/?. &. i/Wfl^-
J)jA.e trfi+S't
0}^y
Carnegie
Institution
OF WASHINGTON
Year Book 66
1966-1967
Library of Congress Catalog Card Number 3-16716
Garamond/Pridemark Press, Baltimore, Maryland
Contents
Page
Officers and Staff iv
Report of the President 1
Reports of Departments and Special Studies 1
Department of Terrestrial Magnetism 3
Committee on Image Tubes for Telescopes 133
Department of Plant Biology 141
Mount Wilson and Palomar Observatories 249
Geophysical Laboratory 317
Department of Embryology 573
Genetics Research Unit 645
Bibliography 673
Report of the Executive Committee 675
Report of Auditors 677
Abstract of Minutes of Sixty-Ninth Meeting of the Board of Trustees 693
Articles of Incorporation 695
By-Laws of the Institution 699
Index 705
in
President and Trustees
PRESIDENT
Caryl P. Haskins
BOARD OF TRUSTEES
James N. White
Chairman
Henry S. Morgan
Vice-Chairman
Garrison Norton
Secretary
Eric Ashby
Amory H. Bradford
Omar N. Bradley
Vannevar Bush
Carl J. Gilbert
Crawford H. Greenewalt
Caryl P. Haskins
Alfred L. Loomis
Robert A. Lovett
William McC. Martin, Jr.
Keith S. McHugh
Margaret Carnegie Miller1
Henry S. Morgan
Seeley G. Mudd
William I. Myers
Garrison Norton
Richard S. Perkins
Elihu Root, Jr.2
William W. Rubey
Frank Stanton
Charles P. Taft
Charles H. Townes
Juan T. Trippe
James N. White
1 Resigned May 5, 1967.
2 Resigned May 5, 19G7; died August 27, 1967.
IV
Trustees (continued)
AUDITING COMMITTEE
Keith S. McHugh, Chairman
Alfred L. Loomis
Juan T. Trippe
EXECUTIVE COMMITTEE
Henry S. Morgan, Chairman
Carl J. Gilbert
Crawford H. Greenewalt
Caryl P. Haskins
Robert A. Lovett
William I. Myers
Garrison Norton
Richard S. Perkins
James N. White
RETIREMENT COMMITTEE
Omar N. Bradley, Chairman
Garrison Norton
Richard S. Perkins
Frank Stanton
FINANCE COMMITTEE
COMMITTEE ON ASTRONOMY
Seeley G. Mudd, Chairman
Amory H. Bradford
Crawford H. Greenewalt
Richard S. Perkins, Chairman
Crawford H. Greenewalt
Alfred L. Loomis
Keith S. McHugh
Henry S. Morgan
COMMITTEE ON BIOLOGICAL SCIENCES
Alfred L. Loomis, Chairman
William I. Myers
Charles P. Taft
NOMINATING COMMITTEE
Carl J. Gilbert, Chairman
Crawford H. Greenewalt
Keith S. McHugh
James N. White
COMMITTEE ON TERRESTRIAL SCIENCES
Juan T. Trippe, Chairman
Richard S. Perkins
Staff
MOUNT WILSON AND
PALOMAB OBSERVATORIES
SIS Santa Barbara Street
Pasadena, California 91106
Horace W. Babcock, Director
Halton C. Arp
Ira S. Bowen, Distinguished
Service Staff Member
Edwin W. Dennison
Armin J. Deutsch
Jesse L. Greenstein
Robert F. Howard
Robert P. Kraft
Robert B. Leighton
Guido Munch
J. Beverley Oke
Bruce H. Rule
Allan R. Sandage
Wallace L. W. Sargent
Maarten Schmidt
Arthur H. Vaughan, Jr.
Olin C. Wilson
Harold Zirin
Fritz Zwicky
GEOPHYSICAL LABORATORY
2801 Upton Street, N.W.
Washington, D.C. 20008
Philip H. Abelson, Director
Peter M. Bell
Francis R. Boyd, Jr.
Charles W. Burnham1
Felix Chayes
Gordon L. Davis
Gabrielle Donnay
Joseph L. England
P. Edgar Hare
Thomas C. Hoering
Thomas E. Krogh2
Gunnar Kullerud
Donald H. Lindsley
J. Frank Schairer
Hatten S. Yoder, Jr.
DEPARTMENT OF
TERRESTRIAL MAGNETISM
5241 Broad Branch Road, N.W.
Washington, D.C. 20015
Ellis T. Bolton, Director
L. Thomas Aldrich, Associate
Director
Roy J. Britten
Louis Brown
Dean B. Cowie
Scott E. Forbush
W. Kent Ford, Jr.
Stanley R. Hart3
David E. Kohne
Richard B. Roberts
Vera C. Rubin
I. Selwyn Sacks
T. Jefferson Smith
John S. Steinhart
Kenneth C. Turner
1 Resigned August 31, 1966.
- Appointed August 1, 1965.
1 On leave of absence from September 1, 1966.
VI
Staff (continued)
DEPARTMENT OF PLANT BIOLOGY
Stanford, California 94305
C. Stacy French, Director
Olle Bjorkman
Jeanette S. Brown
Jens C. Clausen, Emeritus
David C. Fork
William M. Hiesey
Malcolm A. Nobs
James H. C. Smith, Emeritus
DEPARTMENT OF EMBRYOLOGY
115 West University Parkway
Baltimore, Maryland 21210
James D. Ebert, Director
David W. Bishop
Bent G. Boving
Donald D. Brown
Igor B. Dawid
Robert L. DeHaan
Elizabeth M. Ramsey
GENETICS RESEARCH UNIT
Cold Spring Harbor
New York 1172k
Alfred D. Hershey, Director
Elizabeth Burgi
Barbara McClintock
Cytogenetics Laboratory
Ann Arbor, Michigan
Helen Gay
vn
Staff (continued)
OFFICE OF ADMINISTRATION
15S0 P Street, AMI"., Washington, B.C. 20005
Caryl P. Haskins President
Edward A. Ackerman Executive Officer
James W. Boise Bursar; Secretary -Treasurer, Retirement Trust;
Executive Secretary to the Finance Committee
Marjorie H. Walburn Assistant to the President
Donald J. Patton Director of Publications
Eleanor F. Peck Editor
Kenneth R. Henard Assistant Bursar; Assistant Treasurer,
Retirement Trust
Sheila A. McGough Associate Editor
Joseph M. S. Haraburda Assistant to the Bursar
Elizabeth R. Epstein Assistant to the Director of Publications
Marshall Hornblower Counsel
STAFF MEMBERS IN SPECIAL SUBJECT AREAS
Tatiana Proskouriakoff
Anna 0. Shepard
Vlll
Staff (continued)
RESEARCH ASSOCIATES
Carnegie Research Associate
J. D. McGee
Imperial College of Science and Technology, University of London
Research Associates of the Carnegie Institution
Richard A. Chase
Johns Hopkins University
Louis B. Flexner
University of Pennsylvania
Harry E. D. Pollock
Carnegie Institution
C. E. Tilley
Cambridge University
IX
Former Presidents and Trustees
PRESIDENTS
Daniel Coit Gilman, 1902-1904 John Campbell Merriam, 1921-1938 ;
Robert Simpson Woodward, 1904-1920 President Emeritus 19 39-19 A5
Vannevar Bush, 1939-1955
Alexander Agassiz
George J. Baldwin
Thomas Barbour
James F. Bell
John S. Billings
Robert Woods Bliss
Lindsay Bradford
Robert S. Brookings
John L. Cadwalader
William W. Campbell
John J. Carty
Whitefoord R. Cole
Frederic A. Delano
Cleveland H. Dodge
William E. Dodge
Charles P. Fenner
Homer L. Ferguson
Simon Flexner
W. Cameron Forbes
James Forrestal
William N. Frew
Lyman J. Gage
Walter S. Gifford
Cass Gilbert
Frederick H. Gillett
Daniel C. Gilman
John Hay
Barklie McKee Henry
Myron T. Herrick
Abram S. Hewitt
Henry L. Higginson
Ethan A. Hitchcock
Henry Hitchcock
Herbert Hoover
William Wirt Howe
Charles L. Hutchinson
Walter A. Jessup
Frank B. Jewett
Samuel P. Langley
Ernest 0. Lawrence
Charles A. Lindbergh
William Lindsay
Henry Cabot Lodge
TRUSTEES
1904-05
Seth Low
1902-16
1925-27
Wayne MacVeagh
1902-07
1934-46
Andrew W. Mellon
1924-37
1935-61
Margaret Carnegie Miller
1955-67
1902-13
Roswell Miller
1933-55
1936-62
Darius O. Mills
1902-09
1940-58
S. Weir Mitchell
1902-14
1910-29
Andrew J. Montague
1907-35
1903-14
William W. Morrow
1902-29
1929-38
William Church Osborn
1927-34
1916-32
James Parmelee
1917-31
1925-34
Wm. Barclay Parsons
1907-32
1927-49
Stewart Paton
1916-42
1903-23
George W. Pepper
1914-19
1902-03
John J. Pershing
1930-43
1914-24
Henning W. Prentis, Jr.
1942-59
1927-52
Henry S. Pritchett
1906-36
1910-14
Gordon S. Rentschler
1946-48
1920-55
David Rockefeller
1952-56
1948-49
Elihu Root
1902-37
1902-15
Elihu Root, Jr.
1937-67
1902-12
Julius Rosenwald
1929-31
1931-66
Martin A. Ryerson
1908-28
1924-34
Henry R. Shepley
1937-62
1924-35
Theobald Smith
1914-34
1902-08
John C. Spooner
1902-07
1902-05
William Benson Storey
1924-39
1949-66
Richard P. Strong
1934-48
1915-29
William H. Taft
1906-15
1902-03
William S. Thayer
1929-32
1902-19
James W. Wadsworth
1932-52
1902-09
Charles D. Walcott
1902-27
1902
Frederic C. Walcott
1931-48
1920-24
Henry P. Walcott
1910-24
1903-09
Lewis H. Weed
1935-52
1902-04
William H. Welch
1906-34
1938-44
Andrew D. White
1902-03
1933-49
Edward D. White
1902-03
1904-06
Henry White
1913-27
1944-58
George W. Wickersham
1909-36
1934-39
Robert E. Wilson
1953-64
1902-09
Robert S. Woodward
1905-24
1914-24
Carroll D. Wright
1902-08
Under the original charter, from the date of organization until April 28, 1904, the follow-
ing 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.
XI
There is in our community an elderly painter of seascapes who knows
the sea so well that he no longer goes to look at it while he paints. He
dislikes intensely the work of a young painter ivho sets his easel on
the beach and paints things his elder does not remember having seen.
John Steinbeck — Between Pacific Tides
We have all some responsibility for action, some area, however small,
in which each of us and he alone can play the part of agent. There is
a second field, wider and not congruent ivith the first, in which each
of us can contribute to the making of policy. There is a third, wider
still, in which each of us has power to give or ivithhold assent to the
policy decisions of others. There is a fourth, yet wider, in which the
only responsibility of each of us is the neglected but important re-
sponsibility of giving or withholding the trust which supports or
inhibits our fellows in the exercise of their inalienable responsibilities,
as their trust or distrust supports or inhibits us. There is ... a fifth
field . . . the creative function which shapes the work thus and not
otherwise, ivhether the work be a building or an institution, a nation's
history or a human life. Here lies the possibility for the vision that is
manifest, for good or ill, ivhenever a 'state of art' is imposed on a
lstate of nature' . . . the authentic signature of the human mind.
Geoffrey Vickers — The Art of Judgment
. . . the educational and scientific estate . . . gains power in a socially
complex society from its capacity for social invention. And while its
poiver must rest on its ability to attract the support of attached and un-
attached individuals, in the future its numbers will command respect.
John Kenneth Galbraith— The New Industrial State
BASIC QUESTIONS OF THE HIGHEST NATIONAL IMPORT confront US at
present which, as yet, we are not well prepared to understand,
much less to meet. They involve such fundamental issues as the real
nature of our innovative society and the stage of evolution that it may
presently have reached — and the optimum ways in which science and
technology may contribute to its health and growth. What do we truly
most wish to achieve in our current level of governmental support for
science? Is it primarily to accelerate yet further the rate of discovery
and the winning of new knowledge? Is it primarily to encourage and in-
crease the development of additional centers of excellence in research
and in teaching, academic and other, in the nation? Is it primarily to
build yet more rapidly and more massively our resources of scientifically
highly trained and talented young men and women, on whose shoulders,
tomorrow, so much of our national prosperity, and indeed, of national
survival itself, must rest? Is it to architect new and more effective
3
k CARNEGIE INSTITUTION
4
means for providing and managing those massive services of a technical
kind — in health, in commerce, in welfare — the need for which presses
upon us so insistently today and may well prove still more insatiable
tomorrow? Do our primary interests, in fact, include all of these ele-
ments— evidently extremely divergent in their requirements and in-
deed at times conflicting in their demands — and yet more? Are we
primarily interested in that equality of opportunity upon which our
whole philosophy and our very identity as a people have so long rested?
Or are we most concerned with that luminous excellence whose existence
continues to provide today, as it has throughout our past, our best
guarantee of national health and development; upon which our control
of our own destiny has so predominantly rested? Or are we still, as we
have given repeated evidence in the past, equally interested in both?
And if so, how are we going to manage the inconsistencies of action
inherent in those judgments of value : inconsistencies that have always
troubled us, but which now, in our current massive and complex world,
are both far more formidable and demand workable solutions more
importunately than ever before? In such a context, we have rarely been
vouchsafed more illuminating views of the magnitude and the depth
and the complexity of these issues, and of the elements of paradox
within them, than during the year just past.
On December 16, 1966 — the year when the Congress of the United
States finally decided to reject the completion of Project Mohole, the
attempt to drill through the ocean bottom to the region of the Moho-
rovicic discontinuity — the Atomic Energy Commission announced that
the location of its 200-billion-electron-volt accelerator, estimated to
cost $375 million to build and $60 million a year to operate, would be
the village of Weston, Illinois, situated near Chicago. One hundred and
twenty-six proposals for location had been received, submitted from
forty-six of the fifty states. The award followed the longest and most
intense public discussion and political struggle in our history involving
the approval and the locating of a Federal scientific facility, and feelings
and opinions about it were as intense and as mixed as its magnitude and
nature would have predicted. As a climax, it became apparent in mid-
1967 that local authorities in Illinois had failed to give adequate assur-
ance that important social considerations, notably the provision of ap-
propriately integrated housing and living facilities in the vicinity of
the giant machine, would be sufficiently guaranteed. Faced with this
serious omission, and perhaps in some doubt that national expenditures
of this order for this purpose were wholly prudent at a moment in
history when acute perils confronted the nation both abroad and at
REPORT OF THE PRESIDENT 5
home and the demands for men and dollars to meet them were crucial,
Congress debated the matter intensely and at length. For a time, the
Congress seemed inclined to conclude that it was not justifiable to pro-
ceed with the project. Finally — but once again only after intense dis-
cussion— an affirmative judgment was rendered.
In the long view, the correctness of this specific decision seems less
significant than the syndrome of difficulties, trauma, and widespread
uncertainty that attended its taking; in particular it seems less im-
portant than the issues that proved of critical significance in judging
a wholly scientific project which was, however, so costly that it could
be accomplished only with extensive Federal support. Primary among
them, necessarily, were issues related to the order of social imperatives
in our nation and of the economic priorities in our society. Never in our
recent history, perhaps, have the range and character of the problems
surrounding the architecting of Federal support of science on the scale
characteristic of our time been so dramatically underlined in a single
decision — nor, perhaps, so prophetically limned for the future — as in
this year just past. And when these developments are viewed against
the background of certain other factors affecting the Federal support
of science, the outlines of a contemporary dilemma are yet more sharply
focused. Overall expenditures in the United States for research and
development in 1966 have been estimated at $23 billion, with the Fed-
eral Government accounting for slightly over seventy per cent of the
whole. This is a formidable total. But within the past several years,
and notably in the present one, the slope of the curve of growth of
Federal support for research and development has flattened markedly,
approximating a budgeted increase of five per cent. At the same time,
the research costs themselves have risen at least as rapidly, and perhaps
considerably more so. Put in a specific context, these increases seem
even more dramatic. Thus Dael Wolfle has recently estimated that a
dollar spent in research in 1965 purchased only eighty-two per cent
of the technical work of a dollar spent in 1960, only fifty-seven per cent
of that accomplished per dollar in 1955, only forty-one per cent of that
in 1950. Though the nation's expenditures for research and development
were, in dollar terms, eight or nine times as great in 1965 as fifteen
years earlier, yet the actual volume of research effort may have in-
creased by a factor of only three or four. The budgeted increase from
1966 to 1967 in Federal expenditures for research, indeed, barely keeps
ahead of the rise in cost of doing research, and, for all the costs of
Federally supported research and development, it may actually fall
behind.
It is natural to suppose that this slower rate of growth in the dollar
volume of Federally supported research and development in our nation
6 CARNEGIE INSTITUTION
is directly — and only — a consequence of the immediate and importunate
demands of other massive involvements both domestic and foreign,
and that, were these difficulties somehow resolved, older growth rates
would, more or less automatically, be restored. Yet these trends are by
no means confined to our own country, with our own peculiar problems.
Almost identical circumstances, for example, confront Great Britain
today, and it is probable that they are currently besieging all or nearly
all of the industrialized nations of the contemporary world — except
possibly Japan. They suggest that some deeper social circumstances
may underlie the trend. Are there in fact deeper factors that are rele-
vant here? In considering such a difficult, complex, and overween-
ingly important general question, a dimension of critical importance
is provided by some of the events of the year in quite a different dimen-
sion— that of the substantive achievements of science.
Against the background of these difficult and challenging questions
and problems surrounding the Federal support and deployment of huge
sums and programs of science and technology in the nation that have
been emphasized so forcibly during the current year, and by way of
sharpest contrast, it is worth recalling some scientific events in a differ-
ent context which have also marked this year. It was in 1960 that an
attempt was made at the great British radio observatory at Jodrell
Bank to resolve the smallest angular diameters of the then quite novel
radio sources of the sky by interferometric methods with a base line ex-
tended to 32,000 wavelengths. This attempt was, to the surprise of
many, unsuccessful, indicating that some radio sources apparently could
intercept less than three arc seconds in the sky. It was already known
that at least three of these anomalous sources of radio energy appeared
to coincide in their positions in the heavens with three optically iden-
tified starlike objects showing pronounced emission in the blue. But
only when that coincidence had been precisely established was it con-
vincingly demonstrated that the radio emitters and the visible heavenly
objects could be the same. Then came the critical observation of the
spectrum of one of them, 3C 48, by Allan Sandage, which disclosed most
unusual and hitherto unidentified broad emission lines. Three years
later at the Australian radio observatory at Parkes the astronomical
object 3C 273 was identified with a radio source determined in pre-
cision to one arc second in the sky. There followed the striking identi-
fication of its emission lines as of the Balmer series from hydrogen,
but shifted by a factor of 1.158 toward the red end of the spectrum.
Since that germinal discovery, about 150 quasi-stellar radio sources
have been identified, and it is thought that 100,000 radio-quiet quasi-
REPORT OF THE PRESIDENT 7
stellar sources, the "QSO's," ranging down to a magnitude limit of
19.7 in blue light, may exist over the whole sky. To the limit of the
3C catalog, the "quiet" sources outnumber those which are radio
emitters by about five hundred to one. This year the discovery, identi-
fication, and analysis of a new quasar which indicates in its spectrum
the greatest redshif t so far recorded were accomplished by a combination
of studies, on the radio side at the Parkes Observatory in Australia, and
on the optical at the Mount Wilson and Palomar Observatories and the
Lick Observatory in California. Its emission spectral lines are shifted
toward the red by a factor of 3.223. If this redshift by Quasar Parkes
0237-23 is interpreted entirely as a Doppler effect, the quasar stands
as the brightest object known in the universe, and the most distant from
us, receding at 82.4 per cent the velocity of light, compared with a
calculated velocity of 81.4 per cent for the most distant object previously
known, Quasar 1116 + 12.
One of the most arresting features of the quasars has been their strik-
ingly small apparent volumes, considering their immense apparent
distances and their brightness in the sky. Parkes 0237-23, for instance
appears to be of optical magnitude 16.6. A year ago a spectacular dis-
covery, made at the Mount Wilson and Palomar Observatories, revealed
that the light from Quasar 3C 446 had increased in brilliance twenty-
fold over a period measured merely in months. This year the phenome-
non was confirmed by new evidence, again from the Mount Wilson and
Palomar Observatories. Quasar 3C 279 was found capable of increasing
or decreasing in brightness by twenty-five per cent within a period of
twenty-four hours.
What is the nature of these extraordinary objects that are appar-
ently so incredibly distant from us, which appear, comparatively, to
be so small, and which yet radiate such extraordinary amounts of
energy? Calculations involving the apparent distance and the observed
intensity over the radio spectrum of Quasar 3C 273, assigning it an
arbitrary life-span of 105 years, suggest that the energy emitted is
equivalent to that which would be obtained by converting to helium, by
normal solar processes, a mass of hydrogen equivalent to that con-
tained in 109 suns. But in a whole galaxy there are estimated to be only
1011 stars, and even assuming that energy equivalent to that in a whole
galaxy is involved, it is difficult to find a physical process of energy
conversion efficient enough to account for this result.
Do quasars then represent the whole mass of a normal galaxy,
gravitationally compressed within a volume of diameter equal to only
about the distance between our sun and the star nearest us? Or is it
possible, as has recently been inferred from the rapid fluctuation in
their apparent brightness, that quasars are the nuclei of galaxies actu-
S CARNEGIE INSTITUTION
ally being formed, including a number of large exploding stars sur-
rounded by a swarm of cool condensations of stellar material that pass
randomly across the bright source, thus occluding it for us? Is all of
the fantastic redshift of quasars attributable to the normal Doppler
effect, and are the calculated almost incredible distances from us
therefore substantially correct? Or is it possible that some proportion
of the observed redshift is to be ascribed to gravitational effects — and if
so, how much? What is the significance of the recent finding that, in
some of the quasars apparently most distant, the speeds of recession
indicated by the redshifts of their emission and absorption spectra
seem not to agree? Is the assigned lifetime of 105 years too great?
Perhaps no mystery of our universe today is deeper or more strik-
ingly intriguing than that presented by the quasars. Surely no frontier
of physical knowledge in our time offers higher potential for conceptual
advances so great as to be, in effect, revolutions of kind. And in the
opening context, it is especially noteworthy that two of the most spec-
tacular recent advances on that frontier, and perhaps among the most
significant — the demonstration of the greatest redshift value so far
detected, and the demonstration of the extraordinary apparent short-
term variations in energy output of which the nuclei of some quasars
are capable — have been accomplished or consolidated during the past
year.
If the current year has brought critical advances at the frontiers of
our knowledge of physical science and in concepts of our universe on
both the conceptual and the experimental sides, this has been no less
true of research in the areas of life and of mind. Ever since Oparin's
classic work, for instance, first published in book form in the Soviet
Union in 1923 and brought to wider notice in the West with the publi-
cation of the English version of The Origin of Life in 1937, conviction
has been widespread that life as we know it — with its extraordinary
panoply and precision of protein and nucleic acid, its intricate and
wonderfully coordinated systems of information storage and informa-
tion exploitation, of growth and regulation and direction and replica-
tion implicit in the functioning of systems of DNA and RNA and of
the enzymes and their products, had a terrestrial origin in much sim-
pler protobiological systems. Various experimental demonstrations of
the possibility of the synthesis of amino acids from inorganic constitu-
ents in the laboratory, and even of the artificial synthesis of proteins,
have in recent years provided strong reinforcement for such conclusions.
Yet the formulation of thoroughly satisfactory notions of precise
chemical paths through which the earliest precursors of living things
REPORT OF THE PRESIDENT 9
might have arisen lagged well behind the experimental findings. An
early picture of the primordial oceans as a "thick soup," a puree of many
amino acids and related compounds of which the materials of life were
ultimately synthesized, went for some time critically unchallenged.
Similarly, the hypothesis that the primordial terrestrial atmosphere,
before it acquired its present high proportion of oxygen primarily de-
rived from and maintained by the photosynthetic activity of green
plants, consisted largely of methane and ammonia — a concept dating
from an early stage in the development of modern theories about the
origin of life — was held a basic premise. Only in the year just past has
it been convincingly expounded that this assumption seriously con-
tradicts well-established geological evidence that volatile substances,
produced by outgassing from rocks early in the history of the earth,
reacted with an alkaline crust to form alkaline seas and an atmosphere
of which the primary constituents were nitrogen, hydrogen, carbon
monoxide, and carbon dioxide — very different components from those
earlier envisaged. When such a mixture is exposed to radiant energy
in the laboratory, hydrogen cyanide can be produced. And hydrogen
cyanide itself, when irradiated with ultraviolet light, evolves significant
yields of amino acids, the building blocks of the proteins of living things.
Thus, during the year just past, one of the most active and stimulating
frontiers of biological research has been illumined by a concept of
process much more in harmony with views of the chemical nature of
the early earth and its oceans and atmosphere that are widely accepted
by geologists today.
For two hundred years, a curious nervous disease of sheep has been
widely recognized. It has also been greatly feared, for it appears to be
markedly contagious, and once established in an animal its progress
seems inexorable, leading in all but the rarest cases to collapse and
death. In France, the disease has long been known as "la tremblante,"
reflecting the hypersensitivity, the shuddering, and the increasing loss
of muscular coordination that mark its progress, strongly suggestive of
an extensive involvement of the nervous system. In England the same
disease is known as "scrapie" from the intense itching that character-
izes its early stages, resulting in the patchy loss of wool from constant
rubbing against objects of all kinds — a further suggestion of the early
involvement of the peripheral nervous system.
Despite the two centuries that this disease has preyed on men's flocks
and has been a constant dread to the shepherd, despite fifty years of
research into its cause, despite especially intensive investigations of
the past decade, the nature of scrapie remained elusive and unresolved.
10 CARNEGIE INSTITUTION
In 193S two investigators, J. Cuille and P. L. Chelle, reported
that they had been able to infect experimental animals with material
which passed through a filter with pores fine enough to screen the
smallest bacteria. It was then widely believed that the agent of scrapie
was a virus: a virus with a period of incubation extraordinarily long,
varying from four months to four years for sheep, and so influenced
by immunity factors apparently genetically mediated that in any
randomly selected group of exposed animals only five to thirty per cent
ever developed the disease. Yet over a period of ten years studies of
scrapie in tissue culture pursued in several laboratories consistently
failed to demonstrate the existence of any viral agent. Serological
studies failed to demonstrate the production of antibodies in affected
animals, or scrapie antigen in scrapie infective material, regardless of
its nature. Consistently, studies with the electron microscope of tissues
and extracts proved to carry scrapie infection failed to demonstrate the
presence of any virus particles. It was found, moreover, that the infec-
tive agent would withstand exposure to concentrations of formalin fatal
to conventional viruses. In one experiment, indeed, brain tissue taken
from an affected animal and preserved in ten per cent formalin for
periods ranging from six to twenty-eight months proved capable of
transmitting the disease. As added tokens of a refractoriness so great
that it is hard to associate wTith a living organism at all, heating to one
hundred degrees centigrade for half an hour did not destroy the infec-
tious power of affected tissues, nor did exposure to a five-per-cent con-
centration of chloroform at thirty-seven degrees centigrade for thirteen
days, nor to a concentration of phenol of two per cent for the same
length of time under the same conditions. Rapid freezing and thawing,
exposure to ribonuclease or deoxyribonuclease, ether, and chloroform-
methanol were equally without effect.
During this year, experiments have been reported which strongly
suggest that the infective "particle" of scrapie, if such indeed it is, may
be smaller than the smallest virus. A study of size by means of ionizing
radiation, undertaken collaboratively in the Radiopathology Research
Unit of the British Medical Research Council at Hammersmith Hos-
pital in London and the Department of Pathology at the Institute for
Research on Animal Diseases of the British Agricultural Research
Council at Compton in Berkshire resulted in an estimate of a particle of
about seven millionths of a millimeter in diameter — substantially
smaller than the smallest virus particle recorded. Irradiation with
ultraviolet light of a wavelength specifically absorbed by nucleic acids
in exposures that normally destroy their function made no difference
in the infective power of the scrapie particles, suggesting that they
may lack nucleic acid. What, then, is the nature of this seemingly new
REPORT OF THE PRESIDENT 11
kind of disease agent that can be passed from organism to organism by
simple injection, like a normal infection, which obviously can increase
in concentration with time in the infected animal as though it multi-
plied like a virus, and that, like a normal organism of disease, produces
a characteristic and reproducible — and in this case devastating — syn-
drome? There is no clear answer as yet. Only the future can show how
far the repercussions of this new frontier of medicine and of theoretical
biology will develop. But the possibilities range from a further impact
on theories of the origin of terrestrial life to notions of the causes of
abnormal cell development to an enhanced understanding of a group
of dread human diseases the nature of which, today, is little compre-
hended.
One of the major obstacles to understanding how life on earth could
have originated through the chemical paths now generally visualized
is that, though the synthesis of proteins under conditions approxi-
mating those believed to have prevailed on the primitive earth can be
readily conceived, it is less easy to imagine such processes for either
ribonucleic or deoxyribonucleic acid. Yet the very earliest life systems,
however simple and crude they may have been by modern standards,
must nonetheless have been highly adjusted and regulated relative to
nonliving matter, else they could not have persisted at all. What then
could have been the regulators? Could they, in the first instance, have
been relatively simple proteins, performing vital tasks, to be sure, but
much less complex than those that nucleic acids were later required to
handle? An observation made by I. H. Pettison and D. A. Haig in the
course of a study of the behavior of nerve cells from the brains of
animals killed in advanced stages of scrapie and maintained in organ
culture gives pause in this context as well as in that of abnormal
cellular development in general. In several cases dividing nerve cells
were thought to have been detected in these cultures, although nerve
cells in cultures of normal brain tissue of adult animals are ordinarily
of a strictly nondividing habit. It seems possible that the scrapie-
inf ected nerve cells may have acquired an abnormal growth mechanism
at least superficially reminiscent of that of a cancerous cell.
This surprising interpretation has been reinforced by a parallel and
remarkable finding announced this year by Gurdon. When the nuclei
of adult nerve cells from the brain of the clawed toad Xenopus laevis
were injected into developing toad oocytes, it was found that in mature,
unfertilized eggs, DNA synthesis might be resumed by the adult nerve-
cell nucleus, under the influence, apparently, of a cytoplasmic factor
which appears in the egg a few hours after the administration of pitui-
tary hormone to the animal. It has long been a classic concept among
cell physiologists that the power of growth and division is permanently
1J CARNEGIE INSTITUTION
lost at maturity. The idea that it is only inhibited is new and radical.
But it has been foreshadowed for several years by work in the De-
partment of Embryology, described elsewhere in this volume, in which
it has proved possible to "reactivate" the synthesis of DNA in mature
muscle cells infected with Rous sarcoma virus. This effect, moreover,
has been confirmed by Yaffe at Rehovoth in cultures of mature muscle
cells infected with polyoma virus. It seems probable that what we con-
sidered so long to be a permanent loss of the power of specialized cells
to grow and divide in higher animals at their maturity may in fact
represent only the repression involved in an exquisitely regulated cellu-
lar system.
