of America, and it Icarus and Geographos were discovered 

was hoped that it might be held in at the Observatories. High-dispersion 

Pasadena. However, a survey of the hotel spectroscopic studies of Venus and Mars 

situation indicated that to accommodate by Adams and Dunham provide the basis 

locally the more than 1000 members and for the current knowledge of the compo- 

guests who have attended these meetings sition of their atmospheres. Infrared 

in recent years would be impossible. The observations by Nicholson and Pettit of 

1961 General Assembly of the Inter- the lunar surface during an eclipse led to 

national Astronomical Union was there- the concept of a surface covered with 

fore held in Berkeley, between August 15 dust. Recently these infrared lunar 

and 24, the University of California observations were refined by Dr. Shorthill 

acting as host institution. and Mr. Saari of the Boeing Aircraft 

Several international symposia took Company, using the 60-inch on Mount 

place just before or after the Berkeley Wilson. In 1958 and 1960 the 200-inch 

meeting. One of these, on the subject of was used by Dr. Sinton of the Lowell 



Observatory to map the areas on Mars 
that show the absorption bands near 3.4 y. 
which are attributed to organic molecules. 

The development in the last few years 
of rockets capable of going to the neigh- 
borhoods of the moon and the inner 
planets has focused attention on lunar 
and planetary problems. Because of the 
much lower effort and cost required for 
ground-based observations compared 
with observations made from rockets, it 
has become important to push these solar 
system observations to the limits made 
possible with the new photometric and 
infrared techniques developed in recent 

In the past year G. Munch, with the 
assistance of Mr. Robert Younkin of the 
Jet Propulsion Laboratory, has used the 
Cassegrain spectrum scanner to investi- 
gate the monochromatic albedo and the 
total intensity of the absorption bands in 
the spectra of the major planets; Munch 

has also obtained high-dispersion spectra 
of Jupiter, Saturn, and Neptune, which 
have been studied in collaboration with 
Dr. Hyron Spinrad of the same Labora- 
tory. Lines of the hydrogen molecule at 
X6367.80 and X6435.03 were found in the 
spectrum of Saturn, providing the first 
definite evidence for the presence of 
hydrogen in its atmosphere. Dr. Spinrad, 
analyzing the high-dispersion spectra of 
Venus available in the files of the Ob- 
servatories, has found evidence for large 
changes in the apparent temperature of 
the atmosphere. Dr. Bruce Murray of the 
Lunar Research Laboratory at the Cali- 
fornia Institute of Technology has con- 
tinued the studies of the photoelectric 
colorimetry of the moon with the 60-inch 
telescope. He has also developed and 
tested on Mount Wilson a special 20-inch 
infrared telescope which will be used for 
lunar studies at an altitude of 13,000 feet 
on White Mountain. 


After three years in which the rainfall 
on Mount Wilson averaged less than 45 
per cent of normal, the precipitation for 
1961-1962 jumped to 46.14 inches, or 
within an inch of the total for the 
preceding three years. This increased 

rainfall came just in time to avoid a 
serious water shortage on the mountains. 
Solar observations were made on 311 
days; the 200-inch was in use on 287 
nights, the 100-inch on 292 nights, and 
the 60-inch on 265 nights. 


Solar observations were made by 
Cragg, Hickox, and Utter. The numbers 
of photographs of the various kinds taken 
between July 1, 1961, and June 30, 1962, 
were as follows : 

Direct photographs 302 

Ha spectroheliograms, 18-foot focus 270 

K2 spectroheliograms, 18-foot focus 258 

K2 prominences, 18-foot focus 84 
Number of days on which 

magnetograms were obtained 223 

Effective September 1, 1961, a basic 
change was made in the method of 
reduction of the sunspot magnetic data. 
In the past, sunspot groups were num- 
bered and a list of these groups along 

with average magnetic classifications and 
dates of central meridian passage was 
published, or otherwise made available, 
every year or so. More recently, investi- 
gators interested in this information have 
wanted it available rapidly, and they 
have frequently been interested in the 
appearance of the sun on some particular 
day. To meet these needs, and also to 
save time in the reduction process, a list 
is now prepared each month giving daily 
positions and magnetic classifications of 
spot groups for which there are magnetic 
measures. Copies of this list are sent each 
month to interested investigators. As the 
routine observing program, except for the 
magnetograms, has a low priority, the 


sunspot information is not as complete as 
in previous years. Sunspot groups are no 
longer numbered, and no attempt is made 
to keep track of returns. The K2 promi- 
nence patrol was also ended in September. 
Because spot magnetic polarities are 
not observed as often as in the past, and 
because the method of reduction has been 
altered, the tables of sunspot groups and 
classification usually published in the 
Annual Report will be discontinued. 
Magnetic classifications of spot groups 
were made on 161 days from July 1, 1961, 
to June 30, 1962. 

Solar Magnetic Fields 

Howard has completed a preliminary 
study of solar magnetograph observations 
made with very small apertures. Mag- 
netic traces with an aperture about 2 
seconds of arc on a side show root-mean- 
square fluctuations of 8.2 ± 4.4 gauss. 
The autocorrelation function derived 
from these observations shows maxima 
near 16,000 km and 40,000 km and in 
general resembles the autocorrelation 
function that Rogerson derived from 
intensity fluctuations on calcium spectro- 
heliograms. Similar observations made 
recording line-of-sight velocities yield 
root-mean-square fluctuations of 0.39 
db 0.14 km/sec. Observations made 
recording velocities with an aperture held 
fixed show autocorrelation curves that 
are damped cosine curves, indicating the 
presence of oscillations in the solar 
atmosphere. The period observed is 296 
seconds. The spectrum line used for all 
these observations was X5250.218, Fe I. 
Instrumental improvements since these 
observations were completed in the 
summer of 1959 enable us now to make 
much better observations of this type. 

Further observations are planned for the 
near future. 

The daily solar magnetograms, started 
in 1957, constitute a unique series of 
observations giving valuable information 
about daily configurations of the solar 
magnetic fields. Howard has begun an 
extensive study of these records, which 
will include classification of magnetic 
regions and their correlation with optical 
and radio phenomena. The investigation 
starts with the magnetograms from 
August 1959, when the new slant-line 
registration was begun. One interesting 
result that has appeared at this stage of 
the investigation concerns the UM regions 
first discovered by H. W. and H. D. 
Babcock. A large number of UM regions 
have been identified; invariably, at the 
position of the UM region the calcium 
(K2) spectroheliogram for that day shows 
mottlings somewhat brighter than the 
ordinary background. Thus it may be 
possible to detect UM regions over a 
period of fifty years or more using the 
extensive solar plate collection. 

Forbidden Nitrogen Lines in the Solar 

Starting from Vitense's model of the 
solar atmosphere and from a recent 
determination of the abundance of nitro- 
gen by Neven, the intensities of the [N I] 
X 10397 and X 10407 lines have been 
computed by Houziaux. From a com- 
parison of these results with the intensity 
of the weak feature observed at X 10397 
on several spectrograms recently obtained 
by Migeotte at the Jungfraujoch it can 
be concluded that the forbidden doublet 
2 D 5 /2- 2 Pi/2, 3 /2 is present in the infrared 
solar spectrum. The 2 D 3/ 2 — 2 Pi /2, 3/2 
doublet is hidden by a strong line of 
unknown origin. 


An observing program has been started atmospheres. One aspect of the program 

by G. Munch to serve as a basis for a is concerned with the determination of 

reexamination of problems related to the the energy distributions in their spectra 

structure and composition of planetary by photoelectric scanning. The mono- 



chromatic albedos of the planets and the 
total intensity of their absorption bands 
will thus be derived. Observations of the 
integrated light of the major planets have 
been obtained with the Cassegrain scan- 
ner on the 60-inch telescope, from X3400 
out to the long-wavelength sensitivity 
limit of photomultipliers with trialkali 
cathodes. These scans have in part been 
studied by Mr. Robert Younkin of the 
Jet Propulsion Laboratory. Repeated 
tracings of Jupiter obtained during four 
different periods have provided consistent 
results proving that there is not a steep 
fall in the energy distribution shortward 
of X3900, as has been reported in the past. 
During this preliminary work, it has been 
found that the amount of time involved 
in an exhaustive study of the tracings is 
so great that such study is impractical 
without the aid of automatic data 
reduction equipment. An arrangement in 
which the output of the scanner is fed 
directly into a digital voltmeter with 
magnetic tape recording has therefore 
been tested. This will be reduced with an 
IBM 7090 computer. Further observa- 
tions of the variations in color and spectra 
of planetary surfaces will be carried out 
by means of such auxiliary equipment. 

Munch is utilizing the greatly increased 
resolving power and speed of the coude 

spectrographs to take spectra of the 
planets under the highest dispersion 
possible with the purpose of detecting 
new spectral features and studying the 
structure of known ones over the various 
parts of the planetary disks. Plates of 
Jupiter, Saturn, and Uranus in the blue 
and yellow-red regions of the spectrum 
have been obtained. Part of this material 
is being studied in collaboration with Dr. 
Hyron Spinrad of the Jet Propulsion 
Laboratory. In these plates, Spinrad 
verified his discovery of the anomalous 
inclination of the NH 3 lines in the 
spectrum of Jupiter's equator. In Saturn 
it was found that the lines of CH 4 band 
at X6190 have an inclination greater than 
half that of the Fraunhofer lines of the 
scattered solar spectrum, by about 10 
per cent — an amount twice as large as the 
probable error of measurement. In the 
same plates of Saturn two sharp lines of 
X6367.80 and X6435.03 which, with cer- 
tainty, must be identified with the S(l) 
and S(0) lines, respectively, of the 4-0 
quadruple rotation- vibration band of H 2 
were discovered. The possibility of ob- 
serving the quadruple spectrum of H 2 was 
suggested by Herzberg in 1938, but this 
is the first observation of the S(l) and 
S(0) lines of the 4-0 band in any astro- 
nomical object. 


The bright comet Seki-Lines (1962c) of NH 2 and C 2 . The distortion of the 

was observed by Greenstein. It was CN (0, 0) band by resonance fluorescence 

remarkably dust-free after perihelion, the was quite different from that of Comet 

Na I lines were very weak, and the visual Mrkos (1957d). Spectra of 1962c will be 

region of the spectrum consisted largely measured by Greenstein and Arpigny. 


During the report year, 900 spectro- 
grams were taken with the 200-inch 
telescope, 970 with the 100-inch, and 550 
with the 60-inch. 

Chemical Composition of 
Stellar Atmospheres 

The program for the study of the 
abundances of the elements in astronom- 

ical objects continued under the direction 
of Greenstein with the support of the Air 
Force Office of Scientific Research of 

Spectrophotometric analyses of the 
high-galactic-latitude supergiants HD 
161796 (F3 lb) and 89 Herculis (F2 la) 
have been carried out by Searle, Sargent, 
and Jugaku. Comparisons were made with 


the standard low-latitude supergiants 
<p Cassiopeiae (FO la) and a Persei 
(F5 lb). All the elements studied are 
found to have the same relative abun- 
dances in all these stars. Spectroscopic 
absolute magnitudes were derived for the 
high-latitude supergiants using <p Cas and 
a Per as calibration stars. It is concluded 
that both 89 Her and HD 161796 could 
have reached their present heights above 
the galactic plane in times comparable to 
their estimated times of evolution from 
the main sequence if they were expelled 
from the plane at the time of their forma- 
tion with velocities of the order of 100 
km/sec. These results — which are in 
disagreement with those of an earlier 
study of these same stars by Abt — are 
consistent with the view that the high- 
galactic-latitude supergiants are evolved 
runaway stars. 

Additional measurements since last 
year on 3 Centauri A were made by 
Jugaku and Sargent on a Radcliffe 
Observatory coude plate of the far ultra- 
violet for line identifications. Most of the 
40 lines between X3500 and X3100 can be 
identified with Fe II, Mn II, and Ni II, 
although a few fairly strong lines remain 
unidentified. The Be II doublet at X3130 
is absent. A plate of the visual region 
obtained in April 1962 shows that the 
longward shift of X6678 of He I, which 
was interpreted as an isotope shift, is 
still present. 

The abundance analysis of k Cancri, 
B8p, an Mn star, by Jugaku and Sargent 
is progressing. Equivalent widths of 250 
lines in the photographic region have been 
measured. The ratio P III/P II gives a 
value of 0i on = 0.39, which is typical of a 
normal star of spectral type about B7. 
The B — V color also agrees with such a 
temperature. A study of the hydrogen 
lines gives log P e = 1.92. Using these 
values of 6 and log P e , the preliminary 
abundance results are P/Si ^ 1 (as in 3 
Cen A — this means that P is overabun- 
dant by a factor of 100). The identifica- 
tion of a line at X3984 with Hg II by 
Bidelman leads to an overabundance of 

about 30,000 for Hg. Helium is deficient 
by a factor of about 10 (factors of 6 were 
found for 3 Cen A and a Sculptoris). Be 
is overabundant by a factor of 100 
relative to the sun. 

The study of the infrared O I lines in 
the spectra of 20 Ap stars has been 
completed by Searle and Sargent. Oxygen 
is found to be deficient with respect to 
hydrogen by factors ranging from 8 to 
more than 100 in all Ap stars of the 
Si-Eu-Cr, Eu-Cr, Eu-Cr-Sr, and Sr 
classes, whereas in the Mn stars the 
oxygen abundance is normal. Assuming 
that they originated with normal compo- 
sition, the oxygen-deficient Ap stars 
demand that O must have been trans- 
muted into one or more of the cosmically 
abundant elements by an as yet unspeci- 
fied process. 

The infrared N I lines fall at the limit 
of plate sensitivity and for this reason 
have been studied in only four bright Ap 
stars. Two of these, a 2 Canum Venati- 
corum and /3 Coronae Borealis, are 
deficient in O; the remaining two, 
ip Herculis and n Leporis, are Mn stars 
with normal oxygen abundance. In none 
of these stars are the infrared N I lines 
detectable, although they are clearly 
present in the spectra of standard stars 
from B5 to F5. It appears that nitrogen 
is deficient in all four Ap stars by factors 
estimated to be 10 or more. 

Infrared spectrograms, at 20 A/mm, 
have been obtained by Searle and Sargent 
of four bright metallic-line stars and four 
standard stars to study the behavior of 
the O I lines. The metallic-line stars 
selected fall in the two-color and color- 
magnitude diagram among the extreme 
oxygen-deficient Ap stars, but unlike the 
Ap stars their oxygen abundance is 

Spectrograms of 30 Ap stars and 6 
standard stars with types between B5 
and A4 have been obtained at 10 A/mm 
by Searle and Sargent. The pressure- and 
temperature-insensitive ratios of Mg II 
(X4481)/Si II (X4128, X4130) and C II 
(X4267)/He I (X4471) are being studied 



and Balmer line profiles obtained. 

For the Ap stars of earliest type, the 
X4200, Si and Mn classes, there is an 
excellent correlation between the central 
depth of Hy and the U — B color, iden- 
tical with the correlation obtained for 
standard stars. There is no systematic 
difference between the Balmer line profiles 
of the Ap stars and the standard stars, 
and the location of an Ap star in the 
two-color diagram is a good indicator of 
the atmospheric temperature of the star. 

Among the hotter Ap stars there is no 
evidence for Mg abundance anomalies, 
and Si is overabundant (by about X 10) 
only in stars that show the high-excitation 
Si II lines at X4200. Certain sharp-line 
late B-type stars (e.g., 21 Aquilae and 
HD 207840) which have been called Si 
class Ap stars have Si lines of normal 
strength for their colors, and normal 
strength of He, C, O, and Mg. It is 
probable that they are "peculiar" only in 
that they have unusually sharp lines. 

The C II (X4267) strength is normal 
for the color in Ap stars of the Mn class 
and in the so-called Si stars which do not 
show X4200. In these stars, the lines of 
He I are normal or only slightly weaker 
than in normal stars of the same color. 

In the stars that have definite Si over- 
abundance (the X4200 stars), the C II line 
is weak (by a factor of about 5 in the 
equivalent width) and the He I lines are 
very weak (by a factor of about 10 in the 
equivalent width). It seems hard to 
escape the conclusion that the X4200 stars 
are very deficient in helium. 

Spectrograms of 27 Ap stars and 4 
standards at 10 A/mm in the photo- 
graphic ultraviolet have been obtained by 
Searle, Sargent, and Jugaku to study the 
behavior of the Be II doublet at X3130. 
In the cooler Ap stars, the Be II lines, if 
present, are seriously blended, but in the 
hotter Ap stars they are free from serious 
blending. No lines of Be II are to be seen 
in nine X4200 stars observed. Of 10 Mn 
stars, 6 show no trace of Be II, but the 
remaining 4 have very strong lines. In the 
Be-strong Mn stars (112 Herculis, k 

Cancri, /x Leporis, and v Herculis), Be is 
estimated to be overabundant by a factor 
of about 100. All the stars that show 
strong lines of Be II also have lines of 
Ga II in their spectra. However, ir 1 
Bootis, in whose spectrum Ga II lines are 
present, shows no trace of Be II. 

Scans of the continuous spectrum of 
about 20 Ap stars have been made by 
Jugaku and Sargent with the Cassegrain 
scanner. They will be examined in 
conjunction with hydrogen-line profiles 
obtained from coude plates of the same 
stars to see whether there is definite 
evidence that the atmospheres of mag- 
netic stars differ from those of normal 
stars. Preliminary results show that stars 
like a Andromedae (B8p Mn) , which have 
anomalously blue B — V and U — B 
colors, have complete continuous energy 
distributions identical to those of main- 
sequence stars as hot as B5. 

Greens tein, Parker, Wallerstein (Uni- 
versity of California at Berkeley), Heifer 
(University of Rochester), and Aller 
(University of Michigan) have collabo- 
rated in an extensive analysis of three red 
giants with extremely weak lines: HD 
122563, 165195, and 221170. In last year's 
report, HD 165195 and 221170 were 
mentioned as weak-lined G dwarfs. 
Further analysis and the earlier incom- 
pleted work of Greenstein and Aller on 
HD 122563 have shown that these stars 
are, in fact, giants with colors and spectra 
like the stars in globular clusters. In 
many ways, these stars are extraordinary; 
their colors are quite red (B — V ^ -f- 1.0) , 
yet at first glance they could be mistaken 
for F subdwarfs. The results are tempera- 
tures near 4100°K, log P e = -2.5, 
metal/hydrogen ratios 500 times lower 
than in the sun (like the most extreme 
subdwarf). One problem is in the opacity, 
which seems to be largely Rayleigh 
scattering, although the expected colori- 
metric effects are not found. In addition, 
the ratio of iron-group metals to heavy 
elements like Sr, Zr, Ba, Ce, and Eu is 
abnormal, when compared with the sun, 
in that the heavy elements are deficient 


by an additional factor of 50 in HD does not differ from that of the sun by 
122563. This is the first known example more than a factor of 2. 
of large changes within the abundances of Gunn is carrying forward a program of 
the metals. It indicates that the stars studying ' 'strong-line" versus "weak- 
condensed at an early stage in the life of line" field F stars from spectrograms of 
our Galaxy, perhaps 10 7 to 10 9 years (at 20 A/mm dispersion taken with the 
the very latest), and probably 10 7 to 10 8 60-inch. It will be determined whether 
years after the beginning of element the weak-line group can be explained only 
synthesis. Another unusual effect is that in terms of lowered metal abundance 
Eu behaves like the other heavy elements; (relative to hydrogen) or whether differ- 
although Eu in the sun and earth was ences in mean turbulence and mean 
synthesized by the r process of neutron degree of ionization can explain the 
capture, in HD 122563 it was made by existence of the group, 
the s process. There are other elemental 

deficiencies, e.g., V and Mn, of a previ- Lme Blanketing 

ously recorded type, and also evidence Sandage and Smith completed an 

that the heavier elements were synthe- observational study of the differential 

sized in a very metal-poor environment, blanketing effect of weakening the Fraun- 

Gunn and Kraft have completed a hofer lines in stellar spectra. A four-color, 
study based mainly on 200-inch coude broad-band photometric system was de- 
spectrograms of the hydrogen-to-metal vised using an RCA 7263 photomultiplier 
ratio in F-type stars of NGC 752, a cell with an S20 trialkali photocathode. 
galactic cluster of age 1 X 10 9 years. It The system is close to the standard U, B, 
has been suggested from earlier studies of V but adds a fourth color, R, at an 
small-scale spectrograms that NGC 752 effective wavelength of 6800 A. Observa- 
may have a lowered metal abundance tions were made with the Palomar 20-inch 
relative to the sun. If this were true, it telescope of 64 standard stars whose U, 
would mean that as little as 10 9 years ago B, V values are well known and of 32 
different regions of the Galaxy were subdwarfs of intermediate to large ultra- 
forming stars of different metal content violet excess. Three results came from the 
at the same time, though the galactic study. (1) It is possible to transform the 
orbit of NGC 752 is nearly the same as natural photometric system of the S20 
that of the sun. photocathode to the U, B, V system 

In the present analysis, careful atten- (usually observed with an S4 cathode), 
tion was paid to deriving accurate with a systematic accuracy of 0™02 in 
ionization temperatures. With the aid of B — V and 0™05 in U — B. The data 
H7 profiles based on model atmospheres show a nonlinearity in the (u — b) 
(Searle and Oke, 1962) and known natural =f(U — B) transformation curve 
abundances (Parker, Greenstein, Heifer, of amplitude 0™05 which is undoubtedly 
and Wallerstein, 1961), the scale of Ti on due to the different ultraviolet response 
for Hyades F-type stars was first estab- of the 1P21 and RCA 7263 multiplier, 
lished from curves of growth. From the Therefore, precise transformations of S20 
models and H7 line strengths, ionization data to the U — B system must be done 
temperatures for NGC 752 stars were with a nonlinear equation. (2) The effect 
estimated relative to Hyades stars. This of differential line blanketing on the 
method of determining Ti on replaces the positions of stars in the U — B versus 
customary, and somewhat unsatisfactory, V — R diagram is clearly seen. The sub- 
procedure of estimating Tion from T eKC . dwarfs, as expected, stand high by 
Using these temperatures, it is concluded d(U — B) = 0™2. (3) An extension of the 
from a study of curves of growth that the blanketing theory discussed in Year Books 
metal abundance of stars in NGC 752 59 and 60 was made to include the R 



point, and it was shown that the theory 
can with great accuracy correct the sub- 
dwarf positions in the U — B versus 
V — R and in the B — V versus V — R 
diagrams to the position of stars of high 
metal abundance. Therefore, most and 
perhaps all of the previously observed 
peculiarity in the energy distribution of 
subdwarfs over the spectral range X3300 
to X6809 can apparently be explained as 
due to the effects of weak Fraunhofer 
lines on broad-band photometric meas- 
urements. Unpublished data of Sears and 
Whitford at the Lick Observatory suggest 
that this is also true all the way to the 
infrared point at X = 10,000 A on their 
six-color system. 

Spectroscopic scans are being obtained 
by Oke for a selection of very metal- 
deficient stars. Because of the weakness 
of the lines, the absolute energy distribu- 
tion in the spectrum can be accurately 
obtained over a large wavelength interval. 
This facilitates comparisons with model 
atmosphere fluxes. In addition, model 
atmospheres for these stars can be com- 
puted with higher accuracy than for 
corresponding metal-normal stars. The 
hydrogen-line profiles will also be studied. 
Scans of the extremely metal-deficient 
red giant HD 122563 have been obtained 
between X3400 and X8000. Assuming that 
the continuous opacity is due to the 
negative hydrogen ion, comparison with 
model atmospheres gives an effective 
temperature of about 4200°K, in good 
agreement with the value determined by 
Greenstein, Wallerstein, and Parker. 

Color-Magnitude and 
Chemical-Composition Relationships 

Eggen and Sandage reexamined the 
problem of the position of the main 
sequence as a function of chemical com- 
position in the M v , B — V and the M boi, 
log T e diagrams. A sample of stars was 
chosen (1) which were known to be 
dwarfs from spectroscopic luminosity 
criteria, (2) which had trigonometric 
parallaxes larger than 0"034, and (3) for 
which photometry on the U, B, V or U c , 

B, V was available. From this extensive 
material, Eggen and Sandage confirmed 
their previous result (1959) that the 
displacement of a star below the Hyades 
main sequence in the M v , B — V diagram 
is directly proportional to the ultraviolet 
excess. Assuming that the ultraviolet 
excess is due to Fraunhofer-line weaken- 
ing caused by low metal abundance, and 
applying the blanketing theory reported 
in previous years, it was shown that 
differential blanketing corrections move 
all stars, independently of the size of their 
ultraviolet excess, onto the Hyades main 
sequence with good accuracy. In par- 
ticular, 16 extreme subdwarfs with well 
determined absolute magnitudes and with 
excess values averaging 8 (U — B) = 0^21 
were observed to be 1™05 ± 0™04 fainter 
than the Hyades before blanketing cor- 
rections were applied but are moved onto 
the Hyades main sequence to within 
+0™03 ± 0^05 after the corrections were 
made. These results provide the observa- 
tional justification for the main-sequence 
fitting procedure to find distances of star 
clusters where it has always been assumed 
that the main-sequence positions are 
identical in the M v , B — V diagram for 
clusters of different chemical composition. 
These reported results show that the 
assumption is correct, but only if differ- 
ential blanketing corrections are applied 
before the modulus fit is made. 

It was further shown that the large 
scatter in the main sequence, between 
B - V = m 4 and B - V = m 8 of 
trigonometric parallax stars, is due not 
only to errors in the parallaxes but 
primarily to the line-blanketing effect on 
the colors. The distribution of ultraviolet 
excess shows that large variations in 
chemical composition exist among the 
parallax stars closer than 29 parsecs 
(ir > 0"035). The main-sequence scatter 
is markedly reduced when blanketing 
corrections are applied to the observed 
B — V colors. 

The fact that subdwarfs are moved 
onto the Hyades line after applying 
blanketing corrections shows that a 



separate subdwarf sequence does not 
exist in the M ho i, log T e diagram, despite 
the low metal abundance of these stars. 
This remarkable result permits the deter- 
mination of the helium abundance in the 
interior of subdwarfs from the theory of 
the stellar interior. Decreasing the metal 
abundance will, in general, move the star 
below the main sequence. A correspond- 
ing decrease in the helium abundance 
works in the opposite direction. The 
effects can be predicted qualitatively by 
homology arguments, but computed stel- 
lar models are needed for accurate 
abundance determinations. Sandage used 
the models of DeMarque to show that 
the hydrogen (X) , helium ( Y) , and metal 
(Z) abundances by weight of the extreme 
subdwarfs with well determined absolute 
magnitudes are X « 0.95, Y « 0.05, 
Z ^ 0.001, compared with adopted solar 
values of X = 0.65, Y = 0.31, Z = 0.04. 
Subdwarfs are the oldest stars we know. 
Therefore, this result suggests that the 
primeval abundance of both helium and 
the heavy elements was very low, a result 
in agreement with theories of element 
enrichment of the interstellar medium 
with time. The results are not final. 
Interior models with more closely spaced 
abundance differences are needed. Ikco 
Iben of the California Institute Physics 
Department spent two months at Los 
Alamos computing better interior opacity 
values from an IBM 7090 computing 
program developed by A. N. Cox and 
Robert Brownlee. Iben's resulting models 
will be used when they are completed for 
a second solution to the problem. 

From a photometric study of three 
separate samples of main-sequence stars, 
Eggen has found that, judged by the 
distribution of ultraviolet excesses with 
respect to the Hyades stars, two-thirds of 
the stars in the solar neighborhood have 
a higher ratio of metals to hydrogen than 
the sun. The distribution of ultraviolet 
excesses suggests that, if enrichment of 
the interstellar medium has been uniform 
with time, the rate of star formation 
between 5 X 10 9 (formation of the sun) 

and 5 X 10 8 (formation of the Hyades) 
years ago was nearly uniform. 

Color-Spectral-Type Relationships 

Two years ago a report was published 
by Wilson in which it was indicated that, 
among main-sequence field stars later 
than type G5, a considerable spread of 
color for a given type, or of type for a 
given color, was present. The data used 
for this purpose were the old Mount 
Wilson spectral types and the photo- 
electric colors measured by Eggen. Sub- 
sequent spectrograms of some of these 
stars revealed that the Mount Wilson 
spectral types were unreliable. As a 
result, new spectrograms of 10 A/mm 
dispersion have been obtained by Wilson 
for more than 100 of these stars, and new 
types, based on the Yerkes system, have 
been derived. When the revised types are 
plotted against Eggen's colors, a con- 
siderable spread, amounting to 0.2 mag 
at some types, is again found, although 
many stars no longer occupy the same 
locations in this diagram as in the former 

In addition, many members of the 
Hyades cluster were also observed spec- 
troscopically (although at smaller disper- 
sion), and a similar diagram was con- 
structed for them. When these two 
diagrams are compared, the correlation 
between spectral type and color appears 
to be tighter for the Hyades members 
than for the field stars. In the light of 
current knowledge, the simplest explana- 
tion of this result is to suppose that, as 
regards chemical composition, the Hyades 
stars represent a more uniform group 
than the field stars. This conclusion is not 
especially surprising. 

In recent Year Books Wilson has 
reported the discovery of a definite 
relationship between the width of the 
reversals of the H and K lines and the 
absolute magnitude of the star. However, 
the total emission intensity of the H and 
K reversals seems to have no obvious 
correlation with other features of the 
spectrum. This raises the question: why 


is it that among the late main-sequence members of double and multiple systems 

field stars there are stars which appear to have H and K reversals of similar 

spectroscopically identical in every re- strengths when due allowance is made for 

spect except that one has strong central differences in spectral type. The few 

H and K reversals and the other little or exceptions to this rule seem to be re- 

none? In the previously published work stricted to members of systems that are 

mentioned above, it appeared that there themselves short-period spectroscopic 

was a strong tendency for the redder binaries. 

stars of a given type to have stronger Clearly, Wilson's work summarized 
reversals than their bluer counterparts, above is still in a preliminary stage, and 
With the revised types, however, little if definite conclusions should be avoided at 
any of this tendency remains, and thus present. Nevertheless, enough has been 
what appeared to be a promising clue has accomplished to justify pursuing it fur- 
proved to be illusory. ther in attempting to understand its 

As an outgrowth of the present research significance and, perhaps, eventually in 
more definite light is being shed on this making use of it as a tool for the further- 
problem. That H and K reversals are ance of other aims, 
unusually frequent in the Hyades stars 

has long been known and is fully con- Photometry of Stellar Clusters and 

firmed by the present work; indeed, if the Associations 

discussion is restricted to types G5-K0, Sandage and Smith completed the 

inclusive, the fraction of Hyades members study of the color-magnitude diagram of 

in this range which has strong reversals NGC 6712, a globular cluster of relatively 

seems to approach 100 per cent. The same high metal abundance situated in the 

thing is true of main-sequence members Scutum Cloud. A photoelectric sequence 

of the Praesepe cluster. Even in the Coma was determined to V = 17™5, B = 18™5 

cluster, which is relatively poor in known with the 200-inch, and short-exposure 

main-sequence members, the frequency plates were measured for the color- 

of strong reversals appears to be greatly magnitude diagram. It suffices to remark 

in excess of that for the local main- that the earlier conclusion reached from 

sequence objects of the same range of the study of NGC 6536, that the absolute 

spectral types, where the frequency is less magnitudes of the brightest giant stars in 

than 20 per cent. globular clusters are a function of their 

Thus, on the available evidence, stars chemical composition, is confirmed, 

formed in clusters have a much higher Plates of the globular cluster NGC 6712 

probability of possessing strong H and K taken by Sandage were photometered for 

reversals than the noncluster field stars RR Lyrae stars and reduced by Norton 

of the same spectral types. This property and Lynden-Bell. 

may then, perhaps, be thought of as a Sandage continued the photoelectric 

genetic factor that can supply information measurement of faint stars in M 15 and 

of some kind as yet unspecified about the M 92. Complete color-magnitude dia- 

circumstances under which the stars grams for these clusters were prepared, 

were formed. Katem measured a special series of plates 

A natural extension of the investigation from V = l?^ to V = 22™0 in both 

along the line of genetic thinking is to clusters to obtain the main-sequence 

look at the reversals in double and positions using the photoelectric stars as 

multiple main-sequence systems. This standards. The results disagree by about 

has been done for a number of such 0^05 in the color of the main-sequence 

objects, and the results very strongly termination point with the definitive 

support the genetic viewpoint. That is, photometry of M 13 reported several 

there is a very decided tendency for the years ago, and a special program of photo- 


electric intercomparison of the five clus- globular clusters so far analyzed, the 
ters M 3, M 5, M 13, M 15, and M 92 derived ages drop to between 9 and 
was begun. This study must be completed 14 X 10 9 years. The mean absolute mag- 
before the M 15 and M 92 results will be nitude of the RR Lyrae stars should be 
published. adopted as M v = +0.3, M B = +0.5 

A photometric investigation of the mag. 
clusters and surrounding associations of Investigation of clusters of intermedi- 
h and x Persei using combined photo- ate age (10 9 years) and low metal content 
electric and photographic techniques has (3^2 to }/% of normal) has been completed 
been made on the U, B, V system to by Arp. Work continues in the clusters 
V = 17.0 by Wildey. A nonunique epoch near the Galactic nucleus. A very clear 
of star formation is suggested both by an giant branch in the color-magnitude 
apparent fine structure in the bright end diagram of NGC 6838 has been obtained, 
of the C-M diagram and by the presence The brighter sequences appear somewhat 
of main-sequence stars to the photometric similar to those in 47 Tucanae and NGC 
limit even though apparently contracting 6356. The photometric fitting of the main 
stars are found at brighter magnitudes, sequence and derivation of accurate 
Nuclear and gravitational time scales are reddening are now in progress, 
in agreement. The data, when compared Eggen has isolated the "Pleiades 
with the theoretical evolutionary tracks Group," which, together with the Hyades 
of Hayashi and Cameron, suggest that and Sirius Groups, accounts for nearly 
helium burning takes place on the blue 25 per cent of the A-type stars in the solar 
side of the Hertzsprung gap. The differ- neighborhood. Also, two groups of high- 
ences in apparent evolutionary tracks velocity stars, in addition to the Groom- 
between the Galaxy, the Large Magel- bridge 1830 Group, have been found. One 
lanic Cloud, and the Small Magellanic of these, Kapteyn's Star Group, contains 
Cloud are reconfirmed. The two-color at least two RR Lyrae variables, SU 
diagram of the bluest stars is of a gray- Draconis (0?66) and ST Leonis (0 d 48). 
body character. Sp versus (B — V) is The derived median, visual absolute 
mono tonic for all stars. magnitude is +0^8 for both stars. 

The publication of the analysis of the Another possible member of this group is 
color-magnitude diagrams of M 5, which ST Comae Berenices (0™60). The sub- 
indicated ages of 20 X 10 9 years, empha- dwarfs HD 106038, CC835, and CC486, 
sized the serious discrepancy between as well as the horizontal branch A-type 
these ages and those yielded by the star HD 139961, are also members. The 
measurements of cosmological expansion, mean ultraviolet excess of the group is 
In connection with another investigation, -f-0?21, and the variables all have 
Arp has reexamined the assumptions in AaS = 6. No RR Lyrae variables have 
regard to space reddening made in the been identified as belonging to the third 
analysis of the color-magnitude diagrams high-velocity group, which contains the 
of globular clusters. He found the following subdwarfs HD 74000, Ross 626, Ross 
results: (1) The cosecant reddening law 451, Ross 453, HD 108177, and -35°- 
applies to high-latitude globular clusters 14849. 
despite the fact that photometric analysis 

of stars within 500 parsecs shows on the Photometry of Double Stars 

average 0.06 mag less reddening than Eggen began a program of U, B, V 

that given by the law. The implication photometry of the components of visual 

that reddening in the Galaxy is not double stars with the 200 -, 100-, and 

concentrated entirely within a thin sheet 60-inch reflectors. The components of 

needs to be examined further. (2) If the about 100 pairs have so far been observed, 

cosecant reddening law is used for all the including the following of special interest: 



(1) Red dwarf-white dwarf pairs. Ob- 
servations of the components of six wide, 
proper-motion pairs discovered by Giclas, 
Slaughter, and Burnham at the Lowell 
Observatory, G39-27/28, G 14-57/58, 
G87-28/29, G102-39/40, Gil 1-71/72, and 
G61-16/17, have confirmed the discovery 
suspicions that one of the components in 
each pair is a white dwarf. An additional 
pair, G24-9/10, has also been found to 
contain a late-type white dwarf. (2) 
Intrinsic variables. The results for the 
main-sequence companions of the (3 
Cephei variables, Cephei and a Scorpii, 
indicate that the luminosities of these 
variables may be about a magnitude 
fainter than usually supposed. Photom- 
etry of the solar-type, main-sequence 
companions of the Mira variables, R 
Cassiopeiae and RU Cygni, indicates a 
luminosity near — 1 for both variables. A 
weak-lined G-type subdwarf, 10 seconds 
from the RR Lyrae variable AP Serpentis 
(P = d 25), shows an ultraviolet excess 
of 4-0^12 and gives an absolute median 
luminosity near +2 m for the variables if 
the stars form a physical system. (3) A 
wide range of eclipsing stars with visual 
companions is included in the program to 
help in standardizing mass-luminosity 
and radius-luminosity relations. Also, ob- 
servations of the physical companions to 
several bright W Ursae Majoris systems 
indicate that the ultraviolet excesses are 
the same for the variable and the non- 
variable components. The system of VW 
Cephei shares a large space motion with 
HD 199476, although the stars are 
separated in the sky by about 1 degree. 
W Ursae Majoris, BD +55°1351, has a 
faint companion (V E = 12 m 35, B — V 
= -f-1^70), a degree away, with which it 
forms ADS 7497. Although the two com- 
ponents are not physically connected they 
share a common proper motion. The 
radial velocity of the BD star is not avail- 
able. From the companions of these two 
variables, as well as that of AM Leonis 
(ADS 8024), the variables are found to 
lie 0™75 above the main sequence. A faint 
star (V E = 13 m 84; B - V = +0 m 65; 

U — B = +0^11) about 10 seconds from 
UY Ursae Majoris may be physically 
connected to that variable. (4) Structure 
in the mass-luminosity relationship has 
been found for stars with different metal 
contents as judged by their ultraviolet 
colors compared with Hyades stars. The 
colors of all binaries for which orbital 
elements are available are being obtained 
by Eggen with the 20-inch and 60-inch 
reflectors for further study of this 

Photometry of Variable Stars 

Eggen has undertaken three-color 
photometry of all northern cepheids not 
already observed photoelectrically. One 
preliminary result from this program is 
that Baade's faint cepheids in Cygnus are 
reddened by about 1™0. Also, all known 
contact binaries are being observed for 
color. Two variables classified in the 
literature as contact binaries have been 
found to violate the period-color relation 
previously established by Eggen. Exten- 
sive observations of one of these, SZ 
Lyncis, shows it to be a short-period, RR 
Lyrae variable (period near d 13) and not 
a contact binary. A previous conclusion 
that T Tauri variables and contact 
binaries do not coexist in space is appar- 
ently violated by the recent discovery by 
Gotz of V449 Orionis (W Ursae Majoris, 
P = d 44) and V441 Orionis (T Tauri 
Variable) which are separated by less than 
30 minutes. Observations on five nights 
in February with the 100-inch indicate 
that V449 Ori is not a contact binary. The 
star showed very little variation in visual 
magnitude, V E = 15^10 to 15 m 28, or 
(B - V) color, + l m 05 to -fl m 26, but the 
ultraviolet color showed erratic variations 
of more than a magnitude. V441 Ori 
(V E = 14 m 72, B - V = l m 74, U - B = 
-p-l^ll) showed no variation. Other 
variables observed include Nova Orionis 
(1667 and 1894) and X Leonis. The nova 
showed no variation in the blue and visual 
(V E = 14 m 14, B - V = +0 m 48) on five 
nights, but there are erratic variations 
in the ultraviolet. A bright maximum of 



X Leo was observed from February 2 to 5, 

Photometry of the Giclas Proper 
Motion Catalogue 

A routine program of photoelectric 
observation of stars of high proper motion 
taken from the Giclas Lowell Observatory 
Proper Motion Catalogue was started in 
September by Kowal under the super- 
vision of Sandage. The pilot program for 
the discovery of new subdwarfs, reported 
last year, was so successful that this 
regular discovery program was begun. 
Good progress was made during the 
report year with more than 700 stars ob- 
served and reduced in the three colors 
U, B, V. Both the 60-inch and 20-inch 
reflectors were used to make the observa- 
tions. Over 100 new subdwarfs have been 
found. At least 30 of them have the 
extreme line-weakened characteristics of 
the globular-cluster main-sequence stars. 
Sandage continued his routine observa- 
tional program of determining radial 
velocities of the stars in Kowal's lists that 
have ultraviolet excess values greater 
than 8(U - B) = 0^14. This is a standby 
program with the 200-inch coude, which 
is used only when sky conditions prevent 
more critical photoelectric work at the 
prime focus. But poor seeing conditions 
and partly cloudy weather were prevalent 
enough during the winter season so that 
spectra for 35 program stars were ob- 
tained in this way with the 18-inch camera 
giving a dispersion of 18 A/mm. Again, 
the results show that stars with high 
ultraviolet excess values invariably have 
large space velocities relative to the sun. 

The photometric discovery program 
will continue for another year, by which 
time it is hoped that more than 2000 stars 
of the Giclas Catalogue will have been 


Greenstein has completed a velocity 
program on some 150 F-K subdwarfs and 
halo B and A stars. The number of 
spectroscopic binaries found is very low. 
Nevertheless, spectroscopic examination 

of the high-velocity stars revealed many 
types of peculiarities. After several at- 
tempts, an apparently reliable scheme of 
visual classification at 18 A/mm was 
developed with the following results: 
extreme weak-line subdwarfs of F-K 
types, 30 per cent; moderately weak-line 
subdwarfs, 18 per cent; slightly weak-line, 
7 per cent; subgiants or giants, 10 per 
cent; horizontal-branch A-G stars, 4 per 
cent; ionized lines enhanced, 15 per cent; 
CH enhanced relatively to metals, 16 per 
cent. Some of the so-called subdwarfs are 
clearly above the main sequence spectro- 
scopically, and the 10 per cent figure for 
giants or subgiants means that many very 
high space motions are included. The 
radial velocities have internal probable 
errors from 0.5 to 1.3 km/sec. 

Jones has discussed a collection of 200 
plates of late-type subdwarfs taken by 
Greenstein and Sandage with the coude" 
spectrograph of the Hale telescope at 18 
A/mm. An attempt has been made to 
set up a three-dimensional classification 
scheme which estimates the spectral type 
and luminosity of the star independent of 
any weakening of the spectral lines. The 
spectral type is based on four ratios of 
line pairs of differing excitation potential 
and the strength of the hydrogen lines; 
the luminosity is estimated from three 
ratios of ionized to neutral lines in a 
manner closely following the Mount 
Wilson Catalogue. The strength of six of 
the 1.5-volt lines of Fe has also been 
estimated, and the values have been 
combined to form an index of the line 
strength. The principal conclusions are as 
follows: spectral types on this system are 
well correlated with the B — V colors 
corrected for blanketing as far as the 
main sequence is concerned, but stars 
above the main sequence appear to obey 
another relation. Stars whose lines are 
among the strongest for their assigned 
spectral type have types on this system 
that correlate very well with the MK and 
Mount Wilson systems. This correlation 
was used to convert the new types from 
an arbitrary numerical scale to the well 



known A-F-G-K scale. Line weakening 
occurs more frequently for stars earlier 
than G5, where it may amount to 0.7 
logarithmically. Later than G8, it rarely 
exceeds 0.1. Owing to the small amount of 
material, only one curve was derived to 
reduce the luminosity characteristic to 
absolute magnitude based on trigono- 
metric parallaxes, but the absolute magni- 
tudes so derived agree very well with the 
Mount Wilson Catalogue. The Hertz- 
sprung-Russell diagram shows a marked 
main sequence with a large number of 
weak-line stars which appear to define a 
"turnoff" at about G2. There are also 
several subgiants, some with markedly 
weak lines. 

White Dwarfs 

The discovery or spectroscopic con- 
firmation of white dwarfs was continued 
by Greenstein, who finds Feige 22, 24, 
Giclas 21-16, 24-10, 28-13, 29-38, 61-17, 
67-23, 93-53, LDS 235B (helium-rich), 
-37°10500, Wolf 457 (essentially con- 
tinuous) to be of this type. A program of 
observation of white-dwarf members of 
double stars has been started by Green- 
stein in collaboration with the photo- 
metric work of Eggen. One X4670 white 
dwarf, i.e., molecular carbon-rich, is a 
member of a wide binary. 

Faint Blue Stars 

A project was started by Zwicky with 
the aid of a grant from the National 
Science Foundation to obtain spectra of 
special types of blue stars for the purpose 
of determining their radial velocities and 
spectroscopic characteristics. The data 
thus obtained will be used to investigate 
the statistical distribution and proper 
motions of these objects. 

The spectrographic observations were 
made by Berger with the 4-inch camera 
on the 60-inch Cassegrain spectrograph. 
Most of the stars observed are in the list 
of subluminous hot stars discovered by 
Feige. A few additional stars were sup- 
plied by Haro and by the list of blue stars 
published by Cowley. One to four spectra 

of 17 stars have been obtained in the 
magnitude range between 7.5 and 11.3. 
Preliminary measures of part of the 
plates do not indicate a large velocity for 
these blue stars, \V\ < 50 km/sec. 

A search for very faint blue stars 
(15 < m < 19) near the north galactic 
pole has been started by Berger with the 
48-inch schmidt using three-color photog- 
raphy. These observations will be com- 
bined with the observations of 8500 faint 
blue stars near the south galactic pole 
made by Haro and Luyten for an in- 
vestigation of the statistical distribution 
of the halo population. These plates will 
also be compared with the National 
Geographic Society-Palomar Observatory 
Sky Survey plates taken in the early 
1950's for the detection of faint blue stars 
with large proper motions. 

B aimer Lines in Early-Type Stars 

Spectra of 57 B and Be stars have been 
obtained by Houziaux in the region 
X3900 to X3550 in order to study the 
confluence of the Balmer lines. It has been 
found that this emission-free region pro- 
vides a good observational criterion for 
the determination of the gravity of early- 
type emission-line stars. The ratios of the 
intensities between the Balmer lines, 
corrected for atmospheric absorption, to 
the intensity at X3862 are plotted versus 
the principal quantum number. Stars of 
the same spectral type and of different 
luminosities are clearly separated. These 
observational data are now compared 
with the results of flux computation for 
high lines at 41 wavelengths performed 
for 40 model atmospheres in the tempera- 
ture range 9510°K to 29,000°K and for 
log g = 1(1)5. 

Changes have been observed by 
Houziaux in the infrared spectrum of 
Pleione. The strong O I absorptions at 
X7771 and X8446 have disappeared. From 
a spectrophotometric study of the in- 
frared region, it has been shown that the 
O/H ratio during the shell episode was 
0.6 X10 -4 , a value similar to the one 
observed in other B-type stars. 



RR Lyrae Variables 

The photoelectric spectrum scanner has 
been used by Oke in a continued program 
to measure absolute energy distributions 
in the spectra of RR Lyrae variable stars. 
The stars being studied are RR Lyrae, 
SU Draconis, X Arietis, SW Androm- 
edae, and VZ Cancri. Photographic 
spectra with a dispersion of 9 or 18 A/mm 
are also being obtained. The absolute 
energy distributions are compared with 
fluxes computed from model atmospheres 
to obtain values of the effective tempera- 
ture at each phase. These profiles can also 
be used, along with the photoelectric 
scans, to determine space reddening with 
high accuracy. After correction for red- 
dening, RR Lyr, SU Dra, and X Ari all 
appear to have the same temperature 
range, 6000°K to 7500°K. The tempera- 
ture curves, as a function of phase, for 
these three stars are similar but not iden- 
tical. The radial-velocity curves for SU 
Dra and RR Lyr demonstrate conclusively 
that differential radial motions exist 
throughout the atmospheres at all phases. 
Since the continuous opacity changes with 
phase, different mass layers are observed 
with different velocities at various phases. 
Consequently, the observed radial-ve- 
locity curve does not represent the motion 
of the star's photosphere, and integration 
of the velocity curves does not give the 
radius-displacement curve. Therefore, the 
Wesselink-Baade method cannot give 
correct radii for RR Lyrae stars. It is 
found, however, that a modification of the 
method can be used successfully to 
determine the absolute radius. For SU 
Dra the mean radius is 5.2Ro. Using the 
temperatures obtained from the scans, 
this leads to a mean absolute visual 
magnitude of +0.8 ± 0.4. The error can 
be reduced if only differences of absolute 
magnitudes are required. 

The scanner is being used by Oke to 
measure absolute energy distributions of 
suspected horizontal branch stars. These 
measures will be used in conjunction 
with temperatures obtained by fitting H7 

to theoretical profiles to obtain the red- 
dening. The value of the effective gravity 
determined from the scans indicates 
whether or not a star can be a horizontal 
branch object. A comparison will be made 
of effective gravities of RR Lyrae stars 
and nonvariable horizontal branch stars 
to study the effects of the dynamics of 
pulsation on the atmosphere. One star, 
HD 161817, is confirmed to be a hori- 
zontal branch star similar to RR Lyrae 
at maximum light. 


The search for supernovae has con- 
tinued under the direction of Zwicky and 
with the support of the National Science 
Foundation. Between July 1, 1961, and 
May 31, 1962, a total of 15 supernovae 
was discovered at Palomar, all of them on 
plates taken with the 48-inch schmidt 
telescope. Of this number, 4 were dis- 
covered by Humason, 2 by Kearns, 6 by 
Zwicky, and 1 each by H. S. Gates, 
Rudnicki, and Berger. Of these, 2 were in 
the Coma cluster, as identified through 
determination of their symbolic velocities 
of recession. 

The supernova in NGC 4303 developed 
into a type II spectrum after a peculiar 
early behavior. Greenstein found, near 
maximum, large negative displacements 
of the emission lines which were accom- 
panied by very greatly displaced absorp- 
tion edges. Within a month, the emission 
lines became sharper and returned to zero 
velocity. Apparently both absorption and 
emission were formed at the leading edge 
of an opaque expanding shell. It is unclear 
whether the apparent deceleration is real 
or caused by the appearance of the far 
(receding) side of the star, but there is no 
doubt that the velocity spread was 
greatly reduced. 

Observations by Greenstein showed 
that the supernova in NGC 1058 had 
many extremely sharp lines in November 
1961, some accompanied by absorption 
edges. The spectrum resembled that of a 
type II object with low velocity dis- 
persion. H, He I, and C III were present. 


During a secondary light maximum in spectrograph (180 A/mm). Nova Persei 

December a spectrum was obtained at (1901) has a definitely composite spec- 

18 A/mm providing excellent line profiles trum (sdBe + K), and Nova Lacertae 

of the shallow, broad lines. In January (1910) shows some peculiar, as yet not 

1962 the spectrum was like that 75 days understood, variations in the velocity of 

earlier except that the lines had become He II (X4686) in emission. 

sharper again. This object was extremely Nova Sagittae (1913, 1946) proves to 

complex in light and spectral variations, be a spectroscopic binary with the 

and it might be taken to be a distinct shortest known period: 81}^ minutes. 

subclass of type II. This could be detected only by trailing 

Additional spectra of several of the the star over a long slit without repetition. 

brighter supernovae were obtained by An "S wave" was found in the hydrogen 

Zwicky. From a study of the spectra and emission lines superimposed on the ab- 

the light curves, Zwicky believes that it sorption lines of a white dwarf already 

may be necessary to postulate several new found by Greenstein. Meanwhile, 

types of supernovae in addition to types Krzeminski at Lick discovered that Nova 

I and II. The supernova in NGC 1058 Sge is an eclipsing binary, as well. 

may be a representative of a type inter- Greenstein and Kraft, together with 

mediate between ordinary novae and Jon Mathews of the California Institute 

supernovae. Physics Department, collaborated in 

pointing out the possible importance 

U Geminorum Stars (Dwarf Novae) of Nova Sge as a test for the part of 

An extensive study of several U Gem Einstein's general theory that predicts 

variables at minimum light was continued the existence of gravitational waves, 

by Kraft with the prime-focus spectro- However, since the mass ratio is unknown 

graph of the 200-inch. A spectroscopic in Nova Sge, it cannot be determined 

binary orbit for the emission-line com- whether the star emits a significant 

ponent of Z Camelopardalis was obtained amount of gravitational energy. If 5tli/ 

withP = 6.5 hours; SU Ursa Ma j oris was y&2 ~ 5, for example, the emission of 

found to vary in radial velocity, but a gravitational energy is found to be 30 

period has not yet been determined, times that of the luminosity, and the 

Forty-four per cent of the U Gem stars system would collapse in only 20 million 

that can be reached from Palomar have years. The eclipse period would be out of 

been studied so far; all have proved to be phase by a minute in 15 years — an easily 

binaries with P < 9 hours. A study of the detectable quantity, 

motions leads to <Mv> ~ +9.5 at Another object that may be similar to 

minimum. Nova Sge is the suspected U Gem variable 

An hypothesis was advanced suggesting EX Hydrae, for which both Krzeminski 

U Gem stars are descendants of W Ursa and Kraft find P =» 99 minutes. These 

Majoris binaries. The two kinds of vari- two, together with Herbig's similar object 

ables have comparable space distributions VV Puppis, for which P = 100 minutes, 

and velocities. Plausible arguments on might all be emitters of gravitational 

mass transfer between the components of radiation, 
a typical W UMa system show that it 

might well become a U Gem star after ohell stars 

10 7 to 10 8 years. Further plates of 89 Herculis have been 

obtained by Sargent to continue work on 

Old Novae ^he circumstellar envelope. During 1961, 

A search for binary stars of short period the shortward displaced absorption com- 

among old novae has been started by ponents at the Balmer lines weakened 

Kraft using the 200-inch prime-focus considerably in an interval of less than 60 



days. The other peculiar features — the 
emission in the redward wing of Ha, the 
emission at the intercombination lines 
of the neutral metals, and the shortward 
displaced absorption lines at H and K and 
the D lines — did not change. A prelimi- 
nary curve of growth for the circumstellar 
H lines shows that the turbulent velocity 
is very large — greater than 20 km/sec. 

Mass Loss from Stars with Extended 

In connection with his continuing study 
of mass loss from late-type giants, 
Deutsch has observed the differential 
motions occurring in the atmospheres of 
some M-type supergiants. At 4.5 A/mm, 
spectra of \x Cephei (M2 la) reveal the 
contributions of at least five atmospheric 
layers with distinguishable radial veloc- 
ities in the range —29 to +29 km/sec. 
The radial velocities of individual lines 
correlate with equivalent width and 
excitation potential. Asynchronous veloc- 
ity variations can be clearly seen in three 
of the atmospheric layers. 

For the further elucidation of the mass- 
loss process, Deutsch has under con- 
tinuing observation a number of late- 
type giants which are spectroscopic 
binaries. Several of these exhibit circum- 
stellar lines at the D line as well as at 
H and K. 

A reconsideration of the pronounced 
line weakening found in early-type Mira 
variables has led Deutsch to the con- 
clusion that these stars, like others be- 
longing to the halo population, prob- 
ably have metal deficiencies of the order 
of 10~ 2 as compared with the sun or 
other normal stars. The chief source of 
the opacity in these very cool atmospheres 
remains unidentified, however, and until 
it is known the degree of metal deficiency 
will remain uncertain. 

Segregation of Elements in Magnetic Stars 

The problem of the anomalous abun- 
dance of elements in the magnetic stars 
of type A has been investigated by H. W. 

Babcock. These stars generally show an 
abnormally high abundance of several 
elements such as Cr, Sr, Mn, Si, Eu, and 
other rare earths; further, in the out- 
standing spectrum variables, Eu and 
Cr are observed to undergo large varia- 
tions in antiphase. Whatever the basic 
mechanism of this variation, the process 
by which the particular elements are 
segregated and the manner in which this 
segregation is maintained in the face of 
gaseous diffusion are matters of consider- 
able interest. 

In 1947 it was pointed out that many 
of the anomalous elements in the mag- 
netic stars belong to the iron group or 
the rare-earth group, and that their 
atoms generally have large magnetic 
susceptibility owing to the occurrence of 
partly filled internal electron shells. Each 
such atom has a magnetic moment, fi, 
the effective value of which, measured in 
Bohr magnetons, is expressed by the 
product gM, in which g is the Lande 
factor and M is the magnetic quantum 
number. In a magnetic field H, having a 
gradient VH, the atom will experience a 
force gMvH in the direction of stronger 
or weaker field, depending upon whether 
its alignment is parallel or antiparallel to 
the field. The idea that, as a result of 
this force, a selective paramagnetic 
migration of atoms might occur in a 
stellar atmosphere was set aside because 
in thermal equilibrium the magnetic sub- 
levels will be very nearly equally popu- 
lated. Therefore the net magnetic moment 
would be vanishingly small. 

The phenomenon of "optical pumping,' ' 
recently investigated by microwave phys- 
icists, now offers a nonthermal process by 
which atoms in a magnetic field can be 
polarized; i.e., they can be given a prefer- 
ential orientation. It has been found that 
irradiation of atoms by polarized light, in 
a magnetic field, can drastically alter the 
relative population of the magnetic sub- 
levels. This results in a net paramagnetic 
susceptibility and in a migration toward 
stronger or weaker regions of the field 
provided that the magnetic gradient is 



sufficient. It is also known, as a result of 
recent laboratory investigations, that 
polarized atoms have a rather remarkable 
resistance to disorientation by collisions 
with other atoms. This disorientation 
resistance, according to Princeton investi- 
gators, is particularly marked for spheri- 
cally symmetric atoms — those in S states 
— as compared with those in states having 
orbital anisotropy. 

The possible application of the fore- 
going facts relating to optical pumping 
and disorientation resistance has been 
considered with respect to the abundance 
anomalies of the magnetic stars. For the 
elements of the periodic table, the mag- 
netic moment of the ground state of the 
neutral atom, as well as for the first two 
stages of ionization, has been computed. 
Elements whose ground states are not $ 
states have been rejected by reason of 
insufficient disorientation resistance. Then 
elements of very low astrophysical abun- 
dance have been deleted, and, finally, in 
the accompanying tabulation, all re- 
maining atoms having a magnetic mo- 
ment greater than 2 Bohr magnetons 
have been listed. 






























Mo I 





Eii I 

7, 5, 3, 1 


8, 6, 4, 2 

It is seen that a few atoms best fitting 
the conditions for paramagnetic migra- 
tion in a magnetic-field gradient are Cr, 
Mn, Mo, and Eu. With the exception of 
Mo (which has no prominent lines in the 
commonly observed spectral region), 
these elements are known to show striking 
abundance anomalies in the magnetic 

stars. Indeed, Cr and Eu are the most out- 
standing peculiar elements in the mag- 
netic spectrum variables. This result 
lends support to the initial suggestion 
that migration or segregation of para- 
magnetic elements actually occurs in 
magnetic stars, even though this idea 
seems a priori quite unlikely because of 
the requirement of a large magnetic 
gradient if the paramagnetic force is to 
overcome backward diffusion. If the selec- 
tive migration occurs horizontally, an in- 
crease in concentration of one order of 
magnitude over a distance of 10 ll cm is a 
minimum need. The diffusion equation, 
relating the concentration gradient to the 
selective force on a particular kind of 
atom, then shows that a magnetic 
gradient of at least 10~ 3 gauss/cm is 
required. This is about 10 3 times the 
gradient over a large sunspot. If para- 
magnetic concentration of elements actu- 
ally occurs — and no other theory has been 
offered to maintain the segregation of 
particular elements — this will have a 
decided bearing on possible models of 
magnetic stars. 

Radial Velocities of Magnetic Stars 

As a by-product of the study of stellar 
magnetic fields, accurate radial velocities 
have been derived by H. W. Babcock for 
several A-type stars, most having very 
sharp lines. In nearly all cases the spectro- 
grams, made with the 200-inch telescope, 
have a dispersion of 4.5 A/mm. Velocities 
depend on upward of 20 lines measured on 
each plate by Miss Burd. For some 36 
stars brighter than magnitude 6.5, the 
new radial velocities substantially aug- 
ment the data of the Yale Bright Star 
Catalogue. Of these 36 stars, 25 are found 
to have variable velocity. For most of 
them the range is a few kilometers per 
second. For two of these stars, HD 15144 
(HR 710) and HD 187474 (HR 7552), 
which are evidently spectrosopic binaries 
with periods of 2.9978 days and 700 days, 
respectively, Miss Burd has computed 
orbital elements. 


Stellar Polarization gathering power provides an excellent 

signal-to-noise ratio. 

The first form in which stellar polariza- The new polarimeter is of an uncon- 
tion was detected was the circular and ventional design, developed with the 
elliptical type due to the longitudinal intent of overcoming the effects of stellar 
Zeeman effect on line profiles. It is this scintillation and of the nonuniformity of 
that permits the line-of-sight component the cathode of multiplier phototubes, 
of the star's magnetic field to be meas- These effects have placed a serious limita- 
ured. But, if the direction of the stellar tion on precision of measurement with 
field is essentially perpendicular to the existing stellar polarimeters. Scintillation, 
line of sight, the transverse Zeeman effect closely related to astronomical seeing, is 
can be expected. If the absorption lines an intensity fluctuation in the low- 
are numerous and strong, there can occur frequency range ( < 500 cps) . To over- 
a resultant plane polarization of the light come it, the polarization vector of the 
which has been called polarization by light can be resolved into two orthogonal 
magnetic intensification. Owing to an components that are chopped at a 
imbalance in equivalent width between frequency considerably greater than the 
the "perpendicular" and the "parallel" scintillation frequency and admitted al- 
components of the saturated, Zeeman- ternately to a single, stationary photo- 
broadened profiles, an excess of polariza- tube. The instrument employs a slowly 
tion in a plane parallel to the magnetic rotating electrooptic crystal ( ADP plate) , 
field will result. The integrated effect in a excited by a 3500- volt square wave at 
broad region of the spectrum of a star 2000 cps. The crystal becomes birefrin- 
having a field of a few kilogauss may be gent, with a phase shift alternating be- 
of the order of 1 per cent. Plane polariza- tween +90° and —90° for blue light, 
tion due to magnetic intensification This is followed by a fixed circular 
should be observable not only in sharp- analyzer. The plane-polarized components 
line stars but also in those with lines of star light parallel to the two axes of 
broadened by axial rotation. This kind of the crystal are alternately transmitted as 
polarization should be distinguishable these axes alternate between optically 
from the well known interstellar polariza- "fast" and "slow" at the applied fre- 
tion, due to dust grains in space, because quency. The output of the photomulti- 
that is constant, whereas intrinsic stellar plier is amplified, demodulated, and 
polarization may be expected to show filtered by an amplifier of the so-called 
variations in intensity and in position "lock-in" type, which has already proved 
angle as the star rotates or as its magnetic indispensable in other astronomical in- 
field changes. Observation and analysis struments required to measure a weak 
of intrinsic plane polarization should give signal in the presence of noise, 
valuable supplementary data on which to The ADP crystal is rotated at a rate of 
base models of magnetic stars. about 1 turn in 5 minutes. As a result, the 

A sensitive polarimeter for the investi- demodulated signal produces a sine wave 

gation of such effects has been under on a strip-chart recorder. The amplitude 

development during the year by H. W. and phase of the sine wave are readily 

Babcock, after initial tests with a simple related to the percentage polarization and 

rotating analyzer in front of a photom- position of the electric vector in the light 

eter indicated promise for this approach, of the source. Calibration is accomplished 

The instrument is designed to work at the by inserting a depolarizer followed by a 

prime focus of the 200-inch telescope, tilted glass plate designed to introduce 

which is ideal for the purpose because either 1 per cent or 4 per cent polarization, 

there is only one reflection (nearly Tests of the polarimeter during various 

normal) and because the large light- stages of development have shown that 



the 2-kc/sec modulation frequency and 
the very narrow band width of the lock-in 
amplifier result in a satisfactory signal-to- 
noise ratio, and that a precision of a few 
hundredths of 1 per cent can be obtained 
in measures of the polarization of stars 
brighter than about seventh magnitude. 
Observations of fainter sources are gener- 
ally limited by shot noise (randomness of 
arrival of the incoming photons) rather 
than by scintillation. It is satisfying to 
find that no detectable plane polarization 
is introduced by reflection from the 200- 
inch mirror. An upper limit at present is 
0.1 per cent. 

Observations of some 80 stars have now 
been obtained. Practically all show at 
least a small degree of polarization, and 
it is evident that considerable care will 
have to be exercised to maintain standard 
instrumental conditions and consistent 
calibration in the determination of small 
variations due to intrinsic stellar causes. 
Preliminary results for the percentage 
polarization of a few stars are: WY 
Geminorum 1.75, 9 Geminorum 3.26, R 
Leonis 2.47, XX Ophiuchi 5.15, e Ursae 
Majoris 0.06, a 2 Canum Venaticorum 
0.04-0.11, AD Leonis 0.54-4.16, HD 
153882 0.13-0.34, and HD 154445 4.0. 


A program has been initiated by O'Dell 
to test the feasibility of using the doublet 
ratios of [Ar IV], XX4711, 4740, and [S II], 
XX6717, 6731, in a manner analogous to 
[O II], XX3726, 3729, for the determina- 
tion of the electron density and tempera- 
tures in gaseous nebulae. The former 
ratios should be excellent criteria for 
nebulae of densities above 10 4 electrons/ 
cm 3 , where (O II) becomes insensitive. 
It is hoped that forthcoming theoretical 
calculations will aid in placing the 
calibration on a reliable basis. Measures 
are being made on a number of planetary 
nebulae covering a large range in densities 
with both photographic and photo- 
electric spectrographs on the 60-inch and 
100-inch telescopes. 

Although the spectra of the planetary 
nebulae have been the subject of numer- 
ous photographic studies, substantial 
errors exist in the relative intensities of 
the emission lines, due to the inherently 
small photometric range of the photo- 
graphic plate. With this difficulty in 
mind, O'Dell has begun a systematic 
survey of the bright planetary nebulae 
with the photoelectric scanning spectro- 
graph. Particular attention is being paid 
to emission lines that should be indicative 
of the conditions in these nebulae. Since 
the effect of interstellar extinction can 
become large for even moderately 

reddened nebulae, the interstellar extinc- 
tion is being determined from observa- 
tions of Paschen and Balmer series lines 
of hydrogen arising from the same upper 
level. Although the detailed recombina- 
tion theory of Burgess predicts a small 
variation in the ratio of these Paschen 
and Balmer lines with electron tempera- 
tures, the interstellar extinction correc- 
tion should be much more accurate than 
the result by any other current technique. 
As part of this program, the relative 
energy distribution of the continuum from 
the central stars is also being determined. 

The prints of the National Geographic 
Society-Palomar Observatory Sky Survey 
plates were used by Struve to prepare a 
list of 74 interesting nebulae located in 
and near obscuring clouds in the Milky 
Way. Struve finds that the great nebula 
near Antares extends to an angular dis- 
tance of several degrees from the star. 
The reddish glow appears to illuminate 
the southern edge of the long, opaque lane 
which extends from 22 Scorpii toward 
the east. 

Several faint red and blue nebulosities 
were found in the vicinity of p Ophiuchi 
and CoD -24°12684. Several illuminat- 
ing stars are reduced in brightness by 3 
mag, but no luminous nebulosity could 
be assigned to a star whose light is 
dimmed by more than 3 mag. 



The nature of the emission nebulosity 
near a Scorpii confirms the previous result 
that this object does not coincide with the 
reflection nebula belonging to the same 

There is a pronounced tendency for 
luminous nebulosities (probably of the 
reflection type) to be more frequent in 
the Taurus complex of dense clouds than 
in the Ophiuchus-Scorpius complex. The 
quantity discussed is the number of 
nebulosities per square degree of dark 
cloud. Possibly this is connected with the 
fact that the average absorption of the 
clouds is smaller in Taurus than in the 
Ophiuchus-Scorpius complex. 

Parker has continued his study of S22, 
S147, NGC 6888, IC 443, and the Cygnus 
Loop, which are possible supernovae 
remnants. The observing program, which 
has been completed, includes spectro- 
grams and spectral scans of 16 separate 
filaments in these objects. On the pro- 
gram of reduction, relative intensities of 
all lines, corrected under various assump- 
tions for reddening, have been obtained. 
For the condition of collisional excitation 
and ionization, relative intensities have 
been computed for the emission lines in- 
volved for several values of the electron 
pressure and electron temperature. Com- 
parison of the computed and observed 
ratios will enable statements to be made 
about the temperature, density, and 
abundances in the filaments. Monochro- 
matic net fluxes are also being measured 

for the objects, both to investigate the 
mass of the visible filaments and to inves- 
tigate the amount of free-free radio 
emission that could be expected. The 
dynamics of the five objects has also 
been examined. 

G. Munch and Wilson have prepared 
a reply to the criticism raised by K. 
Wurm against the model of the Orion 
nebula proposed earlier by them. The 
lack of agreement between the radial 
velocities of the He I nebular absorption 
lines and the emission lines at the same 
position, the essence of Wurm's criticism, 
is explained in terms of the density fluc- 
tuations existing in the nebula. Wilson 
and Munch have therefore reviewed the 
variety of observational data related to 
such density fluctuations and, within this 
framework, have discussed the radial- 
velocity data provided by their early 
observations. On the whole, they find that 
the model does not need as drastic a re- 
vision as Wurm proposes. 

G. Munch and Dr. A. Unsold observed 
that the star a Ophiuchi, at a distance of 
25 parsecs, shows a K-line component 
undoubtedly of interstellar origin. The 
observations of other near-by stars in the 
same area of the sky lead them to infer 
that the interstellar cloud in front of a 
Oph has linear dimensions no larger than 
1 parsec. The study of this near-by com- 
plex of interstellar matter, which may 
extend right to the sun, is being continued 
by observing additional near-by stars. 


Structure and Internal Motions of the 

A determination of the solar motion 
and the parameter A of differential 
Galactic rotation from cepheid variables 
is being finished by Schmidt and Kraft. 
The total solar motion is found to be 16 or 
17 km/sec, rather less than the value 21 
km/sec adopted by Blaauw and Morgan. 
The value of A is found to be 14 to 15 
km/sec kpc. No reliable determination 
of the curvature of the angular velocity 

curve can be made. A small negative K 
term is found, but the observations are 
equally well represented by a constant 
error of about —3 km/sec in the radial 

The (U, V, TF)-velocity vectors for 
221 well observed dwarf stars have been 
used by Eggen, Lynden-Bell, and San- 
dage to compute the eccentricities and 
angular momenta of the galactic orbits in 
a model galaxy. It is shown that the 
eccentricity and the observed ultraviolet 



excess are strongly correlated. The stars 
with the largest excess (i.e., the lowest 
metal abundance) are invariably moving 
in highly elliptical orbits, whereas stars 
with little or no excess move in nearly 
circular orbits. Correlations also exist 
between the ultraviolet excess and the 
TT-velocity. Also, the excess and the 
angular momentum are correlated; stars 
with large ultraviolet excesses have small 
angular momenta. These correlations 
have been discussed in terms of the 
dynamics of a collapsing galaxy. The data 
require that the oldest stars were formed 
out of gas falling toward the galactic 
center in the radial direction and collaps- 
ing from the halo onto the plane. The 
collapse was very rapid, and only a small 
number X 10 8 years was required for the 
gas to attain orbits in equilibrium (i.e., 
gravitational attraction balanced by cen- 
trifugal acceleration). The scale of the 
collapse was tentatively estimated to be 
at least 10 in the radial direction and 25 
in the Z direction. The initial contraction 
must have begun near the time of forma- 
tion of the first stars, some 10 10 years ago. 

In connection with the study of the 
collapsing galaxy, mentioned above, 
Eggen has prepared for publication a 
catalogue of some 700 stars whose space 
motions with respect to the sun are almost 
certainly greater than 100 km/sec. In 
addition to astrometric and photometric 
data, the catalogue contains the values of 
the "modular velocities" from which new 
space motion vectors can be computed 
for any future changes in the available 
radial velocity, proper motion, or lumi- 
nosity estimates of these objects. 

The globular cluster system in our own 
Galaxy has been reanalyzed by Arp. The 
globular clusters in the Galaxy are shown 
to be 90 per cent discovered. The best 
available moduli with the new RR Lyrae 
zero point (M v = +0.3, M B = +0.5 mag) 
give a distance to the center of the Galaxy 
of Ri = 9.9 kpc d= 0.5 kpc (minimum 
error for an error of 0.1 mag in RR 
Lyrae absolute magnitude) . 

The analysis of the stars in Baade's 

field near the Galactic center has been 
completed by Arp. There is a definite 
giant branch which clearly emerges in the 
color-magnitude diagram and represents 
the population in the nucleus or nuclear 
bulge. This giant branch is not made up 
of globular clusterlike giants. More 
detailed consideration of this diagram will 
be made shortly. 

The material gathered by Guido and 
Luis Munch for the determination of 
motions and distances of faint OB stars 
in directions near that of the Galactic 
center has been studied and reduced. 
All together, they determined U, B, V 
colors and spectral types for 35 stars and 
radial velocities for the 24 objects of 
higher luminosity from 90 plates obtained 
in a variety of dispersions with the X 
spectrograph and the coude spectro- 
graphs. This material is now being pre- 
pared for publication and will be dis- 
cussed with the earlier results obtained 
by them, as well as with the data for the 
interstellar lines in these objects obtained 
during the past few years. 

Rotation and Internal Motions of Galaxies 

The rather extensive observations of the 
internal motions in the elliptical galaxy 
NGC 3115 obtained by Minkowski and 
Oort in 1958 and extended by Minkowski 
during the following observing season 
were discussed by Oort. The principal aims 
were to find the distribution of mass and 
to obtain data on the distribution of 
random motions in the galaxy. It appears 
that, whereas up to a distance of about 60 
seconds from the center along the major 
axis the distribution of mass seems to be 
roughly the same as that of the light, the 
mass density in the outermost parts 
observed decreases more slowly than the 
light density. The ratio of mass density to 
light density, M/L, expressed in terms of 
the mass and light of the sun as units, is 
about 15 in the inner parts, and rises to 
values of 100 in the outer shells. For the 
entire galaxy, M/L was found to be 
about 60. Such a high value has never yet 
been found from observations of rotation, 



presumably because they have never 
reached such large distances. The result 
is important because of its bearing on the 
problem of stability of groups and 
clusters of galaxies. 