There are some resemblances between the behavior of the presumed
infective particle of scrapie and an agent involved in a condition which
has been termed "allergic encephalomyelitis." A small basic protein,
thought by some to be a polypeptide, has been isolated from the central
nervous system of an animal exhibiting nervous symptoms resembling
those of true encephalomyelitis. This protein, when injected into the
footpad of a guinea pig, results two or three weeks later in an acute
reaction in the nervous system of the animal, terminating in paralysis
or even death. This "allergic encephalomyelitis" does not seem to be
infectious, and it may be only a single episode of transfer. In this it is
unlike scrapie. But it is like the scrapie agent in its remarkable resist-
ance to high temperatures, formalin, solvents, and ultraviolet light.
Does it belong in the same class of substances as the scrapie agent?
It seems possible that future findings on the nature of scrapie could
shed light on the fundamental nature of that group of important human
diseases of the central nervous system such as multiple sclerosis where,
as in scrapie, there is slow destruction of that system. Particularly
striking is the possibility that perhaps the most enigmatic human
nervous disease known today, the notorious kuru of the New Guinean
Fore people, which has variously been held to be transmitted by he-
redity, or congenitally, or infectiously, may be illumined by the findings
on scrapie.
At yet another research frontier, where novel theoretical findings
in biology combine with significant potential gains in a different theater
of preventive medicine, this year has once again witnessed striking
events. Although it has been eliminated from a part of the world, the
human body louse still persists in many areas. There it can serve as
the vector transmitting the dreaded typhus or plague from man to
man, as it has done in the scourges of that disease since the dawn of
human history. The louse, and the agent of typhus that it carries, are
REPORT OF THE PRESIDENT 13
as characteristically associated today as they have always been with
hunger, filth, and extreme poverty, and so thrive today in parts of the
world least prepared to meet their onslaught. They continue to be, as
they always have been, the particular accompaniment of the sufferings
and the social disorganization of wars.
Since World War II an intensive battle has been waged against the
human body louse with modern chemical insecticides, particularly
DDT. There have been two consequences. The louse has been eradicated
from many areas. But, like many other insects, the louse is a biochem-
ically versatile organism. In areas where it has not been eliminated,
mutants that successfully resist such agents have largely replaced the
susceptible forms which made up the greatest proportion of the original
populations — as has also happened extensively among the disease-trans-
mitting houseflies and mosquitoes of the world. This development has
created epidemiological emergencies comparable to those produced by
parallel situations among some of the most serious bacterial agents of
disease, where the causative organisms have acquired similar immuni-
ties to the newer antibiotics. Once this has happened, whether among
disease organisms or among the insect vectors that carry them, a way is
reopened to the most dangerous consequences. For our defenses are re-
duced to essentially what they were before the antibiotics and the in-
secticides, on which we have relied so implicitly through the past
decades, were discovered. It is not simple to devise new chemical weap-
ons of comparable effectiveness. More important, it is very likely that,
even when such new weapons are designed and produced, they will once
again gradually become ineffective as further biochemical mutants of
the targets appear and are rapidly selected.
A dramatic new approach announced during the current year could
well foreshadow a major change in the whole method of control of the
vectors of insect-borne disease. More than a decade ago, Carroll M.
Williams of Harvard University discovered in the male of a species
of silkworm moth a rich natural source of a remarkable insect hor-
mone. It serves as one of the elements in a delicately balanced endocrine
system modulating and controlling the complex and precisely adjusted
patterns of development in insects. Its particular function is to main-
tain insect tissue in a youthful, or larval, condition, inhibiting matura-
tion. In the normal insect its action is balanced by another hormone
system, which promotes the maturity of tissues. The opposing systems
ensure the precise structuring of the complex and highly regulated
metamorphosis of the normal insect.
Almost at the time of this discovery, its author predicted that if ever
the naturally occurring juvenile hormone could be made available in
quantity, or if a biologically active analogue could be synthesized arti-
24 CARNEGIE INSTITUTION
ficially at reasonable cost, a new class of insecticides might become
available — a class distinguished from its predecessors by at least three
important characteristics. Unlike substances of the type of DDT, they
would be harmless to organisms outside the class of insects, such as fish,
birds, mammals, and man himself. They might even be made specific
to certain targets, so that beneficial insects need not be indiscriminately
slaughtered with noxious ones. And since these hormones are evolved
by the targets themselves, it is extremely unlikely that the insects could
acquire the immunity to the hormones that they have developed so
effectively for the chemical classes of poisons.
Recently, the first part of this prediction came true. It proved possible
to synthesize an analogue of the insect juvenile hormone which was
biologically active, and also to find cheap and abundant sources of a
natural product. During the year just past, the synthetic hormone was
tested on the human body louse with decisive effects. Eggs continuously
exposed to it failed to hatch. Exposure of the immature insect to mod-
erate concentrations inhibited further development, and heavier dosages
proved fatal to the adults. Thus a potentially powerful weapon against
the further spread of typhus in the world — a possibly invulnerable
weapon which may in time be effectively used also against other men-
acing insect carriers of disease, a weapon which fifteen years ago would
have seemed only generally predictable — has become concrete. That
achievement is surely another major milestone both in biological re-
search and in the control of disease of the year just past.
A keen interest in the nature and workings of the human brain — and
in what has perhaps too often been assumed to be an exact equivalent
of it, the human mind — must be older than civilization, as the numerous
trepanned skulls found in the ancient graves of many lands bear silent,
pointed witness. But it is only within this half century that more
sophisticated biological understanding, the newer and exquisite tech-
niques of neurophysiology, the discoveries of the actions of brain-influ-
encing chemicals applied generally or placed at precise sites on brain-
stem or cortex, and the greatly increased scope and refinement of many
of the modern diagnostic techniques have given us a feeling of increased
promise in that field.
In the last decade, the Age of Computers has affected the nature of
questions asked and the kinds of research undertaken to elucidate them
quite as much in the field of neurophysiology and brain research as in
functional and biochemical genetics, though frequently in very different
ways. We speak often now of computer "communication," for example
— among computers and between computer and operator. Programmed
REPORT OF THE PRESIDENT 15
"languages" for computers, in turn, have become more and more re-
fined and versatile, conveying new degrees of range and precision. To-
morrow it may well be that information can, quite literally, be trans-
mitted with the speed of light, traveling along the coherent beam of a
laser. And no characteristic of the computer that suggests the action of
the brain has been developed more dramatically in recent years than its
"memory." Computer capacities for the storage of information used
to be measured in thousands of bits. Now they are measured in millions,
or even in thousands of millions. Perhaps it is this spectacular advance
in the memory capabilities of computers that has particularly fostered
the idea that machines thus endowed, immensely inferior to brain mem-
ory though we know the endowment to be in a quantitative sense, could
nevertheless by their operations shed some light on the specific nature
of human recollection.
Such an analogy, however, must be regarded with reserve. It would
not be argued that computer coding is a precise analogue of genetic
coding. So it may prove in the case of computer versus human memory,
and to an even greater degree. For the brain and the computer, in last
analysis, are very different things. As Donald MacKay has pointed out,
a primary challenge to research on the nature and the function of the
brain is posed by the very fact of its structural and functional com-
plexity: a structural and functional complexity unexampled elsewhere
in the living world, except perhaps among the great conurbations of
our time. And since even in the greatest urban centers we do not yet
deal with individual aggregates of citizens numbering ten billion or
more, while the human brain does include approximately that number
of individual neurons in its compass, the brain may well be the most
staggeringly complex biological organization in the world. To approach
such a subject with some hope of ultimate real understanding requires
even more than newer, more powerful, more searching techniques.
Over three centuries ago, Descartes was tempted to look at the
mechanisms of body and mind primarily in terms of the clockworks
that were then so novel and so exciting. A decade or more ago, when we
were just beginning generally to appreciate the new and spectacular
developments in telephonic communication, we were constantly tempted
to compare the brain and the nervous system with a telephone network
and a telephone exchange. The comparison clearly stimulated valuable
investigative work on the nature of nerve and brain. But in the end, of
course, it proved much too narrow. So may the comparison ultimately
prove with ideas of "computer thought" and "computer memory," if
conceived in terms of specific models rather than of purely functional
16 CARNEGIE INSTITUTION
parallels. This lesson may be particularly pointed with respect to mem-
ory. For, as MacEay has pointed out, it is the act of remembering, the
process taking place, that we seek to understand, and this may be a
somewhat different goal for the investigator from that of seeking evi-
dence for the physical existence of a record, in the sense that one would
seek it in the storage bank of a computer. This subtle distinction may
in fact be extraordinarily profound. It has been illustrated with unusual
vividness by developments of the year, involving a mode of investigation
that has produced some of the most suggestive findings relating to the
nature of biological memory that have been reported.
Not long ago it was discovered that when mice, trained to a maze
until they had learned it thoroughly and could readily recognize and
again successfully thread it a considerable period after the training,
were injected intracerebrally with the antibiotic puromycin, they suf-
fered apparently complete and permanent oblivion of that training
within ten to twenty hours after the injection. The loss of memory for
a particular routine persisted for at least three months after the treat-
ment. The same effect was later fully confirmed with goldfish similarly
trained to a maze.
Puromycin, a naturally produced antibiotic, powerfully inhibits pro-
tein synthesis. From the beginning this was thought to be a decisive
factor in its action — indeed, this was the property which, on theoretical
grounds, led to its initial selection for the experiment. But puzzling
features appeared. Acetoxycycloheximide and cycloheximide, for ex-
ample, also markedly inhibit protein synthesis. Yet they appeared to be
without effect on memory when similarly injected. It was also later
found that cycloheximide, and also chloramphenicol, could actually pro-
tect against the destruction of memory by puromycin.
The modes of inhibition of protein synthesis of puromycin on the one
hand and of acetoxycycloheximide and cycloheximide on the other are
quite different. Acetoxycycloheximide is known to impede the transfer
of an amino acid from RNA to polypeptide, thus hindering protein
building. Puromycin, however, inhibits protein formation by interfer-
ing with the growth of polypeptide chains in quite a different way. The
absence of an inhibiting effect on memory by cycloheximide, together
with its property of protecting against the inhibiting effects of puro-
mycin, invoked a more complex picture. The new hypothesis suggested
that initial changes within the brain accompanying the fixation of mem-
ory might involve corresponding changes in the amounts of one or more
types of molecules of messenger RNA, altering the rates of synthesis of
one or more proteins essential for memory expression, while these pro-
teins, or possibly their products, might in turn act as inducers of their
related messenger RNA. Thus expression of memory would depend on
REPORT OF THE PRESIDENT 17
changes in protein concentration brought about by corresponding mes-
senger RNA, changes in turn modified by the experience of learning.
Loss of either protein or the corresponding messenger RNA could thus
upset the cycle, with presumed loss of memory. Since puromycin in con-
centrations comparable to those injected is known to inhibit the syn-
thesis of RNA in the cells of vertebrates, this seemed a reasonable
explanation of its memory-destroying power. By the same token, the
protection which cycloheximide is known to give messenger RNA could
account both for its ineffectiveness in destroying memory and for its
memory-protecting action against puromycin.
This reasonable and consistent hypothesis was challenged by a new
finding during the past year. A series of experiments was undertaken
to determine how long puromycin must remain in contact with the
brain to produce complete and long-lasting amnesia. Saline solution was
injected after puromycin treatment, diluting and finally washing out
the drug. Unexpectedly, when the puromycin was removed, the memory
record which had apparently been destroyed proved to have been only
blocked, and was restored. It became evident that mechanisms quite
different from those envisaged earlier must be implicated.
Shortly thereafter results of another approach to the same problem
were described, and a new mechanism was suggested. Recordings of the
electrical activity of the mouse brain taken from the hippocampal region
a few hours after the injection of puromycin gave evidence of the occur-
rence of nervous convulsions which were frequent, but not so severe as
to affect the overt behavior of the animal. When similar recordings were
made from the brains of mice that had received injections of cyclo-
heximide or saline, however, or when cycloheximide injections were
combined with those of puromycin, such abnormal electrical activity
was much attenuated or absent. A conclusion that could be drawn from
these new findings is that the effects of puromycin in producing mem-
ory loss are related primarily to induced abnormalities in the electrical
activity of the brain, involving in fact subclinical seizures; and that
failure in functional recall of the memory store, rather than loss of a
specific memory trace, was the prime factor involved.
Time may show that this picture is not as different from nor so
incompatible with the earlier concepts as might at first appear. Yet,
set in opposition, the two interpretations provide a vivid example of
the kind of contrast between viewpoints emphasizing structure and
viewpoints emphasizing process that may well form a major theme in
our approaches to the brain, and perhaps to biological entities of com-
parable complexity over a wider range, in the years to come.
18 CARNEGIE INSTITUTION
Superficially, the past year seems to have been one of sharp contrasts
on the scientific scene. There has been the satisfaction and the sense of
genuine cultural gain and cultural certainty conveyed by the striking
substantive advances on a multitude of frontiers of scientific discovery
and application typified by these few but representative examples. But
there have also been the prominent difficulties and doubts and uncertain-
ties attending the social scientific concerns of the same year, as con-
comitants of their weight and significance. We have lived with these
latter problems for some years — though they have rarely been so mas-
sive— and they will surely be with us in even more acute form far into
the future. They are concerns of great national import. For not only
do they involve relatively vast sums of public money but, more im-
portant, they deploy great numbers of men and women who, potentially
or actually, are of special significance to our society. In contrast to the
uncommitted ends of purely scientific discovery, the major orientation
of these concerns usually is, as in its very nature it must be, toward
goals more nearly preconceived, and often biased primarily to other
ends than those of scientific discovery per se, or of cultural advance-
ment in the strict sense of cultural content.
It is difficult a priori to imagine a sharper contrast in purpose or
design, or approach, or requirements than that demarking the pursuit
of scientific discovery from the use of the total scientific enterprise as a
new and extremely powerful tool of social change. Yet the two enter-
prises, vividly contrasting as they are in many aspects, not only exist
together but are in fact indissolubly linked. On the one hand, the shape
and the magnitude of exploratory science in our time, for better or for
worse, are deeply affected by the massive deployment of science to social
ends that is so commanding a feature of our day. On the other, science
is, by its very nature, one of the extremely powerful forces for social
change and social molding with which we live. To fail to ponder the
quality of its impact is to be irresponsible. The most important task of
science in our day, in fact, may be to meld these two aspects of the
scientific way without damage to the precious core on which all effec-
tiveness must ultimately rest — the philosophy and the continuing oppor-
tunity of independent, uncommitted scientific exploration. Perhaps the
greatest challenge for our time in this field is to understand truly new
dimensions without discarding or even failing to maintain in pristine
brightness the precious values that we inherit and that are basic to the
significance of science in society. One aid on the long and difficult road
of understanding such relationships and of protecting and conserving
these precious values while bringing them to bear on the overriding
needs of a changing society may be to look on these linked aspects as
both contemporary and sequential stages in a vast social evolutionary
REPORT OF THE PRESIDENT 19
process which exhibits some of the same characteristics as do evolu-
tionary processes elsewhere in nature, and to deal with them in that
light.
So it may be pertinent to recall that it was in February, just one
hundred and ten years ago, that across the mind of Alfred Russell
Wallace, as he lay racked with a bout of severe intermittent malaria at
Ternate in the Moluccas, "there suddenly flashed . . . the idea of natural
selection." During the rest of that bout of ague he thought the notion
through, and made a preliminary draft of it the same evening. During
the following two evenings he composed the famous letter to Darwin
which prompted the recipient, when he read it on June 18, to write on
the same day to Sir Charles Lyell : "I never saw a more striking coinci-
dence. If Wallace had my ms. sketch written out in 1842, he could not
have made a better abstract ! Even his terms now stand as the heads of
my chapters." This was the essay that, with a summary of Darwin's
views, was read as the joint paper before that memorable meeting of the
Linnean Society on July 1, 1858, when the theory of organic evolution
was first officially laid before the world.
Through the century since, we have gradually come to realize how
fundamental were the principles enunciated at that meeting of the
Linnean Society, and how generally they apply in living nature — well
beyond the horizons that could have been even dimly visioned by a
Darwin or a Wallace, or by a generation of their followers. Among
living organisms we are now finding that the processes that Darwin and
Wallace elucidated are displayed at far more elementary levels of
biological organization than those at which they were first observed. At
microscopic and submicroscopic levels within a single cell, indeed, the
evolutionary aspects of living systems are displayed in exquisite detail,
in realms of nature where the very scope of knowledge of Darwin's time
made more than a vestige of appreciation impossible. Even so, the
astonishing circumstance is clear today that the genius of Darwin did
lead him to suspect something of this in the realm of genetics. In that
critical year of 1858 he wrote to Huxley: "Approaching the subject
from the side which attracts me most, viz. inheritance, I have lately been
inclined to speculate, very crudely and indistinctly, that propagation by
true fertilization will turn out to be a sort of mixture, and not true
fusion, of two distinct individuals, as each parent has its parents and
ancestors. I can understand on no other view the way in which crossed
forms go back to so large an extent to ancestral forms." In that
pre-Mendelian age, dominated by notions of blending heredity to which
Darwin himself subscribed in practice, it was a moment when a pro-
20 CARNEGIE INSTITUTION
phetic vision captured one fleeting glimpse of the particulate quality of
inheritance and of the nature of the germ plasm. But not even that
genius, nor yet the insight and the investigations of more than two
subsequent generations of geneticists, were able to bring fully to light
what is so clear today.
It would surely have delighted Darwin immeasurably that evolu-
tionary processes can be seen at work just as surely and just as typically
in the subtle realms of genetic form and structure and genetic mecha-
nism as at the grosser levels of morphology in the organisms whose
heredity they mediate. The early evolution of sexuality in nonsexually
reproducing primitive organisms and the various modes of its determi-
nation, the later reverse evolutionary courses in which bisexuality has
been discarded, culminating in populations of advanced partheno-
genetically reproducing organisms that have appeared again and again
in the history of life in adaptation to special conditions of the environ-
ment ; the structural evolution of the chromosomes themselves from less
precisely disposed genetic material, and the evolution of the complex
and finely regulated processes by which they are proportioned at mitosis
in the division of body cells and between new individuals in the mecha-
nisms of meiosis; the biological "management" by which diversity is
secured in parallel with continuity; the selection of mutations and the
perpetuation of advantageous ones; the manner in which such "acci-
dents" to the chromosome as deletion and translocation are repaired or
put to adaptive ends; the evolution of the diploid from the haploid
chromosome condition and the evolution and adaptive "use" of poly-
ploidy; the development of the controls for the expression of new char-
acters in the plant or animal body with its accompanying evolution of
dominance or recessiveness, and the evolution of sex-limited and sex-
linked inheritance — all these circumstances strikingly repeat and illus-
trate at new levels the grand features of the process. They are being
even more explicitly demonstrated today at the still finer level of
molecular structure. Pioneering studies of the apparently excessive
replication of particular genetic loci in the evolving chromosomes of
some higher organisms, described elsewhere in this Report, emphasize
one aspect of this development. Another is represented by striking
findings concerning the mechanisms through which the effects of partic-
ular genes are reinforced or inhibited or controlled in their appearance
during development by other genes or other chemical entities of molec-
ular dimensions — an approach first conceived some two decades ago by
Dr. Barbara McClintock at the Carnegie Institution, dramatically con-
firmed in bacteria by Jacob and Monod at the Pasteur Institute, and
now so widely recognized in living things that the study of regulator
REPORT OF THE PRESIDENT 21
genes and their modes of action is a heavily populated field of investiga-
tion at the frontiers of biology today.
In the classical case discussed by Jacob and Monod, the product of a
regulator gene, called a repressor, acts by switching off a highly specific
target gene. The control is therefore negative. It might be brought about
directly by the repressor product. Or it might be achieved by an inter-
action of the repressor substance with messenger RNA to block the
translation of the genetic message into specific protein.
During the year investigators have isolated such inhibiting sub-
stances produced by specific repressor genes. It has been possible to
study the action of one of these, the protein made by the Ci gene of
lambda phage, which acts to keep the phage chromosome dormant within
its colon bacterium host until the appropriate time for growth. From
experiments conducted at molecular level it has been possible to demon-
strate that the control is indeed achieved by the direct "occlusion" of a
specific site in the DNA by the repressor substance. It is becoming more
and more clear that actions of this highly specific kind at molecular level
must be responsible for the deployment in time of all the precisely
patterned and adjusted sequential developments in the growing em-
bryos of many-celled plants and animals. The evolution of the molecular
means by which such adjustments are secured displays an intricacy and
a regulated precision to dwarf all that we have known before. Yet it
bears, unmistakably, the marks of the same basic evolutionary processes
that have so impressed us, on the stage of the organism as a whole, since
the days of Darwin.
Even at molecular levels where only the components of life, rather
than the whole living organism are directly involved, evidence of the
exquisite patterns of self -regulation imposed by the requirements of
adaptive evolution are becoming evident. Already the three-dimensional
structures of some biologically critical enzymes have been or are being
completely worked out — the structures of lysozyme, ribonuclease,
chymotrypsin, carboxypeptidase (a protein that digests other pro-
teins) ; of the respiratory enzyme cytochrome c, and the plant digestive
enzyme papain. Max Perutz has recently pointed out that as the modes
of action of such biocatalysts become clear they may prove to be the first
substances of any sort in which the mechanisms of catalysis are fully
understood, although industry has depended heavily upon catalysts in a
myriad of processes for many years.
Even among these biologically vital, and yet nonliving, molecular
entities, self -regulation is extraordinarily precise, suggestively parallel-
ing the self -regulatory mechanisms that evolution has imposed on living
things. Consider, for instance, the cycle through which cytidine tri-
~V CARNEGIE INSTITUTION
phosphate, one of the building blocks for the manufacture of nucleic
acids, is formed. The starting material is aspartic acid or aspartate,
and the catalyst promoting the first step in the five-stage process is the
enzyme aspartate transcarbamylase. The first molecules of aspartic
acid or aspartate that bind to the enzyme change its structure some-
what, making other binding sites on the enzyme more receptive for
aspartate molecules than before and thus accelerating the conversion of
the raw material to the first intermediate product, carbamyl aspartate.
As the reaction proceeds toward its terminal stages and cytidine tri-
phosphate begins to accumulate, however, the final product too be-
comes attached to the enzyme. This sharply diminishes the enzyme
activity, perhaps by competing with the aspartate for binding sites.
Thus, so long as molecules of cytidine triphosphate are accumulating in
excess — that is, so long as the factory for this particular structural
block of a nucleic acid is producing more than is immediately required —
the aspartate transcarbamylase is a less effective catalyst and the
supply of carbamyl aspartate is reduced. Thus the whole manufactur-
ing process is occluded at its beginning. Conversely, as the surplus of
cytidine triphosphate is used up in nucleic acid manufacture and its
concentration in the medium falls, the supply of free carbamyl aspar-
tate is replenished from the population of bound molecules. The reaction
picks up again, until, perhaps, it gets out of step with the general
process once more. In the truest biological sense, this is a mechanism for
the maintenance of homeostasis by negative feedback from the reaction
product, closely simulating living processes and indeed fulfilling the
requirements of such processes.
This is but a single example — and by no means one of the most com-
plex— of a number of self -regulating enzyme systems that have been
analyzed, governing with fine precision an array of biological processes
in living systems. It would be hard to find more dramatic illustrations
of evolution at the most elementary level in the structuring of life where
it has been investigated, a level indeed but newly opened to us. To be
sure, most of the actual steps in such evolutionary processes remain to
be specifically demonstrated, though a few have already been detected
experimentally, such as mutations in the details of allosteric reactions
of enzymes governing metabolic syntheses in certain bacteria. But
whatever the specific paths through which enzymes evolved to govern
the process of life, it seems highly probable that the earliest lacked reg-
ulatory control mechanisms. Once such an enzyme had become firmly
incorporated in an organism, however, and had become indispensable to
its metabolic economy, then the "design" of special active sites on its
molecule, and the selection of the most efficient of these to regulate the
catalyst, must have followed the typical courses of evolution, closely
REPORT OF THE PRESIDENT 23
coupled with the evolution of its "host." In recent years, the discovery
and elucidation of this new setting for biological regulation within the
biological molecule itself has opened vistas that, in their demonstration
of the ubiquity of the processes of evolution through every level of life,
would surely have brought renewed wonder and delight to a Darwin
and a Wallace were they living at this hour.
From this panorama of evolving biological systems, certain common
features of the evolutionary way, at whatever levels and in whatever
contexts they are played out, strike the observer with special force. One
of the most fundamental features, conspicuous at every level, is the
universal circumstance that such evolutionary courses have evidently
unfolded in relatively small successive steps, each new step profoundly
influenced by preceding ones and by all the surrounding requirements
for survival, and each step exquisitely meshed with the total situation.
This is evidently a primary requisite for success and indeed for per-
sistence: the great saltations that some earlier students of animal
evolution expected to find have yet to be satisfactorily demonstrated.
Each new limited change is firmly embedded in the matrix of a prior
context, so that the total pattern of evolution in effect develops as a
mosaic, of which the greater proportion is conservative and older in
design, the smaller, radical and new. But characteristically the older
elements too were sifted earlier in their turn, and selected for adaptive
value with equal rigor. Those that have been tenaciously retained have
been well proven. Great evolutionary changes, for example, have obvi-
ously occurred among the vertebrates since the rise of early fishes. Yet
in neither amphibians nor reptiles nor birds nor mammals nor man have
the great structural principles of the vertebrate eye or the spinal col-
umn or the central nervous system been discarded. Rarely if ever in
successful organic evolutionary change has the baby been thrown out
with — or after — the bath.
Another striking and fairly universal quality of evolutionary change
is that, in order to be successful, the evolving systems must sustain a
tight-woven network of communication permitting the most intimate
and sensitive continuing interactions among their components, whether
at the chemical level, or at that of intracellular organelles, or of cells in
a many-celled system, or of whole organisms in a biological society.
These are qualities of biological evolution to be remembered.
Ever since the days of Herbert Spencer, observers have been struck
by the fact that human societies frequently seem to resemble biological
systems in some aspects of both evolution and structure. The mecha-
nisms involved, of course, have been quite different, as has been pointed
2J, CARNEGIE INSTITUTION
out repeatedly over the years. The capacity to react sensitively to in-
formation, to conserve, communicate, and use it, and constantly to
accumulate new stores is central to successful evolution in either con-
text. But the biological mechanisms of accumulation, selection, use, and
transfer of information differ in obvious and extremely important
practical ways from the mechanisms involved in the accumulation, use,
and transfer of social information by the processes of learning and
through the medium of culture. The potential for what is in effect the
inheritance of acquired characteristics confers upon the evolution of the
culturally mediated society a dynamism, a flexibility, and a versatility
that the biological organism or system can probably never remotely
match. The obvious penalty attached to this versatility and flexibility,
of course, is a degree of vulnerability also seldom matched in the biolog-
ical world, as the disappearance of past great civilizations bears silent,
poignant witness.
But these differences of mechanism, conspicuous and indeed basic as
they are, should not obscure the parallel requirements for success that
face the evolving society and the evolving organism. There are a number
of such similar requirements. They are characteristically elementary,
and by that very token may be especially noteworthy. Thus, precisely as
in the biological world, there is a sharp limit to the rate at which evolu-
tionary change can take place in social evolution, and to the magnitude
of any given step, if the innovations are to be successful and if the
society that they will affect is not to be severely disoriented or even
crippled by the process. As in biological evolution, effective social evolu-
tion must be at once radical and conservative, freely embracing the new
yet scrupulously preserving basic and well-tested elements that have
had a high survival value in the past and which remain relevant to the
present. Like successful biological evolution, successful social evolution
must constantly guard against discarding the essential with the trivial
— against throwing the baby out with the bath.
The more rapid the social evolution, the more imperative becomes this
requirement and the more vigilance is demanded to satisfy it. Through
history this has been one of the most difficult lessons for man to learn.
Repeated failures to understand it or to act upon it adequately must
have accounted for major historical debacles — and indeed for major
distortions of social evolution. It is a danger if anything more acute
today than ever before, because of the immensely increased dynamism
of social movements and evolutionary forces with which we live, and
because the stakes of failure, like those of success, are so much higher
than they have ever been. It would be difficult to find a more apposite
general caveat for our time than this of exercising due care that, in
embracing new and experimental courses on myriad fronts of move-
REPORT OF THE PRESIDENT 25
ment with the ardor that we must, we do not at the same time discard
long-tested values and long-tried adaptive courses which, if they are
lost, will only have, one day, to be re-won — and probably at enormous
cost. This is a consideration that is before us in all our affairs, every
minute of every day.
Finally, it is very clear that, as in biological evolution but to a yet
higher degree, successful adaptive evolution within any society can only
be accomplished in the presence of as complete a nexus of communica-
tion among all parts of the social structure as can possibly be secured.
Only if superior communication prevails can the highly pluralistic social
structure of today be constantly adapted to the selective changes con-
fronting it in the many-sectored environment of society. In the realm of
science, this requirement becomes more difficult, and demands more and
more effort and ingenuity to satisfy, as the volume of scientific litera-
ture burgeons and the ranks of those committed to the scientific way
increase ever further.
With the delineation of these general parameters, we return to issues
raised at the beginning — the vital, challenging, intimately interlocked,
and often also frustrating problems of science policy in American
society, with all their potential opportunities, which events of the
current year illustrate so well, coexisting with the rich harvest of more
evident and satisfying individual conquests at substantive scientific
frontiers that the same year has provided in such impressive measure.
How are the two related?
Throughout the history of science, as John Steinbeck has remarked,
parts of the universe of experience have been rediscovered, redescribed,
reclassified. Again and again an investigator of genius, gazing at the
same world as his predecessor, has lifted a new corner of the veil of
incomprehension and has glimpsed a new scene. In the excitement of the
fresh vision there has typically been, at first, skepticism and controversy
among those to whom the vision has not yet been granted. This, perhaps,
is the most bracing and dynamic and stimulating period. Then, if
acceptance follows, a band of believers slowly accumulates, extending
the new ideas and confirming them in added ways. Gradually this group
of disciples becomes more formalized, and the typical steps of post-
discovery follow — evaluation, taxonomy, rearrangement in the new
world of experience — until, a half generation or so later, a new struc-
ture has supplanted the old. Then it is that the ancient danger, against
which the whole vista of organic evolution warns, becomes particularly
acute. As Thomas Huxley emphasized many years ago, it is the common
fate of knowledge to start as heresy and end as superstition. Character-
■26 CARNEGIE INSTITUTION
istically it is the band of later believers — not the first discoverers — who,
as the new system becomes codified and orthodox, forget the invariants
in the evolution, the older common factors long proved viable that are as
essential in the new structure as they were in the old, and that should
provide a bond between the two. In their enthusiasm for the newly
orthodox, they may press to discard all the features of the older frame:
to start wholly afresh. Ultimately, as patternmaking proceeds in the
new conceptual world, these older invariants, so lightly discarded in the
period of change, typically emerge once more, to reoccupy in the new
system the central position that they held within the old. But time and
expensive effort will have been lost. How much more economical it might
have been had their invariance been earlier recognized!
One such invariant, in the realm of the creation of new knowledge,
stands out particularly when the substantive scientific gains of the year
are viewed against the wider background of the whole technological
activity of the nation and its relation to public policy. Every one of the
substantive advances of the year that has been cited — and many more
could be added — has typically been the work of a single investigator, or
of a small group. None has been achieved by the specifically united
efforts of a massive team — though it has many times been proven that a
close-knit research group can effectively breach the frontiers of new
knowledge when embedded within a large organization, provided that
its own integrity be recognized and rigorously protected.
The exploitation of the implications and the consequences of new
knowledge once won and its application to the purposes of society, may
be, and commonly is, truly the organized communal work of thousands.
The creation of that knowledge is typically the province of a compara-
tive few, dedicated to the life that Newton epitomized when asked how
he made his discoveries : ". . . by always thinking unto them. I keep the
subject constantly before me and wait till the first dawnings open little
by little into the full light." This has been the history of advance in
science since the beginning of science itself. Once, when scientists were a
small group in the society, the work was characteristically that of a
single individual — the Newton, or the Pasteur, or the Michael Faraday.
Today, with the number of scientists so vastly greater, similar pioneer-
ing is more apt to be accomplished by bands of close associates, often
working in parallel with others, at the center of that wider network of
communications that the successful evolutionary pattern demands here
as elsewhere, but themselves small, defined units where flexibility and
the opportunities for intimate, moment-by-moment exchange of in-
formation are at a maximum. This structuring of the research group,
clearly, is an essential core in the evolution of the creation of new
knowledge, and of new viewpoints about the world, which we neglect or
negate at our peril.
REPORT OF THE PRESIDENT 27
Yet the temptation to forget the importance of this invariant is
surely greater today than it has ever been. There are powerful reasons
for this. The acts of creating new knowledge in science on the one hand,
and of consolidating that knowledge and using it for the purposes of the
society on the other, are typically and properly conducted in close con-
junction and form a continuous spectrum. Thus, to the casual observer,
they appear to merge, and to be essentially alike in character and re-
quirements. In fact, at the extremes of that spectrum, genuinely but not
often obviously different processes are invoked which require somewhat
dissimilar approaches. They are commonly shaped by significantly
divergent philosophies and are likely to be carried out by men and
women of rather dissimilar orientation, although the capacities and
training of those men and women may be very similar and despite the
fact that some are able to work in both environments and to discharge
both roles successfully.
Now the instruments for the creation of knowledge, which have re-
mained constant in their basic features over several generations, must
operate effectively today in an environment where the means for the
marshaling and exploitation of that knowledge, once gained, are evolv-
ing radically and rapidly and where experimentation with new forms is
proceeding with unprecedented speed. In this context, science, for indus-
trial nations the world over, has come into a relationship with society
almost precisely opposite from what it was when the mechanisms for
the creation of new knowledge were themselves first evolving — when the
pioneering work of a Galileo or a Leibnitz created little stir at the time
of its accomplishment beyond very limited professional and intellectual
circles. Today the products of this same science have become major
factors in the policies of nations. As Sir Eric Ashby has vividly ex-
pressed the problem, in the broader setting of the whole environment of
universities: "Formerly they (the universities) were detached orga-
nisms, assimilating and growing in accordance with their own internal
laws. Now they have become absolutely essential to the economy and the
very survival of nations. Under the patronage of princes or bishops they
were cultivated as garden flowers, of no more significance to the state's
economy than the court musician. Under the patronage of modern gov-
ernments they are cultivated as intensive crops, heavily manured, and
expected to give a high yield essential for the nourishment of the state."
In such a climate, the protection of the essentials of the research envi-
ronment is both more difficult and doubly demanded.
Ours is a society, originally heavily dependent upon technological
innovation for its very existence, which has now become equally depend-
ent upon the continual creation of new knowledge for its survival, and
C£ CARNEGIE INSTITUTION
which we may truly style innovative in its basic character; a society
in which the costs of some kinds of research have already mounted to
the point where they can be borne only by the Federal Government, and
threaten soon to exceed even that resource; a society in which the twin
streams of the creation and the use of knowledge are now thoroughly
merged. It is an egalitarian society, moreover, where, as de Tocqueville
long ago predicted, free and voluntary associations of citizens flourish
and multiply in every sphere of activity — and ought, as de Tocqueville
said, to fulfill the functions typically assumed by autocratic individuals
or an autocratic bureaucracy in more rigidly structured situations. How,
in such a society, are we to mesh the invariants that are so vital to the
continued creation of new knowledge with the overarching considera-
tions of common purpose that are now so deeply affected by the trends
and developments of science itself? How are we to do this while main-
taining the integrity and independence of the forms and of the very
institutions which are clearly essential to the creation of new scientific
knowledge? How are we to protect, and fully tap, those clear springs
from which high excellence flows — to maintain the sources and to hold
high the selection and training of that precious human reservoir of
extraordinary quality, in science as elsewhere, which has always pro-
vided the ultimate guarantee of our survival — within the massive set-
ting of a technological, pragmatic society of immense power and
resource, whose vast capacities shape these very sources and reservoirs?
And conversely, how is science itself, pragmatically considered, best to
be supplied the most effective channels through which to serve the
national purpose?
To this last question, of course, there is a prior one, itself not yet
clearly answered. How are we to define that national purpose and for-
mulate it coherently with respect to science? How far we are from any
agreed answer to this most basic and difficult of questions the events of
the past year in the arena of publicly supported science give vivid
evidence. As far as any general consensus on the whole social position
and significance of science goes, we are surely living through times as
inchoate and contradictory as any that our society — or perhaps any
society — has breasted, in a world itself grown far more complex and
subtle.
But if our time poses totally novel problems in this sphere, it is also
surely a time of the richest opportunity. Given suitable leadership and
understanding, given a steady and consistent will to surmount the pre-
occupations and the special, short-range interests that distract us, there
is no reason why we should not reap benefits from the interwoven con-
cerns and objectives of science and society in the years to come to dwarf
what wre have known. But the effort will require even more than capable
REPORT OF THE PRESIDENT 29
leadership, and even more than continuing vision to rise above imme-
diate problems. It will require the will and the capacity constantly to
innovate in the field of social mechanisms, constantly to devise and to
try out new patterns, and to discard quickly and decisively those which
prove unsuitable to our purposes.
In this effort, of course, those concerned with the future of science
and public policy — and indeed with science as public policy — are only
called upon to do precisely what America, from the very beginning of
its history, has always been peculiarly adept at doing in a more general
frame : to give full rein to our national talent to be constantly inventive,
and to foster highly pragmatic and selective use of our inventions. In-
deed, the whole challenge to the scientific way in this context offers a
paradigm for the similar challenge confronting the broader intellectual
community of the nation, a challenge epitomized by the whole complex
of the problems facing the universities and indeed of the actual future
role of the universities in American society that so vex us today, but
also extending well beyond this milieu. It is this aspect of paradigm for
a still larger but closely related national problem that lends to these
questions of science and the national policy, of the antitheses between
excellence and equality, of the adequate protection of environments and
modes which we know to be absolutely indispensable to creative thought
in any field, a range and significance even greater than that assured by
their intrinsic importance.
In an economic message to Congress in 1963, President Kennedy esti-
mated that two thirds of the trained people in the nation available for
scientific and technical research were involved, in one way or another,
with the defense, space, or atomic energy programs of the nation. Is this
the optimum distribution of our talent? Or should we devote more
specific attention than we have done to diversification and the means of
bringing it about? Are we making the wisest choices in seeking to
partition our human scientific resources between more highly focused
governmental and more plural civilian enterprises, in the best possible
interest of the ultimate strength of the country? It is a striking circum-
stance that, though Japan spends only about nine tenths of one per
cent of its gross national product on research and development, its
annual rate of economic growth has approximated seven per cent, while
we, who spend more than three times as high a proportion of our own
far greater G.N.P. — nearly three per cent — on research and develop-
ment, are unable to nearly match this growth figure. Does the contrast
to any degree reflect the fact that in our case only about one tenth of this
three per cent, or a total of about three tenths of one per cent, is devoted
SO CARNEGIE INSTITUTION
to our civilian sector, while a large share of the much smaller total
outlay in Japan is so committed? And if there is a real connection, do we
in fact have any other option, given our situation in the world? Is the
case of Japan instructive, or only a special one?
Many more questions might be put within this particular frame of
reference. But the framework itself illustrates with remarkable vivid-
ness a central circumstance confronting alike the specialized areas of
scientific research and the broader intellectual structure of our nation
as a whole, epitomized by the problems facing American universities in
our time. As early as the days of the Manhattan Project it had become
abundantly evident that the expenditures required to carry forward
some kinds of research and development closest to the national interest
were so enormous that they could be met only by the Federal Govern-
ment, and eventually might well exceed even its capacity. The penetrat-
ing estimates of Dael Wolfle earlier cited clearly demonstrate not only
that the aggregate costs of conducting research today in many areas are
well beyond the means of private institutions, but that limitations are
evident even at the Federal level. It may well be that we shall conclude
in the future that some scientific undertakings can be carried forward
only with international participation. In any event, it is already very
clear that such enterprises will permanently require the decisive sup-
port of the Federal Government, and that on a huge scale. Yet, as events
of the year have demonstrated so vividly, discoveries of major impor-
tance on some of the most significant and exciting frontiers of science
may still be achieved at very modest cost, in terms both of money and of
numbers — but not quality — of investigators. This striking paradox is
to be remembered.
For many years now, science has faced the extremely challenging
question of how to advance and guide major projects supported and
operated largely by government — of how to inject scientific judgments
effectively into what must be, in essence, political domains. It is not an
easy task, nor a unidimensional one. It is amply evident that one ele-
ment urgently required is advice and assistance from a multitude of
hands and minds of a wide range of capacities and preparation and
experience, operating at least as much in the private sector as in govern-
ment but bound by a sense of common concern into a diversified and yet
distinctly identifiable whole. In time of full-scale conflict, as in World
War II, when an overriding common peril was universally felt and
clearly identified, when the need for swift and effective action was para-
mount, when the vastly greater part of Federally supported research
and development was sharply oriented to a single unequivocally vital
mission, the direction of such an effort, the mapping of its ends, the
deployment of those engaged in it, constituted a relatively unified mis-
sion, primarily determined by the Executive branch of government.
REPORT OF THE PRESIDENT 31
Even under these circumstances, however, the injection of the scientific
viewpoint at Federal level, the coupling of a scientific effort overrid-
ingly in the national interest and conducted on a national scale with
adequate protection for the modes of operation indispensable to the
productivity of science and scientists, required nothing less than genius
in those who accomplished it. Yet demanding as that task was, the prob-
lem may inherently be yet more difficult in times like the present, when
the effort is not only larger but far more pluralistic, when its goals are
more diffuse, and when it is far more a daily concern of the whole
American people.
Under such conditions, it is in the Congress of the United States as
much as in the Executive branch that the ultimate responsibility for
determining the magnitude and direction of a large part of the Federal
scientific effort properly rests. This circumstance imposes on the Con-
gress a truly awesome responsibility. And it is not one for which Con-
gress has traditionally been particularly well prepared, because hitherto
so few of its members have dealt extensively with scientific issues before
assuming national responsibilities. The year's record of debate on
locating the accelerator, mentioned earlier, illustrates the difficulties
well. Yet progress is surely being made. On the side of government,
many members of Congress have taken great pains to understand the
subtle requirements of the scientific way, to make the acquaintance of
scientists and to listen to their views. On the side of science itself,
various groups especially interested in policy have sought to aid the
Congress, as similar groups earlier took the signal step of serving as
advisors for the Executive.
Science still has a long way to go in these matters. But imperfect as
its leadership now is in this vital sector of our national life, it is indeed
real leadership, and may well prove significant in an even wider theater.
For the experience of science in this field has been longer than that of
most other segments of the intellectual life of the nation. And there are
close parallels between the problems that science has already encoun-
tered and those that currently face the universities as such in basically
similar contexts. There is the same acute financial need that has
brought the Federal Government, vis-a-vis the universities as vis-a-vis
science, into the position of a major arbiter of social change. There is
the same urgency for the institutions that articulate the intellectual
nexus of the nation as a whole, to present a coordinate front in force-
fully representing both the requirements and obligations of their way
of life, and in explaining their proper functions to the patron that they
share with science — the body politic of the nation — who, as in the
special case of science, may sorely need information of this kind. As
with science itself, this is nowhere more vital than in shielding creative
university environments from demands for the execution of too spe-
CARNEGIE INSTITUTION
cifically oriented missions originated from without, and for the per-
formance of particular services which could be better or more appro-
priately rendered elsewhere. There is the same critical national need for
continuing advice from the best minds in the fields of the humanities
and social sciences, whose primary commitments lie outside the sphere
of government, that there is in the case of natural scientists similarly
placed. How great that need may be has been vividly demonstrated
recently in the debacle of Camelot, which surely could hardly have
urred if the most cogent advice in relevant fields had been appro-
priately marshaled. In all these areas the needs, the vulnerabilities, and
the unique potentials of science parallel remarkably those of the rest of
the intellectual nexus of the nation.
In a wider assessment, too, the problems and duties of the community
of science and of the community of universities in the arena of national
policy show remarkable parallels. Each community shares at least three
common relationships with government. Research must be supported
without damaging the research climate; services must be provided to
the government and limits on such services must be set; and, perhaps
most difficult of all, advice must be given by qualified professionals who
can think deeply and prophetically about the future to colleagues on the
fronts of action — colleagues who are far too completely engaged with
the heat and burden of the day to reflect in this fashion, but who can and
must make real the consequences of such thought. How indeed can those
who are able to think most effectively also effectively advise those who
must act with high public responsibility? It is a question daily posed to
science and to the rest of the intellectual world alike.
Science and the humanistic disciplines are at one as primary pro-
ducers of excellence in the society. They are at one, too, as bastions of a
philosophy of equal opportunity. To them alike is posed the challenge of
how best to serve both values without destroying or debilitating either —
a vital, consuming, and extraordinarily difficult assignment that will
stretch as far into the future as we can see. Again, science shares with
the rest of the intellectual world a particularly critical — and most diffi-
cult— task and duty: the task and the duty of preparing the society for
what is to come; of reinforcing its values against the shock of sweeping
change when such change suddenly becomes apparent. Science failed to
do this with respect to the power of nuclear fission, even though the
release of nuclear energy was regarded in some quarters as virtually a
the^ 1 certainty as early as the fourth decade of the century. It
failed for a variety of reasons. The scientists concerned were deeply
absorbed in their work. Early there was some question whether the
REPORT OF THE PRESIDENT 83
process had any practicability. Later there were overriding military
considerations that irrevocably prevented public communication. There
was the ethic then dominant which held that the responsibility of
scientists was entirely to their work, that the social results of that work
were not properly their concern. Surely these were reasons enough. Yet
the ultimate cost of those circumstances was certainly high in terms
both of the intellectual and the institutional unpreparedness of the post-
war society on which the knowledge ultimately broke. It is abundantly
clear that it is not fully paid even today.
It is possible that we are now confronting a situation of equal power
and potential in the biological sciences. The problems inherent in
controlling genetic change in organisms are extraordinarily complex
and difficult— and indeed such controlled change is scarcely feasible in
even the simplest case today. The chromosomes of a single human germ
cell may include in their coded messages on the order of a thousand
million nucleotide base pairs. The task of deleting or mutating a specific
one or even a small group, preselected and at will, seems technically
insuperable. Yet it is currently well known that there are bacterio-
phages capable of deleting from the chromosome of a bacterial host sec-
tions that include genetic information mediating particular processes of
chemical metabolism in that host, which sections, carried by the virus
to a new mutant host of the same species lacking in these particular
capacities, can be incorporated at the proper point in its chromosome
and become functional there, repairing the deficient metabolic mechan-
ism in that cell and in all of its descendants. Bacteria are far simpler
genetically than higher organisms, and such processes of transduction
have never been observed in the cells of mammals, though they have been
sought assiduously there. On the other hand, the evidence is now quite
clear that a large proportion if not all of the forms of life on our planet
use and thus "understand" the same genetic code. And it has been esti-
mated by highly qualified authorities that within a quarter of a century
it may be possible to "program" cells, at least those of bacteria, with
synthetic messages. If and when that is done, the possibility of doing
the same thing with cells of higher organisms might not be too far
removed.
Should such a capability ever be realized, the resulting shock to our
society, if the society were not properly prepared, could be at least as
great as that inflicted by the advent of the atom bomb. And the social
problems that such a technical capacity might pose could be even greater
and more subtle. For, given the awesome power to bring about controlled
genetic change, the paramount question then would be : change for, or
toward, what? It is a question that involves another which hitherto we
have never really had to face — or to evade — because it has never been
34 CARNEGIE INSTITUTION
truly meaningful for us and has seemed, in practice, to lie so far beyond
our ken. It is the question "What, in a specific sense, perhaps even in a
biological sense, does constitute a 'good' society for us: what do we
really want it to be?" If this sort of power were ours, that question
would be posed in the starkest way. Too much emphasis cannot be laid
upon the importance of preparing a society to meet such issues long in
advance of their possible appearance on the horizon. To do this force-
fully and yet soundly, with conviction and yet without sensationalism,
is a task demanding the greatest skill and care both in judgment and in
communication. It is a responsibility shared by scientists and sociolo-
gists and humanists alike, and to be successfully discharged only by
their joint efforts.
There is a final and particularly significant way in which the natural
sciences are at one with the humanities. In Germany it was long held
that exposure to the processes of scientific research ipso facto consti-
tuted a liberal education for the student who himself intended to enter
quite different fields. As with Latin and Greek in the British tradition,
it was held that the rigorous inner logic of the system, its sheer intel-
lectual discipline, its power to identify and to screen minds of great
general potential, its introduction to new realms of concept and new and
difficult techniques of dealing with them, made it a powerful agent for
excellence in education, regardless of the particular ends to which that
education would ultimately be directed.
Similarly, science must share with society as a whole another task of
general humanism, emphasized primarily by science itself. As Edward
Shils has penetratingly noted, the tremendous advances in the life
sciences that characterize our day, with the promise that they hold of
further and even more striking achievements following on a yet keener
and deeper understanding of the human physical condition, also carry a
special risk which we must understand and for which we must be pre-
pared. The prospect of such developments as the substitution of arti-
ficial for worn-out natural vital organs or the replacement of such
organs with other living ones maintained in organ banks, of directed
change in the genetic code or in the expression of genetic factors
through such modifications as could be made possible by our ever-
increasing understanding of protein molecules and particularly of
enzymes, of effective mental therapy or reorientation in desperate cases
that our increasing knowledge of the mind drugs may one day permit us
to achieve, carry an implication beyond that of medical benefit. Inevita-
bly, they also carry the shadowy but sinister possibility of a new kind
of control of one human being over another. And the hint of such a
prospect, whether justified or not, whether real or only imagined, comes
at a time when the experience of widespread violence to human life, and
REPORT OF THE PRESIDENT 35
the memories of that experience on a yet more terrible scale, are all too
vivid. A cumulative effect of such a conjunction of developments could
be that we might in future come to hold individual human life less
sacred than in the past. Such a trend would surely spell one of the worst
social, as well as moral, catastrophes that could befall us. The threat of
such a prospect, however remote, lays the gravest responsibility upon
science and the other disciplines of the intellect alike constantly to labor
at placing new discoveries and newly acquired powers in their proper
setting.
Science and the other intellectual disciplines, the research establish-
ments of science and the universities where scientists are made, the
community of science and the community formed by the educational
fabric of the nation, the scientific nexus and the intellectual nexus as a
whole, clearly share common duties, common opportunities, and common
problems to an extraordinary degree. In some areas the body of science,
by virtue of its own nature and of circumstances beyond its own fore-
telling and control, was touched by the opportunities and immersed in
these problems somewhat earlier than, for example, the corpus of edu-
cation at large. Particularly in its specific relations with government,
scientific research has had a longer history of intensive association than
education, whether in science or in broader realms, and this disparity of
timing has raised some peculiar and often seemingly intractable prob-
lems in the relationships of both science and education to public policy.
But there may be another and salutary aspect of this disparity. The
longer experience of scientific research in this arena may serve as a
paradigm for the parallel but later experience of the educational world
and indeed of the intellectual world as a whole — a paradigm both of
successes and of failures, at once an example and a warning, and above
all, a token for the future. As John Gardner has said: "The nation's
leaders must serve as symbols of the moral unity of society. . . . They
can conceive and articulate goals that lift people out of their petty pre-
occupations, carry them above the conflicts that tear a society apart, and
unite them in the pursuit of objectives worthy of their best efforts." If
the experience of science, both in its successes and in its failures, can
significantly implement these goals, it may truly be accounted of solid
and significant worth for our society and our nation.
We in this tiventieth century have learnt, by bitter experience, that a
climate of liberal, rational opinion is not something which can be
taken for granted, as part of western Europe's intellectual heritage.
It is an ideal to be constantly fought for, ivith unremitting vigilance:
a precious acquisition, all too easily lost.
"Freeing Science From Myth" — Times Literary Supplement
The Year in Review
"Our sun is at least a second generation star. It inherited from pre-
vious stars a rich treasure-house of elements ivhich makes our world
the complex and wonderful environment ive find it to be.
"All of us arc part of that world. We, too, are star dust."
William A. Fowler, California Institute of Technology
The research in the Carnegie Institution for the year just past in-
cluded the uncovering of further information on the still mysterious
astronomical objects known as quasars, the discovery of a probable
physical connection between our Galaxy and another galaxy, the gaining
of new information on the depth and composition of the earth's crust
in eastern North America, the exploration of new vistas in the bio-
chemistry of evolution, a demonstration as to how common marine
sediments form petroleum, the discovery of a way to make some com-
mon plants double their growth by changing their environment, and
the addition of new information on the action of the very important
genes that control other genes. All will be reported upon in this Review.
Such striking findings, which, in one or another form, typify each
year of work in the Institution, by their very nature are deeply exciting.
Yet the less spectacular, but constant and steady progress on difficult
scientific frontiers, the growth of every Staff Member that each year
also typically brings, is no less important. If we were to sum the always
freely chosen individual interests of all our Staff Members, the Insti-
tution's "field" would literally be the history, structure, and organiza-
tion of the universe. This Review will sample the efforts of both Staff
Members and Fellows as they have applied their special talents and
experience to problems that they consider especially attractive and sig-
nificant from this great field.
The Scientist's Concern for History
Although the history of human society is eternally fascinating to
men's minds, it is but a minuscule part of all history. The scientist,
36
REPORT OF THE PRESIDENT 37
too, concerns himself with history — history of a complexity, and scope,
and grandeur that inspire both awe and redoubled curiosity in any
person prepared to investigate it. Some of the grandest and most mem-
orable of all scientific visions have had a strong element of history in
them, like Darwin's and Wallace's on the history of life, or Lyell's and
Agassiz's descriptions of earth history, or Eddington and Jeans on
stellar structure, or Hubble, Baade and Hoyle on cosmology.
The investigation of structure is an interest closely associated with
history. Structure is the stage on which cosmic, planetary, biotic, and
other forms of evolution run their dramatic courses. Thus studies of
the structure of the earth's crust and mantle, or of Galactic structure,
or of intergalactic connections, to mention a few among the examples
in this year's work, all ultimately contribute to an understanding of
the grand history that many scientists find fascinating.
In reading the departmental reports of the Institution this year one
is struck with the number of investigations that are in the context of
this universal history. Such are many of the research programs of the
Mount Wilson and Palomar Observatories in their search for the true
pattern of cosmic history, the programs of the Geophysical Laboratory
and the Department of Terrestrial Magnetism to decipher the history
of formation of the earth's crust and mantle, the studies of paleogeo-
chemistry, experimental taxonomy, the evolutionary aspects of genetics,
and others.
Study of the Earth9 s Crust and Mantle
These interests in history and its structural setting are nowhere
better illustrated than in the geophysical and geological investigations
undertaken within the Geophysical Laboratory and the Department
of Terrestrial Magnetism. Without knowledge of the context of earth
history into which they fit, they easily appear to be a long series of
recondite accounts of specialists' activities. The error of such an im-
pression is very quickly revealed when one examines the relation of
each research program to earth history. Almost every investigation
has a carefully reviewed rationale; it has been specifically designed as
a "probe" into the structure of the earth's crust or mantle, or as an
historical indicator.
It is not so many years ago that historical geology was a subject
somewhat apart from other principal branches of geological studies.
It depended heavily for its progress on paleontology, with assistance
from geomorphology, stratigraphy, and volcanology. These subjects,
of course, continue to illuminate the gradually unfolding history of
the earth. Indeed, the recent and continuing discoveries of micro-
paleontology have extended our knowledge of earth history through f os-
,?< CARNEGIE INSTITUTION
sils to periods once thought unbelievably remote for such a record. But
newer techniques have taken their place alongside the older approaches.
The techniques of experimental petrology are prominent, but isotopic
dating of minerals, seismology, heat-flow studies, crystallography, and
others are making continual contributions. Their importance in ob-
taining the full history of the earth's crust and mantle is very great
because only these techniques can get at the record lying in all of the
igneous and most of the metamorphic rocks that comprise such an over-
whelming proportion of the crust and mantle.
Among the many contributions to our knowledge of earth history pre-
sented in Year Book 66, five illustrate the diversity of the year's work
in geophysics and its relation to the grand design for understanding
the record in the earth's crust and mantle. They are experimental
petrology studies, biogeochemistry, seismologic investigations, earth
heat-flow studies, and isotopic dating studies.
Experimental Petrology. The group of minerals known as pyroxenes
are a major constituent of the earth's mantle1 and also are commonly
encountered in the earth's crust. They are a very diverse group. Al-
though closely related crystallographically they have varied chemical
compositions and textures resulting from different physical conditions
of formation such as temperature, pressure, and oxygen fugacity.
Pyroxenes are associated with basaltic rocks, metamorphic rocks,
layered igneous intrusives, and with some less commonly encountered
rocks thought to have been formed deep within the crust (ultramafic
nodules of kimberlites).2 P. H. Abelson, Director of the Geophysical
Laboratory, observes in his report that petrologists are hopeful that
study of the pyroxenes "will be an extraordinarily useful tool in deter-
mining complex petrogenetic history." A number of studies of the
pyroxenes were carried out at the Geophysical Laboratory during the
year; from them two illustrations are given: study of the phase equi-
libria of two pyroxene minerals known as hedenbergite and wollastonite,
and the electron-probe study of exsolution in pyroxenes.
Some of the older intrusive igneous rocks that outcrop at the earth's
surface are found in the Skaergaard intrusion of Greenland, the Still-
water intrusion in Montana, and the Bushveld complex in South Africa.
Minerals from all of them have been studied at the Geophysical Labora-
tory for their indications of the conditions of mineral formation within
these rock bodies.
D. H. Lindsley, G. M. Brown, and I. D. Muir investigated the phase
1 That part of the earth's interior between the Mohorovicic discontinuity and the core.
^ 2 Ultramafic refers to minerals having- a relatively high content of iron and magne-
sium; kimberlite is a granular rock composed chiefly of pyroxene, olivine (Mg-Fe2Si04)
and biotite mica, found in the diamond "pipes" in South Africa and elsewhere.
REPORT OF THE PRESIDENT
39
relations of two pyroxenes characteristic of the upper layers of the
Skaergaard intrusion, hedenbergite, Ca,Fe(Si206), and a pyroxene
derived from solid solution between wollastonite, CaCSiO.-J, and ferro-
silite, FeSiO.3. By experiment the pressure-temperature conditions
under which inversion (change from one mineral structure to another)
takes place was determined ( Fig. 1 ) . Results from these experiments
were combined with previously obtained experimental data on the in-
version of two other minerals from the same rock body, quartz and
tridymite, both Si02. The combination of these data yields a well-
defined model for the conditions of crystallization of these rocks.
Lindsley, Brown, and Muir conclude that "crystallization of upper zone
c of the Skaergaard intrusion must have taken place at 600 ± 100 bars
and over a temperature interval that included the range 900° to 950 °C."
With these data inferences can be made about the depth within the crust
at which the crystallization took place. The Lindsley-Brown-Muir
IIOO-
o
o
CD
CD
Q_
IOOO
900
Z 3
Pressure, kilobars
Fig. 1. Diagram showing phase relations for green pyroxene from Skaergaard intrusion. Amount
of shading approximately proportional to amount of hedenbergite in run products. Woss, ferrifer-
ous wollastonite (Ca,FeSi206); HDSS/ hedenbergite (Ca,Fe); Trid, tridymite (Si02); Qtz, quartz.
40 CARNEGIE INSTITUTION
experiments might be described as employing the classic methods of
experimental petrology, using pressure-temperature apparatus.
F. R. Boyd and G. M. Brown examined some other pyroxenes with
the aid of an instrument new to the Geophysical Laboratory, the elec-
tron microprobe. This instrument produces a very finely focused beam
of electrons. 1.0 0.1 microns, which, upon impact on a specimen,
generates X rays that can rive accurate information on element com-
position. Very successfully employed for several years in metallurgy,
its use in mineralogy has been limited because of problems attending
the analysis of elements of low atomic number. As F. R. Boyd explains
in a section of his report of this year, the probe permits nondestructive
quantitative analysis of mineral grains at least as small as 8 to 10
microns. He adds, "The degree of homogeneity as well as the compo-
sition of a material can be estimated in probe analysis, and this capacity
will probably revolutionize our concept of equilibrium in mineral
systems."
Boyd and Brown applied the probe to analysis of exsolved3 pyroxene
lamellae from the Skaergaard and Bushveld intrusions. These la-
mellae are considered of particular interest because of their contrast
in calcium content to the crystals within which they occur. The con-
trasting compositions appear to be related principally to the tempera-
tures of crystallization. They are also of interest in determining the
extent of cation migration and ordering in the fine texture of rocks.
Semiquantitative analyses were made of augite lamellae (calcium rich)
in hypersthene (magnesium rich) and hypersthene lamellae in augite.
These 8-to-30-micron-thick lamellae were discovered to have the same
chemical composition as "host" crystals of the same mineral (Fig. 2).
However, the crystal symmetry was found to differ, resembling the
host augite in the case of the hypersthene lamellae (Plate 1). From
these data and those of other experiments Boyd and Brown state that,
for the two minerals considered, extensive cation migration over dis-
tances of 20 to 30 microns took place within the host crystals at tem-
peratures between 1050° and 1000°C, and that migrations over shorter
distances (2 microns) took place below 1000 °C. Although these studies
must be considered as only exploratory, they strongly suggest that the
electron probe may be expected to yield valuable information where its
scanning is applied to the fine structure of diagnostic minerals in rock
bodies.
For a number of years G. Kullerud and his co-workers have pursued
another aspect of experimental petrology that concerns much more
limited sections of the earth's crust, but sections that have been of enor-
3 Solid solutions stable at high temperature can separate into phases with different
chemical compositions on slow cooling. This process is known as exsolution.
Plate 1 . Microphotographs in polarized light of pyroxene crystals from the Stillwater,
Skaergaard, and Bushveld intrusions showing exsolved lamellae.
A. Orthorhombic bronzite crystal with thin lamellae of calcium-rich pyroxene exsolved.
From Stillwater intrusion, Wyoming, United States.
8. Inverted pigeonite" crystal, now orthorhombic hypersthene with thick augite lamellae
in herringbone pattern. Texture suggests exsolution of augite of twinned monoclinic pigeonite,
prior to inversion of the latter to orthorhombic hypersthene. Skaergaard intrusion, Greenland.
C. Inverted pigeonite crystal showing a second, thin set of lamellae exsolved in hypersthene,
presumably after inversion from pigeonite. Bushveld intrusion, South Africa.
D. Complex pattern of exsolution in the hypersthene host of an inverted pigeonite crystal
base). This is additional to the coarse, preinversion lamellae and the parallel thin lamellae
top right). The angle between these two sets of lamellae does not suggest exsolution of
coarse lamellae along the (001) plane of pigeonite; cf. (C). Bushveld intrusion.
Plate 1
Report of the President
REPORT OF THE PRESIDENT
Ul
HwOWl»l^ »"' W* «i
MICRONS
Fig. 2. Electron-probe scans across pyroxene crystals containing exsolved lamellae of another
pyroxene phase. G—J represent scans of augites from the Bushveld formation. Differences in cal-
cium content between "host" and lamellae are depicted. E, scan across a silver lamination in
steel, monitoring the specimen current, and testing sensitivity of the microprobe. Slopes of the
sides of the peak indicate a beam size of 1.0—1.1 microns. F, scan across the same lamination
shown in E, monitoring the FeKa X ray. X-ray spot size is 1.6—1.9 microns.