The preliminary reduction of the Ha 
plates taken by Brandt with the Mount 
Wilson B spectrograph for the purpose of 
redetermining the rotation curve of M 33 
is complete. The rotation curve deter- 
mined optically extends to 30 minutes 
from the center of M 33, but it can be 
extended 60 minutes from the center 
with the aid of the radio observations. 
The shape of the rotation curve is very 
similar to that of M 31 ; hence, the relative 
mass and density distributions should be 
similar. For an assumed distance of 630 
kpc, the approximate mass becomes 
2.3 X 10 10 Mo. 

Emission Nebulae in Galaxies 

Work done by Baade on the precise 
position of all emission nebulae in M 31 
is being prepared for publication by 
Arp. An effort will be made to give illus- 
trative data on the connection between 
these emission nebulae and the spiral 
structure in M 31. 

Another attempt was made by Schmidt 
to determine the helium abundance of the 
interstellar gas in H II regions at various 
positions in the Andromeda galaxy. The 
observations were planned to furnish 
absolute values of the helium-to-hydrogen 
abundance ratio and the diameters of 
both the hydrogen emission region and 
the helium emission region. The first 
objective was attained by plates taken 
in the Orion nebula, where Mathis has 
obtained an absolute calibration. Expo- 
sures on the H II regions in Andromeda, 
of which one was a multinight exposure, 
were carefully guided with the spectro- 
graph slit over the H II region and a 
near-by star. The star spectrum allows 
evaluation of effects of seeing, guiding, 
and scattering in the photographic plate, 
and the diameters of the emission regions 
are corrected for these effects. Once 
these diameters are determined for both 

hydrogen and helium emission, the 
abundance ratio can be determined 
independent of a possible variation with 
abundance ratio of the far ultraviolet 
continua of the exciting stars. The obser- 
vational problem is a marginal one for the 
200-inch prime focus spectrograph, and 
the photographic photometry on the 
plates is difficult. Results thus far give a 
helium-to-hydrogen number ratio of 0.08 
at 89 minutes from the center of M 31, 
0.14 at 70 minutes, and 0.17 at 25 
minutes. Each determination is uncertain 
by a factor between V/2 and 2, so that the 
variation as well as the difference from 
Orion nebulae (0.13) is hardly significant. 
The implication of an essentially constant 
helium abundance would be far reaching, 
as was briefly indicated in the Annual 
Report for last year. 

The observation of the emission lines of 
[O II] in the nuclear region of M 31 was 
continued by G. Munch. An area approxi- 
mately rectangular with dimensions 100 
X 300 seconds centered at the nucleus 
has now been covered with slits at various 
position angles. The emission lines at 
greater distances from the nucleus be- 
come so faint that it is impractical to 
attempt to follow them with existing 
equipment. The material so far collected, 
therefore, will be discussed shortly and 

Munch's lack of success in detecting 
emission lines in the patches which were 
observed by Baade in the disk of NGC 
4594, and which Lindblad considered to 
be H II regions, was reported last year. 
Confirming the tentative explanation 
then given, it has been found that an Ha 
exposure through an interference filter, 
with the /:3.67 focal ratio of the 200- 
inch, does not show such features with 
stronger contrast than broad-band photo- 
graphs do. The patches undoubtedly, 
then, do not have an emission spectrum, 
and their true nature raises an important 
problem. Further work on these objects 
is being planned with a photoelectric 

A systematic search for planetary 



nebulae in other galaxies in the local 
system is being executed by O'Dell, by 
means of photographic plate plus filter 
combinations that give sensitivities 
around the Ha, Nl + N2, and X5400 
regions, since the nebulae should appear 
in the strong emission regions and not in 
the relatively line-free region at X5400. 
Thus far, four suspected nebulae have 
been found in the Leo I system. 

A program of direct photography of 
nearly resolved, Irr, Sc, and late Sb 
galaxies whose redshifts are less than 
2000 km/sec was begun by Sandage using 
a 4 by 4 inch Ha filter of 80 A half -width. 
The purpose is to isolate the H II regions 
and to measure their apparent diameter 
as distance indicators. Absolute calibra- 
tion of the linear diameter will be made 
using galaxies in the local group, such as 
the Large and Small Magellanic Clouds, 
M 33, IC 1613, and NGC 6822, whose 
distances are known from the cepheid 
variables. Preliminary calibration during 
the report year gave the linear diameter 
of the largest H II region as 245 parsecs 
and the mean diameter of the first five 
largest as 175 parsecs. Plates were ob- 
tained of NGC 2403, M 101, NGC 925, 
NGC 2903, and M 51 in a first trial of the 
problem, and it will be feasible to measure 
diameters in galaxies with redshifts as 
large as 2000 km/sec. The ultimate aim 
of the problem is to improve the values of 
the Hubble constant (H). Preliminary 
results using H II sizes determined in 
NGC 925 and NGC 4321 show that H 
lies in the neighborhood of 75 to 100 
km/sec 10 6 psc, in agreement with Sersic's 
earlier study of the H II region problem 
using blue plates of galaxies taken with 
the 200-inch, and in agreement with the 
value of 75 km/sec 10 6 psc which has 
been generally used during the past 
several years. 

Widened spectra of NGC 1068 and 
NGC 4151 have been obtained by 
Sargent at a dispersion of 86 A/mm. 
Exposures of stars whose continuous 
energy distribution have been obtained 
with the Cassegrain scanner have been 

made on the same plates. Line profiles 
have been measured for NGC 1068. In 
NGC 4151 the broad wings at the Balmer 
lines, which were reported by Seyfert, 
extend to ±5000 km/sec. There are no 
absorption lines in NGC 4151 that can 
be attributed to stars. Two strong absorp- 
tion features which occur at X3885 and 
X3732 and which are probably to be 
identified with He I are displaced by 
about —200 km/sec relative to the 
emission line centers. NGC 4151 is seen 
almost face-on, and this indicates that 
material is being expelled from the 
nucleus along the direction of the axis of 

Variable Stars in Galaxies 

Sandage completed the photoelectric 
measurements of selected stars of Baade's 
variable star sequences in IC 1613 in 
order to put Baade's extensive photo- 
metric material of the cepheids in this 
galaxy on the Progson scale. The photo- 
electric sequence extends from B = 12 m 
to B = 22™0, which is not yet faint 
enough to correct the cepheid photom- 
etry at minimum light but is faint 
enough to correct all Baade's data at 
maximum light. This has been done, with 
the result that the slope of the period- 
luminosity relation for IC 1613 is identical 
with that found by Arp for the Small 
Magellanic Cloud. The relation is 

J5 max = 22.61 - 2.25 log P 

which gives an apparent modulus of 
(m — M) B = 24™33, using Kraft's zero- 
point calibration of the cepheids in 
galactic clusters. 

Extensive U, B, V photometry of field 
stars in the direction of IC 1613 indicates 
a reddening due to our Galaxy of E(B — 
V) = 0?03, which gives (m - M) = 
24 m 2 for the true modulus of IC 1613. 
Sandage expects to extend the photo- 
electric sequence to B = 23 ni 5 next 
season. All Baade's material will then be 
published on this photometric scale. 

Extensive material on three variable- 
star fields in M 31, accumulated by 


Baade during the past ten years, is being to be nonvariable. For 22 variables in the 

analyzed by Miss Swope. There are same area, periods could be derived from 

about 120 variables in a field 15 minutes the brightness estimates. Among these 

of arc from the nucleus, 330 variables in 22, 16 belong to Bailey's type (a + b). 

a field 45 minutes from the center, and The mean period of these variable stars is 

more than 50 that are in a field 96 P = 0"?617, almost the same mean period 

minutes south preceding the nucleus. This as Miss Swope found in the Sculptor-type 

last field is the only one which is suitable system in Draco. For 6 variable stars, 

for precision photometry and in which van Agt determined periods shorter than 

there is a photoelectric sequence (Arp, 0?45, but, since the scatter in the esti- 

Year Book 58). In this field there are 20 mated brightness relative to the bright- 

cepheids with both photographic and ness-amplitude is larger for variable stars 

photovisual light curves and also a pre- with smaller amplitude, these periods are 

liminary color-magnitude diagram. Miss established with less certainty. For 4 

Swope finds that the apparent distance variable stars in the group of 28 it has not 

modulus of M 31 is B = 24™75 or V = yet been possible to determine a period. 

24^60, assuming a zero point of the One bright variable star at the northern 

cepheids from Arp and Kraft's period border of the system was also estimated 

luminosity curves. There is probably by way of the eyepiece method. From a 

about 0™15 general reddening due to our limited number of plates, the period of 

Galaxy, which gives a distance modulus this star was determined to be 2?697. A 

corrected for absorption of 24™ 15 in both rough estimate of the maximum and 

B and V for M 31. minimum B magnitude was max = 18.2, 

The slope of the period-luminosity min = 18.7. Scale transfers for blue plates 

curve is essentially the same as that only are available for the system in Ursa 

found by Arp for the Small Magellanic Minor. From a scale transfer of Baum's 

Cloud. sequence in M 13 a provisional magnitude 

Baade also obtained a series of plates of sequence in the Sculptor-type system 

the Sculptor-type system in Ursa Minor, could be established. The median magni- 

These plates are being assessed by Dr. tude of 8 variable stars the periods of 

S. L. Th. J. van Agt of the Leiden which are known and for which the 

Observatory. In this system 91 variable brightness estimates were transferred into 

stars are known. They were discovered on magnitudes is 20™04. Because of the 

plate pairs, each of which covered only scatter in the light curves of the 8 

part of the system. The outlines of the variables, all within 6 minutes of the 

Sculptor-type system as inferred from center of the system, this mean value of 

the positions of the variable stars are the median magnitudes is still rather 

roughly elliptical and indicate dimensions uncertain. 

on the sky of 60 by 22 minutes. About 50 All stars visible on a plate with an 

per cent of the variable stars discovered 103aO emulsion and exposed with a WG2 

in this system are concentrated on the filter in front of the plate were measured 

plates that cover the center of this in the iris photometer of the Leiden 

Sculptor-type system. The number of Observatory. In addition, the same stars 

plates is large enough to determine the were measured on a plate with 103aD 

periods of the cluster-type variable stars emulsion and exposed with a GG11 filter, 

in the central area only. Estimates of the This plate was exposed immediately 

magnitudes for 28 stars in the central after the first one mentioned. No scale 

area were obtained by way of inspection transfers for either of these two plates 

with an eyepiece and comparison with exist. The iris scale readings were used, 

a sequence of close-by comparison stars, therefore, to construct a pseudo-color- 

Two of the investigated stars turned out magnitude diagram for the central part 



of the Sculptor-type system. All stars 
used in this diagram are located within 
305 seconds of the center of the plates 
measured. These plates cover the center 
of the Sculptor-type system, and, since 
the center of the system and the plate 
centers are very close together, the 
pseudo-color-magnitude diagram repre- 
sents the general trend in the central part 
of the system. The pseudo-color-magni- 
tude diagram shows a distinct giant 
branch and a clearly developed horizontal 
branch. In the region where the horizontal 
branch meets the upgoing giant branch, 
the number of stars is considerably lower 
than in the part of the horizontal branch 
immediately before that region. To the 
blue end of the horizontal branch only 
one star was found. The ratio of the 
number of stars at the blue end and at the 
red end of the horizontal branch is about 
1 : 50. The position of the gap of the short- 
period cluster-type variable stars is such 
that only the one blue star mentioned is 
found to the blue side of the gap. 

The faint part of the giant branch sets 
in at about 1% m &g below the horizontal 
branch. At the point of bifurcation there 
is no clear indication of a doubling of the 
giant branch. The plates used for 
the construction of the pseudo-color- 
magnitude diagram were exposed under 
conditions of poor seeing. Much of the 
scatter in the branches of the diagram is 
due to the low quality of the two plates. 
The scatter outside the branches is low; 
therefore, we may conclude that the 
number of faint field stars entering the 
diagram is limited. 

Photometry and Stellar Content and 

The distribution of stars in a number of 
local group dwarf galaxies was studied 
by Hodge. These include the Fornax, 
Sculptor, Leo I, Leo II, Ursa Minor, 
Draco, NGC 147, NGC 185, and NGC 
205 systems. In all, the projected density 
was found to obey Hubble 's law in the 
central regions but to fall off more steeply 

in the outer regions. A cutoff is found 
which corresponds in location to the 
expected tidal cutoff. Photoelectric sur- 
face luminosity distributions in two 
colors were obtained for other dwarf 
galaxies, specifically the irregular systems 
NGC 6822, IC 1613, Sextans, and WLM. 

Photoelectrically calibrated measures 
of the distribution of luminosity and 
color have been made by Hodge for 29 
galaxies primarily of the SO type. The 
results, not yet completely reduced, show 
a clear difference in physical properties 
between Sandage's subgroup in this class 
as given in The Hubble Atlas. 

Two near-by dwarf galaxies were ob- 
served photoelectrically by Baum during 
the report year. The integrated light of 
these very faint tenuous systems must 
be measured by making a series of slow 
scans across them with a photoelectric 
photometer. For this purpose the prime- 
focus photometer at the 200-inch has 
been equipped with a scanning motor. 

The purpose in observing these dwarfs 
is to obtain their color indices and 
absolute magnitudes so as to fit them 
into a color-luminosity diagram for ellip- 
tical galaxies. This diagram divides 
itself into two color-index groups, the 
large elliptical galaxies having color in- 
dices 0.2 to 0.3 mag redder than the dwarf 
systems. All evidence indicates that the 
dwarfs are truly Population II, whereas 
eight-color observations show that large 
ellipticals must be mainly old Population 
I. The newly observed dwarfs add two 
points where they are most needed in the 
difficult part of the color-luminosity 
diagram. One of these, NGC 6822, is 
clearly in the true dwarf class, whereas 
the other, NGC 185, is in the transition 
region between classes. 

Any interpretation of the redshift- 
magnitude relation in terms of cosmologi- 
cal models depends on the way in which 
evolutionary effects are taken into ac- 
count. Since individual stars undergo 
large changes in luminosity and tempera- 
ture as they age, the integrated light of a 
galaxy will tend to change with time. 


Owing to the light-travel time, distant center of an arm, weakest near the inner 

galaxies are seen at an earlier age than edge of an arm, and intermediate near the 

near-by ones. It is therefore necessary to outer edge. The wavelength dependence 

know something about the stellar content is such that asymmetry in the dust lanes 

of galaxies of the kind used for redshift- does not by itself appear to be an ade- 

magnitude work. quate explanation. Evidently the relative 

Photoelectric observations on the eight- number of early-type stars tapers off less 
color system were used in 1958 for com- sharply at the outer edge than at the 
puting a population model for large inner edge. Such would be the situation 
elliptical galaxies. At 4830 A the model if the rotating disk of the galaxy slips 
could be described as 25 per cent Popula- through the arms, dragging them slightly 
tion II plus 75 per cent old Population I. and coiling them up. 
During the current year, a first attempt The evolution of the integrated proper- 
was made by Baum to extend this eight- ties of clusters of stars has been computed 
color population analysis to sample by Arp. The total colors and magnitudes 
regions in spiral galaxies. Experimental of groups of stars of different ages have 
results were obtained by scanning across been derived. The results indicate: (1) 
selected regions of M 74 (NGC 628), The E and SO galaxies contain about 30 
which is a face-on Sc spiral of unusually per cent more K7 dwarfs (or equivalent) 
good symmetry. As expected, the disk by number than galactic or globular 
population underlying the inner spiral cluster populations. This conclusion sup- 
arms has a spectral energy distribution ports evidence from M/L ratios and 
roughly similar to that of an elliptical computations by Spinrad on the com- 
galaxy. The outer regions of the disk, posite spectra. (2) The total increase in 
however, are evidently bluer. The arms magnitude from a "young" E or SO 
themselves, although photographically galaxy to brightness at 10 10 years is 
impressive, contribute astonishingly little, about 4 magnitudes. The brightness-age 
perhaps 10 per cent photo visually, to the curve also enables the evolutionary 
total light of M 74. The width of an magnitude correction for large redshifted 
individual arm is found to increase nebulae to be read off. This completely 
systematically with wavelength, the in- empirical method yields the same K 
frared width being about 1.5 times the correction as was computed by Sandage 
ultraviolet width. for the latest cosmological solution from 

The most interesting feature of Baum's the redshift data. (3) By classifying 

scans is a slight asymmetry in the color galaxies in the two nonevolving param- 

distribution across a spiral arm. This eters of mass and angular momentum 

provides a clue to the solution of an old it is shown that evolution probably does 

dilemma. Since its birth, a typical spiral not take place from one type of galaxy 

galaxy like M 74 has had time for more into another. It is suggested that only 

than 50 revolutions of its disk, but its relatively slowly rotating masses contract 

arms are coiled to the extent of less than into E0 giants, and higher rotation in the 

two visible turns. If the arms rotate with more-flattened galaxy types sets in- 

the disk, the outer ends of the arms must creasingly smaller mass limits. Groups 

share most of the rotation. If, on the other and clusters of galaxies are considered, 

hand, the outer ends of the arms are tied and it is shown that a high density (in a 

to intergalactic magnetic fields, the disk cluster of galaxies) is strictly correlated 

must slip through the arms, which are with E and SO membership, and low 

zones of new-star formation, with only a average density with high membership, of 

small drag. The clue is this: Relative to spirals. This gives a physical explanation 

other colors the strengths of violet and for Hubble's often-made statement that 

ultraviolet are found to be greatest at the spirals tend to be field nebulae. The above 



observations on cluster composition are 
interpreted as confirmation of the hy- 
pothesis about formation. 

Recomputation by Arp of Seare's early 
work using modern magnitude scales 
yields a surface brightness of the Galaxy 
in the solar neighborhood of 8b = 23.8 
mag/sq sec (perpendicular to plane with 
absorption layer allowed for). The com- 

parable surface brightness occurs 11 kpc 
from the center of M 31, assuming cose- 
cant reddening models for both galaxies. 
It is also shown that with such reddening 
models the conclusion of Kron and Mayall 
is incorrect and that the M 31 globular 
clusters are, in fact, the same intrinsic 
color as in our own Galaxy. The following 
comparison has been made: 

Mass, Mq 

Radius (to solar-brightness isophote), kpc 

Number of clusters 

Per cent of gas 


34 X 10 10 




Milky Way 

7 X 10 10 




All these characteristics indicate that 
our own Galaxy is more like an Sc than 
M 31 is, and of course the Milky Way is 
more like an Sb than M 33 is, but it 
seems difficult to classify our own system 
more quantitatively at present. 

Work is being continued by Oke on the 
measurement of absolute energy distribu- 
tions in the spectra of the central regions 
of galaxies. Measurements are now com- 
plete between X3400 and X6000 for about 
20 galaxies. Nearly all observations have 
been made on giant elliptical systems. 
The two brightest galaxies and two 
fainter ones in the Coma cluster have 
been observed. The data are being used to 
compute K corrections for distant galaxies 
and for studying the stellar content of 

Catalogue of Galaxies and of Clusters 
of Galaxies 

Volume I of the Catalogue of Galaxies 
and of Clusters of Galaxies by Zwicky, 
Herzog, and Wild was published in 
October 1961. In the meantime, work on 
volume II has progressed to the point that 
the data on all clusters (about 2500) 
have been analyzed and prepared for 
publication, and the work on the galaxies 
involved is expected to be finished in 
October 1962. Volume II covers the area 
from Milky Way to Milky Way between 
the declinations +15° and +35°. Volume 
III, which covers the south galactic cap 
north of declination —3°, has been 

started. This project was supported in 
part by the National Science Foundation. 

The data of the catalogue have been 
reduced in a dozen different ways and 
have been used to test for interstellar and 
intergalactic absorption. The fact has also 
been confirmed that clustering among 
galaxies is universal and is statistically 
of the same nature at all distances up to 
redshifts corresponding to symbolic veloc- 
ities of recession of the order of 100,000 
km/sec. It has been further confirmed 
that no clusters of clusters of galaxies 

The relative areas of the sky that are 
covered by open, medium compact, and 
compact clusters of galaxies have been 
investigated by Zwicky and Rudnicki. 
The largest clusters are all of the same 
linear size, independent of type and dis- 

The so-called "cluster cell," every one 
of which contains the equivalent of one 
large cluster of galaxies, on the average 
was found to have a diameter of 45 
million parsecs, assuming a redshift 
constant of 100 km/sec per million 

In the catalogue of galaxies by Zwicky, 
Herzog, and Wild the peripheral contours 
used for the delineation of clusters of 
galaxies are the isopheths or equal- 
population contours along which the 
numbers of galaxies per square degree are 
equal to about twice the number of 
galaxies per square degree in the adjacent 



fields. The clusters thus drawn, not in- 
cluding the near-by Virgo cluster, how- 
ever, cover 13 per cent of the 3024 square 
degrees of the sky included in volume I 
of the catalogue. 

In the course of the work on volume II 
of the catalogue it was found that the 
field (of 36 square degrees) centered at 
R.A. ll h 17 m and decl. -r-35°30' (1950) is 
the richest field of galaxies and of clusters 
of galaxies observed thus far, containing 
about 150,000 galaxies and 113 clusters 
of galaxies as counted on a limiting 103aE 
plate (+ red filter) obtained with the 48- 
inch schmidt telescope. 

Internal Motions of Clusters of Galaxies 

About 60 spectra of galaxies in the 
cluster CI 0123-0138 have now been 
photographed by Zwicky, and the follow- 
ing results have been derived: (1) The 
average symbolic velocity of recession is 
V s = 5321 km/sec. (2) The dispersion 
in velocities (radial) is AV S = 406 km/sec. 
(3) The dispersion in V s is essentially con- 
stant from the center of the cluster to the 
periphery, a fact indicating that the 
cluster is stationary and is neither ex- 
panding nor contracting. (4) The disper- 
sion in V s is greater for the fainter 
galaxies than for the brighter ones, 
although the average values of V s are the 

The distribution of the galaxies within 
the cluster indicates an elliptical shape 
of the cluster. There is no indication of 
any rotation, however, from the analysis 
of the radial-velocity data. 

Additional spectra have been obtained 
of member galaxies of clusters in Cancer, 
Hydra I, and the Coma cluster for 

determining velocity dispersions and 
mass-luminosity ratios. 

Redshift-M agnitude Relations 

In principle, there are several observa- 
tional tests for distinguishing between an 
exploding universe and a steady-state 
universe. With methods available today, 
the best test is the relation between the 
redshifts and the distances of large 
clusters of galaxies. More exactly, the 
observable parameters are the redshifts 
and the apparent bolometric magnitudes 
of representative cluster members. 

Photoelectric observations collected by 
Baum since 1955 have been mentioned 
in previous Year Books. The absolute 
amounts of energy received from various 
galaxies are measured photoelectrically 
in eight colors ranging from ultraviolet 
to infrared. Effective wavelengths and 
bandwidths of the eight colors are: 

Color Effective X, A Bandwidth, A 

Ultraviolet 3730 500 

Violet 4335 740 

Blue 5065 430 

Green 5525 470 

Red 6705 850 

Infrared I 7525 600 

Infrared J 8520 800 

Infrared K 9875 1100 

When the resulting spectral-energy dis- 
tributions of galaxies are compared, the 
displacements between them yield both 
their redshift and their relative bolo- 
metric magnitudes. In this way, the red- 
shift-magnitude relation has been ex- 
plored to a much greater distance than 
before. Final photoelectric values for the 
three observed clusters of largest red- 
shift are as follows: 

0024 + 1654 



Symbolic Velocity 
cAX/X, km/sec 


Probable Error 
in Redshift, % 



Cluster 1410 + 5224 is the cluster found 
by Minkowski at the position of Cam- 
bridge radio source 3C295. An optical 
emission line at X5448, presumed to be 

O II 3727, provides a good check on the 
redshift above. 

When these photoelectric redshifts were 
reported earlier, the corresponding bolo- 



metric magnitudes could not be specified 
with the fullest attainable certainty, and 
the apparent shape of the redshift- 
magnitude relation had to be taken as 
tentative. Observations by Baum during 
the report year have been devoted to 
resolving this difficulty. The uncertainty 
arose, not because of limited precision 
in the photometry, but because auxiliary 
data were needed for intercomparing the 
magnitudes in one cluster with those in 
another in the best possible way. During 

the report year, photoelectric and photo- 
graphic observations have been collected 
for constructing the needed luminosity 
functions of seven key clusters of galaxies 
distributed in redshift from AX/X = 0.02 
to AX/X = 0.44. The new photoelectric 
sequences include 57 additional galaxies, 
many of them relatively faint. As before, 
most of the observations were made with 
the pulse-counting photometer at the 
200-inch prime focus. 


Schmidt has engaged in a spectroscopic 
investigation of galaxies believed to be 
connected with radio sources. This pro- 
gram is planned in close cooperation with 
Dr. Thomas Matthews, who made most 
of the identifications. Galaxies connected 
with radio sources 3C33, 88, 98, 171, 198, 
219, 234, 317, 433, 445, from the third 
Cambridge Catalogue, and Coma A were 
investigated. All spectra show emission 
lines, the number ranging from 1 to 13. 
The magnitudes are in the range 15 to 20; 
the redshifts (AX/X) vary from 0.03 to 
0.24. The absolute magnitudes of all 
galaxies are close to —20. The largest 
number of emission lines is seen in 3C234, 
the spectrum of which resembles that of a 
planetary nebula of high excitation. In a 
case like this, the spectrum constitutes a 
strong confirmation of the identification. 
For 3C88, where the spectrum shows 
weak X3727 emission only, the spectrum 
has hardly any confirmatory value. 
Spectra taken of some three or four other 
galaxies identified with radio sources 
show no emission features. At least one of 
these is now known to be a misidentifica- 
tion. It may be estimated conservatively 
that 70 per cent of the galaxies fainter 
than 15 mag that are identified with 
bright radio sources show emission lines 
in the spectrum. A spectrum of the 
galaxy identified with the distant radio 
source 3C295, observed and discussed 
earlier by Minkowski, shows that the 

X3727 emission relative to the continuum 
is rather weak in comparison with that 
observed in some other radio sources. 
Spectroscopic work on some stellar ob- 
jects believed to be connected with radio 
sources is continuing. 

Greenstein obtained spectra of the 
radio source 3C442, which is a double 
elliptical, with only X3727 emission. A 
common feature of the radio galaxies in 
Herculis A, 3C278, and 3C442 is a low 
gradient of surface brightness. The sys- 
tems seem to be ellipticals, but their 
brightness distribution optically is more 
like that of an Sc pole-on system. 

Further spectra of the radio star 3C48 
reveal no changes, and one in the visual 
region showed two more unidentifiable 
emission lines. He II, X5411, was absent. 

Extending the work reported last year, 
Dr. Thomas Matthews and Sandage 
identified two additional radio stars 
similar to 3C48. This brings the total of 
such identifications to three. The objects 
are 3C48, 3C196, and 3C286. The radio 
positions of all three sources were deter- 
mined before optical identification was 
made, and the only object within the 
error rectangle of the radio position is 
stellarlike in each of the three cases. The 
agreement of the radio and optical 
positions is remarkable, being within 4 
seconds of arc in all objects. Photoelectric 
photometry of the stars shows that each 
has very unusual colors. The photometry 



gives V = 16»20, B - V = 0M0, 
U - B = -0 m 59 for 3C48; V = 17^79, 
B - V = m 57, U - B = -0^43 for 
3C196; and V = 17 m 25, B - V = m 26, 
[/ _ ^ = -0?91 for 3C286. Spectra of 
3C196 and 3C286, obtained by Schmidt, 
also show that the stellar objects are 
peculiar and uniquely new. 

The optical flux of 3C48 was found to 
vary from V = 16^02 to V = 16 m 44 over 
the 13-month observation period since its 
initial discovery. The time resolution of 
Sandage 's observations is not great, and 
so nothing is known about variations in 
the order of minutes or hours, but the 
flux does vary from night to night. 
Special observations at radio frequencies 
by Matthews showed that the radio flux 
is constant to within the probable error 
of the determination even though the 
optical flux varies. 

The optical U, B, V measures for the 
three sources were transformed to abso- 
lute flux units and compared with the 
radio data. For 3C48 and 3C196 the 
power spectrum computed from the 
theory of synchrotron radiation was 
shown to predict the U, B, and V values 
to within a few hundredths of a magnitude 
when all but one of the adjustable 
parameters of the theory are determined 
from the radio data. The remaining 
parameter is the critical frequency at the 
high-energy cutoff, and this could be 
adjusted for the excellent fit. This fit 
bridges a gap of more than 20 octaves of 
the power spectrum. But Matthews and 
Sandage were not convinced that this 
agreement shows that the optical radi- 
ation is necessarily due to synchrotron 
emission alone, and the question remains 
open for the future. If the optical flux 
were due to synchrotron emission, the 
energy of the relativistic electrons respon- 
sible for the radiation would be about 

5 bev in a field of 1 gauss, or 50 bev in a 
field of 10~ 2 gauss. 

Additional data at hand suggest that 
future identifications of radio stars can 
be expected, and many of the questions 
raised by these unique objects will 
undoubtedly be better understood in the 
near future. 

One of these sources, 3C48, studied by 
Sandage has also been observed photo- 
electrically by Baum on the eight-color 
system used in connection with the 
redshift-magnitude program. This obser- 
vation permits the optical energy distri- 
bution to be investigated over a broader 
range extending from the ultraviolet to 
the infrared. The eight colors were found 
to fit a slope of —2.25 db per 10 14 cps. 
In absolute terms, the observed flux of 
3C48 at 5490A (5.46 X 10 14 cps) on two 
nights in December 1961 was found to be 
1.04 X 10" 29 watt/m 2 per cps. 

Over the past few years, plates of the 
Crab Nebula have been taken by G. 
Munch with the object of following 
changes in the large-scale structure of the 
amorphous mass emitting synchrotron 
radiation. Particular attention is being 
given to the moving ripples, which at 
irregular intervals appear near the hypo- 
thetical central star. On February 2, 1962, 
a diffuse wisp, about 4 seconds long, at a 
position angle 45° and at about 2 seconds 
distance from the nuclear star, was 
detected for the first time. The other 
moving wisp discovered by Baade and 
discussed by Oort appears somewhat 
farther out. Poor weather and lack of 
observing time prevented following the 
development of this ripple, but its 
observation at a smaller distance from 
the central star points to the likelihood 
that the central star is still very active 
in injecting into the nebula large numbers 
of relativistic particles. 


Stellar Atmospheres and Oke for a range of effective tempera- 

Line profiles of H7 have been computed tures and surface gravities corresponding 
on the Kolb-Griem-Shen theory by Searle to those of F-type stars of different lumi- 



nosity classes, and compared with ob- 
served profiles for (1) the cluster-type 
variables RR Lyrae and SU Draconis, (2) 
two F-type subdwarfs, and (3) F-type 
stars of normal metal abundance. The 
observed and computed profiles are in 
excellent agreement, and the line profiles 
can be used as temperature indicators 
independent of reddening. A comparison 
of the derived Hy temperatures with 
those determined by fitting observed 
absolute energy distributions to fluxes 
computed from model atmospheres shows 
good agreement, provided allowance is 
made for interstellar reddening. 

For the normal metal stars of spectral 
type later than F5, the continuum at H7 
is depressed by line blanketing to such an 
extent that a temperature determination 
by H7 fitting is not possible. For late F- 
and G-type stars, however, Ha profiles 
remain a practicable temperature indica- 
tor, and computations of a grid of Ha 
profiles have been completed. In the late 
G-type stars, the Ha profile is dependent 
not only on temperature but also on 
metal-to-hydrogen ratio. 

Weidemann has analyzed the hydro- 
gen-line profiles and the colors of normal 
white dwarfs of spectral type DA. The 
Kolb-Griem-Shen theory was used, and 
allowance was made for the dependence 
of pressure on depth. The profiles so com- 
puted were essentially independent of the 
H/He ratio and covered a wide range of 
surface gravity and temperature. The 
maximum of the intensity of the hydrogen 
lines is shifted about 4000° hotter than in 
main-sequence stars, and the maximum 
in the Balmer discontinuity about 2500°. 
The effects of the lines on U, B, V colors 
and of the reemitted blanketed radiation 
were estimated. The temperature scale 
derived is close to an earlier estimate by 
Greenstein. Weidemann finds that the 
DA stars range from 18,000°K to 7300°K 
in effective temperature, log g from 7.3 to 
8.5, and masses from 0.25 to 0.S5M o. The 
problem of the DB stars, which show only 
He I lines, was considered briefly; it was 
found that a very high He/H ratio is 

needed to alter the opacity source from 
pure H. Therefore, only if He/H > 10 3 
will the H lines disappear and be replaced 
for hot white dwarfs by He I lines. 

Wright has reported extraordinarily 
low excitation temperatures for normal 
A0 stars. Jugaku explored possible theo- 
retical explanations for this phenomenon, 
which he found also to be present in 
Hunger's analysis of a Lyrae, which gives 
only 5350° for T exc . Jugaku has computed 
the effects of stratification in depth of line 
formation, which corresponds to only 300° 
temperature change, and the model-at- 
mosphere effects which predict T exc = 
8460°K for T eii = 9500°— again too small 
a temperature change to account for the 
low observed T eKC . A possible explanation 
is a large deviation from local thermody- 
namic equilibrium, and another is in- 
crease of turbulent velocity for lines of 
low excitation potential. 

In metal-poor stars of low T, Rayleigh 
scattering becomes important, as Traving 
pointed out for the analysis of globular- 
cluster red giants and as Greenstein and 
Wallerstein find for the metal-poor field 
red giants like HD 122563. Jugaku evalu- 
ated the contribution of metallic absorp- 
tion continua, finding them to be small 
compared with H~ and Rayleigh scatter- 
ing. There is no direct observational sup- 
port for Rayleigh scattering, since it 
would produce a steeply wavelength- 
dependent depression of the blue and 
ultraviolet (i.e., a U — B deficiency in 
metal-poor stars). 

Nishida has studied the problems of the 
structure and evolution of the helium 
stars. A series of models for a helium star 
of 1 solar mass, both in the helium burn- 
ing phase and in the carbon burning 
phase, have been constructed. Results 
show that the evolutionary track lies 
near the location of nuclei of planetary 
nebulae in the H-R diagram, but that it 
does not move to the right across the main 
sequence of the Population I stars. Cal- 
culations of the stellar models for the 
following problems are in progress using 
the IBM 7090: (1) Evolution along the 


horizontal branch of the Population II to be that deficient. Thus, the birth 
stars. (2) Stellar models for stars with luminosity function in the solar neighbor- 
very small masses (less than O.lMo). (3) hood must have been different at earlier 
Evolutionary tracks in the H-R diagram times, specifically such that in the early 
for stars that gravitationally contract to stages of the Galaxy the rate of formation 
become white dwarfs. of stars of 10 solar masses, relative to that 

G. Munch and Dr. R. Kippenhahn of of stars of 1 solar mass, was about 10 or 

the Max Planck Institut fur Astrophysik 20 times larger than it is at present, 
in Munich, Germany, are studying the 

spectrum of late F-type supergiants to Stellar Dynamics 
verify whether the Balmer lines can be Theory predicts that the only exactly 
explained on the basis of models with a steady states of an unrelaxed stellar sys- 
unique temperature as function of depth tern which shows star streaming directed 
in the atmosphere. Suggestions have been to and from the center are those of Ed- 
made in the past to explain a supposed dington's type. These have potentials of 
increase in the strengths of the Balmer the form \p = [£(\) — ij(/jl)]/(\ — ju)> 
lines, as the surface gravity of the stars where A and n are spheroidal coordinates 
decreases at constant effective tempera- and $* and 77 are arbitrary functions, 
ture, in terms of the temperature fluctu- Satisfactory models of this form have 
ations produced by the strong turbulence been discussed by Kusmin. An attempt 
observed in the line contours and curves wa s made by Lynden-Bell to discover 
of growth for such stars. If the absolute whether the light distribution of NGC 
magnitude effect is confirmed, an attempt 4594 could arise from such a system, 
will be made to construct models non- Plates taken by Munch were reduced to 
homogeneous in temperature and to relate relative light intensities which were com- 
the temperature fluctuations to the prop- pared with the predicted functional form 
erties of the turbulent fields. of the projected mass density. Results 

. showed that NGC 4594 cannot be of 

Star Formation Eddingtonian form unless the mass-to- 

Theoretical work on star formation on light ratio varies very considerably and 

a phenomenological basis was continued systematically as a function of distance 

by Schmidt. The main assumption made from the center. 

in these considerations is the absence of Two further stellar dynamical investi- 
systematic radial transport of stars or gas gations were completed by Lynden-Bell. 
in the Galaxy. It appears that the average The first answers the question how long 
past rate of formation of stars around 1 it takes a nonsteady stellar system to 
solar mass is less than three times their approach an unrelaxed steady state and 
present formation rate. The past forma- how it does so in the absence of dissipa- 
tion history of bright, rapidly evolving tion. The time is about 10 times the 
stars can be deduced indirectly from the period of a typical star around the system, 
distribution of ultraviolet excesses in late and the mechanism is Landau damping, 
G-type dwarfs. It can be shown that a well known in plasma physics. The second 
time-independent birth luminosity func- gives a method of solving the self-gravity 
tion implies a certain distribution of equation for flattened steady-state sys- 
excesses, independent of the past forma- terns, and as a result the first exact 
tion rates. Specifically, in this case 72 per theoretical model is produced, 
cent of the late G-type dwarfs would be 

metal-deficient by a factor exceeding 2 Cosmology 

relative to the Hyades. The actual dis- Previous reports have mentioned that 

tribution of excesses shows between 40 a decision between the several proposed 

and 50 per cent of the late G-type dwarfs models of the expanding universe, such 


as exploding world models compared with steady-state model, a galaxy with Z — 0.4 
the steady-state model, can be made if at the present epoch will experience an 
the deceleration of the expansion can be acceleration rather than deceleration, in- 
measured. This is because the deceleration creasing its redshift at the instantaneous 
is caused by the gravitational attraction rate of +9.2 km /sec per million years, 
of matter on individual galaxies. The With present optical techniques there is 
rate of deceleration measures the density no hope of detecting such small changes 
of matter in space, which in turn deter- of redshift for time intervals of less than 
mines the spatial curvature and the 10 million years. If radio techniques are 
intrinsic geometry of space via the field used with observation of the 21-cm H I 
equations of general relativity. Various line, the detection of a frequency shift of 
ways of finding the deceleration param- 3 X 10~ 2 cps per year in a signal of fre- 
eter q , described in Year Book 59, depend quency 2000 Mc/sec is required, which 
on measurement of deviations from line- again appears to be impossible by present 
arity of the redshift-magnitude relation methods. To solve the problem in this 
or a similar relation such as that between way will require, at our present level of 
redshift and apparent diameter. The technology, a precision redshift catalogue 
principle of these measurements is that to be stored away in a stable society for 
one looks back in time as one looks out 10 million years, 
in space and can therefore sample the 

expansion rate of the universe in past Miscellaneous 
cosmic times. This indirect method is 

observationally very difficult, because The problem of the transfer of the 

galaxies with redshifts of the order of radiation in the [O I] line 3 P 2 — Z P\ at 

AX/X = 0.5 must be observed for signifi- 158.13 cm -1 , between the two lowest 

cant answers. Furthermore, as was point- sublevels of the ground state, through 

ed out last year, uncertain corrections the earth's ionosphere, has been studied 

for the evolution of the stellar content of by G. Munch by removing a number of 

galaxies must be made to account for the restrictive assumptions introduced in 

change in absolute luminosity of these previous attempts at a solution. The 

distant galaxies in the light travel time, solution found has been numerically 

which is of the order of 5 X 10 9 years. applied to a model atmosphere by com- 

Sandage looked into the theoretical puting the specific intensity and flux of 

possibility of detecting the deceleration radiation as a function of height. In a 

directly, if a series of redshift measure- paper submitted for publication to the 

ments of a given galaxy were made over Astrophysical Journal, the possibility of 

a suitable time interval. Exact predictions measuring the flux emergent at the top 

of the change of redshift in a given galaxy of the atmosphere, which amounts to 0.1 

with time can be made using the various erg/cm 2 sec, has been suggested as a 

world models. It turns out that this test method for the determination of the 

of world models is a powerful one in kinetic temperature and the oxygen 

principle because the sign of the effect is concentration. 

different for exploding models and the The eventual possibility of carrying 

steady-state model. However, the test is out observations of galactic and extra- 

beyond our present technical capabilities, galactic objects in Lyman-a radiation 

because the effect is extremely small. For was investigated by Munch. It has been 

a galaxy with a present redshift of Z = found that the decay of Lyman-o: through 

AX/Xo = 0.4, the change of redshift with 2-photon emission and dust absorption 

time is only —5.9 km/sec (decelerating) makes it quite unlikely that galactic 

in a million years for the Euclidean model, diffuse Lyman-o; radiation exists, unless 

Similar numbers hold for the hyperbolic the immediate neighborhood of the sun 

and elliptical exploding models. In the is an H II region, as is suggested by the 


emission nebulosities excited by the near- neutral gas and with low dust content 

by stars y Velorum and £ Puppis. Lyman- (as the coronas of M 31 and the Galaxy) 

a emission from the H II regions in may possibly reach the solar system when 

extragalactic systems not surrounded by their redshift exceeds 1000 km/sec. 


The 10-inch ruling engine has been in addition to Mount Wilson and Palomar. 
operation with the new system of inter- During the last year, a number of 
ferometric control. As was described in minor modifications have been made by 
last year's report, this system employs Oke to the photoelectric coude* spectrum 
intermittent spacing, with a fringe clamp scanner on the 100-inch telescope. The 
and differential corrector. Gears are now seeing compensation is now found to be 
on hand for ruling at 407, 610, and 915 quite satisfactory, provided that the see- 
grooves per millimeter. With water cool- ing is average or better and the zenith 
ing of the mercury-198 source for the distance is not more than approximately 
interferometer, the contrast of the fringes 45°. A program has been begun to meas- 
is more than ample for a path difference ure line profiles and equivalent widths in 
of 10 inches. The performance of the the spectra of selected stars. Accurate 
control system has been accurate and photoelectric profiles, such as those of 
reliable, so that the average quality of H7 in A stars, can be used to check the 
the gratings has been raised and the photographic calibration systems used at 
productivity of the machine has been various observatories, 
increased. Virtually theoretical resolving To facilitate the reduction of very 
power is obtained in the higher orders of small-aperture observations made with 
the gratings, and scattered light is held the solar magnetograph, digitizing equip- 
to very low levels. ment has been installed at the 150-foot 

Eleven plane gratings of high quality tower telescope under the supervision of 

were produced by Roberts under the Howard. A shaft encoder digitizes the 

direction of H. W. Babcock in sizes rang- shaft position of a strip-chart recorder, 

ing from 3 by 4 to 6 by 10 inches; most and this information is punched on paper 

were 5 by 8 inches with a spacing of 610 tape at the telescope. These data will be 

grooves per millimeter. fed to a digital computer at a later time, 

Gratings have been delivered to the and the autocorrelation analysis can pro- 

Kodaikanal, Sacramento Peak, Dominion ceed with no intermediate steps in the 

Astrophysical, and David Dunlap Ob- reduction. Thus it will be possible to 

servatories and to the National Bureau accumulate a great number of observa- 

of Standards. The records show that tions with small apertures and greatly 

gratings produced here are now in use at increase the accuracy of the existing 

16 observatories throughout the world in autocorrelation functions. 


The following programs have been photographic photometry, are now avail- 
carried out by guest investigators during able. Except for scale factors that depend 
the report year. upon the richnesses of the clusters, the 

Dr. George O. Abell of the Department luminosity functions are all similar. That 

of Astronomy, University of California, of the Coma cluster is representative. The 

Los Angeles, continued his study of rich number of galaxies brighter than m, 

clusters of galaxies. The luminosity func- N(m), is given, approximately, by 
tions for galaxies in six rich clusters, 

determined by a method of extrafocal log N(m) = constant + s log m 


where s = 0.78 for m < 14.7, and lies in aid of photoelectric observations of the 

the range 0.23 to 0.29 for 14.7 < m < stronger emission lines, an attempt was 

18.3, depending upon what correction is made to reduce the intensities of the lines 

applied for the nonmember galaxies in as measured on the spectrograms to a 

the cluster field. The discontinuity in s is true relative scale for a study of recom- 

due to a maximum in the bright end of bination rates and ionic abundances in 

the luminosity function. If the luminosity the nebula. Further spectrograms of 

functions of the different clusters are shorter exposure and probably additional 

fitted together at this discontinuity, and photoelectric measurements of weaker 

the relative distance moduli so obtained lines will be necessary to obtain an ac- 

are plotted against the known redshifts curate wavelength-dependent intensity 

of the clusters, the root-mean-square calibration. 

velocity dispersion about a straight line Dr. Stanley J. Czyzak of the Aeronau- 

is less than 600 km/sec. On the basis of tical Research Laboratories at the Wright 

the clusters investigated so far, therefore, Patterson Air Force Base continued his 

it appears that the luminosity functions calculations of accurate wave functions 

of clusters can provide good estimates of of various ions of P, S, CI, and A, all of 

their relative distances. which are of astrophysical interest. The 

In the course of this photometry of values for 25 ions of the 3p q configuration 

galaxies Abell made a detailed investiga- were completed. 

tion of some properties of the U, B, V It was now possible to begin a detailed 

system. He has numerically integrated 25 examination of the screening constants, 

stellar energy distributions published by spin-orbit, spin-spin, and spin-other-orbit 

Code and Melbourne against the response calculation for the transition probabilities 

functions of the U, B, V cell-filter com- for the Sp 9 configurations. Spectral data 

binations, through various air masses, of various gaseous nebulae were examined 

The computed variation of extinction for forbidden lines to determine which of 

with color (for a given air mass) is present the transitions had been observed. Also a 

but small for V, is in approximate agree- study was made to determine other 

ment with the variation usually assumed transitions which are significant. In ad- 

for B — V, and is appreciable and non- dition, preliminary calculations of colli- 

linear for U — B. Observed values of the sion cross sections were carried out by 

extinction at Mount Wilson are, on the Dr. Czyzak on ions with q = 1, since 

average, in agreement with those com- these calculations would be simpler than 

puted. In addition, the computed extinc- those for q > 1. 

tions for a given color of star vary non- Dr. H. Gollnow of the Mount Stromlo 

linearly with air mass; as a consequence, Observatory of the Australian National 

the usual procedure for reducing photo- University took spectra of about 20 stars 

electric observations can introduce errors with the coude spectrographs of the 100- 

in U — B colors of as much as 0.1 mag. inch and 200-inch telescopes in a search 

Finally, accurate color equations were for magnetic stars. Dispersions of 4.5 

derived to transfer from instrumental to A/mm and 10 A/mm and a differential 

U, B, V colors, and improved colors of analyzer in front of the slit were used, 

blackbodies have been obtained. The stars were selected between declina- 

Dr. Lawrence Aller of the University tions +25° and —40°, so that their 

of Michigan Observatory obtained a series observations can be continued at the 

of spectrograms of the planetary nebula Mount Stromlo Observatory. About 50 

NGC 7009 with exposures ranging from per cent of the stars show too large ro- 

a few hours to two nights. They revealed tational broadening for the measurement 

a large number of recombination lines of of Zeeman displacements. Of the other 

C, O, Ne, and other elements. With the stars, HD 24712 was studied in some 


detail and a magnetic field varying be- was an extension of an earlier investiga- 

tween +574 and +997 gauss was found, tion during the previous year of the 

The observation of this star will be rotational velocities of the B0-B3 Orion 

continued. stars. The B5-B9 stars were found to 

During the report year, studies of rotate somewhat more slowly on the 

velocity fields in the solar atmosphere average than the general field stars of the 

have been continued by Dr. R. B. Leigh- same type. The maximum rotational 

ton of the California Institute of Tech- velocities were found to occur in the 

nology with the assistance of Robert W. B5-B7 spectral types. The observations 

Noyes and George W. Simon. Special also indicate that there is a smaller per- 

emphasis was given to oscillatory motions centage of slow rotators among the B5-B9 

and large-scale currents discovered with group than among the B0-B3 group, 

the Mount Wilson instruments in 1960. Narrow-band photometric observations 

The main results may be summarized of stars in the star clusters h and % Persei 

as follows: (1) The small-scale velocity and M36 have been obtained by Dr. 

field (1000-5000 km linear dimension) in McNamara with the 60-inch telescope, 

the upper photosphere exhibits a strong A narrow-band photometric study of 

tendency to repeat itself in time with a eclipsing variables has also been initiated 

5-minute period. That is, each local region with the 20-inch telescope at Palomar 

undergoes a quasi-sinusoidal motion Observatory. No results are yet available 

which may persist for several cycles. (2) on these photometric programs. 

The period of the above oscillation does Dr. Walter E. Mitchell, Jr., of the 

not seem to be strictly constant with Perkins Observatory continued the solar 

altitude but tends to decrease by 10 to 20 observations with the Snow telescope 

per cent as one proceeds from the middle during the summer of 1961 with the 

photosphere into the lower chromosphere, assistance of Mr. John C. Muster. Nu- 

(3) The intensity variations in the lower merous improvements were made to the 
chromosphere, as seen at the cores of such telescope and spectrograph. To reduce the 
strong lines as Na X5896, Mg X5173, or scattered light due to Rowland ghosts, an 
similar lines, are also observed to fluctu- arrangement of mirrors and intermediate 
ate in time with the period of the velocity slit was designed to deliver the beam twice 
oscillations. This suggests that the oscilla- to the grating, i.e., to have the spectro- 
tory motions are connected with waves graph act as its own monochromator. A 
which transport energy into the chromo- beam splitter consisting of a plane parallel 
sphere and liberate it as heat or radiation, plate of fused quartz was mounted just 

(4) A network of horizontal currents, inside the entrance slit of the spectro- 
grouped into a system of large-scale graph. The fraction of the beam returned 
(5000-30,000 km) "convective cells," is by this plate was used as a monitoring 
observed. These cells appear to be dis- signal for ratio recording. 

tributed rather uniformly over the solar The double-pass system was employed 
surface, and their correlation properties to make photoelectric tracings of the 
suggest a strong tendency toward an following regions: Na Di and D 2 , Mg 'b\ 
ordered array over distances up to 50,000 Ca II H and K, H/3, H7, H5, Ca I 4226, 
or 100,000 km. and X3570. Throughout, there is a notice- 
Dr. D. H. McNamara of North Ameri- able lowering of central intensities (by 
can Aviation investigated the rotational amounts up to 10 per cent of the con- 
velocities of B5-B9 stars in the Orion tinuum) both as compared with the 
association. The spectra from which the Utrecht Photometric Atlas of the Solar 
rotational velocities were determined were Spectrum and with the Snow single-pass 
obtained with the 16-inch camera of the observations with the same grating. The 
100-inch coude spectrograph. This study region 6700-3900 A was recorded in first 



order with full resolution and with band 
passes of 4, 8, and 16 A. 

Infrared stellar photometric observa- 
tions were obtained by Dr. Mitchell and 
Mr. Philip E. Barnhart with the assist- 
ance of Messrs. John C. Muster, Ronald 
E. Roll, and John H. Hill. Instrumental 
assistance was also provided by Messrs. 
Charles E. Gramm, Anthony J. Prasil, 
and Dr. William H. Haynie of the East- 
man Kodak Company. Preliminary in- 
frared magnitudes were measured for 31 
G, K, and M giants, 6 red supergiants, 
and e Aurigae using an improved East- 
man Kodak-Ohio State University in- 
frared stellar photometer on the 60-inch 
and 100-inch telescopes. The photometric 
system has the following characteristics: 

Magnitude Effective Band 

Designation Wavelength, n Pass, ju 

X 2.2 0.24 

Y 3.7 0.43 

When the observed visual-infrared 
color indices of the measured stars are 
compared with the indices deduced the- 
oretically for the stars considered as 
blackbody radiators, the following con- 
clusions may be drawn: (1) Nonvariable 
giants and supergiants lie, in general, 
close to the theoretical relation; i.e., to a 
first approximation these stars behave as 
blackbody radiators. (2) A few nonvari- 
able giants and supergiants fall unexpect- 
edly far above or below the theoretical 
relation, suggesting large blanketing ef- 
fects on visual magnitude or errors in 
temperature assignments. (3) When ob- 
served at visual magnitudes well below 
maximum, long-period variable stars, 
whose temperatures are derived from 
their spectrum characteristics at mean 
maximum light, show an excessive red- 
dening compared with the theoretical 
relationship; that is, the variability 
occurs almost entirely at wavelengths 
shorter than 2 p. (4) Epsilon Aurigae 
shows an infrared excess of approximately 
1.2 mag, thus supporting the hypothesis 
that it has a large infrared component. 

Attempts were made by Dr. Mitchell 

and Mr. Barnhart to operate a helium- 
cooled Ge : Cd detector for the 8- to 13-/* 
region, but no stellar signal was distin- 
guishable from photometer and telescope 

Dr. Bruce C. Murray of the Lunar 
Research Laboratory at the California 
Institute has continued the program of 
photoelectric colorimetry of the moon 
using the spectrum scanner at the Casse- 
grain focus of the 60-inch. The scanning 
technique initiated during 1960-1961 has 
been perfected, including the successful 
implementation of "lunar rate" for the 
60-inch, to a point where an accuracy of 
0.01 to 0.02 mag has been achieved for the 
eleven independent color values obtained 
from each object examined. Approximate- 
ly fifteen lunar areas of 15 by 15 km size 
have been observed as well as various 
planetary and stellar objects for com- 
parison. The data are in the final phase 
of reduction preparatory to being sub- 
mitted for publication. 

The testing and development of a long- 
wavelength infrared photometer have 
continued during the year, two nights 
during 1961 having been devoted to this 
project at the 60-inch. Recently, however, 
a special 20-inch infrared telescope has 
been designed and built. This telescope 
with a novel optical and photometer 
system has been given preliminary trials 
at Mount Wilson but will later be placed 
in operation at a 13,000-foot site on White 

Dr. Robert L. Wildey and Mr. Howard 
A. Pohn of the Lunar Laboratory have 
initiated a U, B, V photometric program 
to investigate an apparent asymmetrical 
phase lag in the brightness versus phase 
curves of different localities on the moon. 
This phenomenon is apparent in the older 
photographic photometry of the moon; if 
confirmed photoelectrically, it represents 
a most surprising natural phenomenon of 
the moon. 

Observations were continued by Dr. 
Daniel M. Popper of the University of 
California at Los Angeles on the program 
of establishing absolute dimensions of 


stars of various kinds from the analysis of already taken by Dr. G. Waller stein for a 
eclipsing binary systems. Relatively few joint investigation of the velocity curve 
spectrograms were obtained during the of the M component of the system. 
year. Reanalysis has been completed for Dr. H. Spinrad of the Jet Propulsion 
three solar-type eclipsing binaries : VZ Laboratory of the California Institute has 
Hydrae, WZ Ophiuchi, and UV Leonis. analyzed infrared spectrograms of Venus 
The new spectrographic observations with in the plate files. Rotational temperatures 
higher dispersion lead to masses about 30 have been derived from the intensity 
per cent smaller than those obtained pre- distributions of CO 2 rotational lines in 
viously for two of the systems. The the X7820 band. The rotational tempera- 
revised values are more in accord with tures vary from 214°K to 445°K. Total 
the values from visual binaries of the pressures have been obtained from meas- 
same spectral types. The photometric urements of the corrected widths of the 
observations used in the analysis of WZ C0 2 rotational lines; these pressures 
Oph are also new, having been obtained correlate quite well with the rotational 
with the 20-inch at Palomar. A modern temperatures in the sense that the high 
light curve is badly needed for VZ Hyd. pressures correspond to observations of 

The following new results are based on high rotational temperatures, 

incomplete observations. (1) RR Arietis Dr. Spinrad has also found that the 

is a sixth-magnitude K star found to be ammonia and methane rotational lines in 

eclipsing by Archer. The velocity vari- the yellow-red region of the spectrum of 

ation appears to be less than 5 km /sec. Jupiter do not have the expected incli- 

(2) Revised values of the masses of the nation on coude spectra in which the 

K-type giants of RZ Cancri are 3.2 and spectrograph slit was placed along the 

0.5. (Dr. Popper's earlier published values planet's equator. This result is interpreted 

were 0.4 and 2.6.) (3) The D lines of the to mean that these gases are probably 

fainter components, not previously an- not rotating with the same velocity as the 

nounced, have been observed in the Jovian cloud layer. Examination of 100- 

f olio wing eclipsing systems : TW Draconis inch and 200-inch coude" spectra indicated 

(difficult), WW Draconis, RR Lyncis marked variations in the relative inten- 

(metallic-line star; observations com- si ties of the Jovian NH 3 lines near X6460. 

plete), XY Puppis (difficult), and TX Dr. Uli Steinlin continued his observa- 

Ursae Majoris (difficult). tions with the 48-inch schmidt camera to 

Dr. Jorge Sahade of the La Plata obtain material for the program on three- 
Observatory, Argentina, continued his color photometry of the Observatory in 
spectroscopic observations, obtaining Basel, Switzerland. Dr. W. Becker from 
plates of the following objects: (1) The Basel participated in the observations 
eclipsing star V453 Scorpii to supplement from February until April. The observing 
material previously obtained at Bosque program was completed in April with 572 
Alegre with lower dispersion; (2) HD plates taken (416 of them after July 1961) 
188439, an early-type object which Lynds in the following fields: eight Milky Way 
had announced as showing a photometric fields: NGC 1807/17, M37, Great Sagit- 
period of about 9 hours; (3) HD 207739 tarius cloud, Small Sagittarius cloud, 
and AG Pegasi to detect spectral changes, Scutum, Aquila, Lacerta, Cassiopeia; ten 
if any, relative to observations of previous fields in higher galactic latitudes : Selected 
years; (4) the eclipsing system V367 Areas 51, 54, 57, 82, 94, 107, 133, 141, 
Cygni to compare the spectral features 158, and Hyades. 

with those of other systems already Plates have mostly been taken in R, 

investigated; (5) HD 192281 to supple- G, and U for three-color photometry of 

ment observations made in 1960; and some clusters and, above all, of field stars 

(6) 17 Leporis to be used with material in the Milky Way as well as in higher 


galactic latitudes. In some fields, plates present both spectra are being analyzed 
have also been taken in B and V to permit in Kiel. For comparison with HD 161817, 
three-color photometry in the U, B, V the more or less normal stars 5 Delphini 
system as well, and to make possible a A7V, a Ophiuchi A5III, and 111 Herculis 
comparison of the effectiveness of the two A3V had been selected. Delta Del has 
systems. The limiting magnitude lies in quite sharp lines and has since been 
general between 18 m and 19 m . The three- measured for wavelengths and identifi- 
color photometry should provide: (1) cations. The other two stars show strong 
density function and luminosity function rotational broadening. Although a Oph is 
in different directions from the sun; (2) an MK-type star for A5III, the three- 
color-magnitude diagrams of clusters and dimensional Paris classification would 
of clouds of stars within the Milky Way; place it under A5V. This is probably due 
(3) possibly a separation of disk and halo to rotational broadening of the high 
populations in higher galactic latitudes, members of the Balmer series simulating 

About 1500 stars in each of the follow- the Stark broadening in main-sequence 

ing fields have already been measured: stars. This problem is being further 

Selected Area 54, 57, 82, and 107. The analyzed, with J. Kaler (Michigan), 

reduction of these measurements and working at present in Kiel, 
work in other fields is under way at the Spectra for investigating possible differ- 

Basel Observatory. ences in composition connected with 

Photoelectric U, B, V standards for the evolution were taken by Dr. Unsold with 

Basel Observatory program have been the 32-inch camera in the photographic 

obtained with the 100-inch by Dr. A. Th. and visual regions. Five stars with 

Purgathofer of the Vienna University spectral types F5 to G2V of Eggen's 

Observatory. Observations of stars in the y Leonis group, including 5/3 Virginis 

magnitude range from V = 16 m to 18 m (F8), supposed to be metal-superrich, and 

were obtained for most of the Selected two later-type (dK5) stars taken from 

Area fields. O. C. Wilson's "red" and "violet" groups 

Dr. A. Unsold of the University of Kiel, of the main sequence, HD 156026 and 

Germany, in 14 nights of observing, HD 192310, were observed, 
obtained high-dispersion spectra of vari- Some visual test plates of a Cygni A2 

ous groups of stars which might be la showed that the structure of its Ha 

suitable for studying the relations be- emission component has changed con- 

tween chemical composition and evolu- siderably since the last visual plates were 

tion. Most of the plates were taken with taken in 1957. The photometric analysis 

the 32-inch camera of the coude spectro- is being carried through by Dr. Comper 

graph of the 100-inch telescope, and they in Kiel. 

cover the photographic and the visual A considerable number of 100-inch 

regions. coude plates of y Serpentis F6IV-V taken 

HD 161817, usually classified as sdA2, in 1957 by Unsold, in 1959 by Traving, 

is most probably a horizontal branch star, and in 1960 by Bonsack have been 

Its huge space motion, according to analyzed in detail by W. Kegel in Kiel. 

Eggen, is shared by Wilson 10367 = The variation of turbulence with depth 

LPM661, an 11-mag F8 main-sequence turned out to be an essential feature. The 

subdwarf , of which at least the photo- relative abundances of the metals are the 

graphic region could be obtained with the same as in the sun, but, relative to 

16-inch camera. Both stars are obviously hydrogen, all the heavier elements are 

metal-poor. Quantitative comparison of reduced by a factor of about 1.7. That, as 

their chemical composition should give well as the space velocity and the weak 

most interesting indications about evolu- ultraviolet radiation, indicates that y Ser 

tionary events in the red giant or super- is a member of the intermediate Popu- 

giant region of the H-R diagram. At lation II. 


Dr. George Wallerstein of the Astro- Dr. R. v. d. R. Woolley and Mr. C. A. 

nomical Department of the University of Murray of the Royal Greenwich Observa- 

California at Berkeley has been observing tory carried out two programs with the 

K giants in order to obtain abundances of coude spectrograph attached to the 100- 

the elements. The observations of G8-K2 inch reflector. They exposed a number of 

stars in the general field at 6 A/mm in the plates with the 32-inch camera and with 

yellow region are now complete. Stars the 72-inch camera. Some of these were 

with ultraviolet excesses from 0.20 mag exposed to an intensity suitable for the 

to deficiencies of 0.10 mag will be com- measurement of radial velocity; others 

pared with the K0 giants in the Hyades. were more lightly exposed so that they 

Many of the stars included are high- would be suitable for spectrophotometry, 

velocity stars. Some of the other more The radial-velocity plates have been 

interesting stars on the list are a few that measured, and the results will be pub- 

Gyldenkern suspects to be metal-rich lished; the remaining plates have been 

from his narrow-band photometry, and examined with a spectrophotometer. The 

some "4150 stars" that show strong CN spectra of r Ceti, 107 Piscium, and 

in the blue region. In addition, two K0 o 2 Eridani obtained with these plates have 

giants in Praesepe have been observed at been investigated by Dr. Pagel at the 

15 A/mm in the yellow. This work is in Royal Greenwich Observatory, and the 

cooperation with Dr. Heifer of the results have been worked up for a 

University of Rochester. determination of the abundances of ele- 

The high-velocity A star HD 109995 ments in these stars by differential 

has been observed in order to compare it curve-of-growth analysis. 

with Sirius and another high-velocity A Dr. Woolley and Mr. Murray also 

star, 7 Sextantis. A cursory examination carried out a program of direct photog- 

of one 4 A/mm and two 10 A/mm plates raphy at the Cassegrain focus of the 

shows that the lines in HD 109995 are 60-inch telescope. They took repeat plates 

very much weaker than in either of the of a number of cluster fields that had been 

other two stars. observed by van Maanen, including the 

Dr. Wallerstein obtained several plates cluster M 67. In all, 15 fields were 

of 31 Cygni that showed chromospheric photographed; the plates, together with 

Ti II lines as well as the K line. These a selection of van Maanen's first epoch 

plates will be reduced in cooperation with plates, are at the Royal Greenwich 

Dr. Wright of the Dominion Astro- Observatory awaiting measurement for 

physical Observatory. proper motion. 


The Observatories suffered a severe the 200-inch mirror after it had been 

loss in the sudden death on December 26, moved to Palomar. While on leave from 

1961, of Don O. Hendrix, Superintendent the Observatories he ground and figured 

of the Optical Shop. Mr. Hendrix joined the 120-inch mirror of the Lick Observa- 

the staff in 1931 and became Super- tory. The high efficiency of the present 

intendent of the Optical Shop in 1947. He optical equipment of the Observatories is 

developed extraordinary skill in the hand to a large extent due to Hendrix' skill and 

figuring of large nonspherical surfaces ingenuity. 

required in many modern optical designs. Drs. Robert Howard and Olin Eggen 

Among the projects he carried out were joined the staff of the Observatories in 

the optics for the 48-inch schmidt tele- September 1961. Dr. Howard plans to 

scope at Palomar and the corrector plates investigate solar magnetic fields, and Dr. 

for the 15 schmidt cameras on the Eggen has undertaken an extensive photo- 

spectrographs at Palomar and Mount metric program. Dr. Otto Struve became 

Wilson. He also did the final figuring of a member of the staff in March 1962. 



Research Division 

Staff Members 
Halton C. Arp 

Horace W. Babcock, Assistant Director 
William A. Baum 
Ira S. Bowen, Director 
Armin J. Deutsch 
Olin J. Eggen 
Jesse L. Greenstein 
Robert F. Howard 
Robert P. Kraft 
Guido Munch 
J. Beverley Oke 
Allan R. Sandage 
Maarten Schmidt 
Otto Struve 
Olin C. Wilson 
Fritz Zwicky 

Research Associates 
Jan H. Oort 
Kenneth 0. Wright 

Staff Members Engaged in Post-Retirement 
Harold D. Babcock 
Milton L. Humason 
Alfred H. Joy 
Seth B. Nicholson 

Senior Research Fellows 

Rudolph Kippenhahn 
Satoshi Matsushima 
Minoru Nishida 
Evry Schatzman 
Leonard T. Searle 
Volker Weidemann 

Carnegie Research Fellows 
Leo Houziaux 
Charles R. O'Dell 

National Science Foundation Fellows 

John C. Brandt 
Paul W. Hodge 

Research Fellows 

Jacques Berger 
John Hazlehurst 
D. H. P. Jones 
Jun Jugaku 
Donald Lynden-Bell 
Luis Munch 
Konrad Rudnicki 
Wallace L. W. Sargent 
Henrietta Swope 

Research Assistants 
Christine Arpigny 
Jeanne Berger 
Frank J. Brueckel 
Sylvia Burd 
Subhash Chandra 
Jai H. Choy 

Mary F. Coffeen, Librarian 
Thomas A. Cragg 
Donald S. Hayes 1 
Emil Herzog 
Joseph 0. Hickox 2 
Basil N. Katem 
Charles T. Kowal 
A. Louise Lowen 
Joyce E. Sheeley 
Merwyn G. Utter 

Student Observers 

James E. Gunn 
Manuel E. Mendez 
Dimitri M. Mihalas 
Robert H. Norton 
Robert A. R. Parker 
Lewis L. Smith 
Robert L. Wildey 


William C. Miller 

Instrument Design and Construction 

Lawrence E. Blakee, Senior Electronic 

Eileen I. Challacombe, Draftsman 
Floyd E. Day, Optician 
Kenneth E. DeHufT, Machinist 
Robert D. Georgen, Machinist 
Don 0. Hendrix, Superintendent of Optical 

Shop 3 
Melvin W. Johnson, Optician 
Stuart L. Roberts, Instrument Maker 
Bruce Rule, Project Engineer 
Marlin N. Schuetz, Electronic Technician 
Russell R. Van Devender, Jr., Designer and 

Superintendent of Instrument Shop 
James S. White, Electronic Technician 4 

Maintenance and Operation 

Mount Wilson Observatory and Offices 
Paul F. Barnhart, Truck Driver 

1 Resigned March 23, 1962. 

2 Retired September 30, 1961. 

3 Died December 26, 1961. 

4 Resigned December 15, 1961. 



Wilma J. Berkebile, Secretary 

Herbert A. Cole, Laborer 5 

Hugh T. Couch, Carpenter 

Helen S. Czaplicki, Editorial Typist 

Stewart F. Frederick, Janitor 6 

Eugene L. Hancock, Night Assistant 

Mark D. Henderson, Gardener 

Margaret Higgins, Stewardess 

Anne McConnell, Administrative Assistant 

Leah M. Mutschler, Stenographer and 

Telephone Operator 
Bula H. Nation, Stewardess 
Alfred H. Olmstead, Night Assistant 
Arnold T. Ratzlaff, Night Assistant 
Glen Sanger, Janitor 
John E. Shirey, Laborer 
William D. St. John, Janitor and Relief 

Wilma G. Totten, Stewardess 7 
Benjamin B. Traxler, Superintendent 

6 Resigned March 9, 1962. 

6 Resigned October 31, 1961. 

7 Resigned November 15, 1961. 

Palomar Observatory and Robinson Laboratory 

Audrey A. Acrea, Stewardess 
Fred Anderson, Machinist 
Jan A. Bruinsma, Custodian 
Maria J. Bruinsma, Stewardess 
Eleanor G. Ellison, Secretary and 

Arlis R. Grant, Stewardess 8 
Leslie S. Grant, Relief Night Assistant and 

Byron S. Hill, Superintendent 
Helen D. Hollo way, Secretary 
Charles E. Kearns, Night Assistant 
J. Luz Lara, Laborer 
Harley C. Marshall, Office Manager 
Dwight M. Miller, Mechanic 
George W. Pettit, Janitor 9 
Robert E. Sears, Night Assistant 
William C. Van Hook, Electrician and 

Assistant Superintendent 
Gus Weber, Assistant Mechanic 

8 Resigned October 13, 1961. 

9 Retired September 3, 1961. 


Abt, Helmut A., Hamilton M. Jeffers, James 
Gibson, and Allan R. Sandage, The visual 
multiple system containing Beta Lyrae, Astro- 
phys. J., 185, 429-438, 1962. 

Arp, Halton C, The globular cluster M5, As- 
trophys. J., 135, 311-332, 1962. 

Arp, Halton C, The effect of reddening on the 
derived ages of globular clusters and the abso- 
lute magnitudes of RR Lyrae cepheids, Astro- 
phys. J., 135, 971-975, 1962. 

Arp, Halton C, Intrinsic variables and stellar 
evolution, Proc. Symposium on Stellar Evolu- 
tion, La Plata, Argentina, Nov. 7-11, 1960, 
edited by J. Sahade, Observatorio Astro- 
nomico, Secci6n Publicaciones, La Plata, 
Argentina, 1962, pp. 87-117. 

Arp, Halton C, Stellar content of galaxies, 

Science, 134, 810-819, 1961. 
Arpigny, Claude, see Greenstein, Jesse L. 

Baade, Walter, and Henrietta H. Swope, The 
Draco system, a dwarf galaxy (abstract), 
Astron. J., 66, 278, 1961; Astron. J., 66, 300- 
347, 1961. 

Babcock, Horace W., The sun's magnetic field 
(abstract), Science, 134, 1425, 1961. 

Banner, K., W. A. Hiltner, and Robert P. Kraft, 
Colors and magnitudes for 45 cepheids of the 
Northern Milky Way (abstract), Astrophys. 
J., 134, 1026-1027, 1961; Astrophys. J., suppl. 
59, 319-355, 1962. 

Baum, William A., Image converters — A new 
instrumental technique in astronomy, I.C.S.U. 
Rev., 8, 199-201, 1961. 

Baum, William A., The kind of universe implied 
by recent photoelectric redshift observations 
at Palomar, Observatory, 81, 114-115, 1961. 

Baum, William A., Photoelectric test of world 

models (abstract), Science, 184, 1426, 1961. 
Baum, William A., 9a. Sous-commission des 

convertisseurs d'images, Trans. Intern. Astron. 

Union, 10, 143-154, edited by D. H. Sadler, 

University Press, Cambridge, 1960. 
Baum, William A., 9a. Sous-commission des 

convertisseurs d'images, Trans. Intern. Astron. 

Union, 11 A, 34-47, edited by D. H. Sadler, 

Academic Press, London, 1962. 

Bowen, Ira S., John August Anderson, Bio- 
graphical Memoirs, 36, 1-18, published for 
National Academy of Sciences of the United 
States by Columbia University Press, New 
York, 1962. 

Brandt, John C, Helium resonance radiation in 
the night and day sky (abstract), Astron. J., 
66, 279, 1961. 

Brandt, John C, The problem of the Gegen- 
schein, Astron. Soc. Pacific Leaflet 391, 8 pp., 
January 1961. 

Brandt, John C, Interplanetary gas, V, A hy- 
drogen cloud of terrestrial origin, Astrophys. 
J., 134, 394-400, 1961. 



Brandt, John C, Interplanetary gas, VI, On 
diffuse extreme ultraviolet helium radiation in 
the night and day sky, Astrophys. J., 184, 975- 
980, 1961. 

Brandt, John C, On the problem of the distance 
to the center of the Galaxy (abstract), Publ. 
Astron. Soc. Pacific, 78, 324, 1961. 

Brandt, John C, A note on the scale of the 
Galaxy, Publ. Astron. Soc. Pacific, 74, 142-145, 

Brandt, John C., On the interpretation of the 
night sky Lyman a radiation and related 
phenomena, Space Research II, Proc. Second 
Intern. Space Sci. Symposium, Florence, pp. 
624-638, edited by H. C. van de Hulst, D. de 
Jager, and A. F. Moore, North-Holland Pub- 
lishing Co., Amsterdam, 1961. 

Burbidge, E. M., see Wildey, R. L. 

Burbidge, G. R., see Wildey, R. L. 

Cameron, A. G. W., The formation of the sun 
and planets, Icarus, 1, 13-69, 1962. 

Cayrel de Strobel, Guisa, A comparison of photo- 
electric and photographic spectrophotometry, 
Ann. Astrophys., 24, 509-515, 1961. 

Cragg, Thomas, Rotation of Saturn, Publ. As- 
tron. Soc. Pacific, 78, 318-322, 1961. 

Cragg, Thomas, Three new variable stars, Publ. 
Astron. Soc. Pacific, 78, 453-455, 1961. 

Deutsch, Armin J., Non-catastrophic mass-loss 
from stars, Intern. Astron. Union Symposium 
12: Aerodynamic phenomena in stellar atmos- 
pheres, part III, pp. 238-259, edited by R. N. 
Thomas, 1961. 