;j CARNEGIE INSTITUTION
mous importance to human society. These are ore bodies, many of
which are associated with sulfide mineral systems. Experiments have
been concentrated especially on the copper-iron-nickel sulfides asso-
ciated with norite' country rock in the Sudbury district of Ontario,
Canada.
In 1966-1967 the Geophysical Laboratory continued the analysis of
mineral relations within the sulfide systems and the relation of the sul-
fides to adjacent silicates." The investigations of this year extended
knowledge of phase relations of the Sudbury ores through a temperature
ranee from about 1000° to below 400°C. Knowledge of the higher-
temperature behavior of the combined sulfide and silicate systems is
essential to an understanding of the manner of separation of the sul-
fides from the silicates in the original magma, and study of the lower-
temperature phase relations is necessary to account for the specific
character of the final mineral assemblages.
A. J. Naldrett and G. M. Brown studied the iron-magnesium ratio
of the paired pyroxenes from the ore-bearing norite, hypersthene, and
augite, the same minerals studied by Boyd and Brown in their micro-
probe study of lamellae, to establish the temperature of cooling from
magma of the silicates. The ratios found indicate that sulfides and
silicates were exposed together to a temperature of at least 1100°C.
Thus the sulfide ores very probably are formed at or above 1000 °C.
Kullerud and H. S. Yoder, Jr., added further information about the first
formation of the sulfide through study of the mutual solubilities of sili-
cate and sulfide liquids. They found that they do indeed coexist in the
liquid state. They believe that the relation of specific ore deposits to
specific rock types results from a partitioning of metals between the
sulfide and silicate minerals.
J. R. Craig, Naldrett, and Kullerud examined the succession of min-
eral assemblages that form in the solid ore mass as it cools. The wealth
of minerals in these assemblages is illustrated by a partial list of those
encountered in the Sudbury ores: pentlandite, chalcopyrite, cubanite,
violarite, polydymite, millerite, covellite, sillimanite, awaruite, and
idaite, among others. Laboratory study has shown that some of these
minerals are formed at temperatures as low as 400 °C. Thus there now
exists something of a model of the history of formation of one of the
most important ore deposits in the world.
4 A gabbro in which more than half of the pyroxene is orthorhombic. A gabbro is a
plutonir- igneous rock of granitoid texture containing feldspar and pyroxene minerals.
r>. Kullerud, J. R. Craig, and A. J. Naldrett were full-time participants in sulfide
studies. Part-time collaborators included G. M. Brown, S. W. Richardson, H. S. Yoder,
Jr., P. M. Bell, and J. L. England. In addition, G. Moh of the University of Heidelberg
collaborated in some of the work.
REPORT OF THE PRESIDENT U%
Bio geochemistry. Biogeochemistry offers an approach to still other
problems of localized geological history that have unusual interest for
human society. For several years P. H. Abelson and his colleagues at
the Geophysical Laboratory have been examining the diagenesis
(change in form) of organic matter in sediments. Heat treatment of
sediments in the Laboratory can give clues about the organic reactions
taking place under natural conditions. These reactions are of special
concern because of inferences that may be drawn about the forma-
tion of petroleum products within the earth's crust.
This year R. M. Mitterer and T. C. Hoering demonstrated that a
mild heat treatment of the insoluble organic matter (kerogen) in a
Recent marine sediment obtained from the San Nicolas Basin off Cali-
fornia produced a number of substances commonly found in petroleum.
Exposure of a sediment to temperatures on the order of 200 °C for a
few days caused the chemical reduction (hydrogenation or hydrogenoly-
sis)6 of the long, straight-chained structures in kerogen to normal,
saturated hydrocarbons ( Fig. 3 ) . Isoprenoid hydrocarbons and porphy-
rins, molecules with structures characteristic of those used by living
organisms, were also produced from kerogen by such treatment.
Model experiments, in which added olefins and alcohols were reduced
to saturated hydrocarbons when heated with the Recent sediment
showed that such reactions can occur. Mitterer and Hoering conclude
that their experiments provide support for the view that kerogen, a
relatively inert, organic high-polymer, is a major source of the hydro-
carbons found in petroleums.
Explosion Seismology as a "Probe" One of the few probes that can
penetrate anything more than a relatively thin surface layer of the
earth is a seismic wave, the result of a shock at some point within or
on the earth. Seismic waves are, of course, produced by earthquakes,
but they can also be produced by explosions and by various other
methods. For many years explosion seismology has been used with some
effectiveness to obtain information about the crustal structure of the
North American continent, and most recently for the Coastal Plain
and Appalachian Mountain region of the Middle Atlantic states. In
what has been called the East Coast Onshore-Offshore Experiment
(ECOOE) , D. E. James, T. J. Smith, and J. S. Steinhart of the Depart-
ment of Terrestrial Magnetism have collected seismic data that they
have now interpreted in a map of crustal thickness extending from
central North Carolina to central Pennsylvania (Fig. 4). This is the
first three-dimensional model of crustal structure of the eastern North
6 Hydrogenation is the chemical reaction of a substance with molecular hydrogen in
the presence of a catalyst. Hydrogenolysis is the cleavage of a molecule associated with
the addition of hydrogen.
14
CARNEGIE INSTITUTION
8i
O
CO
aj
o
0)
CD
Q
Hydrocarbons- 225 °C
San Nicolas Basin sediment
LJ[a
17
KJ
19
W
u
\J
23
1/
25
W
27
u*
KM
29
[plr*
U
Column temperature 75-325°C 47m in
Fig. 3. Gas chroma tog ram of normal hydrocarbons extracted with benzene-methanol from
San Nicolas Basin, sediment heated to 225°C. Peak numbers indicate the number of carbon
atoms in the hydrocarbon detected.
American area based on physical observation. The map shows a crustal
thickness varying between 30 and 60 km. The thinnest area underlies
Washington, D. C, and adjacent Virginia, and the thickest areas are
found directly beneath the crest of the Appalachians, showing a pro-
nounced "root" structure beneath the mountains. The "root" is broadest
and deepest where the highest elevations occur. Gravity values have
been computed from this model that correlate well in size and distribu-
tion to the actually observed regional gravity anomalies.
James, Smith, and Steinhart infer from their data that the mean
atomic weight of rocks in the upper mantle is lower than that of rocks
in the lower part of the crust. They hypothesize that an upper mantle
composed of peridotite7 or dunite8 and a ferromagnesian (possibly
7 A coarse-grained rock containing a magnesium silicate (olivine) and pyroxene
minerals.
8 A coarse-grained, essentially monomineralic rock, mainly composed of olivine. Some
chromite.
REPORT OF THE PRESIDENT
U5
o.
X
LU
CO
O
r^
c
o
c
D
o
co
<
■u
-o
<A
<D
c
-C
o
*■
o
c
o
CN
■»—
GO
3
,— -*
a
(A
E
o
o
0)
u
E
>»
o
>
_v
o
en
(/>
CO
a)
c
o
**-
-*
V
u
i_
_c
0
■*—
o
c
V)
D
T3
—
<1>
u
0
Q
-^
c
D)
o
U_
E
+ 0 CARNEGIE INSTITUTION
amphibolite") lower crust is consistent with their observations. Their
observations would seem to rule out eclogite,10 a previously hypothesized
component of the upper mantle.
Earth Heat Flow Measurements. Steinhart, Smith, and S. R. Hart,
also of the Department of Terrestrial Magnetism, participated in an-
other type of investigation seeking to delineate crustal structure and
history. This was the measurement of heat flow from the earth's crust
at its surface. The Lake Superior basin was chosen for this investiga-
tion partly because it had been the site of a previous explosion
seismology study. Heat-flow values were calculated for 83 sites in the
lake, thus providing the most comprehensive picture of local and re-
gional heat-flow variation yet obtained for a continental area.11 These
measurements demonstrated also for the first time that large systematic
heat-flow variations on the continent occur on a regional scale. A varia-
tion in flow values by a factor of more than two was found in the Lake
Superior measurements (Fig. 5).
Inferences can be made about the thickness of the crustal structure
from heat-flow measurements because the source of the heat is in radio-
active elements concentrated in the upper crust. Low heat-flow areas
would indicate a thin crust, and vice versa. This interpretation corre-
lates with previous seismic observations. Thus there would seem to
be a thick section of the crust in the central part of the lake and a thin
section at the western end. These observations seem to confirm infer-
ences from seismic observations that the crust below the lake is one
of the thicker parts of the crust of North America. As Smith, Steinhart,
and L. T. Aldrich have hypothesized elsewhere, this unusual crustal
section may be a rift (down-faulted) structure of great age, possibly
having occurred in Precambrian times, making it what might be called
a fossil rift.12
Isotopic Rock Dating. Isotope dating is a final illustration of a method
of seeking out events in earth crustal history. Application of this method
has been undertaken for a number of years by a group composed of
members of the staffs of both the Department of Terrestrial Magnetism
and the Geophysical Laboratory.13
Dating techniques based on interpretation of the decay of radioactive
9 A monoclinic silicate of calcium and magnesium and, usually, one or more other
metals, as iron, manganese, etc.
10 A coarse-grained mafic rock, mainly pink garnet and green pyroxene.
11 Measurements were made from the United States Coast Guard cutter Woodrush,
stationed on the lake.
12 T. J. Smith, J. S. Steinhart, and L. T. Aldrich, "Crustal Structure Under Lake Su-
perior," The Earth Beneath the Continents, Washington, D. C, 1966, p. 196.
13 This year S. R. Hart and L. T. Aldrich of the Department of Terrestrial Magnetism,
T. E. Krogh and G. L. Davis of the Geophysical Laboratory, and F. Munizaga and A. M.
Stueber, Carnegie Institution Fellows, participated.
REPORT OF THE PRESIDENT
U7
u
c
o
u
0)
o
a
E
c
u
_
o
-■
cr
a
.y o
-o —
.£ c
o
o -a
IE
E
= >»
CD
Z "s
■+-
D
CD
E
o
J o
o >
*z o
i2
Supe
w pr
u
o
CD
o c:
"\
q o
£
-■ o
o
o
a> o
u
"? ««
=t ">
0 i!
CD
£ F
,_ _D
*" CD
• D
c £
O >
— d)
rval
land
in 3
= 0
■S "2
> £
C CD
to
£ *S
g is
o
qz v
C Q_
a a
<S®
Fig. 5. Contoured he
indicate location and v>
J>S CARNEGIE INSTITUTION
elements have faced some difficulties in field application because results
can seem conflicting or contradictory. For example, age determinations
on some minerals in a section of Ontario east of Lake Huron, known
as the Grenville province, gave ages of about 900 ± 100 million years
for a mountain-building event. On the other hand, other geological
evidence suggested that the rocks from which these ages were obtained
are much older. Using what is called a whole-rock rubidium-strontium
method of dating, T. E. Krogh and G. L. Davis had established ages of
1500-2000 million years for rocks in the area. It had been assumed that
the 900-1000 million-year event was one in which the rocks of the region
underwent deep burial, intense heating, and plastic deformation, of
which they show unmistakable signs. By the application of carefully
designed sampling techniques to the different rock bodies of the prov-
ince, Krogh and Davis have been able to show that the 900-1000 million-
year metamorphism was preceded by much older metamorphic events.
Rubidium-strontium dating analysis for both granites and paragneisses
of the Province show very close agreement at 1500-1800 million-year
ages, indicating that the Grenville rocks were already intensely meta-
morphosed at the time of the 900-1000 million-year mountain building.
The Astronomer's View
Of all the views of history the grandest, of course, is that of cosmic
history, whose sweep defies imagination. Difficult as it is for a layman
to conceive of a light year (5.88 trillion miles), billions of light years
are part of the everyday thought of astrophysicists and astronomers.
The astronomer's sole source of information is the radiation — visible,
infrared, radio, ultraviolet, X-ray — that can be detected at the surface
of the earth and by space probes. Yet it is a very rich source in astrono-
mers' hands, and from it we now have a tentative conceptual structure
of the universe and its history — comprising events on so vast a scale
that all of human history by comparison is submicroscopic. Much that
is contained in our present concept of the universe derives from infor-
mation obtained on the instruments of the Mount Wilson and Palomar
Observatories, sponsored by the Institution and California Institute
of Technology.
Only a relatively few years ago all astronomy, with the minor excep-
tion of cosmic-ray detection, was optical. But for more than two decades
radio-wave detection has been an increasingly rich source of informa-
tion. Recently X-ray astronomy has been added with the aid of sounding
rockets. Since about 1930, optical astronomy also has been extended
by exploitation of the infrared part of the spectrum. As the year's work
at the Mount Wilson and Palomar Observatories illustrates, the tend-
ency is toward an integration of information from a source object
REPORT OF THE PRESIDENT UO
through as many means of detection as possible. Optical astronomy from
ground-based instruments, however, remains the foundation of our
slowly emerging concepts of what the universe is and has been.
It is most reasonable to assume that the universe is an expanding
one — the more distant an object, the greater its velocity of recession
from the earth. It extends at least several billion light years outward,
but each year seems to bring the discovery of an object with a still
greater redshift14 than that of the previous year. A favorite hypothesis
as to the universe's origin is that of a "big bang," or primeval explosion,
some (possibly 10) billions of years ago. Discoveries at the Observa-
tories gave new information on both these matters, and on others of
cosmic interest.
Quasi-Stellar Sources and Objects. One of the most profound mys-
teries in the sky is that of the quasi-stellar sources (QSSs), whose
extremely high energy output and other unusual qualities were first
detected at the Observatories. One of them, called Parkes 0237-23, was
shown this year to have the largest redshift measured to date. Its emis-
sion line redshift is 2.22, as determined by spectra obtained by H. C.
Arp and T. D. Kinman (Lick Observatory). J. L. Greenstein and
M. Schmidt also determined two separate sets of absorption lines for
Parkes 0237-23 as having redshifts of 2.202 and 1.956. In addition to
their large redshifts, Parkes 0237-23 spectra indicate a relatively high
density for the object and an unusual "double shell" in which ionized
metals are very strong.
A number of other studies added substantially to the facts about
radio-quiet quasi-stellar objects (quasars or QSOs). A. R. Sandage
of the Observatories and W. J. Luyten of the University of Minnesota
continued a study of the faint blue objects found in high galactic lati-
tudes. A 69-object photometric sample indicated that these blue objects
are dominated by radio-quiet quasars and white dwarf stars. Some of
the QSOs showed relatively large redshifts. Sandage and Luyten esti-
mated that the density of QSOs in the star field examined, known as
the Haro-Luyten blue-star field, may be 0.5 per square degree, substan-
tiating an earlier prediction by Sandage that QSOs are very numerous.
Following examination of the Haro-Luyten field, Luyten and Sandage
extended their search for QSOs to seven other fields. Preliminary
analysis of photometric data obtained with the 200-inch telescope for
88 objects indicates that many of the objects in the photometric sample
are good QSO candidates. Sandage and Luyten estimate that there are
at least a hundred thousand QSOs up to a limit at magnitude 19.7 and
that the number is an increasing function of limiting magnitude.
14 The redshift is the displacement toward the red of an observed spectral line as
viewed in a receding source.
50 CARNEGIE INSTITUTION
M. Schmidt studied the space distribution of radio-emitting quasi-
stellar sources and found that their distribution in space increases with
increased redshift. If the density is taken to be 15 at the distance of
redshift 0.5, it is 100 at redshift 1.0.
Knowledge of quasi-stellar sources was also extended in other ways.
Schmidt obtained redshifts for 10 additional QSSs during the year.
J. B. Oke has examined a number of the QSSs for variability. Two
highly variable sources, 3C 279 and 3C 446, have been studied in detail.
Changes in luminosity of as much as 0.25 magnitude per day have been
observed for 3C 279, the visual magnitude of which has varied by
nearly 2.0 during the past year. For both sources the spectral con-
tinuum is somewhat redder when the object is fainter. Two observed
spectral lines (magnesium II of 3C 279 and carbon IV of 3C 446) re-
mained constant, independent of change in the continuum. J. Wampler
of the Lick Observatory and Oke also studied the source 3C 273 in
great detail with photoelectric scanners. They have made an estimate
of electron density in the emitting region of this source, and estimate
it to be at least 106 per cubic centimeter.
Sandage also determined the redshifts of nine new radio galaxies
as part of a program of determining redshifts of all identified radio
galaxies from the Third Cambridge Catalogue, a project in which
Schmidt is also participating.
An X-Ray Star. One of the most unusual astronomical objects studied
during the year concerned the very strong X-ray source Sco X-l, the
position of which had been determined by the X-ray astronomy group
of American Science and Engineering, and Massachusetts Institute of
Technology, in a series of spectacular rocket experiments in 1966 and
early 1967. The optical identification was made at Tokyo Observatory,
Japan, and in a series of photoelectric measurements made by Sandage
with the 200-inch telescope. From night to night variations of 0.8 mag-
nitude were observed, as well as large night-to-night variations in
spectral-line intensities. Optical brightness varied by as much as d=0.03
estimated at about 400 parsecs15 and the X-ray flux was estimated at
about 1037 ergs per second in X rays at wavelengths of 1-10 angstroms.
Later observations of Sco X-l by J. A. Westphal and Sandage sought
time resolutions of the object's emission for periods as short as two
seconds. Using pulse-counting equipment designed and built by the
Astro-electronics Laboratory of the Observatories, their further obser-
vations showed the source to be highly variable in time intervals as
short as one minute. Nonrandom variations with intervals of 4 minutes,
15 One parsec — 3.258 light years.
REPORT OF THE PRESIDENT 51
30 minutes, and 60 minutes were also indicated. J. Y. Jugaku found
evidence for radial velocity changes from night to night.
Redetermination of the Hubble Constant. One of the continuing
problems of cosmological astronomy is the calibration of its distance
indicators. The Hubble Constant, which gives the increase in velocity
of an object's recession from the earth per unit of distance,16 is the
end result of this calibration. A preliminary value for the Constant
was determined many years ago. As a first step in its redetermination,
G. Tammann and Sandage have completed measurements and analysis
of the variable stars in the galaxy NGC 2403. This is a spiral galaxy
and was the first beyond the Local Group of galaxies in which cepheid
variable stars were found soon after the 200-inch telescope began opera-
tion in 1949. Since that time a series of color plates has been made
of the galaxy over a period of years, and a photoelectric sequence be-
tween 1962 and 1965. From this material 56 variable stars have been
located, including 17 cepheids. Sandage has applied five methods to
determine the distance modulus of NGC 2403, using cepheids, the
apparent magnitude of the red supergiant variable stars, the apparent
magnitude of the brightest resolved stars, calibration of bright irregu-
lar blue variables from prototypes in the nearer galaxies M 31 and
M 33, and the angular size of Hydrogen II regions. The five methods
agree remarkably well, and a much more exact distance determination
for this galaxy is expected when the calculations are complete. The
distance to NGC 2403 is considered scientifically important because it
will permit calibration study of other galaxies that will lead directly
to certain galaxies that have measured redshifts. When that is done
an improved value of the Hubble Constant can be calculated.
A Blue Star Anomaly. Evidence of some interest in determining the
validity of the "big bang" cosmological model was obtained by J. Green-
stein and Oke in spectrograms and spectrophotometric scans of blue
stars in the galaxies M 13, M 15, and M 92. According to the "big bang"
model these stars should have a helium-to-hydrogen ratio of nearly 30
per cent. The spectrographic and spectrophotometric results were in
direct contradiction to the model's prediction. Helium I lines are very
weak or absent in stars whose ultraviolet blue visual spectra ( UBV)
demand their presence in the galaxies M 15 and M 92. Earlier observa-
tions, including those of Greenstein and G. Munch, were thereby con-
firmed. It is considered a serious anomaly for the predictions of the
"big bang" model.
Stellar Observations. Two further sets of observations are of particu-
lar interest for what they tell us about the sun and our Galaxy. A. J.
Deutsch has studied for several years a group of stars that he has
16 Present accepted value: 100 km/sec per 106 parsecs (3,258,000 light years).
52 CARNEGIE INSTITUTION
called "blue stragglers." It is his hypothesis that these stars are
metamorphs of solar-type stars. On the basis of observations made of
the blue stragglers he has shown that if the sun's interior contains
as much angular momentum as the blue stragglers did when they were
solar-type stars, the radiation entering the solar hydrogen convection
zone will maintain the sun's present rotation against the torque exerted
by the solar wind. This supports the hypothesis advanced by R. H.
Dicke of Princeton University that the sun's interior rotates much
more rapidly than its surface. Deutsch also accounts for the equatorial
acceleration observed at the sun's surface by postulating that it arises
from the viscous coupling between a rapidly rotating interior and the
slowly rotating hydrogen convection zone of the surface.
E. Becklin and G. Neugebauer studied the general infrared radiating
structure of the central region of the Galaxy and compared it with the
infrared structure of the nucleus of Galaxy N 31. The two galactic
centers were reported to look similar in both shape and brightness in
their infrared structure. Thus our Galaxy has an analogue in this
respect.
Astrophysics and Astronomy at the Department of
Terrestrial Magnetism
The well-known Mount Wilson and Palomar astronomy program has
been supplemented for some years by a modest but imaginative astro-
physical program at the Department of Terrestrial Magnetism. Pri-
marily focused for most of its existence on radio astronomy and image
tube development, the astrophysical activities of the Department of
Terrestrial Magnetism broadened during the year. Optical astronomy
observations, a radio astronomy discovery, a cosmic-ray study, and a
nuclear physics experiment are reported as examples of the range
of interest.
Optical Astronomy. The optical astronomy program of the Depart-
ment of Terrestrial Magnetism centered about the use of the DTM
image tube spectrograph. W. K. Ford, Jr., and Vera C. Rubin report
that a new spectrograph camera, designed by I. S. Bowen, former Di-
rector of the Mount Wilson and Palomar Observatories, gives spectra a
substantially improved quality as compared with earlier equipment.
Among their observations were several probing the nature of "peculiar"
galaxies, subjects of a study at the Mount Wilson and Palomar Observ-
atories described last year. A commonly held hypothesis is that the
morphology of at least some of the peculiar galaxies indicates that they
represent galactic "explosions." Ford and Rubin's observations on the
peculiar galaxies NGC 4038 and NGC 4027 showed that velocity varia-
tions across these galaxies are so small that they cannot in fact be
exploding galaxies.
REPORT OF THE PRESIDENT
53
A Hydrogen "Bridge" from our Galaxy? One of the more interesting
results of the year came from an observation of the southern sky. Sev-
eral years ago a 100-foot parabolic radio telescope was designed by
M. A. Tuve and his associates at the Department, and subsequently
erected at La Plata, Argentina. This is now the principal instrument
of the Instituto Nacional de Radioastronomia, under the direction of
C. M. Varsavsky, a Research Associate of the Institution. F. Strauss,
a student of Varsavsky, has commenced a search for an intergalactic
bridge of neutral hydrogen between our Galaxy and the Magellanic
Clouds, the nearest outside galaxies. Observations have traced an arm
of hydrogen extending from the outer spiral arm of our Galaxy through
at least 60° of galactic longitude. Measurements of the arm's velocity
and position (Fig. 6) strongly suggest a connection of the arm with
the Small Magellanic Cloud. Succeeding observations will show whether
or not the present strong indications of a bridge to the Small Magellanic
Cloud from our Galaxy does exist. If so, it will be the first indication
of any long-period physical connection between this Galaxy and our
galactic neighbors.
/ = 270°
o
O
00
■» II
Sun
(
1
j
c
5
Galactic
center
o
• SMC
11
Fig. 6. Projection of outer arm of the Galaxy and Small Magellanic Cloud on the Galactic
plane, showing extension of hydrogen in the direction of the Cloud. SMC, Small Magellanic
Cloud; Kpc, kiloparsec, or 3258 light years.
Cosmic-Ray Study. Probably the longest continued program of any
scientist in the Institution is the cosmic-ray investigation of S. E. For-
bush of the Department of Terrestrial Magnetism. Forbush's ioniza-
tion-chamber observations and his data interpretations of cosmic-ray
intensity began in 1937. They have been directed partly toward the
diurnal variation of cosmic-ray intensity. His data now represent the
longest series of such measurements available. Forbush's analysis of
the 1937-1965 data, reported this year, provides a value for the average
diurnal variation in cosmic-ray intensity (an amplitude of 0.15 per
cent), and for the first time reliably shows that there is a cyclical
change in the diurnal variation. The amplitude of the cyclical variations
is about 50 per cent of the diurnal amplitude with a period of two solar
CARNEGIE INSTITUTION
cycles, or about 20 years. Zero values for this 20-year wave occurred in
1988, 1948, and 1958, The 195S zero value coincided with the time given
by H. W. Babcock for the reversal of the sun's general magnetic field.
Forbush says that as yet there is no satisfactory theory for the 20-year
variation.
Foil Excitation Experiments. One set of experiments at the Depart-
ment of Terrestrial Magnetism shows that research in atomic physics
need not always depend upon high energy "superequipment." L. Brown,
YV. K. Ford, Jr., Vera C. Rubin, and W. Trachslin report the results
of an experiment in which they use the foil excitation method of
measuring ions in an excited state. The experiment was undertaken
to test if a method could be developed for measuring the lifetimes of
excited states, to be used to determine abundances of elements in stars.
In the experiment a beam of sodium ions from a Van de Graaff acceler-
ator passes through a thin carbon foil. The emerging atoms may be
in any charge state from neutral to completely ionized. If excited, the
atom will radiate under conditions like those of a free atom. The glow-
ing beam that has passed through the foil is imaged on a lens and the
image is then passed into an image tube spectrograph. The procedure
gives a novel slant-lined spectrogram (Plate 2). The charge state of
ions can be determined in a procedure of Trachslin's invention. The
results from the experiment were somewhat unexpected in that only
a few of the lines could be assigned to known transitions in sodium.
It is therefore not yet clear whether the method will eventually afford
data that can be used in determining element abundance in astronomi-
cal objects, but it is evident that a new field of atomic studies relating
to cosmological problems has been opened to investigation.
Biochemical and Biophysical Views on Problems of Evolution
Breathtaking as an excursion into the sweep of cosmic history can
be, most people find new views of the history of life equally fascinating.
At least two windows have been opened to such views in the work of
the year — one in the study of racemization of amino acids in fossil
shells by P. E. Hare and P. H. Abelson of the Geophysical Laboratory;
the other, in the study of "saltatory replication" by R. J. Britten and
D. E. Kohne of the Biophysics Section of the Department of Terrestrial
Magnetism.
Racemization of Amino Acids with Time. For more than a hundred
years it has been known that biological substances may differ in optical
properties. Commonly this is measured by the rotation of light shown
within an instrument called the polarimeter. A substance may be opti-
cally inactive, dextrorotatory (Z)-form), levorotatory (L-form), or
racemized ('equal amounts of L-forms and D-forms) . Most amino acids
Plate 2
Report of the President
O
It.
<&»/ w? SvJf *m
j 4 0 I "****
Plate 2. A part of the foil-excited spectrum
of sodium. Above and below the sodium spec-
trum are comparison lines from an iron-neon
arc, for four of which the wavelength in ang-
stroms is given. The unique characteristics of
the sodium lines are the decreasing intensity of
the lines downstream from the foil and the slant
of the lines caused by the Doppler effect.
5* / 7 ^jjiS|S*j&:
I %J / imm-
-J O
REPORT OF THE PRESIDENT 55
are optically active, and one of the great biochemical puzzles is
why the proteins of all living forms consist principally of L-
amino acids. The universal use of the L-form has long attracted the
attention of scientists interested in the origin and evolution of life.
Why was one stereo-isomer chosen?
Hare and Abelson attacked one facet of this problem during the year
by examining the optical configuration of the amino acids extractable
from Mercenaria shells of three geologic ages : Recent, upper Pleistocene
(^60,000 years), and upper Miocene (10 million years). Nine amino
acids were still detectable in the most ancient shell. As expected, the
amino acids in the Recent shells were almost entirely in the L-form.
But those in the Miocene shells were almost evenly divided between the
L- and L-forms, that is, they were racemized. The Pleistocene amino
acids were partly racemized. One of the most interesting results of
the shell amino acid analysis was the discovery that the fossil shells
contain almost no isoleucine. On the other hand, they contain alloisoleu-
cine, nearly all of it in the L-form. Isoleucine and alloisoleucine have
somewhat different chemical properties. Isoleucine, for example, is
slightly less soluble in water. Alloisoleucine is not found in proteins.
The apparently normal process of racemization that yields products
of nearly identical physical-chemical behavior for most other amino
acids, in the case of isoleucine gives products easily separable by non-
biologic processes.
From these results Hare and Abelson suggest that the primitive
ocean contained a racemic mixture (equal amounts of L-forms and
L-forms) of amino acids. The mystery of how the L-form became the
preferred protein form still remains; indeed, it may be greater than
before. If some unknown process produced a slight preponderance of
either the L-form or the L-form, racemization would restore the equi-
librium. On the other hand, once a route to the selective extraction of
the L-form was assured, then L-form supplies could be replenished by
racemization from the L-form as the L-form was used.
Repeated Sequences in DNAs. The report of the Biophysics Section ir
of the Department of Terrestrial Magnetism commences with these
words: "A central theme runs through this report as it has in past
years : the enormous potential for understanding the relationships and
history of creatures through the similarity of nucleotide sequences in
their DNA."
R. J. Britten and D. E. Kohne report this year on their exploration
of repeated sequences of DNA in different species representing all the
major forms of life. In 1964 Britten and Waring discovered that a
" E. T. Bolton, D. B. Cowie, and R. B. Roberts, Staff Members; D. E. Kohne, Staff
Associate; D. J. Brenner and A. Rake, Fellows; S. Falkow, Visiting Investigator.
56 CARNEGIE INSTITUTION
type of mouse DNA known as satellite DNA was composed of a set of
DNA sequences repeated with astonishing frequency. Indeed, one short
nucleotide sequence was repeated about a million times in this DNA.
This proved to be an extreme example of DNA repetition, but 10,000
and 100,000 sequence repetitions appear to be more common. Britten
and Kohne have pushed forward this year with a versatile and power-
ful set of investigations leading to some hypotheses that may have a
remarkable impact on our view of the evolutionary process in life at
a very fundamental level. The DNAs investigated range all the way
from those of viruses to those of human beings. Fifty-five different
DXAs from the plant and animal worlds were analyzed for the char-
acter of their nucleotide sequences. Among them only the DNA from
viruses, from bacteria (E. coli, Clostridium perfringens) and from
blue-green algae failed to show repetition in the genome. In some cases
i e.g.. salmon sperm) as much as 80 per cent of the cell DNA seems
to be made up of repeated sequences.
In Britten's hypothesis the sets, or "families," of related nucleotide
sequences that will reassociate with each other result from "saltatory
replication'' — events that produce families of hundreds of thousands
of similar nucleotide sequences in the DNA of an organism within
a relatively short time span, perhaps of no more than one generation.
Families of DNA sequences may be shared between or among species,
but diverge more and more as the hypothetical time elapses after speci-
ation increases. In short, relatedness of the DNAs decreases with time
following species branches (Plate 3) .
Britten and Kohne state that divergence among the members of a
family of repeated nucleotide sequences occurs during the evolution of
a species. They believe further that the degree of divergence may be
a measure of the age of the family of sequences. Thus a knowledge of
nucleotide family divergence is of obvious interest in considering the
history of the evolution of life forms. In studies of the reassociation
of nucleotide pairs (homology) of present-day chicken and rhesus
monkey DNAs, Britten and Kohne's results indicate that a large frac-
tion of the two DNAs existed as repeated DNA sequences at the time
those species lines diverged hundreds of millions of years ago.
In addition to the families of repeated sequences found widely dis-
tributed among higher organisms, the DNA of all organisms includes
a component of nonrepeated nucleotide sequences. Both the repeated
and the nonrepeated elements change slowly with time, but Britten
and Kohne's experimental data suggest that the average rate of change
of the repeated sequences is less than that of the nonrepeated.
Britten and Kohne conclude their report with the remark that "the
wide occurrence of families of repeated sequences . . . indicates that
sudden events may be far more important to evolution than we have
Plate 3
Report of the Vreniderd
^■^^^^»MH?F?*bIv^^" :
■ ;F. 11
wststm
IMIII
%•■ ". .v,'?;vi"
IllJjf
CtSSItxfl lift}
WSSM
%MSSMi3wSi$
^KSs4
**« .',
ift^Kli*'
Shifts
■r
. .-
..' ,
;p- - - - - -
4* :&"■'» <-m' ■■:■■'.-■
:, ■ - *
m
ssrilllris
•
Ptii§
l|lipil§f«
90
Plate 3. Schematic diagram suggesting the history of families of repeated DNA sequences.
The right-hand scale indicates the period of time since the saltatory replication that produced
each family. The left-hand scale — as a measure of divergence — represents the thermal
stability of the reassociated pairs of DNA strands formed by the members of a family. The
height of the peaks indicates the amount of DNA of a given age and thermal stability.
REPORT OF THE PRESIDENT 57
heretofore suspected." They suggest the possibility that events can
occur whereby a large number of new genes can arise relatively rapidly
in the genome,18 many of which remain unexpressed in the organism's
phenotype,19 adding to the reservoir of genetic potential. They note
further that "mutation, translocation, and recombination with other
genes would yield a whole range of potential genetic activity [that
could] turn up at times long after . . . the saltation." Thus does the
individual organism carry within itself not only the historical record,
but also the makings of future evolutionary history.
A Widening Interest in Genetics — The Department of
Embryology as an Example
One of the striking implications of the reports of investigators in
biological subjects at the Institution this year is of the key position
now occupied in wider biological realms by subjects once the domain
only of genetics. The investigations of Britten and Kohne illustrate
this. The work of the Department of Embryology, offering another
equally important example of this trend, has focused on a study of
developmental processes in the organism. J. D. Ebert, Director of the
Department, states that even as recently as 1960 it was not uncommon
for embryologists to take the chromosomes and genes as "constants" in
the equation of the development of the organism. Today, however, lead-
ing embryologists everywhere are deeply concerned with the labile role
of genetic materials. As Ebert remarks : "Today it is a cardinal rule
that a detailed knowledge of biochemistry and morphology of normal
development of an organism is essential if we are to understand the
role that genes play in its development."
The bridging of the gap between studies of the gene and those of
the differentiating cell in recent years has been made possible in part
by the rapid progress in knowledge of the molecular basis for gene
action. As Ebert states in his introduction, information contained in
the DNA molecule is first transcribed in the developing organism into
an RNA copy, generally within the cell nucleus. The transcribed (RNA)
copy of the gene then moves to the cytoplasm of the cell, where the
linear information coded in the RNA nucleotides is translated into a
corresponding linear sequence of amino acids. These amino acids in
turn are linked to form polypeptides and are finally converted to pro-
teins. Even though a great deal remains for discovery before a complete
description of the process of translation can be made,20 enough is known
18 A complete set of all the different chromosomes normal to a species. The genetic
content of a cell composes a genome.
19 The expressed characteristics of the individual organism.
20 C. R. Woese, D. H. Dugre, S. A. Dugre, N. Kondo, and W. C. Saxinger, "On the
Fundamental Nature and Evolution of the Genetic Code," Cold Spring Harbor Labora-
tory of Quantitative Biology Symposia on Quantitative Biology SI, 1966, 723-736.
,7> CARNEGIE INSTITUTION
to make this general model a serviceable bridge between the traditional
domains of genetics and embryology. Furthermore, it is now clear that
the transcription of the genetic code in every cell is highly regulated.
There are at least three levels of "controls" of gene action. The trans-
lation of the information contained in its DNA differs from cell to cell
and may differ from time to time in the life history of a given cell.
The frequency with which a given gene is transcribed into RNA may
also vary. Finally, RNA copies of genes may or may not be translated
into proteins. It is clear why exploration of the operation of such regu-
lators of genie expression has become central to the study of develop-
mental biology.
Research in developmental biology treats, in any eukaryotic orga-
nism."' the most complex biochemical systems known. The isolation of
specific systems with measurable relations is not easy, and may be one
reason that the bridge between the gene and the differentiated cell
was so long uncrossed. However, D. D. Brown, of the Department of
Embryology, and his colleagues believe that they have found a model
system for study of the control of gene action in ribosomal RNA
I rRNA) . The organism selected several years ago by Brown as particu-
larly suitable for analyzing this system was the clawed toad Xenopus
laevis. The wisdom of the selection and the ingenuity of the experi-
mental technique have been amply proved by a series of significant
discoveries during recent years. The present year has been typical.
Brown, C. S. Weber, and J. H. Sinclair this year used a molecular
hybridization technique employing Millipore filters to extract from
Xenopus still further significant information about the relation of the
genes (DNA) and ribosomal RNA. Ribosomal RNA can be separated
into three fractions designated, according to rate of sedimentation, as
28S, 18S, or 5S. Brown, Weber, and Sinclair found that the genes for
the 28S and 18S fractions are highly redundant (with about 800 copies
for each type), closely adjacent on a single chromosome.22 The genes
for these types are thought to be alternating. Even more interesting,
these investigators have found that the functionally related genes for
the 5S RNA are not physically linked to the genes involving 28S and
18S RNA. Brown and his colleagues made this discovery by compara-
tive analysis of DNA and rRNA during early stages of growth in an
abnormal mutant of Xenopus lacking a nucleolus in the homozygous
condition.
Using the same methods, Brown, Weber, and Sinclair also showed
that somatic tissues of Xenopus with a wide range of rates of ribosomal
synthesis contain very similar numbers of genes for the three types of
21 An organ i nm having a true nucleus.
22 More specifically an autosome, that is, a nonsex chromosome.
REPORT OF THE PRESIDENT
59
rRNA. However, egg cells have an enormously greater number of genes
for the 28S and 18S RNAs. Thus something of a paradox appears.
Although somatic cells23 synthesize ribosomes at greatly differing rates
with complements of genes that do not vary, the egg cell displays a sev-
eral hundredfold increase of genes for the two types of ribosomal RNA
(Fig. 7). Experiments show that the excess genes or rDNA remain
inert in the cytoplasm during cleavage of the fertilized egg and are
ultimately degraded as the cells of the growing organism proliferate.
Brown, Weber, and Sinclair's observation of the anucleolate mutant
of Xenopus adds to the mounting evidence of the existence and nature
of regulatory mechanisms operating throughout the development of
an organism, "turning on" or repressing rRNA synthesis. Thus the
anucleolate mutant failed to synthesize 5S rRNA although it was
23 Any cell of the body except egg or sperm cells.
0.4
0.2
E
O
C\J
Q
d
0.2
Erythrocyte
Early gastrula
rRNA
Liver
5SRNA
Heterozygote ( l-nu)
2000
000
c:
"e
1/5
o
000
0
10
0
0
Tube no.
Fig. 7. Hybridization of Xenopus laevis rRNA with various DNAs. The undotted curve in each
chart shows the fraction of the genome homologous with 28S rRNA, 1 8S RNA, and 5S RNA
in each type of cell. Even though the genome remains essentially the same, its rRNA expression
differs strikingly among the cell types.
60 CARNEGIE INSTITUTION
demonstrated that the gene for this fraction was present in the anucleo-
late cells. The absence of an activating factor, or the presence of a
repressing factor, must be postulated to account for the behavior of
the 5S RNA gene. If one visualizes the "construction" of a ribosome
that requires all three DNAs, action by the linked 28S and 18S genes
and the independently located 5S genes is required. Although the two
sets of genes are in separate locations, they must be coordinated. The
activator, whatever its nature, appears to be associated with the 28S
and 18S part of the genome. As Ebert observes, the search for the nature
of this mechanism should be very interesting. Its discovery would be
highly significant.
Igor Dawid, also working with Xenopus, has firmly established that
the cytoplasm of its eggs contains substantial amounts of a DNA that is
associated with the cell mitochondria.24 Moreover, there is abundant
evidence that DNA is an integral component of mitochondria in a wide
variety of organisms. It seems possible that this may be another general
characteristic of eukaryotes. Dawid and his colleague, David Wolsten-
holme, state that their results can be "interpreted as suggesting the
preservation of mitochondrial DNA during evolution." In a series of
elegant studies they have also established that mitochondrial DNA in
Xcnopus has the structure of a "twisted circular" molecule (Plate 4).
Genetics at the Genetics Research Unit
Genetics as a subject of research in the Institution goes back almost
to its beginning. No subject in the Institution has had a longer history
of distinguished accomplishment. Indeed, the whole question of the
"bridge" between the genes and the development of somatic tissues,
now opening in an exciting way for developmental biology, has been a
subject of continued research by Barbara McClintock of the Genetics
Research Unit for more than 30 years.
A Control System in Maize. Again this year, Dr. McClintock reports
on yet further analysis of one of the "bridges" between genes and their
somatic expressions in maize : the mechanism of genie control that she
has called the Spm (suppressor-mutator) system. Her report this year
concerns observations of the effects of the Spm system on the gene locus
in maize for anthocyanin synthesis — the production of the blue or red
anthocyanin pigment in the plant body or in seed kernels. Her observa-
tions are of special interest because they demonstrate that a single
control system can provide an astonishing diversity of regulation in
gene expression.
The Spm system includes what Dr. McClintock calls component-1,
the suppressor, and component-2, the mutator. Component-1 might be
24 Organelles found in the cytoplasm of every cell except bacteria and blue-green algae.
Plate 4
Report of the President
*' * * # ,* -«f7 %* felt* •- *%Zf/'\
' ***-f" !':■/ Kf&l+LL. riff >« H *#■ v ' 4* "- > * • ■
^%^&»"%
*^ m %
Plate 4. Electron micrographs of Xenopus laevis (toad) DNA prepared by Wolstenholme.
All micrographs are X 85,000. A-C, Molecules of native DNA from oocyte mitochondria.
A, B, Twisted circles. C, An open circle with a contour length of 5.6 microns.
riate 5
Report of the President
o
-o
c
E
<v
CO
o
<u
u
(J
L-
£
o
a
—
CO
<u
u
c
D
o>
c
^
*»_
-O
o
_Q
>
01
o
'a;
-£!
c
D
D
£
c
o
>_
c
«+-
O
c
D
a;
c
10
<D
E
to
O)
c
"q.
>—
a;
0)
N
'a
c
E
>v
•■♦-
o
D
£
i_
<u
—
>-
<D
_D
u
O
c
o
c
O
1_
D
c
_^
o
c
<D
c
£
_Q>
V
3
E
TJ
a
O
i/)
Im
a
c
D
</>
c
-*—
o
a
.c
r>
.£!
to
Im
£
</>
O
_c
15
</>
4_
D
c
<U
a>
i_
E
D
a>
"D
'q.
<D
-4—
"a
<L>
o
E
o
<i>
^
_c
«-t-
O
«o
i_
<D
<v
~o
i_
_o
0
Q_
_Q
REPORT OF THE PRESIDENT 61
considered the "switching" mechanism for the anthocyanin gene. If it
is active the anthocyanin gene produces a uniform pigmentation; if
it is not active no pigment is produced. However, the type and distribu-
tion of the pigment depend on the activity of component-2. If com-
ponents is inactive and component-1 is active the aleurone layer of
cells of a maize kernel will be lightly pigmented. If component-2 is
activated, deeply pigmented spots appear on the lightly pigmented back-
ground. The size of the superimposed dark pigmentation depends on
the stage of component-2 activation; the earlier the activation, the
larger the number of clonal cells with dark pigment. If component-2
is activated in the gamete all cells are deeply pigmented. Component-2
is called the "mutator" because it induces responses that modify the
organization of the anthocyanin locus and through it the consequent
somatic expression of the anthocyanin gene.
By isolating maize lines that contain one or another of these com-
ponents in the gametes, with the mutator component in at least two
different forms, Dr. McClintock has produced a great variety of pheno-
typic expressions in maize kernels. Among other interesting disclosures
she has shown that the Spm element in a cell may influence pigment
production in neighboring cells, even when no pigment is produced in
the cell itself (Plate 5). Dr. McClintock states that the patterns that
result from the action of the control systems on the anthocyanin genes
"are so varied that they defy a meaningful classification." She adds,
"Only the fact that anthocyanin pigment ... is not vital to the plant
makes it possible to learn about the many kinds of regulation that such
a system can provide." She suspects that what may have been inter-
preted as the expression of different alleles in some past genetic studies
may represent instead different "states" of the gene loci as she has
described them in the case of the anthocyanin genes. She concludes that
the term "alleles" in any specific instance may be ambiguous unless
there is a means of distinguishing between a mutant of the structural
gene itself and a mutant that is produced by a component of a regula-
tory system such as the Spm system.
Phage Lambda and Other Viruses. A. D. Hershey, Director of the
Genetics Research Unit, includes in his report this year two general
reviews, one of the "state of the art" in the biochemical genetics of
bacteriophages and the second, of the meaning of that state for more
comprehensive views of genetics. In the first he summarizes what is
known about sexual and asexual conjugation and genetic recombina-
tion in bacteria, especially as it has been revealed in studies of
Escherichia coli and related species. In the second he discusses the
relations between a particular phage, lambda, and its host bacterium,
E. coli.
<;j CARNEGIE INSTITUTION
Hershey notes that at least seven mechanisms provide opportunities
for exchange of genes in bacteria. None is typically sexual, although
several involve conjugation between differentiated cell types. Genes are
passed from cell to cell by mechanisms that may differ as to origin of
transferred genes, manner of transfer, and mode of replication in the
carrier cell line. One of the most intensely studied of these mechanisms
is transduction, in which extra chromosomal DNA fragments or
viruses may transfer genes from one bacterial line to another. Dr.
Hershey has been particularly interested in description of the genome
of the virus or phage genome.
One class of transducing phage (e.g., coliphage PI) operates by
picking a phage-sized section of bacterial DNA more or less at random
and incorporating it into the phage particle. This DNA segment may
then be transferred (transduced) to another E. coli line, where it is
incorporated into the genome. Hershey notes that this process mimics
sex at the population level, because all bacterial genes are transduced
by PI with similar frequencies.
Phage lambda, which has been a favorite material in Hershey's
laboratory, operates very differently. It specifically transduces the genes
responsible for galactose25 and biotin26 metabolism in E. coli, but no
other genes. Another phage of the same kind, called <£80, transduces
tryptophan-forming genes. Hershey states, "Lambda is important . . .
because it can recombine . . . cellular and viral inheritance in ways
that are fascinating to contemplate and, very likely, of practical im-
portance to humans. "
About 20 genes have been ordered thus far on a "genetic map" for
phage lambda. This is at least half, and possibly more, of the genome
of this phage. Hershey says that the length of the DNA of lambda is
about 44,000 nucleotide pairs, which could accommodate 30 or 40 genes.
Lambda has three capacities of particular interest to geneticists. Like
a typical phage it can infect and lyse bacterial cells, destroying them
and producing numerous phage progeny. But it can also take up
extended residence in the bacterial chromosome, giving rise to viable
host-cell lines of modified inheritance. These bacterial cells are said
to carry prophage and are called lysogenic. Third, lambda can form
transducing phage lines specific to a single gene.
Hershey discusses what he believes to be the most satisfactory model
of the relation of phage lambda to its bacterial host in the transduc-
tion process. The model is that of A. M. Campbell, formerly an Associate
in Research at the Department of Genetics and now at the University
of Rochester Department of Biology. The model derived from Campbell's
Galactose is a constituent of the milk sugar lactose.
26 C10H16N2O8F. Biotin is one of the B vitamins.
REPORT OF THE PRESIDENT
63
discovery, while he was at the Department of Genetics about ten years
ago, of what are now known as the suppressor-sensitive mutants of
lambda. These mutations are of special interest because they can be
found for any gene that has an essential function. Hershey states that
the Campbell model for prophage excision and insertion (Fig. 8) is
strictly analogous to deletion and insertion in all the formal models for
the genesis of any chromosomal rearrangement. Here is one more dra-
matic example of the importance of detailed probing into the structure,
the permutation of forms, the chemistry, and other qualities of what
may superficially seem the tiniest part of the life process. As Hershey
remarks in another connection in his report, "What we see on reflection
is an intimation of history, another witness interrogated as to what
living things are about."
Among the interesting features of lambda transduction brought out
in Hershey's review are : ( 1 ) Crossovers between phage and bacterial
DNAs are nonequational, that is, the length of phage DNA deleted
and of bacterial DNA inserted in the phage chromosome are unequal.
Os
pp
A
N
^ A phage particles
I 1
A R
gal
no
I !
Configuration
at crossover
■P
A
9a
N
Prophage
Xdg phage
particles
Fig. 8. Structural relations of lambda wild-type phage, lambda prophage, and lambda dg
(defective) according to A. M. Campbell's model. Genetic maps are indicated by the phage
genes A, J, N, and R and the bacterial genes gal (galactose) and bio (biotin). Phage and
bacterial components are indicated by shading; terminal cohesive sites by arbitrary symbols:
pp stands for loci of permutation points; ii for loci of prophage insertion sites (loci that become
crossover regions ip and pi in the prophage map).
(tf CARNEGIE INSTITUTION
(2) Measurement of the deleted parts of the phage DNA molecules
suggests that there are about 1100 nucleotide pairs per gene (assum-
ing that all genes in the deleted region are known).
The base content of three distinct sections of the lambda DNA mole-
cule was analyzed in greater detail during the current year than pre-
viously. Elizabeth Burgi, Anna Marie Skalka, and Hershey continued
to use the content of guanine-cytosine as a chemical marker, and found
that the shorter the molecular segment the less it resembles its neigh-
bors in composition. Laura Ingraham and Hershey found that the
measurable base-sequence similarities between lambda phage and E.
coli DNAs were strongest near one end of the lambda DNA molecule
and weakest in the center. They suggest that their results in this case
argue for "a recognition device that does not depend on homology
alone.1' This could possibly be a special enzyme that directs normal
insertion and excision of the prophage.
Dr. Skalka and H. Echols (University of Wisconsin) identified dur-
ing the year two genes in lambda whose primary function may be the
control of transcription. One is concerned with nearly all messenger
RNA synthesis during phage growth, and the other specifically facili-
tates transcription of genes responsible for late functions. R. Werner
demonstrated that the rate of DNA synthesis in another phage, T4,
depends on the number of "growing points" rather than on variation in
rate at a given point. The rate of synthesis at a given point appears to
be constant. He suggests that his results indicate an unanticipated role
for genetic recombination : to distribute "growing points" over newly
synthesized DNA.
Study Systems that Include the Environment
The "bridge" between the gene and the growing organism has been
mentioned as one of the ways of adding to our understanding of the
development of life. One other way of searching for understanding
is to study life in its environment as a cell, as a group of cells, or as
an entire organism.
The Environment in Embryological Study. The Department of Em-
bryology for many years has studied a very special biological environ-
ment, the uterine environment of the mammalian embryo. For several
years investigators, both inside and outside the Department, have
sought ways to sustain the embryo "living free" in culture, detached
from its natural maternal environment. The capacity to culture a
mammalian egg under rigorously defined conditions straight through
from fertilization even to the time of uterine implantation would be a
very important scientific event. D. G. Whittingham, in cooperation with
John Biggers and R. P. Donahue of Johns Hopkins University, has
REPORT OF THE PRESIDENT 65
taken the technique of embryo culture a step further than it had
reached previously. They have succeeded in culturing a mouse zygote
(fertilized egg) to the early two-cell stage in a simple, chemically de-
fined medium and from the late two-cell stage to the blastocyst stage27
under similar conditions. He has discovered, however, that an interrup-
tion of the culture at the early two-cell stage is necessary. The zygote
will not develop into a blastocyst unless it is exposed to the environment
of the fallopian tube28 for a short period between the first and second
cleavages.
Whittingham's experiments have generated some significant infor-
mation about the biochemistry of the development of zygote into blasto-
cyst. He found that the first cleavage division in vitro requires an
energy source, and a very specific energy source, either pyruvate or
oxaloacetate. Cleavage will not take place unless one of the two is pres-
ent in the culture medium. It will not take place in the absence of an
energy source, or in the presence of other compounds like lactate, or
glucose, that might logically seem a source of energy. Whittingham also
found that normally unacceptable energy sources can be used and will
produce cleavage if the follicular cumulus cells29 are placed in the
medium with the one-cell embryos. The follicular cells thus appear able
to metabolize lactate and glucose and another compound, phosphoenol-
pyruvate, into the pyruvate necessary for the fertilized egg's first
cleavage. Pyruvate also appears to be secreted by the oviduct, as was
demonstrated by the cleavage of zygotes in the presence of ovarian
fluid alone, or in an organ culture of the fallopian tube. Thus it has been
established that pyruvate is the essential energy supplier for first
cleavage of the zygote. It was also shown that pyruvate is made avail-
able to the early mouse embryo from two sources, the follicular cumulus
cells and the fallopian tube epithelium. The question of why exposure
to the fallopian tube environment is required in the early two-cell stage,
and only then, has not been answered. It remains a most interesting
problem for further exploration.
Study of Plants in Their Environment. Consideration of the environ-
ment as part of a system for study has long been a subject for the
Department of Plant Biology. Indeed, the Desert Botanical Laboratory,
a predecessor of the Department of Plant Biology, was deeply concerned
with study of the desert environment more than 60 years ago. This
traditional interest of the Department is being carried forward dynam-
ically today by its Experimental Taxonomy Group. Their work is well
27 The blastocyst is a hollow sphere of cells produced by cleavage from the original
zygote just before the migration of cells to specialized sites at which differentiation
within the specific organism commences (gastrulation).
28 The duct leading from an ovary to the uterus.
29 Cells associated with the follicle within which the ovum develops.
66 CARNEGIE INSTITUTION
illustrated by the unexpected results of experiments conducted this year
by 0. Bjorkman and his collaborators. They have found a way to make
many plants, including some common crop plants, grow at as much as
twice the normal rate by a very simple but theoretically startling
expedient. In experiments with beans (Phaseolus vulgaris) and monkey
flowers i Mi mill us cardinalis) it was found that reducing the amount of
oxygen in the air surrounding the plants accelerates their growth.
The average oxygen content of the atmosphere at the earth's surface
is 21 per cent. In Bjorkman's experiments the oxygen content of the air
around the tops of the plants was reduced to either 5 per cent or 2.5
per cent. The roots were given air of normal oxygen content. Bean
seedlings placed in the 2.5 per cent oxygen environment for six days
grew 2.1 times as rapidly as they did in normal air. Growth rates for
longer periods also were notably higher (Plate 6). Monkey flowers
placed in air of the same oxygen content for 10 days increased their
growth by 58 per cent. In a 5 per cent oxygen environment their growth
increase was 90 per cent. A significant exception to the growth increase
pattern appeared in experiments with maize. Corn grown in 5 per cent
oxygen increased its rate of growth by only 16 per cent as compared to
ordinary air.
These experiments were an extension of Bjorkman's earlier work on
the photosynthetic rates of plants maintained in air containing less
than the normal amount of oxygen. As reported last year, Bjorkman
found that many species of plants increased their normal rate of photo-
synthesis by 50 per cent when the oxygen concentration of air around
the leaves was lowered. It is Bjorkman's tentative hypothesis that the
inhibition of photosynthesis caused by oxygen at the level of normal air
is caused primarily "by a back-reaction between a highly reduced
photosynthetic intermediate and molecular oxygen that increases with
0, concentration." It is noteworthy that certain tropical grasses, in-
cluding maize and sugar cane, which lack any major signs of photo-
synthetic inhibition by oxygen, probably use a different "pathway" for
the fixation of carbon dioxide from that employed by most other plants.
C. S. French, Director of the Department of Plant Biology, states in
the Introduction to his Departmental report that "these results open a
new field of inquiry relating to basic differences that have evolved in
the photosynthetic mechanism of higher plants." In addition to its
theoretical significance, Bjorkman's discovery has some immediate
practical value in that it will aid more rapid production of plants for
laboratory studies. It is not too much to hope that increase in plant
growth rate by such a simple change in environment eventually may
have much broader utility.
In a more traditional study Jens Clausen, retired Staff Member of
Plate 6
Report of the President
CT
o
10
c
o
CD
C
T5
C
o
a
</>
0
T3
C
D
a
■a
CD
C
o
c
a;
c
O
CO
>0
a.
REPORT OF THE PRESIDENT 67
the Department of Plant Biology, has published some observations
about the significance of plant species clusters. Clausen states that
modern data indicate that most of the world's tree species are arranged
in clusters of species (cenospecies) . Such clusters circle the earth within
definite latitudinal belts, and are closely enough related to permit
occasional genetic interchange. Examples of these clusters are the larch
genus Larix, which includes six species: the two-needle pine cluster
(type species, Finns silvestris) , the alders (Alnus), the birches
(Betula) , the cypresses (Cupressus-Chamaecyparis) , the beeches
(Fagaceae) , the white oaks {Quercus alba) , and others. Clausen argues
for the application of a conservative concept of species and of genera so
as not to lose the evolutionary significance of the species cluster. He
says that "broad, deep-seated physiological and ecological relationships
. . . built into the various heredity patterns . . . hold the [clusters]
together even on their world-wide migrations."
Progress in Study of Photosynthesis
All higher green plants, the algae, some bacteria, and many simple
marine organisms depend on photosynthesis for their existence. The
entire life pyramid, including all higher organisms, is totally reliant
on this process. Yet photosynthesis appears to be as recalcitrant to
clarification at a fundamental level as it is basic to life on earth. The
Department of Plant Biology has long been heavily committed to the
problem of understanding photosynthesis, at the level of its highly
complex biophysics and biochemistry.
Green chlorophyll, a universal component of the photosynthetic ap-
paratus in green plants, is contained in submicroscopic functional
packages now labeled "photosynthetic units." Hundreds of chlorophyll
molecules are known to comprise a single unit. Light absorbed by any
one of these chlorophyll molecules is transferred from one molecule to
another until it is caught by a particular kind of pigment molecule
known as an "energy trap." Within that trap the energy of a light
quantum is converted to chemical energy. Several hypotheses have been
advanced as to the exact nature of the energy trap, and its precise
manner of functioning is still a matter of lively scientific controversy.
This year D. C. Fork and J. Amesz of the Department of Plant
Biology studied the energy-trapping function in photosynthesis of a
particular pigment known as P700.30 P700 was discovered some years
ago by Bessel Kok of the Research Institute for Advanced Study, Balti-
more, a former Fellow of the Department. P700 has been consid-
ered as one of two possible energy traps within the photosynthetic
30 P700 is a special form of chlorophyll a that acts as a trap for system 1.
0s CARNEGIE INSTITUTION
process known as system 1. The other possible pigment was considered
to be cytochrome /'.
This year Fork and Amesz produced convincing evidence that P700
and not cytochrome / is the energy trap. Experimenting with the red
algae [Iridaea splendens, Schiaymenia pacifica, and Porphyra perfo-
rata I they varied the ratio of the oxidized to the reduced forms of P700
and cytochrome / by exposing the plants to varying light intensities.
For each light intensity the fraction of both cytochrome / and P700
present in the inactive oxidized form was measured. These fractions
were compared with the rate of photosynthesis produced by a given
light intensity. It was found that maximum rates of photosynthesis for
these plants were maintained when cytochrome / was partly inacti-
vated. On the other hand, rates of photosynthesis fell when P700 was
partly inactivated. This is interpreted by Fork and Amesz as meaning
that P700 is the primary energy trap for light absorbed by photo-
chemical system 1.
Another interesting result of the Fork and Amesz experiments was
the discovery that a light quantum traveling into a P700 energy trap
inactivated by oxidation may be passed on to another photosynthetic
unit. They found that there is about a fifty-fifty chance that the travel-
ing light quantum ( exciton ) would move on rather than being converted
into heat and thus wasted.
The Place of Fellows in the Institution
Mindful that "the objects of the Corporation shall be to encourage,
in the broadest and most liberal manner, investigation, research, and
discovery . . . ," the Institution in recent years has evolved post-
graduate scientific training as one of its important functions together
with research.
It is now a commonplace in the scientific world that an essential
part of scientific training occurs in the postgraduate and postdoctoral
years. Universities of course actively conduct this advanced training.
But postgraduate education in the United States has more dimensions
than even universities can give it. Industrial laboratories and engineer-
ing departments of industries, government-operated research institu-
tions— like the National Institutes of Health or the laboratories of the
National Bureau of Standards, for example — privately managed inde-
pendent research institutions, and others, all contribute richly to this
process
Cytochrome / is an iron-containing' hemoprotein found only in plants and related
to other cytochromes; prominent in intracellular oxidations.
-' Carnegie Institution of Washington Articles of Incorporation, Section 2.
REPORT OF THE PRESIDENT 69
Postgraduate training has become so necessary, so ingrained in our
educational structure that postgraduate fellowships have become
numerous; their attainment is now a sough t-f or, even an expected,
event in the young scientist's career. The National Science Foundation
alone supported 405 postgraduate Fellows in its fiscal year 1966, and
285 in 1967. The National Institutes of Health supported 1237 in 1966.
Like so many pragmatic adaptations that distinguish our cultural
history, these fellowships have effectively met a primary need without
involving the trauma of great formal change in our society. They are
an admission that the fresh Ph.D. is not often a mature scientist fully
in command of all the complex tools of his complex metier — litera-
ture, equipment, theory, experimental technique, and professional
communication.
The Carnegie Institution, responding to needs that its senior scien-
tists have long understood, serves as one of these postgraduate training
institutions. During the year the Institution's departments undertook
reviews of their fellowship programs. Although the great value of these
programs was already evident, the place that fellowship work gradually
has come to assume in the life of the Institution still was something of a
surprise when a summary was made. In the period 1952-1967 the
Institution supported 203 Fellows for postgraduate work from its own
funds; funds generously provided by the Carnegie Corporation sup-
ported an additional 62. The National Science Foundation, the National
Institutes of Health, other government agencies, and other foundations
supported still others. Each of these Fellows was with the Institution
at least one full year, and in many cases two, three, or more years,
pursuing his postgraduate education. Many Guest Investigators also
have worked with the Institution's departments during these same
years, sometimes senior in their field, but more often younger scholars
whose period of residence was less than a year.
There are several observations that can be made about the Institu-
tion's remarkably successful fellowship experience of these 15 years.
First, the program has achieved a broad geographical distribution;
Fellows of the Institution in this period came from 25 states of the
United States and 33 foreign countries. Second, a very large percentage
of the Fellows have devoted their subsequent careers to teaching or
other academic work, or to a fundamental research in another type of
research organization. Third, the Fellows as a group have been remark-
ably productive in their subsequent careers, and an important influence
in that productivity was their experience at the Institution. Finally, but
of particular interest, nearly all of the Fellows who came to the
Institution from foreign countries have returned to posts in their own
lands. Of 27 foreign Fellows at the Geophysical Laboratory in the period
CARNEGIE INSTITUTION
mentioned, only one remained in this country at the time of the report,
and he is expected to return home in the near future. Of 40 foreign
Fellows who were in residence at some time during this period at the
Department of Terrestrial Magnetism, only four have remained in the
United States.