Deutsch, Armin J., The prospects of extrater- 
restrial astronomy, Navigation, 8, 12-17, 1961. 

Deutsch, Armin J., P. W. Merrill, and P. C. 
Keenan, Behavior of absorption features in 
the spectra of Mira variables (abstract), 
Astron. J., 66, 282, 1961. 

Deutsch, Armin J., see also Maestre, Leonard A. 

Eggen, Olin J., Stellar groups, Proc. Symposium 
on Stellar Evolution, La Plata, Argentina, Nov. 
7-11, 1960, edited by Jorge Sahade, Observa- 
torio Astronomico, Secci6n Publicaciones, La 
Plata, Argentina, 1962, pp. 233-250. 

Eggen, Olin J., SZ Lyncis: A new ultrashort- 
period RR Lyrae variable, Publ. Astron. Soc. 
Pacific, 74, 159-161, 1962. 

Fowler, William A., Jesse L. Greenstein, and 
Fred Hoyle, Deuteronomy: the synthesis of 
deuterons and the light nuclei during the early 
history of the solar universe, Am. J . Phys., 29, 
393-403, 1961. 

Fowler, William A., Jesse L. Greenstein, and 
Fred Hoyle, Nucleosynthesis during the early 
history of the solar system, Geophys. J ., Roy. 
Astron. Soc, 6, 148-220, 1962. 

Fowler, William A., Jesse L. Greenstein, and 
Fred Hoyle, Nuclear history of the solar sys- 
tem (abstract), Publ. Astron. Soc. Pacific, 78, 
326-327, 1961. 

Gates, H. S., see Zwicky, Fritz. 

Gibson, James, see Abt, Helmut A. 

Gomes, Alercio M., see Humason, Milton L.; 
Zwicky, Fritz. 

Greenstein, Jesse L., Stellar evolution and the 
origin of the chemical elements, Am. Scientist, 
49, 449-473, 1961. 

Greenstein, Jesse L., Radio sources containing 
peculiar ellipticals, Astrophys. J., 135, 679-683, 

Greenstein, Jesse L., and Claude Arpigny, The 
visual region of the spectrum of Comet Mrkos 
(1957d) at high resolution, Astrophys. J., 135, 
892-905, 1962. 

Greenstein, Jesse L., and Fritz Zwicky, Super- 
nova 1959 SN 64, Publ. Astron. Soc. Pacific, 
74, 35-40, 1962. 

Greenstein, Jesse L., see also Fowler, William A.; 
Jugaku, Jun. 

Griffin, Roger F., The positions of optical objects 
in the fields of 42 radio sources, Monthly 
Notices Roy. Astron. Soc, in press. 

Herzog, Emil, and Konrad Rudnicki, Bright 
galaxies with unknown redshifts, Publ. Astron. 
Soc. Pacific, 74, 234-237, 1962. 

Hiltner, W. A., see Bahner, K.; Kraft, Robert P. 

Hodge, Paul W., The Fornax dwarf galaxy, II, 
The distribution of stars, Astron. J., 66, 249- 
257, 1961. 

Hodge, Paul W., The distribution of stars in 
dwarf elliptical galaxies (abstract), Astron. J., 
66, 286, 1961. 

Hodge, Paul W., The distribution of stars in the 
Sculptor dwarf galaxy, Astron. J., 66, 384-390, 

Hodge, Paul W., Studies of the Large Magellanic 
Cloud, VII, The open cluster NGC 1844, 
Astrophys. J., 134, 226-231, 1961. 

Hodge, Paul W., The gravitational stability of 
the NGC 7619 group of galaxies, Astrophys. 
J., 184, 262-264, 1961. 

Hodge, Paul W., Studies of the Large Magellanic 
Cloud, VI, Properties of 1057 field stars (ab- 
stract), Astrophys. J., 184, 666-667, 1961. 

Houziaux, Leo, Stark and Doppler broadening 
for high Balmer and Paschen lines, I, The 
absorption coefficient, Ann. Astrophys., 24, 
541-549, 1961; (abstract) Publ. Astron. Soc. 
Pacific, 73, 328-329, 1961. 

Houziaux, Leo, Mesure et distribution spectrale 
de l'energie recue des etoiles, Scientia, 96, 369- 
375, 1961 



Houziaux, Leo, Forbidden nitrogen I lines in the 
infrared solar spectrum, Z. Astrophys., 53, 237- 
239, 1961. 

Houziaux, Leo, On the infrared spectrum of 
Pleione, Publ. Astron. Soc. Pacific, 74, 250- 
253, 1962. 

Howard, Robert, Solar magnetic fields, Astron. 
Soc. Pacific Leaflet 896, 8 pp., June 1962. 

Hoyle, Fred, see Fowler, William A. 

Humason, Milton L., Photographs of planets 
with the 200-inch telescope, The Solar System, 
vol. Ill, Planets and Satellites, chapter 16, p. 
572 and 15 plates, edited by G. P. Kuiper and 
B. M. Middlehurst, University of Chicago 
Press, 1961. 

Humason, Milton L., C. E. Kearns, and Alercio 
M. Gomes, The 1961 Palomar supernova 
search, Publ. Astron. Soc. Pacific, 74, 215-218, 

Humason, Milton L., see also Zwicky, Fritz. 

Jeffers, Hamilton M., see Abt, Helmut A. 

Jones, D. H. P., The double star ADS 3210, 
01185, Observatory, 82, 29-31, 1962. 

Joy, Alfred H., Ralph Elmer Wilson, Biographi- 
cal Memoirs, vol. 36, pp. 314-329, published 
for National Academy of Sciences of the 
United States by Columbia University Press, 

Joy, Alfred H., Spectroscopic absolute magni- 
tudes: A review, J. Roy. Astron. Soc. Canada, 
56, 67-78, 1962. 

Joy, Alfred H., Paul Willard Merrill, 1887-1961, 
Publ. Astron. Soc. Pacific, 74, 41-43, 1962. 

Joy, Alfred H., Paul Willard Merrill, Quart. J. 
Roy. Astron. Soc, 3, 45-47, 1962. 

Jugaku, Jun, and Wallace L. W. Sargent, The 
spectrum of a Sculptoris, Publ. Astron. Soc. 
Pacific, 73, 249-255, 1961. 

Jugaku, Jun, Wallace L. W. Sargent, and Jesse 
L. Greenstein, An abundance analysis of 3 
Centauri A, Astrophys. J., 134, 783-796, 1961. 

Jugaku, Jun, see also Sargent, Wallace L. W. 

Kearns, C. E., see Humason, Milton L. 

Keenan, P. C, see Deutsch, Armin J. 

Kraft, Robert P., Color excesses for supergiants 
and classical cepheids, V, The period-color and 
period-luminosity relations: a revision, Astro- 
phys. J., 134, 616-632, 1961. 

Kraft, Robert P., Binary stars among cataclys- 
mic variables, I, U Geminorum stars (dwarf 
novae), Astrophys. J., 135, 408-423, 1962. 

Kraft, Robert P., Nova (WZ) Sagittae as a 
binary star (abstract), Science, 134, 1433, 1961. 

Kraft, Robert P., Exploding stars, Sci. American, 
206, 54-63, 1961. 

Kraft, Robert P., and W. A. Hiltner, Color ex- 
cesses for supergiants and classical cepheids, 
VI, On the intrinsic colors and Hess diagram 
of late-type supergiants, Astrophys. J., 134, 
850-860, 1961. 

Kraft, Robert P., see also Bahner, K. 

Maestre, Leonard A., and Armin J. Deutsch, 
List of absorption lines in two ultra-sharp line 
A stars, Astrophys. J., 134, 562-567, 1961. 

Merrill, Paul W., Unidentified lines in spectra of 
sun and stars, Astrophys. J., 134, 556-561, 

Merrill, Paul W., see also Deutsch, Armin J. 

Mihalas, Dimitri M., Light curve of Humason's 
supernova in Virgo (abstract), Astron. J., 67, 
118-119, 1962; Publ. Astron. Soc. Pacific, 74, 
116-124, 1962. 

Minkowski, R., NGC 6166 and the cluster Abell 
2199, Astron. J., 66, 558-561, 1961. 

Minkowski, R., Radio sources, galaxies, and 
clusters of galaxies, Proc. Natl. Acad. Sci. 
U. S., in press. 

Munch, G., and A. Unsold, Interstellar gas near 
the sun, Astrophys. J., 135, 711-714, 1962. 

Nicholson, Seth B., Award of the Bruce Gold 
Medal to Grote Reber, Publ. Astron. Soc. 
Pacific, 74, 183-186, 1962. 

Nicholson, Seth B., and Oliver R. Wulf, The 
diurnal variation of K indices of geomagnetic 
activity on disturbed days in 1940-1948, J. 
Geophys. Res., 66, 2399-2404, 1961; (abstract) 
Science, 134, 1434, 1961. 

Oke, J. B., An analysis of the absolute energy 
distribution in the spectrum of 8 Cephei, As- 
trophys. J., 134, 214-221, 1961. 

Oke, J. B., see also Searle, Leonard. 

Pettit, Edison, Planetary temperature measure- 
ments, The Solar System, vol. Ill, Planets and 
Satellites, chapter 10, pp. 400-428, edited by 
G. P. Kuiper and B. M. Middlehurst, Univer- 
sity of Chicago Press, 1961. 

Preston, George W., A coarse analysis of three 
RR Lyrae stars, Astrophys. J., 134, 633-650, 

Rodgers, A. W., Photoelectric spectrophotom- 
etry of O-type stars, Monthly Notices Roy. 
Astron. Soc, 122, 413-419, 1961. 

Rudnicki, Konrad, see Herzog, Emil. 

Sandage, Allan, The light travel time and the 
evolutionary correction to magnitudes of dis- 
tant galaxies, Astrophys. J., 184, 916-926, 

Sandage, Allan, Photometric data for the old 
galactic cluster NGC 188, Astrophys. J., 135, 
333-348, 1962. 

Sandage, Allan, The ages of M67, NGC 188, M3, 
M5, and M13, according to Hoyle's 1959 
models, Astrophys. J., 185, 349-365, 1962. 



Sandage, Allan, Introductory report, Proc. Sym- 
posium on Stellar Evolution, La Plata, Argen- 
tina, Nov. 7-11, 1960, edited by Jorge Sahade, 
Observatorio Astron6mico, Secci6n Publica- 
ciones, La Plata, Argentina, 1962, pp. 1-22. 

Sandage, Allan, The ages of the oldest stars in 
the Galaxy compared with the cosmic expan- 
sion time, Proc. Symposium on Stellar Evolu- 
tion, La Plata, Argentina, Nov. 7-11, 1960, 
edited by Jorge Sahade, Observatorio Astro- 
n6mico, Secci6n Publicaciones, La Plata, 
Argentina, 1962, pp. 119-135. 

Sandage, Allan, Travel time for light from distant 
galaxies related to the Riemannian curvature 
of the universe (abstract), Science, 184, 1434, 

Sandage, Allan, see also Abt, Helmut A.; Smith, 
Lewis; Wildey, R. L. 

Sargent, Wallace L. W., Circumstellar envelope 
of rho Cassiopeiae, Astrophys. J., 134, 142-160, 

Sargent, Wallace L. W., and Jun Jugaku, The 
existence of He 3 in 3 Centauri A, Astrophys. J., 
134, 777-782, 1961. 

Sargent, Wallace L. W., see also Jugaku, Jun. 

Schmidt, Maarten, Galactic structure at optical 
and radio wavelengths, J. Roy. Astron. Soc. 
Can., 55, 149-166, 1961. 

Schmidt, Maarten, A preliminary discussion of 
helium abundances, Proc. Symposium on Stel- 
lar Evolution, La Plata, Argentina, Nov. 7-11, 
1960, edited by Jorge Sahade, Observatorio 
Astron6mico, Secci6n Publicaciones, La Plata, 
Argentina, 1962, pp. 61-66. 

Schmidt, Maarten, The evolution of the sun's 
neighborhood, Proc. Symposium on Stellar 
Evolution, La Plata, Argentina, Nov. 7-11, 
1960, edited by Jorge Sahade, Observatorio 
Astronomico, Secci6n Publicaciones, La Plata, 
Argentina, 1962, pp. 67-75. 

Searle, Leonard, and J. B. Oke, Effective temper- 
atures of F-type stars using Hy line profiles, 
Astrophys. J., 185, 790-801, 1962. 

Smith, Lewis, and Allan Sandage, Color-magni- 
tude diagram of the strong-line globular cluster 
NGC 6712 (abstract), Astron. J., 67, 121, 1962. 

Swope, Henrietta H., see Baade, Walter. 

Tifft, William G., Multicolor photometry of 
bright galaxies, I, Astron. J., 66, 390-404, 

Traving, Gerhard, Atmospheres of two B-type 
stars in the galactic halo, Astrophys. J., 135, 
439-458, 1962. 

Unsold, G., see Munch, G. 

Wildey, R. L., E. M. Burbidge, A. R. Sandage, 
and G. R. Burbidge, On the effect of Fraun- 
hofer lines on U, B, V measurements, Astro- 
phys. J., 135, 94-108, 1962. 

Wulf, Oliver R., see Nicholson, Seth B. 

Zwicky, Fritz, L'astronomie et la marche dans 
l'univers, I Astronomie, 76, 55-65, 1962. 

Zwicky, Fritz, Die Kerne von Galaxien als neue 
charakteristische Aggregate kosmischer Ma- 
terie, Physik. Blatter, 17, 393-399, 1961. 

Zwicky, Fritz, Cosmic and terrestrial tests for 
the rest mass of gravitons, Publ. Astron. Soc. 
Pacific, 73, 314-317, 1961. 

Zwicky, Fritz, Supernovae and chains of super- 
novae as launchers of stars and galaxies, Publ. 
Astron. Soc. Pacific, 74, 70-73, 1962. 

Zwicky, Fritz, Possible operations on the moon, 
Space Flight, 3, 177-179, 1961. 

Zwicky, Fritz, M. L. Humason, A. M. Gomes, 
and H. S. Gates, Progress report on super- 
novae, Publ. Astron. Soc. Pacific, 73, 351-354, 

Zwicky, Fritz, see also Greenstein, Jesse L. 

Geophysical Laboratory 

Washington, District of Columbia 
Philip H. Abelson 




Experimental Petrology 56 

Pyroxenes 56 

The join diopside-Ca Tschermak's 

molecule at atmospheric pressure . 56 
Phase relations in the system 
CaMgSi 2 q6-CaAl 2 Si06-Si0 2 at 
low and high pressure .... 59 
X-ray data for diopsidic pyroxenes . 61 
Melting relations in the system 

diopside-anoi thite-silica ... 65 
The system MgSi() 3 -CaMgSi 2 0^ . . 68 
The system diopside-enstatite-silica . 75 
Preliminary results on melting rela- 
tions of synthetic pyroxenes on the 
diopside-hedenbergite join ... 82 

Metamorphic petrology 82 

Metamorphic reactions involving two 

volatile components 82 

Synthesis and stability of anthophyllite . 85 
Quartz-chlorite assemblages in the 
system MgO-Al 2 3 -Si0 2 -H 2 . . 88 
Alkali-rich igneous rocks and minerals . 91 
The system _ Na 2 0-Al 2 3 -Fe 2 3 -Si0 2 

and its bearing on the alkaline rocks . 91 
Peralkaline residual liquids: some 

petrogenetic considerations ... 95 
The system nepheline-diopside ... 96 
A reconnaissance of the systems 
acmite-diopside and acmite-neph- 

eline 98 

Accessory minerals 100 

Investigations in the system FeO- 

Fe 2 3 -Ti0 2 . . . . . . .100 

Magnetite-ilmenite relations . . . 100 
Application of experimental data to 

natural minerals - 105 

Relations between ilmenite, hema- 
tite, magnetite, and rutile . . . 106 
Stability relations of dravite, a 

tourmaline 106 

Mantle minerals 107 

Effect of pressure on the melting of 
pyrope 109 

Statistical Petrography 112 

Sanidine phenocrysts in some peralkaline 

volcanic rocks 112 

Bulk analyses and norms . . . .114 
Feldspar phenocrysts of the analyzed 

rocks 115 

Projection of results into "petrogeny's 
residua system" 116 

Variance relations in some published 
Harker diagrams 118 

The treatment of FeO and Fe 2 3 in 
Harker diagrams 119 

On the relative scarcity of intermediate 
members in the oceanic basalt-tra- 
chyte association 121 

Granite in Port Clyde peninsula . . . 123 

Feldspar in the granite of the Port Clyde 
peninsula 126 

Two-mica granite and hornblende-biotite 
granite 128 

Crystallography 130 

Relationships between crystal structure 

and crystal morphology 130 

Lattice constant refinement . . . .132 
The crystal structure of sillimanite . .135 
The crystal structure of Fe mica . . .139 
On the transitions of bornite . . . .139 

Ore Minerals 142 

The Mo-S system 143 

The Fe-Ni-S system 144 

Liquid immiscibility 144 

Pentlandite stability relations . . .146 
Bravoite stability relations . . . .149 

The Fe-Mo-S system 150 

The Cu-Ni-S system 151 

The Fe-Ni-As system 152 

The Cu-Fe-S system ...... 154 

Pyrrhotite-pyrite-chalcopyrite relations 154 
Exsolution textures and rates in solid 

solutions involving bornite . .155 

Exsolution textures 155 

Rates of exsolution 156 

Chalcocite-chalcopyrite assemblages . 157 
Heating experiments on natural born- 

ites . . . 159 

Method for mixing liquids at controlled 

temperatures 160 

Pyrrhotite from Tern Piute, Nevada . .161 

Stony Meteorites 163 

Iron Meteorites 165 

The system Fe-Ni-S 165 

The system Fe-Ni-P 166 

Higher phosphides in the system Fe-P . 166 

Geothermal Calculations 168 

The Ages of Rocks and Minerals . . .173 
Geographic distribution of mineral ages 
in the central portion of North 

America 173 

Ages of minerals from the Coutchiching 

sediments, Rainy Lake, Ontario . .176 
Age relation between the Karelian and 
Svecofennian orogenies in Finland . .178 

Organic Geochemistry 179 

Paleobiochemistry 179 

Thermal stability of algae . . . .179 
Fatty acids in sedimentary rocks . .181 
The isolation of organic compounds 
from Precambrian rocks .... 184 
The biogeochemistry of the stable iso- 
topes of carbon 187 

The isotopic composition of the carbon 

of fatty acids 187 

The stable isotopes of carbon in the 
carbonate and reduced carbon of 
Precambrian sediments . . . .190 

Miscellaneous Administration . . . .192 
Institute on Isotopes and Radioactivity . 192 
Journal of Geophysical Research . . .192 

Lectures 192 

Penologists' Club 193 

Summary of Published Work . . . .194 

Bibliography 201 

References Cited 202 

Personnel 208 

Carnegie Institution of Washington Year Book 61, 1961-1962 


The Geophysical Laboratory continues fields of sillimanite, corundum, and 

its diversified program of studying the probably also "/? alumina" on the 

earth through application of physical liquidus near its composition. Pressure 

science and mathematics. Superficially, greatly increases the maximum amount 

the areas of effort during the report year of alumina that the pyroxene can accom- 

appear to be very similar to those of the modate in its structure ; lime Tschermak's 

preceding period. There was activity in molecule (CaAl 2 Si0 6 ) is stable up to 

experimental petrology, statistical petrol- 1500°C at 20 kb, and there is complete 

ogy, crystallography, ore minerals, mete- solid solution between this phase and 

orites, geothermal calculations, the ages diopside. This work clearly shows that 

of rocks and minerals, and organic important changes in melting relations 

geochemistry. There were, however, sub- are produced by pressure and that phase 

stantial shifts of emphasis within the diagrams determined at atmospheric pres- 

program. For instance, much of the work sure cannot be applied to the production 

this year in experimental petrology was of magma at great depths in the earth, 

focused on the pyroxenes, and greater Boyd and Schairer present final results 

emphasis was placed on phase equilibria on the system MgSi0 3 -CaMgSi 2 06. Most 

at higher pressures. A substantial effort mafic rocks contain two pyroxenes, and 

was expended on studies of the miner- this binary system is fundamental to an 

alogy of meteorites. New investigations understanding of the mineralogy and 

of organic geochemistry were initiated, genesis of these rocks. It was found that 

including analysis of Precambrian carbo- the solvus intersects the solidus over a 

naceous materials. composition interval of 35 weight per 

The Laboratory is continuing its broad cent, so that solid solution between the 

program of studying the phase relations Ca-rich and Mg-rich pyroxenes is much 

in basalts and their derivative rocks. The more restricted than was previously 

alkali-type basalts in particular have been thought. Evidence for a hitherto unrecog- 

examined from several viewpoints this nized form of Mg-rich pyroxene was 

year. The fundamental joins nepheline- found in runs at temperatures above 

diopside, acmite-diopside, and nepheline- 1385°C. 

acmite, bearing on alkali-type rocks, were Yoder and Tilley continued with 

worked out by Schairer, Yagi, and Yoder. Schairer their heating experiments at 1 

Studies in the system Na 2 0-Fe203-Al 2 03- atmosphere on natural basalts and nat- 

Si0 2 have been carried out by Schairer ural coexisting pyroxenes, and completed 

and Bailey in view of its importance to the preparation of their extensive mono- 

the peralkaline derivative rocks. graph, "The origin of the basalt magmas: 

Work by Clark, Schairer, and de An experimental study of natural and 
Neufville on the composition plane synthetic rock systems." 
CaMgSi 2 6 -CaAl 2 Si0 6 -Si0 2 at low and One of the crucial systems in petrology, 
high pressures represents the first exten- diopside-forsterite-silica, has been re- 
sive study of the effect of pressure, investigated in part by Schairer and 
transmitted by an inert medium, on Yoder. The revisions are of exceptional 
melting and subsolidus relations in a import to the general differentiation of 
portion of a complex quaternary system, magmas and the nature of pyroxenes that 
Some of the phase relations at 20 kb are crystallize in them. Many refinements 
totally different from those at atmos- occur within a very small temperature 
pheric pressure. Anorthite melts incon- interval, and by close temperature con- 
gruent ly at this pressure, and there are trol and calibration seven isothermal 



sections in 5° intervals were mapped out. ilmenite- magnetite intergrowths may 

Turnock has surveyed melting relations form by oxidation. 

of synthetic pyroxenes on the join The most important natural compound 

diopside-hedenbergite, using controlled of boron is tourmaline. The stability 

atmospheres with fixed partial pressures range of one of the end members of 

of oxygen. tourmaline, dravite, has been outlined 

In his studies of metamorphic reactions under hydrous conditions, special care 
Greenwood has made two contributions, being taken to retain the sodium and 
one theoretical and the other experi- boron. The work was carried out mainly 
mental. He has derived equations for by Robbins, of the National Bureau of 
equilibrium boundaries between reacting Standards, working during his free time 
phase assemblages in systems that contain at the Laboratory, with the help of Yoder. 
two volatile components and are subject Boyd and England have found that 
to variation of pressure and temperature, pyrope garnet melts incongruently to 
These equations have the same form as spinel and liquid in the same pressure 
the usual expressions for crystal-liquid range as that in which basalts are be- 
equilibria but do not carry the restriction lieved to form in suboceanic areas. Spinel 
that the relative proportions of the two contains no silica, and so the liquid that 
volatile components are limited by the forms by this reaction is oversaturated. 
proportions of the other components. The Hence, the incongruent melting of pyrope 
effect of removing this restriction is to may explain the development of over- 
make stable many reactions that would saturated basalts from mantle rocks 
normally be regarded as metastable. He believed to be undersaturated. The 
has also been able to delineate the upper classic reaction that has been used io 
and lower stability limits of anthophyllite explain the development of oversaturated 
and observe the nucleation kinetics of the lavas is the incongruent melting of 
mineral. enstatite, discovered by Bowen in runs 

Many of the principal mineral assem- at atmospheric pressure. However, Boyd 

blages found in low-grade metamorphic and England showed last year that 

rocks include chlorite and quartz. Previ- enstati ( te melts congruently at pressures 

ous synthesis studies had indicated that greater than those present at moderate 

these are not compatible minerals. New depths in the earth's crust, 
studies by Fawcett and Yoder, involving In statistical petrography, studies of 

experiments of long duration, have sue- the phenocryst-groundmass relations in 

cessfully resolved this apparent conflict peralkaline lavas (Zies and Chayes) and 

with nature. of the relations between two-mica and 

Lindsley has shown that synthetic biotite-hornblende granites in the Port 

titaniferous magnetites in equilibrium Clyde peninsula (Suzuki and Chayes) 

with ilmenite contain little or no ilmenite have been continued, with results of 

in solid solution at temperatures below considerable interest. Chayes has now 

1000°C; compositions of these magnetites fully substantiated his suggestion of last 

lie close to the magnetite-ulvospinel join, year that in the typical Harker array the 

The titanium content of the magnetites variance of silica is approximately an 

is strongly dependent on oxygen f ugacity order greater than any other variance ; in 

as well as on temperature and hence 23 of 25 published arrays for which 

cannot be used as a simple geologic calculations have been completed the 

thermometer. Hydrothermal oxidation of variance of silica is larger than the sum 

magnetite-ulvospinel solid solutions yields of all other variances. This great excess 

ilmenite-magnetite intergrowths textur- of silica variance is the most important 

ally similar to natural occurrences, sup- single influence on the correlations that 

porting the hypothesis that many natural characterize a Harker array. 



J. D. H. Donnay (Johns Hopkins Uni- 
versity) and G. Donnay have applied 
their second generalization of the law of 
Bravais to the elucidation of the external 
forms of ionic crystal structures. Their 
study of the mineral barite, the morphol- 
ogy of which had heretofore remained 
unexplained, leads to the new concept of 
"centers of charges." These turn out to 
be the equipoints of the bond assem- 
blages. This punctualization of charges 
is apparently the key to the morphologies 
of ionic crystals, the interpretation of 
which is a generalization of that of the 
NaCl morphology. 

Burnham has developed a least-squares 
technique for refinement of lattice con- 
stants of crystals, which has been 
programmed for the IBM 7090 digital 
computer. The procedure, already em- 
ployed successfully by other staff mem- 
bers, has the following features. It is 
applicable to crystals of any symmetry 
and will accept data, from cards or tape, 
either as angle measurements for any 
wavelength or in the form of calculated 
d values. Observations may be weighted 
according to any scheme, and up to nine 
systematic correction terms may be 
included with each observation. 

Morimoto has studied the transition 
mechanisms by which the three poly- 
morphic forms of bornite are inter- 

Studies of many facets of the ore 
minerals continue. Increased emphasis is 
being placed on application of laboratory 
findings to ores. However, the program is 
still predominantly a laboratory investi- 
gation of phase equilibria. The Mo-S 
system was studied by Morimoto and 
Kullerud, who found that the "Mo 2 S 3 " 
phase is only stable above 605°C. Very 
important results have been obtained on 
the Fe-Ni-S system, in which at high 
temperatures Kullerud found liquid im- 
miscibility over a large region extending 
across the sulfur- rich part of the system. 
In other studies he also showed that 
pentlandite breaks down at 610°C and 
that bravoite is only stable below 137°C. 

In the Fe-Mo-S system Kullerud and 
Buseck found that the minerals pyrite 
and molybdenite are stable together 
below 726°C. The Cu-Ni-S system is 
being investigated by Moh and Kullerud, 
who have finished the 600°C isothermal 
section. Buseck studied the Fe-Ni-As 
system and has synthesized the new 
mineral oregonite. The solid solutions in 
the uni variant region containing pyrite, 
pyrrhotite, chalcopyrite, and vapor were 
examined by von Gehlen and Kullerud. 
They found that at 600°C application of 
the pyrite-pyrrhotite thermometer gives 
temperatures about 50°C lower when 
chalcopyrite is present than when it is 
absent. Brett studied exsolution textures 
from solid solutions involving bornite 
(digenite-bornite, chalcocite-bornite, chal- 
copyrite-bornite) . His results indicate 
that textural evidence alone does not 
permit drawing of reliable conclusions 
about the thermal history of the minerals 
that form these solid solution pairs. 

Two studies were directed at problems 
involving meteorites. Clark, studying the 
system Fe-Ni-P, has demonstrated that 
the Fe/Ni ratio of the schreibersite, 
(Fe,Ni) 3 P, in equilibrium with both 
kamacite and taenite changes measurably 
with temperature. The phosphide is a 
common constituent of iron meteorites, 
and its composition, along with the 
compositions of the alloy phases, will help 
to trace the history of these extraterres- 
trial bodies. A curious, but simple, 
relationship that has emerged from the 
synthetic system is that the ratio 
Ni/(Fe -f- Ni) is the same in the schrei- 
bersite as in the taenite (7 alloy) with 
which it is in equilibrium. Ramdohr has 
examined polished sections of more than 
a hundred stony meteorites. He has 
identified more than twenty minerals y 
half of which had not been seen previously 
in stony meteorites. Native gold was seen 
in specimen. In addition, he observed 
twelve new minerals that have not been 
fully identified but whose composition 
can be partially inferred from associations 
with known substances. Noteworthy 



structural and textural relations were also 
seen, including localized droplets of fused 
troilite and iron. 

In a continuation of the theoretical 
geothermal studies of the past few years, 
the effect on heat flow at the surface 
produced by very high thermal conduc- 
tivity at depth in the earth has been 
investigated by Clark. He has found that 
under certain circumstances high conduc- 
tivity at depth reduces the surface flux. 
Under other conditions the opposite 
effect is produced. The possibility of 
variable conductivity in the earth intro- 
duces an ambiguity in the interpretation 
of heat-flow measurement in addition to 
ambiguities caused by lack of precise 
knowledge of the distribution of radio- 
activity and the initial temperature. A 
second geothermal investigation concerns 
the cooling of the deep mantle. It has 
been found that appreciable cooling could 
take place if the initial thermal gradient 
were sufficiently steep, but it is not yet 
clear whether such a steep gradient is 
tolerable on other grounds. 

Zircon age studies have been made by 
Davis on the ancient igneous and sedi- 
mentary rocks at Rainy Lake, Ontario. 
The results indicate that all the zircons 
crystallized about 2750 million years ago. 
The rocks were eroded to form sediments, 
which were subsequently metamorphosed 
about 2600 million years ago. 

Additional age determinations by Til- 
ton and Kouvo in Finland show that the 
Karelian and Svecofennian orogenies 
occurred at about the same time, although 
geological evidence suggests that the 
Svecofennian orogeny is the older. 

A geochronological map of the United 
States and southern Canada, based on 
several hundred mineral age determina- 
tions, has been constructed. The Pre- 
cambrian rocks occur in belts or zones, 
with younger rocks on either side of an 
old central belt. 

Abelson and Parker have isolated 
saturated fatty acids including stearic, 
palmitic, and myristic from rocks as old 
as 500 million years. Very recent sedi- 
ments contain these same entities and 
virtually no unsaturated types. This 
relationship differs sharply from that 
noted in algae, which ostensibly are the 
major source of organic matter in 
sediments. Parker has isolated pure fatty 
acids from algae and found that their 
C 13 /C 12 ratios differed from the C 13 /C 12 
ratio of total cell. Different types of fatty 
acids from the same organism have the 
same C 13 /C 12 ratio. 

Hoering has studied geochemical evi- 
dence for the existence of life in Pre- 
cambrian rocks. The fractionation of the 
stable isotopes of carbon into a C 13 -rich 
carbonate fraction and a C 13 -depleted 
reduced fraction, which is characteristic 
of sedimentary rocks of known biological 
association, was found to exist in some of 
the oldest known sedimentary rocks, 
including the Bulawayan limestone, which 
has a minimum age of 2.7 billion years. 
Hoering also has isolated a number of 
organic compounds from these Precam- 
brian rocks. The chemicals are similar to 
those that have been found in coal. Both 
these results are consistent with the 
existence of life and photosynthesis during 
early Precambrian times. 



The Join Diopside-Ca Tschermak's 

Molecule at Atmospheric Pressure 

John de Neufville and J. F. Schairer 

In order to study the extent of the 
substitution of A1 2 3 in diopside, Hytonen 

(Year Book 60) prepared three series of 
compositions on the plane enstatite- 
wollastonite- corundum. During the past 
year his work on the diopside-Ca Tscher- 
mak's molecule (CaAl 2 Si0 6 , henceforth 
abbreviated CTs) series has been extend- 
ed to the CTs composition. Figure 1 is a 
temperature-composition section at at- 



1600° - 

Ca Mg Si 2 6 

30 40 50 60 

Wt. %CaAI 2 Si0 6 

90 CaTschermak's 
Ca AI 2 Si0 6 

Fig. 1. Temperature /composition plot of data obtained on the join diopside-Ca Tschermak's 
molecule (CaAl 2 Si0 6 ) at 1 atmosphere. Abbreviations for phases encountered: Di S8 , diopside solid 
solution; Mel 8S , melilite solid solution; Ak 8 9Gehn, etc., specific melilite composition in terms of weight 
per cent akermanite and gehlenite; Sp, spinel; Fe, forsterite; An, anorthite; Al, "(3 alumina" and/or 
corundum; L, liquid. 

mospheric pressure along this join. It has 
been constructed using Hytonen's unpub- 
lished data on eleven compositions and 
new data on four more aluminous 
compositions. Hytonen's X-ray deter- 
minative work on diopsidic pyroxenes has 
been extended to more aluminous 
pyroxenes and melilite solid solutions in 
polyphase assemblages. The Di-CTs join 
intersects at a high angle the "grossu- 
larite"-"pyrope" join studied under sim- 
ilar conditions by Chinner and Schairer 
(Year Book 59). Thus it continues the 
assault on uni variant lines, invariant 
points, and solidus volumes in the silica- 
poor part of the CaO-MgO-Al 2 3 -Si0 2 
quaternary system. Quenching experi- 
ments on the Di-CTs compositions at 
temperatures between the liquidus and 
the solidus have given new data on the 
temperatures of three quaternary invari- 
ant points and on the positions of critical 
planes separating the seven solidus vol- 
umes encountered. 

The maximum degree of stable pyrox- 
ene solid solution is estimated indirectly 
to be greater than 40 mole per cent A1A1 

for (Ca,Mg)Si. Sakata (1957) observed a 
continuous shift in pyroxene lattice 
parameters in day-long runs at 1200°C on 
compositions between diopside and 
Di 6 oCTs 4 o. We find that the same 
extreme pyroxene solid solution phases 
form and persist indefinitely at somewhat 
higher temperatures in the presence of a 
small amount of liquid. The pyroxene 
crystallizing in Di 6 oCTs 4 o at 1 atmosphere 
must lie off the Di-CTs join, probably on 
the diopside-MgAl 2 Si0 6 (Mg Tscher- 
mak's molecule)-CaAl 2 Si0 6 plane, because 
it coexists with two relatively magnesium- 
poor phases, anorthite and akermanite. 
Lattice parameters determined by Sakata 
(1957) closely fit the lattice parameter 
curves (fig. 4, p. 63) at Di 6 oCTs 40 for 
pyroxenes crystallized directly on the 
Di-CTs join at 20 kb. Present evidence 
indicates, accordingly, that the Di 6 oCTs 40 
pyroxene crystallized at atmospheric 
pressure has approximately the same 
degree of A1A1 for MgSi substitution as if 
it lay on the Di-CTs join at that bulk 

In molar coordinates the Di-CTs join 



is parallel to the akermanite-gehlenite 
series of melilites within the tetrahedron 
defined by the four oxides, both series 
involving the exchange of A1A1 (gehlenite 
and CaAl 2 Si06) for MgSi (akermanite and 
diopside) . Melilites are encountered in six 
of the seven solidus volumes cut by the 
Di-CTs join, and they range in compo- 
sition from pure gehlenite to pure or 
nearly pure akermanite. Ervin and Os- 
born (1949) have determined the d values 
of several X-ray reflections as a function 
of composition in this series. Only the 
strongest peak, (211), may be used for 
X-ray determinative work in polyphase 
assemblages. Although the change in 
d(2ii) from akermanite (2.871 A) to 
gehlenite (2.846 A) is linear, it is very 
small. Thus, an estimated possible meas- 
uring error of ±0.02° in 20 corresponds to 
about ±7 weight per cent akermanite. 
Nine melilite compositions determined by 
this method are shown in figure 2. 

The melilite composition changes as a 
function of bulk composition in the three- 
phase assemblages Mel ss + An + Sp, 
Mel ss + An -f- Fo, and Mel ss + Di ss + 
An. It is invariant in the four-phase 
assemblages Geh + Sp + An + "alu- 

mina," Ak 69 Geh 3 i + An + Sp + Fo, and 
Ak 89 Geh u + An + Di ss + Fo. The 
melilite composition can be calculated as 
a function of bulk composition in the 
three-phase regions where melilite co- 
exists with phases of fixed compositions, 
that is in Mel ss + An + Fe and in 
Mel ss + An + Sp. It cannot be uniquely 
calculated in three-phase assemblages 
containing Mel 8S + An + Di ss . To calcu- 
late the melilite composition in four-phase 
solidus assemblages, the critical planes 
bounding the tetrahedron must be pre- 
cisely located; only for Geh + Sp + An 
+ "alumina" was it possible to do this. 
Where these compositions could be 
calculated as a function of bulk compo- 
sition they are shown on figure 2 as solid 
lines. In all other assemblages the melilite 
compositions have been estimated from 
the approximate position of critical 
planes, and are dotted. 

At least three invariant points have 
been encountered in the phase-equi- 
librium studies of the Di-CTs series 
compositions. All of them lie outside 
their respective four-phase volumes and 
are reaction points. The Geh + "alumi- 
na" + Sp + An + L reaction point lies 

-2.870 0< 




-2.850 = 

Diopside '0 

Co Mg Si 2 6 

20 30 40 50 60 

Wt. % Co AI 2 Si0 6 



90 Co Tschermok's 
Co AI 2 Si0 6 

Fig. 2. Plot of melilite compositions in solidus assemblages versus bulk composition along join 
diopside-Ca Tschermak's molecule (CaAl 2 Si0 6 ). Solid lines, calculated theoretical melilite compo- 
sitions; dashed lines, estimated theoretical melilite compositions; dots, compositions actually deter- 
mined by measurement of d( 2 in of the melilites; bars, estimated possible error in determinations 
(±0.02° 20 = ±7 weight per cent akermanite). Abbreviations as in figure 1. 



within the Mel 68 + An + Sp volume at a 
temperature of 1360° =fc 5°C. This is only 
in fair agreement with De Vries and 
Osborn (1957), who measured a value of 
1350°C for this point. " Alumina" refers 
to corundum and/or "(3 alumina." These 
phases often occur together, although 
"(3 alumina" is predominant near liquidus 
temperatures, and only corundum peaks 
are observed in X-ray patterns of solidus 

The Sp + An + Fo + Ak 69 Geh 3 i + L 
invariant point lies within the Mel ss + 
An + Fo volume, and its temperature is 
1225° =fc 5°C. The Di ss + An + Fo + 
AkggGehn -f- L invariant point also lies 
within the same volume. Its temperature 
is not known precisely. It is drawn in 
figure 1 at 1225°C, which probably repre- 
sents a maximum value. Since these two 
invariant points have closely similar 
temperatures and compositions, it is 
difficult to decipher their mutual relation- 
ship. If they have the same temperature 
and different compositions, it is likely 
that some compositions along the Di-CTs 
join will pass through neither point and 
will crystallize directly to Mel ss + An + 
Fo without forming any pyroxene or 
spinel. This possibility is depicted on 
figure 1. Another, equally likely, possi- 
bility is that the reaction point at which 
pyroxene is consumed occurs at a slightly 
lower temperature than the reaction 
point at which spinel is consumed. If this 
is so, some compositions along the join 
may pass through both points as they 
crystallize, before winding up as a mixture 
of melilite, anorthite, and forsterite. 
Since compositions along the "grossu- 
larite"-"pyrope" join (Chinner and 
Schairer, Year Book 59) appear to raise 
but not answer the same questions, the 
answers can be supplied only by the study 
of compositions lying off these joins in the 
Ak-Ak 70 Geh 3 o-An-Fo volume. 

Chinner and Schairer observed that 
several compositions on the "grossular- 
ite"-"pyrope" join crystallized an alumi- 
nous pyroxene that reacted with liquid at 
lower temperatures to produce a melilite- 

anorthite-forsterite assemblage. They 
suggested that extensive fractionation of 
diopside crystallized from a basaltic melt 
as the result of limestone syntexis would 
enrich the melt in A1 2 3 . Aluminous 
pyroxene could then store up CaO and 
A1 2 3 , which would contribute to the 
formation of melilite as pyroxene redis- 
solved in the magma during the final 
stages of crystallization. This mechanism 
is in complete qualitative agreement with 
the classic contamination sequence de- 
scribed by Tilley and Harwood (1931) at 
Scawt Hill. It also receives excellent 
qualitative confirmation by the relations 
observed on the Di-CTs join. The analo- 
gous reaction point in this system is 
Di S8 + AksgGehn + An + Fo + L, at 
which melts do indeed consume pyroxene 
and form melilite and other phases. This 
is presumably the same reaction point 
encountered by Chinner and Schairer. 
Their mechanism, however, has at least 
one serious limitation in the application 
to the melilite rocks of Scawt Hill. The 
pyroxene at this reaction point in the 
synthetic system has been shown to 
contain about 40 mole per cent A1A1 for 
(Ca,Mg)Si. This is more than twice as 
much AI2O3 as is found in Scawt Hill 
aluminous pyroxenes (Tilley and Har- 
wood, 1931). Thus the analogy between 
the synthetic and the natural pyroxene 
reaction point is less direct than Chinner 
and Schairer inferred. 

Phase Relations in the System 

CaMgSi 2 & -CaA hSiO G -Si0 2 

at Low and High Pressure 

Sydney P. Clark, Jr., J. F. Schairer, and 
John de Neufville 

There is substantial indication that 
basaltic magmas are generated in the 
mantle, perhaps at considerable depths. 
Some of this evidence is seismic, some 
geothermal, some geologic. With the 
exception of the seismic activity associ- 
ated with Hawaiian eruptions, it is 
indirect and perhaps capable of other 


interpretations. But enough evidence corundum, forsterite, the melilites aker- 
points in the same direction to make a manite and gehlenite, wollastonite, diop- 
study of the effect of pressure on melting side, enstatite and its polymorphs, the 
relations in systems of petrological inter- aluminosilicates andalusite, kyanite, sli- 
est worth while. Furthermore, an upper limanite, and mullite, the garnets pyrope 
limit to the temperature in the mantle is and grossularite, cordierite, anorthite, 
set by the liquidus of whatever material and the polymorphs of silica. During the 
is down there, for superheated liquid must report year a new phase, with the 
either move relatively rapidly toward the composition of lime Tschermak's mole- 
surface or lose its superheat by reaction cule (CaAl 2 Si0 6 ), was synthesized for the 
with surrounding solid material. Effects first time at a pressure of 20 kb. 
of pressure on liquidus relations must be Because of the importance of this 
studied in systems of moderate complex- quaternary system, much previous work 
ity before inferences about melting in the has been done to elucidate phase relations 
mantle can be drawn with any confidence, in it. Most were investigations of lines 
It has been found that a pressure of 20 kb and planes joining two or three of the 
produces large effects on the liquidus that phases listed above. In this way the 
could not have been predicted from data tetrahedron is crossed in many directions, 
obtained at atmospheric pressure alone, and, given enough such studies, it should 

The quaternary system CaO-MgO- be possible to deduce with high precision 

Al 2 3 -Si0 2 , which contains the join the quaternary equilibrium relations at 

CaMgSi 2 6 (diopside)-CaAl 2 Si0 6 (lime atmospheric pressure. 

Tschermak's molecule)-Si0 2 , is of great For initial study in this system at high 

importance, because it is sufficiently pressure we selected the join diopside- 

complicated to represent qualitatively anorthite. As Bowen recognized, this is a 

the phase relations of the basic igneous simple, pseudobinary representation of 

rocks, particularly basalts, and rocks many basalts and diabases. The system 

arising from their metamorphism. The was first shown not to be truly binary by 

main constituents commonly present in Osborn (1942). We expected that the 

such rocks and absent from this quater- nonbinary behavior would be accentuated 

nary system are iron in both its valence by pressure, and this has proved to be so. 

states, soda, water, and to a lesser extent Part of the join is quaternary because of 

K 2 0, Ti0 2 , and MnO. In this simplified the incongruent melting of anorthite at 

system it is impossible to study such high pressures. We have, however, only 

important relationships as the effects of studied compositions lying in the plane 

the fugacities of water and oxygen or diopside-lime Tschermak's molecule-sil- 

changes in composition of feldspars, ica, which contains the join diopside- 

Experimental difficulties occasioned by anorthite. 

the various possible oxidation states of For purposes of orientation it is helpful 

transition elements and the volatility of to consider the composition plane ensta- 

alkalies and water at high temperatures, tite-wollastonite-corundum (fig. 3). All 

however, are avoided. Despite the sim- phases shown in the figure lie precisely in 

plifi cations, a number of reactions of this plane; none are projected. Of 

petrological importance take place in this particular interest are the intersecting 

system; because of this and its relative joins diopside-lime Tschermak's molecule 

chemical tractability the system is and grossularite-pyrope. The pyroxene 

well suited for a beginning to the study join is characterized by complete solid 

of complex chemical equilibria at high solution at 20 kb; the garnet join, by 

pressures. complete solid solution above 30 kb. 

Among the important phases lying in Magnesian Tschermak's molecule, shown 

this system are the oxides spinel and on the diagram, has never been synthe- 











Fig. 3. Solid phases in the plane CaSi0 3 - 
MgSi0 3 -Al 2 3 . 

sized, although Boyd and England (Year 
Book 59, p. 49) have made enstatites with 
at least 15 mole per cent A1 2 3 in solid 

The plane in figure 3 contains a number 
of phases with a striking variety of crystal 
structures, all characterized by a metal- 
to-oxygen ratio of 2 :3. Their densities are 
closely correlated with the structure. 
Densities of diopsidic pyroxenes men- 
tioned in the ensuing paragraphs have 
been calculated from X-ray data dis- 
cussed below. Cell edges of the grossular- 
ite-pyrope series of garnets were given 
by Chinner, Boyd, and England (Year 
Book 59, p. 77), and the densities of other 
phases were taken from the literature, 
using X-ray data whenever possible. 

Wollastonite and the pyroxenes are 
chain-type silicates. The lightest, wollas- 
tonite, has a density slightly greater than 
2.9 g/cm 3 . The density of enstatite is 
3.212 g/cm 3 , that of diopside is 3.281 
g/cm 3 , and that of lime Tschermak's 
molecule is 3.437 g/cm 3 . The garnet 
structure is composed of isolated silica 
tetrahedra, connected by irregularly co- 
ordinated cations. The density of pyrope 
is 3.566 g/cm 3 ; that of grossularite is 
3.603 g/cm 3 . The densest structure is that 
of the closely packed oxide corundum, 
4.02 g/cm 3 . 

It has been suggested that pyroxenes 
might undergo transitions to the corun- 

dum structure at very high pressures, and 
this inversion has been reported in 
MgGe0 3 (Ringwood and Seabrook, 1962). 
It does not seem to have been remarked 
that garnets also have the metal-to- 
oxygen ratio appropriate to undergo a 
transition to a corundum structure. Such 
an inversion may take place deep in the 
transition zone in the mantle. 

Two other comparisons of density are 
interesting to make. The first is between 
the density of crystalline lime Tscher- 
mak's molecule (CaAl 2 Si0 6 ) and the 
densities of its low-pressure breakdown 
products gehlenite (p = 3.038 g/cm 3 ), 
anorthite (p = 2.765 g/cm 3 ), and either 
"|8 alumina" or corundum. The density of 
"/3 alumina" is not well known, but 
neither alumina phase is present in large 
amounts. The mean density of the break- 
down products cannot be far from 2.9 
g/cm 3 . Lime Tschermak's molecule is 18 
per cent denser than this. The density 
change between the pyroxene and garnet 
at the intersection of the two joins shown 
in figure 3 is 6 per cent (3.368 versus 3.592 
g/cm 3 ). 

X-ray data for diopsidic pyroxenes. In 
order to set up suitable determinative 
procedures for complex solid solutions 
such as those shown by diopsidic py- 
roxenes, careful crystallographic work 
must be done. The fine-grained nature of 
synthetic crystals precludes single-crystal 
studies, and care must be taken that 
determinative peaks on powder patterns 
can be unambiguously indexed. Other- 
wise errors from effects of preferred 
orientation may influence measurements 
of unresolved multiple reflections. 

In crystals of low symmetry it is all but 
impossible without the aid of a high-speed 
computer to be sure that all indexing 
allowed by the space group has been 
compared with the observed reflections. 
Only by being certain that all possibilities 
have been considered can one be sure that 
a reflection is not multiple. Such pre- 
cautions have not always been taken in 
the past. 

All data processing was carried out on 



an IBM 7090 digital computer using 
programs written by Charles W. Burn- 
ham. His program for calculating unit- 
cell parameters by least squares is 
described elsewhere in this report. His 
program for calculating d values permit- 
ted by the space group from the param- 
eters of the unit cell was used in indexing 
powder patterns. 

TABLE 1. Miller Indices and d Values of 

Reflections Used in Calculating Unit-Cell 

Parameters of Diopsidic Pyroxenes 

d Value 



Lime Tschermak's 
























1 . 9468 




The starting point in our investigation 
was a carefully indexed powder pattern 
of pure diopside. This pattern was com- 
pared with patterns obtained on material 
prepared by completely crystallizing 
glasses on the join diopside-lime Tscher- 
mak's molecule at 20 kb. The positions of 
the peaks were found to shift smoothly as 
a function of composition from one end 
of the join to the other. No peaks 
appeared that could not be traced into 
their counterparts in the diopside pattern ; 

this plus optical examination provides 
evidence that only one phase, a pyroxene, 
was present in these runs. As a check, the 
complete pattern for the composition 50 
per cent diopside, 50 per cent lime 
Tschermak's molecule, was calculated. 
No unexpected interferences between 
peaks were found. The reflections used 
and their d values for diopside and lime 
Tschermak's molecule are given in table 
1. These reflections were chosen because 
they can be indexed unambiguously and 
are sharp and strong — an important 
feature if they are to be used for deter- 
minative purposes in mixtures of phases 
that do not contain very much pyroxene. 

The first three reflections listed in 
table 1 fall at 20 angles less than 31° for 
copper radiation. Hence the d values 
cannot be determined with high accuracy. 
The parameters of the unit cells were 
calculated by least-squares adjustment 
both with and without these peaks. The 
resulting parameters do not differ sig- 
nificantly, but the standard errors are 
usually smaller if the low-angle peaks are 

The unit-cell parameters of lime 
Tschermak's molecule and diopside are 
given in table 2, along with parameters 
for diopside from other observers. The 
agreement is good. The change of 
parameters along the joins diopside-lime 
Tschermak's molecule and diopside- 
enstatite is shown in figures 4 and 5. The 
data in figure 5 were obtained by applying 
the procedures described above to a series 
of glasses that had previously been 
crystallized at 1 atmosphere. Compo- 
sitions containing more than 40 per cent 

TABLE 2. Unit-Cell Parameters of Lime Tschermak's Molecule and Diopside 

Lime Tschermak's 


(Sakata, 1957) 

(H. H. Hess, 

a, A 

9.615 ±0.003 

9.745 ±0.001 



b, A 

8.661 ±0.002 

8.925 ±0.001 



c, A 

5.272 ±0.003 

5.248 ±0.001 



0, deg 

73.88 ±0.03 

74.13 ±0.01 



V, A 3 

421.79 ±0.28 

439.08 ±0.07 











74.25 - 




435 [ 





25 50 75 

Wt % CaAI 2 Si0 6 


Fig. 4. Unit-cell parameters along the join diopside-lime Tschermak's molecule. 



o 9.70 


+ -+-- 


o <t 




°< 5.25 





en 73.75 


| 73.50 

^ 73.25 






o< 435 







40 60 

Wt % Enstatite 
Fig. 5. Unit-cell parameters along the join diopside-enstatite. 



enstatite do not crystallize to a single larger than those found by other workers 

phase under these conditions. to cast serious doubt on the determinative 

Along the join lime Tschermak's mole- curves given by Hytonen and Schairer. 
cule-diopside, the substitution is Al-Al for Melting relations in the system diopside- 
Mg-Si. One would expect that replacing anorthite-silica. Liquidus data for this 
an Mg atom with a relatively small Al system at atmospheric pressure are shown 
would cause a and b to decrease. Likewise, in figure 6. Dots indicate the compositions 
replacing an Si atom with a relatively studied by the quenching method. Except 
large Al atom in the silica chains would for compositions near the diopside-silica 
cause a slight increase in c. These are the join, the figure has approximately the 
observed effects. /3 changes little in this appearance of the simplest type of 
series. The internal consistency of the ternary diagram, that is one in which 
data for this parameter, i.e., the lack of only three pure solid phases exist and 
scatter of the points about the curve in liquid miscibility is complete. That this 
figure 4, is remarkable considering the is only approximately true was first 
scale of the diagram. The parameter that shown by Osborn (1942), who demon- 
changes most is b, and hence reflections strated that the join diopside-anorthite is 
with large k are most satisfactory for not binary owing presumably to solid 
determinative purposes along this join. solution of alumina in the pyroxene. This 

Volumes in this solid solution series result has been confirmed by Hytonen 

depart systematically from a straight line and Schairer ( Year Book 60) . To obtain 

connecting the end members in a way more precise information on the compo- 

that implies that they are nonlinearly sition of the pyroxene, careful X-ray 

related to composition. The departure work was done on a composition lying on 

from linearity, although apparently real, the diopside-anorthite join that was 

is not large. A straight line would fit the equilibrated with liquid at 1260°C and on 

data within 0.5 per cent. a composition lying in the ternary plane 

Edges of the unit cell change little that was equilibrated with liquid at 

along the diopside-enstatite join; the most 1220°C. In both, the departure of the 

conspicuous feature of figure 5 is the large unit-cell parameters from those of pure 

decrease in /3 with increasing content of diopside was small ; it was greater for the 

enstatite. The volume is essentially linear composition crystallized at the higher 

with composition over the limited range temperature. Hytonen and Schairer (Year 

of the data. Book 60, p. 137) indicate that at 1135°C 

There is a systematic difference be- in this system (a temperature well below 

tween our results and those of Hytonen the solidus) the pyroxene contains about 

and Schairer (Year Book 60, p. 136). They 3 per cent lime Tschermak's molecule, 

based a determinative procedure for They considered it probable that this 

diopsidic pyroxenes on the positions of amount of solid solution was metastable, 

the (150) and (510) reflections. We did and our results suggest the same. Because 

not read (510) because of possible inter- of the small shift in properties relative to 

ferences with (422) and (332), but we can experimental error, it is not possible to 

calculate its position from our data. For determine the direction in which these 

both reflections our 20 angles are about pyroxenes differ from pure diopside. 

0.1° larger than those reported by In all the sixteen compositions within 

Hytonen and Schairer. By assuming a the triangle diopside-anorthite-silica, the 

value for 0, it is possible to calculate a third solid phase first appeared on cooling 

and b for diopside from their data. Using at temperatures between 1218° and 

the extreme values of /3 in table 2, it is 1225°C. This implies that the system 

found that a = 9.751 to 9.755 A and diopside-anorthite-silica is very nearly 

b = 8.937 A. These values are sufficiently ternary, and that the stable pyroxene 



Si0 2 

TWO_ '» 

10 20 30 40 50 60 70 80 90 



Weight per cent CaAI 2 Sl °s 

Fig. 6. Equilibrium diagram for the system diopside-lime Tschermak's molecule-silica at 1 

must lie close to the plane of figure 6 at 
1222°C. The X-ray evidence implies that 
it is essentially pure diopside. The 
piercing point, or ternary eutectic, must 
be close to or at the thermal maximum on 
the quaternary univariant line connecting 
two quaternary eutectics. At one, wollas- 
tonite, diopsidic pyroxene, anorthite, and 
a silica phase coexist with liquid, and at 
the other enstatitic pyroxene, diopsidic 
pyroxene, anorthite, and a silica phase 
coexist with liquid. Determination of the 
composition of the latter eutectic is of 
great geologic significance, since it repre- 
sents the goal of crystallization of a 
simplified silica-saturated basalt at low 

High-pressure studies of the liquidus in 
this system have been carried out in a 
"single-stage" type of apparatus similar 
to that described by Boyd and England 
in Year Book 60. Results at 20 kb are 
shown in figure 7 ; a large number of runs 
have also been made at 30 kb, but this 

work is not yet ready for presentation. 
In all the work described the load pres- 
sure has been decreased by 3 per cent to 
allow for the effect of friction. 

The accuracy with which temperature 
can be measured is much lower at high 
pressures than at atmospheric pressure. 
At high pressures the uncertainty in 
temperature ranges from d=10°C in 
favorable cases to d=20°C or so. These 
estimates are based on the internal 
consistency and the reproducibility of 
some of our results. There is in addition 
a correction for the systematic effect of 
pressure on the emf of a thermocouple; 
this has been omitted because the elusive 
problem of quantitative determination of 
the correction remains to be successfully 
attacked. In contrast, at atmospheric 
pressure an accuracy of d=2°C can be 
achieved with care. 

The eutectic temperature in the binary 
system diopside-silica is raised by slightly 
more than 200°C by a pressure of 20 kb. 



This is essentially the same as the change 
in melting point of diopside itself. The 
composition of the eutectic is not measur- 
ably affected by pressure. In this system, 
as in all the work at 20 kb, quartz is the 
silica phase stable on the liquidus. The 
effect of pressure on the two-liquid region 
in this system has not been investigated. 

The system anorthite-silica is not 
binary at 20 kb because of the incongru- 
ent melting of anorthite, probably to 
corundum + liquid (Boyd and England, 
Year Book 60, p. 119). Between the fields 
of corundum and quartz on the liquidus 
there is a field of sillimanite. The temper- 
ature of lowest point on the liquidus, 
between the quartz and sillimanite fields, 
is 1540°C, and the composition is 48 
weight per cent Si0 2 . At atmospheric 
pressure the binary eutectic lies at 1368°C 
and*59 weight per cent Si0 2 . (Both silica 
contents are determined relative to lime 
Tschermak's molecule.) 

Changes produced by pressure in the 
system diopside-anorthite are greater 

than in the other limiting systems. Not 
only does anorthite melt incongruently at 
high pressures but also the amount of 
alumina in the pyroxene increases dra- 
matically. At compositions near anorthite, 
corundum and "/? alumina" both appear 
at high temperatures, with and without 
other crystalline phases. One of these 
alumina phases must be metastable on 
the liquidus, but it is not clear which. 
There is some evidence that, although 
corundum is stable at the anorthite 
composition, "(3 alumina" is the stable 
liquidus phase at neighboring magnesian 
compositions. It will be difficult to work 
out the correct relationship between these 
phases because of the stubbornness with 
which they both persist metastably. 

The nature of the minimum on the 
liquidus in this system has not yet been 
determined. It may be a cusp, resembling 
a eutectic, or it may be a smooth trough, 
depending on whether the minimum lies 
within the pyroxene field or at its 
boundary. Figure 7 is drawn as if this 

Si0 2 

20 Kilobars 

20 CaAI 2 Si 2 ( 


CaMgSi 2 0g 

CaAI 2 Si0 6 

Weight per cent 
Fig. 7. Equilibrium diagram for the system diopside-lime Tschermak's molecule-silica at 20 kb. 


minimum were a cusp at the boundary of kb is not even qualitatively similar to the 

the field, but future work may indicate system at atmospheric pressure, and 

the need for modification of this feature quantitative differences in melting be- 

of the diagram. The temperature and havior occur at all compositions. The 

composition of this point are 1480°C and most striking new features caused by 

71 weight per cent anorthite. At atmos- pressure are the incongruent melting of 

pheric pressure this point lies at 1274°C anorthite, the appearance of sillimanite 

and 43 weight per cent anorthite. on the liquidus, the appearance of quartz 

The complex relations at high pressures on the liquidus above 1000°C, and 

found in the systems diopside-anorthite extensive solid solution in the pyroxene, 

and anorthite-silica continue into the None of these effects occurs at atmos- 

triangle of figure 7. The fields adjacent to pheric pressure, and none of them could 

the anorthite composition have not yet have been inferred without high-pressure 

been fully delineated. There must be a experimentation. 

field of sillimanite, one of corundum, There is an interesting possible geo- 

probably one of "/3 alumina," and, near logical consequence of the shift in 

the piercing point, one of anorthite itself, composition of the piercing point with 

Although pyroxene, anorthite, and quartz pressure. If a small amount of liquid were 

are the solid phases present at the piercing formed by fractional fusion at 20 kb in 

point, the relationship there is not this system, it would have the approxi- 

ternary. There is a melting interval of mate composition diopside 2 2-lime Tscher- 

about 30°C. This point contains about 10 mak's molecule 4 2-quartz 36 . If this liquid 

weight per cent more anorthite than its were then decompressed suddenly, per- 

counterpart at atmospheric pressure, and haps by rapid upward intrusion, it would 

its temperature is raised about 125°C by arrive in a superheated condition and the 

20 kb. This is somewhat less than the composition of the liquid would be well 

increase in the minima in the diopside- inside the anorthite field at low pressure, 

anorthite and anorthite-silica systems. The liquid would crystallize large quan- 

That pyroxenes grown in this system tities of feldspar before other solid phases 

at 20 kb do not lie on the join diopside- appeared, which suggests a mechanism 

lime Tschermak's molecule is shown by for the origin of anorthosites. It is to be 

the failure of compositions, as determined expected that in the system albite- 

by X rays, to bear the relations to each diopside-silica the piercing point will 

other demanded by principles of phase behave in a similar way because of solid 

equilibria, and by the fact that different solution of jadeite in the pyroxene and 

parameters of the unit cell have values the eventual disappearance of albite from 

that would correspond to different the liquidus. An important unexplored 

amounts of lime Tschermak's molecule in question is the behavior of intermediate 

solid solution. Correction for enstatite in plagioclases ; it is not yet known whether 

solid solution, determined from /5, im- the mechanism outlined can produce 

proves the internal consistency of the feldspars of the compositions found in 

data, but the remaining discrepancies are anorthosites. 
probably large enough to be considered 

real. Presumably there is also magnesian The System MgSi0 3 -CaMgSi 2 6 

Tschermak's molecule (or corundum) in F R Boy ^ ^ and j p Schairer 
solid solution in the pyroxene. 

These results should dispel any doubts Mineral assemblages containing two 

that pressure, even in the absence of pyroxenes are of almost ubiquitous 

volatile constituents, can profoundly occurrence in mafic and ultramafic igne- 

affect phase diagrams. In part of the ous rocks. The two pyroxenes are usually 

range of compositions, the system at 20 a calciferous pyroxene, augite or ferro- 



augite, and a lime-poor hypersthene or 
pigeonite. Such pyroxenes show a wide 
variation in Mg/Fe ratio together with a 
more limited variation in Ca/(Mg + Fe). 
Understanding of the equilibria between 
such pyroxene pairs is of great petrologic 
interest, and the simplest system through 
which the problem can be approached is 
the join MgSi0 3 -CaMgSi 2 6 . 

Liquidus-solidus relations along this 
join were determined many years ago by 
Bowen (1914). Atlas (1952) was the first 
to study the subsolidus equilibria, and by 
means of fluxes he located the solvus and 
showed that two pyroxenes coexist at all 
temperatures below 1350°C. The crest of 
the solvus as determined by Atlas was 
shown to be about 50° below the solidus 
curve ; within this 50° interval a complete 
solid solution between MgSi0 3 and 
CaMgSi 2 6 seemed to exist. Atlas showed 
that orthorhombic MgSi0 3 was stable at 
temperatures up to 985°C. Although 
clinoenstatite is commonly obtained in 
runs on MgSi0 3 composition quenched 
from above 1000°C, Atlas argued that 
protoenstatite was the stable form in this 
range and that clinoenstatite formed in 
the quench. High-temperature X-ray 
studies by Foster (1951) showed that 
both orthorhombic MgSi0 3 and clino- 
enstatite could be inverted to proto- 
enstatite at temperatures above 1275°C. 
These studies proved that protoenstatite 
has a stable field at high temperature. 

Boyd and Schairer (Year Book 56) 
determined the solvus on this join by 
both dry and hydrothermal techniques. 
We found that the solvus intersected the 
solidus over a composition interval of 
about 15 weight per cent, so that a 
complete solid solution between 
CaMgSi 2 6 and MgSi0 3 does not exist at 
any temperature. 

Scatter of our preliminary results along 
the part of the solvus curve that defines 
the limit of solubility of CaMgSi 2 6 in 
protoenstatite led to a further investiga- 
tion of this part of the system. Results 
obtained this year indicate that the 
composition interval over which the 

solvus intersects the solidus is much 
wider than was previously indicated. Also 
evidence was found for an additional form 
of Mg-rich pyroxene, stable above 

Figure 8 shows the liquidus and 
subsolidus equilibria for the system 
MgSi0 3 -CaMgSi 2 6 . Quenching data for 
an extensive series of compositions along 
the join locate the liquidus temperatures 
and the equilibrium between crystals and 
liquid. Small circles indicate the tempera- 
tures as determined. Liquidus tempera- 
tures and the temperature of appearance 
of a Mg-rich pyroxene were determined 
for some compositions by Bowen (1914). 
Our results are in complete agreement 
with these data within the error of 
measurement. Bowen's data are given in 
figure 8 as triangles. 

Several of the crystal -liquid fields on 
the diopside side of the equilibrium 
diagram are so small that they cannot be 
shown on the scale of figure 8. This part 
of the diagram is expanded in figure 9. 
There is a series of solid solutions with a 
temperature minimum on the melting 
and freezing curves. These relations are 
interrupted by the incongruent melting of 
pyroxenes to forsterite and liquid. The 
composition of the binary reaction point 
forsterite + diopside -f- liquid was deter- 
mined as En 23 . 25 Di 76 .75 (weight per cent). 
The temperature is 1389° ± 2°C. Bowen 
(1914, p. 233, fig. 18) inferred these 
relations. Our data also tie in very closely 
with those of Schairer and Yoder on the 
system forsterite-diopside-silica given 
elsewhere in this report (pp. 75-82). 

Compositions on the join MgSi0 3 - 
CaMgSi 2 6 that are crystallized dry in 
the two-phase field form cryptoperthitic 
intergrowths of protoenstatite and diop- 
side. X-ray methods are therefore neces- 
sary to fix the composition of individual 
phases and to locate the solvus bounda- 
ries. Since protoenstatite inverts to 
clinoenstatite in the quench, the X-ray 
data are given for clinoenstatite. The 
shift of several reflections with compo- 
sition is sufficiently large to fix the 




T 1 1 1 1 1 1 1 1 r 


1 1 1 1 1 r 


Di ss +Fo + L 


i° ° A 

±2 *D JO D D / 

Di„+ L 

J L 

J L 

MgSi0 3 10 20 30 40 50 60 

Weight per cent 

J _L 


J L 



90 CaMgSi 2 6 

o Quench data, dry 

a Quench data , dry , Bowen (1914) 

a Single phase run, dry 

[] Single phase run , 500 bors H 2 

O Point on solvus boundary from dry run or runs 

Point on solvus boundary determined by 
homogenizing pyroxenes, dry 

Point on solvus boundary from runs at 500 bars H 2 
Single phase run, 1000 bars H 2 
i Two phase run . 1000 bars H 2 

Fig. 8. Liquidus and subsolidus equilibria in the system MgSi0 3 -CaMgSi 2 6 . Some liquidus 
points determined by Bowen (1914) are shown along with the data obtained by the authors. The 
value of 1025°C given for the inversion of rhombic MgSi0 3 to protoenstatite is based largely on 
extrapolation of preliminary high-pressure results. See figure 9 for an expanded view of the phase 
relations on the diopside side of the diagram. 





I39!.5 C 

- 10 


+ oo 


DU+Fo + L DU+L DU+L 






MgSi0 3 


70 80 90 CaMgSi ? 

Weight per cent 

Fig. 9. Expanded view of the phase relations in a part of the system MgSi0 3 -CaMgSi206. See 
figure 8 for the entire diagram. 

composition of a phase within ±2 per 
cent. Figure 10 shows the diffractometer 
patterns of the 220 peaks for a series of 
compositions across the solvus. These 
runs were made dry, without flux, and 
were held at 1365°C for 2 weeks. In the 
single-phase regions, the 220 reflection 
forms a sharp, single peak, but compo- 
sitions within the two-phase field show a 
double reflection, indicating the presence 
of two intimately intergrown pyroxenes. 

Figure 11 shows measurements of the 
220 reflection for a series of compositions 
from pure MgSi0 3 to En 6 5Di 3 5. Silicon 
was used as an internal standard for these 
measurements. The points for runs at 
1365°C shift progressively with compo- 
sition from pure MgSi0 3 to a bulk 
composition of En 8 oDi 2 o. For composi- 
tions richer in diopside than En 8 oDi 2 o, the 
composition of the enstatite is fixed at 
En 78 Di 2 2 independent of the bulk compo- 



T= 1365° 

Di 25 En 75 Di 30 En 70 Di 40 En 60 0' 50 En 50 Di 6c£ n 40 D'TO^O 

I* 29 

Fig. 10. Tracings of the 220 reflection from X-ray diffractometer patterns of a series of runs 
across the solvus in the system MgSi0 3 -CaMgSi 2 6 . The 220 reflection is a sharp, single peak in the 
single-phase regions of solid solution bordering the solvus. Within the solvus the 220 reflection splits 
into a doublet indicating the presence of a cryptoperthitic intergrowth of a Ca-rich diopsidic pyroxene 
and a Mg-rich clinoenstatite. 











i — : — ! — : — i — i — r 

-i — ; — i — r~— t — r~i — i — i — l I — r 

ill — ! — r~i — r-r 

O 13S5° 
© 1300° 

e J?ooiQ 

1365 # 

I l l l 

O 0-'^2I 


I I I I I I I I I I I I I I I I I I I I I I I I I 1 I I I I I ! I I I I I I I I 


MgSi0 3 

20 25 30 

Weight per cent 

35 40 

CoMgSi 2 6 >■ 

Fig. 11. Shift of the 220 reflection in clinoenstatites in the system MgSi03-CaMgSi 2 6 . Runs in 
the single-phase field at 1365°C fall on a smooth curve when plotted against the bulk compositions 
of the runs. Runs in the two-phase field at various temperatures fall off this curve, indicating that 
diopside is present as well as clinoenstatite and that the composition of the clinoenstatite is fixed at 
constant temperature, independent of the bulk composition of the run. 



sition; diopside is present as a phase in 
these runs, and the Mg-rich pyroxene is 
saturated. Two-phase runs at 1300° and 
1250°C are also shown in figure 11, and 
the points on the solvus determined by 
these data are plotted in the equilibrium 
diagram, figure 8. 

Reversals have been obtained at 1365°C 
for both sides of the solvus. At 1365°C 
the points on the solvus obtained by 
unmixing solid solutions are En 78 Di 2 2 and 
En 35 Di 6 5. Values obtained by homoge- 
nizing pyroxenes previously unmixed at 
lower temperatures are En 78 .5Di 2 i. 5 and 
En 34 Di 6 6. It has not proved possible to 
reverse the solvus at temperatures below 
1365°C. At most temperatures, however, 
two or more bulk compositions within the 
two-phase field gave the same value for 
the solvus. This is a strong presumption 
of equilibrium though not a proof of it. 

The solvus curve defining the limit of 
solubility of MgSi0 3 in diopside is little 
changed from our preliminary diagram 
(Year Book 56), although additional data 
have been incorporated. Present results, 
however, show that the solubility of 

CaMgSi 2 6 in protoenstatite is much 
more restricted than was indicated by 
our preliminary data. The intersection of 
the solvus curve with the solidus on the 
MgSi0 3 side is at a composition of 
En 75 Di 2 5 rather than En 55 Di45. The error 
in our preliminary results developed 
chiefly because runs were too short. The 
time required to reach equilibrium on the 
MgSi0 3 side of the solvus is 2 to 4 times 
as long as on the diopside side. Present 
results on the MgSi0 3 side are based on 
runs of 2 to 6 weeks. 

A curious phenomenon has been found 
in the temperature interval between 1365° 
and the solidus at 1405°C. Measurements 
of the 220 reflections for a series of 
compositions from MgSi0 3 to En 6 oDi 40 
crystallized at 1395°C are shown in 
figure 12. These points show a progressive 
shift of 220 with composition out to a 
bulk composition of En 75 Di 25 . Points for 
compositions richer in diopside fall off the 
curve and indicate that the solvus for 
this temperature is at En 76 Di 2 4. This 
point fits well with the points on the 
solvus obtained in the range 1250° to 

i i i i i i i i i i i i i i i i i i i | i i i i i i i i i i i i i i i i i i i i i i i i 

MgSi0 3 

i i i t I i i i i I i t i i I i i i i I i i i i I i i i i I i i i i I i t i i I i i i i 

10 15 20 25 

Weight per cent 



CoMgSi 2 6 

Fig. 12. Shift of the 220 reflection in Mg-rich pyroxenes crystallized at 1395°C. The curve for 
1365°C runs is reproduced from figure 11. 



1365°C (fig. 11). However, 220 measure- 
ments for pyroxenes in the single- phase 
field at 1395°C fall on a curve different 
from the curve obtained for 1365°C runs. 
The 1365°C curve is reproduced in figure 
12, and the difference can be seen to be 
insignificant for pure MgSi0 3 but appre- 
ciable for bulk compositions containing 

Full X-ray patterns of these 1395°C 
runs show that most of them are clino- 
enstatite but that some of them are 
orthorhombic enstatite. The position of the 
220 reflection does not seem to be 
significantly influenced by whether the 
crystal form is orthorhombic enstatite or 
clinoenstatite. There can be no doubt 
that the orthorhombic enstatite in these 
runs formed in the quench, inasmuch as 
the orthorhombic form has been proved 
to be unstable at temperatures above 
approximately 1000°C (see below). We 
have obtained orthorhombic enstatite in 
the quench in a considerable number of 
runs over a range of bulk compositions 
in this system but always at temperatures 
above 1385°C. We have, however, never 
observed it to form in the quench in runs 
on pure MgSi0 3 composition. 