Within the 15 years there were 132 Fellows at the Department of
Terrestrial Magnetism and the Geophysical Laboratory. Sixty-five of
the 132 now are in university teaching positions; 42 of the 65 have full
professorships or other permanent posts, and all but a few of these 42
are still below age 40. Fifty-four of the 132 are in posts where they
carry on academic research; 6 have devoted themselves to applied
arch. Thus, 90 per cent of the scholars who have been selected for
postgraduate training at the Institution occupy key positions in the
future progress of science. Among these scholars are a number who are
already considered to be among the leaders in their specialities. The
same judgment can be made for each of the other Departments of the
Institution. The Department of Embryology, for example, has had
about 70 Fellows in this period, including those supported by the United
States Public Health Service and other outside sources of funds. Nearly
all have taken up careers in teaching or academic research, and they
include an impressive share of the younger leaders in this field.
The productivity of these men is attested in part by the positions they
now occupy and their scientific reputations. Another measure of pro-
ductivity is given by papers describing the results of their experimental
and theoretical work published in professional journals. This record has
been truly gratifying to all whose vision endorsed and encouraged the
fellowship program at the Institution. The Geophysical Laboratory, for
example, compiled a bibliography of "Papers Arising in Part or Com-
pletely from Work Performed by Visiting Staff." The 41 Fellows listed
in this bibliography published 134 such papers between 1952 and 1967.
Philip Abelson, Director of the Geophysical Laboratory, comments in
his report on Fellows and Guest Investigators that they have, singly or
in collaboration, participated in 70 per cent of the investigations de-
scribed in the most recent annual report of the Laboratory. I can offer
no more telling observation on the place that postgraduate training has
in the professional life of the Institution's more permanent staff.
The Fellows at the Institution are expected not only to participate in
current research, but also to master pertinent subjects in which
their previous training was inadequate. For example, Fellows at the
Geophysical Laboratory frequently commence their work with inade-
quate backgrounds in physical chemistry. As Abelson states, "They
attain mastery of relevant aspects of physical chemistry during their
tenure." At almost any time of year that might be chosen, it is likely
REPORT OF THE PRESIDENT 71
that one or more seminars on special subjects of training interest will
be going on at the Institution, including such topics as computer pro-
gramming, genetic coding, irreversible thermodynamics, or quasars.
Without exception the Departments of the Institution consider it im-
portant that Fellows undergo intellectual growth during their tenure.
One former Fellow recently wrote about his experience at the Institu-
tion, "It has enabled many of us to acquire skills and insights not
available in even the best-equipped graduate schools. . . ."
There is an equivalent benefit for the Institution in this process, of
course. Thus Ellis Bolton, Director of the Department of Terrestrial
Magnetism, writes from the point of view of the staff, "Our Fellows
have had . . . and will have the profoundest kind of influence upon the
work of the Department as it progresses." The best of the fellowship
experiences, and this includes a majority, have left both Staff Member
and Fellow intellectually richer.
We expect to continue to provide these opportunities for both staff
and Fellows at a rate that members of our staff believe is feasible. No
less than by its substantive contributions to our knowledge of the his-
tory and structure of the universe, the Institution would wish to be
known by these results of the compelling personal experiences of young
scientists.
Losses . . .
It is with great sadness that I record here the death of a former
Trustee, Elihu Root, Jr., on August 27, 1967, in New York City, at the
age of 86. Mr. Root had just retired as Trustee of the Institution at the
Annual Meeting on May 5, 1967. The occasion marked the first time
since the founding of the Institution that there has been no member of
the Root family on the Board of Trustees. Mr. Root's father, the Secre-
tary of State and the Secretary of War, was elected in 1902 to serve on
the Institution's original Board. He remained in that position until his
death in 1937, when Elihu Root, Jr., was elected a Trustee.
During his 30 years as a Trustee, Elihu Root, Jr., took an extraor-
dinary part in the affairs of the Institution. From 1950 to 1956 he
served as Chairman of the Board. After his work in that capacity was
completed, he continued most actively in the affairs of the Institution,
both on the scientific and administrative sides. Rarely indeed did he
miss a meeting of the Executive Committee, or of the Finance Com-
mittee, on which he was also active over many years. He was a constant
inspiration to his colleagues on the Board, and to the Administrations
that served with him.
Elihu Root was born in New York City in 1881. Like his father and
CARNEGIE INSTITUTION
his son, he was graduated with a B.A. degree from Hamilton College in
Clinton, New York. After earning an M.A. from Hamilton, he went on
to Harvard Law School where he received the LL.B. degree in 1906.
Shortly after his graduation he founded the firm of Root, Clark and
Bird with two of his young law school classmates. One of his first clients
as a young attorney was Andrew Carnegie, a man who, as Mr. Root
said, "did not hold lawyers in awe." The firm of Root, Clark and Bird
ultimately became one of the most distinguished in the nation and
claimed many notable personalities among its clients. Mr. Root stayed
with the firm, which eventually became Dewey, Ballantine, Bushby,
Palmer and Wood, until 1954, when he became counsel to the firm of
Cleary, Gottlieb, Steen and Hamilton.
Although his law practice was a busy and stimulating one, Mr. Root
found time to serve simultaneously as a director of several large corpo-
rations, including the Mutual Life Insurance Company of New York
and American Telephone and Telegraph.
A patron of arts and letters as well as an artist himself, Mr. Root
was a trustee of the Grand Central Art Galleries, New York's Metro-
politan Museum of Art, and the New York Public Library. With char-
acteristic thoughtfulness, he and his brother Edward assembled a
notable collection of oil paintings and prints, which they donated to
Hamilton College to be borrowed by the students for their rooms.
Twice Mr. Root's career was interrupted by what he called "the
unfortunate business of making war." During World War I he saw
active service as a major in the 304th Infantry in France. And at home
he was a pioneer in the Plattsburgh Movement, established to set up
voluntary military training camps for civilians. By the time war was
formally declared, it had furnished thousands of trained junior officers
to the nation. During World War II Mr. Root served as a member of
the U. S. Army Air Corps strategic target board, which advised on
points where enemy economies should be attacked. For this work he was
awarded the Medal for Merit by President Truman.
These are but a representative few of the achievements and activities
that occupied Elihu Root's busy life. His absence will be keenly felt by
all who knew him, and his memory will be treasured for many years to
come.
With equal sadness I record here the passing of another fellow
Trustee, another former Chairman of the Board of the Institution, and
a dear friend — Barklie McKee Henry. He died on September 4, 1966, at
Blue Mountain Lake in the Adirondacks. His personal interest in the
Institution was constant and intense; his wise counsel was invaluable
and inspiring.
REPORT OF THE PRESIDENT 73
Mr. Henry became a Trustee in 1949 and at once assumed a most
active part in guiding the Institution and shaping its philosophy in the
context of its role as a unique research organization. In 1961 he became
Chairman of the Board of Trustees and remained in that position until
1965. Through all the years of his Trusteeship he brought the formi-
dable range of his talents and abilities to bear on the fortunes of the
Institution. No problem was too difficult for him to tackle, nor any detail
too minute to escape his notice. He was a frequent and welcome visitor
both in the Office of Administration and at the various departments of
the Institution. Staff Members looked forward particularly to his visits
and the opportunity to seek his thoughtful, understanding, and knowl-
edgeable advice.
Mr. Henry was born in Ventnor, New Jersey, in 1902, and was grad-
uated cum laude from Harvard University in 1924 with an A.B. degree.
After some further study at Balliol College, Oxford University, he be-
came managing editor of the magazine, Youth's Companion. A year
later he joined the Guaranty Company of New York.
In 1930 Mr. Henry left his business career to devote himself fully to
those things in life that he valued most. Among these was his interest
in a rapprochement between medicine and psychiatry, which he did
much to encourage through his work with the New York Hospital. After
serving the Hospital in various other capacities, he became its president,
and was influential in strengthening the psychiatric services it offered.
The inventive side of his personality served him well in many instances.
For example, as president of the New York Hospital he saw to it that
the grim bars in the psychiatric ward were replaced by mesh screens,
now in wide use.
In the early days of the Depression he was active in the Emergency
Unemployment Relief Organization in New York, an unusually success-
ful self-help undertaking at the local level. He later managed the merger
of two similar organizations to form the Community Service Society,
the largest private family welfare agency in the United States, of which
he became the first president.
In 1940 he went on active service with the Navy and during the war
commanded first a coastal mine sweeper, then a submarine chaser, and
finally a destroyer escort. Later, he was called upon by Secretary of the
Navy Forrestal to help with contract terminations.
In addition to all these activities he was a trustee of the Pierpont
Morgan Library, the American Academy in Rome, the Rockefeller
University, the John Hay Whitney Foundation, and the Institute for
Advanced Study in Princeton.
A list of achievements, however, does not do justice to Barklie
Henry's life. For a list does not intimate the manner in which these
:: CARNEGIE INSTITUTION
achievements were brought about. It does not reveal the vibrance or the
resolution he brought to all things he encountered.
Barklie Henry is sorely missed for his skill and his acumen, his in-
sight and his brilliance, his understanding and his great warmth. But
most of all he is missed as a friend and a companion, because above all
else, he was a profoundly humane human being.
It is with a very special sense of loss that I must record the resigna-
tion from the Board of Trustees of Mrs. Margaret Carnegie Miller,
which was accepted, though reluctantly and with keen regret, by the
Board at its meeting on May 5, 1967. Mrs. Miller is deeply missed. But
we are particularly happy that she remains near us, accessible and
available to give of her advice to the Institution that she has known and
watched from its very beginning.
Although with the retirement of Dr. Barbara McClintock on June 30,
1967, the Institution lost as a Staff Member one of the most distin-
guished American geneticists of all time, I am particularly happy to
record that she has consented to return as the third Distinguished
Service Member in Institution history, and will continue her research
program without interruption at the Genetics Research Unit in that
capacity.
The Achievement Award of the American Association of University
Women was presented to Dr. McClintock in 1947. On April 30, 1965,
Cornell University bestowed the honorary office of "Andrew D. White
Professor-at-large" on Dr. McClintock, and on April 24, 1967, she was
awarded the Kimber Genetics Medal of the National Academy of
Sciences "for a series of brilliant and significant studies on chromosome
structure and function. "
And Gains . . .
We are fortunate indeed this year in welcoming two new Trustees of
high distinction — Eric Ashby and William McChesney Martin, Jr., both
elected to the Board on May 5, 1967.
Sir Eric Ashby, F.R.S., was born in Bromley, Kent, England, in 1904,
and was educated at the City School of London and the Imperial College
of Science, University of London. During the two following years he
was a Commonwealth Fund Fellow at the University of Chicago and
served as a Fellow at the Desert Laboratory of the Carnegie Institution.
In 1931 he returned to England to become Lecturer at the Imperial
College of Science, and Reader in Botany at Bristol University. Later
REPORT OF THE PRESIDENT 75
he served as Professor of Botany successively at the University of
Sydney in Australia and at the University of Manchester in England.
From 1950 to 1959 he was President and Vice-Chancellor of the Queen's
University, Belfast, Northern Ireland.
In 1942 Sir Eric conducted an inquiry for the Prime Minister of
Australia into the enlistment of scientific resources during World War
II. He was thus instrumental in organizing scientists for the war effort
in Australia. In 1945 and 1946 he was Counsellor and Charge d'Aff aires
at the Australian Legation in Moscow.
His contributions to education are unique. He has pioneered the
development of higher education in Africa, India, and Australia, as well
as in Europe. From 1958 to 1960 he was Chairman of the now famous
"Ashby Commission" from which the blueprint for higher education in
Nigeria was derived.
Sir Eric is at present Vice-Chancellor of Cambridge University. He
continues to serve as Honorary Advisor to the Nigerian National Uni-
versities Commission and as a member of the Council of the University
College of Makerere, Uganda.
William McChesney Martin, Jr., at present Chairman of the Federal
Reserve Board, began his career in the Bank Examination Department
of the Federal Reserve Bank in St. Louis in 1928. A year later he joined
A. G. Edwards and Sons, also in St. Louis, and was a partner in that
firm from 1931 until 1938. In 1931 he also became a member of the New
York Stock Exchange, and in the same year edited and published
Economic Forum, a scholarly journal published quarterly. He remained
with the Exchange until 1941, serving successively as Member, Chair-
man of the Board, and the youngest President in the history of the
Exchange.
In 1941 Mr. Martin joined the Army, and by 1945 had risen from
private to colonel. In that year he was appointed to the Board of Direc-
tors of the Export-Import Bank, and in the two succeeding years served
as Chairman of its Board. In 1949 he became the United States Execu-
tive Director of the International Bank for Reconstruction and Develop-
ment, and also served for a year as Assistant Secretary of the Treasury.
In 1951 he was appointed Chairman of the Federal Reserve Board.
Mr. Martin has received honorary degrees from Yale, Amherst, and
Harvard, and from Washington University, St. Louis. He serves cur-
rently as Trustee of Yale University, The Johns Hopkins University,
and the Foreign Service Educational Foundation.
The Institution eagerly anticipates a long and pleasurable associa-
tion with both Sir Eric Ashby and Mr. Martin.
76 CARNEGIE INSTITUTION
It gives me the greatest pleasure to record the several honors that
have come to Staff Members during the past year.
Dr. Allan R. Sandage of the Mount Wilson and Palomar Observa-
tories was awarded the Gold Medal of the Royal Astronomical Society
for his fundamental work on stellar evolution and the history of the
Galaxy, and for his further work in providing optical data basic to
modern cosmology. In addition, Dr. Sandage was invited to be the
Sigma Xi National Lecturer in October 1966, the Holiday Lecturer
for the American Association for the Advancement of Science in
December 1966, and the Vanuxem Lecturer of Princeton University in
February 1967; and he was the recipient of honorary Sc.D. degrees
from the University of Chicago and the University of Illinois.
Dr. Alfred D. Hershey of the Genetics Research Unit received the
honorary degree of Doctor of Science from the University of Chicago on
May 5, 1967, at a special convocation celebrating the seventy-fifth
anniversary of the University.
Dr. James D. Ebert of the Department of Embryology was elected to
membership in the National Academy of Sciences, and was also elected
to membership on the Board of Scientific Overseers of the Jackson
Laboratory.
Robert P. Kraft of the Mount Wilson and Palomar Observatories
was an invited lecturer for the American Astronomical Society at the
summer meeting at Yerkes Observatory in June 1967 on Stellar Ro-
tation, Stellar Evolution, and the Conservation of Angular Momentum.
On December 16, 1966, Dr. Donald D. Brown of the Department of
Embryology was named "Maryland's Outstanding Young Scientist of
the Year" by the Maryland Academy of Sciences.
The American Society of Zoologists elected Dr. David W. Bishop of
the Department of Embryology as Chairman of their Division of
Development Biology.
Reports of Departments
and Special Studies
Department of Terrestrial Magnetism
Committee on Image Tubes for Telescopes
Department of Plant Biology
Mount Wilson and Palomar Observatories
Geophysical Laboratory
Department of Embryology
Genetics Research Unit
Department
of Terrestrial Magnetism
Washington, District of Columbia
Ellis T. Bolton
Director
L. Thomas Aldrich
Associate Director
Carnegie Institution Year Book 66, 1966-1967
Contents
Introduction 7
Geophysics 7
Geomagnetism 8
Cosmic-ray program 8
Electrical conductivity studies of the mantle under the Andes ... 12
Explosion seismology 13
Time-term analysis — ECOOE 13
Wide-angle reflections and amplitudes from ECOOE 24
Late arrivals and amplitudes 25
Earthquake seismology 28
Q for P waves in the mantle 28
Shear-wave absorption in the upper 130 km 29
A path for high-frequency shear (S) waves in the upper mantle ... 31
Converted waves from the interface at a depth of 410 km .... 32
A search for small, deep earthquakes in the Andes 35
On the spatial distribution of earthquakes near San Juan, Argentina . 37
Isotope geology 42
Potassium, rubidium, and strontium in ultramafic rocks and minerals . 42
Geological history of the Grenville province 44
Heat flow 52
Technique and results 52
Astrophysics 57
Optical astronomy 57
Radio astronomy 59
Northern hemisphere 59
Southern hemisphere 61
Nuclear physics 62
Elastic scattering of polarized protons on nuclei of spin zero .... 62
Elastic scattering of polarized protons on deuterons 64
Atomic physics 66
Atomic and ionic spectroscopy with foil excitation 66
Biophysics 68
An instructive glossary . 68
Repeated nucleotide sequences 73
Time course of DNA reassociation 76
Precision of sequence matching and fractionation 78
Rate of divergence of nucleotide sequences during evolution .... 80
Saltatory events of replication 83
Chromosome pairs 86
A source of divergence? 87
Intramolecular heterogeneity of the DNA of temperate bacteriophages . 88
Nucleotide sequence relationships among phages <£80 and A, and E. coli . 95
DNA-DNA reactions involving identical DNAs 96
Heterologous DNA-DNA reactions 96
Qualitative aspects of microbial DNA duplexes 106
Memory and learning mechanisms 118
- Cited 124
Bibliography 125
Personnel 129
Department of Terrestrial Magnetism
i iiiii
NGC 4038
O
00
II
CK
III
II II IIIII IIIII III II
] IIIII
III llll IIIII III I
I IIIII
II IIIII 1 111
Spectra of peculiar galaxy NGC 4038, taken with the DTM image tube spectrograph at Kitt
Peak National Observatory. Line emission arises from bright knots in galaxy; these knots may
be identified by insets at right showing orientation of spectrograph slit across galaxy. Emission
lines going completely across spectra are due to night-sky radiation.
INTRODUCTION
The thrust of the Department of we eschew cooperation with others,
Terrestrial Magnetism's efforts con- or fail to rise to the opportunities
tinues to emphasize fundamental presented by sometimes very large
work in physics. Our activities con- undertakings such as the seismic
cern astrophysics, biophysics, and studies of the East Coast reported
geophysics. Within each of these very here. Whenever we become involved,
encompassing areas we have chosen however, the individual scientist is
to explore limited problems that are intimately involved in every phase of
both significant and approachable by the activity from, for example, dig-
a small staff. We continue also to ex- ging holes for seismometers, to writ-
press ourselves through personal re- ing programs for the computer and
search in which the individual bringing wisdom to bear on the
scientist, perhaps in concert with a meaning of the results. That this is
colleague or two, becomes deeply im- a satisfying and valid pursuit, and
mersed in his quest and remains a successful way of research in all
unbuffered from gathering the ma- the areas we investigate, is keenly
terials he needs or from using the illustrated in the scientific reports
tools he requires. This is not to say that follow.
GEOPHYSICS
L. T. Aldrich, R. Cabre, S.J., M. Casaverde, G. L. Davis S. del Pozo, L. Fernandez, S.J.,
S. E. Forbush, E. Gajardo, A. A. Giesecke, Jr., S. R. Hart, D. E. James, T. E. Krogh,
H. Marconi, F. Munizaga, P. N. S. O'Brien, A. Rodriguez, G. Saa, I. S. Sacks, R. Salgueiro,
T. J. Smith, J. S. Steinhart, A. M. Stueber,
S. Suyehiro, and F. Volponi
The geophysical description of our
planet involves many of the mathe-
matical and physical tools at man's
disposal. Our report describes the
efforts of a dozen scientists — Staff
Members, Fellows, and Research As-
sociates of the Department — to use
these tools creatively. The report
year started auspiciously with the de-
livery of a "small" digital computer
system (IBM 1130). It has been a
much-used part of the efforts de-
scribed below and has extended our
capability most significantly.
The range of topics undertaken by
this group is in itself a demonstration
of the varying needs of modern geo-
physics. The statistical study just
completed of the diurnal variation of
the cosmic-ray flux bombarding the
earth demonstrates the great value of
a 30-year collection of carefully made
observations, faithfully obtained and
reported by colleagues in Christ-
church (New Zealand), Cheltenham,
Fredericksburg (United States), and
Huancayo (Peru). The time-term
analysis of some 800 seismic observa-
tions made during the summer of
1965 demonstrates the complexity of
the requirements for describing the
structure of the earth's crust over a
large area, our good fortune in hav-
ing friendly observers to obtain the
data, and the need for an interested
government agency to support the
activities of the North American Ex-
plosion Seismology Group.
Intensive efforts in measuring
thermal gradients, thermal conduc-
tivity, and the bottom temperature
regime of Lake Superior have pro-
s
CARNEGIE INSTITUTION
vided a new appreciation of the vari-
ations in this parameter in sup-
posedly stable shield areas of the con-
tinents. For the first time, from these
measurements one can plot meaning-
ful contour maps of the rate at which
heat tlows from the earth's surface.
The study will be extended to the
remainder of Lake Superior this
summer with the helpful collabora-
tion of the U.S. Coast Guard.
In the area of isotope geology a
continuing study of the North Amer-
ican Precambrian shield is reported
and isotope data are given for ultra-
mafic rocks in limited areas of North
Carolina.
Significant results have been ob-
tained from the operation of a net of
four seismic stations with high dy-
namic range and wide frequency re-
sponse. The net extends from New
Guinea through Chile and Peru to
Washington. The original purpose of
the net has yet to be realized, but
already definitive evidence for a ve-
locity discontinuity in the mantle
under South America has resulted
from a study of deep-focus earth-
quakes. This net will have been
doubled by the end of August 1967.
GEOMAGNETISM
Cosmic-Ray Program
S. E. Forbush
Tht variation, with a period of two
solar cycles, in the cosmic-ray diurnal
anisotropy, 1937—1965, and the super-
posed variations correlated tvith mag-
netic activity. Annual means, 1937-
19 65, of the cosmic-ray diurnal ani-
sotropy component, in the asymptotic
direction 128° E of the sun, contain
a well determined wave with a period
of 20 years, which is twice the solar
cycle period of 10 years for the in-
terval 1937-1 905. This variation is
independent of magnetic activity,
and, when removed, the residual var-
iations combined vectorially with
those in the asymptotic component
38° E of the sun (in which there is
no 20-year wave) give variations
which are principally in the asymp-
totic component 90° E of the sun.
Yearly means of this resultant com-
ponent are well correlated (r =
+ 0.75) with magnetic activity U0
and, on the average, vanish for U0 =
0. U0 is the absolute value of the
southward geomagnetic component of
the so-called equatorial ring current
(Forbush1) . The amplitude of the 20-
year wave is 60% of the amplitude
for the 1937-1965 average diurnal
anisotropy 90° E of the sun. The
wave passes through zero in the
middle of 1958 near the time shown
by Babcock2 for the reversal of the
sun's general magnetic field.
In the Archimedean spiral streams
from the sun, Ness and Wilcox3 found
a decided tendency for the magnetic
field, in the plane of the ecliptic, to
be oriented parallel or antiparallel to
the theoretical direction (135°) pro-
posed by Parker.4 This is approxi-
mately the direction of the cosmic-
ray diurnal anisotropy component
with the 20-year variation. Wilcox
and Ness5 found that, within sectors,
the field directions were consistent
with the outward or inward solar
fields over the solar region from
which the sectors originated. These
facts may bear on the mechanism
underlying the 20-year variation in
the cosmic-ray diurnal anisotropy.
These results are based on the
pressure-corrected ion-chamber data
gathered at Cheltenham-Fredericks-
burg and at Huancayo for the period
1937-1965, and at Christchurch for
the period 1937-1961. To eliminate
the so-called local temperature effect
which remains after correcting to
constant barometric pressure, the de-
viation of each yearly mean vector
from a 25-year mean, in the 24-hour
LMT harmonic dial, is obtained for
each station. These yearly deviation
DEPARTMENT OF TERRESTRIAL MAGNETISM 9
vectors for Huancayo and for Christ- fitted to N_i6, the yearly deviations
church are normalized to those for from the 25-year mean, of the diurnal
Cheltenham by a clockwise rotation anisotropy 128° E of the sun. The
of 37° and 20°, respectively, and by amplitude of the 20-year wave is zero
multiplying respective amplitudes by in P_10, the yearly deviations from
0.91 and 1.00. Components of these the 25-year mean of the diurnal ani-
yearly deviation vectors at Chelten- sotropy 38° E of the sun. The devia-
ham and of the normalized deviation tions (N-ie — W) of points from the
vectors for Huancayo and for Christ- wave in Fig. 1 are correlated (r —
church are taken on a right-handed +0.75) with U0 and with yearly
set of axes, Pe and Ne, with the P0 means of P_1C (r = +0.65), which
axis making an angle (9, in degrees, are also correlated with U0 (r =
positive clockwise, with the observed 0.59). The components parallel, P36,
25-year average vector for Chelten- and perpendicular, N36, to the asymp-
ham, which has its maximum at 13.8 totic direction 90° E of the sun are
hours local solar time in the 24-hour obtained from P_16 and (N-1G — W).
harmonic dial. The mean value of N3G is zero, and
An analysis of variance for the 25 the correlation between 2V36 and U0 is
values of P0 and N0 at the three sta- negligible.
tions shows that the variance among The correlation (r = +0.75) be-
years for P0 and for N0 is at least 20 tween P36 and U0 is shown by the
times the residual variance, definitely open circles in Fig. 2. Making use of
indicating that the diurnal anisotropy these correlations, the values of iV_16
was not constant from 1937 to 1961. in Fig. 1 are corrected to U0 = 0 (no
The residual variance corresponds to magnetic activity). These corrected
a standard deviation of 0.017% for values of AT_16 are well fitted in Fig. 3
the statistical uncertainty of single by the 20-year wave used in Fig. 1.
yearly values of P0 and of N0 at each This fit indicates that the 20-year
station. The corresponding standard variation is independent of magnetic
deviation estimated from the varia- activity. Using the above correla-
bility, within months, of daily mean tions, and assuming that the total
24-hour waves at Cheltenham is diurnal anisotropy 90° E of the sun
0.014%. It may be compared with vanishes for U0 = 0 (i.e., no mag-
the value 0.011% that would result netic activity), the resulting ampli-
if the only uncertainties were those tude of the diurnal anisotropy 90° E
due to the finite sampling or "count- of the sun, averaged during the
ing rate" of the detector. This shows period 1937-1965, is 0.12%.
that the so-called temperature effect As indicated above, the 1937-1965
is effectively eliminated by using the mean of N36 is zero and the correla-
deviations of the yearly means of the tion between yearly means of N36 and
diurnal variation from a 25-year of U0 is not significant. Furthermore,
average. It also shows the excellent the variance of annual means of N3G
agreement between the values of P0 is only 23% of that for P36. This was
and N0 from the three stations. Thus the basis for assuming heretofore
yearly values of Pe and Ne, averaged that the variability of P36 arises from
for all stations, are used hereafter; variations, from the 25-year average,
for three stations (1937-1961) and of the component in the diurnal ani-
for two (1961-1965) the respective sotropy 90° E of the sun. Similarly,
standard deviations are 0.010% and the time of maximum of the 1937-
0.012% for Pe and for Ne. 1965 average vector, say C1S (ampli-
Figure 1 shows the 20-year wave tude 0.12%), for the total diurnal
10
CARNEGIE INSTITUTION
1940
1950
1960
Fig. I. Twenty-year wave fitted to deviations, N_u, of yearly means from the 1937—1961
average of the diurnal anisotropy component in the asymptotic direction 128° E of the sun.
anisotropy is taken as 18.0 hours
local asymptotic time. For each sta-
tion, the local solar time of maximum
of the vector C18 determines the addi-
tive correction required for geomag-
netic deflection if, in fact, the local
asymptotic time of maximum for C1S
is 18.0 hours.
Those corrections, in hours, are:
for Godhavn (not otherwise used
herein), 1.2; Cheltenham, 1.7; Christ-
church, 2.7; and Huancayo, 4.1. The
values are reasonably compatible with
those obtained from calculations of
^omagnetic deflections by Venkate-
san and Dattner" and by McCracken
et al.7 This correlation indicates that
the local asymptotic time of maxi-
mum for CV-, is near 18.0 hours.
For Huancayo the difference, in the
24-hour harmonic dial on local solar
time, between the observed 1937-
1965 average diurnal variation vector
and the vector (d8/0.91) gives the
average vector for the so-called tem-
perature effect. This finding is in fair
agreement with preliminary values
from a few yearly mean differences
between the diurnal variation in the
ion chamber and that from Simpson's
IGY neutron monitor.
These results comprise a self-con-
sistent, statistically satisfactory ac-
count of the variability of yearly
means of the diurnal anisotropy and
of its average for the period 1937-
1965.
Monte Carlo experiments with
filters. It has been indicated by some
investigators that using digital com-
DEPARTMENT OF TERRESTRIAL MAGNETISM
11
o
o
<-t-
o
OlP
P,.= -12.9+ 0.29 Un
36 r=+0.75 °
Legend: o Annual means
• 3-Year means
Except for 1937-38
20
40
60
U0inX
Fig. 2. Correlation between magnetic activity, Uq, and the deviations P^6> of yearly means from
the 1937—1961 average, of the diurnal anisotropy component in the asymptotic component 90° E
of the sun.
puters to "filter" cosmic-ray data
would improve the signal-to-noise
ratio for the diurnal variation. In
collaboration with Dr. Martin Pome-
rantz and Dr. S. P. Duggal of the
Bartol Research Foundation, Monte
Carlo experiments to test this sug-
gestion were conducted on the Bartol
computer.
Bihourly values were computed
from a synthetic 24-hour variation
having constant phase and amplitude.
To each of these bihourly values,
comprising about a "year," was added
a random variate from a population
with zero mean and known variance.
With the computer these values
were "filtered" by each of three dif-
ferent filters designed to remove
noise and to pass the synthetic diur-
nal variation. For each "day" the
24-hour wave was determined by har-
monic analysis of bihourly values.
Statistical analyses of the results (in-
volving about one year of synthetic
data for each filter) demonstrated
that the reliability of the average
wave was the same whether the data
were filtered or not. Thus, unless the
data contain persistent waves with
frequencies that are not harmonics
of the 24-hour wave, filtering does
not improve the results obtained for
the 24-hour wave. Since there is no
evidence for such waves in cosmic-
ray data, filtering results in no im-
provement.
Observations and reductions of
data. Cosmic-ray ionization chambers
were operated throughout the report
year at Huancayo and at Fredericks-
burg, Virginia. Scalings and reduc-
tion of records have been maintained
current for both stations.
Cooperation in operation of cosmic-
ray meters. Grateful appreciation is
12
CARNEGIE INSTITUTION
O
o
-10
1940 1950 1960
Fig. 3. Twenty-year wave from Fig. 1 fitted to N_i6, corrected to Uq — 0.
expressed to the U.S. Coast and Geo-
detic Survey and the staff of its mag-
netic observatory at Fredericksburg
for efficient operation of the meters
during the past report year, and to
the Government of Peru and the
Director and staff of the Instituto
Geofisico del Peru for making cos-
mic-ray records from Huancayo
available. /
Electrical Conductivity Studies
of the Mantle under the Andes
S. E. Forbush, M. Casaverde, R. Salgueiro,
S. del Pozo, and L. T. Aldrich
The report year has been largely
a period of preparing equipment and
recording sites for observations.
Some data have been obtained on lo-
cations distant from the Andean pla-
teau, but the reduction of this infor-
mation is still to be completed. The
field season beginning July 1, 1967,
is expected to provide data from the
ten portable Askania variographs,
three DTM versions of the vario-
graph developed at Gottingen Uni-
versity, and two or three la Cour in-
struments with increased sensitivity
in their D components. This equip-
ment will be thinly deployed over an
area of about 12° latitude by 12°
longitude in Peru and Bolivia. It is
hoped that the study may also be ex-
tended into northern Chile this year.
This array of stations, designed in
consultation with Dr. U. Schmucker,
should provide the broad outlines of
the region of anomalously high elec-
trical conductivity as a minimum. If
a sufficient number of stations can be
operated simultaneously, they will
also provide an extensive survey of
the equatorial ring current.
The Askania instruments have
been revised at the Department to
have a more positive and accessible
drive system, which will also provide
as an integral part of the system a
12-hour switch to give positive indi-
DEPARTMENT OF TERRESTRIAL MAGNETISM
13
cation of the D trace. The chronome-
ters for all the instruments have been
revised to provide hour marks and
15-minute marks using low enough
power to be operated by the main As-
kania battery. This revision of the
chronometers has included a new
waterproof housing to maximize their
usefulness at jungle stations. The
DTM version of the Gottingen vario-
graph has been designed to provide
minute marks for easy digitizing
(and reduced power requirements)
as well as hour and 15-minute marks.
Each timing system is independent,
so the measurement of this crucial
element is relatively certain.
EXPLOSION SEISMOLOGY
T. J. Smith, J. S. Steinhart,
P. N. S. O'Brien, and D. E. James
For the past several years, much of
the effort at the Department in ex-
plosion seismology has been directed
toward the effective utilization of
time-term analysis in the hope of
producing realistic representations of
crustal structure. Dissatisfaction
with traditional plane-layered models
as a means of describing the crust
led us to apply time-term analysis to
data from the Lake Superior experi-
ment (see Year Books 62, 63, 64) .
This method proved so successful in
delineating crustal structure that an
even more ambitious time-term anal-
ysis was planned for the East Coast
Onshore-Offshore Experiment
(ECOOE). A description of the ex-
periment and a map showing the lo-
cations of shots and stations were
given in Year Book 65. Reexamina-
tion of the Lake Superior results al-
ready reported (see Year Book 64)
has continued.
Time-Term Analysis — ECOOE
D. E. James, T. J. Smith, and J. S. Steinhart
This study was focused upon the
problem of obtaining a three-dimen-
sional representation of crustal struc-
ture beneath the Middle Atlantic
states and investigating the extent to
which Bouguer gravity anomalies
predicted by the seismic structure
can be correlated in size and shape
with observed gravity anomalies.
Subcrustal velocity. Provision is
made in the time-term analysis to
compute subcrustal velocity. It is,
however, impossible to determine ve-
locities by the time-term method
when the profile is completely or al-
most completely unreversed. In the
strict sense, only the ECOOE South-
ern Line observations were single-
sided; the Northern Line, while heav-
ily unbalanced toward shots at sea,
does have many observations of the
Schuyler shots and hence is partially
reversed. Thus, the two shot se-
quences represent somewhat different
problems and are treated as two dis-
tinct groups of data.
The subcrustal velocity for the
Northern Line was found to be 8.15
km/sec by the usual time-term
method. Because there is no direct
way of computing the Southern Line
velocity by time-term analysis, we in-
vestigated the behavior of time terms
when the refractor velocity was con-
strained to incorrect values. By this
means we sought to develop a set of
criteria by which we could determine
the refractor velocity for a single-
ended profile. A dipping-layer time-
term model, originally scaled to Lake
Superior P1 structure, was con-
structed and the time terms were cal-
culated using different constrained
velocities. The result is shown in Fig.
4. Station 16 was used both as a shot
point and as an observing site, and
the shot was observed by sites 1
through 15. As the velocity is con-
strained to incorrect values, the time
terms undergo a distortion which
produces a discontinuity between the
shots and the stations.
The Southern Line time terms
were computed for the range of con-
14
CARNEGIE INSTITUTION
E
Model 1S66
* Shots
o Stations
25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9
f 500
Vn ■ 5.5 km/sec
15.07 km
7 6 5 4 3 2 1 400300200100
7.95 km
///tii
V,= 6.5 km/sec
50 km
Fig. 4. Constrained velocity set of time terms for Lake Superior model for velocities of 6.30,
6.50, and 6.70 km/sec.
strained velocities from 8.06 to 8.50
km sec. The time terms of the four
on-line stations nearest shore and a
selected set of on-line shots were
plotted in the same fashion as the
model time terms. The Schuyler time
term is taken to be 3.67 seconds on
the basis of the Northern Line re-
sults. The fact that the Schuyler time
term is fixed is central to this tech-
nique. The results for the entire set
of constrained velocities are shown
in Fig. 5. It is clear from the figure
that a demand for time term conti-
nuity from shot sequence to stations
sharply limits the range of acceptable
velocity values.
On the assumption that the struc-
ture is smoothly varying from sta-
tions to shots (less than 30 km sep-
arate the most easterly station from
the most westerly shot), a velocity
close to 8.25 km/sec appears to be
most appropriate for the Southern
Line data. Although this velocity
value depends upon the normalizing
value of the Schuyler time term, it
varies only between about 8.22 and
8.28 for possible Schuyler time terms
as computed for the Northern Line.
It is clear that a velocity of 8.15 km/
sec is incompatible with the condition
of shot-station time-term continuity.
The 0.10 km/sec difference in Pn
velocity between the Northern and
Southern Lines is not particularly
surprising. Herrin and Taggart8 have
shown that large changes in Pn veloc-
ity do occur over comparatively small
areas, and there is no a priori reason
for expecting the Pn velocities for the
Middle Atlantic states to be particu-
DEPARTMENT OF TERRESTRIAL MAGNETISM
15
Coast Line
Fig. 5. Constrained velocity set of time terms for Southern Line in-line shots and stations.
larly homogeneous. The Pn velocities
for the final time-term solutions are
given in Table 1, together with the
total number of observations and the
rms residual for all observations.
Structural distortion of time terms.
Asymmetry of the recording network
will clearly produce distortions in the
time terms if the structure being in-
vestigated is sufficiently complex. To
examine the size and nature of these
distortions, the single-layer geomet-
rical model shown in Fig. 6 was con-
structed to the scale of the East Coast
experiment. Least-time ray paths
were measured graphically to obtain
a set of travel times corresponding
to first-arrival refracted events, and
these synthetic observations were
then subjected to time-term analysis.
The resulting model time terms are
also shown in Fig. 6 along with the
"true" time terms given by
where H is the depth measured along
a normal from the surface site to the
interface. There are regions in which
the normal is not unique, but the
admissible choices yield results that
are identical for purposes of this
study. It will be seen from the figure
that station 109, which models the
Schuyler site, has been taken as a
shot point as well as a receiving sta-
tion and thus serves as a normaliza-
tion point for the model results.
This simple study indicates clearly
TABLE 1 . Velocity, Number of Observations, and RMS Residual
for Northern and Southern Line Observations
Shot Sequence
Velocity
No. of Obser-
vations
RMS Residual,
seconds
Northern Line
Southern Line
8.15±0.05
8.25 ±0.05 (est.)
437
428
0.3788
0.2832
16
CARNEGIE INSTITUTION
ECOOC P„ Model 01o6
True time terms
Computed time terms
normalized to Schuyler)
0 Stations
* bhots
'
East
5 E
V =6.50 km/sec
b*
Vn= 8. 20 km/sec
50 c
-• — o-
117 116 115 114 113 112 111 110 109 108 107 106 105 104 103 102 101 100 13579
I 100km I (Schuyler) 2468,°
li
12
Fig. 6. Structural model and synthetic time-term results showing structurally induced time-term
distortions for single-ended profiles.
an approximate algorithm to compen-
sate for structural distortion in cases
where a similar geometry is appli-
cable and calculated time terms are
available. Each of the three segments
a, b, and c of the time-term graph
can be associated with a segment of
the interface whose slope is deducible
from the slope of the corresponding
time-term segment. The depth of the
interface a' is easily determined,
since the associated time terms have
been subjected to no structural dis-
tortion. The depth of the b' segment
can be calculated by requiring that
the intersection of the time-term seg-
ments a and b correspond to the in-
tersection of the interface segments
a/ and b'. The final segment c' is po-
sitioned similarly under the criterion
that the intersection of b and c be
related by refraction paths to the in-
tersection of b' and c'. The details of
this procedure, though difficult to de-
scribe, are straightforward and well
defined, and in the model study they
yield a corrected structure that is in
satisfactory agreement with the orig-
inal travel-time model. From the cor-
rected structure so derived, it is now
possible to deduce true time terms
and determine the corrections to be
applied to the calculated values. Ob-
served departures from the idealized
piecewise linear interface can be in-
terpreted as local perturbations upon
the approximating model.
The true set of time terms in Fig.
6 represents an idealized sampling of
calculated ECOOE time terms over
an approximate east-west section in
the central part of the area. By the
DEPARTMENT OF TERRESTRIAL MAGNETISM
17
procedure described above, a set of
true ECOOE time terms can be found
from the calculated time terms. It is
evident that there are many possible
schemes for taking any one calculated
time term into the set of true time
terms; however, the unique peak in
the time terms can be translated in
only one way. For the particular set
of calculated ECOOE time terms, the
observed peak goes into the true peak
by means of a translation vector
whose components are a spatial offset
of 72 km shotward and a time-term
decrease of 0.46 second. All time
terms associated with the "root"
structure have been offset by this
amount with the exception of the
time terms near the eastern discon-
tinuity, which have been offset by 40
km but are otherwise unaltered. The
structure in the northwestern part
of the area was sufficiently different
from that implied by the initial model
that a second structure was modeled.
Here the translation vector has a spa-
tial offset of 65 km shotward and a
time-term decrease of 0.52 second.
The estimated time-term error due to
structure is less than ±0.2 second.
The modified time-term map from
which crustal thicknesses have been
derived is shown in Fig. 7.
Mean crustal velocities. Because of
the paucity of Pt data for the East
Coast experiment, it has not been
possible to determine mean crustal
velocities directly from the ECOOE
data by the time-term method (see
the section ''Late Arrivals and Am-
plitudes" for a discussion of East
Coast mean crustal velocities deter-
mined from wide-angle reflections) ;
however, a number of crustal veloci-
ties determined by other investigators
from profile analysis are shown in
Fig. 7. Some of these values are from
earlier Carnegie data.9' 10 For the re-
maining velocity values, we are in-
debted to members of the U.S. Geo-
logical Survey and to Brian Lewis of
the University of Wisconsin. From
these few mean crustal velocities, ve-
locity contours have been drawn on
the assumption that the lines of equal
velocity approximately parallel the
strike of the Appalachian structure.
The data are adequate for the south-
ern part of the area of study and do
not indicate striking lateral changes
in mean crustal velocity, save for the
overall decrease that occurs between
the crest of the Appalachian Moun-
tains and the coastal area. For the
northern and central parts of the
area, there are virtually no data, and
the assumption that the velocities are
constant parallel to the structure is
a major source of uncertainty.
A number of isolated mean crustal
velocities have been measured since
the computation of the crustal thick-
nesses presented in this report. In
the central and northern parts of the
area they indicate agreement with
our values, but there is some sugges-
tion that the crustal velocities of the
Piedmont, Coastal Plain, and conti-
nental shelf in the southern part of
the area are lower than those used in
the calculations of crustal thickness.
The crustal structure that has been
calculated from the modified time
terms is shown in Fig. 8. Crustal
thicknesses based on time terms are
contoured by solid lines. The dashed
contours are crustal thicknesses that
have been inferred from other seis-
mic work or are based upon time-
term values calculated at the edges of
the area studied. The dashed lines
are included only because gravity
studies discussed later in this report
require closed contours. The mean
crustal velocities used in the calcula-
tions are contoured in Fig. 7. Because
mean crustal velocities are not well
known over the area studied, it is
expected that significant deviations
from the contoured velocity values
may occur locally. Time-term varia-
tions caused by local deviations in
IS
CARNEGIE INSTITUTION
E
k.
<D
•*-
O
E
UJ
o
o
u
■D
0>
O
E
a
o
E
3_
=>
o
"c
o
U
N
DEPARTMENT OF TERRESTRIAL MAGNETISM
19
o.
E
o
u
>
c
a
o
"E
o
u
o
_r
a
Q
E
i_
o
o
E
*^
LU
o
o
u
"O
o
la
a
c
3
v.
u
oo
6)
20
CARNEGIE INSTITUTION
• ail CTUSta] velocity result in spuri-
ous changes in crust al thickness. For
areas, particularly along the
at, this effect may be important.
The structure contoured in Fig. 8
a number of Interesting features.
.1 thicknesses range from 30
in one area on the Coastal Plain
I Piedmont to about GO km at one
nt under the Appalachian Moun-
neral, crustal thicknesses
th the Piedmont are not rapidly
chai . An exception to this pat-
vident in the area around
>n, D.C. In this area, there
many data, and crustal thick-
are shown to change very
idly over the Piedmont and Coastal
Plains. The trend of comparatively
large crustal thicknesses along the
st appears firmly established.
Time terms are few on the western
k of the Appalachian Mountains,
: the contours drawn in that area
uncertain.
The most striking feature of the
map is the prominent "root" struc-
ture beneath the Appalachians. This
• tends to lie almost directly be-
neath the trend of highest elevation
in the Appalachian Mountains, and
appear.- to be broadest and deepest
under the mountains in the southern
1 of the region where the highest
elevations occur (about 5000 feet).
In the northern part of the area
Med, the thickest part of the crust
Is to lie slightly to the east of the
of highest elevation.
G >ity anomalies a nd seismic
crui structure. Over the years,
many attempts have been made to
ncile seismic results with gravity
M In general these attempts
have not been very successful, partly
■ it is difficult to make mean-
ingful direct comparisons between
mic profile models and gravity
anomalio .
By using gravity computations
based upon the crustal structure de-
rived from the time terms, we have
attempted to ascertain the range of
velocity-density relationships that
yield an acceptable correlation be-
tween gravity anomalies predicted
from seismic structure and the ob-
served anomalies. In the absence of
detailed information about the veloc-
ity structure of the crust, the choice
of a velocity-depth relationship is
somewhat arbitrary. There are, how-
ever, a number of constraints im-
posed upon the velocity-depth func-
tion: (1) It must yield the mean
crustal velocities from which depths
were calculated. (2) It must be com-
patible with the results from existing
profiles. (3) The velocity for the up-
per part of the crust (below the first
2-3 km) should be near 6.10 km/sec,
as this value has been found for much
of the East Coast area,9'10'15"17 al-
though there remains some discrep-
ancy between this value and the ve-
locity values determined by shallow
refraction work. Within the limits of
the constraints, we have assumed as
a model of velocity increase within
the crust V = 6.10 + 0.017Z (see
"Late Arrivals and Amplitudes"),
where Z — depth in kilometers.
A subcrustal velocity of 8.20 km/sec
is assumed over the entire map area.
Bouguer gravity values for various
velocity-density relationships have
been computed at 50-km grid inter-
vals over the map area. The gravity
values have been adjusted so that the
computed anomaly is equal to the ob-
served anomaly at Chesapeake Bay
(grid point 1100, 500 in Fig. 9A),
and this crustal section is taken to be
the reference section for the area.
Predicted gravity values in the region
of the Coastal Plains have been cor-
rected for the effect of the low-density
sedimentary rocks.
It remains only to determine the
velocity-density relationship that best
fits the East Coast area. Three
straight-line velocity-density relation-
DEPARTMENT OF TERRESTRIAL MAGNETISM 21
ships cited in the literature were crease may be systematic in the lower
used: Woollard's relationship, dp/dV part of the crust, or it may simply
— 0.233; Birch's constant mean represent a discontinuity in the mean
atomic weight relationship, d„/dV = atomic weight between the rocks of
0.316; and Thompson and Talwani's the lower crust and those of the up-
(1966) approximate gabbro-dunite per mantle. The second alternative is
line, dp/dV = 0.194. All the relation- suggested particularly by the fact
ships are of the form p = a + bV, that the line connecting values for
where p = density, V = compres- gabbroic rocks to those of dunites
sional velocity, and a and b are con- and peridotites cuts across lines of
stants. Bouguer gravity anomaly mean atomic weight and has a slope
values for these velocity-density re- dp/dV ~ 0.20. Thus, because the best
lations have been computed at 50-km fit between computed and observed
grid intervals over the area studied, gravity values is obtained from a
The results establish that any dp/dV of about 0.19 to 0.23, one con-
straight-line relationship with a value elusion might be that gabbroic rocks
of dp/dV greater than about 0.23 or make up the lower part of the crust
less than about 0.19 yields predicted and peridotite, the upper mantle. As
gravity anomalies that are respec- the slope dp/dV between gabbroic or
tively too large or too small relative less mafic rocks and eclogite is well
to the observed anomalies. The grav- above 0.23, it appears unlikely that
ity values obtained by using the the upper mantle consists of any sig-
velocity-density relationship derived nificant fraction of eclogite.
by Thompson and Talwani (1966) Discussion. From this study we
are contoured in Fig. 9 (A) over most may summarize as follows: (1) A
of the area of study; with the excep- difference of 0.1 km/sec in the meas-
tion of the edges and local anomalies, ured Pn velocity exists between the
the discrepancy between computed Southern and Northern Lines. (2) A
and observed gravity values is less three-dimensional model of crustal
than 20 mgal. A simplified form of thickness has been obtained which
part of the Bouguer gravity map of shows that crustal thicknesses vary
the United States is displayed in Fig. from about 30 km in some areas over
9(B). the Coastal Plains and continental
The fact that the Woollard curve shelf to nearly 60 km in a few locali-
has about as large a value of dp/dV ties along the crest of the Appalach-
as can be tolerated raises interesting ian Mountains. (3) Three-dimen-
questions and imposes some strict sional gravity computations based
limitations upon the range of possible upon the seismic structure yield grav-
materials that constitute the lower ity values that correlate well in size
part of the crust. On the basis of and shape to the observed regional
Birch's results, the low value of dp/ Bouguer anomaly for values of dp/dV
dV suggests that the mean atomic between 0.20 and 0.23. (4) The fact
weight must decrease in the region of that 0.23 is about as large a value of
the lower crust and upper mantle. dp/dV as can be accommodated by the
Recent work by Knopoff18 indicates data, strongly suggests that the mean
that the slope Birch derived for lines atomic weight of subcrustal rocks is
of constant mean atomic weight is lower than that of rocks in the lower
too low; if this proves to be correct, part of the crust. A mafic (possibly
the decrease in mean atomic weight amphibolite) lower crust with an up-
must be even greater than that im- per mantle composed of peridotite or
plied from Birch's curves. The de- dunite is one model consistent with
22
CARNEGIE INSTITUTION
DEPARTMENT OF TERRESTRIAL MAGNETISM
23
o
IN
0
=>
O
3
o
CO
L_
0)
M-
D
Q
^r
LU
N.
o
8f
LU G
^~o
-£ O
^£
<L> .
D
Jr> Q-
o
0
rx
So
u
"O
D C
o o
1—
M- >
o 5
0
0.°°
00
go
t\
£■ o
*> o
o fc
oSO
>- 00
o
o
0) •
3 3
00
05 ^
3 ^
o -Q
* -o
M- *
° a
c E
o o
•s u
o
k.
Csj
00
0) t/i
> <u
■o S
<D CO
E^
a <d
.§*e
co 3
O
o
■C
CO
cQ .T
. .«+-
o o
. a
U) D
LE £
24
CARNEGIE INSTITUTION
this conclusion: models in which ec-
logite is the rock type of the upper
mantle are not.
Whether the measured differences
in P., velocity for this area are the
result of genuine velocity inhomoge-
neity in the mantle or simply the con-
sequence oi different average ranges
of observations is a question difficult
to resolve. The data indicate, how-
ever, that, even if the apparent Pn
velocity increases with distance, the
increase must be very small. More-
over, there is some suggestion from
previous profile studies that real lat-
eral subcrustal velocity variations do
exist in the area between the South-
ern and the Northern Lines. Work of
Tatel et al.9 in the area of the Ches-
apeake Bay and the Patuxent River
indicated a Pn velocity near 8.1 km/
sec. whereas preliminary results from
the North Carolina experiment of
1962 suggested a Pn velocity near 8.3
km sec.:9 Thus the data appear to
favor lateral velocity inhomogeneity
in the upper mantle, the velocities
being near 8.15 km/sec for the
Northern Line and near 8.25 km/sec
for the Southern Line.
Wide-Angle Reflections and
Amplitudes from ECOOE
P. A7. S. O'Brien
The records were divided into two
groups, those from the Northern Line
of shots and those from the Southern
Line. For each group the times t of
the late arrivals were squared and
plotted against the square of the ob-
serving distance x.
For the Southern group the plotted
points fell on two parallel straight
lines whose intercepts on the t2 axis
were in the ratio of 4/1. This pro-
vides almost incontrovertible evi-
dence that the two sets of points
relate to a primary and a first-order
multiple from a single plane reflec-
tor. The estimated value of the mean
overburden velocity is 6.30 km/sec,
and this should be quite reliable un-
less there are extreme, systematic
variations in the near-surface veloci-
ties. The mean reflector depth was
31.8 km. Because the primary reflec-
tions were observed out to a range
beyond 400 km, it is possible to place
an upper limit to the mean velocity
gradient in the crust, for the ray-
leaving the shot point must have been
reflected at the point 200 km away
from the shot and 32 km beneath it.
If the velocity gradient had been
more than 0.01 sec-1, no ray leaving
the shot point would have reached
that reflection point.
The few records from the North-
ern Line examined showed mainly
multiple reflections — the observing
ranges were greater than for the
Southern Line — which gave a mean
overburden velocity of 6.50 km/sec
and a mean depth of 34 km. The few
primary reflections observed were
fully consistent with these values.
The measured amplitudes were
highly variable. However, it was quite
clear that Pn could not be a simple
head wave, as the recorded amplitude
was 30 times that calculated for a
range of 200 km and 300 times as
great as that calculated for the ex-
treme range of 700 km. The simplest
way to explain these large amplitudes
is to involve a velocity gradient in
the uppermost part of the mantle.
A gradient of 5 — 10 X 10^3 sec-1
would be sufficient.
The amplitude of Pn decreased by
a factor of 2 when the distance was
increased from 200 km to 700 km.
If this attenuation was due entirely
to absorption, the Q of the upper
mantle would be about 1100. If Q has
a lower value than this, the velocity
gradient beneath the Mohorovicic
discontinuity must increase with
depth.
The velocity gradients quoted are
inconsistent with those assumed in
the time-term analysis, but no at-
DEPARTMENT OF TERRESTRIAL MAGNETISM 25
tempt has yet been made to remove of refracted rays so that the time
the inconsistency. terms are independent of azimuth
and refer to a point on the refractor
Late Arrivals and Amplitudes almost directly beneath the station.
In the structure obtained by Smith
et al.21 the maximum variation of the
P. N. S. O'Brien
Selected records from the Lake time term with azimuth was about
Superior Experiment of 1963 and the 1.0 second (12.5 km) and the maxi-
East Coast Onshore-Offshore Experi- mum offset of the refractor point
ment of 1965 were reexamined, pri- from the station was about 80 km.
marily to identify reflections and to It therefore seemed desirable to re-
measure P-wave amplitudes. The analyze the first arrival times with-
identification of reflections is impor- out the assumption that each station
tant because analysis of their time- had a unique time term. A more im-
distance curves enables an estimate portant reason for the reanalysis was
to be made of the mean velocity over- that the presence of structure en-
lying the reflector. If the reflector is abled the mean overburden velocity
also a refractor giving first arrivals, to be estimated from the first arrival
it is possible to convert the first ar- times themselves. A direct check
rival time terms into depths without could therefore be obtained on the
having to use qualitative arguments very high velocity used in the con-
to place limits on the overburden ve- version of time terms to refractor
locity. Measurement of P-wave am- depths.
plitudes is important because ex- The method for obtaining over-
pected values may be calculated for burden velocity may be understood
the crustal models proposed from in- by reference to Fig. 10. The refrac-
terpretation of the time-distance tor structure is reflected in the travel
data; and if they are not consistent times at stations offset toward the
with those measured, the model is shot point by an amount h tan Qc,
obviously unsatisfactory. where h is the refractor depth and 6C
Neither experiment was designed is the critical angle. By matching up
for either of the above purposes; the time-distance graphs obtained
nevertheless, the results of this re- from shooting in opposite directions,
examination have been informative, an estimate may be made of 2h tan 6C,
Some work was also done to finish and this, together with the estimated
the analysis and interpretation of a delay time (h cos 0C/V, where V is the
few seismic records obtained in the mean overburden velocity) , enables h
Italian Alps.20 and V to be determined.
The Lake Superior Experiment of It was possible to match up three
1963. The continuing work on these separate structures on the upper re-
records falls into three main parts: fractor; they gave estimates of the
reanalysis of the first arrival time- mean overburden velocity of 5.13,
distance data; analysis of the late 5.44, and 5.50 km/sec. If it is as-
arrival time-distance data ; and anal- sumed that the overburden is not
ysis of the amplitudes. homogeneous but consists of a layer
Reanalysis of the first arrival of 4.0 km/sec (sediments) overlying
time-distance data. These data have one of 6.1 km/sec (volcanics), the
already been exhaustively analyzed mean vertical velocities are reduced
by the time-term method.21' 22 But this to 4.81, 5.03, and 5.28 km/sec. There
method assumes that the refractor was only one well-defined structure in
has a plane boundary within the cone the arrivals from the Mohorovicic
26
CARNEGIE INSTITUTION
■2h, tan 9
s
r* — 2 h, tan?c— H
Distance
Ground surface
Refractor
Fig. 10. Use of offset in obtaining mean overburden velocity.
discontinuity. Matching it up indi-
cated that the mean velocity beneath
the upper refractor was about 6.68
km sec.
The reinterpretation of the first
arrivals has not yet been completed
in detail, but the final structure
should not depart much from that
shown in Fig. 11. It should perhaps
be emphasized that the reinterpreta-
tion was based on exactly the same
time-distance information as the
original time-term analysis and that
any differences relate simply to the
added constraints of the present
method, namely, that delay times
should refer to positions on the re-
fractor and not to positions on the
ground surface, and that refractor
velocities should be obtained only
from reversed shooting across the
same portion of the refractor. The
second constraint accounts for the
fact that the upper-mantle velocity is
estimated to be 8.21 and not 8.07
km /sec as found by the method of
time terms.
Analysis of the late arrival time-
distance data. The records were so
dominated by reverberation in the
lake that it was not often possible to
pick a clear onset to any late arrival.
However, nearly all records taken at
distances less than about 450 km
showed one, and sometimes two, late
arrivals whose amplitude was usually
5 to 10 times that of the first few
cycles of the first arrival.
The only records from the west end
of the lake that were examined were
those from Victor, Yankee, X-ray,
and Tyro. All of them exhibited a
late arrival which was clearly a re-
flection from the Mohorovicic dis-
continuity and whose time-distance
curves indicate a mean crustal veloc-
ity of 6.75 km/sec. The records from
X-ray and Tyro also showed an ear-
lier late arrival which probably re-
fers to an intracrustal discontinuity.
Its time-distance curves indicate
mean velocities of about 6.0 km/sec
for X-ray and 6.3 km/sec for Tyro
and mean depths of about 12 km and
23 km, respectively. However, allow-
ance for the considerable thickness
DEPARTMENT OF TERRESTRIAL MAGNETISM
27
C
Range in km from Otter
o g -2
o .5 o
o
o
^1-
t — r
o
o
CN
8
t — r
if
o
35 5.28 km/sec 45 48 28 5.03 km/sec 2221 56 59 18 16 4.31 km/sec 10
10
20
E
f 30
O
Q
?0- —- -0?
6.71 km/sec
40
50
60
.A—
a
o
8.21 km/sec \
Southern line
Ray path L— *-R
Ray path R— »~L
Main line
Time-term analysis (Smith et a!., 1966)
Fig. 1 1 . Lake Superior crustal structure from first arrivals.
of sediments would reduce the esti-
mated depths and slightly increase
the estimated velocity of the material
immediately above the reflector.
So far, Woodsman and Otter Cove
are the only stations at the eastern
end of the lake whose records have
been examined. They show a reason-
ably late arrival which is most simply
interpreted as coming from a reflec-
tor whose depth increases from about
15 km near the Keweenaw Peninsula
to about 25 km at a position two
thirds of the way from the tip of the
Peninsula to Otter Cove. This inter-
pretation is still highly tentative and
needs to be checked against records
from other stations. The records also
show a very few later arrivals that
are consistent with an identification
as reflections from a depth of about
53 km.
Analysis of the amplitudes. The
amplitudes of the late arrivals were
broadly consistent with their inter-
pretation as reflections, but the first
arrivals from the Mohorovicic dis-
continuity were too large to be due
to head waves from a refractor with
a constant velocity. Velocity gra-
dients of about 5 X 10-3 sec-1 would
be sufficient to explain the observa-
tions.
Bancroft23 has presented P-wave
spectral amplitudes as a function of
range and has clearly shown that
maximum amplitudes were obtained
for shots fired at depths in the range
130-159 meters. O'Brien24 has used
the spectral amplitudes to estimate a
Q value of 830 ± 30 for the upper
crust and has shown that the opti-
mum shooting depth corresponds to
the depth at which the bubble pulse
"frequency' ' becomes equal to the res-
onant frequency of the water layer.
This depth, for Nitramon WW-EL
fired on an acoustically hard bottom,
is equal to 143 W0-18 meters, where W
is the charge weight in tons.
28 CARNEGIE INSTITUTION
EARTHQUAKE SEISMOLOGY determine the absorption of compres-
. a; ; 5. ,/.. /.. Fernandez, sional wave energy in the upper half
H. Marconi, A. Rodriguez, 0f the mantle. To this end, the fre-
°> amI F- quency spectrum at the Department,
Yolponi
at a distance of 50°, and that at
O for P W WES IN the Mantle Cuzc0> pf u> at a stance of 8°, were
compared.
I. S>. Sacks The amplitude ratio of two fre-
The large For avian earthquake of quencies A and f2 at a seismograph
r 17. 1966, has been studied to A is
mM .4(A) S(f2) R(f1) -.(/,-Aln
M, = _ • • exp
A(f,) S(f,) R(f2) QiVt
where S(f) is the source function, The frequency independent param-
R(/) is the receiver function includ- eters such as the geometrical effect,
ing seismograph and crustal transfer spherical spreading, and focusing,
functions, /\ is the distance from cancel out in this equation ; i.e., abso-
earthquake to seismograph, and Vi is lute amplitudes are not considered,
the effective velocity along the path. If we now take the ratio of M for
The ratio /• V, equals tlt the travel two stations
time to station 1.
S(A) R1(f1) /-tt(A-A)A
exp
3A S(A) RAf.) \ ft /_g1(/1)Jg,(/,)
M2 5(A) Ro(h) I -*(f*-f i)U \ Ri(f*)R*(fi)
— . — • exp
S(f-2) R2(f2)
/-7T (A-/x) A
exp [ ,
\ Qi ) \ Q2 / (1)
If identical seismographs are used at 14-second window was used. The
the two stations, the R(f) factors broadband tape channel was played
are mainly the crustal transfer func- through Kronheit band-pass filters,
tions. Fernandez has calculated and To get detail in areas where the spec-
measured transfer functions for var- trum is very steep, e.g., at 1.3 cps,
ions crusts. The amplitude of R is three filters were used in series. The
roughly proportional to the imped- mean amplitude in the time window
a nee contrast (p]V1/p2V2) at the was determined by means of a pla-
Moho discontinuity and is usually nimeter. The filters were considerably
about 0.8. Therefore, R,R2 probably broader than the equivalent filters
varies between 0.6 and 1.5 over the that would have been used with con-
frequency range. The above range of ventional Fourier analysis over the
R I: is based on a time window of same time window. The spectrum ob-
in finite extent. As the time window is tained is therefore somewhat
reduced, the amplitude of R de- smoothed, showing the gross struc-
creases. Kanamori28 has shown that ture but not the microstructure.
use of a relatively short time window, The two spectra are shown in Fig.
about 10 sec, reduces the amplitude 12. The energy arriving at Cuzco has
of R considerably. In the analysis of components above the ground noise
the Peruvian earthquake example a up to 11 cps, but the signal at the
DEPARTMENT OF TERRESTRIAL MAGNETISM
29
Q.