Dry runs at all temperatures below 
1365°C in the single-phase field have 220 
spacings that fall on the 1365°C curve. 
These runs are normal clinoenstatite. A 
limited number of runs made above 
1395°C seem to fall on the 1395°C curve, 
whereas runs at 1385°C scatter in be- 
tween. These data suggest that there is 
a hitherto unrecognized form of Mg-rich 
pyroxene, stable above 1385°C. This form 
inverts in the quench to a distorted 
clinoenstatite or, sometimes, to ortho- 
rhombic enstatite. 

High-temperature X-ray studies are 
needed to confirm or disprove this 
suggestion. It is difficult to reach tem- 
peratures above 1350° to 1400°C with 
diffractometer heating stages, but the 
problem is being investigated in the 
laboratory of J. V. Smith, of the Univer- 
sity of Chicago. 

The orthorhombic enstatite ;=± proto- 

enstatite inversion is extremely sluggish 
at atmospheric pressure. Pure ortho- 
rhombic enstatite, prepared hydrother- 
mally, has been heated for more than 2 
months at 1080°C without change. Partial 
conversion to protoenstatite was observed 
in a dry run for 3 months at 1100°C. 
Addition of Na 2 W0 4 as a flux lowers the 
temperature at which orthorhombic en- 
statite will invert to protoenstatite. With 
Na 2 W0 4 , partial inversion of the ortho- 
rhombic form was observed at tempera- 
tures as low as 1025°C However, we were 
unable to convert clinoenstatite or proto- 
enstatite to orthorhombic enstatite at 
any temperature at atmospheric pressure, 
even with the use of Na 2 W0 4 as a flux. 

The presence of H 2 increases the 
reaction rate somewhat. Orthorhombic 
enstatite can readily be prepared from 
MgSi0 3 glass under hydro thermal con- 
ditions, but not from clinoenstatite or 
protoenstatite. A reversal of the transi- 
tion over a temperature interval of about 
50° was accomplished in runs at 500 bars 
H 2 which were also fluxed with Na 2 W0 4 . 
In evaluating the hydro thermal data, 
however, account must be taken of the 
effect of pressure on the inversion. 

Experiments made in single-stage appa- 
ratus (Boyd and England, Year Book 60) 
have shown that the orthorhombic ensta- 
tite ^ protoenstatite inversion is very 
sensitive to pressure. Pressure favors the 
orthorhombic form, and preliminary data 
indicate that the slope of the transition 
curve is about 75°/kb. A reversed bracket 
on the transition was obtained at 1525°C 
and 6.7 ± 0.6 kb. Extrapolation of pre- 
liminary hydrothermal and high-pressure 
data indicates an inversion temperature 
at atmospheric pressure of about 1025°C, 
in agreement with the runs at atmospheric 
pressure that were fluxed with Na 2 W0 4 . 
This value is in rough agreement with the 
inversion temperature of 985°C deter- 
mined by Atlas (1952) with LiF flux. 

The solubility of diopside in ortho- 
rhombic enstatite was determined by 
hydrothermal runs at 1000 bars H 2 in 
the temperature range 800° to 1000°C. It 



proved impractical to use X-ray methods 
on this part of the solvus. The runs were 
made long enough so that the presence 
or absence of diopside could be estab- 
lished by microscopic examination. 

Attempts to locate the solvus in the 
range 1000° to 1250°C on the MgSi0 3 
side of the diagram by crystallization of 
runs with Na 2 W0 4 flux were not success- 
ful. The flux differentially dissolves Si0 2 
and CaO, and so the products of these 
runs were usually pyroxenes + forsterite. 
As long as the two pyroxenes are on the 
join MgSi0 3 -CaMgSi 2 06 their mutual 
solubility should not be influenced by the 
presence of the forsterite. However, the 
220 spacings of the clinoenstatites in these 
runs indicated that they contain virtually 
no diopside. The results of the fluxed runs 
are inconsistent with the dry data at 
1250° to 1400°C and inconsistent with the 
usual form of a solvus curve. A check on 
these results was attempted by making a 
hydro thermal run at 1150°C and 500 bars 
H 2 0. In spite of a quartz buffer around 
the run it was severely desilicated, and 
the products were clinoenstatite + diop- 
side + forsterite. Again the 220 spacing 
of the clinoenstatite indicated that it 
contained virtually no diopside. Hydro- 
thermal runs on the CaMgSi 2 6 side of 
the solvus gave results in excellent 
agreement with dry runs, but for the most 
part these runs were shorter and at lower 
temperature, and desilication was not a 
problem. Various explanations are possi- 
ble for the failure of the fluxed runs to 
give consistent results. It may be that a 
variation of the (Mg + Ca)/Si ratio in 
the pyroxene is responsible. Evidence for 
the existence of such a variation in 
orthorhombic enstatite in high-pressure 
runs has been described (Boyd and 
England, Year Book 59). 

The System Diopside-Enstatite-Silica 
J. F. Schairer and H. S. Yoder, Jr. 

New studies on the join diopside- 
enstatite (see pp. 68-75) indicate that 
the solid solution of these pyroxenes is 

not complete and that a large solid 
immiscibility gap exists at the solidus. 
Additional relations on the pyroxene 
liquidus of the diopside-enstatite-silica 
system are thereby introduced that were 
not distinguishable by Bowen with the 
techniques available to him in 1914. 
Difficulties arose in the new studies of 
diopside-enstatite near the solidus be- 
cause of the complex changes within a 
small temperature interval, and it was 
realized that some advantage was to be 
gained by studying the pyroxene relations 
in the presence of the additional compo- 
nent silica. For these reasons a revision 
of the system diopside-enstatite-silica 
was undertaken. 

The revised liquidus diagram of the 
diopside (CaMgSi 2 6 )-forsterite (Mg 2 - 
Si0 4 ) -silica (Si0 2 ) system of which diop- 
side-enstatite-silica is a part is shown in 
figure 13. The data in the diopside- 
forsterite-enstatite (MgSi0 3 ) portion are 
those of Bowen (1914) with a suggested 
revision on the diopside solid solution 
(Di ss ) -forsterite (Fo) boundary curve. A 
minimum relation is proposed instead of 
the continuous drop of temperature to the 
CaMgSi 2 6 -Mg 2 Si0 4 join, which has been 
shown to be binary. The temperatures in 
the region of the proposed minimum are 
within experimental error, and the exact 
relations cannot be elucidated with 
present techniques. The two-liquid region 
is taken from the work of Greig (1927). 
The significant additions to Bowen's 
study are the realization of a boundary 
curve separating the fields of Di ss and 
Pr 8S (protoenstatite solid solution) and a 
minimum on the Di ss -Tr (tridymite) 
boundary curve. At point A the reaction 
Fo + L = Diss iakes place at a tempera- 
ture of 1405° ± 2°C. The Fo-Pr 8S 
boundary curve at temperatures higher 
than point A is the well known reaction 
curve involving Fo + L = Pr ss . The 
boundary curve Fo-Di ss at temperatures 
below point A is also a reaction curve for 
a part of its traverse to the minimum in 
the CaMgSi 2 6 -Mg 2 Si04-MgSi03 compo- 
sition triangle and involves Fo + L = 



CaMgSi 2 6 


Mg 2 Si0 4 

I7I3±5 # 


Weight per cent 

Fig. 13. Revised liquidus diagram at atmospheric pressure of the diopside-forsterite-silica 
system. Data in diopside-forsterite-enstatite portion are those of Bowen (1914); two-liquid region 
based on work of Greig (1927). 

Di 88 . At point B the reaction Pr ss + L 
= Di ss + Tr proceeds at a temperature 
of 1374° ± 2°C. There are reasons (see 
pp. 68—75) to believe that proto- 
enstatite is not the correct crystal 
structure of the solid solutions crystal- 
lizing on the liquidus labeled Pr ss . The 
powder X-ray diffraction patterns of 
MgSi0 3 -rich pyroxenes quenched from 
above about 1370°C have unique charac- 
teristics that cannot be specifically 
assigned to the now recognized forms of 
MgSi0 3 . They may be related to, but are 
different from, what Glasser and Osborn 
(1960) referred to as "high enstatite." It 
is to be understood that until more 
definitive data are at hand the crystal 

structure of the MgSi0 3 -rich pyroxene 
crystallizing in the fields labeled "Prss" 
is open to question. 

Because of the many significant 
changes that take place between the 
temperatures 1410° and 1370°C, isother- 
mal sections were studied in 5° intervals. 
Most of these are presented in figures 14 
to 20, on which are plotted the bulk 
compositions of runs carried out. Runs on 
other bulk compositions, at slightly 
higher and lower temperatures, of course, 
contribute to fixing the relationships. 

The relations at 1410°C, shown in 
figure 14, represent a temperature imme- 
diately above the first critical change. 
Only MgSi0 3 -rich pyroxenes, Pr ss , are 



stable. Tridymite is believed to be the 
stable phase of Si0 2 , but cristobalite is 
most often obtained metastably. Points 
C and D lie on the Fo-Pr S8 and Pr ss -Tr 
boundary curves of figure 13, respec- 
tively. The dashed crystal-liquid tie line 
in the field marked Pr ss + L in figure 14 
and similar tie lines in fields involving 
solid solutions in subsequent figures are 

At 1405°C, figure 15, the intersection 
of the pyroxene solvus and the solidus 
takes place. The reaction is Fo + L — » 
Di ss . The composition of Di ss is marked 
by the letter G, about Di 5 8Pr 4 2, and is the 
maximum content of MgSi0 3 that Di can 
contain in solid solution. The maximum 
amount of CaMgSi 2 6 held in solid 
solution by Pr, H, is also reached at this 
temperature; it is estimated to be about 
24 weight per cent Di. The point E lies at 
the junction of the Fo-Pr S8 , Fo-Di ss , and 
Pr 8S -Di ss boundary curves of figure 13. 

Point F marks a position on the Pr S8 -Tr 
boundary curve of figure 13. All bulk 
compositions in the triangle Mg 2 Si0 4 -(z- 
H become crystalline at essentially this 

Lowering the temperature to 1400°C, 
figure 16, gives rise to five new fields: 
Fo + Di S s, Di ss + L, Fo + Di 8 s + L, 
Pr S8 + Di S8 + L, and Fo + Pr 88 + Di 88 . 
Points /, /, and K lie respectively on the 
boundary curves of Fo-Di ss , Pr ss -Di 88 , and 
Pr 88 -Tr of figure 13. 

In figure 17 are given the relationships 
found at 1390°C. A second field of 
Digs + L has evolved. Neither the extent 
of solid solution in the very CaMgSi 2 6 - 
rich pyroxenes nor the precise limits of 
the crystal + liquid field were deter- 
mined. The points M, N, and lie 
respectively on the boundary curves 
Fo-Di 88 , Pr ss -Di ss , and Pr S8 -Tr of figure 13. 

In the temperature interval 1390° and 
1385°C (compare figs. 17 and 18) the 

CoMgSi 2 6 

Mg 2 Si0 4 

Weight per cent 
Fig. 14. Phase relations of diopside-forsterite-silica system at 1410°C. 



CaMaSi 2 6 

Mg 2 Si0 4 MgS,0 3 Si0 2 

Weight per cent 
Fig. 15. Phase relations of diopside-forsterite-silica system at 1405 °C. 

CaMqSi 2 6 

Mg 2 Si0 4 



Weight per cent 
Fig. 16. Phase relations of diopside-forsterite-silica system at 1400 °C. 




Di,.+ L 

Mg 2 Si0 4 



Weight per cent 
Fig. 17. Phase relations of diopside-forsterite-silica system at 1390 °C. 

CoMgSi 2 € 



Mg 2 Si0 4 MgSi0 3 ~'" 2 

Weight per cent 
Fig. 18. Phase relations of diopside-forsterite-silica system at 1385 °C. 



CoMgSi 2 6 

Mg 2 Si0 4 MgSi0 3 

Weight per cent 

Fig. 19. Phase relations of diopside-forsterite-silica system at 1375 °C. 

CaMgSi 2 6 


Mg 2 Si0 4 MgSi0 3 Si0 2 

Weight per cent 
Fig. 20. Phase relations of diopside-forsterite-silica system at 1370°C. 



remaining liquids in the CaMgSi 2 6 - 
MgSi0 3 -Mg 2 Si0 4 part of the system 
crystallize; presumably the last liquid is 
consumed at a minimum on the Fo-Di S8 
boundary curve. In addition, all bulk 
compositions on the join MgSi0 3 - 
CaMgSi 2 6 become crystalline. The reac- 
tion relationship of Fo + L — » Di ss 
terminates at various temperatures along 
the Fo + Di S8 + L curve, depending on 
the bulk composition. The points P and 
Q of figure 18 are on the Pr ss -Di ss and 
Pr ss -Tr boundary curves, respectively, of 
figure 13. No attempt was made to show 
the tie lines in the Di ss + L region because 
of the wide spread in possible orientations. 
Figure 19 portrays the relations at 
1375°C, indicating the nature of the 

closure of the lowest temperature liquids 
and the precursory conditions of the 
Pr 8S -f- L — > Diss + Tr reaction. The 
points R and S lie respectively on the 
boundary curves Pr ss -Di ss and Pr 88 -Tr of 
figure 13. Attention is called to the new 
fields Di S8 + Tr + L and Di ss + Tr that 
result from the complete crystallization 
of compositions on the CaMgSi 2 6 -Si0 2 
join at 1376° ± 2°C. 

All compositions in the system 
CaMgSi 2 6 -MgSi0 3 -Si0 2 are completely 
crystalline at 1371° ± 2°C, and the nature 
of the system is shown in figure 20 for a 
temperature of 1370°C. The important 
points X and Y are approximately 
Di 24 Pr 76 and Di 6 4Pr 3 6, respectively. With 
the exception of polymorphic transitions 













Clino- pyroxene 

» pyroxene + liquid 
« wollastonite + liquid 

• pyroxene + wollastonite 

+ liquid 

Mg 5 Ca 5 Si0 3 

30 50 

Mol per cent 


Fe 5 Ca 5 Si0 3 

Fig. 21. Preliminary T-X section across the join Mgo. 6 Cao.6Si0 3 (diopside)-Fe .6Ca .5SiO 3 
(hedenbergite). Total pressure = 1 atm. Partial oxygen pressure is in equilibrium with iron + wiistite. 



and exsolution phenomena little change 
in the character of the isothermal sections 
takes place with further lowering of the 

The application of the revisions of the 
diopside-forsterite-silica system to peno- 
logical problems is of exceptional import. 
No attempt will be made here to evaluate 
the new implications. Light is cast on the 
presence or absence of hypersthene in 
natural rocks (see Tilley, 1961), the 
reaction relations of olivine with liquid 
to produce hypersthene in some cases and 
augite in others, the phenocryst-ground- 
mass relations of hypersthene, pyroxene 
zoning and exsolution, and many aspects 
of fractionation in magmas and the 
evolution of derivative magmas. 

Preliminary Results on Melting Relations 

of Synthetic Pyroxenes on the 

Diopside-Hedenbergite Join 

A. C. Turnock 

A study of melting relations of the 
Mg-Fe-Ca pyroxenes, with compositions 
in the three-component system MgSi0 3 
(En)-FeSi0 3 (Fs)-CaSi0 3 (Wo), has been 
started with compositions along the join 
diopside (MgCaSi 2 6 )-hedenbergite 
(FeCaSi 2 6 ) using a controlled-atmos- 
phere quenching furnace with a total 
pressure of 1 atmosphere and a partial 
pressure of oxygen that would be in 
equilibrium with Fe -f- Fei_ x O. The 
oxygen pressure was regulated by mixing 
carbon dioxide and carbon monoxide 
(Darken and Gurry, 1945). 

A diagram of the experimental results 
is presented in figure 21. There is a 
complete series of monoclinic pyroxenes 
from diopside to hedenbergite, but the 
Fe-Mg substitution causes important 
changes in the stability of the pyroxenes. 
The Mg-rich pyroxenes, as shown on the 
left-hand side in figure 21, melt through 
a temperature range given by the solidus 
and liquidus curves, and this part of the 
diagram is essentially binary. The effect 
of iron content in lowering the tempera- 

tures of the solidus and liquidus is 
pronounced. Pyroxenes richer in the 
hedenbergite molecule than about 60 per 
cent, however, will not melt but convert 
to wollastonite solid solution. These two 
phases may be polymorphs across the 
range Hed 76 to Hed 100. The wollas- 
tonite solid solution persists metastably 
at lower temperatures, and in the diagram 
the two curves that define its subsolidus 
conversion to pyroxene are based on the 
reverse reaction, pyroxene — ■» wollastonite 
solid solution. Time studies of this 
reaction satisfactorily showed that the 
transition interval was not occasioned by 
incomplete reaction, and the two curves 
intersect the liquidus at positions that 
satisfy boundary points for the field 
"pyroxene + wollastonite ss + liquid." 

Wollastonite solid solution melts 
through an interval of about 90°C. In the 
low-temperature part of the field "wollas- 
tonitess + liquid" there is probably 
another field, "wollastonite ss + liquid + 
tridymite." Small amounts of a silica 
phase have been observed, but there is 
not yet enough information to draw in a 
field boundary. 

Metamorphic Petrology 

Metamorphic Reactions Involving 
Two Volatile Components 

H. J. Greenwood 

Many metamorphic reactions involve 
more than one volatile or mobile compo- 
nent and are therefore influenced by 
pressure, temperature, and the compo- 
sition of the coexisting fluid phase. The 
equilibrium relationships may be por- 
trayed in a variety of ways, for example, 
by plotting the chemical potentials of the 
mobile components against one another 
at constant temperature and pressure 
(Korzhinskii, 1959; Zen, 1961). Alterna- 
tively, the situation may be represented 
on an isobaric T-x diagram, on which are 
plotted the temperature and the compo- 
sition of the coexisting fluid phase, the 
other components being regarded as 



nonvolatile or immobile. This kind of 
diagram has some advantages over the 
chemical potential, or \ii versus /*/, 
diagram, not the least of which is its 
direct use of the measurable variables 
temperature, pressure, and composition. 

Equations have been derived for the 
equilibrium boundaries between reacting 
phase assemblages in such systems. These 
have the same form as the usual expres- 
sions for crystal-liquid equilibria, but 
they do not carry the restriction that the 
relative proportions of the two volatile 
components are limited by the propor- 
tions of the other components. The effect 
of removing this restriction is to make 
stable many reactions that would nor- 
mally be regarded as metastable. 

The slope of an equilibrium boundary 
for a reaction taking place at constant 
pressure in the presence of a one-phase 
binary fluid having zero enthalpy of 
mixing is 

\dx 2 /p AS \x 2 xj 

where x 2 is the mole fraction of component 
2 in the fluid and v 2 is the stoichiometric 
coefficient of component 2 in the reaction. 
All reactions that take place in such 
systems can be expressed by the general 

aA -^bB + v\ + v 2 

in which a moles of solid phases A react 
to give b moles of solid phases B and v\ 
and v 2 moles of the volatile components 
1 and 2, respectively. The equation of the 
reaction should be written so that 

\v\\ + | v% | = 1 and vi + v 2 ^ 

to make the stoichiometric coefficients 
equivalent to the mole-fraction compo- 
sition of the gas given off in the reaction. 
Inspection of these equations reveals 
several points of interest to metamorphic 
petrology. If v 2 = 0, 

If pi = 

dT \ RT ( J_ 

,dx 2 / P AS \x 2 


If Vi > and v 2 > 0, 

(dT/dx 2 ) P = (T max ) 

where v x = x h v 2 = x 2 . (3) 

If vi = —v 2 , equal amounts of compo- 
nents 1 and 2 appear on opposite sides of 
the reaction, and their entropies tend to 
cancel, making AS for the reaction small 
and (dT/dx 2 )p correspondingly large. 
Accordingly, as 

A£->0, (dT/dx 2 )p-><x> (4) 
If -1 < v x < 0, 1 > v 2 > 

(l^il < W), 

+ 00 > 

If 1 > Vi > 0, - 1 < v 2 < 


— 00 < 

(U)p < S(-ir) (6) 

The importance of these rather terse 
statements to metamorphic petrology can 
best be appreciated by examining some 
geologically interesting reactions that 
lend themselves to this treatment. Equa- 
tion 1 describes a reaction in which only 
component 1 is given off (H 2 in fig. 22). 
As an example of such a reaction we might 

Ca 2 Mg 5 Si 8 22 (OH) 2 


2CaMgSi 2 6 + 


dT\ = RT 

,dx 2 ) P AS 


3MgSi0 3 + Si0 2 + H 2 

Enstatite Quartz 

Equation 2 describes a reaction in which 
only component 2 is given off (C0 2 in 
fig. 22). Example (see fig. 22): 

MgC0 3 -> MgO + C0 2 

Magnesite Periclase 

Equation 3 describes a reaction in which 
both volatile components are given off, 
such as 

iCa,Mg 5 Si 8 22 (OH) 2 + |CaC0 3 + 

Tremolite Calcite 



|Si0 2 

£CaMgSi 2 6 + |C0 2 + |H 2 


(See fig. 22, T max at x C o z = 0.75.) 
Equation 4 describes a reaction such as 

Mg(OH) 2 + C0 2 


MgC0 3 -f H 2 


(See fig. 22, vertical boundary.) 

Equation 5 describes a reaction such as 
|CaMg(C0 3 ) 2 + |Si0 2 + iH 2 -> 



fCa 2 Mg 5 Si 8 22 (OH) 2 + |CaC0 3 + |C0 2 

Tremolite Calcite 

Equation 6 describes a reaction such as 
4H 4 Mg 3 Si 2 9 + 9CaC0 3 + 5C0 2 -> 

Serpentine Calcite 

Ca 2 Mg 5 Si 8 22 (OH) 2 + 


7CaMg(C0 3 ) 2 + 7H 2 



H 2 

025 0.5 0.75 

Mole fraction C0 2 


Fig. 22. Diagrammatic sketch illustrating 
the six types of crystal-vapor equilibrium reac- 
tions in binary gas mixtures. H 2 is component 
1, and CO2 is component 2, of the equations. 

We may, for the sake of discussion, 
regard these reactions as models of 
metamorphic isograds. The most obvious 
feature is that an isograd defined on the 

basis of a reaction that evolves one 
volatile component may cross an isograd 
defined on the basis of a reaction that 
evolves the other volatile component. In 
addition, a plot like figure 22 may be 
regarded as a map of an area that has a 
gradient in the proportions of C0 2 and 
H 2 across it at a large angle to the 
thermal gradient. If such an area could 
be found in the field, containing rocks of 
suitable compositions, it should be pos- 
sible to demonstrate the crossing of 
isograds. In reactions like the formation 
of diopside from tremolite, calcite, and 
quartz, it is clearly of great importance 
to know something of the composition of 
the fluid phase in equilibrium with the 
minerals before coming to any conclusion 
about the temperature of metamorphism, 
even assuming some knowledge of the 
total pressure. 

Reactions like those described by 
equations 4, 5, and 6 are perhaps the most 
interesting of all when they are regarded 
as isograds. Their steep slopes in T-x 
plots like figure 22 show that the progress 
of many such reactions is affected more 
by the composition of the coexisting fluid 
phase than by either temperature or 
pressure. This observation leads directly 
to the concept of an isograd that is 
essentially neither isotherm nor isobar 
but that provides a firm limit on the 
composition of the fluid with which the 
minerals of the rock could have been in 
equilibrium. It cannot be too strongly 
urged, therefore, that when an isograd is 
under discussion the chemical reaction be 
precisely defined. 

Experiments are now under way that 
will fix the positions of reactions of the 
sort just discussed in the system MgO- 
CaO-Si0 2 -H 2 0-C0 2 . The apparatus is 
essentially the same as was used in an 
earlier investigation of the system 
NaAlSi 2 6 -H 2 0-argon (Greenwood, 1961) 
in which the solid phases are held in open 
capsules in a bomb containing a mixture 
of C0 2 and H 2 0. Pressure and tempera- 
ture are measured, and the composition 
of the gas is analyzed at the end of each 



run. The stability of wollastonite has 
been studied rather fully, and preliminary 
data are now available on a number of 
other equilibria. Figure 23 shows the 
stability relations of wollastonite in 
mixtures of C0 2 and H 2 at 1000 and 
2000 bars. All the reactions shown repre- 
sent reversals of the equilibrium. The 
data are in good agreement with those 
of Harker and Tuttle (1956), assuming 
that the CO 2 and H 2 mix ideally. This 
apparent close approach to ideal mixing 
is probably illusory, because it seems 
likely that the gas mixture contains three 





£ 500 



P = 2000 bars 

Calcite + Quartz 


0.25 0.50 0.75 

Mole fraction CO2 


Fig. 23. Stability relations of calcite, quartz, 
and wollastonite in mixtures of H 2 and C0 2 . 
Circles, 2000 bars; rectangles, 1000 bars. 

molecular species rather than two. Re- 
action between C0 2 and H 2 to produce 
H2CO3 could easily produce the same 
effect as ideal mixing of C0 2 and H 2 on 
a solid-gas equilibrium. The accumulation 
of more data on mineral equilibria in the 
mixtures will allow direct estimation of 
the extent of reaction between H 2 and 
CO 2 . In addition to the wollastonite 
reaction, preliminary runs indicate that 
the reaction of talc, calcite, and quartz to 

give diopside occurs at a lower tempera- 
ture than the wollastonite reaction. 

iMg 3 Si 4 O 10 (OH) 2 + |CaC0 3 + JSi0 2 -» 

Talc Calcite Quartz 

3CaMgSi 2 6 + iH 2 + |C0 2 

According to equation 3 this reaction 
curve must pass through a maximum in 
temperature where #002 = 0.75. At a 
total pressure of 1000 bars the tempera- 
ture of this maximum has been deter- 
mined to be 600° =fc 25°C, at least 25° 
lower than the wollastonite curve at this 
composition, confirming the field obser- 
vation that diopside can be formed at 
lower temperatures than wollastonite. 

Synthesis and Stability of Anthophyllite 
H. J. Greenwood 

Pure magnesian anthophyllite, though 
of limited natural occurrence, has been 
the subject of considerable attention in 
the geological and geochemical literature. 
A significant part of this interest may be 
traced to the classic paper by Bowen and 
Tuttle (1949) on the system MgO-Si0 2 - 
H 2 0, describing experiments in which 
they were unable to demonstrate the 
stability of this mineral. Since that time 
there have appeared a number of papers, 
both theoretical and experimental, agree- 
ing with their conclusion that the mineral 
is not stable in the presence of excess 
H 2 0. Anthophyllite has been under study 
at this Laboratory for more than three 
years. Last year (Greenwood, Year Book 
60, p. 105) the existence of a stability 
range in the presence of excess H 2 was 
indicated. Continuation of the study has 
produced enough data to allow detailed 
discussion of the upper and lower stability 
limits and nucleation kinetics of the 
mineral. Fyfe (1962) has recently given 
an account of some experiments with 
unanalyzed natural materials, which also 
indicate that the mineral has a range of 
stability in the presence of excess H 2 0, 
but which do not define the reactions by 



which anthophyllite may form from 
lower-temperature assemblages including 

Synthesis of anthophyllite is not easy. 
Most of the hydrothermal experiments 
that have failed to produce the mineral 
have failed because of its extreme 
reluctance to nucleate, even well within 
its own field of stability. Glasses, oxide 
mixes, and mixtures of the other minerals 
in the system MgO-Si0 2 -H 2 in various 
proportions do not crystallize directly to 
anthophyllite, even when maintained 
within the anthophyllite stability field 
for periods as long as 4 months. The use 
of solutions of MgCl 2 and of HC1 did not 
seem to facilitate the nucleation. The only 
way in which anthophyllite could be 
produced in the absence of preexisting 
nuclei was by the metastable decompo- 
sition of talc at 1000 bars and 830°C for 
a period of 20 hours. This procedure for 
obtaining starting materials with which 
to test the stability of anthophyllite was 
employed throughout most of the investi- 
gation. The materials so obtained are 
known to be of the pure magnesian 
anthophyllite composition; they have a 
refractive index of ftz = 1.615 and give 
all the characteristic X-ray reflections of 
natural anthophyllite. 

This metastable nucleation of antho- 
phyllite about 80°C above its upper 
thermal stability limit at 1000 bars is 
readily explained on the basis of crystal 
structure. It seems unlikely that the 
mineral could form in the complete 
absence of any nuclei so far above its 
point of thermal breakdown. The require- 
ment that talc be used as the starting 
material suggests that the talc is supply- 
ing nuclei having the anthophyllite 
structure. Disintegration of the sheets of 
tetrahedra in the talc structure into 
strips could provide the necessary struc- 
tural units having a an th = c ta ic and c an th A 
frtaic = 30°. Rate studies support this 
conclusion. Talc held under these condi- 
tions breaks down rapidly (half-life 73^2 
hours) to anthophyllite, protoenstatite, 
and quartz. Anthophyllite increases faster 

than the other reaction products until 
16J/2 hours have elapsed, after which time 
it decreases, becoming undetectable after 
120 hours. The final products are the 
phases stable at 830°C and 1000 bars, 
quartz and enstatite. Von Gehlen (1962) 
has shown that talc heated at 1300°C at 
atmospheric pressure is transformed into 
protoenstatite and quartz, with the 
protoenstatite oriented in the same way 
with respect to the parent talc structure 
as is postulated here for anthophyllite. 

Rate studies on the decomposition of 
an analyzed natural anthophyllite (A1 2 3 , 
1.94; FeO, 11.12; CaO, 0.64 weight per 
cent) at 1000 bars have shown that for 
times up to 40 days at 800°C (50°C above 
its stability limit) the amount of break- 
down is barely perceptible, although at 
850°C it is complete in 5 days. The 
excellent crystallinity and relatively 
coarse grain size of the natural material, 
perhaps together with its departure from 
the pure magnesian end member, evi- 
dently make it very slow to react, and 
extrapolation indicates that 4 or 5 months 
would be required to decompose the 
mineral near the equilibrium curve. 

Starting materials for the runs used to 
define the limits of stability of the 
amphibole were prepared in the manner 
described. Oxide mixes on each of the 
bulk compositions MgO-Si0 2 , 7MgO- 
8Si0 2 , and 3MgO«4Si0 2 were separately 
crystallized well inside the stability field 
of talc, and then given the heat treatment 
to produce anthophyllite, protoenstatite, 
and quartz from the talc, together with 
the other phases inherited from the 
crystallization in the talc field. These 
starting materials consisted of various 
mixtures of forsterite, enstatite (proto 
and clino), quartz, cristobalite, antho- 
phyllite, and talc. This is an obvious 
disequilibrium mixture on the requisite 
bulk composition containing as nuclei all 
the phases to which the mixture could 
finally crystallize at equilibrium. Runs 
are from 3 to 4 months in duration, at the 
end of which time the reactions are from 
about 30 to 100 per cent complete. 













MgO p o En Si0 2 




Temperature ,°C 

Fig. 24. Stability relations of talc, anthophyllite, and enstatite 

Depending on the bulk composition, the 
final products consist of mixtures of any 
pair of the phases forsterite, orthoensta- 
tite, anthophyllite, talc, quartz. Proto- 
enstatite, clinoenstatite, and cristobalite 

The experimental results are shown in 
figure 24. All runs shown represent 
reversals of the reactions represented by 
the equilibrium curves. The arrows beside 
the run symbols indicate the direction 
from which the equilibrium was ap- 

proached. Shorter runs in which no 
reaction occurred are not considered 
significant and are not reported. Both the 
upper and the lower stability limits of 
anthophyllite have been determined by 
reversing the reactions at several points, 
and they are considered to represent 
stable reversible equilibria. The upper 
stability limit of talc and the lower limit 
of enstatite in equilibrium with H 2 are 
indicated by the two dashed curves. Both 
reactions must occur in this narrow 


interval, but it has not been possible to inary literature survey reveals only a 

determine their relative positions. small number of chemical analyses of 

In summary, anthophyllite has a range chlorites from low-grade metamorphic 

of stability in the presence of excess H 2 0; rocks. Turnock (Year Book 59) showed 

and talc, anthophyllite, and H 2 can that iron chlorite may exist in equilibrium 

coexist in stable equilibrium over a with quartz up to almost 600°C at a total 

narrow temperature interval. pressure of 2000 bars. The magnesium 

analogues of these chlorites may be 

Quartz-Chlorite Assemblages in the System represented in the system MgO-Al 2 3 - 

MgO-Al 2 Os-Si02-H 2 Si0 2 -H 2 0, studied by Yoder (1952) and 

J.J. Fawcett and H.S. Yoder, Jr. Roy and Roy (1955); neither of these 

studies, however, shows chlorite and 

Investigation of synthetic systems quartz as a compatible mineral pair in 

closely related to low-grade metamorphic the temperature range 450° to 900°C at 

rocks offers many opportunities to the 15,000 psi (Yoder) or 130° to 1300°C 

experimental petrologist. Experimental between 5000 and 30,000 psi (Roy and 

study of the low-grade metamorphic Roy). 

rocks is handicapped by the slow rates of As natural occurrences indicate that 

reaction, which are probably due to the chlorite and quartz may constitute a 

absence of a liquid silicate phase so that stable assemblage, a series of experiments 

diffusion of ions must take place either in have been performed in an attempt to 

the solid state or, more likely, through a define a stability field for Mg chlorites 

gas phase. To understand chemical and quartz. Starting materials for the 

reactions in low-grade metamorphic rocks experiments were glasses whose compo- 

and quantitatively evaluate prevailing sitions, shown in figure 25, for the most 

physical conditions it is important to part lie on the anhydrous join anthophyl- 

study in the laboratory the phase rela- lite (7MgO-8Si0 2 )-Mg gedrite (5MgO- 

tions of the minerals and groups of 2Al 2 3 -6Si0 2 ); the glasses were prepared 

minerals that play a significant role. under the supervision of Dr. Schairer in 

Chlorite and quartz are two of the most connection with a study of these amphi- 
common minerals present in low-grade boles. Other glasses were supplied by 
metamorphic rocks. Indeed, they charac- Schairer and by Yoder. The compositions 
terize the chlorite zone of progressive plot between the composition of the 
metamorphism and often persist into the chlorite solid solution series and quartz as 
biotite and even garnet zones (Barrow, projected from H 2 onto the face 
1893; Tilley, 1925; Mason, 1962). In MgO-Al 2 3 -Si0 2 , in the system MgO- 
terms of the facies concept of meta- Al 2 3 -Si0 2 -H 2 0. Determinative runs were 
morphism the quartz-chlorite assemblage made at pressures of 2 and 5 kb in cold- 
characterizes in particular the quartz- seal hydrothermal bombs, the duration of 
albite-muscovite-chlorite subfacies of the the runs varying from 1 to 6 weeks. At 
greenschist facies (Fyfe, Turner, and 2 kb and temperatures of 400° and 600°C 
Verhoogen, 1958, p. 218). The large talc was found to grow readily, in the 
volume and wide distribution of these early stages of the runs, with charges 
rocks on the earth's surface suggest that containing less than 15 per cent A1 2 3 . 
these two minerals may exist together in Some of the talc reacts only very slowly 
equilibrium over a considerable, though to produce a more stable mineral assem- 
as yet unspecified, range of temperature blage. Talc also grows from compositions 
and pressure. containing more than 15 per cent A1 2 3 , 

Hutton (1940) suggested that the but the smaller amounts produced from 
chlorites of low-grade metamorphic rocks these compositions are consumed corn- 
are dominantly iron rich, but a prelim- pletely by reaction in less than 3 weeks. 



Si0 2 












A! 2 3 

Weight per cent 

Fig. 25. Composition of glasses (+) used in the determination of the quartz-Mg chlorite stability 
field. The projected composition of Al montmorillonite should plot on top of pyrophyllite and not 
kaolinite, as shown. 

Preliminary results are illustrated in 
figure 26. Quartz and chlorite are stable 
together up to almost 600°C, but the 
projected shape of the chlorite-quartz 
field is a reflection of the stability of the 
chlorite solid solution series. At lower 
temperatures (450°C) a wide range of 
chlorite compositions is stable; with 
increasing temperature the range becomes 
narrower. The chlorite coexisting with 
quartz at the maximum temperature of 
the quartz-chlorite field contains about 
20 weight per cent A1 2 3 . An increase or 
decrease in the A1 2 3 content of the 
chlorite results in a reduction of the 
maximum temperature of the quartz- 
chlorite stability field. Montmorillonite 
crystallized at temperatures below 450°C. 
Talc formed metastably in the less 
aluminous compositions in the quartz- 

chlorite field, but in the longer runs 
(6 weeks) quartz and chlorite represent a 
more stable assemblage. Many runs 
produced the 7 A polymorph of the 
chlorites (aluminous serpentine, Yoder, 
1952; septechlorite, Nelson and Roy, 
1958), but most runs of 3 weeks or longer 
produced the 14 A chlorite. 

Chlorite and quartz coexist, together 
with a third phase (talc or cordierite) on 
either side of the quartz-chlorite stability 
field. These two small fields are limited at 
higher temperatures by the field of talc 
+ chlorite + cordierite and, for more 
Si0 2 -rich compositions than those in the 
section of figure 26, by talc + cordierite 
+ quartz. 

At temperatures in the 700° to 850°C 
range a careful search has been made 
for the phases in the anthophyllite 












o o o 


O + 


O o 


o o 

7MgO-8Si0 2 -H 2 

6Mg0-AI 2 3 -7Si0 2 H 2 
MgSi <^-r AIAI 

5MgO-2AI 2 036Si0 2 H 2 

Fig. 26. Phase relations along the extended join anthophyllite (Mg 7 Si 8 022(OH) 2 )-Mg gedrite 
(Mg 5 Al 4 Si 6 22 (OH) 2 ) in the system MgO-Al 2 3 -Si0 2 -H 2 0. 

(Mg 7 Si 8 22 (OH) 2 )-Mg gedrite (Mg 5 Al 4 - 
Si 6 2 2(OH) 2 ) group. Amphiboles of un- 
known composition have been synthe- 
sized, but they have been shown to be 
unstable. In view of the existence of a 
stability field for pure anthophyllite 
(Greenwood, this report), work on the 
aluminous anthophyllites will be con- 
tinued in the coming year. 

To test the results obtained with 
synthetic starting materials, naturally 
occurring minerals have been used as 
starting materials for some experiments; 
mixtures of talc + cordierite have been 
converted to chlorite + quartz + minor 
talc within the quartz-chlorite field of the 
synthetic materials, and similarly mix- 
tures of quartz + chlorite have been 
converted to cordierite + talc + quartz 
within the stability field of that assem- 

blage, as indicated by runs using glasses. 

An increase in pressure from 2 to 5 kb 
raises the temperature of maximum 
stability of the quartz-chlorite assemblage 
from about 600° to 625°C. It is of interest 
to note that the quartz-chlorite reaction 
curve is about 100°C below the maximum 
stability of clinochlore at 2000 bars, 
whereas the analogous muscovite-quartz 
curve is only 15°C below the upper 
stability limit of muscovite (Yoder and 
Eugster, 1955). The effect of pressure on 
the lower stability limits of the quartz- 
chlorite stability field has not yet been 

All runs below 450°C produced a 
montmorillonite phase, but there is, on 
theoretical grounds, a narrow tempera- 
ture interval between the montmorillonite 
field and the quartz-chlorite field in which 


the stable assemblage is quartz or talc + due to complex reactions involving chlo- 

montmorillonite + chlorite. The precise rite, quartz, or both. 

phase relationships have not yet been 

worked out at lower temperatures. Alkali-Rich Igneous Rocks 

Nelson and Roy (1958) determined the AND Minerals 
maximum stability of the chlorite solid 

solutions in the absence of quartz to be The System Na20-Al 2 0s-Fe 2 03-Si02 and 

710°C at 1000 atm. The maximum Its Bearing on the Alkaline Rocks 

stability of the chlorite + quartz assem- Jm Fm Schairer and D. K. Bailey 
blage was found in the present work to be 

675°C at 2 kb. Although the pressures The alkaline rocks, by virtue of their 

are not equivalent, the data are in accord uncommon chemistry and mineralogy — 

with the general rule that heterogeneous with essential amounts of feldspathoids 

reactions must take place within the and alkali pyriboles in a wide range of 

stability fields of the reactants and proportions, and often with minor 

products. The present data indirectly amounts of rare minerals — have engaged 

suggest that the maximum stability of the attention of petrographers and ana- 

the alumina-poor and alumina-rich chlo- lysts to an extent that belies their 

rites is lower than that suggested by quantitative importance; consequently, a 

Nelson and Roy. The breakdown products large body of information is available on 

of clinochlore and amesite obtained by them, and the problem of their origin has 

Nelson and Roy (1958) do not appear to provoked much speculation and argu- 

represent equilibrium assemblages. Runs ment. It does not follow, however, that 

of much greater duration than those the considerable attention devoted to 

carried out by Nelson and Roy on the these rocks has been misdirected, for the 

chlorite breakdown may yield assem- typical alkaline centers, in common with 

blages different from the forsterite, talc, kimberlites and carbonatites (with which 

and spinel obtained by those authors. alkaline rocks are frequently associated), 

The information on the limits of are restricted to the stable continental 

stability of the quartz-chlorite mineral areas of the earth's crust, and the rocks 

assemblage may be applied in a general are probably the surface expression of 

way to the conditions of formation of deep-crustal and subcrustal activity in 

potassium-deficient low-grade metamor- epeirogenic zones. The natural compo- 

phic rocks. As the chlorite minerals in the sitional ranges in the alkaline rocks, and 

rocks always contain iron, however, the the awesome range of rock names, make 

data can only indicate maximum tern- generalizations about their composition 

peratures in the pressure range under difficult, but a large proportion of the 

consideration. Moreover, the effect of rocks are essentially assemblages of 

other phases such as muscovite, biotite, nepheline (or related feldspathoids such 

feldspar, or epidote on the quartz-chlorite as sodalite and cancrinite) , sodic pyroxene 

stability field is unknown. (or amphibole) , and alkali feldspar, thus 

It is clear from these results that falling into two broad groups: ijolites 

previous investigators had not made runs (nepheline-pyroxene rocks) and foyaites 

of sufficient duration to obtain the (nepheline-f eldspar-pyroxene rocks) . Both 

quartz-chlorite assemblage. It can now types bulk largely in alkaline complexes, 

be concluded that the synthetic studies are commonly associates of carbonatite, 

support the field observations that Mg- and hence figure prominently in theories 

rich as well as Fe-rich chlorites can of the origin and the differentiation of 

coexist with quartz over a large P-T these rocks. The study of the system 

range. Limitation of the coexistence of Na 2 0-Al 2 03-Fe203-Si02 offers an oppor- 

quartz and chlorite in the natural rocks is tunity to observe the essential rock- 



Fe 2 Oj 














51 e_ 

6 45_ 


.5No 2 Fe 2 3 -8Si0 2 
6No 2 4Fe 2 O s 5S>0 2 



2. 5Na 2 Fe 2 3 8Si0 2 

3. (No 2 4SI0 2 ) COMPN. 





A_AI 2 OyFe 2 O3-(Na 2 4Si0 2 ) 





F_NEPHELINE-ACMITE - (No £ 4Si0 2 ) 


Fig. 27. The system Na20-Al 2 03-Fe 2 03-Si02 to show relations of the compounds and joins 
studied. Three faces of the tetrahedron have been laid flat in the plane of the base. 

forming minerals nepheline, acmite, and 
albite in equilibrium with liquids the 
compositions of which are analogous to 
those of natural alkaline rocks. 

Study of equilibrium within the quater- 
nary system Na 2 0-Al 2 3 -Fe 2 3 -Si0 2 at 1 
atmosphere pressure began with an 
examination of the join jadeite-acmite in 
1948-1949 (Year Book 48, p. 32) as part 
of a more general study of the stability 
relations of jadeite. This join is not 
binary, and it was found that the primary 
phase for compositions acmite 100-20 was 
hematite or hematite-corundum solid 
solution, and that compositions 
jadeite 100-80 gave nepheline-albite solid 
solution as the primary phase with 
hematite-corundum solid solution as the 

second phase. This meant that relations 
in the join could be described correctly 
only in terms of the quaternary system, 
and work was started in five joins in this 
system the following year. 

Phase-equilibrium data for the bound- 
ing ternary system Na 2 0-Fe 2 3 -Si02, 
establishing the incongruent melting of 
acmite, have been published by Bowen, 
Schairer, and Willems (1930) and for the 
system Na 2 0-Al 2 3 -Si0 2 by Schairer and 
Bowen (1956). Because reduction of 
Fe 2 3 to FeO increases with temperature, 
only the parts of the system Na 2 0-Al 2 3 - 
iron oxide-Si0 2 with low liquidus temper- 
atures can be treated as essentially 
quaternary and in the system Na 2 0- 
Al 2 3 -Fe 2 3 -Si0 2 . Fortunately this low- 


temperature region embraces the compo- — is due to the incongruent melting 

sitions of greatest geological interest, relationship of acmite, the primary phase 

approximately within the volume sodium volume of hematite thus extending 

metasilicate-acmite-nepheline-silica. The through this join. 

position of this volume within the tetra- It seemed most likely, from geometric 

hedron Na20-Al 2 03-Fe20 3 -Si02 can be considerations, that the join nepheline- 

seen from figure 27. The positions of the acmite-Na 2 0»4Si0 2 would intercept the 

five joins first studied can also be seen univariant line nepheline-acmite-albite- 

from this figure; they are Na 2 0«4Si0 2 - liquid, and that in addition this plane 

Al 2 3 -Fe 2 3 (which includes the acmite should contain the ternary reaction point 

and jadeite compositions), nepheline- nepheline-acmite-hematite, which is also 

silica-acmite, nepheline-sodium disilicate- of particular interest in alkaline rock 

acmite, albite-sodium disilicate-acmite, problems. During the past few months 

and nepheline-5Na 2 0*Fe 2 3 -8Si0 2 -ac- thirty-three compositions have been pre- 

mite. Brief progress reports on these joins pared, and work on this join is nearing 

were given in Year Books 49, PP- 46-47, completion; the preliminary equilibrium 

SO, pp. 53-54, and SI, pp. 52-53, noting diagram is given here as figure 28. The 

the existence of various eutectics and piercing points of two univariant lines, 

piercing points without further details, acmite-albite- quartz-liquid and nephe- 

Clearly, however, the system has a line-acmite-albite-liquid, and the ternary 

bearing on problems other than jadeite reaction point nepheline-acmite-hematite 

stability, which was the initial stimulus have been located, at temperatures of 

for its study. One such is its application 758°, 845°, and 905°C, respectively. 

to studies of alkaline rocks, and it was We intend to publish the complete 

during a joint visit to the alkaline diagrams for the six joins and the 

rock/carbonatite complex of the Fen area crystallization flow diagram in the near 

in southern Norway that we decided to future, when we hope that the bearing of 

extend the work on this system. the results on the crystallization history 

From the data on the first five joins in of related alkaline rocks and other rocks 

the system it is possible to deduce a such as peralkaline granites and rhyolites 

crystallization flow diagram, but there can be discussed. Some of the petrological 

were no observational data on the implications are already apparent and 

important univariant line nepheline- will be of interest to geologists working 

acmite-albite-liquid. Its existence and on alkaline rock problems ; for this reason 

extent have to be inferred from other a few of them are stated briefly here, 

data and the geometry of the system. The univariant line along which the 

Such a univariant line should link the three important rock-forming minerals 

quaternary reaction point hematite-neph- nepheline, acmite, and albite crystallize 

eline-acmite-albite-liquid and the quater- in equilibrium with liquid spans a rela- 

nary eutectic point nepheline-acmite- tively large composition and temperature 

albite-sodium disilicate-liquid, between interval with the composition of the 

which there is a considerable composition liquid moving toward a quaternary 

interval and a temperature interval of eutectic where the fourth solid phase, 

200°C (from 915° ± 5°C to below 727°C). sodium disilicate, begins to form. Liquid 

From the point of view of the alkaline compositions along this line range from 

rocks it was important to have an those analogous to ijolite to a foyaitic or 

intersection of this univariant line to nepheline syenitic composition, approx- 

confirm these deductions. That such an imately from liquids containing 40 to 10 

intersection had not been found in any of per cent potential acmite : the tempera- 

the previous five joins — notably the ture interval is from 915° ± 5°C to below 

composition plane nepheline-acmite-silica 727°C. This means that the three min- 



Na 2 0-AI 2 3 -2Si0 2 

" 1526*2° 


Undersaturated 60, 

Oversaturated ' c °'eA 


Na 2 0-4Si0 2 '" 

JO 867*3° 40 50 60 

Weight per cent 




Na20Fe 2 3 4Si0 2 

Fig. 28. Preliminary equilibrium diagram of the join nepheline-acmite-(Na 2 0-4Si0 2 ). This is a 
portion of the join nepheline-hematite-(Na 2 0-4Si0 2 ). 

erals, in various proportions, can exist in 
equilibrium with a wide range of liquid 
compositions over a long temperature 
interval. If similar conditions pertain to 
magmas of analogous composition there 
would be ample scope for differentiation 
and separation of residual liquids by any 
of the normally invoked physical proc- 
esses. Such a condition is entirely in 
keeping with the wide compositional 
ranges seen in alkaline complexes and 
even, in some hand specimens and 
outcrops, by segregation of early-formed 
minerals. Equilibrium between nepheline- 
acmite-albite and a liquid containing 
potential sodium disilicate is consistent 
with the observation that, of the analyses 
in Washington's (1917) tables that give 

sodium metasilicate in the norm, about 
half are of unsaturated alkaline rocks, 
but probably more important is the fact 
that the residual liquid is becoming 
progressively enriched in sodium disilicate 
with increasing crystallization. In nature 
the residual liquid would also become 
increasingly enriched in volatiles, and 
production of such residual fluids would 
explain the almost invariable alkali 
metasomatism of country rocks around 
alkaline intrusions. 

The liquid phase at the ternary reaction 
point nepheline-acmite-hematite- liquid 
has the composition of a simplified ijolite, 
figure 28. This point should be a tempera- 
ture maximum on the corresponding 
univariant line, such that liquids with an 



initial composition on the more siliceous 
side of the composition plane move 
toward the quaternary reaction point 
nepheline-acmite-albite-hematite, i.e. have 
a foyaitic trend, whereas those on the 
opposite side move toward a quaternary 
eutectic at which 5Na20«Fe203*8Si02 
crystallizes with the other three phases. 
This increasingly basic liquid trend, with 
60 per cent or more acmite in some 
compositions, may be compared with the 
ijolite-melteigite trend seen in natural 
rocks. This indication of the possibility 
of an ijolitic liquid's having two possible 
divergent differentiation trends depend- 
ing on small compositional variations is 
interesting, but an even more interesting 
feature is that there appears to be very 
little temperature difference between the 
ternary reaction point nepheline-acmite- 
hematite and the quaternary reaction 
point nepheline-albite-acmite-hematite . 
This means that only slight fluctuations 
in physical conditions or composition will 
determine the differentiation trend of the 
liquid toward either foyaitic residual 
liquids or mafic melteigitic liquids. 

The critical consideration in regard to 
all the above suggestions is whether the 
equilibrium relations at 1 atmosphere are 
seriously changed under pressure and in 
the presence of volatiles, and it is planned 
in the coming year to investigate the 
effects of water pressure on some of the 
critical joins in the system. 

Peralkaline Residual Liquids: Some 
Petrogenetic Considerations 

D. K. Bailey and J. F. Schairer 

It was noted in the above discussion 
of the anhydrous system Na 2 0-Al 2 3 - 
Fe 2 3 -Si0 2 that the residual liquids 
resulting from crystallization of nephe- 
line, albite, and acmite became increas- 
ingly enriched in sodium disilicate. In 
nature this enrichment would be expected 
to be concomitant with enrichment in 
volatiles, and fluids of this type would be 
expected to react with wall rocks con- 
taining A1 2 3 and Si0 2 to form nepheline, 

nepheline-albite, or albite, depending on 
the proportions of A1 2 3 and Si0 2 present. 
For convenience of discussion the simplest 
expressions of the reactions are given in 
the equations 

Na 2 0-2Si0 2 + A1 2 3 -> nepheline (1) 

Na 2 0.2Si0 2 + 2Si0 2 + A1 2 3 -> 

(nepheline + albite) (2) 

Na 2 0-2Si0 2 + 4Si0 2 + A1 2 3 -> 

albite (3) 

In pelitic wall rocks reaction 2 would 
have the most general application, the 
silica and alumina balance corresponding 
to that of the clay minerals. Calcareous 
and ferruginous pelites might be expected 
to push the reaction even further in the 
direction of greater production of nephe- 
line. It is not inconceivable, for instance, 
that nephelinization such as that demon- 
strated at Bancroft, Ontario (Tilley, 
1958), could result from reaction of 
sodium silicate-bearing fluids and impure 
aluminous members of the limestone 
series; such a reaction would require far 
less transfer of material than that needed 
for metasomatic replacement of pure 
limestone. With higher Si0 2 -Al 2 3 ratios 
in the country rocks the reactions would 
trend toward equation 3, giving eventu- 
ally only feldspathization of country 
rocks. The generation of residual fluids 
rich in alkali silicate may thus be seen to 
offer a simple explanation of the meta- 
somatism around alkaline intrusions. 

The parent peralkaline undersaturated 
liquid giving rise to these fluids might 
arise by partial melting of alkali basalt. 
Bowen (1945, p. 88) pointed out that 
residual liquids from fractionation of 
basalt might become enriched in sodium 
silicate by operation of the "plagioclase 
effect," and he was well aware that 
reaction of such liquids with aluminous 
wall rocks would tend to form nepheline 
and albite. (Certainly such liquids would 
account for adinole formation at dolerite- 
shale contacts and might play a part in 
the formation of some spilites.) The 



converse of residual liquid origin — partial 
melting of alkali basalt — could yield the 
same reactive undersaturated liquids and 
perhaps give rise to nepheline-syenite 
complexes with no associated basalt. 

It is less obvious perhaps that over- 
saturated compositions could jdeld resid- 
ual liquids capable of producing meta- 
somatic effects similar to those derived 
from unsaturated liquids. In the system 
Na 2 0-Al 2 3 -Fe 2 Q 3 -Si0 2 , quartz, acmite, 
and albite also crystallize in equilibrium 
with a liquid becoming enriched in sodium 
disilicate, the quaternary eutectic lying 
to the silica-poor side of the join acmite- 
nepheline-Na 2 0-4Si0 2 (fig. 28), which 
means that liquid compositions near this 
point can be expressed in terms of the 
molecules acmite, albite, sodium disili- 
cate, and silica, the amount of silica being 
less than that required for a composition 
Na 2 0*4Si0 2 . If residual fluids of this 
nature were to react with sediments in 
which the Si0 2 /Al 2 3 ratio was 2 or less, 
such as bauxitic, calcareous, or ferrugi- 
nous pelites, conditions intermediate 
between equations 2 and 3 would obtain, 
and it would be possible to have neph- 
elinization of certain rocks around a 
peralkaline granite intrusion. 

Peralkaline residual liquids might re- 
sult from a variety of originally over- 
saturated compositions, for, as Tuttle and 
Bowen (1958, pp. 84-87) have indicated, 
fractionation of liquids on the alkali side 
of the albite-orthoclase-quartz section in 
the system Na 2 0-K 2 0-Al 2 3 -Si0 2 would 
be expected to yield residuals rich in 
alkali silicates and volatiles. They point 
out that such liquids escaping from a 
cooling granite "may affect granitization 
of the adjacent rocks, providing the rocks 
have appropriate composition," but their 
data indicate that the content of alkali 
silicates in such liquids is such that 
reaction with aluminous country rocks 
could yield undersaturated mineral as- 
semblages in the manner described above. 
These considerations only serve to under- 
line a point that perhaps receives too 
little emphasis, namely that the system 

NaAlSi0 4 -KAlSi0 4 -Si0 2 ("petrogeny's 
residua system") is a residua system only 
for compositions that are subaluminous, 
in the sense used by Shand ; compositions 
off this plane, on the alkaline side, would 
be expected to fractionate to liquids rich 
in alkali silicates. It follows that partial 
melting of basement or buried sediments, 
with a bulk composition slightly less than 
subaluminous, should first yield such a 
liquid fraction. Usually this liquid, during 
its uprise, would be expected to react 
with country rocks to produce granite, 
either magmatically or metasomatical- 
ly, or it might produce a more diffuse 
regional metasomatism; but in an alumi- 
nous environment it should also be 
possible for undersaturated assemblages 
to result. 

The System Nepheline-Diopside 
J. F. Schairer, Kenzo Yagi, 1 and H. S. Yoder, Jr. 

In view of the significance of the 
system nepheline-diopside as a principal 
join of the petrologically important 
tetrahedron diopside-nepheline-forsterite- 
quartz, further study was initiated in 
1950-1951 (Year Book SO, p. 54), and it 
has been continued intermittently in 
subsequent years. The system was found 
to be of such complexity that over five 
hundred runs have been made in an effort 
to determine the stability regions of the 
various mineral solid solutions. Some 
forty years ago the system nepheline- 
diopside was investigated by Bowen 
(1922) in connection with the genesis of 
alnoitic rocks of the Province of Quebec, 
Canada. The subsolidus relations were 
not established at the time because of the 
difficulties in the determination of minute 
crystalline phases under the microscope 
without the aid of X-ray techniques. 

As is evident from an inspection of 
figure 29, the system is not binary, and 
should be considered a join in the quinary 
system Na 2 0-CaO-MgO-Al 2 3 -Si0 2 . All 
the crystalline phases obtained are solid 

1 Tohoku University. 







— -o.«^Nej$+OI + Mel + Di„+ L 

\ Di„+OI+Mel+Ne„+L " 

0i„ +Mel +Ne, $ +L 

Ne s , + Mel + Di M 


N020-AI 2 3 -2Si02 














40 50 

Weight per cent 
Fig. 29. Pseudobinary diagram of equilibrium in mixtures of nepheline and diopside 


solutions, and their precise compositions 
have not been determined ; however, they 
may be designated by the principal end 
member present. The phases are diopside 
solid solution (Di ss ), nepheline solid solu- 
tion (Ne ss ), carnegieite solid solution 
(Cg ss ), melilite (Mel), olivine (01), and 
liquid (L) . In some regions of temperature 
and composition a specific solid solution 
may have a fixed composition. This 
conclusion is based on the principle that 
a solid solution of fixed composition will 
appear or disappear during cooling at the 
same temperature from a range of 
compositions. For example, in compo- 
sitions rich in the diopside component, it 
is seen that 01, Mel, and Ne ss appear 
successively at specific temperatures. It 
may be concluded that in the designated 
ranges of bulk composition the solid 
solutions involved here were of fixed, but 

unknown, composition. Such solid solu- 
tions, believed to be of fixed composition, 
are underlined in the figure. 

It was not possible to fix the solidus of 
the system with assurance because of the 
difficulty of recognizing small amounts of 
glass in the quench products or, for some 
compositions, the presence or absence of 
small amounts of olivine. In addition, 
crystal growth was sluggish and equi- 
librium could not be established with 
certainty. All runs having more than 5 
per cent crystals were examined with 
powder X-ray diffraction techniques. 

The new results reaffirm Bowen's 
observations that melilite and olivine 
separate from liquids whose total compo- 
sition can be expressed as a mixture of 
nepheline and diopside. Olivine appears 
to react with liquid until consumed, 
producing diopside solid solution (see also 



Schairer and Yoder, 1960, on the system 
nepheline-diopside-silica) and melilite. 
The intimate association of nepheline, 
clinopyroxene, and melilite in lavas 
suggests that melilite may indeed be a 
differentiation product of an alkali basalt 
magma. Further information is needed on 
the composition of the melilites that 
crystallize in the join nepheline-diopside 
and on the crystallization relations in the 
portion nepheline - diopside - albite-f orster- 
ite of the simplified basalt tetrahedron 
nepheline - diopside - f orsterite - quartz of 
Yoder and Tilley (Year Book 59, p. 67). 

A Reconnaissance of the Systems 
Acmite-Diopside and Acmite-N epheline 

Kenzo Yagi 

The main constituent molecules of the 
pyroxenes in the alkaline rocks are 
diopside, hedenbergite, acmite, and jade- 
ite. Jadeite is present only in minor 
amounts in most of these pyroxenes. 
When the compositions of these pyrox- 

enes are plotted in the triangular diagram 
diopside-hedenbergite-acmite (+ jadeite) 
the points line in a zone extending from 
the diopside corner through the center of 
the triangle to the acmite corner. This 
suggests complete solid solution between 
diopside and acmite and extensive solid 
solution with hedenbergite (see Yagi, 
1953). Nepheline syenites and related 
rocks have nepheline and acmitic pyrox- 
ene as the most important constituent 
minerals in addition to alkali feldspars. 
To investigate the relations in alkaline 
rocks a study of the joins diopside- 
hedenbergite-acmite and nepheline-diop- 
side-acmite is necessary. 

First the join nepheline-diopside was 
studied ; the results are given elsewhere in 
this report (pp. 96-98). The results on 
the join acmite-diopside are given here as 
figure 30. There is a complete series of 
solid solutions between acmite and diop- 
side. Bowen, Schairer, and Willems (1930) 
showed the incongruent nature of the 
melting of acmite at 990°C to hematite 


1 1 1 

111111 ^-^ 

Liquid ^-^^^"^ / 

^-— -"■"*' s 

_— -— s 

_o— " S 


""""^ Pyroxene + Liquid ^ 


^^\^ ^tf 

** - 

Hematite + Liquid ^x^^sq 


- ^r 

^v *»» 

jr Pyroxene + Hematite 


/^ + Liquid ^* 







1 ! 1 

i i i i i i 






- 900 

- 800 

Acmite 10 

Na 2 0Fe 2 3 -4Si0 2 









Fig. 30. 

Weight per cent 
Equilibrium diagram for the join acmite-diopside. 

CaO-MgO-2Si0 2 








990 ±5° 



Carnegieite + Liquid 

Carnegieite + Hematite + Liquid 

■— O o — — — — — 

Carnegieite + Nepheline + Liquid 

Hematite + Liquid 

Nepheline + Hematite + Liquid 

Acmite + Nepheline 








Acmite |0 

Na 2 0Fe 2 3 -4Si0 2 








Weight per cent 
Fig. 31. Equilibrium diagram for the join acmite-nepheline. 

90 Nepheline 
Na 2 0-AI 2 3 -2Si02 

and liquid. In the join acmite-diopside all 
compositions with 40 per cent or more 
acmite melt incongruently, and the join 
is not binary. Attention is called to the 
fact that some of the iron is always 
present as ferrous iron, although most of 
the iron in these melts is ferric iron. 
Therefore, the system is never truly 
binary even in the portion richer in 
diopside than 40 per cent, and there is 
always a small amount of liquid (glass) 
present in the region labeled pyroxenes. 
The results of a study of the join 
acmite-nepheline are given here in figure 
31. A very wide primary field of hematite 
appears on the liquidus surface as a result 
of the incongruent melting of acmite. 
The system is not binary. The phases 
present are acmite, hematite, carnegieite, 
nepheline, and liquid. There is a narrow 
region of coexistence of nepheline and 

carnegieite, suggesting that they are solid 
solutions with a narrow range of compo- 
sitions. Hematite present in the melts 
varies in color from deep reddish brown 
to pale brown, suggesting differences in 
composition perhaps due to the presence 
of A1 2 3 in solid solution in the hematite. 
The precise compositions of solid solu- 
tions have not been determined. Mixtures 
of acmite and nepheline begin to melt at 
about 908°C. 

Studies of the join acmite-diopside- 
nepheline in progress at Tohoku Univer- 
sity are nearly completed. They will be 
presented during the next year. Most of 
the studies on acmite-diopside and 
acmite-nepheline were made at Tohoku 
University, but some of the quenching 
experiments were run at the Geophysical 
Laboratory in December 1960 and Jan- 
uary 1961. 



Accessory Minerals 

Investigations in the System 

D. H. Lindsley 

The system FeO-Fe 2 3 -Ti0 2 contains 
several phases of geologic interest (fig. 
32). In addition to the Ti0 2 minerals 
rutile, anatase, and brookite, the im- 
portant phases are: 

1. The rhombohedral hematite-ilmen- 
ite (aFe 2 3 -FeTi0 3 ) series, with com- 
plete solid solution above ^950°C (Car- 
michael, 1961), referred to as the a series 
by Verhoogen (1962). 

2. The cubic magnetite-ulvospinel 
series, lying on the binary Fe 3 04-Fe 2 Ti04 
join, with complete solid solution above 
~600°C (Vincent, Wright, Chevallier, 
and Mathieu, 1957), called /3 series or /S 
spinels for convenience. 

3. The cation-deficient spinels that lie 
on the Fe 2 3 and Ti0 2 side of the 
Fe 3 4 -Fe 2 Ti0 4 join, called 7 spinels by 
analogy with maghemite (7Fe 2 3 ). 

4. The pseudobrookite series, between 
Fe 2 Ti0 5 (pseudobrookite proper) and 
FeTi 2 6 , an unnamed end member not 
known to occur in nature. Complete solid 
solution in this series is found above 
1150°C (Akimoto, Nagata, and Katsura, 
1957). The term 7 spinel is used only for 
convenience in reference, and does not 
imply existence of the hypothetical 
7FeTi0 3 end member that has been 
postulated for these spinels (e.g., Nicholls, 
1955). Inasmuch as these spinels appa- 
rently form by oxidation of spinels, the 
concept of cation deficiency is more useful 
and more valid than that of 7FeTi0 3 
solid solution. 

Magnetite-ilmenite relations. Magnetite 
grains containing lamellae of ilmenite 

Ti0 2 
(Rutile, Anatase, Brookite) 

FeTi 2 5 



Fe 2 Ti0 5 



Fe 3 4 

Mol per cent 


Fig. 32. Phases in the system FeO-Fe 2 3 -Ti0 2 . Temperatures of complete solid solution (heavy 
lines) in the /3 series, a series, and pseudobrookite series are approximately 600°, 950°, and 1150°C, 
respectively. The join magnetite-rutile (dashed line) is found at low temperatures. 



oriented in the (111) planes of the host 
are found in a variety of rocks and ores. 
The ilmenite lamellae have been widely- 
interpreted as due to exsolution from 
original ilmenite-magnetite solid solu- 
tions. From a study of natural specimens 
Ramdohr (1955) concluded that the 
parental phase could contain up to 50 
mole per cent ilmenite. Several workers 
have noted, however, that ilmenite- 
magnetite intergrowths cannot be homog- 
enized by heating to 1000° to 1200°C if 
bulk composition is maintained; it has 
therefore been suggested that the original 
phase contained ulvospinel rather than 
ilmenite in solid solution and that 
ilmenite is formed by oxidation of 
ulvospinel. Ramdohr (1953) pointed out 
that ulvospinel exsolves in the (100) 
planes of magnetite and that ilmenite 
formed by oxidation of such ulvospinel 
lamellae has oblique extinction — a mode 
of occurrence very different from that of 
most ilmenite-magnetite intergrowths. 
Several workers have also suggested that 
primary magnetite-ulvospinel solid solu- 
tions may be oxidized to y spinels, which 
then break down to ilmenite-magnetite 
intergrowths. The oxidation hypothesis 
has gained added support from recent 
experimental data (Webster and Bright, 
1961; R. Taylor, unpublished Ph.D. 
thesis at Pennsylvania State University) 
and from theoretical considerations (Ver- 
hoogen, 1962), which indicate that the 
stable solubility of ilmenite in magnetite 
even at 1200° to 1300°C is much too small 
to explain observed amounts of ilmenite 
in natural ilmenite-magnetite inter- 
growths. 2 Results of the current investi- 

2 The extensive solubility of ilmenite in 
magnetite reported by Schmahl, Frisch, and 
Hartgartner (1960) at 1000°C is questionable 
because their experimental method could not 
distinguish between mixtures and solid solutions 
of ilmenite and magnetite, and no X-ray or 
optical observations were reported. The limits of 
solid solution shown in their phase diagram are 
based on Ramdohr's estimates from natural 
occurrences and hence cannot be used as inde- 
pendent evidence on the extent of natural solid 

gation support a variant of the oxidation 
hypothesis and show that natural tex- 
tures as well as assemblages can be 
explained by that hypothesis. 

Reagents used were Fisher " certified' ' 
Fe 2 C>3 and Ti0 2 , and United Mineral and 
Chemical Corporation 99.999 per cent Fe 
sponge. Before weighing, the Fe sponge 
was analyzed for 2 , and appropriate 
corrections were made in the proportions 
of Fe and Fe 2 3 . Material was mixed by 
grinding under acetone or toluene to 
inhibit further oxidation. Single-phase 
starting materials for hydrothermal ex- 
periments were synthesized from the 
mixes by heating at 1000°C in evacuated 
silica glass tubes or at 1200°C in Alundum 
crucibles in a controlled atmosphere of 
N 2 + H 2 . Homogeneity was checked by 
optical and X-ray examination. 

Stability relations were determined by 
the hydrothermal buffer technique of 
Eugster, using the buffers wlistite-mag- 
netite ( WM) , fayalite-magnetite-quartz 
(FMQ), nickel-nickel oxide (NNO), and 
magnetite-hematite (MH) to control 
oxygen fugacity (/o 2 ). Oxygen fugacities 
of these buffers as functions of tempera- 
ture and total pressure can be derived 
from the expressions given in table 3. 
Alloying of Fe from the charge with Pt 
containers at low oxygen fugacities was a 
problem in early hydrothermal buffer 
experiments. Wrapping the charge in Ag 
foil or using Ag instead of Pt containers 
prevents Fe loss but introduces additional 
disadvantages. A. Muan (unpublished 
data presented in a Penologists' Club 
lecture) has shown that Ag-rich Ag-Pd 
alloys have melting points higher than 
that of pure Ag, but are still almost 
immiscible with Fe. The high permea- 
bility of Ag-Pd alloys to hydrogen makes 
them ideal as charge containers for 
buffered hydrothermal experiments. Most 
of the data here presented were obtained 
from runs made in Ag 7 oPd 3 o (weight per 
cent) containers. Temperatures were 
measured to ±2°C and regulated to 
d=2°C. At and below 800°C most runs 
were made at 2 kb total pressure; at 



TABLE 3. Calculations of Oxygen Fugacities of Buffers as Functions of Temperature 

and Total Pressure 

T in °K;/o 2 and P to t in bars. From Eugster and Wones, 1962. 

T , A , „ , „ (Ptot - 1). 

^vgjoz — - 


T J-> T ^ 






Wiistite-magnetite (WM) 
Fayalite-magnetite-quartz (FMQ) 
Nickel-nickel oxide (NNO) 
Magnetite-hematite (MH) 





higher temperatures, lower pressures 
(usually 1 kb) were used to protect the 
pressure vessels. All compositions are 
given in mole per cent unless otherwise 

A series of standard /3 spinels ranging 
from Mtioo to Mt 20 Usp 8 o were synthesized 
hydrotherapy at 800°C using the WM 
buffer, each composition being made in 
duplicate or triplicate. Ten measure- 
ments of the (333, 511) peak, using 
internal standards of CaF 2 or Si with Fe 
radiation on a Phillips powder X-ray 
diffractometer, were averaged for each 
sample. Pure ulvospinel, which is not 
stable at the lowest /o 2 attainable with 
available buffers, was synthesized at 
1200°C in a controlled atmosphere of N 2 
and H 2 . Because the (311) peaks of CaF 2 
and Si interfere with the (333, 511) peak 
of ulvospinel, quartz was used as an 
internal standard. Comparison of results 
obtained with samples of intermediate 
compositions using CaF 2 , Si, and quartz 
standards showed no detectable differ- 
ences in 20 values. The 26 (Fe KaJ versus 
composition data were plotted for use as 
a determinative curve. Corresponding 
unit-cell edges are given in figure 33, with 
the data of Akimoto, Katsura, and 
Yoshida (1957) for comparison. Repeated 
measurements of both standards and 
unknown specimens showed a repro- 
ducibility of ±0.01° 26; this internal 
consistency permits determination of 
compositions to at least d=2 mole per cent 
despite any systematic errors that might 

affect the absolute accuracy of the X-ray 
data. Unit-cell edges are believed to be 
accurate to ±0.001 A. 

Data are now available on the compo- 
sitions of titaniferous magnetite in equi- 
librium with ilmenite for the buffers 
NNO, FMQ, and WM. Starting materials 
were either /3 spinels alone or /? spinels + 
ilmenite. The following reactions take 
place during buffered runs : 

Ulvospinel-rich /3 spinel + 2 

= Fe-Ti spinel + ilm S8 (1) 

Magnetite-rich /3 spinel + ilm 

= Fe-Ti spinel + ilm ss + 2 (2) 

(Ilm ss means ilmenite with Fe 2 3 in solid 
solution; for the buffers NNO, FMQ, and 
WM this Fe 2 3 content is less than 15 
per cent. Fe-Ti spinel is used as a general 
term to indicate either /3 or 7 spinels.) 
For each temperature and oxygen fugac- 
ity the unit-cell edges of the Fe-Ti 
spinels formed in reactions 1 and 2 are 
nearly identical and are always inter- 
mediate between those of the initial /3 
spinels, suggesting that the Fe-Ti spinels 
have approached the composition of the 
spinel that is in equilibrium with ilmen- 
ite ss . However, the unit-cell edges alone 
cannot yield unique compositions if the 
product spinels are 7 phases, which they 
must be if there is solid solution of 
ilmenite in the spinel. Akimoto, Katsura, 
and Yoshida (1957) have determined cell 
edges of and 7 spinels; their data 
permit determination of composition 



from the cell edge if the Fe/Ti ratio of 
the spinel is known. The presence of 
ilmenite in a run, however, prevents 
identification of the Fe/Ti ratio of the 
product spinel with the known ratio of 
the entire charge. Chemical analysis of 
the spinel is likewise not feasible when 
ilmenite is present. The amount of 
ilmenite in solid solution with the Fe-Ti 
spinels of reactions 1 and 2 must therefore 
be determined indirectly. 

The equilibrium compositions of the 
products of reactions 1 and 2 are uniquely 
fixed at constant pressure by the tempera- 
ture and oxygen fugacity; the relative 
amounts of the products are determined 
by the bulk composition (or by the Fe/Ti 
ratio in f 0z buffered systems) of the 
starting materials. For each fixed P, T, 

and /o 2 there must exist a bulk compo- 
sition for which the Fe-Ti spinel formed 
will coexist with an infinitesimal amount 
of ilmenite ss . That critical composition 
can be estimated for reaction 1 by making 
a series of runs in which the Fe 2 Ti0 4 
content of the starting /3 spinel is suc- 
cessively lowered until no ilmenite is 
found in the products. The composition 
of the spinel thus formed can be deter- 
mined from the unit-cell edge and by 
chemical analysis. No differences in unit- 
cell edge were detected before and after 
runs at several critical compositions, 
indicating within the accuracy of the 
X-ray data that the product spinel is 
essentially a binary p spinel of the 
starting composition. 

One composition indicated by the 




o< 8.48 


"S 8.46 

5 8.44 



Fe 3 4 

a This Investigation 
• Akimoto ei at, 1957 

Mol per cent 

Fe 2 Ti0 4 

Fig. 33. Plot of composition versus unit-cell edge for magnetite-ulvospinel solid solutions (/3 
spinels). Vertical extent of data points shows uncertainty in cell-edge determinations; horizontal 
extent not significant. Compositions Mtioo to Mt 2 oUsp 8 o made hydrothermally at 800°C; pure 
ulvospinel made at 1200°C in N 2 + H 2 mixture. Data of Akimoto, Katsura, and Yoshida (1957) 
from samples made in evacuated silica glass tubes are shown for comparison. 



X-ray data was checked as follows. At Fe-Ti spinels in equilibrium with ilmenite 

960°C and the /o 2 of the NNO buffer, the can be determined from the unit-cell 

spinel in equilibrium with ilmenite ss is edges. 

Mt 5 2Usp 4 8 by X-ray determination. Runs The available data indicate negligible 

were made on spinels of that composition solid solution of ilmenite in magnetite, 

at 960°C using the NNO, FMQ, and WM That ilmenite-magnetite intergrowths (as 

buffers, and at 800°C using the WM well as ilmenite-magnetite assemblages) 

buffer. Part of each sample was removed can be formed by reaction 1 is shown in 

for X-ray and optical examination, and figure 34, plate 1. Pure ulvospinel was 

the remainder, about 100 mg, was dried 
for 16 hours at 200°C in N 2 for chemical 
analysis. No ilmenite was detected opti- 
cally or by X ray. Ferrous iron was 
determined by the modified Pratt method 
(table 4). The determinations of FeO are 

held at 1000°C at the f 0z of the Stellite 
bomb (roughly equal to that of the 
NNO buffer) for 3 hours. The resulting 
texture — lamellae of ilmenite ss in the 
(111) planes of a Mt 5 oUsp 5 o solid solu- 
tion — closely resembles natural textures. 

TABLE 4. FeO Contents and Unit-Cell Edges of Some Fe-Ti Spinels in Equilibrium 

with Ilmenite s3 




Run No. 

T, °C 



Per Cent 

Per Cent 

























Theoretical FeO for Mt £ 

2Usp 4 8 



minimum values, as any alloying of Fe Reactions 1 and 2, carried out at a 

with the charge container, incomplete series of temperatures for each of several 

drying of the sample before weighing, or buffers, permit us to bracket the compo- 

oxidation of the solution before titration sitions of /3 spinels that are in equilibrium 

would tend to reduce the value for FeO. with a phases. Data obtained using the 

Evidently the spinel in equilibrium with buffers NNO, FMQ, and WM are 

ilmenite ss at 960°C and the fo z of the presented in figure 35; in all runs with 

NNO buffer deviates from a binary /3 these buffers the coexisting a phases are 

spinel by approximately 1 mole per cent ilmenite rich. Exact compositions of the 

FeO, a deviation that is not detected by a phases are now being determined. Runs 

the X-ray method used. The maximum are also being made using the buffers MH 

deviation of the Mt 5 2Usp 4 8 composition and MnO-Mn 3 4 . Figure 35 is simply a 

from the binary join is less than 0.5 mole graphical representation of the compo- 

per cent FeO at 800°C with the WM sitions of /3 spinels that coexist with 

buffer and should be further lowered with 
decreasing temperature. The apparently 
stoichiometric binary compositions indi- 
cated by X-ray data for other spinels in 

ilmenite ss for given temperatures and 
buffers; it is not a phase diagram, 
inasmuch as the compositions of the 
ilmenites are not represented. For a given 

equilibrium with ilmenite will be checked temperature and buffer fo 2 , a spinel of a 

by similar analysis. It seems justified to composition to the right of the appropri- 

conclude that any deviation is small and ate curve will break down by reaction 1 

that the approximate composition of to yield two phases: ilmenite ss plus a 






Reaction 2 Reaction I 

NNO ► < 

FMQ > <3 

WM ► *m 




Fe 3 4 





Mol per cent 

Fe 2 Ti0 4 

Fig. 35. Composition of /3 spinel in equilibrium with ilmenite 88 as a function of temperature and 
the oxygen fugacities of three buffers. NNO, nickel-nickel oxide buffer; FMQ, fayalite-magnetite- 
quartz buffer; WM, wustite-magnetite buffer. Reaction 1, ulvospinel-rich /3 spinel + 2 = Fe-Ti 
spinel + ilmenite 88 . Reaction 2, magnetite-rich /3 spinel + ilmenite = Fe-Ti spinel + ilmenite 88 + 2 . 
Point A is discussed in the text. 

spinel whose composition is indicated by 
the curve. Any spinel to the left of the 
curve is stable by itself but in the presence 
of ilmenite will form by reaction 2 the 
spinel indicated by the curve. For 
example, at 1000°C, the spinel Usp 5 5 
(point A in fig. 35) is unstable at the /o 2 
of the NNO buffer and will oxidize to 
Usp 5 i + ilmenite S8 . The same spinel (A) 
with the FMQ buffer is stable down to 
910°C; at successively lower temperatures 
spinel A is unstable and breaks down to 
spinels successively richer in magnetite, 
plus ilmenite ss . At the lower oxygen 
fugacities of the WM buffer, spinel A 
would remain stable upon cooling until 
the magnetite-ulvospinel solvus is reached 
(^500° to 550°C according to the data of 
Vincent, Wright, Che vallier, and Mathieu, 

It is well to recall that the curves in 

figure 35 are drawn not at constant 
oxygen fugacity but at the fugacity of 
each buffer, which varies with tempera- 
ture (see, for example, Eugster and Wones, 
1962). The data are nevertheless sufficient 
to establish the principle that the 
composition of (3 spinel in equilibrium 
with ilmenite S3 is strongly dependent on 
oxygen fugacity as well as on tempera- 

Application of experimental data to 
natural minerals. Available experimental 
data in the ternary system FeO-Fe 2 3 - 
Ti0 2 indicate that the stable solubility of 
ilmenite (= cation deficiency) in Fe-Ti 
spinels is small at and below magmatic 
temperatures, although metastable cation- 
deficient Fe-Ti spinels are easily made by 
oxidation of /? spinels in air at 400° to 
550°C. It thus seems likely that natural 
cation-deficient Fe-Ti spinels ("titano- 



maghemites") form metastably by oxi- 
dation at moderate temperatures, say 
below 600°C. (However, many natural 
7 spinels contain minor amounts of Mg, 
Mn, Al, V, and Cr, and the experimental 
data from the pure synthetic system 
cannot rule out the possibility that the 
presence of these elements might stabilize 
the cation-deficient structure.) As cation- 
deficient spinels are less dense than the 
equivalent assemblages /? spinel + a 
phase, high pressure should inhibit their 
formation in plutonic rocks. 

The experimental data presented here 
do not disprove the theory that ilmenite- 
magnetite intergrowths form by exsolu- 
tion from a primary ilmenite-magnetite 
solid solution, but they strongly support 
the alternative hypothesis that such 
intergrowths result from the oxidation of 
magnetite-ulvospinel solid solutions. An 
intermediate 7 phase may form in some 
volcanic and hypabyssal rocks, but direct 
oxidation to magnetite-rich /3 spinel + 
ilmenite S8 seems likely in plutonic rocks. 
In rocks where oxygen fugacity remains 
sufficiently low upon cooling, little or no 
ilmenite will form and ulvospinel-magne- 
tite intergrowths may result. 

Relations between ilmenite, hematite, 
magnetite, and rutile. Attempts to deter- 
mine the hematite-ilmenite solvus hydro- 
thermally have been unsuccessful because 
no buffer is available with an oxygen 
fugacity at which hematite ss and ilmen- 
ite ss can coexist. Initial compositions of 
Hem 50 Ilm5o are oxidized to hematite ss + 
pseudobrookite ss (or hematite ss + rutile) 
by the MH buffer, and are reduced to 
magnetite ss + ilmenite ss by the NNO 
buffer. If both hematite ss and ilmenite S8 
can coexist at equilibrium for a given 
temperature, the/o 2 ranges at which each 
is stable must overlap ; the zone of overlap 
must lie between the oxygen fugacities of 
the MH and NNO buffers. As the 
compositions of the coexisting hematite S8 
and ilmenite ss move farther apart upon 
cooling, the corresponding f 0z range of 
mutual stability probably decreases. It 
is possible that at low temperatures (say 

below 200° to 400°C) there is no f 0z at 
which both hematite ss and ilmenite as can 
coexist at equilibrium. Under this hypoth- 
esis hematite-ilmenite intergrowths ex- 
solved at low temperatures are meta- 
stable. They may form because less 
energy is required for migration of Fe 
and Ti in the inherited oxygen framework 
of the original phase than for complete 
reorganization into new phases such as 
magnetite plus rutile. In this regard it is 
significant that many low-grade meta- 
morphic rocks contain the assemblage 
magnetite + rutile, which is chemically 
equivalent to hematite + ilmenite. The 

Fe 3 4 + Ti0 2 = Fe 2 3 + FeTi0 3 (3) 

has a small positive free energy, AG(S) = 
+ 1 to +2 kcal, over the temperature 
range 100° to 1200°C, according to the 
best available data. A(?(3) is probably 
smaller than the uncertainty involved in 
its derivation; if, however, its sign is 
correct it accounts for the assemblage 
magnetite + rutile in low-grade meta- 
morphic rocks. At higher temperatures 
the free energy of mixing of Fe 2 3 
becomes sufficient to favor ilmenite- 
hematite solid solutions over the magnet- 
ite + rutile assemblage. 

Stability Relations of Dravite, 
a Tourmaline 

C. R. Robbins 3 and H. S. Yoder, Jr. 

The most abundant and geochemically 
most important of the boron minerals are 
the tourmalines. They are found in a 
variety of igneous, metamorphic, and 
sedimentary rocks of all ages. Their 
authigenic formation at low temperatures 
in some limestones and sandstones is of 
particular interest. 

Chemically, tourmalines are complex 
borosilicates of variable composition, the 
variation resulting from the large number 
of substitutions permitted by the struc- 
ture. Preliminary calculations suggest 

3 U. S. National Bureau of Standards. 



that tourmalines may be described as 
isomorphous mixtures of several end 

A number of the tourmalines have been 
synthesized from rather complex systems, 
but their stability relations were not 
determined. The objective of the present 
study is the determination of the pressure- 
temperature stability range of a tour- 
maline of specific composition, the iron- 
free end member dravite, NaMg 3 Al 6 B 3 - 
Si 6 2 7(OH) 4 . This composition was chosen 
because the crystal structure studies of 
Hamburger and Buerger (1948) had 
established the ideal formula. It was also 
of interest to relate this composition to 
the petrologically important system 
MgO-Al 2 3 -Si0 2 -H 2 0. 

For this work a glass of the requisite 
anhydrous composition was carefully 
prepared in several steps to avoid loss of 

decomposition products are cordierite, 
liquid, gas, and a crystalline phase that 
has not yet been identified. Sporadic 
occurrences of trace amounts of spinel, 
mullite (?), and, once, of sapphirine have 
been observed microscopically in the 
dissociation products of both the natural 
tourmaline and the glass. These dissoci- 
ation products are fine grained and 
frequently occur as inclusions in the 
cordierite or glass. They may well be the 
result of leaching. 

Above 895°C and 5000 bars both glass 
and natural tourmaline form the assem- 
blage kornerupine + sapphirine + liquid 
+ gas. Previous synthesis of kornerupine 
is unknown to the writers. The phase is 
well crystallized, and its X-ray pattern 
agrees well with that of a natural mineral 
from Kazebanza, Quebec (U. S. N. M. 
no. 106.774). 

TABLE 5. Comparison of Indices of Refraction and Unit-Cell Dimensions of Synthetic 

Dravite and Dobruva, Austria, Tourmaline 

a, A 

c, A 



Synthetic dravite 
Dobruva tourmaline 





Na 2 and B 2 3 . In addition, a natural 
tourmaline from Dobruva, Carinthia, 
Austria (U. S. N. M. no. 103.791), was 
selected for comparative studies. A 
chemical analysis of the material by 
H. B. Wiik showed it to be exceptionally 
close to the ideal dravite composition. 
Indices of refraction and unit-cell dimen- 
sions of synthetic dravite and the 
Dobruva tourmaline are given in table 5. 

Preliminary results of this study at 
various temperatures, and water pres- 
sures up to 5000 bars, are summarized in 
figure 36. The part of the curve below 
500 bars was calculated from the inte- 
grated Clausius-Clapeyron equation. It is 
evident that dravite is stable over a wide 
range of temperatures and pressures. 

At temperatures above 865°C and 
pressures up to 2000 bars the main 

At 925°C and 5000 bars kornerupine is 
no longer stable, and the phases coexisting 
in equilibrium are sapphirine + liquid + 
gas. Inclusions observed in the sapphirine 
appear to be minute spinel octahedra and 
probably result from leaching. 

Although these results are only pre- 
liminary, it would appear that they will 
have application to natural occurrences, 
since the associations tourmaline + cor- 
dierite, tourmaline + kornerupine, and 
cordierite + kornerupine + sapphirine 
(Ussing, 1889) are known. 

Mantle Minerals 

F. R. Boyd, Jr., and J. L. England 

The discovery that pressures as low as 
5 kb cause enstatite to melt congruently 
(Boyd and England, Year Book 60) raises 













i 1 1 1 — i — i — i — I — i — rr 

i — • — r 

i — ■ — r 

-i 1 r 





* i • 




* B 9 » 

--1 tJ L 

400 500 600 700 800 900 1000 1100 1200 1300 1400 

Temperature ,°C 
Fig. 36. Preliminary P-T diagram of the system dravite-water. 

a problem in accounting for the formation 
of basaltic magma that is oversaturated 
in silica. Basalts in the Pacific Ocean 
basin are sometimes considerably over- 
saturated in silica. For example, the 
degree of oversaturation of the primitive 
shield basalts of Hawaii ranges up to 
about 6 weight per cent Si0 2 (Powers, 
1955, p. 81). Nevertheless, there is 
considerable evidence to indicate that the 
mantle rocks from which these basalts 
were derived contain olivine. The prin- 
cipal minerals in rocks in the upper 
mantle are probably enstatite, diopsidic 
pyroxene, olivine, and py rope-rich garnet. 
It is impossible to derive a liquid 
oversaturated in silica by partial fusion 
of a mixture of pyroxene, olivine, and 
garnet in the absence of an incongruent 
melting reaction. For many years it was 
thought that the incongruent melting of 
enstatite found by Bowen and Andersen 
(1914) at atmospheric pressure provided 

a mechanism for generating oversaturated 
liquids from olivine-bearing parent rocks 
at depth. High-pressure data, however, 
indicate that this reaction must be 
restricted to relatively shallow depths in 
the crust. Estimates of the depth of 
formation of basaltic lava in the ocean 
basins range from 50 to 100 km where the 
pressure is in the range 15 to 30 kb. Since 
experimental data have shown that 
enstatite melts congruently in this range, 
an alternative explanation for the compo- 
sition of these oversaturated basalts must 
be sought. 

A possible alternative is that an 
incongruent melting reaction involving 
garnet is effective at depth in the mantle. 
Boyd and England suggested in Year 
Book 60 that pure pyrope must melt 
incongruently and that this reaction 
might also be present in more complex 
natural melts. The melting relations of 
pyrope have been restudied with im- 


proved techniques, and it has been found boundary (A in fig. 37) is essentially the 

that pyrope does melt incongruently over same as that given in our preliminary 

a wide P-T range. As is described here- diagram in Year Book 58 except that no 

after there are at least three incongruent friction correction has been made for the 

melting reactions for pyrope in the present results. Experience has shown 

pressure range 25 to 36 kb. Above 36 kb that the friction in single-stage runs at 

pyrope melts congruently. high temperature is less than was initially 

The principal incongruent melting estimated. The pressure on the run is 

reaction of pyrope is to spinel + liquid, believed to be within d=5 per cent of the 

Since spinel contains no silica, the liquid load pressure. The results shown for 

that forms in the incongruent interval temperatures above 1500°C are new and 

contains more silica than pyrope compo- were obtained by techniques developed 

sition does. In the melting of pure pyrope in a study of the melting curves of albite 

the composition of the liquid lies in the and diopside. Details of these techniques 

three-phase field pyrope + Al-enstatite + have been published recently (Boyd and 

quartz. If this incongruent melting reac- England, 1962). 

tion were present in more complex Liquid of pyrope composition cannot 

natural systems, oversaturated basalts be quenched to a glass over most of the 

could theoretically be generated by investigated P-T range. Recognition of 

fractionation of liquid from a partly the various melting reactions, therefore, 

melted garnet peridotite. By analogy depends on textural differences in the 

with the melting relations for pure pyrope runs. Fortunately, pyrope itself will not 

it would be expected that the incongruent form in the quench. Runs quenched from 

melting would be present over a restricted the fields above curves B, C, and D 

depth range. Pyrope-rich garnet would crystallize in the quench to assemblages 

not be stable in the melting interval at consisting wholly or largely of aluminous 

lower pressures and would melt congru- enstatite. 

ently at higher pressures. For pure pyrope At temperatures below curve E and at 
the depth range would be from about 75 pressures below curve A pyrope compo- 
to 105 km, but it would probably be sition crystallizes to a fine-grained mix- 
shallower in a natural system. ture of aluminous enstatite, sapphirine, 

The phase relations determined in the and sillimanite. Enstatite and sapphirine 

pyrope study are a further demonstration can be distinguished in X-ray diffrac- 

of the importance of high pressure in tometer patterns of such runs, and they 

modifying crystal-liquid equilibria in can be recognized under the microscope, 

silicate systems. In the pressure range 20 Aluminous enstatite forms about 80 per 

to 40 kb a liquid of pyrope composition cent of the products. The presence of 

can be crystallized by at least five sillimanite is known from the study of 

different paths, depending on the pres- other compositions in the system MgO- 

sure. It is not certain that the incongruent Al 2 3 -Si0 2 in this P-T range, but the 

melting of pyrope is a significant feature amount that forms in a run on pyrope 

of magma differentiation at depth in the composition is too small to detect by 

mantle, but there is little question that optical or X-ray methods, 

pressure will prove to have a major At 1615°C and 24.5 kb there is a 

influence on such differentiation. pronounced break in the boundary of the 

pyrope stability field. This break is 

Effect of Pressure on the Melting of Pyrope bel j wed to be due to the intersection of a 

melting curve in the breakdown products 

Phase relations for pyrope composition field with the pyrope subsolidus boundary. 

in the pressure range 15 to 50 kb are Most likely, though not certainly, the 

shown in figure 37. The subsolidus melting curve (E in fig. 37) is the solidus 



curve for the breakdown products. It is 
known that aluminous enstatite is a 
stable phase in the P-T field between 
curves E and F, but it is not known 
whether E marks the disappearance of 
sapphirine or sillimanite. Runs quenched 
from pressures and temperatures imme- 
diately above or below curve E have 
essentially identical X-ray patterns and 

are indistinguishable under the micro- 

Curve F in figure 37 is the liquidus 
curve for aluminous enstatite. In runs 
quenched from below curve F the ensta- 
tite forms a fine-grained mosaic of 
crystals. In runs quenched from above 
curve F the enstatite crystallizes as 
coarse blades with undulate extinction 

i i i i i i i i i i i i i i i i i i i 

| i i i i i i i i i | 

i i i i i i i i 

i i i i i i i 













L,QUID JU - x / 
/ L---i^rK x D 








+ i : 


Mg 3 Al2Si 3 0| 2 


i i i 

J— l—L 

\ i i t l 

I I I 

i i t I I i I I I I I I I t I 

I I I I 


20 30 

Load pressurewkilobars 



Fig. 37. The stability field of pyrope garnet. Synthetic pyrope was used as starting material for 
all runs shown except for the two runs indicated by open circles on curve A. In these two runs pyrope 
formed from seeded, crystalline breakdown products. The dashed curve, H, is a nucleation boundary; 
pyrope forms readily from glass or crystalline starting materials at pressures higher than curve H f 
but it will not nucleate in the P-T field between curves A and H. For a further discussion of the 
nucleation problem, see Boyd and England, Year Book 58. 


under crossed nicols. This texture is should react to form some other phase, 

characteristic of enstatite crystals that A mixture of 20 per cent crystalline 

have formed in the quench in high- sapphirine + 80 per cent pyrope glass 

pressure runs (Boyd and England, Year and a mixture of 24 per cent crystalline 

Book 60). spinel + 76 per cent crystalline pyrope 

In the P-T field bounded by curves F, were run at temperatures midway be- 

C, and G the quench crystals of enstatite tween curves F and G at the pressures 

contain scattered, subhedral grains of an 21.5 kb and 28.7 kb. The products of 

isotropic phase with a refractive index these runs looked identical with those of 

appreciably higher than that of the runs made in this P-T field on pyrope 

enstatite in which they are poikilitically composition except that the concentra- 

enclosed. These isotropic grains have tion of the primary, isotropic phase 

rounded to rectangular shapes when seen poikilitically enclosed in enstatite quench 

under the microscope and are 1 to 5 crystals was greatly increased. X-ray 

microns in diameter. The texture of the patterns of these runs showed that the 

runs quenched from this field is strikingly products obtained with both mixes were 

similar to that of the granules of primary spinel + enstatite. Since sapphirine was 

forsterite embedded in quench enstatite converted to spinel in these experiments, 

found in runs on MgSi0 3 composition the primary phase in the field bounded by 

cooled from a temperature within the curves F, C, and G is proved to be spinel, 

incongruent melting interval at atmos- The melting relations shown in figure 

pheric pressure. The isotropic granules 37 only partly define the melting of 

decrease in abundance as the temperature pyrope composition in the pressure range 

is raised in the interval between curves F 15 to 29 kb. There must be at least one 

and G. Runs quenched from above curve more curve than is shown in the P-T 

G contain only glass and/or quench range below curve F. Quenching diffi- 

crystals of enstatite metastably rich in culties and the small amounts of phases 

A1 2 3 . other than aluminous enstatite that are 

The quantity of the isotropic primary present on pyrope composition prevented 

phase in the field bounded by curves F, the identification and location of this 

C, and G is insufficient to show on an curve. Study of the melting of a variety 

X-ray diffractometer pattern. Optical of compositions in MgO-Al 2 3 -Si0 2 in 

properties obtained indicate that it must this P-T range would undoubtedly clarify 

be either spinel or sapphirine. The the picture, but the tendency of these 

refractive indices of spinel and sapphirine compositions to crystallize in the quench 

are similar. Sapphirine has a low bire- remains a formidable problem, 

fringence, but it appears isotropic in The melting curves of the breakdown 

grains only a few microns in diameter. A products (E, F, and G) intersect the 

test was devised, however, that showed pyrope stability field and give it a faceted 

the primary phase to be spinel. boundary. The principal incongruent 

At constant temperature and pressure, melting reaction is to spinel + liquid, but 

changing the proportions of phases in at least two other reactions in which 

equilibrium will not change the kinds of pyrope melts to Al-enstatite + liquid + 

phases present or their compositions, other crystalline phases must be present 

Hence, if the primary isotropic phase was in the pressure range 25 to 29 kb. Above 

spinel, it would be possible to add spinel 36 kb the melting is congruent. The 

to the mixture of spinel and liquid on maximum incongruent melting interval 

pyrope composition without changing the at constant pressure is about 90° at 25 kb 

phase relations. If the primary phase was and diminishes as the pressure is raised 

spinel and sapphirine was added, the until the melting becomes congruent, 

sapphirine would not be stable and The pressure interval over which the 


melting is incongruent is 11 kb, corre- to 5. 5°/kb. The only other silicate melting 

sponding to a depth interval in the upper curve thus far determined in the pressure 

mantle of about 30 km. range above 30 kb is diopside. The slope 

The average slope of the pyrope solidus of the diopside curve in the range 35 to 50 

curves (B and C) in the incongruent kb is 6.9°/kb. These slopes are substan- 

melting interval between 25 and 36 kb is tially less than was earlier estimated for 

about 16.5°/kb. Above 36 kb, where the most silicates on the basis of data 

melting is congruent, the slope decreases obtained in a lower pressure range. 


o • 7 • D 7 . . lected by H. S. Washington, was kindly 

banidine Jrhenocrysts in borne , , J X , ,, & TT ~ -* T ^ i 

Peralkaline Volcanic Rocks r Jl leaSed *?. us £ ^ U " S : Natlonal 

Museum. I he other three specimens were 

F.Chayes and E.G.Zieswith X-ray data by collected by Chayes, in company with 

Professor S. Vardbasso and Dr. A. Atzeni, 

The petrologist often uses bulk chem- of the University of Cagiiari. 

ical analysis as in some sense a substitute The Paris de Besa trachyte flow, 

for modal analysis, and recent improve- situated about 5 km west of the town of 

ments in modal analysis have prompted Ales and described briefly by Atzeni 

a revival of RosiwaFs countersuggestion (1959), is exposed by a small window 

that chemical composition be inferred through the post-Miocene basalts on the 

from modes. There are circumstances in southeastern flank of Monte Arci. It is a 

which the first procedure is unavoidable, fine-grained blue-gray rock studded with 

and there are also circumstances in which numerous blocky phenocrysts of glass- 

the second seems very convenient. We clear sanidine. As Atzeni remarks, these 

hope the work reported here, part of a sometimes contain cores of oligoclase. 

long-range and rather general study of The transition from oligoclase core to 

rhyolites and trachytes, persuades the sanidine mantle may be either blurred 

reader that much may also be gained by and gradual or sharp and abrupt. In the 

using analytical chemistry and petrog- former case the crystals usually show 

raphy as supplements to rather than highly undulant extinction, and the 

substitutes for each other. "core" is likely to have a jagged outline 

In the current report year we have marked by many reentrants ; in the latter 

completed examination of four peralka- there is no suggestion of strain or replace- 

line specimens and the feldspar concen- ment, and the sharply euhedral core 

trates prepared from them. Siliceous usually shows polysynthetic twinning, 

lavas of this type provide an excellent — There are also occasional phenocrysts of 

perhaps the best — opportunity to study acmitic diopside, sublenticular clots of 

the relation between crystal composition tridymite, and irregular inclusions of 

and bulk composition in a natural other, possibly cognate, volcanic rocks, 

"system" closely resembling the experi- As in most Sardinian volcanics so far 

mentalists' version of "petrogeny's re- collected in this project, the feldspar 

sidua." The specimens include a trachyte phenocrysts show little indication of 

from Paris de Besa, Sardinia, a por- alteration, although joint surfaces 

phyritic pantellerite from Pantelleria, and throughout the rock are usually stained 

two comendites from the type localities yellowish brown and in thin section 

Le Commende and Le Fontane, Isola San similar staining sometimes occurs in 

Pietro, Sardinia. The pantellerite, col- phenocrysts. Carlsbad twins are common, 


but in our specimens no other variety of trifling amount. Despite much effort we 

twinning has been observed in the were unable to isolate enough of either 

sanidine, which is also free of micro- and cossyrite or acmite for analysis. The mode 

cryptoperthitic intergrowth. Its optic of Washington's porphyritic pantellerite 

angle is variable but always very small, from Gelkhamar (U. S. N. M. no. PRC 

(Unless our material is entirely atypical, 2000) is shown in column 2 of table 6. 
Atzeni's identification of the alkali feld- The comendites of Isola San Pietro 

spar of this rock as microcline is erro- were discovered by Bertolio and described 

neous.) The mode of our specimen (no. in considerable detail by Johnsen (1912); 

25B10) is shown in column 1 of table 6. since Johnsen's work no new information 

TABLE 6. Modes of the Analyzed Specimens 



PRC 2000 













Tridymite and others 








25B10, trachyte, Paris de Besa, Sardinia. 

PRC 2000, pantellerite, Gelkhamar, Pantelleria. 

39B2, comendite, "Commende type," Capo Sandolo, Isola San Pietro, Sardinia. 

40B5, comendite, "Fontane type," Le Fontane, Isola San Pietro, Sardinia. 

The porphyritic pantellerite is H. S. on these interesting rocks appears to have 
Washington's specimen from Gelkhamar, been published. Our specimen 39B2 is 
Pantelleria, described and analyzed by from a roadside exposure about 500 
him (Washington, 1914). A molar excess meters east of the lighthouse at Capo 
of alkalies over R2O3, signified in the Sandalo and about 1 km west southwest 
CIPW system by the appearance of Ns of Le Commende, from which the rock 
(sodium metasilicate) in the norm, has type takes its name. It is a completely 
long attracted attention to the Pantel- devitrified blue-gray glass containing 
lerian lavas. This specimen was one of numerous phenocrysts of bipyramidal 
two recently reanalyzed ; for a comparison quartz and blocky, water-clear sanidine, 
of the new and old analyses see Zies the sanidine usually showing a pro- 
(1960). The feldspar analysis, made at nounced schiller. The matrix consists of 
the same time as the bulk analysis, spherulitic masses of extremely fine- 
appears here for the first time. The grained quartz, feldspar, and an acicular 
principal phenocryst of this specimen is green mineral, which may be either acmite 
sanidine, called soda microcline by Wash- or arfvedsonite. Specimen 40B5 is from 
ington. In a careful examination of two the quarry at Le Fontane, about 500 
thin sections and of many granular miles southwest of the town of Carloforte. 
products obtained at various stages of the It is a bluish gray glass, closely matching 
sample preparation no second feldspar the description of Johnsen's "Fontane 
was observed. Phenocrysts of bipyram- type" comendite. Both quartz and feld- 
idal quartz are abundant; phenocrysts spar phenocrysts are similar in appear- 
other than quartz and sanidine, chiefly ance to those already described from Le 
cossyrite and acmite, are present only in Commende, though much less abundant 



than in our particular specimens from 
that locality. Prominent in 40B5 are 
stringers of a dense black glass distributed 
through the rock in conspicuously laminar 
fashion. The groundmass is glass showing 
little evidence of devitrification. The 
modes of specimens 39B2 and 40B5 are 
shown in columns 3 and 4 of table 6. 

Bulk analyses and norms. Analyses and 
CIPW norms of the four specimens, as 
well as of the material forming one of the 
dark stringers in 40B5, are shown in 
table 7. It will be noted that although all 
the rocks are peralkaline Ns appears in 
only one of the four norms. This is, of 
course, the pantellerite from Gelkhamar, 
Pantelleria; the large amount of norma- 
tive Ns shown in the original analysis of 
this specimen was the principal occasion 
for its reanalysis. Ns is recorded in seven 
of the ten available peralkaline norms of 
the Pantellerian lavas (Washington, 
1914), the first and still the most extreme 

example of molar excess of alkalies over 
ferric oxide and alumina. 

Ns is not present in either of our 
comendite norms, occurs in only one of 
the norms of the seven comendite 
analyses given by Johnsen (1912), and is 
evidently both uncommon and quanti- 
tatively insignificant in the type locality 
of comendite. This contrast between 
comendite and pantellerite is perhaps 
particularly striking because in the petro- 
graphic literature the names are often 
used interchangeably. The differences 
between our two comendites and the 
Gelkhamar pantellerite are about what 
would be anticipated from inspection of 
earlier analyses. In connection with these 
particular rocks, however, such an inspec- 
tion raises more problems than it solves. 
Even if we agree to ignore sampling 
difficulties, which are unusually acute, 
and the total amount of information, 
which is, as usual, rather small, Johnsen 

TABLE 7. Bulk Analyses and Norms 
(Specimens as identified in table 6; 40B5 inc. is fragment of a stringer in 40B5.) 














Si0 2 











A1 2 3 











Fe 2 3 















































Na 2 








K 2 










H 2 0+ 











H 2 0" 










Ti0 2 









Zr0 2 








P 2 5 











so 3 









C1-H 2 sol. 








C1-H 2 msol. 











for CI 








Sought but not found. 



points out that his analyses of comendite 
differ markedly from the earlier, incom- 
plete analyses of Bertolio, Washington 
points out that his analyses of pantel- 
lerite differ markedly from the earlier, 
incomplete analyses of Forstner, and 
there is now reason to suspect that 
Washington's Ti0 2 estimates were sys- 
tematically high (Zies, 1960, p. 306), an 
analytical bias that would necessarily 
generate overestimates both of the 
amount of Ns and of the frequency of its 
occurrence in any set of norms. Like so 
many of the problems of modern descrip- 
tive petrography, satisfactory comparison 
of these two rock types will require a 
manifold increase in the number of rock 
analyses with no sacrifice, and preferably 
with some improvement, in their quality. 
Feldspar phenocrysts of the analyzed 
rocks. In all four specimens the principal 
phenocryst is sanidine, characterized by 
very small optic angle, apparent mono- 
clinic symmetry of X-ray powder spectra, 
absence of multiple twins, and lack of 
cryptoperthic structure or anorthoclase 
type grill. The geological occurrence is of 
course the classic one for sanidine; the 
blocky, sharply euhedral habit and the 
glassy, often transparent character of 
the crystals are equally appropriate. 
Despite careful search, no other feldspar 
has been identified in three of the 
specimens; in the Paris de Besa trachyte, 
as already noted, sanidine crystals some- 
times contain cores of oligoclase. The 
X-ray spectra of the Paris de Besa and 
Gelkhamar feldspars show very little 
submicroscopic unmixing, whereas those 
from the Capo Sandalo and Le Fontane 
comendites seem to be almost entirely 
unmixed, the powder patterns being 
interpretable as mixtures of nearly pure 
Or and Ab. This unmixing can be 
detected only by X ray. It is therefore 
rather startling to discover that in the 
very year in which Laue first predicted 
that crystals ought to diffract X rays 
Johnsen (1912, p. 6) unhesitatingly 
attributed the schiller of the comendite 
sanidine to incipient unmixing, which, 

carried to completion, would finally 
transform an "unstable monoclinic" crys- 
tal into a "stable triclinic" one. The line 
between science and prescience is some- 
times very hard to draw ! 

Analyses and norms of the best feldspar 
concentrates that could be obtained from 
each specimen by magnetic and heavy- 
liquid separation are shown in table 8. 


Analyses and Norms of Four 


AJkali Feldspars 


(Specimens as identified in table 6.) 


PRC 2000 



Si0 2 





A1 2 3 





Fe 2 3 














_ : 





Na 2 





K 2 





Ti0 2 






H 2 Of 



















































* Sought but not found. 

f Samples dried at 110°C before analysis. 

Results of fragment counts made on 
three of the analyzed specimens are 
recorded in table 9; in the feldspar from 
40B5 no quartz or plagioclase was found, 
and the amount of groundmass adhering 
to sanidine grains was too small to 
estimate by this technique. In the counts, 
incidentally, the precision of ratios of 



TABLE 9. Fragment Counts of Analyzed 

Feldspar Concentrates 

(Specimens as identified in table 6.) 


PRC 2000 


Alkali feldspar 











Count length 




minerals to each other is about that 
appropriate to the count length, but, 
since i 'groundmass" occurs almost entire- 
ly as thin discontinuous margins about 
sanidine in all four concentrates, estima- 
tion of the amount of groundmass from 
the number of grains in which it is 
observed requires an "adjustment" or 
"correction" factor that is not much 
better than a shrewd guess. The values 
given are probably overestimates. 

Although FeO appears in barely more 
than trace amounts in the analyses of 
table 8, Fe 2 3 is present in quantities far 
greater than can reasonably be attributed 
to visible impurities or analytical error. 
In all four feldspars a strong buff tint can 
be produced by heating for 10 minutes 
or less at ^850°C in air, and discharged 
by heating over a Meker blast for a few 
minutes at ^4100°C. In our view most 
of the Fe 2 3 in these analyses must be 
regarded as part of the feldspar, probably 

proxying for A1 2 3 , as already suggested 
by Johnsen (1912, p. 6). 

It will be noted that quartz and 
hematite are present in all four and a 
little corundum in two of the feldspar 
norms. In three of the specimens the 
content of normative quartz is roughly 
comparable to the modal amounts shown 
in table 9. The agreement is far from 
exact, but the presence of measurable 
amounts of modal quartz sharply limits 
the use of these analyses as a basis for 
speculation about the nature of the 
silica-cation balance in alkali feldspars. 
We may point out, however, that despite 
diligent search no quartz at all was noted 
in feldspar 40B5, yet the norm shows 2.53 
per cent. Most of the norms of literature 
analyses of alkali feldspar so far examined 
show normative quartz in comparable or 
greater amounts, amounts large enough 
so that they could scarcely be overlooked 
by the petrographer or fabricated by the 
chemist. Many also show appreciable 
amounts of normative hematite and 
corundum. Despite serious analytical and 
sampling uncertainties in available data 
it seems to us that the possibility of 
systematic departure from the assumed 
1:1:6 ratio of RO:R 2 3 :Si0 2 in alkali 
feldspar deserves more than casual con- 

Projection of results into "petrogeny's 
residua system." Ternary coordinates of 
the four rock-feldspar phenocryst pairs 

TABLE 10. Rock and Feldspar Compositions Projected into the Ternary System Q-Or-Ab 
(Numbered specimens as identified in table 6; Commende and Fontane from Johnsen, 1912.) 


PRC 2000 39B2 










38.0 35.8 
36.0 34.4 
26.0 29.8 





Or (analytical) 
Or (X-ray) 


37.4 39.9 
35 37 






Fig. 38. Data of table 10 plotted in Q-Or-Ab diagram. JC, rock anal. p. 11, no. 2, feldspar anal. 
p. 5, no. 1; JF, rock anal. p. 22, no. 4, feldspar anal. p. 19, no. 1, in Johnsen (1912). Other specimens 
as identified in table 6. 

described in this note together with two 
evidently similar pairs taken from John- 
sen (1912) are listed numerically in table 
10 and shown graphically in figure 38. 
The last line of the table gives compo- 
sitions of alkali feldspar determined in 
the way described by Tuttle and Bowen 
(1958, pp. 11-13) on specimens homoge- 
nized at 850°C and room pressure for 24 
hours. As far as could be determined from 
the X-ray powder spectra the materials, 
which were initially monoclinic, were 
completely homogenized by this treat- 
ment. We could find no significant 
differences between Or content estimated 
on specimens treated in this way and that 
obtained from specimens heated hydro- 
thermally for extended periods of time. 
It will be noted that all four estimates of 
Or by X ray are lower than the relevant 

Or/ (Or + Ab) ratios calculated from the 
analytical data. We suspect that the high 
Fe 2 3 and excess Si0 2 already noted may 
be responsible for this discrepancy, since 
the determinative curves are developed 
for synthetic materials as free of Fe and 
as close to the 1:1:6 ratio as possible. 

In figure 38 a line connects each 
projected bulk composition with its 
appropriate feldspar, our data being 
shown with solid dots and lines, the two 
pairs taken from Johnsen by open circles 
and dashed lines. Our Fontane specimen 
obviously checks Johnsen's very closely, 
and the two Commende pairs are also in 
good agreement. The differences between 
the slopes of these lines may appear small 
in ternary projection but are in fact very 
large. In compositions distributed along 
the line connecting the Pantellerian 


feldspar and its host rock the slope of the of rock analyses, as portrayed graphically 

regression of Or as a function of Ab or Q in the Harker variation diagram. 4 Be- 

would be zero and the slope of the regres- cause of algebraic restrictions arising from 

sionofAb as a function of Q would be — 1. the cloture property, co variances and 

In compositions distributed along the line correlations are not independent of 

connecting our Fontane feldspar with its variances, as is normally assumed in 

host rock, on the other hand, the slope of statistical (or other) testing. The larger 

the regression of Or as a function of Ab the variance of a particular variable, the 

would be +1 and the slope of the more strongly negative are its expected 

regression of either Ab or Or as a function correlations with other variables, and as 

of Q would be — %. long as no variance is greater than the 

The relations between phenocryst and sum of the other variances all expected 

host implied by these two pairs are thus correlations are negative. The particular 

very different. In the first the normative effect of variance or co variance of most 

groundmass feldspar must be much importance for an appreciation of the 

richer in Or than the phenocryst feldspar ; Harker diagram may be illustrated by a 

in the second the two must be nearly very simple model. 

identical in composition. From data Let us suppose an M-variable closed 

already given we may estimate that the array in which the parent variances, 07 2 , 

normative Or content of groundmass of variables X, are equal for 2 $ j ^ M, 

feldspar in the Gelkhamar pantellerite is but in which ci 2 9* aj 2 . Further, we 

61.2 compared with 37.4 in the pheno- specify that, although ci 2 is potentially 

crysts. In the Capo Sandolo comendite, variable, the system remains stable 

on the other hand, the Or content of the during any particular sampling or set of 

normative groundmass feldspar is 47.7 as samplings. Under these conditions the 

compared with 39.9 in the phenocrysts, expected correlation between X\ and X 3 

and in the Fontane comendite the in such a sampling is 

comparable values are 50.7 and 48.4, this , 

last pair probably differing by less than pl > " °" l/o "^ 1 ~ M) W 

the total experimental error. The possible an d that between any pair of variables 

importance of this distinction between no t including Xi is 

pantellerite and comendite is obvious; P 

fractionation of the Pantellerian type _ ± ± <Tl (2) 

would involve extensive end -stage enrich- 2 — M L (M — 1 ) o"y 2 - 

ment in potassium, whereas fractionation Nqw fa b definition nonne gative, and 

of the comenditic type would proceed ft can be ghown ( for instance> chayeS; 

with no notable shift in the Na/K ratio lQm) ^ < {M _ l} We are thug 

It is equally obvious that a discussion of concemed with variations in Plj and Pjk 

this problem based on data from three or ^ arige ag ft consequence of 

four specimens is scarce y more than idle itti to in the range 

speculation. Once more the desirability 01 ^ <- <(M—Vi 

a drastic increase in analytical potential ^substitution"^ these values in 

becomes apparent. equations 1 and 2 we have 

Variance Relations in Some Published ^ piy > — 1 1 ,^ 

Harker Diagrams —1/(M — 2) ^ p jk < +1) 

F. Chayes ^ s Gl m0 ves from its lower to its upper 

Much of the interest in petrographic limit > the negative correlation to be 

closed arrays centers on relations between 4 The diagram known by Harker's name seems 

silica and other essential oxides in suites to have been invented by Iddings (1892). 



expected between X\ and Xj becomes 
progressively stronger while the initially- 
strong negative correlation between Xj 
and X k becomes progressively weaker; if 
a i 2 > (M — l)aj 2 the expected value of 
the latter correlation is positive. For the 
present we specify no mechanism by 
which to manipulate the variances, 
arguing only that, if variance relations of 
the specified sort did in fact occur, the 
correlations to be expected in the absence 
of any other relation between the vari- 
ables would be dictated by them. 

That the principal negative correlations 
of the Harker diagram seemed intimately 
related to the variances in the fashion 
suggested had already been noted 
(Chayes, 1962), but in fact this inference 
was largely based on examinations of 
graphs. During the report year calcula- 
tions were carried through on twenty-five 
suites of analyses of volcanic rocks that 
had served as the basis of published 
Harker diagrams. Satisfactory description 
of the results calls for bibliographical and 
other detail not appropriate in a report 
of this sort. Certain of the findings are so 
extreme, however, that further work is 
hardly likely to lead to substantial 

In every suite, for instance, the vari- 
ance of silica was considerably larger than 
that of any other oxide. In twenty-three 
of the twenty-five suites the variance of 
silica was larger than the sum of all other 
variances, both exceptions being suites of 
oceanic lavas, assemblages to which the 
Harker diagram is rarely applied. The 
ratio of silica variance to the sum of other 
variances is so far never larger than 3.38; 
its average value, 1.95, taken as an 
estimate of the ratio <ri 2 /(rj 2 (M — 1) 
would lead to pu ~ —0.63 in equation 1, 
if we count the variables in the way pro- 
posed in last year's report. Although this 
is hardly more than a very crude approx- 
imation, the variance relations are clearly 
such as to require very strong negative 
correlation between silica and all other 
oxides that contribute materially to the 
total variance of a Harker array. These 

oxides are, in order of increasing average 
variance, Fe 2 3 , A1 2 3 , FeO, MgO, CaO; 
curiously, and perhaps significantly, this 
is also in order of increasingly strong 
negative correlation with Si0 2 . Since 
Ti0 2 , Na 2 0, and K 2 do not contribute 
materially to the total variance, the effect 
of cloture on their correlations with Si0 2 
should be negligible. Although Ti0 2 is 
usually negatively correlated with Si0 2 , 
it is well known that the correlations of 
Na 2 and K 2 with Si0 2 are nearly 
always strongly positive. Systematically 
strong positive correlations involving 
Si0 2 thus emerge only where the variance 
relations permit. 

A considerable excess of silica variance 
over the sum of other variances seems to 
be characteristic of the continental basalt- 
andesite-dacite-rhyolite association ; if the 
basic and acid parts of such suites are 
considered separately, the excess variance 
of silica usually persists in the rhyolite or 
dacite-rhyolite and, to date, always 
persists in the basalt-andesite parts of the 
assemblages. An apparently similar ex- 
cess is not uncommon in suites of the 
oceanic basalt- trachyte association but 
seems to be generated here by the 
grouping of analyses that do not belong 
together, as suggested in a later section 
of this report. At any rate, in the basaltic 
portions of oceanic basalt-trachyte suites 
silica variance never exceeds the sum of 
the other variances and, indeed, is rarely 
the largest variance. (Unfortunately, 
there are very few oceanic suites in which 
nonbasalts are sufficiently numerous to 
make separate computation worth while.) 
A full account of variance-covariance 
computations in volcanic suites is now in 

The Treatment of FeO and Fe 2 Os in 
Harker Diagrams 

F. Chayes 

In most Harker diagrams only one Fe 
oxide is shown; customarily, a new 
variable is formed by adding the adjusted 
weight of one of the oxides to the posted 



amount of the other in each analysis, viz., 


(Fe 2 3 )r = Fe 2 3 + l.llFeO 
(FeO) r = FeO + 0.901Fe 2 O 3 


Al though modern justifications of it are 
rarely explicit, the practice itself was 
proposed by Iddings (1892) in the first 
publication containing what we would 
now call Harker diagrams. In this 
pioneering discussion Iddings often shows 
total Fe as FeO, but always shows FeO 
and Fe 2 3 separately as well. (In the first 
text treatment of the subject, however, 
Harker [1909] shows the iron oxides 
separately in only six of the twelve 
diagrams he presents.) 

From close examination of the "very 
carefully executed analyses of the rocks 
from the region of the Yellowstone 
Park," Iddings argues (1892, p. 153) that 
"In this group of rocks the reciprocal 
behavior of the ferrous and ferric oxides 
is one of the most marked chemical 
features" and concludes "it seems highly 
probable that during the differentiation 
of the magma all of the iron existed in the 
ferrous condition . . . and that subse- 
quently it was in part more highly 
oxidized, so that the more ferric oxide 
was produced the less ferrous remained." 
Before hastily concluding that this dictum 
implies negative correlation between FeO 
and Fe 2 3 , the reader is urged to examine 
figure 39, a display of the data upon 
which it is based. The correlation between 
FeO and Fe 2 3 is actually positive, 
though very weak. Denoting Si0 2 by x, 
Fe 2 3 by y, and FeO by z, the Iddings' 

* 4 



u. 2 


• • • 

• m 

2 3 4 
FeO. wt.% 

Fig. 39. FeO and Fe 2 3 in volcanic rocks of 
Yellowstone Park (data from table 1 of Iddings, 

data give r yz = +0.246. The partial 
correlation is very different, viz., r yz . x = 
— 0.690. For a fixed silica content there is 
indeed some tendency for FeO and Fe 2 3 
to vary inversely in the sample, but the 
tendency hardly seems strong enough to 
warrant either Iddings' detailed specu- 
lations about a pooled Fe variable or the 
tacit conviction of modern petrographers 
that no other Fe variable is desirable in 
Harker diagrams. 

There is, nevertheless, strong negative 
correlation between both iron oxides and 
silica in this earliest "Harker array," as 
in so many of its successors. In fact, for 
the Iddings data r xy = 0.643, r xz = 
-0.834, and r x(y+z) = -0.925. This last 
correlation is, I believe, the common 
though rarely stated occasion for pooling 
the Fe oxides into a single variable. We 

Fig. 40. Histogram of product moment correlations between FeO and Fe 2 03 in the raw data of 
twenty-five published Harker diagrams. 


all like points that lie fairly close to authenticity is beyond any reasonable 

fairly simple curves, and in most Harker doubt. (For a review of the oceanic lava 

arrays the linear correlation between associations, see, for instance, Tilley 

silica and some form of the sum of the [1950], or Turner and Verhoogen [1951, 

iron oxides will be stronger than that pp. 124-155].) 

between silica and either of the iron We are equipped with a full comple- 

oxides separately. The explanation offered ment of names for lavas intermediate in 

by Iddings is only one of a large class of composition between basalt and trachyte 

hypotheses compatible with this relation- — trachybasalt, trachydolerite, trachy- 

ship ; not all members of this class require andesite, kohalaite, mugearite — but the 

inverse or "reciprocal" variation of the rocks themselves seem not at all common 

iron oxides, whether the correlation on oceanic islands. This apparent scarcity 

implied is total or partial. of intermediate members, noted by a 

The correlation between FeO and Fe 2 3 number of petrologists, has been perhaps 

in Harker arrays is in fact extraordinarily most dramatically presented by Barth 

variable. Figure 40 is a histogram showing (Barth, Correns, and Eskola, 1939, p. 65). 

the distribution of r y » in the twenty-five Citing Hawaii as an example, Barth 

arrays for which computations have so remarks that in the intrapacific province 

far been completed. Observed relations rocks containing between 53 and 58 per 

between FeO and Fe 2 3 run the gamut cent of silica seem to be completely 

from strong negative correlation to lacking. Since Barth wrote, however, 

virtually perfect positive correlation. To three Hawaiian specimens with silica in 

the extent that the Harker diagram is the forbidden range have been reported, 

intended to provide a condensed descrip- two from Maui (Macdonald and Powers, 

tion of the data the use of a single Fe 1946, pp. 119 and 122) and one from 

variable will often be either misleading or Oahu (Tilley, 1950, p. 41). Although the 

uninformative. It would be preferable to search for them was perhaps in part 

return to the practice of Iddings; the stimulated by Barth 's remark, these new 

separate oxides of iron should be retained finds do render his example strictly 

as Harker variables whether or not some incorrect. And whether the precept it 

form of pooled Fe variable is also con- illustrated was ever correct will depend 

structed. on whose analyses are to be discarded. 

From elsewhere in the intrapacific prov- 

On the Relative Scarcity of Intermediate ince there are— and were at the time of 

Members in the Oceanic Basalt-Trachyte Barth's writing— at least seven other 

Association analyses of lavas falling in the forbidden 

„ Ch silica range, one from Easter Island and 

two each from Samoa, the Marquesas, 

There seems to be no question that and the Society group. But there are also 

basalt is by an enormous margin the at least twenty analyses of Pacific lavas 

principal oceanic lava and that trachyte in which 58 < Si0 2 < 63, so that, al- 

— often moderately feldspathoidal in though lavas with silica in the range 

either norm or mode — is a poor but 53-58 are perhaps not so rare in this 

uncontested second. The next most region as Barth suggested, analyses of 

important oceanic lava is probably pho- them are nevertheless considerably less 

nolite, and where phonolite is abundant, common than those of true trachytes. In 

as in Tahiti or Reunion, for instance, other oceans this central minimum is not 

hypabyssal or even plutonic feldspath- quite so clear cut; in the Indian Ocean, 

oidal rocks may occur. Rhyolite is on for instance, available data suggest no 

the whole very uncommon, though there shortage of intermediate silica values for 

are indeed a few occurrences whose the Kerguelen archipelago but at Reunion 



and Mauritius the situation is about like 
that in the Pacific. In the Atlantic a 
"Barth gap" seems to occur, though very 
weakly, at St. Helena, but on Ascension 
and in the Canaries the shortage is most 
evident in the 58-63 per cent Si0 2 range, 
while there are many analyses in the 
range 63-68 per cent. 

Table 11 shows the incidence of silica 
values in classes whose width is Y§ of the 
range observed in each island or island 
group. It will be noted that zeros occur 
only in classes 4 and 5, and that with two 
exceptions l's are confined to classes 3, 4, 
and 5. The column totals of the table 
leave little doubt that, for the array as a 
whole, the inference of some kind of 
hiatus — either a minimum or an outright 
discontinuity — in the parent distribu- 
tion (s) is almost unavoidable. 

Is it possible that this preponderance 
of trachyte over trachyandesite is merely 
a consequence of traditional cabinet- 
specimen sampling aimed at collecting 
rare and unusual material? On the 
assumption that the various lavas can be 
adequately identified in the field, it does 
not seem at all likely that collectors 
would have ignored material as rare and 
interesting as trachyandesite. If, as is 
suggested in many of the source papers, 
the hand-specimen distinction between 
trachyte and basalt is in fact difficult and 
unsatisfactory, there is even less likeli- 
hood of a serious sampling bias. There 
seems no reason to doubt that in this 
respect the distribution of analyses 
reflects, approximately at least, the dis- 
tribution of rocks, and that lavas inter- 
mediate in composition between basalt 

TABLE 11. Frequency of Silica Values in Sixths of Group 



for Larger 



Oceanic Basalt-Trachyte Suites 










Pacific Ocean 

Kohala-Hulalai 1 








Georgian and Society Islands 2 









Marquesas 3 









Samoa 4 








Easter 5 
















Atlantic Ocean 

Ascension 6 









St. Helena 7 







Canary Islands 8 



















Azores 9 
















Indian Ocean 

Reunion 10 








Mauritius 11 








Kerguelen 12 

























Sources: 1, Washington (1923); 2, Iddings and Morley (1918) and Lacroix (1923, pp. 279-289); 
3, Chubb (1930); 4, Daly (1924); 5, Bandy (1937); 6, Daly (1925); 7, Daly (1927); 8, Fuster, Ibarrola, 
and Lobato (1954); 9, Berthois (1953); 10, Lacroix (1923, pp. 227-237); 11, Walker and Nicolaysen 
(1954); 12, Edwards (1938). 



and trachyte appear to be less abundant 
than trachytes because they are in fact 
less abundant. One is tempted to point 
out imposing continental analogies: the 
lavas of eastern Otago, East Africa, the 
Iki Islands of Japan, etc. It is important 
to realize, however, that, although the 
data as a whole appear to indicate a 
minimum in the frequency distribution of 
silica in oceanic lavas, from only one island 
group do we have enough analyses to provide 
a reliable test for the significance of observed 
departures from uniform density in classes 
8, 4, 5, and 6 of table 111 

This is a regrettable and highly 
unsatisfactory state of affairs, one that 
should be remedied at the earliest 
opportunity. With the increasing funds 
available for oceanographic research it is 
to be hoped that the lavas of the oceanic 
islands will soon receive from ocean- 
ographers the same kind of attention 
space scientists are devoting to meteor- 

Granite in Port Clyde Peninsula 
Y. Suzuki and F. Chayes 

It is commonly supposed that there is 
a gradual transition from true granites to 
gabbros, with parallel tendencies toward 
increase of plagioclase over potash feld- 
spar, increase of An over Ab in plagio- 
clase, increase of color index, and decrease 
of quartz content. Our area, however, 
provides no support for this classical 
notion of a compositional continuum. 
Rather, there is a very strong suggestion 
that the two principal facies, biotite- 
muscovite granite and biotite-hornblende 
granite, are quite distinct and readily 

In the peninsula stretching from Rock- 
land southwestward to Port Clyde, 
Maine, there are many excellent expo- 
sures of granitic and dioritic rocks. These 
rock types outcrop within the outlined 
areas of figure 41, and sample localities 
are marked. They are intrusive into 
Paleozoic sediments. The unmarked area 
is underlain by Paleozoic rocks or glacial 

drift. Along the shoreline outcrops are 
abundant, although often deeply weath- 
ered. The area was once the center of a 
large quarrying industry. 

Following up some earlier studies of 
the quarries by Chayes, Suzuki spent 
about a month during the summer of 1960 
attempting to sample the intrusive com- 
plex systematically on a 1-km grid. The 
final distribution of specimens is shown in 
figure 41. Despite the enormous amount 
of natural and artificial outcrop in the 
area, the sample density varies greatly 
over the grid. By conventional standards, 
however, we have an unusually large 
sample of the complex upon which to base 
our report. 

Most of the outcrop area is underlain 
by fine-grained two-mica granite. The 
northeast part of the complex consists 
of coarse-grained hornblende-biotite gran- 
ite. Trondhjemite and biotite-quartz 
diorites are mostly confined to the 
western border. 

The average values for the two main 
granite facies are shown in table 12. The 
first column is based on modal analyses 
of thirty-nine specimens, the second on 
modes of twelve specimens. By conven- 
tional variance analyses all differences 
are highly significant. Although the 
differences in the two average modes are 
in the expected directions, their numer- 
ical values hardly suggest the expected 
continuity. The color index of the 
hornblende facies is only 4.1 per cent 
greater than that of the muscovite facies, 
but its quartz content is 8.0 per cent less. 
Differences between the other major 
constituents are, similarly, very much 
larger than the color-index difference. 
The real clue to the situation seems to be 
the presence or absence of hornblende — 
sometimes of very little hornblende. 

Although plagioclase is dominant over 
potash feldspar in all the hornblende- 
bearing granites, the most calcic plagio- 
clase so far encountered is An 35, and the 
quartz content of these rocks is usually 
well within the granite range (fig. 42). 

Rocks with no potash feldspar at all 



3 Km. 

Fig. 41. Sample localities in the Port Clyde peninsula complex. Double circle, muscovite granite; 
open circle, muscovite-biotite granite; cross circle, biotite granite; solid circle, hornblende-biotite 
granite; triangle, trondhjemite; diagonal cross, quartz-biotite diorite. 

may or may not be conspicuously inter- 
mediate between granite and gabbro in 
quartz content. But rocks containing even 
a little potash feldspar are, in this 
respect, not transitional at all. Rather, 
they are simply granitic. 

The plagioclase of all facies of the 
complex exhibits mild but persistent 
zoning, so that reliable estimation of its 
An content is difficult. The work on this 
problem has already been described ( Year 

Book 60, p. 169). Here it is only necessary 
to point out that, although in broad 
outline the results are compatible with 
the proposed transitional relation, there 
are striking exceptions, and the total 
range of An content is rather small. The 
plagioclase of the two-mica granites is 
oligoclase, or occasionally andesine. In 
the hornblende-biotite granite it is sodic 
andesine, rarely oligoclase. 

Sharp contacts between sizable masses 



TABLE 12. Modal Compositions of Muscovite-Biotite Granite and Hornblende-Biotite 

Granite of Port Clyde Peninsula, Maine 

Muscovite-Biotite Granite 

Hornblende-Biotite Granite 






Potash feldspar 













Color index 





Sample size 





Fig. 42. Modal ternary ratios of Port Clyde peninsula specimens. Symbols as defined in figure 41. 


of the two principal types of granite are modal or mineralogical properties of 

nowhere exposed, but we have so far these rocks. 

found no evidence for compositional The Ab content of a grain of potash 
gradation between the two. Although feldspar may receive contributions from: 
reliable identification cannot always be (1) plagioclase not removed by the 
made in hand specimen, microscopic separatory procedure, (2) perthitic inter- 
examination nearly always permits ready growths, (3) Ab in solid solution. Source 
assignment of an outcrop to one or the 1, contamination, can be held suitably 
other of the two main classes. Difficulties low in most of the Port Clyde peninsula 
arise only when both muscovite and rocks if only small amounts of concen- 
hornblende are absent, and such rocks trate are required. Optical properties 
are rare. measured on individual fragments or 

The exact relation between hornblende- parts of fragments of perthitic inter- 

biotite and muscovite-biotite granite in growths are concerned only with source 3, 

the Port Clyde area is still not known, but in estimates of composition by X-ray 

but it seems quite clear that some type of powder techniques preliminary heat treat- 

geochemical or stratigraphic discontinu- ment 5 converts contribution from source 2 

ity separates them. Further, there is into contribution from source 3. If the 

field evidence — inclusions, schlieren, brec- perthite appears to be of replacement 

ciation, etc. — that the relation of the origin, the composition determined by 

granites to the diorites and gabbro- X ray after heat treatment may be 

diorites of the complex involves extensive mineralogically interesting but petro- 

hybridization. graphically uninterpretable. If, as in the 

It is to be noted, too, that it is the Port Clyde area rocks, there seems no 

two-mica granite, not the hornblende- reason to suppose other than an exsolu- 

biotite granite, that is usually involved in tion origin for the perthites, the compo- 

this migmatization. Indeed, the large sition determined by X ray after homog- 

mass of hornblende-biotite granite in the enization (and inversion) is properly 

northeastern part of the outcrop area is regarded as an estimate of the compo- 

separated by the two-mica granite from sition of the alkali feldspar before 

the principal outcrop area of the dioritic exsolution, and this is a valuable datum, 

facies, which lies to the southwest. We should also like to know the Ab 

Thus, although the notion of a com- content of the K-feldspar phase now 

plete compositional continuum might visible in the rock. At present the only 

provide here, as elsewhere, a convenient methods purporting to give this informa- 

nomenclature and classification for the tion are optical, and they are exceedingly 

various facies of the complex, genetic rough. The biggest index difference 

inferences drawn from or based upon such between pure orthoclase and pure albite, 

a concept would be misleading. for instance, is 0.017 (Tuttle, 1952). With 

measurements subject to an uncertainty 

of, say, 5 in the fourth place, it is obvious 

Feldspar in the Granite of the Port Clyde t h at jf tne apparent difference between 

Peninsula ^wo observations is not zero it cannot be 

Y. Suzuki l ess than 3 per cent Ab. Whereas the 

X-ray procedure gives an average value 

This section describes variations in Ab for the homogenized specimen, the index 

content of potash feldspar in the granites measurements give a minimum value for 

of the Port Clyde peninsula and reports Or from the maximum gamma found in a 

attempts to determine whether such particular sample. 

fluctuations appear to be systematically s At 800°C, 1 kb, 1 week, water saturated, 

related to variations in other measurable after grinding. 



In table 13 the average Or per cent as 
determined by X ray is shown in column 
2, and in column 3 the maximum observed 
gamma index is recorded, followed by an 
estimate of the equivalent Or content 
from Tuttle's diagram (1952, p. 559). The 
fourth column gives the angle from which 
the "triclinicity" of Goldsmith and Laves 
(1954) is computed, and the fifth the 
average An content of accompanying 
plagioclase, determined by measurement 
of X-ray powder diagrams. The An 
content of accompanying plagioclase is 
clearly smaller in two-mica granites than 

in hornblende granites. The Goldsmith- 
Laves angle is somewhat larger in the 
two-mica granites, the largest value in 
the hornblende-biotite granites being less 
than the smallest in the two-mica granite. 
The difference between minimum and 
average Or per cent in alkali feldspar is 
large enough to suggest that the K 2 
content of this mineral varies from grain 
to grain. In several specimens its hetero- 
geneity seems beyond reasonable doubt, 
and it is curious that three of the five 
hornblende-biotite granites fall in this 
category. It is also curious that, although 

TABLE 13. Composition of Alkali Feldspar and Accompanying Plagioclase in Granite 

of the Port Clyde Peninsula 

Average An 


Average Or 

Or Per Cent from 

Angle between 

Per Cent in 

Per Cent 

Minimum y 

20 131 and 131 





Two-Mica Granite 

































1 . 5260 














































Hornblende-Biotite Granite 









1 . 5265 






















Muscovite Granite 













Biotite Granite 













Two-Mica Granite (minor dike) 









TABLE 14. Or in K Feldspar and An in Plagioclase by Rock Types 

(Data of table 13) 

No. Samples 

Average Or 

Per Cent in 

Potash Feldspar 

Minimum Or 

Per Cent in 

Potash Feldspar 

Average An 
Per Cent in 

All Data 

Standard deviation 



Two-Mica Granite 



Standard deviation 





Hornblende-Biotite Granite 

Standard deviation 






in the two-mica granites both the modal 
content of K feldspar and the minimum 
Or content of K feldspar show strong 
negative correlation with An in accom- 
panying plagioclase, the average Or per 
cent in K feldspar does not. Table 14 
shows, by rock type, the average values 
of columns 2, 4, and 6 of table 13. 

Certain of the classical "gradations" 
appear to be present within the mica 
granites. There is, for instance, a mark- 
edly inverse variation between the 
amount of K feldspar in the mode and 
the average An content of the plagioclase 
in the rock; since the total feldspar 
content is relatively stable it is not 
surprising then to find rather strong 
positive correlation between An content 
of plagioclase and plagioclase content of 

Although the Goldsmith-Laves angle of 
microcline increases almost linearly with 
increase of An content in plagioclase of 
the hornblende-bio tite granite, there is a 
strong suggestion of an opposite trend in 
the two-mica granites ; at present we have 
no explanation to offer for either of these 
effects. Although average Or in K feldspar 
does not appear to be significantly 
correlated with any other sample statistic, 
there is a fairly strong inverse variation 
between minimum Or in K feldspar and 
average An content of plagioclase. 

Two-Mica Granite and Hornblende-Biotite 

Y. Suzuki 

The marked modal differences between 
the two-mica and hornblende-biotite 
granites of the Port Clyde peninsula 
prompted a literature search for quanti- 
tative modal data about other closely 
associated granites of these two types. As 
might have been expected, this search 
was unsuccessful; the only detailed com- 
parison between these rock types possible 
at present is one that utilizes chemical 
analyses, and such a comparison is now 
in progress. 

To qualify for inclusion, an analysis (1) 
must be of a rock called granite in the 
source publication, (2) must list deter- 
minations of the nine essential oxides, 
(3) must be accompanied in the source 
publication by a "qualitative" mode or, 
at least, a list of essential minerals. (This 
list must of course show that it belongs 
in one of the two groups under dis- 

The need for the third requirement is 
obvious. The second was adopted largely 
as a means of eliminating partial analyses, 
on the perhaps questionable assumption 
that if a rock is not sufficiently interesting 
to warrant a full analysis it may also fail 



TABLE 15. Averages of Muscovite-Biotite and Hornblende-Biotite Granites 

Muscovite-Biotite Granite 

Hornblende-Biotite Granite 







Si0 2 





A1 2 3 





Fe 2 3 




















Na 2 





K 2 





Ti0 2 





Sample size 



to rate a good partial analysis. In fact, 
however, FeO and Fe 2 3 turn out to be 
of major importance. 

The desirability of the first requirement 
will be immediately apparent only to 
readers who have attempted to make use 
of any of the standard petrographic 
classifications in a study of granitic rocks. 
Although the classifications without ex- 
ception assign very broad compositional 
limits to "granite" — up to 80 per cent 
Si0 2 in CIPW, or as little as 5 per cent 
quartz in Johannsen, for instance — 
petrologists have customarily used the 
term in a much more restrictive sense. 
Our interest here is with real rocks that 
have actually been described as granite. 

To date, 78 analyses satisfying all three 
requirements have been found: 32 from 
North America, 16 from Finland, and 30 
from Japan. Means and standard devi- 
ations of the nine essential oxides are 
shown in table 15. The difference be- 
tween silica averages is suggestive, where- 
as that between the ferrous oxide averages 
is decisive. The range of Si0 2 in the 
hornblende-biotite granites is 64.47-76.68, 
and that of the muscovite-biotite granites 
is 67.20-75.86. There is thus a suggestion 
that FeO and Si0 2 may be merely 
compensating for each other. This, how- 
ever, is by no means the whole story. In 
subsets containing, respectively, all anal- 
yses with (a) more than 69 per cent Si0 2 , 

TABLE 16. Effect of Various Silica Restrictions on Average Compositions of 
Muscovite-Biotite (A) and Hornblende-Biotite (B) Granites 

of Si0 2 

More than 69.00 
Per Cent 

More than 72.00 
Per Cent 

Between 71.00 and 
74.00 Per Cent 

Per Cent 







Si0 2 







A1 2 3 







Fe 2 3 




























Na 2 







K 2 







Ti0 2 







Sample size 









(b) more than 72 per cent silica, and (c) 
between 71 and 74 per cent silica, the 
amount of FeO is significantly greater in 
the hornblende-biotite granites. These 
calculations are summarized in table 16. 
One-third of the hornblende-biotite 
granites and two- thirds of the muscovite 
granites contain less than 2 per cent FeO, 

and no significant differences between 
these "less than 2 per cent FeO" sub- 
groups were found. For these speci- 
mens the distinction between hornblende- 
biotite and biotite-muscovite granite is 
thus primarily physical rather than 
chemical. Work on this problem is 


Relationships between Crystal Structure 
and Crystal Morphology 

J. D. H. Donnay G and G. Donnay 

The second generalization of the law of 
Bravais, which was reported under this 
heading in last year's report (Year Book 
60, pp. 208-214), has now been success- 
fully applied to the unraveling of a 
particularly challenging morphology, that 
of the mineral barite. Ever since Mallard 
(1879, p. 318) and Friedel (1904, p. 339) 
applied the classical law of Bravais to 
barite, the morphological development of 
this species has remained an enigma to 
the present day, even though the crystal 
structure has been known for a long time 
(James and Wood, 1925). The first 
generalization (1937) of the law of 
Bravais is powerless, as was shown by 
Hartman and Perdok (1955) and by 
Seager (1959); the consideration of pseu- 
doperiods (Hartman, 1961) was helpful 
but not entirely satisfactory. 

Barite has an ionic crystal structure. 
It is well known that in certain simple 
ionic structures ions of equal charges but 
of opposite signs play the role of equiv- 
alent points when the law of Bravais is 
called upon to explain the morphology. 
The punctualization of the ionic charges 
is the basic postulate in this interpre- 
tation. Friedel made use of it in the 
classical case of NaCl, the crystal 
structure of which is governed by a 
face-centered cubic lattice, but where the 
morphology is controlled by the primitive 

6 The Johns Hopkins University. 

cubic lattice, with half the cell edge, that 
is obtained when all the ions are replaced 
by unit charges concentrated in the nodes 
of this new lattice. The sign of the charge 
can be disregarded because the strength 
of the bond Na + -Cl~ is equal to that of 
the bond Cl~-Na + . Friedel explains the 
morphology of calcite in the same way: 
the rhombohedral lattice whose nodes 
carry the double charges (either positive 
or negative) is the morphological lattice 
obtained by applying the law of Bravais 
of 1849; in this case, not only elementary 
ions (Ca ++ ) but complex ions (SO4)™ as 
well are punctualized. The application of 
the second generalization of the law of 
Bravais to the problem of barite has led 
us to a new type of punctualization, 
namely that of pairs of neighboring ions 
with the same sign. The reasoning pro- 
ceeds as follows. 

Let us start with the structural space 
group of James and Wood (1925), Pnma, 
with axial ratios a : b : c = 1.6304:1:1.3136. 
We first note that the dominant general 
form z receives the symbol (211) in this 
structural setting, whereas it should be 
symbolized (111) from the morphological 
point of view; it would then correctly 
define a primitive morphological lattice. 
The conclusion is that all (hkl) faces in 
the structural notation must obey the 
criterion "h even," which implies the 
halving of the structural a unit length in 
the three-dimensional bond assemblage. 
This agrees with a previous result of 
Hartman and Perdok (1955), that the 
energy period of the bond chain along the 
x axis is a/2. 



The zone of the (Old) faces is a simple 
zone with (Oil) dominant, which requires 
the corresponding reciprocal-lattice net 
to be primitive from the point of view of 
morphology, that is to say, of bonds. But 
the structural net b*c* has its mesh 
centered, owing to the n glide plane. We 
must, therefore, postulate additional 
extinction criteria that will require both 
k and I to be even. The dominant face will 
have to be symbolized (022). To the 
reciprocal net (26*, 2c*) there should 
correspond a direct net (b/2, c/2) that 
will express the periodicity of the two- 
dimensional bond assemblage of the 
projection of the crystal structure onto 
the yz plane. Turning now to the known 
barite structure (fig. 1 of James and 
Wood, 1925), we actually observe the 
predicted net if we replace by equivalent 
points the pairs of neighboring projected 
ions with the same sign. 

The zone of the (hOl) faces is also a 
simple zone with unit face dominant, 
which must obey the criterion "h even" 
(see above) and the additional criterion 
"I even" in order that (202) be the 
dominant face. As a group, the faces 
(hOl) do not occur frequently enough to 
have their indices co-prime: multiplying 
all the indices by 2 makes the faces in this 
zone recede to their correct ranks in the 
list of decreasing frequencies predicted by 
the generalized law of Bravais. Whereas 
the structure, projected onto the zx plane, 
has a primitive net with mesh ca, the 
two-dimensional bond assemblage should 
have a primitive mesh (c/2, a/2). This 
predicted mesh can indeed be recognized 
in figure 1 of James and Wood (1925) 
after the ca projection is suitably ex- 
tended. The pairs of ions with the same 
sign that are to be punctualized are not as 
obvious on inspection as in the be 
projection: there are two kinds of pairs 
of Ba ions and two kinds of pairs of S0 4 
ions, according as the line segment that 
connects the two ions in a pair slopes to 
the right or to the left. 

The zone of the (kkO) faces is a simple 
zone with (210) dominant. The condition 

"h even," which is imposed on all (hkl) 
faces, is also the criterion of the structural 
a glide plane, which requires the unit 
length a to be halved in the xy projection 
of the structure. This halving holds for 
morphology too: this zone does not yield 
any information other than the prediction 
that both the projected structure and the 
corresponding two-dimensional bond as- 
semblage have the same periodicity. As 
shown in figure 1 of James and Wood 
(1925) no punctualization of ions or pairs 
of ions can be found to define a mesh 
other than (a/2, b). 

In addition to planar projections, we 
must consider linear projections, on the 
coordinate axes. Such a projection of the 
crystal structure has a one-dimensional 
bond assemblage, whose period may be 
the same or smaller. The relative im- 
portances (frequencies of occurrence) of 
the pinacoids constitute the experimental 
data: we observe that c is the most 
frequent, and a the least frequent. 

According to the structural space group 
Pnma, the linear projections onto the 
coordinate axes x, y, z have periods a/2, 
b/2, c/2, respectively. This would imply 
the sequence a(200), c(002), 6(020) as the 
order of importance of the pinacoids, 
which is contrary to facts. To express the 
fact that a is the least frequent of the 
pinacoids, we must write it a (400), so 
that the predicted sequence becomes 
c(002), 6(020), a(400). This, in turn, 
requires that the one-dimensional bond 
assemblage of the projection of the 
structure onto the x axis have period a/4. 
This prediction can be checked in figure 1 
of James and Wood (1925) : the projected 
charges along the a length look as follows : 

+ + + + + + 

where the first and the last pair of positive 
signs are separated by translation a. If 
all pairs of equal signs are considered 
equivalent and replaced by points, the 
length a is divided by 4. 

We must now check that the other two 
linear projections of the structure are not 
divided by 4. This is immediately 



apparent for the b axis (fig. 1; James and 
Wood, 1925). Here, equal numbers of 
plus signs and minus signs are projected 
on the same points at y = and y = %; 
the structure is composed of electrically 
neutral planes; b is halved, both for the 
projected (linear) structure and for its 
one-dimensional assemblage. The situ- 
ation is not so clear for the a axis. Here 
the centers of the barium and sulfur 
atoms do not lie exactly in the same plane 
(parameters that should ideally be equal 
to 3^3 are found to be 0.333 and 0.305 by 
James and Wood, who place their origin 
at a center of symmetry). Although the 
structure cannot be said to consist of 
electrically neutral planes, it nevertheless 
results from the stacking of neutral 
layers, with period c/2. This period 
controls the linear bond assemblage in 
keeping with the morphological symbol 
(002) of the basal pinacoid. 

Finally we must justify the punctual- 
ization of pairs of ions. If we replace by a 
single central charge the two charges of a 
pair of ions with the same sign, we must 
introduce a compensating quadrupole 
that consists of the original two charges 
and two opposite charges placed in the 
center of the pair, next to the punc- 
tualized charge. Then we see that the 
bonds between successive equipoints of 
the bond assemblage are indeed rigorously 
equal in strength. Consider, for instance, 
along the a length, a first pair of Ba ions, 
followed by a pair of S0 4 ions, itself 
followed by a second pair of Ba ions. The 
interactions to be taken into account 
between two successive pairs of ions are 
of four kinds: charge-charge, charge- 
quadrupole, quadrupole-charge, and 
quadrupole-quadrupole. These interac- 
tions between the first Ba pair and the 
S0 4 pair are equal, each to each, by 
symmetry, to the interactions between 
the SO 4 pair and the second Ba pair. The 
centers of charge can thus be considered 
equivalent points from the point of view 
of bonding. 

We are indebted to Dr. H. F. Hameka, 
Johns Hopkins University, for suggesting 

to us the consideration of the quadrupole. 
The above results were given in the 
special issue of Kristallografiya published 
in honor of Professor N. V. Belov. 

Lattice Constant Refinement 
Charles W. Burnham 

Practically all phases of experimental 
mineralogy require knowledge of precise 
crystallographic lattice constants. Such 
values form the basis of detailed three- 
dimensional crystal structure refinements 
as well as studies of subsolidus phase- 
equilibrium relationships. To place pre- 
cise lattice constant determination on a 
routine basis a least-squares technique 
for lattice constant refinement has been 
developed and programmed for the IBM 
7090 digital computer. 

The refinement procedure is complete^ 
general in the. following respects: 

1. It is applicable to crystals of any 

2. It will accept data, from cards or 
tape, either as angle measurements for 
any wavelength or in the form of calcu- 
lated d values. 

3. Observations may be suitably 
weighted according to any scheme. 

4. Up to nine systematic correction 
terms may be included with each obser- 
vation. Each term consists of an unknown 
refmable parameter and a coefficient 
whose form may be any one of five 
different types. The functional form of 
each type of coefficient is programmed in 
a separate subroutine to suit individual 
experimental conditions. 

To allow for systematic errors, Bragg's 
law is modified to include an error in 0: 

n\/2d = sin (0 + Ad) (1) 

In practice n is absorbed by the reflection 
indices and will not appear in subsequent 
equations. Following the method of 
Cohen (1935), equation 1 is squared and 
expanded in a Taylor series retaining 
terms not involving powers of the error, 



= sin 2 + sin 20 Ad (2) 

The term A0 contains all systematic 
errors; it can be expanded to separate n 
distinct types of errors : 

= sin 2 + 2 sin 20 Ad k (3) 



The error terms, Ad k , are of the form 
X k f k (6), where X k is an experimental 
factor whose value is initially unknown 
and is to be refined, and 7^(0) is a function 
of 0, which generally vanishes at 6 = t/2 
(Buerger, 1942; Klug and Alexander, 

Cohen's method may be generalized by 
introducing reciprocal lattice notation: 



X 2 

= ~T (Thkl'Thkl) (4) 

Here r hk i is the reciprocal lattice vector 
for the reflection hkl. When the dot 
product is evaluated in terms of reciprocal 
lattice constants equation 3 is expanded 
and rearranged to give 

h 2 a* 2 + k 2 b* 2 + Z 2 c* 2 + 2/*/ca*6* cos 7* 

+ 2hla*c* cos |8* - - 2klb*c* cos a* 


, f /MV 4 sin 2 6 



9k(6) = -fk(6) (4/X 2 ) sin 26 (6) 

and e represents random error in the 

Equation 5 can be transformed to a 
linear equation in terms of the variations 
of the parameters by expansion in a 
Taylor series about a set of trial reciprocal 
lattice constants and experimental un- 
knowns. If only the first two terms of the 
expansion are retained, the transforma- 
tion yields 

Qcalc + S ^~ dClj + S ^v~ ^* 

fc=i dX k 

= (Jobs + 6 (7) 

where Q ca ic represents the left side of 
equation 5 evaluated using the trial 

parameters, the a 3 - are the reciprocal 
lattice constants, and 

Qobs = (4 sin 2 O bs)A 2 

Since equation 7 is linear in terms of the 
parameter variations, 8a 3 - and 8X k , a set 
of m of these equations, one for each 
observed 6, can be solved by standard 
least-squares techniques (Whittaker and 
Robinson, 1944) for the parameter vari- 
ations, provided that m ^ n + 6 (in the 
triclinic case). The refinement program 
generates and inverts the least-squares 
normal equations matrix according to a 
method developed by Busing and Levy 

When each observation is weighted in 
proportion to its reliability, equation 7 is 
multiplied by -\/wi, and the least-squares 
procedure minimizes ^Wiu 2 . The two- 


term Taylor expansion is exact for 
orthogonal unit cells, hence the least- 
squares parameter shifts, when algebra- 
ically added to the trial parameters, will 
yield a set of lattice constants for which 
the random errors are minimized. Two, 
or perhaps three, consecutive cycles in 
which the corrected parameters from the 
preceding cycle make up the new set of 
trial parameters may be required for 
complete convergence in the nonorthog- 
onal crystal systems. 

Since the residual, e, represents the 
difference between an observed and a 
calculated Q, the standard deviation of a 
measurement of 6 must be converted to 
the equivalent standard deviation of Q. 
The program automatically computes the 
proper least-squares weight for each Q 
according to 

Vw Q = - - = - A — _._ 0/> (8) 

<jq 4o-0 sin 2d 

The least-squares standard error of fit, 
corresponding to the standard error of an 
observation of Q of unit weight, is com- 
puted after each cycle of refinement 
according to 


S W i** 


L m 

n J 




where m is the number of observations 
and n is the total number of varied 
parameters. The variance-covariance ma- 
trix, \Vi\, of the varied parameters is 
obtained from 

\Vi\ = <ro\B\-i (10) 

where \B\ is the n X n least-squares 
normal equations matrix. The standard 
errors of the varied parameters are, of 
course, the square roots of the diagonal 
terms of \Vi\. 

Following each least-squares cycle the 
new values of the direct lattice constants 
and the unit-cell volume are evaluated 
using standard formulas (Buerger, 1942). 
The direct lattice constant variance- 
covariance matrix, | V a \ , is obtained from 
the reciprocal variance-covariance matrix 
according to (D. Handwerker, personal 
communication, 1962): 

\V d \ = \D\ \V r \ \D\ T (11) 

where \V r \ is the 6X6 reciprocal lattice 
constant variance-covariance matrix, con- 
taining terms from \Vi\ plus appropriate 
zeros for nontriclinic cases, and 

D\ = 

da da 



dy dy 



The standard errors of the direct lattice 
constants correspond to the square roots 
of the diagonal terms of \V d \. The 
standard error of the unit-cell volume is 
evaluated in an analogous manner: 

erV = 

'E\ IF, 



where \E\ is the row vector containing 
the partial derivatives of V with respect 
to the direct lattice constants. 

The printed results from each refine- 
ment cycle include a list of observed and 
calculated d values, the residuals (d bs — 

dcaic) and (Qobs — Qcaic), and the weighted 
residuals (d ohs — d ca i c )/<r d and (Q obs — 
Ocaic)/^ based on the trial parameters. 
The least-squares results contain the 
reciprocal lattice constant and systematic 
correction term experimental parameter 
shifts and standard errors in addition to 
the direct lattice constant shifts and 
standard errors. The asymmetric part of 
the direct lattice constant variance- 
covariance matrix is made available for 
subsequent inclusion in interatomic dis- 
tance and angle error computations. 

To illustrate the results obtained with 
this procedure, table 17 lists refined 
lattice constants for kyanite, Al 2 Si0 5 
(triclinic). A single crystal of kyanite 
from Burnsville, North Carolina used for 
intensity measurement for structure re- 
finement (Burnham, 1962), was used to 
obtain precision Weissenberg (Buerger, 
1937) photographs about the a, b, and c 
axes. Of the 79 film measurements 
employed in the least-squares analysis 23 
were of Old reflections, 20 were of hOl 
reflections, and 36 were of hkO reflections. 
Since the precision Weissenberg film 
measurement, /, is linearly related to 0, 
and all measurements were considered to 
have equal precision, all O bs were 
weighted unity. Column 1 of table 17 
lists the results obtained when systematic 
correction terms were included to com- 
pensate for film shrinkage, specimen 
absorption, and camera eccentricity. The 
coefficients, g(6), of equation 5 were 
assigned the following forms (Buerger, 
1942) : 

Film shrinkage : 

<7(0) 8 hr = ^ (| - 0) sin 26 
Absorption : 


Eccentricity : 



0(0) abs = ^ cos 2 6 sin 20 

^(0) ecc ==^-sin 2 20 

Separate film shrinkage and absorption 
corrections were applied to data from 



TABLE 17. Kyanite (Al 2 Si0 5 ) Lattice Constants 

Seven Systematic 



No Systematic 



Precession (Skinner, 

Clark, and Appleman, 


a > A 

7.1192 ±0.0005 

7.1197 ±0.0004 

7.121 ±0.002 


7.8473 ±0.0004 

7.8479 ±0.0003 

7.846 ±0.002 

c, A 

5.5724 ±0.0006 

5.5736 ±0.0004 

5.577 ±0.005 

a, deg 

89.977 ±0.005 

89.969 ±0.006 

89.97 ±0.08 

0, deg 

101.121 ±0.005 

101.126 ±0.006 

101.15 ±0.08 

7, deg 

106.006 ±0.003 

106.001 ±0.003 

106.00 ±0.08 

Unit-cell volume, A 3 

293.16 ±0.06 

293.28 ±0.03 


different films. One eccentricity term was 
applied to all observations. Complete 
least-squares convergence was attained 
after two iterations. 

Column 2 of table 17 contains the 
results obtained with the same data using 
no systematic correction terms. Column 
3 lists the results obtained by Skinner, 
Clark, and Appleman (1961) with quartz- 
calibrated precession data from another 
specimen of Burnsville kyanite. 

It must be emphasized that the least- 
squares standard errors represent the 
precision attainable with a specific set of 
data. The precision will, of course, vary 
with the ratio of observations to refinable 
variables. The values are, in general, 
conservative, since they implicitly involve 
all correlation, or parameter interaction, 
effects. They should not be construed, 
however, as the accuracy to be expected 
when several sets of parameters obtained 
by the same or different X-ray techniques 
on the different samples are compared. 

The Crystal Structure of Sillimanite 
Charles W. Burnham 

Details of the crystal structure of 
sillimanite are essential to an understand- 
ing of the crystal chemical relationships 
between the Al 2 Si0 5 polymorphs (anda- 
lusite, sillimanite, kyanite). Because of 
the extreme similarity of their X-ray 
diffraction patterns, a well determined 
sillimanite structure must, in addition, 
form the basis of detailed studies of the 
structures of mullites of various compo- 

sitions. A three-dimensional refinement of 
the previously determined sillimanite 
structure (Taylor, 1928; Hey and Taylor, 
1931) was undertaken with single-crystal 
counter diffractometer data measured on 
a small cleavage fragment of clear 
sillimanite from LaBelle County, Quebec. 
The unit-cell dimensions of this specimen 
were refined to the following values: 
a = 7.4856 ± 0.0006, b = 7.6738 ± 
0.0003, c = 5.7698 ± 0.0008 A. 

Preliminary least-squares refinement of 
the structure (Burnham, 1961) reduced 
the unweighted disagreement factor R to 
10.3 per cent and the weighted (root- 
mean-square) R to 4.8 per cent. At that 
stage agreement between observed and 
calculated structure factors for the sub- 
structure reflections {I even) was excellent 
(table 18) whereas that for reflections 
with I odd indicated almost complete lack 
of complement structure convergence. 
Attempted refinement of disordered mod- 
els and a noncentrosymmetric model 
failed to improve complement structure 
agreement but had little adverse effect on 
the substructure R value. This indicated 
that the substructure reflections are very 
insensitive to minor structural changes 
and that substantial convergence of the 
complement structure will be required 
before the details of the structure can be 

During the past year the sillimanite 
study has been continued. Analysis of the 
refinement procedure demonstrated that 
convergence had not been attained be- 
cause of strong mathematical interactions 



TABLE 18. Sillimanite Disagreement Factors, R 

Hey and 
Taylor (1931), 






Unweighted R 

Weighted R 

z\\Fo\ - \F e \r 

'Hw(\Fo\ - \Fc\) 

ZwFo 2 

Even-level unweighted R 
Odd-level unweighted R 










between structure parameters. Least- 
squares correlation coefficients between 
pairs of atomic coordinates whose differ- 
ences determine the complement struc- 
ture varied from —0.66 to —0.81. 

Observed structure factors had been 
assigned least-squares weights in inverse 
proportion to their variances as deter- 
mined by counting statistics. The distri- 
bution of sillimanite structure factor 
magnitudes in reciprocal space is not 
random; reflections on odd reciprocal 
lattice levels normal to the c axis receive 
intensity contributions from the comple- 
ment structure alone. Because of the 

resulting unfavorable counting statistics 
the average weight assigned to these 
observations was 0.07, compared with the 
average weight of 0.20 assigned to the 
substructure observations on even recip- 
rocal lattice levels. 

Underweighting of the critical class of 
observations proved to be the primary 
cause of the strong parameter inter- 
actions. When all structure factors with 
measurable values were assigned weight 
1.0 and those whose values were below 
the minimum observable value were 
assigned weight 0.01, all structure param- 
eters were effectively uncoupled. Com- 

TABLE 19. Sillimanite Atom Coordinates 


Hey and 





Taylor (1931) 



O a : 













O b : 















O c : 













O d : 




































Al 2 : 
















plete convergence was attained after nine 
additional least-squares cycles during 
which all atomic coordinates and aniso- 
tropic temperature factors were varied. 
The final R values, listed in table 18, 
confirm Pbnm as the correct sillimanite 
space group. 

The refined atomic coordinates are 
compared with those of Hey and Taylor 
(1931) in table 19. Although refinement 
produced significant coordinate shifts, it 
did not alter the basic geometrical 
relationships between coordination poly- 
hedra. Chains of slightly distorted alumi- 
num octahedra run parallel to the c axis 
and are supported by double chains of 
aluminum and silicon tetrahedra. Differ- 
ences in interatomic distances (table 20) 

TABLE 20. 

Sillimanite Interatomic 










Si tetrahedron 





Si-O c 




Si-O d 




O a -O d 




O a -O c 




O c -Od 




O d -O d ' 




Al tetrahedron 

Al 2 -0 6 




Al 2 -O c 




Al 2 -O d 




6 -0<* 




O b -O c 




o c -o d 








Al octahedron 













Oa-O b 




Oa-Ob" (shared) 2 



Oa-O d 








0d'"-0 o 




Od"'-0 6 




* Atoms designated with a single prime 
represent transformation of the unprimed atom 
in the same coordination polyhedron according 
to x' — x, y' — y, z' = }/2 — z. Double primes 
represent transformation according to x" = — x, 
y" = —y, z" = l /2 -\- z. Triple primes represent 
transformation to a centrosymmetric equivalent. 

show that the distribution of aluminum 
and silicon in the tetrahedra is ordered. 
Figure 43 illustrates the bonding within 
and between coordination polyhedra. 

The tetrahedral double chains are of 
particular crystal chemical interest. Each 
double chain may be thought of as a 
continuous series of four-membered rings, 
each ring containing two silicon and two 
aluminum tetrahedra in the sequence 
Si-Al-Si-Al. The Si-O c -Al 2 bond angle of 
171.6° and the Si-O d -Al 2 bond angle of 
114.4° control the basic configuration of 
the ring. Whereas the O-Si-0 tetrahedral 
angles are close to ideal (107.4° to 111.3°), 
the Si-0 bond distances show that the 
silicon atom is not at the center of its 
tetrahedron but is measurably displaced 
toward O c . The aluminum tetrahedron is 
more irregular, but, again, the cation is 
closest to O c . 

The average Si-0 distance is 1.615 A, 
and the average tetrahedral Al-0 distance 
is 1.770 A. Smith and Bailey (1962) show 
that this average Si-0 distance is close to 
that for other silicates in which three 
corners of each tetrahedron are shared 
with other tetrahedra. They also suggest 
that the average Al-0 distance is close to 
the extrapolated value for layer silicates. 
If these are to be accepted as "expected" 
averages for sillimanite, the anomalous 
positions of the cations must be explained. 

The anisotropic temperature factors 
for O c indicate a vibrational configuration 
corresponding to an oblate spheroid 
whose circular equator lies in the plane 
normal to the mirror plane containing Si, 
Al 2 , and O c , and is essentially parallel to 
the bisector of the Si-O c -Al 2 angle. The 
root-mean-square amplitude of vibration 
in the equatorial section is 0.12 ± 0.01 A; 
that normal to this section and directed 
toward Si and Al 2 is 0.06 ± 0.02 A. The 
equivalent isotropic temperature factor 
calculated for O c is 0.86, as compared 
with values ranging from 0.35 to 0.50 for 
the other three oxygen atoms in the 
asymmetric unit, and 0.36 to 0.42 for the 
oxygen atoms in andalusite (Burnham 
and Buerger, 1961). 



Fig. 43. Projection on (001) of the refined structure of sillimanite showing cation-anion bonds. 
The z coordinate of each atom is given beside its designation. 

Since there is a local charge imbalance 
of —0.25 on O c , the abnormally short 
bond distances could be attributed to 
excess Coulomb attraction between the 
cations and O c . The indicated vibrational 
anisotropy may consequently arise as 
compensation for an electron-density 
distribution related to anomalous bond 
character not considered in the spherical 
scattering factor curves used in refine- 
ment. This explanation, however, does 
not appear to be consistent. The Al 2 -0& 
distance is 0.04 A larger than the Al 2 -O c 
distance, yet 0& also bears a charge 
deficiency of —0.25. The shortest Si-0 
bond in andalusite is directed to an 
oxygen with a charge excess of 0.2, and 
the two shortest Al-0 bonds in the five- 
coordinated aluminum group of andalu- 
site involve oxygens with charge imbal- 
ances of —0.4 and +0.1. 

Alternatively, the indicated thermal 
motion of O c may represent an average 

electron-density distribution of atoms 
with normal vibration amplitudes but 
different time-average coordinates in 
different unit cells. Under this hypothesis 
the actual position of O c from ring to ring 
is displaced from the coordinates listed in 
table 19 to positions within the equator 
of the oblate vibrational spheroid but not 
necessarily on the mirror plane. An 
approximation to the resulting effect on 
bond distances can be computed by 
averaging the distances over the indicated 
thermal motion assuming the cation and 
anion to vibrate independently. Averag- 
ing increases the Si-O c distance to 1.576 A 
and the Al 2 -O c distance to 1.732 A. Both 
increases are significant relative to the 
standard errors of the bond distances, but 
are not sufficiently large to normalize the 

Further studies of this important 
crystal chemical problem are now under 
way. The positional variation hypothesis 



will be tested by examining the structure 
at very low temperatures. If the oxygen 
atom, O c , is, in fact, statistically distrib- 
uted, the apparent thermal motion should 
not diminish with decreased temperature. 
If, however, the large temperature factor 
actually represents thermal vibration, it 
will be measurably reduced at low 

The Crystal Structure of Fe Mica 
N. Morimoto, J. D. H. Donnay, 7 and G. Donnay 

Work on synthetic iron mica {Year 
Book 60, p. 214) has been continued. The 
crystal structure was determined by the 
three-dimensional least-squares methods. 
The computations were carried out at the 
National Bureau of Standards, with the 
help of Dr. Helen Ondik, on the IBM 
7090, using the modified Busing program. 
The intensity data, without absorption 
correction, gave R — 0.23, including non- 
observed reflections, and R = 0.13, 
excluding nonobserved reflections, after 
four cycles of least-squares refinement. 
The absorption correction was then 
applied to the data by means of the 
program of C. W. Burnham. The cor- 
rected data gave R = 0.21 or 0.09, 
according as the nonobserved reflections 
were or were not included, after three 
cycles of refinement. These last compu- 
tations were performed on the IBM 7090 
of the Johns Hopkins Computing Center, 
using the Trueblood program as modified 
by Koenig with different isotropic tem- 
perature factors for the different atoms. 
The work is still in progress. 

On the Transitions of Bornite 
N. Morimoto 

The transition mechanisms of the three 
polymorphic forms of bornite (Morimoto 
and Kullerud, 1961) were studied from 
the structural viewpoint. 

The crystal structure of the high- 
temperature form is essentially the anti- 

7 The Johns Hopkins University. 

fiuorite structure, only slightly more 
complicated. The sulfur atoms occupy 
the nodes of the cubic face-centered 
lattice with a = 5.50 A, being cubically 
close-packed. Each sulfur tetrahedron, on 
the average, contains % of a metal atom. 
This fractional atom is itself statistically 
distributed over twenty-four equivalent 
sites inside the sulfur tetrahedron. Thus, 
in the whole unit cell, six metal atoms 
are statistically distributed over 24 X 8 
= 192 sites. 

The cubic edifice of the metastable form 
is a result of twinning of a large number 
of small domains in eight different 
orientations. Each such crystal has a 
rhombohedral cell with a r h = 6.70 A and 
a = 33°32'. 

The structure of this rhombohedral 
form can be derived from that of the 
high-temperature form considered along 
the body diagonal (111) of the cube 
(fig. 44). All the sulfur atoms stay in 
place, retaining the cubic close packing. 
Of the four sulfur tetrahedra sites, two 
do not change at all. One becomes vacant, 
and the metal atom that occupied it in 
the high-temperature form is redistrib- 
uted among the other three sites. The 
corresponding three sulfur tetrahedra 
now contain one full atom apiece. To 
compensate for the vacant site, the last 
metal site is slightly displaced. The 
statistical distribution of % of a metal 
atom among twenty-four possible sites 
inside each sulfur tetrahedron changes to 
the statistical distribution of one metal 
atom among four possible sites. 

Figure 45 shows the structural relations 
between the high-temperature and the 
metastable forms, both of which consist 
of layers parallel to (lll) r h- Two struc- 
tures are built on the basis of the cubic 
close packing of the sulfur atoms. The 
statistically distributed metal atoms are 
represented as bands. 

The distance between the Mi layer and 
the sulfur layer becomes shorter in the 
metastable form, suggesting the possi- 
bility that the Fe atoms concentrate in 
Mi layers. Although the structure of the 







18.95 A 






Fig. 44. Derivation of the structure of metastable form from that of high-temperature form. 



1 6.95 A 





■ • . 




































-* 1 


0.75 M 

0.75 M 
0.75 M 

0.75 M 
0.75 M 

0.75 M 
0.75 M 

0.75 M 
0.75 M 

0.75 M 
0.75 M 

0.75 M 

1 r 1 

, ; \s ... 


•••' : . '"''.•• " " . - 

• ' . ■,. " \ '"' - - : ■ .'.' 

[ i 1 
















1 — - 





















Fig. 45. Layer structures of the three polymorphic forms of bornite. 

low-temperature form was not actually 
determined, it seems likely that the metal 
atom, statistically distributed at the four 
corners of a tetrahedron in the metastable 
form, will occupy one of the four sites in 
the low-temperature form. The stoichio- 
metric composition confirmed for most 
natural bornite suggests that the Fe 
atoms must take some definite positions, 
the Mi positions, which are closer to the 
sulfur atoms than other metal positions. 
This relation in the low-temperature form 
is shown by lines in figure 45. 

The arrangements of the metal vacant 
layers change their orientations according 
to a simple twin law in the metastable 
and possibly in the low-temperature 
forms. Domain structures always take 
place on transition from the high- 
temperature to metastable forms. This 
indicates that the metal vacant layers 

cannot stably keep their orientation over 
a long distance. The diffracted X rays 
from each crystal ( = domain orientation) 
are not coherent with those from other 
crystals. The domains themselves, how- 
ever, must be small, since the twins 
cannot be recognized as such by direct 
methods of observation. The volumes of 
the different domain orientations must be 
nearly equal so as to give cubic or 
tetragonal symmetries. 

The twinning found in the metastable 
bornite is different from usual twins in 
that most of the atoms (Si, Sn, Mn, and 
Mm) build a continuous periodic struc- 
ture throughout the whole edifice, so that 
the twin relations apply only to the Mi 
and the vacant M positions. Such slight 
structural rearrangements take place that 
the transition heat should be very small 
and the transition unquenchable. 



Increased emphasis has recently been the Fe-S system at temperatures below 

placed on applications of the synthetic 200°C. 

systems to ores. Specimens have been The mineral bravoite was found to be 

systematically collected in a number of stable below 137°C by experiments 

mines, and their mineral assemblages employing the method of mixing aqueous 

have been studied in polished sections and liquids. Bravoite is a common product of 

by X rays. Employment on these mineral alteration in numerous ores, and, since 

assemblages of the geological thermom- its thermal stability is now known, its 

eters that we have developed in recent presence may serve as a valuable geo- 

years has produced interesting new logical indicator. These experiments also 

information about the formation of the indicate that the a(Ni,Fe)i^J3 phase is 

ores but has also brought forth new stable at least down to 150°C, and they 

problems that demand solution. have further produced data enabling us 

Progress in these studies frequently to draw the phase diagram for the system 
depends on development of new research at 150°C. Differential thermal analysis 
methods and their subsequent systematic and high-temperature X-ray diffraction 
employment to produce a steady flow of studies of synthetic as well as natural 
new data. The investigations started with pentlandite (Fe , Ni) 9 S 8 have demon- 
the most basic systems and have pro- strated that pentlandite is stable only 
gressed step by step to those sufficiently below 610°C, at which temperature it 
complex to include the most important decomposes to pyrrhotite (Fei_ x S) and 
minerals of many common ore types. To the Ni 3 ±xS 2 phase. This mineral corn- 
date twelve binary and twelve ternary monly occurs with pyrrhotite in ultra- 
systems have been studied, and rapid basic rocks. The sulfides are believed to 
progress is being made in quaternary have segregated as liquid drops from the 
systems. rock magma. Earlier interpretations of 

Laboratory experimentation has re- the ore assemblages were based on the 
cently been facilitated by the develop- assumption that the pentlandite-pyrrho- 
ment of a simple apparatus for mixing tite pair is stable to at least 850°C. The 
aqueous solution in closed systems and at new information on pentlandite stability 
controlled elevated temperatures. The relations necessitates reinterpretation of 
upper temperature limit of this method field occurrences and has significant effect 
in investigated systems slightly overlaps on the theory of formation of such ores, 
the lower temperature limit for attain- Differential thermal analysis experiments 
ment of equilibrium in corresponding dry have also demonstrated that a two-liquid 
systems. The mineral assemblages and + vapor region extends across the sulfur- 
solid solution compositions obtained in rich part of the ternary Fe-Ni-S system 
the overlapping temperature range after at temperatures above 991°C for the Ni-S 
years of heating of the dry systems are side and above 1083°C for the Fe-S side, 
identical with those developed in a few The upper stability curve of linneite 
hours in the aqueous systems. Therefore, (Co 3 S 4 ) has been determined up to 2000 
when such identity can be established, bars. In the presence of vapor this 
this method can be employed to deter- mineral decomposes to Coi-^S and CoS 2 
mine the phase relations in many impor- at 665°C. This study, similar to the study 
tant systems at low temperatures and by of poly dy mite, is of interest because the 
experiments lasting only a few hours each, reaction Co 3 S 4 ^ Coi-^S + CoS 2 involves 
This method has already been applied to only solids. 

parts of the Fe-Ni-S system and is now In the literature several phases have 

being used to clarify phase relations in been reported to exist in the Mo-S 



system. However, only hexagonal molyb- 
denite (MoS 2 ) has been established as a 
mineral species. Two MoS 2 forms were 
made synthetically, one rhombohedral at 
low temperatures and one hexagonal at 
elevated temperatures. Apparently the 
rhombohedral form is metastable. The 
only other phase obtained in the system 
is a monoclinic compound of approxi- 
mately M02S3 composition. This phase is 
not stable below 610°C, where Mo and 
MoS 2 are stable together. 

The phase relations in the Cu-Ni-S 
system have been studied at 600°C. No 
ternary compound occurs, and solid 
solutions extend only very short distances 
into the ternary system from the binary 

The Fe-Ni-As system has been studied 
at 800°C. Two ternary phases occur: 
(Fe,Ni)As 3 solid solution, which is not 
found in nature, and an intermediary 
solid solution, (Ni,Fe)2As, corresponding 
to the mineral oregonite. Extensive solid 
solutions exist between some of the 
phases, for instance between FeAs and 

Studies of the Fe-Mo-S system have 
centered on the stability relations of the 
pyrite (FeS 2 ) -molybdenite (MoS 2 ) min- 
eral pair. These two minerals are stable 
together below 726°C. At this tempera- 
ture invariant conditions exist in the 
system, and the five phases pyrite, 
molybdenite, pyrrhotite, liquid, and va- 
por are all stable. Above the invariant 
point pyrite is no longer a stable phase in 
the presence of molybdenite, and pyrrho- 
tite-molybdenite becomes the stable min- 
eral pair. 

Investigations of the complicated sys- 
tem Cu-Fe-S have shown that at various 
temperatures the pyrrhotite compositions 
of the pyrite-pyrrhotite-chalcopyrite as- 
semblage are significantly different from 
those of the pyrite-pyrrhotite assemblage. 
These results indicate that pyrrhotite 
temperatures determined on ores con- 
taining chalcopyrite as well as pyrrhotite 
and pyrite are from 45° to 60°C lower 
than those that would have been obtained 

had chalcopyrite not been present. 

Exsolution textures developed on cool- 
ing of synthetic bornite-type solid solu- 
tions have been correlated to those found 
in ores. This study indicates that the 
thermal history of an ore body cannot be 
surmised from the presence of exsolution 
lamellae of one mineral in another. 
Exsolution lamellae, as shown in labora- 
tory experiments, may indicate rapid 
cooling, which probably does not take 
place in ore deposits, or they may origi- 
nate from a solid solution of low concen- 
tration that cooled at a slow rate so that 
the degree of supersaturation was always 
relatively low. 

Chalcocite-chalcopyrite assemblages 
are sometimes observed in ores. These 
minerals are incompatible at high tem- 
peratures but, owing to the variation in 
the metal-to-sulfur ratio in the chalco- 
pyrite field, may form a stable assemblage 
at very low temperatures. 

The phase relations determined on 
synthetic systems have been applied to 
systematically collected ore specimens 
from many localities. Polished-section 
studies of minerals from the copper 
deposits of the Keweenaw peninsula 
revealed the presence of several important 
minerals not previously reported from 
this district. These mineral associations 
give valuable information about the phase 
relations in the ternary system Cu-Ni-As. 

Sphalerite-pyrrhotite and pyrrhotite- 
pyrite temperatures have been deter- 
mined from numerous samples from the 
Brabant Lake, Saskatchewan, ores; from 
the Ducktown, Tennessee, mines; from 
the Elisabeth Mine, Vermont; from the 
Outukompu district in Finland; and from 
Sulitjelma, Norway. 

The Mo-S System 

A r . Morimoto and G. Kullerud 

Study of the Mo-S system by quench- 
ing, microscope, and X-ray methods was 
initiated primarily to elucidate the phase 
relations between molybdenite (MoS 2 ), 



the most important source of molyb- 
denum, and the other phases reported in 
the system. The information obtained 
will serve as a necessary basis for studies 
of more complicated systems involving 
molybdenum and sulfur, such as Mo-Fe-S. 

Among the many reported phases in 
this system, only molybdenite, the hex- 
agonal form of molybdenum disulfide, is 
established as a mineral species. Recently, 
molybdenum sesquisulfide (Mo 2 S 3 ) was 
confirmed as a stable phase above 1Q00°C, 
coexisting, depending on the composition, 
with Mo or MoS 2 and sulfur vapor 
(McCabe, 1955; Stubbles and Richard- 
son, 1960). A new form of MoS 2 with 
rhombohedral symmetry was synthesized 
at about 900°C (Bell and Herfert, 1957). 

Rhombohedral MoS 2 has the cell 
dimensions a = 3.16 ± 0.1 A and 
c = 18.37 ± 0.03 A. The c translation is 
1 J/2 times as long as that of the hexagonal 
form. The crystal structure, given by Bell 
and Herfert and later revised by 
Semiletov (1962), has the same kind of 
layer structures as the hexagonal form, 
where Mo atoms are in triangle prisms of 
S atoms. Mo 2 S 3 has monoclinic symmetry 
with a = 8.6335, b = 3.208, and c = 6.092 
A, and (3 = 102°43 / . In this compound, 
however, Mo atoms are coordinated by 
octahedral arrangements of S atoms 
(Jellinek, 1961). 

Mo 2 S 3 appears to be stable only above 
610° ± 5°C. When the elements are used 
as starting materials, Mo 2 S 3 appears 
above 610° ± 5°C. Below this tempera- 
ture Mo and MoS 2 are obtained. But 
when Mo and MoS 2 are the starting 
materials, Mo 2 S 3 is not obtained even 
after 30 days at 650°C. On the other 
hand, once Mo 2 S 3 is formed it does not 
break down even after being heated for 
1 month at 600°C. The reaction rates of 
the system are so slow that equilibrium 
assemblages are not obtained even at 
800°C in a reasonable time. Above 900°C, 
however, equilibrium is usually estab- 
lished in less than 1 week. The exact 
composition of the Mo 2 S 3 phase, deter- 
mined at 935°C, was found to be Mo 2 . eS 3 , 

which deviates slightly from the stoichio- 
metric ratio. Measurements of the posi- 
tions of reflections in X-ray powder 
patterns of "Mo 2 S 3 " grown in equilibrium 
with Mo and of those of "Mo 2 S 3 " grown 
in equilibrium with MoS 2 give identical 
results, indicating a very limited solid 
solution, if any, in this phase at 935°, 
800°, and 700°C, the temperatures of 
these experiments. 

According to Semiletov, the structural 
differences between the hexagonal and 
the rhombohedral forms of MoS 2 can be 
explained by assuming different stacking 
orders of S-Mo-S layers. MoS 2 synthe- 
sized below 900°C gives X-ray powder 
diffraction patterns with broad peaks, 
and, in general, the lower the temperature 
of synthesis the broader are the peaks. 
These poorly defined peaks do not fit 
exactly either with those of the hexagonal 
form or with those of the rhombohedral 
form and are on the whole similar to 
diffraction effects commonly attributed 
to stacking faults in layered structures. 
Above 900°C the peaks become sharp and 
distinctly show the hexagonal pattern. 
Natural MoS 2 always shows the hexag- 
onal form, and, once synthesized, the 
hexagonal form of MoS 2 does not change 
to the rhombohedral form or to any 
intermediate form even at low tempera- 
tures or after prolonged heating. We 
believe that the rhombohedral form is 
metastable throughout the entire tem- 
perature range. 

Experiments designed to determine 
possible solid solution on either side of 
MoS 2 composition showed that, within 
the limits of error of our methods, M0S2 
is stoichiometric. 

The Fe-Ni-S System 
G. Kullerud 

Liquid immiscibility . Liquid immisci- 
bility between sulfides and silicates has 
been postulated as a mechanism for the 
enrichment of many important ores 
through magmatic segregation with the 
sulfides separated from the silicate magma 



by gravity settling. These sulfides consist 
mainly of mixtures of pentlandite 
(Fe,Ni) 9 S 8 and pyrrhotite (Fei_ x S), 
which are sulfur poor compared with 
sulfides of other types of deposits. It was 
suggested in last year's report that the 
metal-rich sulfide drops not only separate 
by gravity settling from a silicate 
magmatic solution but may, even before 
this event, have formed through liquid 
immiscibility among the sulfide phases. 
This view is supported by results of 
recent investigations in the ternary 
system Fe-Ni-S. 

A region of liquid immiscibility was 
found by Kullerud and Yund (1962) to 
exist in the Ni-S system above 991°C and 

over a composition range from 54.5 to 
more than 97 weight per cent S. Kullerud 
(Year Book 60) reported the existence of 
a liquid immiscibility region above 
1083°C and over a composition range 
from 46.2 to more than 95.5 weight per 
cent S in the binary system Fe-S. 
Additional differential thermal analysis 
experiments on ternary compositions 
have now shown that the liquid immisci- 
bility region extends across the Fe-Ni-S 
system. Figure 46 shows the results 
obtained for various amounts of sulfur in 
a section in which the Fe/Ni ratio is 
constant (61.4 Fe, 38.6 Ni weight per 

In all experiments with more than 


30 35 X 40 


Weight per cent 

Fig. 46. Phase relations in the section from Fe,Ni alloy with 38.6 per cent nickel to sulfur. Only 
the part containing more than 30 per cent sulfur is shown. 



41.5 weight per cent S a thermal effect, 
caused by the breakdown of nickel- 
bearing pyrite (Clark and Kullerud, 
Year Book 58), was observed at 729°C. 
At this temperature invariant conditions 
exist, and the five phases, nickel-bearing 
pyrite, iron-bearing vaesite, hexagonal 
(Fe,Ni)i_a;S solid solution, liquid, and 
vapor, are all stable. A second heat effect 
observed at 862°C in all experiments with 
41.8 weight per cent or more sulfur 
demonstrated the disappearance of iron- 
bearing vaesite from the section. 

Above this temperature divariant con- 
ditions exist. The phases are now 
(Fe,Ni)i_a;S solid solution, liquid, and 
vapor. Since the liquid and vapor both 
contain more than 99.9 weight per cent 
sulfur the section is now binary. Below 
862°C it is, of course, not binary but 
represents only a projection onto a phase. 

Stoichiometric (Fe,Ni)S in the section 
is stable below 860°C. Above this 
temperature the solid solution becomes 
metal deficient even in the presence of 
excess (61.4 Fe, 38.6 Ni) alloy. The 
melting relations are similar to those of 
the Fei-zS and aNii-^S solid solutions. 

The maximum melting point is at 
1074°C, where the solidus and liquidus 
curves intersect at about 40.5 weight per 
cent S. Mix crystals of this composition 
are the only ones that melt directly to a 
liquid of the same composition as the 
solid. The corresponding maximum melt- 
ing point in the Fe-S system is at 1192°C 
and about 38.1 weight per cent S, and in 
the Ni-S system at 992°C and about 38.2 
weight per cent S. Thus in the ternary 
system a curve marking maximum melt- 
ing of the (Fe,Ni)i_ x S solid solution 
series is slightly concave toward the 
sulfur corner (see fig. 47). In a T-X plot 
this curve also slopes uniformly without 
a maximum or a minimum. In all 
experiments with 49.1 weight per cent or 
more sulfur a heat effect was also recorded 
at 1028°C. The liquidus curve on the 
sulfur side of the maximum melting point 
recorded for various compositions was 
found to reach 1028°C when there was 


Weight per cent 


Fig. 47. Liquid immiscibility in the Fe-Ni-S 
system is shown in the upper part of the diagram. 
The heavy line extending across the system from 
1192°C on the left to 992°C on the right indicates 
compositions and temperatures of maximum 
melting of the (Fe,Ni)i_ I S solid solution series. 

about 51.0 weight per cent S. In all 
experiments with 51 to 97 weight per cent 
S a single strong peak was recorded at 
1028°C in addition to the heat effects at 
726° and 826°C. 

The liquid immiscibility region in this 
section, therefore, exists above 1028°C 
and over a composition range extending 
from about 51 to more than 97 weight 
per cent S. 

Pentlandite stability relations. This min- 
eral, our most important source of nickel, 
usually occurs in intimate association 
with pyrrhotite, often in oriented inter- 
growths that presumably are produced by 
exsolution. It is found in basic rocks like 
norites and may well be derived from such 
rocks by magmatic segregation. Pent- 
landite, (Fe,Ni) 9 S 8 , has cubic symmetry. 
The literature reports its melting point 
at about 875°C. It is readily synthesized 
in quenching experiments in closed, 
evacuated silica tubes at temperatures 
above 500°C. Below this temperature 
reaction rates are slow and considerable 
time is required to obtain a homogeneous 



product. X-ray diffraction patterns of 
these materials are invariably identical to 
the X-ray diffraction pattern of natural 
pentlandite regardless of the temperature 
of synthesis. On studying synthetic 
pentlandite in polished sections and by 
using oil immersion, however, pronounced 
differences in textures were observed 
between those synthesized at 500° to 
6Q0°C and those synthesized at 700° to 
800°C. The lower-temperature products 
appeared homogeneous when studied by 
means of both X rays and the microscope, 
whereas the higher-temperature materials 
displayed distinct textures due either to 
inversion or to breakdown in the solid 

Since microscopical studies alone could 
not explain the texture variations, differ- 
ential thermal analyses were tried; the 
results are given in figure 48. A few 
milligrams of Lake Toxaway quartz 
served as internal standard. The high-low 

inversion in this material appears at 
573°C both on heating and on cooling. 
On the left side of figure 48 are shown the 
heating curves (bottom) and cooling 
curves (top) recorded for synthetic pent- 
landite of (Fe,Ni) 9 S 8 composition in 
which the Fe:Ni ratio equals 1. A very 
strong thermal effect appears at 610°C on 
heating and at 609°C on cooling. A second 
strong peak was recorded at 862°C on 
heating and at 863°C on cooling. The 
temperature at which this peak occurs 
coincides more or less with the melting- 
point temperature of about 875°C given 
for pentlandite in the literature. Com- 
parison of the two peaks shows that the 
one at 610°C is at least as strong as that 
produced by the melting process. There- 
fore, the lower-temperature effect cannot 
readily be explained as the result of a 
polymorphic inversion but rather indi- 
cates the breakdown of the pentlandite 

573°C ! MV > 

«%7^°r 609 C 

* 6I0°C 
i i 

863*C 57 io c 
,862 P C 

881 °C 

Fig. 48. Differential thermal analysis curves of synthetic pentlandite, heating curve bottom left 
and cooling curve top left; and of natural pentlandite, heating curve bottom right and cooling curve 
top right. Heating and cooling rate was 3°C per second in all experiments. The small peak at 573°C 
is due to inversion in quartz, which was used as internal standard. 



DTA curves on pentlandite from the 
bottom of Frood Mine, Sudbury, are 
shown on the right side of figure 48. The 
heating curve is on the bottom and the 
cooling curve on the top. Lake Toxaway 
quartz was again used as internal 
standard. The heating and cooling rates 
of 3°C/min and all other experimental 
conditions were the same as those for the 
synthetic material. 

The first strong exothermal peak, which 
in the synthetic material appeared at 
610°C on heating, is recorded at 613°C on 
both heating and cooling of the natural 
pentlandite. The second heat effect is 
recorded at 864°C on heating and at 
881°C on cooling. The small disturbances 
recorded between these peaks (see right 
side of fig. 48) may be due to accessory 
minerals in the natural sample. 

To determine whether breakdown ac- 
tually occurs at 610°C, X-ray diffraction 
films were made with a high-temperature 
X-ray camera first at room temperature, 
then at about 600°C, then at about 650°C, 
and finally again at room temperature. 
To avoid oxidation the pentlandite 
specimen was kept in a sealed silica tube 
constructed for this purpose. The first 
exposure gave the pentlandite pattern 
with no other reflections. The second also 
gave the pentlandite pattern but the 
reflections were considerably displaced 
from their positions on the film taken at 
room temperature. The displacements 
indicate a much larger unit-cell size at 
600°C than at room temperature. The 
thermal expansion of pentlandite appears 
significantly larger than that reported for 
any other sulfide. The exact thermal 
expansion is being determined. The films 
made at 650°C contained none of the 
pentlandite reflections. Instead they 
showed all the stronger reflections of 
hexagonal pyrrhotite and all the reflec- 
tions of the high-temperature Ni 3±x S2 
phase described by Kullerud and Yund 
(1962). On cooling to room temperature 
the pattern obtained was again that of 
pentlandite ; no other reflections appeared. 

In the Fe-Ni-S system pentlandite lies 

on a straight line from the (Fe,Ni)i_ x S 
to the Ni 3±a; S2 solid solution. In figure 49, 
which shows the breakdown of pentland- 
ite schematically, (Fe,Ni)i_ x S is on the 
left side. The Ni content of the 
(Fe,Ni)i_ x S mix crystals and the Fe 
content of the Ni 3±x S phase at the 
temperature of the breakdown have not 
yet been accurately established. This 
section is not binary because of the 
variable metal-to-sulfur ratios of the end 
members. Pentlandite and pyrrhotite are 
stable together below 61Q°C. Pentlandite 
and heazlewoodite are stable together 
below about 550°C. Stability depends in 
part on the Ni-to-S ratio: if this ratio is 
high the phase may invert to Ni 3±x S 3 at 

E 600 


1 1 1 

1 1 1 i I I 

hotite + Ni 3+r S2 




pentlandite + Ni 3 » x Sg 







,) glO 20 30 

40 50 60 70 80 90 m; 
(Fe,N,) 9 S 8 Nl 3 

Atomic per cent 

Fig. 49. Schematic illustration of the break- 
down of pentlandite to pyrrhotite + the Ni 3±a: S; 
phase at 610° ± 2°C. 

535°C; if it is low, inversion may take 
place at 524°C (Kullerud and Yund, 
1962). Solid solution of iron in heazle- 
woodite may also affect the temperature 
of inversion significantly. Above the 
temperature of inversion but below 610°C 
pentlandite is stable with the unquench- 
able Ni 3±x S 2 phase. Above 610°C pyrrho- 
tite and the Ni 3±x S 2 phase are stable 
together. The binary Ni 3±x S 2 phase melts 
incongruently at 806°C (Kullerud and 
Yund, 1962) to liquid + aNii_ s S. 

Very slight disturbances are noticed in 
the heating and cooling curves in the 820° 
to 830°C region. They are too small to be 
caused by incongruent melting of Ni 3±x S 2 . 
In DTA experiments on S3mthetic mix- 



tures of FeS and (Fe,Ni) 9 S 8 in the 1:1 
weight per cent ratio the heat effects 
were again recorded at 610° and 862°C. 

It is probable that a considerable 
amount of pyrrhotite is soluble in the 
Ni3 ±x S 2 phase and that the melting 
temperature of that phase increases to 
862°C with increasing pyrrhotite content. 

Many ores containing pentlandite 
formed originally much above 600°C. 
Pentlandite, therefore, is a phase that 
must have formed during the cooling of 
the ore bodies. This new information has 
important bearings on the interpretation 
of mineral assemblages containing pent- 

Bravoite stability relations. Bravoite, 
(Fe,Ni)S 2 , is a typical low-temperature 
mineral. It is commonly found as an 
alteration product of pentlandite, and it 
often forms by alteration of linnaeite. It 
occurs as pore fillings and in cavities in 
many lead-zinc ores; it occurs in certain 
sediments; and it has recently been 
reported (Ramdohr and Kullerud, Year 
Book 60) as a secondary phase in certain 
chondritic meteorites. In hydro thermal- 
type deposits bravoite is one of the 
youngest minerals and often is associated 
with older minerals like pyrite, chalcopy- 
rite, millerite, linnaeite, and polydymite. 
In ores believed to have formed through 
magmatic differentiation of sulfide melts 
bravoite, one of the youngest minerals, is 
associated with older pyrrhotite, chal- 
copyrite, pentlandite, platinum minerals, 
etc. Bravoite in such ores is formed by 
alteration of older minerals through the 
action of water. 

It was important to investigate the 
stability field of bravoite because of its 
wide geological range of occurrence. This 
was first attempted (Clark and Kullerud, 
Year Book 59) by the dry method 
involving the heating of mixtures of iron, 
nickel, and sulfur in silica tubes. But 
bravoite did not form in these experi- 
ments, which owing to slow reaction rates 
could not be performed below 200°C. 
Next, wet chemical methods were at- 
tempted. Bravoite was precipitated at 

room temperature with ammonium poly- 
sulfide from aqueous solutions containing 
weighed amounts of dissolved ferrous 
ammonium sulfate and nickel sulfate. 
These precipitates, which were exceed- 
ingly fine grained, were next heated in 
silica tubes with a slight excess of 
ammonium polysulfide at specified tem- 
peratures and for specified periods of 
time. The products of the heating 
experiments were readily identified in 
X-ray diffraction patterns and, much less 
readily, in polished sections. 

The results of dozens of experiments 
are shown in figure 50. On heating, 


± 300 



8 12 16 20 24 28 32 36 40 44 48 

Time in days 

Fig. 50. The curve shows the rate of break- 
down of precipitated bravoite at various tem- 
peratures. It is practically parallel to the 
horizontal axis after 36 days and indicates that 
bravoite is stable below about 140°C. 

bravoite breaks down to pyrite + vaesite ; 
the rate of breakdown is given by the 
curve. Below the curve bravoite persists, 
and above it, has decomposed. Extrapola- 
tion of this curve to the point where it 
parallels the time axis indicates that 
bravoite is stable below approximately 
140°C. This method is very time- 
consuming, and since the reactions cannot 
be reversed the temperature derived by 
extrapolation of the rate curve may be 
significantly too high. To save time in this 
kind of experimentation and to assure 
equilibrium conditions, a simple method 



was devised by which the two solutions 
could be heated separately to the desired 
temperature and then mixed; it is 
described in a separate section. The first 
experiments were performed at 200°C. 
The solutions were heated separately to 
this temperature and then mixed. The 
immediate reaction taking place on mix- 
ing was strongly exotherm and manifested 
itself by raising the temperature in the 
reaction vessel by about 5°C. The 
temperature gradually decreased to 200°C, 
and the vessel was thereafter kept at that 
temperature for 1 hour. The products 
were pyrite and vaesite, which by X rays 
were found to have identical cell dimen- 
sions and the same compositions as pyrite 
and vaesite synthesized together at 
20Q°C by dry experimentation over a 
period of 550 days. The identities of the 
products obtained by these two methods 
at one and the same temperature are 
encouraging and indicate that equilib- 
rium data may be obtained by the 
mixing-of-solutions method at tempera- 
tures too low for dry synthesis. Additional 
experiments showed that bravoite is 
stable below 137° ± 6°C. The reversi- 




°. '40 


B 120 

g. 100 
£ 80 


py + vs. 

137 ±6 

FeS 2 + bv 

bv + vs. 

FeS 2 10 20 30 40 50 60 70 80 90 N iS 2 
Weight per cent 

Fig. 51. Bravoite is stable below 137° ± 6°C 
as determined in experiments involving mixing 
of liquid at elevated temperatures. Below 137°C 
bravoite is stable with FeS 2 (pyrite or marcasite) 
or with vaesite, depending on the bulk compo- 
sition. Above this temperature pyrite and vaesite 
form a stable mineral assemblage. 

bility of the reaction 2(Fe,Ni)S 2 ^ FeS 2 
+ NiS 2 was demonstrated. In one experi- 
ment performed for this purpose the 
liquids were mixed at 150°C, where 
pyrite + vaesite form. Then the tempera- 
ture was lowered to 130°C and main- 
tained for 72 hours. After this period of 
time, bravoite was detectable in X-ray 
diffraction powder patterns. 

In figure 51 the stability of bravoite is 
shown in relation to pyrite and vaesite. 
Bravoite and FeS 2 (marcasite or pyrite) 
are stable together below 137°C and form 
a common mineral assemblage in nature. 
Bravoite and vaesite are also stable 
together below 137°C. This assemblage 
was previously not known to exist in 
nature, but we have now found it in 
specimens from southeast Missouri. 

The Fe-Mo-S System 
G. Kullerud and Peter R. Buseck 

Minerals in the Fe-Mo-S system are 
pyrite (FeS 2 ) and pyrrhotite (Fei_ x S) 
along the Fe-S join, and molybdenite 
(M0S2), our most important source of 
molybdenum, on the Mo-S join. In 
addition, a phase of approximately Mo 2 S 3 
composition occurs in the synthetic 
system but has not been established as a 

The phase relations between pyrite and 
molybdenite are of immediate interest 
because these two minerals occur together 
in the majority of the ores mined for 
molybdenum. Pyrite is stable to 743°C, 
where it melts incongruently to pyrrho- 
tite + liquid. Pure molybdenite is stable 
to about 1350°C. 

Since pyrite and molybdenite are 
stoichiometric compounds as closely as 
can be determined by our methods, the 
join FeS 2 -MoS 2 is essentially binary even 
in the presence of excess sulfur. This 
sulfur is added to avoid decomposition of 
FeS 2 at temperatures below its stability 
limits through loss of sulfur to the vapor 
phase. Mixtures containing excess sulfur 
were heated at 700°C and lower tempera- 



tures for extended periods of time. Pyrite 
and molybdenite remained stable together 
in all these experiments. Determinations 
of the cell dimensions by means of X-ray 
difTractometer methods of both phases 
before and after heating showed no 
measurable change in either pyrite or 
molybdenite. In polished sections pure 
synthetic pyrite appears identical with 
pyrite heated together with molybdenite, 
and pure synthetic molybdenite appears 
identical with molybdenite heated with 

To determine the solubility of pyrite in 
molybdenite and that of molybdenite in 
pyrite, synthetic FeS 2 and MoS 2 were 
heated together at 724°C for 11 days. 
Subsequent measurements of d 3 n of the 
pyrite with Si internal standard gave 
a = 5.418 ± 0.002 A, which is identical 
with the values given by Swanson, 
Gilfrich, and Ugrinic (1955) and Kullerud 
and Yoder (1959) for pure FeS 2 . Measure- 
ments of dooe of molybdenite after being 
heated with FeS 2 using Si0 2 as internal 
standard gave c = 12.294 A, which is 
identical with the value of c = 12.295 A 
given by Swanson, Gilfrich, and Ugrinic 
for pure MoS 2 . These results indicate that 
very little if any MoS 2 is soluble in FeS 2 
at 724°C, and that very little if any FeS 2 
is soluble in MoS 2 at the same tempera- 
ture. This conclusion is based on the 
assumption that if solid solubility existed 
in either phase it would have measurable 
effects on the lattice dimensions of the 
host materials. 

In DTA experiments on various FeS 2 - 
MoS 2 mixtures, all with excess sulfur, a 
strong thermal effect was recorded at 
726° rt 3°C both on heating and on 
cooling. This is the maximum tempera- 
ture at which pyrite and molybdenite can 
coexist as a mineral pair in the presence 
of vapor. Above this temperature pyrrho- 
tite and molybdenite form the stable 
mineral association. The five phases 
pyrrhotite, pyrite, molybdenite, liquid, 
and vapor are all stable at 726°C, and 
invariant conditions, therefore, exist in 
the ternary system at this temperature. 

This invariant point is situated on the 
FeS 2 -MoS 2 join at about 95 weight per 
cent FeS 2 . This join is binary below the 
726°C invariant temperature. 

Tpie Cu-Ni-S System 

G. Moh and G. Kullerud 

The phase relations in this system have 
been studied at 600°C in evacuated, 
sealed silica tubes. The phases that occur 
are chalcocite (Cu 2 S), digenite (Cu 9 S 5 ), 
and covellite (CuS) along the copper- 
sulfur join; heazlewoodite (Ni 3 S 2 ) and the 
high-temperature Ni 3±a; S 2 phase as well as 
Ni 7 Se, millerite (NiS), aNii_ x S, polydym- 
ite (Ni 3 S 4 ), and vaesite (NiS 2 ) on the 
nickel-sulfur join. At 600°C the only 
stable binary phases are chalcocite, 
digenite, Ni 3±x S 2 , aNi x _ x S, and NiS 2 . 
There are no ternary compounds. The 
limited solid solutions among the stable 
phases and their stability relations are 
shown in figure 52. At 600°C complete 
solid solution exists between digenite and 
chalcocite, which we will refer to as the 
chalcocite solid solution. However, this 
solid solution does not extend very far 
into the ternary system. Experiments 
with mixtures of members of the chalco- 
cite solid solution and Ni 3±x S 2 , Nii_ x S, or 
NiS 2 showed that the ternary solid solu- 
tion extends much less than 0.5 per cent 
toward Ni 3±x S 2 and less than 1 per cent 
toward both NiS and NiS 2 . NiS 2 is a 
stoichiometric compound (Kullerud and 
Yund, 1962) that takes 1.0 per cent Cu 9 S 5 
into solid solution at 600°C. The aNii_ x S 
phase forms solid solution with the 
chalcocite solid solution. This solubility 
is very low in the nickel-deficient part of 
the Nii_ x S solid solution but increases as 
the nickel deficiency decreases and is 
about 1.3 per cent at the point of stoichi- 
ometry. The Ni 3±x S 2 phase that forms 
the most extensive binary solid solution 
of all the compounds in this system also, 
expectedly, forms the largest ternary 
field. It extends about 3.5 per cent toward 
Cu 2 S and 2.5 per cent toward Cu. The 
solubility of sulfur is too small to be 



600 °C 

Cu 9 S, 
Cu 2 S 

3±X S 2 

Weight per cent 

Fig. 52. Phase relations in the Cu-Ni-S system at 600°C. All phases or phase assemblages coexist 
with vapor, and the vapor pressure is that of the system. 

measured in Cu or Ni or in the Cu-Ni solid 
solution series. 

Since all experiments were performed 
in rigid tubes vapor is present in equi- 
librium with all phases or phase assem- 
blages given in figure 52. The univariant 
assemblages are: chalcocite s.s., vaesite, 
liquid, and vapor; chalcocite s.s., vaesite, 
aNii_a;S s.s., vapor; chalcocite s.s., 
aNii-sS s.s., Ni 3±a; S2 s.s., vapor; chalco- 
cite s.s., Ni 3±a; S2 s.s., CuNi alloy, vapor. 
This CuNi alloy contains about 68 per 
cent Ni as determined from tie-line 
intersections and by X rays. Divariant 
regions are digenite, liquid, vapor; chal- 
cocite s.s., vaesite, vapor; chalcocite s.s., 
aNii_ x S s.s., vapor; chalcocite s.s., Ni 3±x S2 
s.s., vapor; chalcocite s.s., CuNi alloy 
(ranging in composition from pure Cu to 
CuNi with about 68 per cent Ni), vapor; 
Ni 3±a; S2 s.s., NiCu alloy (ranging in 
composition from pure Ni to NiCu with 
about 32 per cent Cu), vapor. 

The Fe-Ni-As System 

Peter R. Buseck 

Eleven binary compounds occur in the 
system Fe-Ni-As. Along the nickel-arsenic 
join these include Ni 5 _zAs 2 in both a 
stable (jS) and a metastable (/3 r ) form; a 
low-temperature, possibly metastable, 
phase of approximate composition Ni 2 As 
(Heyding and Calvert, 1957); NinAs 8 , 
corresponding to the mineral maucherite ; 
Nii_zAs, corresponding to the mineral 
niccolite; and NiAs 2 in two polymorphs 
that correspond to the minerals rammels- 
bergite and pararammelsbergite. The 
compounds Fe 2 As, FeAs, and FeAs 2 are 
stable along the Fe-As join; of these com- 
pounds only FeAs 2 has an established 
mineral equivalent, loellingite. Complete 
solid solution exists above 912°C between 
Fe and Ni. Below this temperature the 
solid solution is limited, with stable 


a (bcc) and y (fee) phases and a meta- contains Co. The composition of the 

stable a 2 (bcc) phase occurring. A binary other ternary phase, the " intermediary 

phase, Ni 3 Fe, is stable below 503°C solid solution," is not so well known, 

(Hansen and Anderko, 1958) ; awaruite is primarily because of the difficulty of 

its mineral equivalent. At least two detecting it optically. In reflected light it 

ternary compounds exist in the system, is white, has a high reflectivity, is weakly 

Of these, the (Fe,Ni)As 3 phase is not anisotropic, and is practically indistin- 

established as a mineral species, whereas guishable from synthetic maucherite and 

the other, having an approximate compo- j3Ni 5 _ x As 2 . For this reason the determina- 

sition (Fe,Ni) 2 As, evidently corresponds tion of its solid solution range is based 

to the mineral oregonite. Knowledge of largely on X-ray diffraction studies. 

the phase relations in this system is an Unfortunately, this method is not precise 

important step in our efforts to gain because a phase present in only a few per 

understanding of the conditions prevail- cent cannot be detected. However, the 

ing during formation of the "magmatic (Fe + Ni)/As ratio is about 2, and its 

segregate" type of ore. Moreover, this is solid-solution range extends from at least 

one of the bounding systems of the very Ni/Fe = 5 to Ni/Fe = 1. 

important quaternary system Fe-Ni-As-S, The X-ray pattern of the "intermedi- 

which includes many minerals of common ary solid solution" corresponds closely to 

occurrence both in arsenide and sulfide- that published by Ramdohr and Schmitt 

type deposits, and thus provides a link (1959) for the mineral oregonite, found 

between the two. with awaruite in Josephine County, 

The system was studied at 800°C in Oregon. The composition given by Ram- 
evacuated, sealed silica glass tubes in dohr and Schmitt for oregonite, however, 
which vapor was always present. Owing does not lie within the range of the 
to the very slow reaction rates of the synthetic solid solution field at 800°C, 
arsenides, high temperatures are required but it is known only from X-ray fluores- 
for the attainment of equilibrium within cence. At the low temperature at which 
a reasonable period of time. All the the oregonite presumably formed, the 
results were obtained from experiments ''intermediary solid solution" may extend 
involving heating for at least 1 month, to the oregonite composition. 
During this period the materials were A number of the tie lines for the system 
reground one or more times to facilitate Fe-Ni-As that have been located are 
the reactions. listed below. Those described by Rose- 

Because of the sluggish reaction rates boom (1958) for the higher arsenides are 

and the appreciable solid solution between not included. Tie lines run from a 2 Fe-Ni 

many of the phases some of the tie lines to Fe 2 As, to the "intermediary solid 

have so far been located only approxi- solution," and to j8Ni 5 _ x As 2 . Others ex- 

mately. Much effort was devoted to tend from the "intermediary solid solu- 

verifying the stability and determining tion" to maucherite, j8Ni 5 _ x As 2 , Fe 2 As, 

the composition of the "intermediary" FeAs, to both the Fe-rich and Ni-rich 

or (Fe,Ni) 2 As solid solution. sides of the solvus for the solid solution 

Both ternary phases form extensive series between FeAs and niccolite, and 

solid solution. The field of stability of the toward niccolite solid solution. 

(Fe,Ni)As 3 phase does not extend to There are several univariant regions, 

either of the binary end members. Its Those containing a 2 Fe-Ni also include 

stability range has been investigated by aFe and Fe 2 As, Fe 2 As with "intermediary 

Roseboom (1962) and Pleass and Heyding solid solution," and "intermediary s.s." 

(1962). The compound (Fe,Ni)As 3 is with jSNis-^As-i. Other univariant regions 

analogous to the mineral skutterudite, containing the "intermediary s.s." are 

though apparently skutterudite always those with Fe 2 As and FeAs, jSNi 5 _ a; As 2 



and maucherite, and two phases of the 
FeAs-niccolite solid solution series. 

There are also a number of prominent 
divariant regions such as the one between 
/3Ni 5 _ a; As 2 and 7Fe-Ni and those between 
the "intermediary s.s." and Ni-bearing 
FeAs as well as Fe-bearing niccolite. 

Grains whose compositions lie within 
the two-phase regions between the "inter- 
mediary s.s." and Fe-bearing niccolite or 
Ni-bearing FeAs commonly display a 
distinctive myrmekitic texture. It is 
extremely fine grained and visible only 
under very high magnification; even 
then the phases present cannot be 
optically identified. However, it appears 
that in some samples both host phases 
contain these intergrowths, whether as 
the result of decomposition of a solid 
solution on quenching or as the result of 
a liquid's having been present is not clear. 

Evidence, as yet inconclusive, has been 
found for the existence of a third ternary 
phase situated in the field bounded by the 
uni variant region containing c^Fe-Ni and 
the "intermediary s.s." together with 
Fe 2 As on the one hand and /3Ni 5 _ x As 2 on 
the other. 

At lower temperatures awaruite (Ni 3 Fe) 
appears as a phase. Oregonite, in its only 
known occurrence, is associated with 
awaruite. As the tie lines at 800°C run 
from a 2 Fe-Ni to /3Ni 5 _ x As 2 , a switch in tie 
lines is necessary to establish oregonite 
and awaruite as a stable mineral pair at 
low temperatures. 

The Cu-Fe-S System 

Pyrrhotite-Pyrite-Chalcopyrite Relations 
K. v. Gehlen 8 and G. Kullerud 

The composition of pyrrhotite when 
deposited in equilibrium with pyrite is a 
useful indicator of the temperature con- 
ditions that existed when the assemblage 
formed, provided that the pyrrhotite 
maintained its original iron-to-sulfur ratio 
during the subsequent cooling process. 

8 University of Erlangen-Ntirnberg. 

Most ores that contain these two min- 
erals also contain additional minerals such 
as chalcopyrite, sphalerite, galena, and 
often small amounts of magnetite, some 
of which are known to form measurable 
solid solution with one or both members 
of the pyrrhotite-pyrite pair. The pyrrho- 
tite solid solution is then no longer binary 
but becomes ternary or even more 
complex, and application of the phase 
relations in the strictly binary synthetic 
system Fe-S to such ores becomes 
hazardous. Chalcopyrite is perhaps the 
most common mineral occurring with the 
pyrrhotite-pyrite pair, and pyrrhotite can 
take significant amounts of copper into 
solid solution. Therefore, it was of 
interest to investigate whether pyrrhotite 
compositions, as determined by di 2 
measurements, in ternary pyrrhotite- 
pyrite-chalcopyrite assemblages at con- 
trolled temperatures would coincide with 
the values given by Arnold for binary 
pyrrhotite-pyrite assemblages. Mixtures 
of iron, copper, and sulfur with bulk 
compositions inside the univariant pyr- 
rhotite-pyrite-chalcopyrite field were 
heated in evacuated, sealed silica tubes at 
various temperatures. Measurements of 
the dw2 value of the pyrrhotite synthe- 
sized in equilibrium with pyrite and 
chalcopyrite at 600°C gave 2.0532 ± 
0.0017 A. Pyrrhotite synthesized in equi- 
librium with pyrite in the absence of 
chalcopyrite, at the same temperature, 
gives d w2 = 2.0497 =b 0.0007 A according 
to Arnold (1962). This value is identical, 
within the limit of error of our measure- 
ments, with our results on pyrrhotite 
synthesized in equilibrium with pyrite at 

The difference in d values of pyrrho- 
tites synthesized with and without chal- 
copyrite present is significant to geo- 
logical thermometry. Application of the 
d-versus- composition and T-versus-com- 
position curves by Arnold would only in 
the second case lead to the correct 
temperature estimate of 600°C. The d 
value of pyrrhotite coexisting with chalco- 
pyrite and pyrite would indicate a 


temperature of only about 550°C. A chalcocite and bornite-chalcopyrite, sug- 

positive correction of about 50°C would gests that lamellae may be retained as an 

therefore be required at this temperature, exsolution texture only when the solid 

The magnitude of this correction solution is cooled from above the solvus 

depends on the variation with tempera- in relatively short periods. Complete 

ture of the copper content of the pyrrho- migration to grain boundaries or a mutual 

tite phase and the composition of the boundary texture results if the solid 

chalcopyrite phase. That the required solution is cooled over longer periods, 

correction diminishes with decreasing The common occurrence of exsolution 

temperature is probable but remains to lamellae in ore sulfides has led some 

be shown. recent investigators (e.g., Lyon, 1959) to 

The solubility of copper in pyrrhotite conclude that some ore bodies cooled from 

exceeds 3 weight per cent at 700°C (Yund 600° to 200°C in a matter of minutes or 

and Kullerud, Year Book S 9), and we have hours. Clearly, masses of ore, some 

now determined the solubility at 600°C comprising millions of tons, cannot cool 

to be about 2 weight per cent. by conduction at so rapid a rate. 

Sphalerite commonly coexists with the The present study was initiated to 

pyrrhotite-pyrite mineral pair; however, investigate this paradox and to gain a 

the solubility of ZnS in pyrrhotite is better understanding of exsolution tex- 

negligible even at very high temperatures tures. 

(Kullerud, 1953). For this reason the A study of the literature of metallurgy 

presence of sphalerite should not measur- and solid-state physics (e.g., Geisler, 

ably affect the pyrrhotite-pyrite solvus. 1951; Baker, Brandon, and Nutting, 

A situation very similar to that for 1959) reveals that exsolution lamellae 

sphalerite exists for galena, which is also need not necessarily be formed only by 

a common mineral in pyrrhotite-pyrite rapid cooling. Lamellae are retained as 

ores. the stable exsolution texture if the degree 

On the contrary, the presence of iron of supersaturation is low, in other words, 

oxides with the pyrrhotite-pyrite assem- if the solid solution is initially dilute or 

blage may affect the pyrrhotite compo- the cooling rate is slow, 

sition significantly since pyrrhotite is very The solid solutions involving bornite 

susceptible to oxidation (Kullerud, Year (digenite-bornite, chalcocite-bornite, and 

Book 56). Numerous such ores contain chalcopyrite-bornite) in the system Cu- 

small amounts of magnetite. The effect Fe-S were chosen for experimental study 

of its presence on the pyrrhotite geo- because phase relations are fairly well 

logical thermometer may be significant, understood, because extensive solid solu- 
tion occurs, and because exsolution 

_, , . _ . _ . « , . , textures in this system are commonly 

Exsolution 1 extures and Rates in bolid ■ 

Solutions Involving Bornite Lamellae ' have now been obtained in 

P. R. Brett runs cooled at rates as low as 3°C per day 

for 6 months. As a general rule, the more 

Exsolution textures. Very little is known concentrated the initial solid solution, the 

of the contribution of diffusion and less common are lamellae as the final 

exsolution to the formation of textures exsolution product. Lenses or mutual 

during the cooling of ores. There has been boundary textures are the end products 

little systematic experimental work to of exsolution in the more concentrated 

back up the interpretation of such solid solutions. Lamellae were often 

textures. observed in combination with a mutual 

All previous work on exsolution tex- boundary texture, suggesting that either 

tures in sulfides, mainly on bornite- some of the lamellae did not coalesce to 


form irregular grains or there was a late rigorous theory of kinetics, for a reaction 

stage of formation of lamellae. rate in the solid state is dependent not 

In addition to lamellae and mutual only on such variables as concentration, 

boundary textures, "veining" and "re- temperature, and pressure but also on the 

placement" relations were occasionally rate of nucleation, diffusion, recrystalliza- 

observed. The "veining" textures result tion, etc. Moreover, the difficulties in- 

from the exsolved phase depositing along volved in determining the exact time for 

a continuous series of grain boundaries a reaction to proceed to a certain point 

(fig. 53, pi. 2). The "replacement" are considerable. The composite rate 

textures, in which the exsolved phase cannot be quantitatively considered by 

appears to replace the host phase, were separate treatment of each process, for 

observed in chalcopyrite exsolved from the effect of individual variables cannot 

bornite (fig. 54, pi. 2). Eutectoid textures be isolated. Nevertheless, such studies 

observed as products of exsolution in the can at least ascertain whether reequili- 

bornite-digenite and bornite-chalcocite bration in the system studied can be 

pairs are similar to those commonly expected during the slow cooling of 

present in these minerals in ores. natural mineral assemblages. The time 

The results of this study indicate that taken for a reaction like exsolution to 
few interpretations from textural evi- proceed to a certain point can be deter- 
dence may be made on the thermal mined only by measurement of the 
history of minerals that form solid change in a composition-dependent prop- 
solution pairs. Exsolution lamellae can erty such as cell edge, hardness, or 
indicate extremely rapid cooling (which magnetic susceptibility, 
is not to be expected in mineral deposits) The rates of exsolution in the solid 
or cooling of a solid solution of a relatively solution field bornite-digenite-chalcopy- 
low initial concentration, in which the rite in the system Cu-Fe-S were chosen 
cooling rate was such that the degree of for the present study. This system was 
supersaturation was never high. Veining, selected because phase relations are 
pseudo-replacement, and mutual bound- relatively well known and because there 
ary textures can occur as products of have been suggestions in the past that 
exsolution. The utmost caution must be solvi in this system, when determined, 
taken in the interpretation of the textural would be useful for geologic thermometry, 
relations between any minerals that may The change in composition during 
form solid solution pairs. exsolution could be ascertained by meas- 

Rates of exsolution. It has long been uring the a cell edge of exsolved bornite. 

suspected that many ore minerals and This cell edge varies markedly with the 

mineral assemblages remain unchanged Cu/(Cu + Fe) ratio. Solid solutions of 

during the cooling period, in this way various compositions about bornite along 

retaining the evidence required to infer the bornite-digenite and bornite-chal- 

the conditions during ore deposition. It copyrite joins were prepared at 700°C. 

is on the supposition that many systems They were then annealed or cooled to 

do not equilibrate with falling tempera- 400°, 300°, 200°, and 50°C in various 

ture that the principles of geothermom- times. If the cell edge of the bornite was 

etry are based. constant for a given temperature below 

An understanding of the extent of the solvus regardless of its original 

equilibration of mineral systems can be composition, equilibrium was assumed to 

obtained only by the study of reaction have been attained. 

rates at different temperatures and All runs above 50°C were held at 

pressures. Unfortunately, data on rates of temperature for 2 months or more, and 

solid-state reactions such as exsolution equilibrium was attained in all. When the 

cannot be considered in terms of a solid solutions were cooled to 50°C in 3 


months, equilibrium was also attained; rapidity in nature. Unfortunately, these 

the same was observed when they were systems are useless as potential geother- 

cooled from 600° to 50°C in 1 hour. mometers, because the exsolved phase 

When solid solutions of chalcopyrite in consistently migrates out of the host 

bornite were annealed at 100°C for 2 mineral, making reconstitution of the 

weeks, equilibrium was attained, whereas original solid solution impossible. The 

digenite in bornite, held at 50°C for 3 minerals with greatest potential as geo- 

months, approached equilibrium within thermometers are therefore the most 

approximately 3 weight per cent. When refractory ones (such as arsenides, oxides, 

solid solutions of either chalcopyrite or pyrite, and sphalerite) and the ones most 

digenite in bornite were cooled from above difficult to study, 
the solvus to 50°C in 7.5 minutes, 

equilibrium was closely approached, the Chalcocite-Chalcopyrite Assemblages 

variation in cell edge being 10.944 to P R Brett 
10.950 A (±0.005 A). 

All the runs in which disequilibrium In the course of the investigation of 

was most pronounced were those in which bornite-chalcopyrite exsolution textures, 

the original solid solution was dilute. In bornites with maximum sulfur content 

many runs the dilute solid solution were prepared along the bornite-chal- 

exhibited no exsolution at all after being copyrite join at 700°C (see Year Book 59, 

cooled or annealed in spite of the fact that figs. 43-45). Chalcopyrite exsolved from 

more concentrated solid solutions ex- the bornite on annealing or cooling to 

solved to equilibrium. Doubtless a nucle- 50°C ; another phase was seen in amounts 

ation problem is involved; the more insufficient for determination by X-ray 

concentrated solids have a greater degree diffraction. By reason of the geometry of 

of supersaturation at the annealing the Cu-Fe-S phase diagram, the appear- 

temperature, hence have a greater ten- ance of the phase, and the composition of 

dency to nucleate. the runs, this phase is probably chalcocite. 

It may be concluded from the study The tie line chalcopyrite-chalcocite is 

that rates of exsolution (and indirectly of possible only because chalcopyrite always 

solid diffusion) are rapid in this part of contains less sulfur than is indicated by 

the Cu-Fe-S system. Complete equilibra- its stoichiometric formula, so that chal- 

tion would be expected in nature in times cocite, bornite, and chalcopyrite are not 

of the order of 1 hour. exactly collinear. 

To check the extent of equilibration of Chalcopyrite formed at 700°C is more 

natural bornites the a cell edges of deficient in sulfur than chalcopyrite 

thirteen bornites from nine different formed at lower temperatures (Yund and 

localities and environments were meas- Kullerud, Year Book 59). Accordingly, 

ured by means of the X-ray diffractom- chalcopyrite exsolving from a chalcopy- 

eter. Except for the anomalous red bed rite- bornite solid solution must become 

bornites mentioned elsewhere in this more sulfur rich as cooling proceeds. The 

report, all cell edges correspond to those bornite in equilibrium with the exsolving 

of stoichiometric bornite (10.950 =fc 0.005 chalcopyrite must therefore become in- 

A). This is further evidence that equi- creasingly poor in sulfur (and iron) as 

librium is complete in nature. exsolution proceeds, and must form a 

The fact that equilibration can occur chalcocite-bornite solid solution that 
in so short a time casts grave doubts on breaks down at low temperature, 
the use of this part of the system for Chalcocite-chalcopyrite has often been 
geothermometry. The majority of sulfide observed as a natural assemblage, par- 
systems, like the minerals in the system ticularly under supergene conditions. In 
Cu-Fe-S, also equilibrate with great view of the collinearity mentioned above, 



it has invariably been taken to be a 
disequilibrium assemblage. The present 
work suggests that this need not neces- 
sarily be so. 

Studies along the join bornite-digenite 
previously reported by Kullerud were 
continued down to 50°C. The join 
bornite-digenite was found to exist at 
least to 50°C. Therefore, a pyrite- 
chalcocite join is impossible below the 
bornite-digenite immiscibility gap. Above 
this solvus bornite, digenite, and chal- 
cocite form a complete solid solution field 
(Yund and Kullerud, Year Book 59), 
again prohibiting a chalcocite-pyrite join. 

The assemblage pyrite-chalcocite has 
commonly been reported in copper-iron 

were not formed simultaneously. The 
chalcocite was probably formed at low 
temperatures where reaction rates are 

The persistence of pyrite with chalco- 
cite in both hypogene and supergene ores 
can only be ascribed to the lack of 
reactivity of pyrite. In view of this 
inertness, all sulfide assemblages involv- 
ing pyrite cannot definitely be regarded 
as equilibrium assemblages until more 
conclusive evidence is available. 

The relationships between composi- 
tions and the a cell edge of bornite solid 
solutions were determined by measuring 
the 20 values of the (440) reflection on the 
X-ray diffractometer. Provided that the 





"i 1 - I " 

-j | 



in «or> 


i l _ 1 

I 1 


Cu 5 FeS 4 








>-CuFeS 2 _ x 

Moi per cent 

Fig. 55. Variation in the a cell edge of bornites with maximum sulfur content in the bornite- 
chalcopyrite solid solution field at 700°C. 

sulfide deposits. Although much of this 
chalcocite may be misidentified digenite, 
there are some well authenticated exam- 
ples of chalcocite-pyrite. The assemblage 
must be due to the breakdown of a 
chalcocite-digenite-bornite solid solution 
that formed in grain boundary equilib- 
rium with pyrite. A chalcocite-pyrite 
hypogene assemblage should therefore 
always be accompanied by bornite and/or 
digenite, as at Butte. 

Assuming equilibrium during deposi- 
tion, a pyrite-chalcocite assemblage with- 
out accompanying digenite or bornite is 
thus an indication that the two minerals 

bornite solid solution is of maximum 
sulfur content, the relation between a and 
Cu/(Cu + Fe) is linear. This was demon- 
strated by Kullerud in his studies on the 
bornite-digenite join {Year Book 59). The 
relationship has now been verified for 
bornite containing chalcopyrite in solid 
solution (fig. 55). However, the present 
investigation revealed that a decrease in 
sulfur content of less than 0.5 weight per 
cent can increase the cell edge by as much 
as 0.015 A. The cell edge of sulfur- 
deficient bornite is dependent on (1) the 
Cu/(Cu + Fe) ratio, (2) the sulfur 
content, (3) the quenching procedure, 


(4) the type of grinding used in prepa- studied by means of an electron micro- 
ration of the diffractometer mount. For scope (at the National Bureau of Stand- 
this reason caution should be exercised in ards) , but no submicroscopic phase of the 
applying data published on bornite cell type suggested by Takeuchi and Nambu 
edges to natural bornite. (1956) was noted. 

In general, bornite cannot be chem- 

Heating Experiments on Natural Bormtes ically anaIyzed accurately because small 

P. R. Brett inclusions of other sulfides cannot be 

Bornite can take some chalcopyrite eliminated. A sample of the "anomalous" 

into solid solution, but this is totally bornite was examined on the electron 

exsolved on cooling, even if the cooling probe at the U. S. Geological Survey; this 

time is only some few minutes (Brett, technique gave analyses with such a large 

this report). When heated to tempera- standard deviation that the results could 

tures below 400°C, however, certain not be applied in the present study, 

natural bornites exsolve up to 25 volume The possibility that oxidation could be 

per cent chalcopyrite (Wandke, 1926; responsible for the formation of chalco- 

Takeuchi and Nambu, 1956; Prouvost, pyrite in some heated bornites seemed 

1960; McCauley, 1961). This situation is worth examining. Two samples of Bristol 

anomalous in view of the slow cooling bornite were finely ground and exposed 

that presumably occurs in nature. to the air for 5 days and 4 months, 

Bornite from various localities was respectively. There was no detectable 

heated in evacuated silica glass tubes at change in weight on oxidation. Bornite 

temperatures ranging from 75° to 600°C. from Bristol was chosen because of its 

Bornites from Moonta, South Australia; purity and because it did not exsolve 

Messina, South Africa; Bristol, Connec- chalcopyrite on heating. Approximately 

ticut; and Magma, Arizona, remained 2 volume per cent chalcopyrite was found 

unchanged, but those from Similkaween, as irregular blebs when the oxidized 

B. C, Beaverdell, B. C, and red bed bornite was heated at 270°C for 1 hour, 

copper localities in Utah exsolved chalco- After 1 hour at 300°C synthetic stoichio- 

pyrite up to 25 volume per cent. The metric bornite oxidized in the same way 

Utah bornites exsolved chalcopyrite after contained small amounts (less than 1 per 

only 10 minutes at 400°C, but no exsolu- cent) of a very fine-grained second phase, 

tion occurred even after 10 days of possibly chalcopyrite. 

heating at temperatures lower than 75°C. The cell sizes of all bornites were 

The exsolved phase was identified as measured both before and after heating, 

chalcopyrite both optically and by its with results shown in table 21. It is 

three principal X-ray diffraction peaks, apparent from table 21 that the normal 

Anomalous unheated bornite was bornites have a close to that of stoichi- 

TABLE 21. Cell Edge of Bornite from Various Localities before and after Heating 




before Heating, 
±0.005 A 

after Heating, 
±0.005 A 

Magma, Ariz. 



Bristol, Conn. 



Messina, South Africa 



Red bed bornite,* Utah 



Red bed bornite,* Utah, 




Oxidized Bristol, Conn., 





Synthetic Cu 5 FeS 4 



* Exsolves chalcopyrite on heating. 



ometric bornite, whereas bornites that 
exsolve chalcopyrite may have a con- 
siderably smaller cell edge before heating 
and on heating revert to a cell size similar 
to that of stoichiometric bornite. The 
small cell cannot be ascribed to oxidation, 
because the cell edge of oxidized Bristol 
bornite before heating is greater than that 
of the unoxidized bornite. 

When heated at 700°C and quenched 
the Utah bornite contained no exsolved 
chalcopyrite, but the cell size had reverted 
to that of stoichiometric bornite. The low 
value of a for the anomalous bornite 
evidently cannot be ascribed to a high 
iron content. It has been mentioned 
elsewhere in this report that the lower the 
sulfur content of a bornite the greater the 
cell edge. Possibly the anomalous bornites 
contain more sulfur than stoichiometric 
bornite. Heating at 700°C would probably 
expel the excess. At any rate, this is a 
reasonable explanation of the behavior of 
the Utah bornite. 

It is suggested that at temperatures 
between 75° and 400°C these sulfur-rich 
bornites break down to chalcopyrite, 
stoichiometric bornite, and chalcocite. 
The chalcocite compensates for the 
increased copper content of the bornite 
caused by the exsolution of chalcopyrite. 
A phase resembling chalcocite or digenite 
has been observed with chalcopyrite in 
heating experiments on anomalous born- 
ites by Greig (personal communication, 
1961) and Prouvost (personal communi- 
cation, 1962). 

Yund (personal communication, 1962) 
reports that synthetic sulfur-rich bornites 
with Cu/Fe ratio equal to or less than the 
stoichiometric ratio exsolve chalcopyrite 
when heated. This is additional evidence 
that the anomalous bornites are anom- 
alous not because they are iron rich or 
contain oxygen but because they are rich 
in sulfur. The conclusion is by no means 
proved, however. 

There is good evidence that the 
anomalous bornites never attained the 
temperature of approximately 75°C dur- 
ing their formation or later, as it is 

possible to cause chalcopyrite to exsolve 
from them at this temperature. 

Method for Mixing Liquids at 
Controlled Temperatures 

G. Kullerud 

In the past, studies of mineral assem- 
blages in aqueous solutions have ordi- 
narily been conducted by mixing the 
liquids at 25°C then slowly heating the 
mixture to a specified elevated tempera- 
ture at which it was kept for desired 
periods of time. In this procedure 
precipitation often takes place as soon as 
the solutions are mixed, and the phases 
formed at room temperature or during 
the heating period may persist metastably 
for considerable lengths of time. A 
disadvantage of the method is that 
equilibrium cannot be proved to have 
existed in any one experiment. The 
results of many earlier studies on the 
pyrite-marcasite relations in which this 
method was used are probably examples 
of the shortcomings of the procedure. 

Progress in our studies of dry systems 
is commonly hampered by slow reaction 
rates. At low temperatures these rates are 
usually so slow that dry experimentation 
is out of the question because of the time 
involved in obtaining equilibrium. Since 
many minerals in nature form at low 
temperatures and are often stable only in 
the region not available to dry synthesis, 
it is desirable that we develop other 
methods by which equilibrium can be 
obtained in a short time. 

Bravoite is an example of a low- 
temperature mineral that we could not 
synthesize by the dry method. Phase 
equilibrium in this part of the Fe-Ni-S 
system could be obtained only down to 
200°C, and this accomplishment required 
13 months. 

The new method allows separate 
heating to a preassigned temperature of 
two or if necessary more liquids. One 
liquid is sealed into a long evacuated 
Pyrex tube; the other is poured directly 
into a cylindrical Teflon container that 


has a pressure-tight closure. The Pyrex The tungsten-copper deposits at Tern 

tube is then also inserted into the Teflon Piute, Lincoln County, Nevada, occur in 

tube, and the pressure seal is closed. This a skarn aureole surrounding a small 

unit, which has a thermocouple well granodiorite stock. One of the mines, the 

similar to that of cold seal and Tuttle Free Tunnel, was studied in detail because 

bombs, is next placed in a horizontal of its rather complex and varied ore 

preheated furnace and heated to the assemblages. All the metallic minerals 

temperature of the experiment. Teflon occur in a diopside, andradite skarn, 

softens on heating but readily withstands which separates the barren intrusion from 

the internal pressure to 200°C under unmineralized limestone and, locally, 

these conditions. When the temperature hornfels. 

of the experiment has been reached the The metallic minerals occur as dissem- 

liquids are mixed by exerting pressure on inations or small lenses, veining being 

the Teflon tube, by means of a specially extremely minor. With the exception of 

constructed pair of pliers with jaws, to pyrite, chalcopyrite, and scheelite, all of 

the point where the Pyrex tube shatters, which occur throughout the aureole, the 

The reaction that immediately takes minerals are roughly thermally zoned. In 

place is recorded by the thermocouple, the "inner," formerly hotter, portions of 

which shows a rapid increase in tempera- the aureole, in approximately paragenetic 

ture of as much as 5°C. Thereafter the sequence are molybdenite, pyrrhotite, 

temperature drops back over the next 10 magnetite, and marcasite. Near the 

minutes or so to that recorded before the limestone in the "outer," formerly cooler, 

liquids were mixed. The Teflon tubes are parts, are sphalerite, galena, galenobis- 

kept in the furnace at temperature from mutite, cosalite, and native bismuth. 

1 to 100 hours, depending on the tern- Except for sphalerite these minerals are 

perature of the experiment after mixing sparse. 

of the liquids, to produce well crystallized Pyrrhotite was one of the first metallic 

materials. The products are filtered, minerals to form, and, as such, it was 

washed, and studied by means of X-ray deposited during the earliest, hottest 

diffraction patterns and in polished stages of the mineralization period. It is 

sections. most prominent along the granodiorite- 

skarn contact but occurs in decreasing 

Pyrrhotite from Tem Piute, Nevada amounts farther from the granodiorite. 

The pyrrhotite is commonly associated 
with and generally surrounds eunearal 

The composition of hexagonal pyrrho- crystals of pyrite, thereby indicating its 

tite when in equilibrium with pyrite has later origin. In an attempt to determine 

found considerable use as a geological its temperatures of formation all available 

thermometer. Efforts to apply the ther- pyrrhotite was sampled and examined in 

mometer to pyrrhotite-pyrite assem- the laboratory. 

blages from the Tem Piute district Arnold (1962) demonstrated that the 
encountered difficulties because at this di 02 spacing of hexagonal pyrrhotite is a 
locality the pyrrhotite is monoclinic. function of its composition. The compo- 
Arnold and Reichen (1962) suggested that sition is dependent on the temperature of 
the thermometer may be valid even for formation provided that the pyrrhotite 
such assemblages and that the compo- formed in equilibrium with pyrite and 
sition of monoclinic pyrrhotite can be that it did not reequilibrate with de- 
determined by the standard X-ray creasing temperatures. Most of the Tem 
method if the specimen is first inverted Piute pyrrhotite was sampled within 1 
to the hexagonal form by heating in mm of pyrite, and all such samples have 
vacuo. similar dio 2 values. It is therefore assumed 



that these pyrrhotites formed in equi- 
librium with pyrite. Pyrrhotites that did 
not form close to pyrite have different 
compositions from those in contact with 
it. This would presumably not be so had 
all the pyrrhotite reequilibrated as tem- 
peratures fell during cooling. 

In the inversion of monoclinic pyrrho- 
tite to the hexagonal form the time and 
temperature allowed for annealing are 
critical ; with too long an annealing period 
or too high a temperature the pyrrhotite 
reequilibrates, and with too short an 
annealing time or too low a temperature 
it does not invert. To determine the 
optimum time and temperature for 

temperature of the hexagonal-monoclinic 
transition lies below 260°C, but at the 
same time the reaction is too slow for 
quick annealing at temperatures below 
300°C. As the samples inverted rapidly 
but did not reequilibrate in 0.1 hour at 
346°C the other Tern Piute specimens 
were annealed under these conditions. 

Several pyrrhotites were sampled from 
specimens containing no pyrite. In speci- 
mens that contained appreciable pyrite 
the pyrrhotite was extracted with a 
dentist's drill kept in contact with pyrite 
at all times so that no pyrrhotite was 
collected farther than 1 mm from pyrite. 

Table 22 lists the results of X-ray 

TABLE 22. Average 20 (1O2) and Corresponding d ( i 02 ) of Tern Piute Pyrrhotite 







44.03 6 


2.0563 ±0.0002 


43.96 6 


2.0594 ±0.0004 


44.00 2 


2.0578 ±0.0002 




2.0572 ±0.0002 




2.0601 ±0.0002 




2.0623 ±0.0005 

Pyrite absent 




2.0636 ±0.0003 

Pyrite absent 




2.0574 ±0.0002 


43.93 9 


2.0606 ±0.0004 

Pyrite absent 


43.88 2 


2.0631 ±0.0002 

Pyrite absent 




2.0601 ±0.0002 

* Standard error of eight or more successive oscillations. 

annealing, natural pyrrhotite was finely 
ground under acetone and concentrated 
magnetically; replicate runs prepared 
from this material were heated in 
evacuated silica glass tubes. 

Runs were annealed for 0.1 hour at 
different temperatures. At 700°C and 
555°C the pyrrhotite-pyrite reaction is 
sufficiently rapid for the samples to have 
reequilibrated. Two runs heated at 346° 
and one at 455°C inverted but did not 
have time to reequilibrate. They indicate 
the same composition within the limits 
of error of the method. A sample heated 
at 300°C did not invert in 0.1 hour but 
did in 1}^ months. Likewise, one at 
260°C did not invert within \ x /i months 
but did within 1 year. Clearly the 

measurements on a number of Tern Piute 
pyrrhotites, all of which were originally 
monoclinic. Numbers 3 and 4 are from 
the same sample; they provide almost 
identical results. Likewise, all the pyr- 
rhotite samples that were adjacent to 
pyrite have very similar di 2 values and, 
had they been hexagonal when collected, 
would indicate temperatures between 
455° and 510°C. Those that were not in 
contact with pyrite have consistently 
larger dw 2 values and, had they also been 
originally hexagonal, would correspond to 
minimum temperatures between 390° and 
450°C. Although very reasonable for 
contact metasomatic deposits such as 
Tern Piute, these temperatures must be 
regarded as tentative. At present it is not 


clear that the relations between hexagonal material yield sound estimates of tem- 

and monoclinic pyrrhotite are such that peratures of formation, 
measurements on inverted monoclinic 

P. Ramdohr 9 and G. Kullerud 

During this past year more than a structure and containing Fe-C-S was 

hundred stony meteorites have been observed in 10 per cent of the meteorites, 

studied in polished sections in addition to Daubreelite, FeCr 2 S 4 , is also present in 

those described in last year's report. The about 10 per cent of the specimens, 

following opaque and semiopaque min- Sphalerite, ZnS, occurs in trace amounts 

erals have been identified: Minerals only. Chalcopyrite, CuFeS 2 , was observed 

containing elemental iron include a iron in a few meteorites, and pyrite, FeS 2 , was 

(kamacite) with variable Ni content, identified only once. 

Fe-Ni solid solutions (taenite) with the Besides these minerals a number of 

structure of 7 iron, and intergrowths of new ones were observed in small amounts 

the a and 7 phases, plessite. Cohenite and mostly in single meteorites. These 

(Fe 3 C) occurs only in a few stony phases are referred to by the letters A 

meteorites and in small amounts. Schrei- through L. For most of them the compo- 

bersite (Fe 3 P) is widely distributed in sitions are partly or completely unknown 

small amounts. A new mineral, which by although their major constituents can 

synthesis was found to have the compo- often be deduced from the mineral 

sition (Ni , Fe) 2 S and which we refer to as assemblages with which they are associ- 

the Henderson phase, was observed in ated. Mineral A is strongly anisotropic 

three meteorites. Graphite (C) occurs in and has a dark yellow-green color. It 

about one- tenth of the specimens. Native almost invariably occurs as lenses or 

copper (Cu) is commonly observed, but lamellae in daubreelite and only rarely is 

in trace amounts. Native gold was found independent of troilite. Its optical 

observed in only one specimen. Troilite properties indicate that it has a pseudo- 

(FeS) is present in all specimens examined hexagonal orthorhombic symmetry, and 

and is frequently the most abundant it may be a transformation product of 

opaque mineral. Chalcopyrrhotite, (Fe , daubreelite. Mineral B occurs inter- 

Cu,Ni,Zn)S, a cubic high-temperature layered with mineral A and appears to 

solid solution, was observed in about have formed from it, not directly from 

one-third of the specimens. Valleriite daubreelite, with which it is also closely 

occurs as a disintegration product of associated. This mineral may be a 

chalcopyrrhotite and as an exsolution terrestrial alteration product, although 

product of pentlandite. Pentlandite, the neighboring minerals, some of which 

(Fe,Ni) 9 S 8 , is present in about one-fourth are very susceptible to weathering, show 

of the meteorites examined. Oldhamite, no sign of alteration. 

(Ca,Fe,Mn)S, is limited to meteorites Mineral C is olive-brown, weakly 

that are highly reduced or that have a reflecting, and apparently isotropic. It is 

high sulfur content. A new (Fe,Mg,Mn, commonly, but not always, associated 

Ca)S phase similar to oldhamite but with with daubreelite. Mineral D is colorless 

much higher reflectivity is rather com- and transparent with high refractive 

mon. Alabandite, MnS, was not observed, index. It replaces ilmenite and chromite 

A new mineral with a hexagonal layer and is always associated with chromite. 

Mineral E is dark brown and occurs with 

9 University of Heidelberg. troilite. It is relatively soft, is isotropic, 



and shows traces of internal reflections. 
It contains exsolution bodies of troilite. 
Mineral F is white and resembles certain 
terrestrial arsenides and sulfarsenides 
such as loellingite and arsenopyrite. This 
mineral is isotropic and has good cleavage 
parallel to (111). It probably contains 
arsenic. Mineral G, a light blue mineral, 
occurs in association with troilite. It is 
relatively soft, is isotropic, and shows 
traces of internal reflections. It contains 
troilite exsolution bodies and is pre- 
sumably a sulfide, but is optically differ- 
ent from any known sulfide. Mineral H 
has a yellow-gray color and almost 
metallic characteristics. It is anisotropic 
with greater reflectivity than chalco- 
pyrite but with paler colors. Its prop- 
erties differ from those of any previously 
described mineral. Mineral I is colorless; 
it is of the spinel type and often contains 
ilmenite exsolution lamellae. It is trans- 
parent, its refractive index is in the range 
1.8-1.9, and its hardness is greater than 
that of olivine but lower than that of pure 
Mg-Al spinel. This mineral was found by 
synthesis to have the composition 
Mg2Ti0 4 . Mineral K has a very dark gray 
color and is definitely isotropic. It is 
sometimes partly rimmed by troilite, and 
its optical properties indicate that it is a 
sulfide. It may possibly be a member of 
the (Fe,Cu,Zn)S mineral group. Mineral 
L is strongly pleochroic and is commonly 
intergrown with mineral A. This inter- 
growth indicates that mineral L may be 
hexagonal. Its hardness is moderate and 
similar to that of troilite. 

In contrast to the numerous sulfides 
observed in stony meteorites the common 
oxides are limited to chromite, magnetite, 
and ilmenite. 

With few exceptions the stony meteor- 
ites are more uniform in their silicate 
mineralogy than terrestrial rocks. The 
major silicates are olivine and pyroxene 
(usually orthopyroxene) , with minor 
amounts of plagioclase. Glass is fre- 
quently present in small amounts ; quartz 
and tridymite are rare. The individual 
meteorites commonly display complex 

mineralogical relationships. A chondrule 
or even a zone within a chondrule may 
represent local equilibrium. Sometimes 
reactive gases apparently brought about 
changes in mineralogy. Evidently an 
increasing degree of reduction was accom- 
panied by increasing temperature, result- 
ing in changes of both mineralogical 
composition and grain size and in 
elimination of brecciation. Complex ge- 
netic histories are displayed in many 
meteorites through spontaneous melting 
processes resulting in droplets of nickel- 
iron, sulfides, and probably glass. In 
many specimens the mineralogy indicates 
that sulfur has been introduced in some 
form. In such meteorites FeS is observed 
to replace Fe, sometimes even taenite. 
Since nickel is less reactive than iron with 
respect to sulfur, the introduction of 
sulfur leads to relative enrichment of 
nickel with a consequent decrease in the 
amount of kamacite. Continued intro- 
duction of sulfur leads to complete dis- 
appearance of the metal phase and often 
produces pentlandite. Further complica- 
tions include the presence of reactive 
hydrocarbons in many meteorites. 

Noteworthy structural and textural 
phenomena are: fusion on "dislocations," 
manifested by the occurrence of droplets 
of fused troilite and iron in varying 
amounts in veinlets ; spontaneous melting, 
resulting in patches of glass with fused 
droplets of sulfide or iron in the interior 
of many meteorites, not related to the 
fusion crust, to heat developed on impact, 
or to brecciation or sintering and, 
therefore, distinctly different from fusion 
on "dislocations"; exsolution, which is 
common, as for instance ilmenite from 
chromite, chalcopyrite from troilite, mag- 
netite from olivine ; mechanical distortion 
and recrystallization, evident in very 
many meteorites; terrestrial weathering 
effects, observed in numerous stony 
meteorites, the products of which, such 
as magnetite, can sometimes be mistaken 
for primary components. 

The new minerals discovered in stony 
meteorites occur in amounts too small to 


permit standard chemical analysis and tialities in this field. The new phases can 

almost invariably also in too small also be synthesized if the major constitu- 

amounts to allow investigation by X-ray ents can be surmised from the mineral 

powder diffraction methods. The electron paragenesis, and our efforts in this 

probe, however, shows promising poten- direction are increasingly successful. 


S. P. Clark, Jr. 

During the past several years, work has 

been done in the systems Fe-Ni-S and The System Fe-Ni-S 
Fe-Ni-P for the primary purpose of 

finding the compositions of troilite, As was stated in last year's report ( Year 

(Fe,Ni)S, or schreibersite, (Fe,Ni) 3 P, in Book 60, p. 184), the amount of nickel in 

equilibrium with both kamacite (a alloy) the troilite in equilibrium with both 

and taenite (7 alloy) at various tempera- kamacite and taenite is small. Since the 

tures and low pressures. Knowledge of 7 structure in the alloy cannot be 

these compositions as a function of quenched in the range of temperatures 

temperature provides an indication of the over which investigations have been 

temperature of formation of iron meteor- made, the composition of the troilite 

ites which supplements the one provided cannot be closely determined by the 

by the Fe:Ni ratios of the two alloy standard methods of phase equilibria. If 

phases. Disagreement between these ther- the charge contains the metastable a 2 

mometers might indicate a lack of phase, which forms from the 7 alloy on 

equilibrium among the phases present, quenching, an upper limit to the possible 

This, coupled with textural observations, nickel content of the troilite is found by 

might give important information about projecting the line connecting the bulk 

the cooling histories of the meteorites, composition of the charge with the 

Alternatively, such a discrepancy might composition of the a alloy to the FeS-NiS 

result because the iron meteorites were join. The composition of the a alloy is 

formed at high pressure, which would be taken from the known system Fe-Ni, 

interesting to know. assuming it to be unaffected by sulfur. 

Besides the work on schreibersite, This is plausible because sulfur is prac- 
X-ray studies of the higher phosphides of tically insoluble in nickel and raises the 
iron, Fe 2 P and FeP, have been made, a-7 transition in iron by only 3°C. 
They were stimulated by the discovery The sensitivity of the method depends 
by Chao, Adler, Dwornik, and Littler on the limiting amount of the a 2 phase 
(1962) that metallic spherules in some that can be detected. The distinction 
tektites consist of kamacite plus a second between a and a 2 is based on the sharp- 
phase, which they tentatively identified ness of the X-ray reflections in the back- 
as a phosphide. Under the microscope reflection region. For the a phase, the 
this other phase is highly anisotropic, K Q1 and K« 2 reflections of (220) are easily 
which is inconsistent with the optical and sharply resolved; for a 2 they are 
properties of ordinary schreibersite. The blurred into a single diffuse reflection, 
studies of the higher phosphides were Since the cell dimensions of the two 
made in order to help in the identification phases are nearly the same, it is difficult 
of this phase. to detect even moderate amounts of one 



in the presence of the other. The X-ray 
reflections are superimposed. 

Runs that limit the possible compo- 
sitions of troilite in equilibrium with 
kamacite and taenite at 800° and 700°C 
and low pressure are shown in table 23. 
The run at 800°C limits the NiS content 
of the troilite at that temperature to 
0-0.3 weight per cent. Similarly at 700°C 
the amount of NiS present in the troilite 

error of less than 0.01 the ratio Ni/(Fe + 
Ni) is the same in the schreibersite as in 
the 7 alloy with which it is in equilibrium. 
This is true for ratios up to at least 0.25 
and at temperatures of both 700° and 
800°C. It is somewhat surprising in view 
of the strong segregation of nickel in the 
metal phase in alloy-sulfide and alloy- 
silicate systems of iron and nickel. A 
possible interpretation is that the metal- 

TABLE 23. 

Data Fixing the Compositions of Troilite in Equilibrium with 
Kamacite and Taenite 

T, °C 

Bulk Composition of Run, 
Fe Ni 

wt. % 



Duration of 
Run, days 






a 2 


must lie between and 0.5 weight per 
cent. These quantities of nickel are so low 
that troilite does not appear to be suitable 
for use as a thermometer. The accuracy 
with which the desired equilibrium com- 
positions can be determined in the 
laboratory is too low for this purpose, 
although the picture could possibly be 
changed by use of the electron probe 

The System Fe-Ni-P 

The nickel content of the schreibersite 
in equilibrium with kamacite and taenite 
is much larger than that of the com- 
parable troilite, and it changes demon- 
strably with temperature. At 800°C the 
Ni/(Fe + Ni) ratio of this schreibersite 
is between 0.065 and 0.10, and at 700°C 
it lies between 0.125 and 0.15. Further 
work is required to fix these compositions 
more closely. Present results are con- 
sistent with the assumption that the 
phase diagram of the Fe-Ni system is 
significantly affected by phosphorus. This 
is known to be so for iron; the system 
Fe-P is of the "Y-loop" type. 

An interesting result of these investi- 
gations is that within an experimental 

phosphorus bond is nearly metallic in 
schreibersite. That a metallic form of 
phosphorus can be made at high pressures 
is at least consistent with metallic 
behavior of the phosphorus atom in the 
schreibersite lattice. 

Higher Phosphides in the System Fe-P 

In addition to the work on schreiber- 
site, some of the properties of Fe 2 P and 
FeP have been investigated. The powder 
diffraction patterns of these phases have 
been completely indexed out to the 
minimum d values observed with Fe K a 
radiation, and the optical properties of 
the solid phases formed by quenching 
liquids in this system and those formed 
by growth in the solid state have been 

Fe 2 P grown in equilibrium with FeP at 
1000°C has unit cell parameters measur- 
ably smaller than those of Fe 2 P equili- 
brated with Fe 3 P at the same tempera- 
ture. This indicates that at high tem- 
perature Fe 2 P departs from stoichiometry. 
The possibility of lack of stoichiometry in 
FeP has not yet been investigated. 

The fact that liquids in this system are 
known to be easily supercooled suggested 



that metastable solid phases might be 
formed on quenching liquids. Charges of 
composition 89.5 weight per cent Fe, 10.5 
weight per cent P, and 73.3 weight per 
cent Fe, 26.7 weight per cent P, were 
fused at 1070° and 1300°C, respectively, 
and quenched by dropping into ice water. 
These compositions are close to the 
eutectics between Fe and Fe 3 P, and Fe 2 P 
and FeP, respectively. The temperatures 
are a few tens of degrees above the 
eutectic temperatures of 1050° and 
1262°C (Hansen and Anderko, 1958). 

These runs yielded the phases to be 
expected if equilibrium had been reached, 
as shown by the X-ray patterns of the 
charges. The first produced Fe 3 P and 
metal, and the second Fe 2 P and FeP. The 
optical properties of the quenched charges, 
however, are strikingly different from 
those of the same phases when grown by 
combination of the elements at subsolidus 
temperatures. Both Fe 3 P and FeP were 
highly anisotropic under the microscope 
and exhibited properties corresponding to 
the description of the unknown phase in 
the metallic spherules in tektites de- 
scribed by Chao, Adler, Dwornik, and 
Littler (1962). These properties are in 
sharp contrast to the properties of these 
phases when grown at lower tempera- 
tures. The difference is possibly due to 
strains in the lattice. It does not appear 
to affect the X-ray properties. 

On the basis of its X-ray properties, the 
unknown phase in the tektites can be 
identified as an iron-rich schreibersite. 
The d values of the reflections observed 
by Chao, Adler, Dwornik, and Littler 
(1962) are compared with those of Fe 3 P 
in table 24. The spacing of the (411) 
reflection indicates a Ni/(Fe + Ni) ratio 
of about 0.05 according to the determina- 
tive curve given in Year Book 60, page 
184. This estimate may be somewhat high 
because of absorption. 

The agreement between the d values in 
columns 1 and 3 of table 24 is excellent. 
The identification of the last reflection in 
the table as (402) must be considered 
somewhat uncertain, however, since this 

TABLE 24. Comparison of the X-Ray 

Properties of the Unknown Phosphide 

and Fe 3 P 

Unknown Phase 
(Chao et al., 1962) 

Fe 3 P 

























is one of the weaker reflections in the 
schreibersite pattern. It is surprising that 
Chao et al. should observe this reflection 
and not stronger ones such as (510) or 
(132). The other reflections listed in the 
table are among the strongest ones in the 
schreibersite pattern. None of the lines of 
Fe 2 P or FeP has d values close to 1.600 A. 
These results indicate that little new 
information about the origin of tektites 
can be inferred from the presence of 
optically anisotropic schreibersite. It is 
clear from the glassy nature of these 
bodies and the spherical shape of the 
metallic particles that they have been 
melted and then relatively rapidly 
quenched. These seem to be the condi- 
tions necessary to produce the observed 
phosphide. Of greater interest is the new 
evidence that the unknown phase is 
indeed schreibersite. Tektites are com- 
monly thought to be the result of the 
"splash" produced by the impact of an 
iron meteorite. The principal disagree- 
ment about their origin centers around 
whether the impact occurred on the earth 
or the moon. In either event the metallic 
spherules are presumably part of the 
meteorite itself, and as such they may be 
virtually identical in all tektites produced 
by a given impact. Hence the content of 
minor elements like sulfur, phosphorus, 
or carbon in the spherules should help in 
determining whether or not there was a 
multiplicity of falls in regions of complex 
strewn fields such as southeast Asia and 




S. P. Clark, Jr. 

The past year has witnessed a striking 
reawakening of interest in terrestrial heat 
flow stimulated in part by geothermal 
investigations in the Pacific Ocean basin. 
This work, which in recent years has been 
carried on mainly by R. P. von Herzen 
at the University of California at La 
Jolla, has shown much fine-scale irregu- 
larity, which must in part at least be real. 
For example, the extremely high heat 
flows on the East Pacific Rise now appear 
to be confined to two relatively narrow 
zones trending parallel to the crest of the 
Rise. The sharpness of these features is 
suggestive of volcanic origin; in any event 
their cause must lie at very shallow 
depths. The interesting question whether 
similar features exist in continental 
regions cannot be answered with present 
data. A number of proposals for drilling 
holes for geothermal purposes have been 
submitted to the National Science Foun- 
dation, and it is to be hoped that the 
observational basis of this subject can be 
greatly broadened in the next few years. 

Theoretical investigations of subjects 
related to earth temperatures, such as 
those described in Year Books 59 and 60, 
have been continued. This type of work 
forms an essential background for the 
interpretation of geothermal results. The 
studies have been facilitated by the 
replacement of the IBM 704 digital 
computer at the National Bureau of 
Standards by the more powerful IBM 
7090, decreasing the expense and labor 
involved in treating the rather cumber- 
some problems that have been considered. 
Most investigative effort has been de- 
voted to the effect on surface heat flow 
of very high thermal conductivity at 
depth and to the cooling of a uniform, 
nonradioactive earth. 

In some of the cases considered below, 
radioactive generation of heat is involved. 
It is assumed that 40 per cent of the 
present heat production is by uranium, 

40 per cent by thorium, and 20 per cent 
by potassium — proportions similar to 
those commonly observed in terrestrial 
rocks. Account is taken of radioactive 
decay by fitting the decay curve of such 
an assemblage of radioactive isotopes with 
a single decay constant. This approxi- 
mation, which is amply accurate for 
present purposes, leads to a fourfold 
reduction in machine time. 

The effect of high thermal conductivity 
at depth has been investigated as 
described in Year Book 59 (p. 144). We 
consider a sphere composed of an outer 
shell with finite and constant thermal 
conductivity surrounding a central region 
with infinite conductivity. This gives a 
rough upper limit to the effect of such 
processes as radiative transfer, which lead 
to high conductivities at high tempera- 
tures and hence imply high conductivity 
at great depths. The model has obvious 
imperfections: the thickness of the outer 
shell must be set arbitrarily, the thermal 
gradient must vanish in the central 
region, and a discontinuous change in 
properties is introduced at a level where 
the properties of the real earth are likely 
to be continuous. But this approach has 
the great advantage that it leads to linear 
equations, and the radiogenic heat and 
heat flow can be clearly and uniquely 
separated from the thermal effects of 
initial heat. 

The first theoretical investigation of 
heat flow involved calculations of the 
thermal flux from an earth of constant 
properties with radioactive elements dis- 
tributed uniformly throughout a surface 
layer of variable thickness (Clark, 1961). 
It was found that the flux, when regarded 
as a function of thickness of the radio- 
active layer, passed through a broad 
maximum at moderate thicknesses and 
decreased when the thickness exceeded 
500 km. MacDonald (1961) later pub- 
lished a similar calculation for a non- 



linear model in which radiative transfer 
was taken into account. His results show 
a monotonic rise in heat flow with thick- 
ness of the radioactive layer. Since there 
is no satisfactory way of separating 
radiogenic flux from that due to initial 
heat in MacDonald's problem, and since 
he took very high initial temperatures 
(1880°C at a depth of 100 km), it seems 
worth while to examine further the case 
of an earth with a perfectly conducting 
center. This examination should reveal 
the reasons for the qualitative differences 
between the uniform and nonlinear 

In the first problem considered it was 
assumed that radioactivity was uniformly 
distributed through the outermost 500 
km of the earth. The initial temperature 
was taken to be zero, and the thickness 
of the outer shell of finite conductivity 
was allowed to range from 200 to 500 km. 
Results are shown in figure 56; as the 
thickness of the outer shell is increased, 
the curve levels off and asymptotically 
approaches a value of about 1.1. The 
pronounced minimum in the curve of 
figure 56 is perhaps surprising at first. It 

200 400 

Thickness of shell, km 


Fig. 56. Ratio of surface heat flow to present 
heat production as a function of thickness of 
outer shell of finite conductivity. Radioactivity 
uniformly distributed through outermost 500 km. 

results from conduction of heat toward 
the earth's center as well as toward the 
surface. Downward conduction is most 
efficient if the shell is 200 to 300 km thick. 
The temperature reaches a fairly pro- 
nounced maximum at shallow depths 
(fig. 57), but this maximum does not 













Depth, km 

Fig. 57. Temperatures for outer shells 200, 500, and 6371 km thick. Initial temperature zero 
Radioactivity uniformly distributed through outermost 500 km. 



exist if plausible nonzero initial tempera- 
tures are adopted. In that case the 
concentration of radioactive heat pro- 
duction near the surface tends to lessen 
the amount of cooling at shallow depths; 
the consequences are discussed below. 

In another set of calculations, the 
effect of a central region of infinite con- 
ductivity on the cooling of a nonradio- 
active earth was investigated. The initial 
temperature was assumed to be of the 
form To + mx, where x is depth and To 
and m are constants. The thermal 
conductivity does not enter this problem 
explicitly, and it is convenient to consider 
the thermal gradient at the surface rather 
than heat flow as a function of the thick- 
ness of the outer shell. Results are shown 
in figure 58, where the various curves are 
labeled with appropriate values of T 
and m. 


to 4 - 


i r 

j L 

200 400 600 800 

Thickness of shell, km 

Fig. 58. Thermal gradient at the surface as a 
function of the thickness of the outer shell of 
finite conductivity. No heat production. Initial 
temperature To + mx. Numbers beside curves 
give values of To and m. 

From figure 58 we see that the thermal 
gradient at the surface, and hence the 
heat flow, is sensitive to the thickness of 
the outer shell if T is large and m is small. 
The gradient is large for small thicknesses 

of the shell and decreases markedly as 
the thickness is increased. For small T 
and large m there is a very much weaker 
effect in the opposite sense. 

In the foregoing examples the problem 
is put somewhat differently from the way 
it appears in the published work cited 
above. Here the independent variable has 
been the thickness of the outer shell, a 
parameter most closely related to the 
assumed effectiveness of processes such as 
radiative transfer which lead to high 
conductivity at high temperatures. Figure 
59, however, shows heat flow as a function 
of thickness of the radioactive layer for 
several thicknesses of the outer shell. T 
was taken to be 1800°C, and m 0.8°C/km; 
these constants lead to initial tempera- 
tures close to those tabulated by Mac- 

These results indicate that MacDonald's 
findings of an increase in heat flow with 
increasing thickness of the radioactive 
layer is due to nonlinearity in his earth 
model. It appears to result from the 
reduced cooling caused by shallow radio- 
activity. Thickening the radioactive layer 
maintains near-surface temperatures at 
higher values, causing high thermal 
conductivity because of radiative trans- 
fer. This produces an effect analogous to 
reducing the thickness of the outer shell 
of finite conductivity. Radioactive heat- 
ing in effect promotes the escape of initial 

In the linear case the T term contrib- 
utes between 30 and 40 per cent of the 
total flux, the lowest proportion corre- 
sponding to the thickest outer shells. If 
this term were cut in half, which is a 
plausible adjustment, the extreme ratios 
of heat flow to heat production shown in 
figure 59 would be reduced to about 0.6 
and 1.1. MacDonald's estimates of the 
contribution of initial heat to the flux at 
the surface led to values less than 25 per 
cent, which seem too low by contrast with 
the present results. 

The results given above extend the 
previous study of the effect of depth of 
burial of radioactivity on surface heat 



o 1.4- 


o 1.2- 

1 1 1 1 1 1 



1 1 1 1 1 1 

200 400 600 

Thickness of radioactive layer, km 

Fig. 59. Ratio of heat flow to heat produc- 
tion as a function of the thickness of the radio- 
active layer. Initial temperature 1800 + 0.8z. 
Numbers beside the curves give thicknesses of 
the outer shell of finite conductivity in kilo- 

flow to cases involving further variable 
parameters. The effect of high thermal 
conductivity at depth proves to be 
greater than that of changing the thick- 
ness of a surface layer of radioactivity, 
especially if the initial temperature is 
high. These results point up our inability 
to find a connection between radioactive 
heat production and heat flow at the 
surface without precise hypotheses about 
thermal properties and initial tempera- 
tures in the mantle. 

A second major field of investigation 
has been the cooling of a uniform, 
nonradioactive earth. Interest in this 
problem arises mainly from its geomag- 
netic implications, discussed below. Initial 
temperatures in the earth are assumed to 
be of the form m(R n — r n ), where R is 
the radius of the earth, r is the radial 
coordinate, and m and n are constants. 
Initial temperatures for n ranging from 
1 to 4 and for a central temperature of 
500°C are shown in figure 60, and the 
amount of cooling in 5 X 10 9 years is 
shown in figure 61. For the higher values 
of n, the cooling is greatest near the 
surface and becomes small at great 
depths. For n = 1, however, the cooling 
is nearly independent of depth, and even 

increases slightly toward the center. 
These results agree with the earlier 
conclusions (Year Book 59, p. 146) that 
the cooling of the deep layers cannot 
exceed 100° or 200°C on this model. These 
data show that cooling cannot be large at 
depth unless initially the thermal gradient 
was relatively steep. 

The geomagnetic importance of this 
problem arises from the fact that forceful 
arguments can be made in support of the 
notion that the earth's magnetic field 
results from fluid motions in the outer 
core. The simplest way to produce such 
motions proves to be thermal convection. 
The temperature at the boundary of the 
inner core is probably fixed by latent heat 
of crystallization, and it becomes neces- 
sary to find conditions under which the 
temperature at the outer boundary of the 
core remains steady or decreases slightly 
with time. If this condition is not met, an 
adiabatic gradient will not be maintained 
and thermal convection will cease. 

Two processes tend to heat the lower 
mantle : conduction of heat from the core 
down the adiabatic gradient, which is 
presumed to exist; and residual radio- 
activity in the mantle itself. If conditions 
can be found under which cooling more 
than offsets these sources of heating, an 
obstacle in the path of dynamo theories 
of the magnetic field will be removed. 

The usual way of estimating initial 
temperatures in the mantle is to assume 
that they correspond to some melting 
curve. Empirically the most satisfactory 
such curve is given by the Simon equa- 
tion, which predicts a very small thermal 
gradient in the lower mantle. But this 
prediction does not take account of the 
transition zone between 400 and 1000 km. 
Evidence that phase changes are respon- 
sible for this region is accumulating, and, 
if they are, the melting curve should 
steepen in this range of depth. Further 
work will be required to show whether 
steep gradients can persist throughout 
the lower mantle and prevent any rise in 
temperature near the boundary of the 





1000 1500 2000 2500 3000 

Depth , km 

Fig. 60. Four cases of initial temperature considered. Reading from bottom to top, curves are 
for n = 1, 2, 3, and 4, respectively. 















i i 


\N = 4 


s^ N = 3 " 

&■" 1 


N = 2 

500 1000 1500 2000 2500 3000 

Depth, km 

Fig. 61. Cooling of nonradioactive earth. Initial temperatures shown in figure 60. 



G. R. Tilton, G. L. Davis, S. R. Hart, 10 and L. T. Aldrich 10 

Some drill and bore 
The solid earth, and from the strata there 
Extract a register, by which we learn 
That he who made it, and revealed its date 
To Moses, was mistaken in its age. 

Cowper, The Task 

Knowledge of the earth's crust has been given showing the extent of the observa- 
extended into the past by interpreting tions as of 1957. So much more informa- 
the isotopic ages of minerals. Some age tion is now available that it seems 
measurements give new dimension to appropriate to discuss the problem again, 
existing geological concepts; others may Our interest has been particularly 
allow a choice between conflicting ideas stimulated by additional age investiga- 
or provide a basis for new consideration tions in the southwestern United States. 
unhampered by preexisting conceptions. The results summarized in table 25 and 
An example of the extension of knowledge figure 62 indicate that crystallization of 
is afforded by an outline map of the igneous rocks occurred approximately 
central part of the North American 1300 to 1500 m.y. ago in an area extending 
continent. A regular pattern of ages has from southeast Missouri to eastern New 
been developed from the age measure- Mexico. The earlier investigations of 
ments by a number of laboratories in the Aldrich, Wetherill, Davis, and Tilton 
United States and Canada. No similar (1958) and Giletti and Damon (1961) 
regularity is apparent on other conti- have reported similar ages in western 
nents. Geophysical implications of these Arizona and Colorado, 
results await further study. The relation When these ages are compared with 
of measured ages to a geologic problem others from central North America it is 
at Rainy Lake on the Minnesota-Ontario seen that older ages occur to the north 
boundary is being studied. Zircon meas- and west of the 1300-1500 m.y. rocks; 
urements are being made in an attempt younger ages, to the south and east. The 
to find ages predating a 2600-million- distribution of ages is shown in figure 63, 
year-old metamorphic event. In Finland based on a survey of ages in the literature, 
the approximate contemporaneity of the Some of the localities have been more 
Karelian and Svecofennian orogenies has thoroughly studied than others. In favor- 
been established, although geologic evi- able places ages from zircon, mica, and 
dence had earlier been interpreted to show feldspar are in agreement ; in less favor- 
that the Svecofennian belt is the older. able ones, only K-Ar or Rb-Sr ages have 

been measured for a single mineral. The 

Geographic Distribution of Mineral Ages Paleozoic ages (200-450 m.y.) from the 

in the Central Portion of North America Appalachian chain that serve to define 

the <0.5-m.y. zone have been omitted 

One important application of the dating f or simplification; likewise the post- 
of rocks is to ascertain the ages and their p rec ambrian ages from the Rocky Moun- 
geographic distribution in the crystalline ta i n area m t h e western United States 
basement rocks of the continent of North have not been shown. 
America. Information about this problem Figure 63 shows that the age measure- 
has been accumulating from many labora- ments j n central North America can be 
tories at an ever-increasing rate over the 
past decade. In Year Book 57 a map was 10 Department of Terrestrial Magnetism. 




TABLE 25. Mineral Ages from the Central and Southwestern United States 


Mineral and 

Age, million years 

Sr 87 Ar 40 Pb 206 Pb 207 Pb 207 Pb 208 

Rb 87 K 40 U 238 U 235 Pb 206 Th 232 

St. Francis Mts. 
M-5 Fredericktown, Mo. 

M-16 Granite, Mo. 

Decaturville Uplift 
M-20 Decaturville, Mo. 

Arbuckle Mts. 
M-23 Tishomingo, Okla. 

Sandia Mts. 
A-26 Albuquerque, N. M. 

Muscovite (P) 1430 1405 
Zircon (G) 

Microcline (G) 1300 

Biotite (G) 1320 1280 

970 1120 1425 1230 

Muscovite (P) 1445 1290 
1350 1250 
1340 1300 

Zircon (G) 
Biotite (G) 

Zircon (G) 
Biotite (G) 

970 1080 1320 1200 
1120 1250 1475 1290 

P, pegmatite; G, granite. 

V.CHITA MTS "X . FRB , ER , CK ...HSjSt^ «T S 

Fig. 62. Locations of samples from the central and southwestern United States. 



Fig. 63. The distribution of ages in crystalline rocks from the central part of North America. 
Circles represent ages that are within the limits specified on the map for a particular zone. Crosses 
are ages that are not within the limits. 

grouped by age and geographic location 
in such a manner that few exceptions are 
found. In some cases, such as the 300-350 
m.y. ages for biotite from Precambrian 
gneisses in the western part of the 
Appalachian belt, the exceptions have 
obvious explanations. These ages reflect 
Appalachian metamorphism. Others, such 
as the HOO-m.y.-old granite at Pikes 
Peak, Colorado, seem to represent iso- 
lated intrusions of younger bodies of rock. 
The area comprising much of the states 
of Minnesota and Wisconsin, the northern 
peninsula of Michigan, and part of 
Ontario is a "mixed age zone" in which 
ages similar to those in each of the 
surrounding zones can be found. In 
general, the present results confirm and 
extend the pattern of age distribution 
given in Year Book 57. 

There is as yet no evidence that the 
regularities in the occurrence of age zones 

found for the central part of North 
America will be found on other conti- 
nents. On the contrary, such data as 
exist for Europe, Africa, and Australia, 
although perhaps less extensive than 
those for North America, indicate rather 
complex patterns of age distribution. 

The geophysical significance of the age 
distribution in figure 63 is a matter for 
further study. Taken at face value the 
pattern suggests that the continent has 
increased in extent over geologic time, 
but the results do not constitute proof of 
this. For a land mass to grow at the 
expense of an ocean basin it is necessary 
to form a crust some 30 to 40 km thick. 
At present it is not certain that the age 
distribution in North America applies to 
such a thick layer of rock. This factor, 
together with the lack of similar regu- 
larity in age distribution on other 
continents, indicates a need for caution in 


interpreting the results. Whatever the than any age found previously. That the 

interpretation, the distribution pattern zircon is old was indicated by the 

shows an impressive regularity. Pb 207 -Pb 206 age of 2760 m.y. The search 

, __. . „ . _, .. . _ . for very old rocks, as well as the need to 

Ages of Minerals from the Coutchiching establisn the mec hanisms of loss of 

Sediments, Rainy Lake, Ontario daughter elements, was stimulated by 

In the vicinity of Rainy Lake, at these results, and rocks of the area were 

International Falls and Fort Frances on collected. 

the Minnesota-Ontario border, A. C. New age measurements have been 

Lawson mapped a series of metamor- made in an effort to determine the time 

phosed sedimentary strata that he named intervals involved in the formation of 

Coutchiching, lying below a series of this rock sequence. The initial studies 

metamorphosed volcanic rocks (Kee- have been on the mineral zircon, because 

watin) . Circular bodies of granite-gneiss past work has shown that zircon ages are 

are enclosed by the Coutchiching. There but little affected, if at all, by the forces 

is lack of agreement among geologists attendant upon regional metamorphism. 

whether they are intrusive granites, Consequently, zircons preserve a record 

mantled gneiss domes, or paragneisses of an initial crystallization. They survive 

derived by intense metamorphism of detrital cycles because of their physical 

Coutchiching sediments. properties, thus providing some clues to 

In Year Book 59 the ages measured for the source of the sediments. The ages of 

a single zircon sample from the Coutchi- micas, feldspars, and amphibole minerals 

ching metasediments at Rainy Lake, are much more sensitive to the effects of 

Ontario, were reported. The age pattern geological cycles. 

was very discordant, so much so that The ages measured are given in table 

when examined from the viewpoint of 26. The first sample is the one to which 

continuous loss of lead by solid diffusion reference has already been made. An even 

(Tilton, 1960, and Year Book 59) the very more discordant pattern was found for 

old age of 3800 m.y. was derived — older sample CC 35, implying an impossibly 

TABLE 26. Zircon Ages, Ontario 

No. Rock and Location 




Age, million years 

p b 206 


p b 207 


p b 207 
p b 206 

Pb 208 

r pj 1 232 






















































RL 109 Coutchiching, Rainy Lake, Ont. 

CC 21 Coutchiching?, Rainy Lake 

(or granite gneiss) 
CC 22 Coutchiching?, Rainy Lake 

(or granite gneiss) 
CC 26* Coutchiching, Rainy Lake 
CC 29 Gneiss, Rainy Lake 
CC 20 Keewatin, Rainy Lake 

CC 35 Granite, Bad Vermilion Lake 

CC 33 Granite, Bad Vermilion Lake 
CC 43 Granite, Saganaga Lake, Minn. 

* Biotite from CC 26: Rb-Sr age, 2510 m.y. 



old age when corrected for loss of lead by 
continuous diffusion. Careful study of 
this sample, as well as the earlier one, 
revealed the presence of an impurity in 
both (15 per cent in CC 35, 5 per cent in 
RL 109). The impurity did not yield an 
X-ray pattern, and positive identification 
has not yet been made. The ages of these 
two samples result from the analysis of a 
mixed system, not comparable with that 
of the other zircons. That this can be so 
is evident from figure 64, representing all 
the zircon data on a concordia diagram. 

The preliminary conclusions that can 
be drawn from the results of the zircon 
analyses are: 

1. The Pb 207 -Pb 206 values for all the 
Rainy Lake zircons lie within the range 
2620-2750 m.y., a very narrow range in 
view of the geological complexity of the 
area. The ages are only a little greater 
than the mica ages from the area. 

2. The source rocks for the zircons in 
the sediments crystallized earlier than 
2600 m.y. ago, and possibly earlier than 
2750 m.y. 



OCIear. somples 
©Impure somples 

U 235 

Fig. 64. Parent-daughter ratios for zircons from Rainy Lake, Ontario, compared with the curve 
calculated for loss of lead by continuous diffusion for 2750 m.y. 

The least-squares line of best fit to the 
points, excluding the two questionable 
samples, coincides with the essentially 
linear part of the continuous diffusion loss 
curve for zircons crystallizing 2750 m.y. 
ago. The impure samples lie off this line. 
Another sample of the Bad Vermilion 
Lake zircon, CC 33, obtained from a 
different part of the granite, gave a 
pattern conforming to that of the rest of 
the zircons. 


A single Keewatin sample is not 
significantly different in age from the 
Coutchiching zircons. 

4. The discordant ages of the pure 
samples can be explained by loss of lead 
by continuous solid diffusion. 

The results can be explained in two 
ways. The zircons may have crystallized 
at a single time about 2700 m.y. ago in 
the source for all the Coutchiching 
samples, or else older rocks or sediments 



were so strongly metamorphosed that the 
zircon ' 'clocks" were completely reset 
2600-2700 m.y. ago. 

Age Relation between the Karelian and 
Svecofennian Orogenies in Finland 11 

Two orogenic belts with distinctly 
different trends have long been recognized 
in the Precambrian rocks of Finland. The 
Karelian belt extends with a north- 
northwest trend from Lake Ladoga in 
southeastern Finland to Finnish Lapland, 
whereas the Svecofennian belt extends 
across southern Finland in a general 
east- west direction. Many geologists be- 
lieve that the Svecofennian belt is older 
than the Karelian. The evidence in 
support of this view has recently been 
summarized by Eskola (1961). The 
principal observations are that the trend 
of the Karelian belt seems to interrupt 
that of the Svecofennian belt and that 
granites and granite-gneisses similar to 
those found in the Svecofennian belt 
occur as blocks in the basement complex 
of the Karelian belt. The basement 
complex in the Svecofennian belt is 
nowhere recognized with certainty by 

Kouvo (1958) reported several mica 
and zircon ages for the intrusive rocks in 
both belts and found ages of 1750 to 1850 
m.y. in each, in agreement with the few 
measurements of earlier workers. Some 
geologists have accepted the viewpoint 
that intrustion of rocks occurred simul- 
taneously in the two belts; others, notably 
Eskola (1961), have advanced another 
interpretation. Eskola suggested that the 
influence of metamorphism on the "older" 
Svecofennian belt at the time of intrusion 
of rocks of the "younger" Karelian belt 
was sufficiently strong to erase the 
existing age record in the different 
minerals. This will be called a "rejuvena- 
tion hypothesis." 

Wetherill, Kouvo, Tilton, and Gast 
(1962) found a Pb 207 -Pb 206 age of 2240 

11 In collaboration with Olavi Kouvo, Geo- 
logical Surve}- of Finland, Otaniemi. 

m.y. for zircon from a Svecofennian 
schist near Tampere, 100 miles northwest 
of Helsinki, showing that complete era- 
sure of ages had not taken place 1800 
m.y. ago in the Svecofennian belt. Since 
no age determinations had been made on 
zircon from the intrusive rocks in this 
area, the question of rejuvenation 
throughout much of the Svecofennian 
belt was not resolved by this result. The 
possibility existed that neither the sedi- 
ments nor the intrusives were completely 
regenerated 1800-1900 m.y. ago in the 
Tampere area, but were in the other 
areas studied. This postulate has been 
shown to be most unlikely by work in the 
past year. 

At Tampere a Svecofennian granodio- 
rite intrudes graywacke and phyllitic 
schists. The body is approximately 20 km 
in diameter with numerous dikes and 
stringers cutting the surrounding sedi- 
ments. Many large outcrops of granodio- 
rite occur in the area, so that it is possible 
that this body is part of a considerably 
larger mass. Zircon age determinations 
have been made on three samples: a 
specimen of granodiorite taken about 1 
km from the observed contact with the 
schists; schist A, collected about 2 km 
from the contact with the granodiorite; 
and schist B, collected 5 km from the 
contact. Zircons from these specimens 
differ in appearance in that rounding is 
more frequent in the samples from the 
schist than from the granodiorite. Obser- 
vations on 200 crystals from each sample 
indicated that 75 to 85 per cent were 
rounded in the schist samples, only 5 per 
cent in the granodiorite. The granodiorite 
contained several per cent of crystals with 
length-to- breadth ratios of 3 to 5; such 
elongated crystals were not observed in 
the schists. The zircons from the grano- 
diorite and schist are dissimilar in form 
and appear to represent two distinctly 
different populations. The age of the 
granodiorite zircon should not be appre- 
ciably influenced by zircon from the 

The age results are given in table 27. 



TABLE 27. Ages for Zircon from a Svecofennian Intrusive and the Neighboring Schists 


i, ppm 

Age, : 




p b 206 


p b 207 


p b 207 
p b 206 

p b 208 

Schist A 
Schist B 







The zircon from the granodiorite has 
nearly concordant age values compatible 
with a time of crystallization about 1900 
m.y. ago. This is in agreement with the 
results of Kouvo on other Svecofennian 
intrusive rocks. On the other hand, the 
Pb 207 -Pb 206 age values for the zircons from 
the schists are distinctly older, showing 
that the rejuvenation hypothesis does not 
apply in this area. These ages are strong 
evidence that intrusion of igneous rocks 
occurred about 1900 m.y. ago in both the 
Svecofennian and the Karelian belts and 
that the two orogenies are therefore 
approximately contemporaneous. 

The data are also pertinent to the 
problem of discordant lead ages for 
zircons. The temperature conditions un- 
der which zircons lose lead are not well 
understood and are based on studies of 
zircons that have undergone regional 
metamorphism. Here the magnitudes of 
time and temperature are poorly known. 
The present observations show that 

intrusion of sizable masses of rock can 
occur without completely erasing the age 
record in zircon in the immediately 
surrounding rock. Knowledge of the true 
age of the zircons would permit a more 
restrictive statement to be made about 
the amount of lead loss. At present these 
data do not uniquely determine the age 
or ages of the zircons from the schists. 
The zircons may have been derived from 
a source somewhat older than shown by 
the Pb 207 -Pb 206 age values, perhaps 2300 
m.y. old. Alternatively, if some loss of 
lead from the zircons did occur at the 
time of granodiorite intrusion 1900 m.y. 
ago, or if sources of more than one age 
contributed zircon to the schists, some or 
all of the zircons might be considerably 
older than 2300 m.y. Wetherill, Kouvo, 
Tilton, and Gast (1962) found 2700-m.y.- 
old rocks in the pre-Karelian basement 
complex; rocks of this age might have 
contributed zircons to the schists at 



The fatty acids are major components 
of all living matter and are among the 
more thermally stable organic substances. 
In principle, they can survive at low 
temperatures for billions of years and 
thus might be found in sedimentary rocks 
that have been deposited since the origin 
of organisms employing fatty acids. Fatty 
acids have been postulated to be the 
major material from which some petro- 
leum hydrocarbons are formed. 

We have extracted and identified fatty 

acids from recent and ancient rocks. In 
an attempt to provide a background for 
interpreting our observations we have 
also conducted laboratory studies of the 
stability of the crude components of 
living matter at elevated temperatures. 
These will be described first. 

Thermal Stability of Algae 
P. H. Abelson 

When organic detritus is deposited in 
anaerobic sediments a number of mecha- 
nisms act to alter or destroy it, including 
biological activity and chemical changes 



resulting from interaction among the 
components and degradation due to the 
intrinsic instability of organic matter. 
Unusual circumstances may shield the 
detritus from most of these factors ex- 
cept intrinsic chemical instability, which 
thus sets an upper limit to the long- 
term survival of components. Thermal 
stability can be estimated by laboratory 
experiments on pure compounds at 
elevated temperatures coupled with ap- 
plication of the Arrhenius equation to 
extrapolate to ambient temperatures. For 
saturated fatty acids this procedure 
yields decomposition times of many 
billions of years. In nature, however, most 
of the organic matter is degraded more 
rapidly, and chemical interactions surely 
play an important role. To investigate 
them, work was started last year on 
thermal stability of the components of 
algae. These studies have been extended 
to include an examination of the changes 

C-16 Sat. 

in major biochemical components and a 
more detailed look at the fate of fatty 

Chlorella pyrenoidosa was incubated 
both wet and dry in the absence of oxygen 
at temperatures of 190° and 142°C. The 
product was split into major fractions by 
the conventional procedure employed for 
fresh tissue. This has obvious drawbacks 
since after degradation the fractions no 

TABLE 28. Thermal Degradation of Algae 








Cold TCA (H 2 0- 

soluble fraction) 










Hot TCA 

(nucleic acid) 















18 Days 




C-16 Sat 


3 Days 190° C 

Fig. 65. Gas-liquid chromatograms of methyl esters of fatty acids extracted from Chlorella. The 
algae were exposed to heat for varying periods of time. The bottom curve is for a control specimen. 


longer behave exactly like those from As we shall see, these thermal tests agree 

fresh material. However, we can make only in part with what has been observed 

useful comparisons, obtain a feeling for in nature, and it appears that additional 

what is happening, and gauge the role of mechanisms operate in the sediments to 

chemical interactions and degradation in destroy unsaturated fatty acids, 
altering the organic sediments. Results of 

such experiments are displayed in table Fatty Acids in Sedimentary Rocks 

28. It may be noted that with prolonged p H AMson and p L Parker 
incubation a major amount of the organic 

matter is converted to a black insoluble Fatty acids have been extracted from 

residue similar to the kerogen of sedi- rocks ranging in age from recent to 500 

mentary rocks. Even with short exposures m.y. old. The most abundant species seen 

significant changes occur, including the were the saturated acids, stearic (C-18), 

virtual disappearance of nucleic acid, palmitic (C-16), and myristic (C-14). 

Increase in the cold trichloroacetic acid Large qualitative and quantitative differ- 

extract (water-soluble fraction) probably ences between the contents of source algal 

arises from breakdown products of other detritus and the residual carbonaceous 

fractions. Detailed examination of the material of the sediments have been 

protein and lipide fractions provides more noted. Most striking is the absence of 

detail on what has occurred. Some of the unsaturated fatty acids in even recent 

more unstable amino acids disappear as samples. 

expected, but even the more stable ones Many samples have been examined, of 

like alanine vanish at a faster rate. Thus, which the following are typical: (1) recent 

when incubated in dilute solution, pure mud from Gulf Coast off Port Aransas, 

alanine has a half -life of 2 X 10 7 sec at Texas, collected by P. L. Parker; (2) 

190°C. When it is incubated as part of recent mud from San Nicolas Basin off 

protein of algae the half-life diminishes to Southern California collected by K. 0. 

10 6 sec at 190°C. Emery; (3) core from Pedernales, Vene- 

The fate of some of the individual fatty zuela, 5000 years old, furnished by John 

acids was also examined. The striking M. Hunt; (4) Green River shale from 

feature was the relatively rapid rate of Mahogany ledge about 40 m.y. old, 

disappearance of the more highly unsat- collected near Rifle, Colorado, by P. H. 

urated substances. This is illustrated in Abelson; (5) sample from Alun shale, 

figure 65, which displays chromatograms Sweden, approximately 500 m.y. old, 

of methyl esters of fatty acids extracted furnished by Gosta Salomonsson of the 

from heated and from control Chlorella. Swedish Shale Oil Company. 

The chromatogram of the control Samples were ground in a ball mill if 

specimen reveals a substantial content of necessary to attain small particle size, 

an unsaturated C-18 fatty acid containing treated with aqueous HC1, filtered, dried, 

three double bonds (-6H). On heating at and extracted in a Soxhlet extractor. The 

190°C for 3 days this compound largely crude product consisted mainly of highly 

disappears, and it has practically van- colored tarry materials amounting to 1 

ished after 18 days. The C-18 compound to 10 per cent of the weight of organic 

with two double bonds is less sensitive, matter in the original sample. In turn the 

but it also disappears after the longer desired fatty acids constituted as little as 

incubation. The saturated and mono- 0.1 per cent of the crude extract. To 

unsaturated acids were about equally obtain reasonably resolved peaks from 

resistant, the saturated apparently being gas-liquid chromatography, partial purifi- 

the more enduring. Under favorable cation was essential. This included chem- 

thermal conditions both these classes of ical refining and solvent extraction. The 

compounds could last millions of years, saturated fatty acids can withstand rather 



Fig. 66. Chromatogram of esters of fatty acids extracted from surface mud collected at Port 
Aransas, Texas. 

Fig. 67. Chromatograms of esters of fatty acids extracted from a grab sample from San Nicolas 
Basin (lower curve) and a core from Pedernales, Venezuela. 

drastic oxidizing and reducing treatment, 
during which the tars are destroyed. The 
methyl esters of fatty acids are much 
more soluble in petroleum ether than the 
tars are. Various combinations of these 
treatments were employed, and their 
effectiveness was monitored by radio- 
active tracers. In the exploratory stages 
special care was taken to preserve 
unsaturated acids. Later it became clear 
that saturated fatty acids were present 
in substantial quantities in old rocks, and 

procedures were modified accordingly. 

Cooper (1962), who has recently re- 
ported on the occurrence of fatty acids in 
rocks and petroleum reservoir waters, 
employed urea adduction in his studies. 
This procedure concentrates fatty acids 
with respect to unsaturates and tars but 
with the small quantities of saturated 
acids available tends to lead to relatively 
large losses. 

In figures 66 to 69 are shown chromat- 
ograms of fatty acids obtained from rocks 




Fig. 68. Chromatograms of esters from Green River shale. The upper curve was obtained from 
material that had a special treatment with concentrated HI. 


Fig. 69. Chromatogram of esters of fatty acids extracted from Alun shale. The crude acids were 
treated with alkaline permanganate to partly free the fatty acids of tars. 



of a variety of ages ranging from a few 
years to 500 m.y. The chromatograms are 
not strictly comparable because of differ- 
ences in chemical processing. Neverthe- 
less, there is considerable similarity in 
these traces. All indicate the presence of 
C-14, C-16, and C-18 saturated fatty 

In the recent sediments palmitic acid 
(C-16) was the major constituent. This is 
in keeping with its ubiquitous occurrence 
and large abundance in present-day 
organisms. Oleic acid (C-18 — 2H), which 
is a major constituent of algae and which 
readily survives thermal tests, is not 
present. In the older rocks stearic acid 
(C-18) was relatively more important. We 
believe that this may or may not imply 
differences in the utilization of fatty acids 
at an earlier period. There are many 
mechanisms that could lead to relative 
losses of one or another of the constitu- 

The amounts of fatty acids extractable 
from old rocks are relatively small, 
2 X 10 -4 to 10~ 5 gram per gram of 
organic matter. The same is also true of 
recent sediments. Since the original 
detritus might have contained 10 to 20 
per cent fatty acids, a rather dramatic 
change has occurred. Only 1 part in 1000 
of these acids apparently remains in an 
extractable form after a short period of 
exposure to the anaerobic environment. 

To investigate this low yield, tracer 
experiments employing C 14 -tagged pal- 
mitic acid were carried out. These 
experiments checked the efficiency of 
procedures once acids were freed from the 
matrix; they could not shed light on the 
efficiency of the original extraction. They 
showed that a labeled substance could be 
added to a mud and be recovered in a 
crude extract, and that purifications 
could be carried through without undue 

The disappearance of most of the fatty 
acid or its tight binding to the matrix 
when detritus is converted to kerogen 
thus remains an important but unsolved 
problem of organic geochemistry. 

Another significant aspect is that the 
fatty acid content of organic matter does 
not change much with time from the 
present to the 500-m.y.-old specimen. 
This gives hope that even older occur- 
rences may be found. 

The Isolation of Organic Compounds from 
Precambrian Rocks 

T. C. Hoering 

The ultimate fate of most organic 
materials is to be oxidized to carbon 
dioxide. Some organic substances escape 
this fate by being buried in sediments. A 
few remain as compounds similar to those 
of the original living cells. Thus amino 
acids, carbohydrates, fatty acids, and 
pigments have been found in sedimentary 
rocks. However, the majority of the 
organic compounds in rocks have been 
converted to an insoluble substance 
known as kerogen. 

At 25°C organic substances are un- 
stable with respect to decomposition into 
methane, carbon dioxide, and graphite. 
At this temperature reactions leading 
toward these products require times as 
long as billions of years. The so-called 
■ 'graphite" of Precambrian sedimentary 
rocks may contain intermediate molecules 
in the chemical pathways of the decom- 
position of kerogen. 

It is the purpose of this work to con- 
sider the chemical nature of the carbon 
of Precambrian sedimentary rocks and to 
see whether any recognizable organic 
compounds can be isolated from it. Any 
such organic compounds need not bear 
much resemblance to the chemical com- 
ponents of living cells, but as the nature 
and transformations of kerogen are 
gradually understood they may give some 
insight into the existence and the nature 
of Precambrian life. 

The reduced carbon of Precambrian 
rocks is reminiscent of high-rank anthra- 
cite coal, and therefore some of the 
techniques for the elucidation of coal 
structure were employed. The reactions 
used included (a) oxidation and recovery 



of aromatic and aliphatic acids, (6) 
thermal pyrolysis and isolation of ali- 
phatic and olefmic hydrocarbons, (c) re- 
duction with anhydrous hydrogen iodide 
and identification of saturated hydro- 
carbons, (d) solvent extraction followed 
by spectroscopy of the extracts. 

For experimental simplicity, much of 
the work was done on massive graphite 
of Precambrian age. The samples include 
the following: 

1. Michigami coal from the Iron River 
formation of northern Michigan. It has 
been described by Tyler, Barghoorn, and 
Barret (1957). Samples were collected by 
E. S. Barghoorn and P. H. Abelson. 

2. Anthroxolite from Sudbury, On- 
tario, Canada (Thompson, 1956). The 
samples were collected by P. H. Abelson. 

3. Graphitic material from the Soudan 
iron mine, Oliver Mining Company, 
Soudan, Minnesota. Samples were col- 
lected by F. L. Klinger. 

4. Carbon leader from the Main Reef 
series, Transvaal, South Africa. Sample 
was donated by P. Ramdohr. 

Some work was done also on the finely 
dispersed carbon of the Gunflint chert, 
the Bulawayan limestone, and the Trans- 
vaal dolomite. These rocks are described 
in another section of the writer's report. 

Oxidation of coal by alkaline potassium 
permanganate is a well known reaction. 
The products are a mixture of benzene 
polycarboxylic acids (Holly and Mont- 
gomery, 1956). Figure 70 is a drawing of 
a paper chromatogram of the aromatic 
acids isolated from the oxidation of the 
carbonaceous material from the Soudan 
iron mine. The acids on the chromato- 
gram appeared as dark blue and fluores- 
cent spots when viewed under ultraviolet 
light. The ultraviolet adsorption spec- 
trum of an aromatic acid from one of the 
spots of the chromatogram is shown in 
figure 71; it is typical of this class of 
compounds. Through a comparison of the 
rate of migration of known substances on 
paper chromatograms, the presence of 
benzenetricarboxylic, benzenetetracar- 
boxylic, and benzenepentacarboxylic 

-< — Ammonia-Ethonol 


















O o O 

o % 

Fig. 70. Paper chromatogram of the aromatic 
acids from the oxidation of Michigami coal. A 
mixture of 1 part of coal with 1.6 parts of KOH 
was renuxed with excess KMn0 4 for 24 hours. 
The solution was acidified, treated with SO 2, and 
evaporated to dryness. The solids were extracted 
with diethyl ether. The extract was separated by 
two-dimensional paper chromatography accord- 
ing to the procedure of Germain (1959). The 
separated acids gave a deep blue color or a bright 
fluorescence when viewed under ultraviolet light. 
A comparison of Rf values and colors under 
ultraviolet light, with known acids, indicated the 
presence of benzene polycarboxylic acids. 

220 230 240 250 260 270 280 290 

Fig. 71. The ultraviolet adsorption spectrum 
of an aromatic acid from the oxidation of 
Michigami coal. The spot numbered 2 in the 
paper chromatogram shown in figure 70 was 
eluted with dilute sodium hydroxide, and the 
ultraviolet adsorption spectrum was taken. 



acids was indicated. Benzenehexacar- 
boxylic acid (mellitic acid) was identified 
in all samples, but as this compound can 
be made from purely inorganic graphite 
its presence is of little significance to this 
research. The oxidation products were 
also examined for low-molecular- weight 
aliphatic acids, but only acetic acid was 

The pyrolysis of coals is a well studied 
process. Figure 72 is a tracing from the 
gas-liquid chromatography separation of 

Time >- 

Fig. 72. Gas-liquid chromatogram of the 
hydrocarbons from the pyrolysis of Michigami 
coal. Samples of graphite were pyrolyzed in a 
vacuum, and the gases were pumped off for 
chemical analysis. The temperature was raised 
gradually. The gases given off below 250°C were 
due to adsorbed air. At 300°C hydrocarbon gases 
began to be evolved, and above 600°C molecular 
hydrogen was observed. The gases were trans- 
ferred to a temperature-programmed gas-liquid 
chromatograph and separated with a 6-foot 
silicone rubber packed column. This figure shows 
a typical chromatogram with gases from methane 
through pentane being observed. 

the hydrocarbons derived from the heat- 
ing of Michigami coal in a vacuum. All 
the hydrocarbons from methane through 
pentane have been identified. 

Thermal pyrolysis of coal is very 
destructive to any structure and does not 

give much insight into the nature of the 
organic substances present. The chemical 
reduction and liquefaction of coal may 
be more informative. Figure 73 shows a 
typical mass spectrometric analysis of the 
saturated hydrocarbons obtained from 
the action of anhydrous hydrogen iodide 
on Michigami coal. A mixture of hydro- 
carbons from methane through hexane is 
indicated. The orders of magnitude of the 
yields of hydrocarbons liberated by 
pyrolysis and reduction range from 10 to 
100 parts per million of starting rock. 

The exhaustive extraction of coal by 
basic solvents such as pyridine has long 
been a means of isolating organic sub- 
stances. Infrared adsorption spectra of 
organic molecules are very specific for the 
types of chemical bonds contained in 
them. Figure 74 is an infrared adsorption 
spectrum of the organic substances 
extracted from the carbonaceous ma- 
terial of the Transvaal dolomite by 
pyridine. The presence of methylene 
groups (-CH2-) is the most conspicuous 
feature of this spectrum. 

The chance of contamination is an 
ever-present danger in the search for 
trace amounts of organic substances in 
material that has had such a long history 
as the rocks studied in this work. Con- 
tamination in the chemical reagents and 
water used for the work can be tested by 
running the appropriate blanks. Airborne 
dust or pollen is another source of 
contamination. By using a number of 
different procedures and by looking for a 
number of different organic substances, 
we can hope to decide whether laboratory 
contamination is a problem. Natural 
contamination of the rock during the long 
period from its deposition in the Pre- 
cambrian to the present is much more 
difficult to evaluate. 

The results obtained so far support the 
premise that the so-called "graphite" of 
Precambrian rocks was originally kerogen. 
If so, it is of interest to ask whether the 
organic compounds that formed this 
kerogen were the product of living cells 
or whether they could represent abio- 




Fig. 73. The mass spectrum of the hydrocarbons from the treatment of Precambrian "graphite" 
with anhydrous hydrogen iodide. Ten grams of Michigami coal was placed in a bomb, and 50 grams 
of anhydrous hydrogen iodide was distilled in. The bomb was heated to 180°C for 16 hours. Substances 
volatile at 100°C were distilled off the reaction mixture, and the iodine and hydrogen iodide were 
removed. The gases were separated into fractions by gas-liquid chromatography, and various frac- 
tions were admitted into the mass spectrometer for analysis. This figure shows the mass spectrum of 
a mixture of hydrocarbons obtained in this manner from Michigami coal. A mixture of hydrocarbons 
from methane through pentane is shown. 

e io 

Wavelength in microns 


Fig. 74. Infrared adsorption spectrum of pyridine extract from Precambrian Transvaal dolomite. 
The "graphite" from the Transvaal dolomite was exhaustively extracted in a Soxhlet apparatus with 
pyridine. The pyridine was evaporated, and the resulting oil was pressed into a KBr pellet. The 
sharp adsorptions at 3.4-3.5, 6.8-6.9, and 14.0 microns are characteristic of isolated methylene 
(-CH 2 -) groups. The broad adsorptions at 5.5-5.6 microns are suggestive of substituted aromatic 

logically produced organic compounds 
from a period that preceded terrestrial 
life. A number of the rocks studied have 
textures that are generally described as 
due to colonial algae. The carbon isotope 
studies reported by the writer here 
indicate that photosynthesis was occur- 
ring during the time of their formation. 
Thus the evidence is in favor of the 
existence of biological activity very early 
in the Precambrian era. 

The Biogeochemistry of the Stable 
Isotopes of Carbon 

The Isotonic Composition of the Carbon of 
Fatty Acids 

P. L. Parker 

Nier and Gulbransen (1939) first 
measured variations in the C 13 /C 12 ratios 
of naturally occurring carbon. They noted 
that plant and animal carbon was 


slightly depleted in C 13 compared with recovered. A small amount of the total 

the inorganic carbon of limestone. Craig sample was analyzed to locate and 

(1953) in a detailed survey of variations identify the fatty acids present. Then a 

in the relative abundance of the carbon large sample was injected into the 

isotopes confirmed and expanded this instrument, and the pure ester of each 

observation. In both these studies the fatty acid was collected as it streamed out 

whole plant or animal was combusted to of the detector. Small glass tubes passing 

carbon dioxide, and so the measured ratio through a paper cup of dry ice served as 

represents an average of the many collectors. To obtain 2 or 3 mg of ester it 

different chemical compounds present in was necessary to repeat the collection two 

living matter. Living matter can be or three times. The glass tube containing 

broken down into different types of the sample was placed directly in the 

chemical compounds and the C 13 /C 12 combustion line, and the sample was 

ratios of these compounds compared, burned to C0 2 . This C0 2 was used for the 

Abelson and Hoering (1961) carried the mass analysis. The results of the mass 

study of the C 13 /C 12 variations to the analyses are expressed in terms of <5C 13 , 

molecular level for several amino acids the parts per thousand difference between 

isolated from a variety of photosynthetic the C 13 /C 12 ratio of the sample and a 

organisms. Measurements of C 13 /C 12 reference material, 
ratios of fatty acid molecules relative to 
the C 13 /C 12 ratios of organisms from 

which the acids were isolated are de- dC u = ^"'wi/nu^''' reference X 1000 

scribed in the following. ^ '^ ^f^nce 

The total lipide was Soxhlet- extracted A negative 5C 13 value indicates that the 
from the samples with methanol and sample contains less C 13 than the stand- 
chloroform. The extract was taken to ard; a positive value, that it contains 
dryness on a steam bath under a stream more. 

of dry nitrogen. The residue was saponi- In view of the complex physical manip- 
fied for 2 hours with a 5 per cent solution ulations and chemical procedure it was 
of potassium hydroxide in methanol and necessary to run a number of control 
acidified with sulfuric acid. A few milli- experiments to ensure that the isotope 
liters of water were added, and the fatty fractionation measured was not thereby 
acids were extracted from the mixture brought about. The esterification reaction 
with chloroform which was then taken to was shown not to fractionate isotopes by 
dryness. The water-free residue was a comparison of stearic acid with methyl 
esterified by the boron trifluoride method stearate made from the stearic acid. If 
(Metcalfe and Schmitz, 1961). This final the acid is taken as 0.0 per mil the ester 
solution was a complex mixture of the is —0.5. The isotope effect in the gas 
methyl esters of several fatty acids as chromatography was measured by corn- 
well as any material that happened to paring the ester before and after chro- 
follow the chemical procedure. Final matography. If the ester before chro- 
purifi cation and separation of the mixture matography is taken as 0.0 the ester after 
of esters into specific esters was brought chromatography and 100 per cent collec- 
about by high-temperature gas chro- tion is —0.4; after only 50 per cent 
matography. collection the ester is +4.8 (100 per cent 

The chromatographic analysis was collection was used throughout this work) . 
performed with an 8-foot copper column Isotope fractionation due to the pro- 
packed with diethylene glycol succinate cedure is less than 1.0 per mil. On the 
(LAC 3R 728) on acid-washed chromo- basis of repeated runs with the same 
sorb-P. A thermal conductivity detector starting material the overall error is 
was used so that the samples could be estimated to be 1.0 per mil. 



By means of these techniques the 
isotopic compositions of the fatty acids 
of two algae, a marine grass, and a 
plankton tow were measured. The results 
are given in table 29. Without exception 
the fatty acids were found to be signifi- 
cantly depleted in C 13 as compared with 
the whole cell. Variations between differ- 
ent fatty acids from the same organism 
are too close to experimental error to be 
considered significant. 

Chlorella was grown in the laboratory 
in a solution of the type described by 
Sorokin and Krauss (1958), 5 per cent 

TABLE 29. 5C 13 of Fatty Acids 






Carbon as 

Cell as 



Chlorella pyrenoidosa 

(inorganic carbon 

was tank CO2) 

Total cells 









Stearic plus oleic 



Ulva sp. (inorganic carbon 

was sea-water carbon) 

Total cells 


















Thalassia sp. (inorganic 

carbon taken as sea- 

water carbon) 

Total cells 





















Plankton, mostly euphau- 

siids (inorganic car- 

bon taken as sea- 

water carbon) 

Total cells 






Palmitic plus 




Stearic plus oleic and 




CO2, 95 per cent air, agitation, and 
constant illumination. Table 29 shows 
that the fatty acids of Chlorella are about 
4 per mil depleted in C 13 as compared 
with the total cells. The Chlorella used in 
the present work was grown in the same 
way and had the same carbon isotope 
ratio as the Chlorella used by Abelson and 
Hoering 2 years before. According to 
Abelson and Hoering the total amino 
acids of Chlorella are enriched in C 13 by 
3 per mil relative to the total cells. Thus 
the depletion in C 13 of the fatty acids is 
balanced by the enrichment in C 13 of the 
amino acids. The isotope variations are 
in the right direction and of the magni- 
tude to yield a material balance. 

The other three samples, from the 
ocean, were collected by the Marine 
Laboratory of the University of Miami. 
Again, the fatty acids of these three 
samples are depleted in C 13 relative to the 
whole organism. Ulva is a marine alga 
that grows attached to rocks along the 
coast. Thalassia is a marine "grass" that 
grows in great abundance in the shallow 
bays of the Gulf and Atlantic coast. The 
euphausiids are small animals that live 
in the open sea. 

The isotope fractionation in the forma- 
tion of the fatty acids is in the same 
direction for all three of the photosyn- 
thetic organisms. If the <5C 13 of the feed 
C0 2 is taken as 0.0, the 5C 13 values of the 
fatty acids for all the plants fall between 
— 17 and —25, suggesting that the 
biochemical reactions involving isotope 
fractionation in going from C0 2 to fatty 
acids are similar for all the plants studied. 

Petroleums are depleted in C 13 relative 
to modern organisms (Silverman and 
Epstein, 1958). Petroleum derived in 
large part from fatty acids, which have 
been shown to be generally depleted in 
C 13 relative to whole organisms, might 
reflect this depletion in C 13 . This is 
probably too simple a picture, because 
the organic molecules enriched in C 13 
must also be accounted for. Nevertheless, 
knowledge of the isotopic composition of 
specific types of molecules from living 



organisms may give some clues about the 
ultimate fate of the many different types 
of organic molecules trapped in the 

The Stable Isotopes of Carbon in the 

Carbonate and Reduced Carbon of 

Precambrian Sediments 

T. C. Hoering 

When inorganic carbon is fixed by plant 
photosynthesis there is an isotope effect 
and living cells have a lower concentra- 
tion of the heavy isotope of carbon than 
the carbon dioxide or bicarbonate ion of 
their environment. The analyses of the 
C 13 /C 12 ratio of a large number of 
carbonates and reduced fossil carbons 
have been published. Ages of these 
samples range from recent back to the 
Cambrian. In general, the C 13 /C 12 ratio 
of carbonates is 1.02 to 1.03 times that of 
the associated reduced carbon, undoubt- 
edly because of the isotope fractionation 
during photosynthesis. 

Wickman (1941) and Rankama (1948) 
have proposed taking the isotopic compo- 
sition of the graphite in very old rocks as 
an indication of biological or nonbio- 
logical origin. Their reasoning was criti- 
cized by Craig (1954) for a number of 
reasons, including their rather arbitrary 
grouping of carbon isotope ratios into 
biological and nonbiological. 

The purpose of the present work was 
to measure the isotopic composition of the 
carbon in coexisting carbonates and 
reduced carbons in some of the very 
oldest rocks of the Precambrian. The 
existence of isotope fractionation between 
the oxidized and reduced forms of carbon 
in a rock that has had a mild thermal 
history, especially if the magnitude of the 
fractionation is similar to that found in 
rocks of known biological association, 
suggests that photosynthesis was occur- 
ring during the time of deposition of 
the rock. In Precambrian rocks, in 
which fossil evidence is meager or non- 

existent, such geo chemical studies give 
especially important evidence on the 
record of early terrestrial life. 

The following criteria were set up for 
the selection of samples: (a) they are 
sedimentary rocks that have suffered as 
low a degree of metamorphism as possible ; 
(6) they have presumptively remnant 
algal structures; (c) their minimum age 
can be estimated from the isotopic ages 
of neighboring igneous rocks. On this 
basis, the following rocks were used: 

1. The Gunfhnt chert from near Port 
Arthur, Ontario. This rock, carefully 
described by Tyler and Barghoorn (1954), 
has a minimum age of 1.7 b.y. Structures 
contained in it have been described as 
filamentous blue-green algae. The samples 
were collected by E. S. Barghoorn. 

2. The algal limestone from the Belt 
series of Glacier Park, Montana. They 
contain structures described as colonial 
algae and are documented by Fenton and 
Fenton (1937). General aspects of the 
geology of the region are described by 
Ross (1954). A minimum age of 1.2 b.y. 
is suggested (Tilton and Davis, 1959). 
The samples were collected by P. H. 

3. Domed algal growths of the Dolo- 
mite series, from near Schmidt's Drift, 
Union of South Africa. The structures 
have been described by Young and 
Mendelsohn (1948), and isotopic ages 
measured by Nicolaysen (1958) have set 
a minimum age of 2.0 b.y. 

4. The Bulawayan limestone of the 
Zwankendaba series, from Bulawayo, 
Southern Rhodesia. The algal stromato- 
lites in these rocks have been described 
by McGregor (1940), and the rocks have 
a minimum age of 2.7 b.y. (Holmes, 
1954). These are taken by many geolo- 
gists to be among the oldest known 
sediments. Samples were collected by 
I. Goldberg. 

5. The Randville dolomite of the Iron 
River formation from near Crystal Falls, 
Michigan. The rocks have been described 
by James (1958). They have a minimum 
age of 1.5 b.y. The samples were col- 



lected by H. James and P. H. Abelson. 
Thin sections have been cut from the 
carbonate rocks for microscopic examina- 
tion. Typically they consist of partly 

6C 13 = 

\C 12 /x \C 12 / 


X 1000 

\S^ /^ ) standard 

recrystallized calcite or dolomite with There is clearly a large difference in the 

black specks of dispersed carbon in them, carbon isotope ratio between the oxidized 

In some, the black particles have been and the reduced forms of carbon. The 

concentrated along grain boundaries of dC u of the reduced carbon tends to be 

the recrystallized material. slightly more negative than is reported 

The carbon dioxide for the isotope for coals of more recent ages, possibly 

analysis of the carbonate fraction was because of isotope fractionation during 

generated by treating with concentrated the transformation of organic material 

phosphoric acid. The reduced carbon while stored in the sediments, 
fraction was isolated from the carbonate A hypothesis is that the carbonate and 

fraction by treating with hydrochloric the carbon are related to each other by 

and hydrofluoric acid. The resulting some inorganic process. The reduction of 

insoluble residue contained the reduced carbon dioxide by magmatic gases to give 

carbon, pyrite, and insoluble metal graphite or the interaction of carbon 

TABLE 30. Isotopic Composition of the Carbon in Precambrian Rocks 


5C 13 Carbonate 5C 13 Reduced Difference 

Gunflint chert 




Algal limestone, Glacier Park 




Algal domes, Dolomite series, S. Africa 




Bulawayan limestone 




Randville dolomite 




fluorides. It was combusted to give dioxide and methane to give graphite 

impure carbon dioxide. This gas was would involve high temperatures. These 

unsuitable for isotope analysis and was processes would yield isotope fractiona- 

purified by gas-solid chromatography tion, the heavy isotope concentrating in 

with a heated column of silica gel and the carbon dioxide and the light isotope 

with helium as the sweep gas. After the in the graphite phase. The rocks used in 

eluted carbon dioxide passed through a the present study, however, show no 

conductivity cell and its response was evidence of exposure to such high 

measured, it was frozen from the helium temperatures. It would have to be a 

stream by passing it through a trap coincidence that the distribution of the 

cooled with liquid nitrogen. This purifi- isotopes is so similar to that found in 

cation requires only about 5 minutes and rocks of known biological origin, 
yields very pure carbon dioxide, suitable The results of this work are consistent 

for the mass spectrometer. with a model of the existence of photo- 

The results of this experiment are synthesis and biological activity in the 

shown in table 30 and are expressed in oldest rocks of the Precambrian era. The 

parts per thousand difference in the experiments described in another part of 

C 13 /C 12 ratio of the sample and a standard the report on the isolation of organic 

material, NBS Isotope Reference Sample compounds from the carbon of Precam- 

20. brian rocks give support to the model. 




Institute on Isotopes and Radioactivity 

A week-long institute, or special course, 
"Isotopes and Radioactivity," designed 
to acquaint secondary school science 
teachers of the Washington area with the 
role of radioactive isotopes in science and 
civil defense, was held from October 30 
to November 3 at the Administration 
Building of the Carnegie Institution of 
Washington. It attracted much favorable 
attention from the press, radio, and 
television, and drew enthusiastic praise 
and thanks from the participants. 

Conceived by Philip H. Abelson, Pres- 
ident of the Washington Academy of 
Sciences, the Institute was sponsored by 
the Academy and the Joint Board on 
Science Education. At the request of Dr. 
Abelson, the morning-lecture and after- 
noon-laboratory curriculum was organ- 
ized by Ralph T. Overman, Chairman of 
the Training Division of the Oak Ridge 
Institute of Nuclear Studies. About 140 
teachers from parochial, private, and 
public schools were released from their 
classrooms to take this intensive course, 
one or two from each school. Their classes 
were met by scientists and engineers who 
had volunteered through the Joint Board 
to substitute for them. 

The Institute is discussed in more 
detail in an article, by Frank L. Campbell, 
which appeared in the December 1961 
issue of the Journal of the Washington 
Academy of Sciences. 

Journal of Geophysical Research 

The Journal of Geophysical Research is 
published monthly by the American 
Geophysical Union with P. H. Abelson 
(Geophysical Laboratory) and J. A. 
Peoples, Jr. (University of Kansas), as 
coeditors. About half of the editorial 
work, including manuscripts on upper 
atmosphere and space, as well as some of 
the papers involving geochemistry, are 
handled at this Laboratory. The Journal 

is regarded by many as the world's 
leading geophysical publication. 

Though publishing about 5400 pages a 
year, the Journal has one of the fastest 
publication times among scientific jour- 
nals. This accomplishment is due to the 
effective efforts of Dr. and Mrs. Peoples 
at Kansas, and the cooperation of Mrs. 
Lucile Stryker and Miss Mary Jane Miles 
of Carnegie Institution, and Mr. A. D. 
Singer and Miss Marjorie E. Imlay of the 
Geophysical Laboratory. 


During the report year staff members 
and fellows were invited to present 
lectures as follows : 

As the recipient of the Regents' Dis- 
tinguished Alumnus Award for 1961- 
1962, P. H. Abelson addressed a gathering 
at the Washington State University on 
April 5, 1962. At the American Associ- 
ation for the Advancement of Science 
meetings in Denver he participated in the 
Extraterrestrial Biochemistry and Bi- 
ology Symposium and the Symposium on 
Geochemical Evolution — the First Five 
Billion Years. He delivered the Retiring 
President's Address before the Washing- 
ton Academy of Sciences and the Sigma 
Xi Lecture at the Institute of Biosciences, 
Florida State University; and he partici- 
pated in the Panel Discussion on the 
Chemical Origin of Life before the 
Chemical Society of Washington. Dr. 
Abelson also lectured to the Department 
of Botany, University of Missouri; the 
Applied Physics Laboratory, Johns Hop- 
kins University; the Medical School at 
Georgetown University; the Institute for 
Space Studies, New York City; Research 
Associates at the National Institutes of 
Health; and the National Academy of 
Sciences at its annual meeting in Wash- 
ington, D. C. 

F. R. Boyd lectured at the Department 
of Geology, Pennsylvania State Uni- 



C. W. Burnham gave two talks to the 
Geology Department at the University of 
Minnesota and addressed the Washington 
Crystal Colloquium at the National 
Bureau of Standards. 

S. P. Clark, Jr., lectured at the College 
of Mineral Industries, Pennsylvania State 
University; the Department of Geology, 
University of Minnesota; the Institute of 
Geophysics, University of California at 
Los Angeles; and the National Academy 
of Sciences Summer Study Session on 
Nuclear Processes in Geology, Woods 
Hole, Massachusetts. 

G. Donnay gave a lecture on color 
symmetry groups at the Mineralogical 
Institute of the University of Tokyo, 

H. J. Greenwood delivered two lectures 
at the Department of Geology, California 
Institute of Technology. 

T. C. Hoering addressed the Research 
and Development Laboratory of the Gulf 
Oil Company, Pittsburgh; the Depart- 
ment of Botany, University of Maryland ; 
and the National Academy of Sciences 
Summer Study Session on Nuclear Proc- 
esses in Geology, Woods Hole, Massa- 
chusetts. He also participated in the 
Symposium on the Biogeochemistry of 
the Isotopes of Sulfur at Yale University. 

G. Kullerud lectured at the National 
Research Council, Ottawa, and the 
Departments of Geology at Lehigh 
University and McGill University. He 
also gave a series of five talks at the 
Department of Geology, Queen's Univer- 
sity, Kingston, Ontario, and two lectures 
at the Department of Geology, University 
of Western Ontario. 

N. Morimoto lectured at the Depart- 
ments of Geology at the University of 
California, Berkeley, and the University 
of California, Los Angeles. 

P. L. Parker addressed the Department 
of Zoology, Cornell University, and the 
Institute of Marine Science, University 
of Texas. 

H. S. Yoder, Jr., gave three lectures at 
Clemson College and one at the Lamont 
Geological Observatory of Columbia Uni- 

versity. During a visit to Japan, sup- 
ported in part by the National Science 
Foundation, he gave lectures at the 
International Symposium on Volcanology 
held in Tokyo and the Departments of 
Geology of Hokkaido, Tohoku, and 
Kyoto Universities. He also spoke on 
high-pressure techniques at symposia in 
Kyoto and Osaka sponsored jointly by 
the Department of Geology of Kyoto 
University and the Matsushita Electric 
Industrial Company. 

Penologists' Club 

Six meetings of the Penologists' Club 
were held at the Laboratory this year. 
The following papers were presented: 

"The system Fe-Zn-S; a preliminary report 
after five years," by Paul Barton and Pete 
Toulmin (U. S. Geological Survey). 

"The petrology of the Rainier underground 
tests," by D. E. Rawson (Lawrence Radi- 
ation Laboratory). 

"New observations on the opaque minerals 
of stony meteorites: Facts without hypoth- 
eses," by Paul Ramdohr (University of 
Heidelberg and Geophysical Laboratory). 

"Penological applications of the electron 
probe," by S. 0. Agrell (Cambridge Uni- 
versity) . 

"Field and laboratory observations pertain- 
ing to the origin of granite pegmatites," by 
R. H. Jahns (Pennsylvania State University). 

"Some applications of sedimentary petrol- 
ogy to layered intrusions," by E. Dale 
Jackson (U. S. Geological Survey). 

The Summary of Published Work 
below briefly describes the papers pub- 
lished in scientific journals during the 
report year. In addition, the following 
papers are now prepared for publication: 
P. H. Abelson, "Geochemistry of amino 
acids"; P. H. Abelson, "Paleobiochem- 
istry"; R. G. Arnold, R. G. Coleman, and 
V. C. Fryklund, "Temperature of crystal- 
lization of pyrrhotite and sphalerite from 
the Highland-Surprise Mine, Coeur 
d'Alene District, Idaho"; F. Chayes, 
"Numerical correlation and petrographic 
variation"; G. A. Chinner and J. F. 
Schairer, "The join CasALSisO^- 



Mg3Al 2 Si 3 0i2 and its bearing on the 
system CaO-MgO-Al 2 3 -Si0 2 at atmos- 
pheric pressure"; L. A. Clark, "X-ray 
method for rapid determination of sulfur 
and cobalt in loellingite" ; S. P. Clark, Jr., 
"Temperatures in the continental crust"; 
B. R. Doe, "Relationships of lead isotopes 
among granites, pegmatites, and sulfide 
ores near Balmat, New York"; H. J. 
Greenwood and H. L. Barnes, "Binary 
mixtures of volatile components"; G. 
Kullerud, "Sulfide research"; G. W. 
Morey, "The action of water on calcite, 
magnesite, and dolomite"; N. Morimoto, 
"On the transition of bornite"; N. 
Morimoto and G. Kullerud, "Poly- 

morphism in digenite"; J. V. Smith and 
W. Schreyer, "Location of argon and 
water in cordierite"; G. R. Tilton, G. W. 
Wetherill, and G. L. Davis, "Mineral ages 
from the Wichita and Arbuckle Moun- 
tains, Oklahoma, and the St. Francis 
Mountains, Missouri"; A. C. Turnock 
and H. P. Eugster, "Fe-Al oxides: Phase 
relationships below 1000°C"; D. R. 
Wones, "Phase equilibria of 'ferriannite/ 
KFe 3 +2 Fe +3 Si3O 10 (OH) 2 "; H. S. Yoder, 
Jr., and C. E. Tilley, "Origin of basalt 
magmas: An experimental study of 
natural and synthetic rock systems"; 
R. A. Yund, "The system Ni-As-S: Phase 
relations and mineralogical significance." 


(1352) Heat flow in the Austrian Alps. S. P. 
Clark, Jr. Geophys. J., 6, 54-63, 1961. 

Data on underground temperature obtained 
during the construction of the Arlberg and 
Tauern tunnels in Austria have been combined 
with measurements of the thermal conduc- 
tivity of 42 samples of rock from near the 
tunnels to calculate the terrestrial heat flow. 
The value in the Arlberg is found to be 
(1.9 + 0.2) X 10- 6 cal/cm 2 sec; that in the 
Tauern, (1.8 ± 0.2) X 10~ 6 cal/cm 2 sec. The 
new results are in good agreement with the 
value 1.9 X 10~ 6 cal/cm 2 sec found earlier in 
the Loetschberg tunnel in Switzerland, and 
indicate that relatively high geothermal fluxes 
extend into the eastern Alps. The high flux 
can be attributed to radioactive heat gener- 
ation in a thickened crust. 

(1353) Ponctualisation des charges dans les 
structures cristallines du type ionique. 
J. D. H. Donnay and G. Donnay. 
Compt. Rend., 253, 291-292, 1961. 

Pairs of neighboring ions with the same sign 
are replaced by points, in which is concen- 
trated the total charge of the two ions. Such 
points, regardless of their sign, are equivalent 
as far as morphology is concerned. This 
punctualization is performed on various pro- 
jections of the crystal structure of barite 
(planar, onto coordinate planes; linear, onto 
coordinate axes). 

(1354) A structural explanation of the poly- 
morphism and transitions of MgSiOg. 
W. L. Brown, N. Morimoto, and J. V. 
Smith. /. Geol, 69, 609-616, 1961. 

Differences between the polymorphs of 
MgSiOa consist essentially of different ways of 
stacking slabs of Si03 chains, and transitions 
between the polymorphs may be effected by 
movements of chains by two-thirds of the 
2-axis spacing, together with associated dis- 
placements of Mg atoms by one-third of c. 
The transitions from proto- to rhombic 
enstatite and from proto- to clinoenstatite 
involve the same percentage of displaced 
atoms, but, because the displaced atoms are 
distributed more uniformly in the second 
transition, it is thought that a nucleus of 
clinoenstatite will propagate more easily than 
one of rhombic enstatite. This suggestion is 
consistent with the rapid metastable forma- 
tion of clinoenstatite at low temperatures and 
with the sluggish formation of rhombic 
enstatite (often very disordered) from proto- 
enstatite. Shearing stress should favor the 
formation of clinoenstatite in conformity with 
the experiments of Turner et al., and thus it 
may be a ''stress mineral" in the sense of 
Harker. Highly complex schemes for arranging 
the Si0 3 chains are possible, and, as an 
example, three possible sequences are proposed 
for the enstatite with a 36 A a axis described 
by Bystrom. 



(1355) Compositions and structural states of 
anhydrous Mg-cordierites : A re-investi- 
gation of the central part of the system 
MgO-Al 2 3 -Si0 2 . W. Schreyer and J. 
F. Schairer. J. Petrol, 2, 324-406, 

The central portion of the system MgO- 
Al 2 3 -Si0 2 has been studied with the aim of 
determining the range of solid solution as well 
as the stability limits of the various structural 
states of the ternary compound cordierite. The 
previously suggested limited solid solution 
between cordierite of the composition 2MgO- 
2A1 2 3 • 5Si0 2 (2:2: 5) and Si0 2 is now believed 
to exist only metastably. Between 800° and 
1300°C the composition of cordierite was 
found to be invariably 2MgO-2Al 2 3 -5Si0 2 . 
Above 1300°C, however, there is evidence for 
the existence of limited solid solution in 
cordierite (2:2:5) toward a theoretical com- 
pound "Mg-beryl" (3:1:6). The existence of 
cordierite solid solution at liquidus tempera- 
tures has an important bearing on the melting 
relations of many compositions within the 
system. Because of this solid solution the 
courses of crystallization of melts consisting of 
normative cordierite (2:2:5) and small 
amounts of MgSi0 3 , for example, have to 
follow parts of the boundary curve between 
the cordierite and spinel fields with these two 
phases coprecipitating over a limited range of 
temperatures. The dividing line between 
compositions that complete their crystalliza- 
tion at the ternary eutectic forsterite + proto- 
enstatite -f cordierite + liquid, 1364° ± 3°C, 
and those that complete their crystallization 
at the ternary eutectic protoenstatite + cor- 
dierite + tridymite + liquid, 1355° ± 3°C, 
was formerly considered to be the join 
MgSi0 3 -cordierite (2:2:5). Because of solid 
solution in cordierite coexisting with liquid 
this dividing line is displaced slightly in the 
direction toward more siliceous bulk compo- 
sitions. Furthermore, the temperature maxi- 
mum along the boundary curve cordierite + 
protoenstatite + liquid cannot lie at the 
intersection of this boundary curve with the 
join MgSi0 3 -2:2:5, but must lie with the tie 
line MgSi0 3 -cordierite S8 . The position of this 
temperature maximum thus moves closer to 
the ternary eutectic protoenstatite -f- cor- 
dierite + tridymite + liquid. Temperatures 
and compositions of some of the invariant 
points in the system have been redetermined. 

On the basis of Miyashiro's distortion index, 
the structural states of the cordierites synthe- 

sized are subdivided into high-cordierite, 
intermediate-state cordierite, and "low"- 
cordierite. High-cordierite was obtained in all 
compositions at any temperature as the first 
form of cordierite to crystallize. With con- 
tinued heating at appropriate temperatures, 
this metastable high-cordierite was found to 
go over gradually through intermediate-state 
cordierite to the stable form "low"-cordierite. 
The rate of this transition varies with bulk 
composition and generally increases with 
temperature. In contrast to this metastable 
behavior are the stable relations among the 
polymorphs, which were found to be a func- 
tion of temperature as well as total bulk 
composition of the cordierite-bearing mixtures. 
In bulk compositions with low Al 2 3 /Si0 2 
ratios {% 1:5) high-cordierite was not found to 
be a stable phase at any temperature; in bulk 
compositions with intermediate Al 2 3 /Si0 2 
ratios high-cordierite is stable only in the 
presence of much liquid; in those with high 
Al 2 3 /Si0 2 ratios ( > 1 : 2 : 5) a stable transition 
from "low"-cordierite to high-cordierite takes 
place at subsolidus temperatures. This rela- 
tionship is considered indirect evidence that 
Al/Si ordering is the principal cause of the 
transition from high- to "low"-cordierite. 

Owing to solid solution the transition from 
"low"- to high-cordierite in the presence of 
liquid, for certain bulk compositions with 
intermediate Al 2 3 /Si0 2 ratios, takes place in 
a manner that cannot be described by a 
varying distortion index. For this reason a 
new variable, the intensity index, defined as 
* = /(5ii+42i)//(i3i), is introduced, which is zero 
for high-cordierite solid solutions and 1.15 to 
1.35 for "low"-cordierite. 

The sensitive dependence of the structural 
behavior of cordierite on its chemical environ- 
ment excludes the possibility of using this 
property as a geologic thermometer to a very 
large extent. Experimental investigations on 
cordierite-bearing synthetic "haplobuchites," 
as well as on a fused shale from the Bokaro 
coalfield in India, revealed that high-cordierite 
is not a stable phase for these bulk compo- 
sitions at any temperature. Natural cordierites 
with structural states close to, or identical 
with, high-cordierite, which have been found 
in rocks formed at high temperatures (buch- 
ites, etc.), are believed to be metastable 
products of crystallization. They are preserved 
because the duration of heating was not 
sufficient to produce the stable low-tempera- 
ture form. Petrographic and X-ray studies 



show that there is a close relationship between 
the distortion index and the degree of perfec- 
tion of the crystal form of cordierites in these 
rocks. On the basis of these results it seems 
possible to use the structural state of cor- 
dierites, at least qualitatively, as a geologic 
timer for the crystallization history of the 
enclosing rock. 

(1356) A redetermination of equilibrium rela- 
tions between kyanite and sillimanite. 
S. P. Clark, Jr. Am. J. Sci., 259, 
641-650, 1961. 

The equilibrium curve between kyanite and 
sillimanite has been established by quenching 
experiments at temperatures between 1000° 
and 1500°C and pressures between 17 and 24 
kb. The curve is given by the expression 
P = 4.1 + 13.2 X 10- 3 T, where the pressure, 
P, is in kilobars and the temperature, T, is in 
degrees Centigrade. There is some evidence 
that the phase boundary may depart from 
linearity at low temperatures, but no quanti- 
tative estimate of the amount of curvature 
can be obtained from present data. 

If kyanite forms stably in nature, pressures 
of nearly 10 kb are required. This is equivalent 
to the weight of about 30 km of overburden. 
Such great depths of burial are not required if 
pressure is contained by the strength as well 
as by the weight of the overlying rock. It is 
suggested that "tectonic overpressures" of a 
kilobar or more may exist in rocks undergoing 

(1357) Metastable solid solutions with quartz- 
type structures on the join Si0 2 - 
MgAl 2 4 . W. Schreyer and J. F. 
Schairer. Z. Krist., 116, 60-82, 1961. 

Various members of a series of metastable 
solid solutions with a quartz-type structure 
and with compositions between Si0 2 and 
MgAl 2 4 have been synthesized from glass. 
Increasing amounts of Mg +2 and Al +3 in the 
quartz structure cause a gradual contraction 
parallel to, and a gradual expansion perpen- 
dicular to, the c axis. Siliceous members of the 
series are optically positive, and less siliceous 
members negative; for a member with about 
73 weight per cent Si0 2 the birefringence is 
zero. The refractive indices of the solid 
solutions increase with decreasing Si0 2 con- 
tent. Members with less than about 92 weight 
per cent Si0 2 exhibit high-quartz structures 
even at room temperature, whereas more 

siliceous members go through an inversion to 
a low-quartz structure when quenched to 
room temperature. The temperature of this 
inversion is lower than that of pure quartz 
(Si0 2 ) as a result of the presence of Mg +2 and 
A1+ 3 in the structure. 

(1358) Phase relations in the system Ni-As. 
R. A. Yund. Econ. Geol, 56, 1273- 
1296, 1961. 

Phase relations in the system Ni-As were 
determined in rigid silica glass tubes, in 
collapsible gold tubes, and by differential 
thermal analyses. The system includes the 
well established minerals maucherite (NinAss) 
niccolite (Nii ±I As), and the NiAs 2 polymorphs 
rammelsbergite and pararammelsbergite. 

A phase with the composition of Ni 3 As 
(dienerite) could not be synthesized, and if 
this phase exists it must be stable only below 
200°C. Ni5_ x As 2 is stable to approximately 
993°C and has a large variation in its Ni/As 
ratio. Maucherite, which is essentially re- 
stricted to NinAss composition, melts incon- 
gruently at 830° ± 5°C to niccolite plus a 
liquid. The existence of a metastable form of 
NinAss appears to be likely. 

Niccolite, which is stable to 962° ± 3°C, 
also has a large variation in its Ni/As ratio. 
The niccolite solvus between NiAs and NiAs 2 
is not useful as a geothermometer, however, 
since it is nearly vertical in the temperature 
range of geologic interest. The pararammels- 
bergite-rammelsbergite inversion was found to 
occur at 590°C under the vapor pressure of 
the assemblage when pure NiAs 2 is in equi- 
librium with niccolite. The inversion tempera- 
ture is raised 22°C/1000 bars, giving a AH of 
0.57 kcal/mole at 590°C. When pure NiAs 2 is 
in equilibrium with metallic arsenic instead of 
niccolite, the inversion temperature is approx- 
imately 8°C higher. Investigation of the 
inversion temperatures of natural specimens 
of rammelsbergite and pararammelsbergite 
shows that solid solution of elements such as 
Fe, Co, and S may lower the inversion by 
more than 100°C. 

(1359) Molar volumes and thermal expansions 
of andalusite, kyanite, and sillimanite. 
B. J. Skinner, S. P. Clark, Jr., and D. E. 
Appleman. Am. J. Sci., 259, 651-668, 

Precise measurements of unit-cell param- 
eters of four andalusites, four sillimanites, and 



five kyanites from different localities lead to 
the following molar volumes at 25°C: anda- 
lusite, 51.550 + 0.011 cm 3 /mole; sillimanite, 
49.918 ± 0.015 cm 3 /mole; kyanite, 44.116 + 
0.021 cm 3 /mole. 

Unit-cell parameters at high temperatures 
were measured with a heating stage on an 
X-ray difTractometer. From these data the 
molar volumes and thermal expansions of all 
three minerals were obtained between 25° and 

(1360) Woodring Conference on Major Bio- 
logic Innovations and the Geologic 
Record. P. E. Cloud, Jr., and P. H. 
Abelson. Proc. Natl. Acad. Sci. U. S., 
47, 1705-1712, 1961. 

The Woodring Conference was held at Big 
Meadows Lodge, Skyline Drive, Virginia, 
June 14-16, 1961. It was attended by twenty- 
three biologists and geologists. The conference 
was a multidisciplinary approach to major 
biological innovations in the context of the 
geologic record, and with emphasis on the 
nature, manifestations, and timing of events 
leading to the first Metazoa. This report 
describes the proceedings of the meeting and 
includes an excellent bibliography. 

(1361) The system NaAlSi 2 6 -H 2 0-argon : To- 
tal pressure and water pressure in 
metamorphism. H. J. Greenwood. J. 
Geophys. Res., 66, 3623-3946, 1961. 

Phase equilibrium in metamorphic rocks is 
affected by temperature, pressure, the pro- 
portions of nonvolatile components, and the 
chemical potentials of the reacting volatile 
components. Theory interrelating these vari- 
ables has been tested by studying the reaction 
analcite — > albite -f- nepheline + water in the 
presence of mixtures of water and argon. New 
data on the system Ar-H 2 permit calculation 
of the composition of the water-argon mixture, 
which should equilibrate with the phase 
assemblage analcite + albite + nepheline. 
Experimental determination of this compo- 
sition as a function of pressure at constant 
temperature is in good agreement with the 

(1362) Stability relations of glaucophane. W. 
G. Ernst. Am. J. Sci., 259, 735-765, 

Stability relations have been determined for 
glaucophane [oNa 2 Mg3Al 2 Si8022(OH)2] + ex- 

cess vapor and for quartz + glaucophane + 
vapor by means of conventional hydrothermal 
techniques. The high-temperature stability 
limit of this amphibole ranges from 850°C at 
175 bars vapor ( = total) pressure to 868°C at 
2000 bars Pvapor- Neither differential stress nor 
high pressures are necessary for the formation 
of glaucophane. The presence of excess silica 
lowers its high-temperature stability limit 
only 3° to 6°C. 

Unusually large enthalpy values for the 
reactions glaucophane — * forsterite + ensta- 
tite + albite -f- vapor and quartz + glauco- 
phane — > enstatite + albite + vapor (330 + 
60 and 320 + 60 kcal/mole, respectively) can 
be explained only in part by the change in 
coordination of aluminum from 6 in glauco- 
phane to 4 in albite. The entropy of glauco- 
phane at 864°C and 1000 bars vapor pressure 
is 150 + 50 cal/deg/mole. 

Optical properties of synthetic glaucophane 
agree well with data for natural specimens. 
Unit-cell dimensions of the synthetic material 
are slightly larger than those of natural 

The experimental investigation indicates 
that glaucophane is stable over a wide range 
of physical conditions, given appropriate 
chemical conditions. Bulk compositions rich in 
soda and magnesia and poor in lime relative to 
alumina should favor production of glauco- 
phane. The rare occurrence of such chemical 
environments severely restricts the crystal- 
lization of glaucophane in nature. 

(1363) Annual report of the Director for 

(1364) Age measurements on rocks from the 
Finnish Precambrian. G. W. Wetherill, 
O. Kouvo, G. R. Tilton, and P. W. Gast. 
J. Geol, 70, 74-88, 1962. 

New mineral age measurements are reported 
from several subdivisions of the Finnish 
Precambrian. Samples of zircon, feldspar, and 
muscovite collected from the gneissose pre- 
Karelian basement area in eastern Finland 
indicate an age of about 2700 m.y. for these 
rocks. In contrast, biotite ages from the same 
rocks agree at 1800 m.y., presumably repre- 
senting the effect of the orogeny at this time. 

Measurements on samples of mantled gneiss 
domes within the Karelian belt give feldspar 
and zircon ages supporting the correlation of 
these rocks with the pre-Karelian basement 



to the east, and again the biotite ages represent 
the time of the 1800-m.y. orogeny. These 
results are closely analogous to data previ- 
ously reported for mantled gneiss domes near 
Baltimore, Maryland. 

Additional measurements on the younger 
Precambrian rocks of Finland confirm earlier 
data indicating an age of around 1800 m.y. 
for plutonic rocks associated with both the 
Svecofennian and Karelian orogenic belts. 

(1365) Polymorphism in bornite. N. Mori- 
moto and G. Kullerud. Am. Mineralo- 
gist, 46, 1270-1282, 1961. 

Synthetic Cu 5 FeS4 and natural bornite were 
observed in three crystalline modifications: 
(1) a high-temperature form, o face-centered 
cubic, with a = 5.50 ± 0.01 A, Z = 1, and 
probably antifluorite structure; (2) a meta- 
stable form, cubic, FdZm or F43m, with 
a = 10.94 ± 0.02 A, Z = 8; (3) a low- 
temperature form, primitive tetragonal, space 
group P!2iC, pseudo-i42d, with a = 10.94 ± 
0.02, c = 21.88 ± 0.04 A, Z = 16. The high- 
temperature form is nonquenchable and is 
stable only above 228° ± 5°C (for synthetic 
materials). The metastable form appears on 
rapid cooling from temperatures above that 
of the polymorphic inversion; it changes to 
the low-temperature form slowly at room 
temperature. The low-temperature and the 
metastable forms are closely related in crystal 
structure, as shown by their similar intensity 
distributions in X-ray patterns. Twinning of 
the tetragonal form about a threefold twin 
axis [221 J accounts for other previously 
reported "modifications." 

(1366) Arsenopyrite crystal-chemical relations. 
N. Morimoto and L. A. Clark. Am. 
Mineralogist, 46, 1448-1469, 1961. 

The composition of naturally occurring 
arsenopyrite varies from about FeAso.gSi.i to 
FeAsi.iSo.9, as indicated by the more credible 
published chemical analyses and one new 
analysis. Analytical errors probably account 
for any apparent deviations of the Fe/(As+S) 
ratio from 1:2. 

Five arsenopyrites of different compositions 
were studied by single-crystal X-ray methods. 
The changes caused by increasing arsenic 
content are (1) the triclinic symmetry 
approaches monoclinic and (2) metrically the 
cell approaches the orthorhombic. These 
pseudosymmetries give rise to two types of 

twinning. Although refinements of the arseno- 
pyrite crystal structure by means of (hOl) and 
(hkO) data were hampered by twinning, the 
atomic coordinates obtained in this investi- 
gation confirm those of Buerger. The inter- 
atomic distances Fe-As, Fe-S, and As-S are 
2.35, 2.25, and 2.33 A, respectively. 

Indexed X-ray powder data are given. The 
metrically monoclinic cell constants for six 
analyzed arsenopyrites relate linearly to 
arsenic content and inversely to sulfur content. 
Provided the combined minor element content 
is below 1 per cent, the curve d 1S i = 1.6106 -+- 
0.00098:r, where x is the arsenopyrite arsenic 
content in atomic per cent, enables rapid 
determination of arsenopyrite compositions to 
within 1 atomic per cent. 

(1367) Stability relations of the ferruginous 
biotite, annite. H. P. Eugster and 
D. R. Wones. J. Petrol, 3, 82-125, 

Annite, KFe 3 AlSi 3 Oio(OH) 2 , a member of 
the iron biotites and the ferrous analogue of 
phlogopite, has been synthesized and its phase 
relations have been determined as functions 
of temperature, fugacity of oxygen (/o 2 ), and 
total pressure (P to tai ~ Ph 2 o + Pn 2 ). A 
method for controlling / 02 at high total 
pressures is described, and data for the 
"oxygen buffers" used are given. Buffers range 
from quartz -f- iron -f fayalite assemblages 
(low /o 2 ) to magnetite-hematite assemblages 
(high /o 2 ). Optical properties and unit-cell 
dimensions of synthetic annites depend on the 
conditions of synthesis. 

By recalculating published analyses of 
natural iron-rich biotites it can be shown that 
a constant hydrogen content cannot be 
assumed for such biotites. Oxidation may have 
occurred by drying at 115°C. Octahedral 
occupancy therefore cannot be calculated from 
such data. 

Phase relations of annite are presented in 
2070 and 1035 bar sections. Depending on 
fo z -T values, annite was found to decompose 
to one of the following assemblages: hematite 
-f- sanidine, magnetite + sanidine, fayalite -f- 
leucite + kalsilite, iron + sanidine. All de- 
compositions are dehydration and redox 
reactions and are sensitive to changes in / H2 o 
and/o 2 (or/ H2 o and/ H2 ). At 2070 bars total 
pressure annite + magnetite + sanidine can 
coexist between 425° and 825°C, depending on 
the magnitude of /o 2 . 

In the presence of quartz the stability field 



of annite is more restricted. Phase equilibria 
in the system KAlSi0 4 -Si0 2 -Fe-0 2 -H2 have 
been summarized schematically. 

Wherever possible, thermodynamic extrap- 
olations are made to test the internal 
consistency of the data. Enthalpies of forma- 
tion are calculated for both annite and 
phlogopite. Ranges of /o 2 values in nature as 
well as mechanisms for changes in /o 2 are 
investigated. It is useful to distinguish be- 
tween assemblages that are internally buffered 
with respect to /o 2 changes and those that are 
not buffered. The applications of individual 
reactions involving annite to specific geologic 
problems are discussed with respect to 
igneous, metamorphic, and sedimentary rocks. 

(1368) The Ni-S system and related minerals. 
G. Kullerud and R. A. Yund. J. 
Petrol, 3, 126-175, 1962. 

The system Ni-S has been studied sys- 
tematically from 200° to 1030°C by means of 
evacuated, sealed silica glass tube experiments 
and differential thermal analyses. Compounds 
in the system are Ni 3 S2 (and a high-tempera- 
ture, nonquenchable Ni 3±x S 2 phase), Ni 7 Se, 
Nix_ x S, Ni 3 S 4 , and NiS 2 . The geologic occur- 
rence of the minerals heazlewoodite (Ni 3 S 2 ), 
millerite (/sNii_*S), polydymite (Ni 3 S 4 ), and 
vaesite (NiS 2 ) can now be described in terms 
of the stability ranges of their synthetic 

Hexagonal heazlewoodite, which is stoichio- 
metric within the limit of error of the experi- 
ments, inverts on heating to a tetragonal or 
pseudotetragonal phase at 556°C. This high- 
temperature phase (Ni 3±a; S 2 ) has a wide field 
of stability, from 23.5 to 30.5 weight per cent 
sulfur at 600°C, and melts incongruently at 
806° ± 3°C. The /3Ni 7 S 6 phase inverts to 
aNi 7 S6 at 397°C when in equilibrium with 
Ni 3 S 2 and at 400°C when in equilibrium with 
aNiS. Crystals of aNi 7 S 6 break down to 
Ni 3 _ x S 2 + «NiS at 573° ± 3°C. The low- 
temperature form of Nii_ x S, corresponding to 
the mineral millerite, is rhombohedral, and the 
high-temperature form has the hexagonal 
NiAs structure. Stoichiometric NiS inverts at 
379° ± 3°C, whereas Nii_ x S with the maxi- 
mum nickel deficiency inverts at 282° ± 5°C. 
The Nii_ x S-NiS 2 solvus was determined to 
985° ± 3°C, the eutectic temperature of these 
phases. Stoichiometric NiS is stable at 600°C 
but breaks down to Ni 3 _ x S 2 and aNii_ x S 
below 797°C, whereas aNi^S with 38.2 

weight per cent sulfur melts congruently at 
992° ± 3°C. Vaesite does not vary measurably 
from stoichiometric NiS 2 composition and 
melts congruently at 1007° ± 5°C. Polydym- 
ite breaks down to aNii_ x S + vaesite at 
356° + 3°C. Differential thermal analyses 
showed the existence of a two-liquid field in 
the sulfur- rich part of the system above 991°C 
and over a wide compositional range. 

(1369) Equilibrium relations between pyrrho- 
tite and pyrite from 325° to 743°C. 
R. G. Arnold. Econ. Geol, 57, 72-90, 

The pyrrhotite solvus that represents the 
compositions of pyrrhotite coexisting in equi- 
librium with pyrite was determined in the 
temperature range 325° to 743°C by experi- 
ments conducted in sealed, evacuated, silica 
glass capsules and at pressures equal to the 
pressure of the vapor in equilibrium with the 
condensed phases. Experiments conducted in 
sealed, collapsible gold tubes demonstrate that 
confining pressures of 2000 bars do not 
measurably affect the position of the solvus 
below 670°C. 

The compositions of synthetic hexagonal 
pyrrhotite were measured within ±0.13 
atomic per cent Fe with the aid of an X-ray 
determinative curve that relates d(102) to 

X-ray powder data and a general descrip- 
tion are given for an unidentified lamellar iron 
sulfide phase occurring in rapidly quenched 
iron-deficient pyrrhotite. 

Temperatures of crystallization of ten 
natural pyrrhotite-pyrite assemblages are 
estimated by means of the pyrrhotite solvus. 
The temperature of crystallization of sphal- 
erite coexisting with pyrrhotite and pyrite in 
four of these samples was also measured. With 
very few exceptions the estimates obtained 
from the two methods agree well within the 
experimental error. 

(1370) Measurement of the metal content of 
naturally occurring, metal-deficient, 
hexagonal pyrrhotite by an X-ray 
spacing method. R. G. Arnold and 
L. E. Reichen. Am. Mineralogist, 47, 
105-111, 1962. 

It is shown on the basis of fourteen chem- 
ically analyzed pyrrhotites that the metal 
content of metal-deficient natural pyrrhotites 
may be measured to ±0.25 atomic per cent 



by means of an experimentally derived X-ray 
determinative curve, provided that the com- 
bined concentration of nickel, cobalt, and 
copper in solid solution is less than about 0.6 
per cent by weight. 

(1371) Metastable osumilite- and petalite-type 
phases in the system MgO-Al 2 3 -Si0 2 . 
W. Schreyer and J. F. Schairer. Am. 
Mineralogist, 47, 90-104, 1962. 

Two new compounds have been synthesized 
metastably in the system MgO-Al 2 3 -Si0 2 . 
One has a structure similar to that of osumilite 
and other related phases, such as the synthetic 
compound Na 2 0-5MgO- 12Si0 2 . It has a 
composition along the line Si0 2 -MgAl 2 4 , 
probably close or equal to MgO-Al 2 3 -4Si0 2 
as deduced from the phase assemblages. On 
the other hand, the measured mean index of 
refraction (1.535), according to the Gladstone- 
Dale relationship, suggests a composition 
containing less Si0 2 , such as 4Mg0-4Al 2 3 - 
7Si0 2 . The other compound, whose composi- 
tion is unknown, yields a powder X-ray 
diffraction pattern similar to those of petalite, 
Li 2 0-Al 2 3 -8Si0 2 , and lithium disilicate, 
Li 2 0-2Si0 2 . The two metastable phases form 
during devitrification of glass of certain bulk 
compositions at relatively low subsolidus 
temperatures. Upon further heating they are 
gradually replaced by assemblages that are 
more stable for these bulk compositions and 
include cordierite and a silica modification 
with or without protoenstatite. 

(1372) A titaniferous basalt from the Island of 
Pantelleria. E. G. Zies. J. Petrol, 8, 
177-180, 1962. 

A new analysis of a highly titaniferous 
basalt from the Island of Pantelleria, first 
described by H. S. Washington, is presented. 
The new values for both Ti0 2 and A1 2 3 differ 
appreciably from Washington's and produce 
marked changes in the calculation of the 
CIPW norm. The analytical procedures by 
which the new values were obtained are given 
in outline. 

(1373) Centers of charges inferred from barite 
morphology. J. D. H. Donnay and 
G. Donnay. Soviet Phys. Cryst., 6, 
679-684, 1962. 

A comparison between the crystal structure 
of barite and its morphological development 
leads to the concept of centers of charges. The 

centers of charges act as equivalent points in 
the bond assemblage that controls the 

(1374) Mineral ages from the Appalachian 
province in North Carolina and Ten- 
nessee. G. L. Davis, G. R. Tilton, and 
G. W. Wetherill. J. Geophys. Res., 67, 
1987-1996, 1962. 

Age measurements are given for nine zircons 
and nine micas from the Appalachian orogenic 
zone in western North Carolina and eastern 
Tennessee. These measurements provide fur- 
ther evidence for the existence of crystalline 
rocks as old as 1000 m.y. in the area. A still 
older age of 1300 m.y. is found for zircons from 
two gneissic rocks; these older zircons are 
probably detrital. All the zircons have 
discordant ages. The discordances are com- 
patible with loss of lead by continuous 
diffusion or episodic loss as a result of Paleo- 
zoic metamorphism. Possible difficulties in 
ascribing the discordances solely to episodic 
loss during Paleozoic metamorphism are 
pointed out. The problem of loss of lead 
during fusion of zircon has been studied; 
losses are shown to be negligible. 

( 1 376) Skutterudites (Co, Ni, Fe) As 3 _ x ' Compo- 
sition and cell dimensions. E. H. 
Roseboom, Jr. Am. Mineralogist, 47, 
310-327, 1962. 

Skutterudites were synthesized by heating 
mixtures of Co, Ni, Fe, and As at 600° to 
800°C in sealed, evacuated silica tubes. The 
resulting phases were identified by powder 
X-ray diffraction methods and by ore 

Analyses of natural skutterudites by 
numerous workers have indicated nonstoichio- 
metric compositions with a deficiency of As. 
It has been suggested that the As deficiency 
may be due to the presence of other phases as 
impurities. In the present study, synthetic 
cobalt skutterudite was found to have a small 
but real As deficiency, even in the presence of 
crystalline As, but this deficiency is too small 
to account for the large deficiencies indicated 
in many analyses of natural skutterudites. 

Natural skutterudites are known to vary 
widely in their Co, Ni, and Fe content, but 
pure Ni and Fe members are unknown. The 
same is true for synthetic skutterudites. The 
limits of solid solution vary little with tern- 



perature between 600° and 800°C, and most 
natural skutterudites fall within the limits of 
solid solution observed for the synthetic 

The cell edges of twenty-six synthetic 
skutterudites with nickel content equal to or 
greater than iron content are related to 
composition by the function a = 0.1240X — 
0.0246F + 8.2060, where a is the cell edge in 
k, X is the mole ratio Ni/(Co -f- Ni + Fe), 
and Y is the mole ratio Fe/(Co + Ni + Fe). 
The function describes the measurements to a 
standard deviation of 0.00086 A. The cell 
edges of thirteen analyzed natural skutteru- 
dites of other workers show fair agreement 
with the synthetic ones, and are described by 
the above function to a standard deviation of 


0.0097 A. The deviations of the measured cell 
edges of natural skutterudites from cell edges 
computed using the function are not demon- 
strably due to differences either in (As + S)/ 
(Co -j- Ni + Fe) ratios or in total sulfur 

(1377) Erzmikroskopische Untersuchungen an 
Magnetiten der Exhalationen im Valley 
of the 10,000 Smokes. P. Ramdohr. 
Neues Jahrb. Mineral., Monatsh., 49-59, 

These fumaroles produced locally large 
quantities of loosely coherent crystals of 
magnetite. Analyses by E. G. Zies revealed 
the presence of substantial amounts of Zn, Cu, 
Pb, Mn, Ni, Co, Mo, and Sn. Various writers 
have thought of these metals as existing in an 
anomalous form of mixed crystals in the 
magnetite in spite of the fact that Zies gave 
evidence and expressed the opinion that most 
of the metallic constituents were present as 
sulfides. Actually the magnetite contains the 
following sulfides: FeS, FeS 2 , CuFeS 2 , chalco- 

pyrrhotite, bornite, Cu 2 S, Cu, ZnS, MoS 2 , 
FeAsS. Besides that there is zincite; only a 
part of Zn and Mn are in magnetite itself. 
Paragenetically, that assemblage is of interest 
for ore deposits of exhalative origin in general. 

(1382) Phase equilibria in silicate systems at 
high pressures and temperatures. F. R. 
Boyd, Jr. In Modern Very High Pres- 
sure Techniques, edited by R. H. 
Wentorf, Jr., Butterworths, Washing- 
ton, D. C, pp. 151-162, 1962. 

Studies of mineral equilibria at high pres- 
sures yield data for estimating the conditions 
of formation of igneous and metamorphic 
rocks. These data also provide a basis for 
speculation about the mineralogy of rocks in 
the earth's mantle. In the pressure range up 
to 50 kb, experiments are most easily and 
accurately made with single-stage apparatus. 
Pressures up to 100 kb can be obtained with 
two-stage apparatus in which the piston is 
supported by a KBr cell compressed to about 
20 kb. 

Most silicates whose atomic structures are 
relatively open networks invert or break down 
to denser phases in the pressure range 10 to 
30 kb. The quartz-coesite inversion is a 
chemically simple example, and the P-T curve 
for this reaction in the temperature range 700° 
to 1700°C is given. Minerals with more closely 
packed atomic structures are stable to much 
higher pressures, but some inversions in these 
minerals have been discovered. The inversion 
of the olivine Fe 2 Si0 4 to a spinel form is 
briefly discussed. Few measurements of the 
effect of pressure on the melting relations of 
silicates have thus far been made, although 
such data will have important geologic appli- 
cations. A preliminary melting curve for 
diopside to 35 kb is given. 


Abelson, P. H., see Cloud, P. E., Jr. 

Appleman, D. E., see Skinner, B. J. 

Arnold, R. G., Equilibrium relations between 
pyrrhotite and pyrite from 325° to 743°C, 
Econ. GeoL, 57, 72-90, 1962. 

Arnold, R. G., and L. E. Reiehen, Measurement 
of the metal content of naturally occurring, 
metal-deficient, hexagonal pyrrhotite by an 
X-ray spacing method, Am. Mineralogist, 47, 
105-111, 1962. 

Boyd, F. R., Jr., Phase equilibria in silicate 
systems at high pressures and temperatures, 
in Modern Very High Pressure Techniques, 
edited by R. H. Wentorf, Jr., Butterworths, 
Washington, D. C, pp. 151-162, 1962. 

Brown, W. L., N. Morimoto, and J. V. Smith, 
A structural explanation of the polymorphism 
and transitions of MgSi0 3 , J. GeoL, 69, 609- 
616, 1961. 

Clark, L. A., see Morimoto, N. 



Clark, S. P., Jr., Heat flow in the Austrian Alps, 
Geophys. J., 6, 54-63, 1961. 

Clark, S. P., Jr., A redetermination of equilibri- 
um relations between kyanite and sillimanite, 
Am. J. Sci., 259, 641-650, 1961. 

Clark, S. P., Jr., see also Skinner, B. J. 

Cloud, P. E., Jr., and P. H. Abelson, Woodring 
Conference on Major Biologic Innovations and 
the Geologic Record, Proc. Natl. Acad. Sci. 
U. 8., 47, 1705-1712, 1961. 

Davis, G. L., G. R. Tilton, and G. W. Wetherill, 
Mineral ages from the Appalachian province 
in North Carolina and Tennessee, J. Geophys. 
Res., 67, 1987-1996, 1962. 

Donnay, G., see Donnay, J. D. H. 

Donnay, J. D. H., and G. Donnay, Ponctualisa- 
tion des charges dans les structures cristallines 
du type ionique, Compt. Rend., 253, 291-292, 

Donnay, J. D. H., and G. Donnay, Centers of 
charges inferred from barite morphology, 
Soviet Phys.-Cryst., 6, 679-684, 1962. 

Ernst, W. G., Stability relations of glaucophane, 
Am. J. Sci., 259, 735-765, 1961. 

Eugster, H. P., and D. R. Wones, Stability 
relations of the ferruginous biotite, annite, 
J. Petrol, 3, 82-125, 1962. 

Gast, P. W., see Wetherill, G. W. 

Greenwood, H. J., The system NaAlSi 2 6 -H 2 0- 
argon: Total pressure and water pressure in 
metamorphism, J. Geophys. Res., 66, 3923- 
3946, 1961. 

Kouvo, O., see Wetherill, G. W. 

Kullerud, G., and R. A. Yund, The Ni-S system 
and related minerals, J. Petrol., 3, 126-175, 

Kullerud, G., see also Morimoto, N. 

Morimoto, N., and L. A. Clark, Arsenopyrite 
crystal-chemical relations, Am. Mineralogist, 
46, 1448-1469, 1961. 

Morimoto, N., and G. Kullerud, Polymorphism 

in bornite, Am. Mineralogist, 46 } 1270-1282, 

Morimoto, N., see also Brown, W. L. 

Ramdohr, P., Erzmikroskopische Untersuchung- 

en an Magnetiten der Exhalationen im Valley 

of the 10,000 Smokes, Neues Jahrb. Mineral., 

Montash., 49-59, 1962. 
Reichen, L. E., see Arnold, R. G. 
Roseboom, E. H., Jr., Skutterudites (Co,Ni,Fe)- 

As 3 ~:r: Composition and cell dimensions, Am. 

Mineralogist, 47, 310-327, 1962. 
Schairer, J. F., see Schreyer, W. 

Schreyer, W., and J. F. Schairer, Compositions 
and structural states of anhydrous Mg- 
cordierites: A re-investigation of the central 
part of the system MgO-Al 2 3 -Si0 2 , J. Petrol., 
2, 324-406, 1961. 

Schreyer, W., and J. F. Schairer, Metastable 
solid solutions with quartz-type structures on 
the join Si0 2 -MgAl 2 4 , Z. Krist., 116, 60-82, 

Schreyer, W., and J. F. Schairer, Metastable 
osumilite- and petalite-type phases in the 
system MgO-Al 2 3 -Si0 2 , Am. Mineralogist, 47, 
90-104, 1962. 

Skinner, B. J., S. P. Clark, Jr., and D. E. 
Appleman, Molar volumes and thermal expan- 
sions of andalusite, kyanite, and sillimanite, 
Am. J. Sci., 259, 651-668, 1961. 

Smith, J. V., see Brown, W. L. 

Tilton, G. R., see Davis, G. L.; Wetherill, G. W. 

Wetherill, G. W., O. Kouvo, G. R. Tilton, and 
P. W. Gast, Age measurements on rocks from 
the Finnish Precambrian, J. Geol., 70, 74-88, 

Wetherill, G. W., see also Davis, G. L. 

Wones, D. R., see Eugster, H. P. 

Yund, R. A., Phase relations in the system Ni-As, 
Econ. Geol, 56, 1273-1296, 1961. 

Yund, R. A., see also Kullerud, G. 

Zies, E. G., A titaniferous basalt from the Island 
of Pantelleria, J. Petrol, 3, 177-180, 1962. 


Abelson, P. H., and T. C. Hoering, Carbon- 
isotope fractionation in formation of amino 
acids by photosynthetic organisms, Proc. Natl 
Acad. Sci. U. S., 47, 623-632, 1961. 

Akimoto, S., T. Katsura, and M. Yoshida, 
Magnetic properties of TiFe 2 04-Fe 3 4 system 
and their change with oxidation, J. Geomagnet. 
Geoelec. (Kyoto), 9, 165-178, 1957. 

Akimoto, S., T. Nagata, and T. Katsura, The 
Ti-Fe 2 6 -Ti 2 Fe0 5 solid solution series, Nature, 
179, 37-38, 1957. 

Aldrich, L. T., G. W. Wetherill, G. L. Davis, and 
G. R. Tilton, Radioactive ages of micas from 
granitic rocks by Rb-Sr and K-A methods, 
Trans. Am. Geophys. Union, 39, 1124-1134, 

Arnold, R. .G., Equilibrium relations between 
pyrrhotite and pyrite from 325° to 743°C, 
Econ. Geol, 57, 72-90, 1962. 

Arnold, R. G., and L. E. Reichen, Measurement 
of the metal content of naturally occurring, 
metal-deficient, hexagonal pyrrhotite by an 



X-ray spacing method, Am. Mineralogist, 47, 

105-111, 1962. 
Atlas, L., The polymorphism of MgSi0 3 and 

solid state equilibria in the system MgSi0 3 - 

CaMgSi 2 6 , J. Geol, 60, 125-147, 1952. 
Atzeni, A., Sopra una particolare colata post- 

miocenica delFedificio vulcanico del M. Arci, 

Rend. Seminar. Fac. Sci. Univ. Cagliari, 

29{Y 2 ), 1-15, 1959. 
Baker, R. G., D. G. Brandon, and J. Nutting, 

The growth of precipitates, Phil. Mag., 4j 

1339-1345, 1959. 
Bandy, M. C., Geology and petrology of Easter 

Island, Bull. Geol. Soc. Am., 48, 1589-1610, 

Barrow, G., On an intrusion of muscovite biotite 

gneiss in the southeast Highlands of Scotland, 

Quart. J. Geol. Soc. London, 49, 330-358, 1893. 

Barth, T. F. W., C. W. Correns, and P. Eskola, 

Die Entstehung der Gesteine, Springer- Verlag, 

Berlin, 1939. 
Bell, R. E., and R. E. Herfert, Preparation and 

characterization of a new crystalline form of 

molybdenum disulfide, J. Am. Chem. Soc, 79, 

3351-3355, 1957. 
Berthois, L., Contribution a F etude lithologique 

de l'Archipelo des Acores, Comun. Serv. Geol. 

Port., 84, 1953. 
Bowen, N. L., The ternary system diopside- 

forsterite-silica, Am. J. Sci., 88, 207-264, 1914. 

Bowen, N. L., Genetic features of alnoitic rocks 
at Isle Cadieux, Quebec, Am. J. Sci., 3, 1-34, 

Bowen, N. L., Phase equilibria bearing on the 

origin and differentiation of the alkaline rocks, 

Am. J. Sci., 248A, 75-89, 1945. 
Bowen, N. L., and 0. Andersen, The binary 

system MgO-Si0 2 , Am. J. Sci., 37, 487-500, 


Bowen, N. L., J. F. Schairer, and H. W. V. 

Willems, The ternary system Na 2 Si0 3 -Fe 2 3 - 

Si0 2 , Am. J. Sci., 20, 405-455, 1930. 
Bowen, N. L., and O. F. Tuttle, The system 

MgO-Si0 2 -H 2 0, Bull. Geol. Soc. Am., 60, 439- 

460, 1949. 
Boyd, F. R., and J. L. England, Effect of pressure 

on the melting of diopside, CaMgSi 2 Oc and 

albite, NaAlSi 3 08, in the range up to 50 kb, J. 

Geophys. Res., 67, in press, 1962. 
Braun, F., Morphologische, genetische und 

paragenetische Trachtstudien an Baryt, Neues 

Jahrb. Mineral., Geol., Beilage Bd. A, 65, 173- 

219, 1952. 

Buerger, M. J., The precision determination of 
the linear and angular lattice constants of 
single crystals, Z. Krist, A, 97, 433-468, 1937. 

Buerger, M. J., X-Ray Crystallography, John 
Wiley & Sons, New York, 1942. 

Burnham, C. W., The structures and crystal- 
chemistry of the aluminum-silicate minerals, 
Ph.D. thesis, Massachusetts Institute of 
Technology, 1961. 

Burnham, C. W., Refinement of the crystal 
structure of kyanite, Z. Krist., in press, 1962. 

Burnham, C. W., and M. J. Buerger, Refinement 
of the crystal structure of andalusite, Z. Krist., 
115, 269-290, 1961. 

Busing, W. R., and H. A. Levy, A procedure for 
inverting large symmetric matrices, Comm. of 
the Assoc, for Computing Machinery, in press, 

Carmichael, C. M., The magnetic properties of 
ilmenite-hematite crystals, Proc. Roy. Soc. 
London, A, 263, 508-530, 1961. 

Chao, E. C. T., I. Adler, E. J. Dwornik, and J. 
Littler, Metallic spherules in tektites from 
Isabela, Philippine Islands, Science, 135, 97- 
98, 1962. 

Chayes, F., On correlation between variables of 
constant sum, J . Geophys. Res., 65, 4185-4193, 

Chayes, F., Numerical correlation and petro- 
graphic variation, /. Geol., 60, 440-452, 1962. 

Chubb, L. J., Geology of the Marquesas Islands, 
Bernice P. Bishop Museum, Honolulu, Bull., 
68, 1930. 

Clark, S. P., Jr., Heat flow from a differentiated 
earth, J. Geophys. Res., 66, 1231-1234, 1961. 

Cohen, M. U., Precision lattice constants from 

X-ray powder photographs, Rev. Sci. Instr., 6, 

68-74, 1935. 
Cooper, J. E., Fatty acids in recent and ancient 

sediments and petroleum reservoir waters, 

Nature, 193, 744-746, 1962. 

Craig, H., The geochemistry of the stable carbon 
isotopes, Geochim. Cosmochim. Acta, 3, 53-92, 

Craig, H., Geochemical implications of the 
isotopic composition of carbon in ancient rocks, 
Geochim. Cosmochim. Acta, 6, 186-196, 1954. 

Daly, R. A., The geology of American Samoa, 
Carnegie Inst. Wash. Publ. 840, pp. 95-142, 

Daly, R. A., The geology of Ascension Island, 
Proc. Am. Acad. Arts Sci., 60, 1-80, 1925. 

Daly, R. A., The geology of Saint Helena Island, 
Proc. Am. Acad. Arts Sci., 62, 31-92, 1927. 

Darken, L. S., and R. W. Gurry, The system 
iron-oxygen, I, The wtistite field and related 
equilibria, J. Am. Chem. Soc, 67, 1398-1412, 

De Vries, R. C, and E. F. Osborn, Phase equi- 
libria in high alumina part of the system 
CaO-MgO-Al 2 3 -Si0 2 , J. Am. Ceram. Soc, 40, 
6-15, 1957. 



Edwards, A. B., Tertiary lavas from the Ker- 

guelen Archipelago, Rept. B.A.N.Z. Antarctic 

Res. Expedition, ser. A, 2, pt. 5, 72-99, 1938. 
Ervin, G., and E. F. Osborn, X-ray data on 

synthetic melilites, Am. Mineralogis