E
o
2.0
Frequency (cps)
Fig. 12. Spectra of the compressional wave arrivals from the Peruvian earthquake of Octo-
ber 17, 1966, recorded at DTM in Washington, D.C. (upper) and at Cuzco, Peru (lower). Only
the slopes of the curves are significant, not their positions. The absolute amplitude at Cuzco is in
fact considerably larger than at Washington.
Department dropped below the noise
at 3.5 cps, owing partly to the greater
attenuation of the longer path and
partly to local noise at this station.
The dominant feature in the DTM
spectrum is the steep decline beyond
1.1 cps and the flattening out beyond
1.5. There is some indication of a
similar feature but of smaller ampli-
tude on the Cuzco spectrum, which
suggests that source radiation might
account for some of the effect. Equa-
tion 1 cannot be solved unless some
estimate is made of Qlt the Q on the
path to Cuzco. The simplest model
would be a homogeneous model in
which Qi = Q2. In this case the
mantle Q = 1750. It is suggested
from the work of Sumner26 that the
Qp for the upper couple of hundred km
may be nearer 1000, in which case the
mantle Qp — 1500. These figures are
based on the mean slope of the spec-
tral data. Confidence limits on this
determination would be misleading at
this stage because the two seismo-
graphs and the earthquake were not
along the same great-circle path and
we have therefore made the assump-
tion that the ratio [S(/i)]/[S(/2)] is
independent of azimuth.
Shear Wave Absorption in the
Upper 130 km
sSKP-SKP. A technique is de-
scribed that enables the Qs structure
of a region to be determined by means
of the spectral data from a single
well-situated vertical component seis-
mograph. The factor by which the
shear- wave amplitude decreases at
frequency / is exp (-irft/Qs) where t
is the travel time to the observation
point and Qs is a dimensionless, fre-
quency-independent number. The ray
paths of the phases used are shown
in Fig. 13. The spectrum of the sSKP
wave train is divided by that of the
SKP arrival. The ray paths are very
similar for all the path below the
earthquake, and any path effects or
receiving station effects therefore
cancel out. The spectrum of sSKP is
modified by three effects that do not
affect the SKP spectrum:
1. Possible differences in the radi-
ation of the earthquake in the SKP
and sSKP directions. The directions
are nearly antipodal, and the radia-
tion is therefore expected to be simi-
lar.
2. Filtering effect of the earth's
30
CARNEGIE INSTITUTION
Fig. 13. Ray paths of sSKP and SKP. The two paths below the earthquake are very similar.
crust at the point of reflection. This
effect is small, provided that the fre-
quency window is adequate. The
transfer function (amplitude versus
frequency) of a single-layer crust,
considering the interference between
reflections and multiple reflections
from the Moho and the surface, has
spectral peaks that repeat every V/2h
cps where the V is the shear velocity
in the crust and h is the crustal
thickness. The effect, which can be
seen in Fig. 14 at the lower frequency
end of the spectrum, is to superim-
e an oscillation on the Q effect.
3. The effect of anelasticity, i.e.,
the greater absorption of high fre-
quency energy due to the finite Q of
the region between the earthquake
and the surface. The ratio of the
spectral amplitudes over a frequency
range A/ of the two phases is R =
exp (-irAft/Q) , where t is the time
the sSKP phase spends in the region
above the earthquake. If the fre-
quency range A/ is large enough,
the absorption will dominate the first
two effects. Plate 1 shows the verti-
cal-component short-period and long-
period records from the Port Moresby
standard seismograph for an earth-
quake at a depth of 130 km in
South America. The spectral ratio of
the SKP and sSKP phase is shown
in Fig. 14. It is necessary in any
analysis of this type to be sure that
all the frequency components plotted
are in fact above both ground noise
and digitizing noise. An equivalent
time section of the record just before
the sSKP arrival (which is the
smaller one) was therefore analyzed
to determine the background noise.
All the points appearing in Fig. 14
DEPARTMENT OF TERRESTRIAL MAGNETISM
9i
Qsfor upper 130 km
0.8 0.9 1.0 1.1 1.2
Frequency,
Fig. 14. The spectral ratio of the arrivals shown in Fig. 15. The frequency spectrum of sSKP
is divided by the frequency spectrum of SKP. Points are plotted only where the signal-to-noise
ratio at that frequency is greater than 2/1. The open circle at f = 0.45 cps has a S/N of 1.5.
The square X, and closed circles at f > 0.8 cps refer to the time window used — the square, 7.5
sec; the X, 1 0 sec; and the closed circle, 1 5 sec.
are those with a signal-to-noise ratio
greater than 2/1. The noise in this
case is mainly the coda of the SKP
arrival, as can be seen in the short
period (SP) record of Plate 1. A
rather poor point, S/N ^ 1.5/1, was
all that could be determined in the
frequency range 0.2 - 0.8 cps. The
time window used was 44 sec on the
LP record. Three time windows were
tried on the SP record — 7.5, 10, and
15 sec. The resultant spectra were
sensibly the same for the three win-
dows, suggesting that interference
effects from reflections were small.
The Q value satisfying the data is
300. The uncertainty in this value
may be seen from the lines represent-
ing different Q values drawn in Fig.
14. All points lie between 200 and
400. The value is considerably higher
than that reported by Anderson et
al.27 from surface waves, which was
110 for the same depth. It does, how-
ever, apply to a different region from
that considered in their report.
A Path for High-Frequency Shear
(S) Waves in the Upper Mantle
The Qs for the upper 600 km of
the earth was determined to be 160
from a study of shear waves reflected
and refracted at the core as reported
in Year Book 62. It was suggested
that not all of this upper region had
this low Q, but that a relatively thin
zone, possibly of the order of 100-
200 km thick, had a Q value well be-
low 100, while the remainder of the
region had a much higher Q. It
seemed plausible that the "earth-
quake free" region in South America
between depths of 350 and 500 km
might be associated with a region of
low rigidity, owing possibly to par-
tial melting, and that this region
would have a low Q for S waves.
A number of broadband, wide-range
DO
CARNEGIE INSTITUTION
seismographs were installed to study
the Q structure in greater detail.
As was reported in Year Book 65,
the spectrum of the direct S wave
recorded from a deep earthquake
showed severe attenuation of the
higher frequencies, as would be ex-
pected from a path with a Qs of 160.
Later in the 5 train, however, higher
frequencies do appear. Recordings of
a deep earthquake made with three
different frequency passbands are
shown in Plate 2. The compressional
arrival maintains its envelope at all
frequencies; i.e., the duration of the
arrival at the lower frequencies is
similar to that at the higher fre-
quencies, as would be expected from
a simple source and a simple path to
the seismograph. The low-frequency
shear arrival is also very simple and
of similar duration to the P. At the
high frequency, there is no trace of
this clear S wavelet at the expected
arrival time, but 15 seconds later
some high-frequency activity starts
and persists without significant am-
plitude variation long after the low-
frequency trace amplitude has be-
come very small. The low-frequency
S amplitude drops by more than a
factor of 3 after 10 seconds, whereas
the high frequency shows no diminu-
tion for well over a minute. The
following tentative conclusions are
drawn:
There is a path for high-frequency
& through an otherwise highly ab-
sorptive region. Oliver, Sykes, and
Isacks of the Lamont Observatory
recently reported (Trans. Am. Geo-
phys. Union, March 1967) such a
path from the deep earthquakes in
the Fiji region to their station NIU.
If such a path exists in South Amer-
ica it presumably reaches the higher
Q region at some distance from To-
conce. Assuming that it travels then
to the station at a velocity of 4-5
km/sec, the S (direct) — S (high-
frequency) time of 15 seconds is
equivalent to a distance of 60-75 km
from the station. The long duration
of the high-frequency late-arriving S,
however, is unexplained.
Converted Waves from the Inter-
face at a Depth of 410 km
A rather clear arrival has been
recognized between the P and S ar-
rivals from South American deep
earthquakes. It has been identified as
a shear-to-compressional wave con-
version at a depth of about 410 km.
Figure 15 shows the ray paths. The
Seismograph
,nferf°ce at + An
Earthquake depth > 400 km
Fig. 1 5. Ray paths of direct P, direct S, and converted S-to-P arrivals.
DEPARTMENT OF TERRESTRIAL MAGNETISM
33
identification was made as follows:
Waveform comparison. The direct
shear waves from a deep source have
the higher frequencies severely at-
tenuated by the low-Q region through
which they travel. In Year Book 62
we reported a mean Qs value of 160
for the upper 600 km. The waveforms
could be meaningfully compared only
at periods longer than 2 sec. Plate
2, a and &a is a radial component
seismogram which shows P, the new
phase, and S clearly. The peaks and
troughs of the new phase are marked
A through D in Plate 2, a and 2, aa
and the correlated peaks on the S
phase are similarly marked (2, a).
Plate 3 is another example of the
phase, again from a radial component
record. The S waveform has been
determined both from direct S and
ScS as recorded on the Cuzco broad-
band large dynamic range seismo-
graph, and found to be similar to the
converted phase wavelet shown in
Plate 3. The similarity to the S
wavelet suggests that the phase was
radiated as S. Since it arrives well
before the direct S, it must have
traveled a considerable portion of
its path as the higher-velocity P.
It arrives at the seismograph as P,
appearing only on the vertical and
radial components but not on the
transverse. The position of the inter-
face relative to the hypocenters was
determined by studying earthquakes
occurring at varying depths. The time
difference between P and the SP
phase decreases as the earthquake
depth decreases, indicating that the
interface is above the earthquakes
studied, which ranged between 526
and 649 km. Figure 16 shows theo-
retical travel-time differences be-
tween the SP phase and P for various
depths over the distance range for
which the phase exists. The travel
times have been constructed by means
of the Jeffreys-Bullen tables. It is
known that the difference from the
Jeffreys-Bullen travel times is quite
large for shallow earthquakes in this
general region, but the agreement is
surprisingly good for deep earth-
quakes. Figure 17 shows these resid-
uals for two deep earthquakes. The
earthquakes were located with tele-
seismic data, i.e., few of the stations
whose residuals are shown in the
figure were used to locate the earth-
quakes. Since the theoretical SP
Distance, degrees
Fig. 16. Theoretical travel-time differences between the direct compressional arrival and the
S-P converted wave. Curves are shown for three earthquake depths. The travel times for other
depths can be interpolated using the differences which are marked between the curves.
34
CARNEGIE INSTITUTION
+5
C2
— 1
E
•
•
•
• •
° o
1
May 13d 02h 23m
(19.3s, 63. 8w)
111.
ii i a d
Co $ °
H = 589, M =
5.1 '
5
o
-5
'©"To_
-T-O'
--CDOr
July 30d 02h 11"
(22.8s , 63. 7w)
H = 526,M = 4.5
10
15
20
Distance in degrees
Fig. 17. Travel-time residuals of two deep earthquakes used in the converted S-P wave study.
Residuals are shown from the Jeffreys-Bullen travel times. Dots are readings from high-elevation
seismographs which generally have high positive residuals. The earthquakes were located by
stations mainly at teleseismic distances. The residuals shown should therefore be independent of
the hypocenter location procedure.
travel times were calculated from the
Jeffreys-Bullen tables for depths
greater than 400 km only, they may
be moderately reliable. Table 2 shows
the residuals of the observed SP
phase from the theoretical. The
agreement is satisfactory.
Figure 18 shows the relative am-
plitude of this converted phase as a
function of angle incidence. Table 3
gives the angle of incidence for vari-
ous epicentral distances and depths.
The amplitude reaches a maximum
just before the phase disappears. In
the record at a distance of 12°, which
has an angle of incidence at the inter-
face of 34°, and therefore near max-
imum amplitude, the SP phase is
considerably larger than direct P. At
a distance of 14°, the phase has dis-
appeared. The angle of incidence is
greater than 34° and, as can be seen
from Fig. 21, the converted phase
does not exist.
As yet, there are not sufficient data
to determine the velocity contrast at
the interface from the amplitude be-
havior of the SP phase. Preliminary
indications are that it must be of the
same order as the Moho discontinu-
ity, i.e., a velocity contrast of the
order of 10%. Frequencies as high as
TABLE 2. Residuals from Theoretical SP-P Travel Times, Based on an
Interface Depth of 413 km
Observed
Travel
Residual,
Times,
Observed-
Depth,
Distance,
SP-P,
Calculated,
Station
Region
km
degrees
seconds
seconds
ANT
W. Brazil
649
12.85
42.5
1.4
ZON
Argentina
589
7.25
24.5
3.1
ZON
S. Bolivia
526
9.78
14.5
-1.0
TCC
S. Bolivia
526
4.17
16.0
-3.4
DEPARTMENT OF TERRESTRIAL MAGNETISM
35
10
15 20 25
Angle of incidence, degrees
30
35
40
Fig. 18. The relative amplitude of incident shear (S) to refracted compressional (P) waves
(after McCamy, Meyer, and Smith, 1962).28 The different curves are for different velocity con-
trasts. Curve 13 has a contrast of 17%; curve 15, of 11%; and curve 17, of 6%. Preliminary
data favor curve 15.
TABLE 3. Angle of Incidence at Interface for Different Depths and
Epicentral Distances*
Angle of Incidence,
degrees
Epicentral Distance, degrees
Depth 667
603
540 km
13.8
19.6
25.5
31.7
32.7
33.4
2.2
3.5
5.2
8.05
8.88
12.0
2.05
3.22
4.70
7.50
8.33
11.42
1.85
2.90
4.35
7.05
7.88
11.0
The depths are the same as used in the theoretical travel times in Fig. 16.
2 cps in the SP phase (see Plate 2, c)
suggest a very sharp interface with a
transition zone thickness of a few-
kilometers only.
A Search for Small, Deep
Earthquakes in the Andes
Shigeji Suyehiro
To identify small, deep earthquakes
in South America a quadripartite net
about 1 km in size was set up in
Toconce, Chile, and one in Cuzco,
Peru. During the period of observa-
tion the U.S. Coast and Geodetic Sur-
vey identified two fairly large, deep
earthquakes in northern Argentina
within about a month by teleseismic
observations, but neither foreshocks
nor aftershocks associated with these
earthquakes were observed by the
36
CARNEGIE INSTITUTION
quadripartite net. The lack of small,
deep earthquakes during the period
of observation may indicate that
there are comparatively few small,
deep earthquakes in comparison with
shallow earthquakes.
In the course of the quadripartite
observation in Toconce, systematic
azimuthal distortions of apparent ve-
locity and direction of propagation
were noted.
The apparent velocity is plotted
against the direction of propagation,
both having been reduced to the level
plane. Frequently, earthquake waves
traveling from south to north give
much lower velocities than those
traveling in the opposite direction.
The distortion cannot be attributed
to an actual distribution of earth-
quake hypocenters, such as deep ones
to the north and shallow ones to the
south, which would give the observed
result without assuming any local
anomaly. Nor can it be attributed to
the difference in weathering at quad-
ripartite points because all four seis-
mometers were installed on the same
solid bedrock. Actually the deep
earthquakes in northern Argentina,
which should give an apparent veloc-
ity of 12-14 km/sec, gave only 7-9
km/sec. Shallow earthquakes to the
north, which should give low veloci-
ties of the order of 8 km/sec, gave,
without exception, the much higher
apparent velocity of more than 20
km/sec. In Fig. 19, therefore, the
ratio of observed to calculated appar-
ent velocity is plotted for earthquakes
of more than 300 km in epicentral
distance located by the U.S. Coast
and Geodetic Survey or by G. Saa,
S. J., using the Carnegie local net. The
calculation is based on the Wadati-
Oki travel times, and nearer earth-
quakes of less than 300 km in epi-
central distance were not plotted, for
a small shift in epicenter will greatly
change the calculated apparent ve-
locity. Figure 19 indicates that the
observed distortion is caused by some
local structure beneath the array
rather than by any actual distribu-
tion of hypocenters. Figure 20 shows
the distortion observed in the direc-
tion of propagation.
The simplest way of explaining
such azimuthal distortion is to as-
sume an interface of 30° in dipping
angle where velocities in the surface
are 4.5 km/sec, and those of the
second layers, 6.5 km/sec.
This dipping interface of the order
of 30° must be local. To interpret the
crustal structure under a certain
area, a much larger array is required.
Small arrays must be used for study-
ing local seismicity rather than local
structure.
-o
>
o
O
o
o
4,
3,
2.
1.
° n
Q_ U'
£- 0.
o
_o
o
>
c
o
5
•
5
•
•
5
0
•
ft
►
•
8
•
..
7
*
••
•
•
••
■ a
5
-180 -140 -100 -60 -20 0 20 60
W N
Azimuth (degrees)
100 140 180
Fig. 19. Ratio of observed to calculated apparent velocity plotted against azimuth in To-
conce. Calculation was made from epicentral distance and depth, based on Wadati-Oki ve-
locity distribution, for reported earthquakes, epicentral distance of which is more than 300 km.
DEPARTMENT OF TERRESTRIAL MAGNETISM
37
flj
JO
.y
D
o
O
c
o
_•
o
o
o
o
■4—
t.
o
<i>
F
CO
-4—
o
D
n
1_
«-»-
o
O
CL
<J
.O.
O
Q
c
o
o
.— to
■o 5
_y
U
U
60°
40°
20°
0°
20°
40°
60°
•
«
/
/
1
— <■
•
\
*
••
N
1
/
/
• N
*
/
/
>--.
../
p
•
•
®
160° -120° -80°
W
■40° 0° 40c
N
Azimuth
80° 120° 160°
Fig. 20. Deflection of observed from calculated direction of propagation plotted against
calculated direction of propagation in Toconce for located earthquakes of A > 300 km by U.S.
Coast and Geodetic Survey or by Carnegie net.
On the Spatial Distribution of
Earthquakes Near San Juan,
Argentina
F. Volponi and H. Marconi
The problem of the spatial distri-
bution of seismic activity has been
discussed in numerous papers, many
of which had another purpose of
study. The following report examines
some experimental data on this sub-
ject obtained in a restricted region
on the eastern flank of the Andean
cordillera that divides Chile and Ar-
gentina. This region lies between 31°
and 32° south latitude and 68° and
69°30' west longitude (Fig. 21). Two
large earthquakes have occurred here
— one, of magnitude 7.8 on the Rich-
ter scale, on January 15, 1944, and
another, of magnitude 7, on June 10,
1952. The present activity of the area
may be judged by the fact of the
occurrence of 12 perceptible events
in the first 6 months of 1966.
The equipment for the study con-
sisted of single, vertical, Wilson-
Lamison type seismometers with elec-
tronic amplification and visible re-
cording. Amplification was 105-106 at
10 cps. Timing was obtained from
crystal chronometers, and second
marks were registered on the trace,
making it possible to measure abso-
lute time with a precision of 0.1 sec.
The stations were located (as indi-
cated in Fig. 25) at Zonda, Pie de
Palo, Hualilan and Leoncito. The
principal station at Zonda also con-
tained many other seismometers of
different constants and amplifica-
tions.
Also, as early as 1958 an excessive
number of seismic events compared
with the total number recorded had
been observed with values of S — P
of about 12 sec. As seen in Fig. 22 (A)
the frequency distribution of S — P
values for a period of 18 months
shows a large excess in the interval
10 to 14 sec. These data were ob-
tained on a vertical Benioff-type seis-
mograph with T0 — 0.8 sec and with
38
CARNEGIE INSTITUTION
70° 65° 60°
Fig. 21. Location of zones studied.
a magnification of 28,000 for signals
at 1 Hz. The magnitude range of the
recorded events was from 2.5 to 4.5.
The curve in Fig. 22(B) was ob-
tained from a long-period (T0 = 50
sec) horizontal seismograph with a
magnification of 35,000 for signals
of period of 1 Hz and 70,000 for 2-Hz
signals. It is seen that a slightly
larger number of events (166) in the
magnitude range from 2 to 4 were
recorded than in the earlier study.
Finally, the curve in Fig. 22(C)
gives the distribution resulting from
the consideration of 537 events of
magnitude between 1 and 3 recorded
on the Carnegie vertical seismograph
with T„ = 1 sec and magnification
280,000 at 5 Hz.
The three curves are very similar.
The figure shows a small concentra-
tion of activity in the range 5 to 20
km (S — P = 1 to 4 sec) followed
by an almost total absence of events
before beginning the more important
concentration for S — P values be-
tween 10 and 13 sec. Not only does
this maximum occur at 12 sec, but
also the percentage of events with
this S — P time is nearly identical
in all three distributions.
The events occurring in the shorter
S — P interval were well recorded at
Zonda; a number were also recorded
at Pie de Palo, but only a few were
registered at Hualilan or Leoncito.
The seisms of the second group (S —
P = 12 sec) , in contrast, were well
recorded at all four stations with
only a slight variation in intensity.
The gain at each station was ad-
justed according to the noise level
encountered, but the sights were spe-
cially chosen to permit them to oper-
ate at the highest possible gain.
Two properties of the events with
S — P = 12 sec were studied:
their spatial distribution, and evalu-
ation of the constants m and K, in
the Ishimoto-Iida relationship
nA*
K
(2)
DEPARTMENT OF TERRESTRIAL MAGNETISM
39
c
=>
u-
0)
30 -
20
10
c
I 30
D
O
20
10
30
20
July 1959 to Feb. 1961
119 Events
M = 2.5 - 4.5
T0 = 0.8 second
Vertical
1
i i I i i i r
i i i i
n
B
ii i i
Jan. to June 1966
166 Events
M = 2 -4
T0 = 50 seconds
Horizontal
Lri
Jr"^-T
I I I I I III
I I I I
I I I l
I III
Jan. to June 1966
537 Events
M= 1 -3
T0 = 1 second
Vertical
15
S - P (seconds)
Fig. 22. Frequency distribution of S-P values at Zonda station.
where n is the frequency of seismic
events and A their maximum ampli-
tude. Figure 23 shows a plot of fre-
quency versus amplitude for ampli-
tudes from 1 to 25 mm at Zonda sta-
tion. If the events are grouped in
amplitude at 5-mm intervals and
plotted as log N versus log amplitude,
Fig. 24 results. The linear fit of the
figure implies the validity of the
Ishimoto-Iida relationship. From
equation 2
log N + mlogA = log If (3)
where K and m are constants ; m is
the slope of the line in Fig. 24 and
has the value 1.6, giving a value of
Kx = 3.6 X 10-4.
Of the 340 seisms with S — P
times in the range of 10-13 sec,
about 40 had data that permitted the
precise location of their hypocenters.
Examples of these data are shown in
Plate 4. The precision of the hypo-
central location is believed to be with-
in 2 km in any direction. The plot
of the determinations is shown in
Fig. 25. Their locations fall in a nar-
40
15-
E 10-
a
Z
5-
I l i I I — I I I I l l I l
CARNEGIE INSTITUTION
1 I I I I I I I I I I I "T
^in
J_J I 1 L_L
Mill I i i i
25
30
5 10 15 20
Amplitude (mm)
Fig. 23. Plot of frequency versus amplitude at Zonda station (amplitude 1—25 mm]
4 6 8 10
Log of the amplitude
Fig. 24. Verification of formula n = KA^71 and emission of the constants m and K.
DEPARTMENT OF TERRESTRIAL MAGNETISM
41
31°00'
O
31°30'
Hualilan
• Zonda
Piede Palo
20 40 Km
I —i
Leoncito
69°
West longitude
68c
£
West
Surface
East
20
40
60
•
•
—
80
"
•
•
•
•
•
•
• • •
• • •
—
—
100
•
• •• •
• •
•
• •
• 9
Fig. 25. Distribution of hypocenters from seismograms shown in Plate 4.
row band about 30 km in the north- In the vertical direction they are all
south direction and cover the full 100 located roughly between 70 and 100
km of the net in the east-west direc- km, and it is presumed that this ac-
tion. The band probably extends far- tive volume is part of the upper
ther, both to the east and to the west, mantle.
42
CARNEGIE INSTITUTION
In summary, a volume of seismic
activity has been studied near San
Juan, Argentina, which has the fol-
lowing* properties:
1. The trace amplitude, frequency
of occurrence relationship can be
characterized by the Ishimoto-Iida
equation
seisms mm1-6
nA1G = 34 X 10-
km3 year
2. The volume studied is 30 X 30
X 100 km3 and lies with its axis nor-
mal to the Andean cordillera between
Chile and Argentina.
3. About 30 ^c of all the local
earthquakes recorded in this area
100 x 100 X 100 km3 lie in this
active volume.
4. Xo surface expression can be
correlated with this activity in the
upper mantle.
ISOTOPE GEOLOGY
S. R. Hart, T. E. Krogh, G. L. Davis,
L. T. Aid rich, F. Munizaga, and
A. M. Stueber
Potassium, Rubidium, and
Strontium in Ultramafic Rocks
and Minerals
A. M. Stueber
Ultramafic rocks are of importance
in the earth sciences because they are
commonly considered to be the domi-
nant material of the earth's mantle.
Furthermore, ultramafic rocks are
generally regarded as having some
sort of genetic relationship with ba-
salts.
The determination of the potas-
sium, rubidium, and strontium con-
tent as well as the strontium isotopic
composition of ultramafic rocks and
minerals from a small area of west-
ern North Carolina has been under-
taken. Prior work of this nature on
ultramafic rocks has been primarily
in the form of survey studies which
have made significant contributions
but at the same time have posed sev-
eral equally significant and as yet
unanswered questions. It has been
shown, for example, that the Sr87/
Sr86 of alpine ultramafic rocks of
wide geographic distribution are
rather variable and definitely higher
than the SrS7/Sr86 of basalts, but the
possible variability of SrS7/Sr86 in a
single alpine ultramafic intrusion has
not been investigated. Similarly, it
has been shown that the strontium
concentrations of ultramafic rocks
are generally low, but at the same
time variable over as much as an
order of magnitude. However, the
distribution of the Sr, as well as the
K and Rb, within the several mineral
phases of ultramafic rocks has not
as yet been established. Finally, it ap-
pears that there is no general pro-
portionality between Sr87/Sr86 and
Rb/Sr of alpine ultramafic rocks. The
possibility that such a proportionality
does exist within a single intrusion,
and that an Rb-Sr isochron age can
therefore be determined, remains to
be tested.
A detailed and systematic sampling
program was carried out on five
closely related alpine ultramafic in-
trusions in western North Carolina:
Balsam Gap, Dark Ridge, Addie-
Webster, Buck Creek, and Day Book.
Samples were selected on the basis
of variable mineral composition and
variable degree of serpentinization.
More than 100 grams of each of these
samples were ground with care to
avoid contamination and were inti-
mately homogenized in an attempt to
improve reproducibility of K, Rb, and
Sr analyses, and mineral separations
were carried out on a few selected
samples. The results of the analyses
which have been completed thus far
are presented in Table 4.
It is clear that the concentrations
of K, Rb, and Sr are quite low in
olivine; enstatite tends to contain
more K and Rb, but not significantly
more Sr, than olivine; diopside will
DEPARTMENT OF TERRESTRIAL MAGNETISM
43
TABLE 4. Potassium, Rubidium, Strontium, and the Isotopic Composition of Strontium
mafic Rocks and Minerals from Western North Carolina
Ultra-
K,
Rb,
Sr,
ppm
ppm
ppm
K/Rb
Rb/Sr
Sr37/Sr36
Webster pyroxenite
16.0
0.115
5.34
139
0.022
16.1
0.122
5.27
132
0.023
0.71 05
Webster enstatite
8.27
0.124
0.348
67
0.356
...
7.95
0.130
0.213
61
0.610
. . .
Webster diopside
16.2
0.150
9.08
108
0.017
13.2
0.125
9.05
106
0.014
0.7090
Addie peridotite
12.2
0.048
2.06
253
0.023
0.7109
8.81
0.032
2.13
273
0.015
. . .
Addie olivine
2.94
0.026
0.376
113
0.069
...
3.17
0.019
0.391
167
0.049
. . .
Addie pyroxenes
34.0
0.077
7.18
441
0.011
(enstatite + diopside)
34.4
0.083
6.83
414
0.012
0.7087
Dark Ridge peridotite
138
0.223
1.17
619
0.191
• • •
146
0.236
1.09
618
0.217
• • •
145
0.222
1.05
652
0.211
0.7241
Dark Ridge olivine
13.3
0.094
0.190
141
0.495
. . .
12.7
0.085
0.174
150
0.491
0.7238
Dark Ridge enstatite
57.4
0.262
0.430
219
0.609
61.3
0.257
0.494
238
0.520
0.7243
Dark Ridge diopside
743
1.04
5.01
714
0.207
0.7224
705
0.999
5.03
706
0.198
0.7233
* Estimated uncertainty in Sr87/Sr86 after normalizing to Sr86/Sr88 = 0.1194 is ±0.001.
accept relatively large amounts of K,
Rb, and Sr. Indeed, diopside gener-
ally has an order of magnitude more
Sr than either enstatite or olivine.
This is reasonable in view of the
tendency for Sr to substitute for Ca,
which is present in significant
amounts only in the diopside.
In some cases the reproducibility
of K and Rb concentrations is quite
good, whereas in others there seems
to be room for improvement. Because
K and Rb vary sympathetically (as
indicated by the rather good repro-
ducibility of K/Rb over a wide range
of values), and because differences
between duplicate analyses are sim-
ilar on a percentage basis at both
high and low concentration levels, it
is thought that differences in dupli-
cate analyses for K and Rb, at least
in mineral separates, are the result
of inhomogeneous distribution of K
and Rb within the minerals. How-
ever, differences in duplicate analyses
for whole rocks may be the result of
inhomogeneous distribution of min-
eral phases within the sample, in
spite of the attempts to eliminate this
factor.
From these analyses of mineral sep-
arates it seems reasonable to conclude
that the concentrations of K, Rb, and
particularly Sr, in an ultramafic rock
are very critically controlled by the
amount of diopside present in the
rock.
The K/Rb of these rocks and min-
eral separates are plotted as a func-
tion of K concentration in Fig. 26.
Stueber and Murthy29 have shown
that the K/Rb in ultramafic rocks
generally vary between 200 and 500
and seem to be unrelated to K con-
tent. Such a conclusion would apply
in a general way to the data pre-
44
CARNEGIE INSTITUTION
5;:
700
600
500h
400
300
200
100
! 1
Dark ridge peridotite
Addie peridotite
1 1
Di
■-
V- ebster pyroxenite
-
En* D!
-
_ t
1
-i ^"
^
I.
1 >^
-
-/dT0£
--
En | |
IV \
1
100
200
K, ppm
300 700
Fig. 26. Potassium-rubidium ratio as a func-
tion of potassium content of minerals from
ultramafic rocks from western North Carolina.
sented here. However, if each ultra-
mafic rock is considered as a separate
system it can be seen that the K/Rb
invariably increase with increasing K
content. Such behavior is opposite to
the trend observed in igneous rocks
in general, and is similar to that re-
ported by Hart30 for the ultramafic
samples from St. Paul's rocks. More-
over, for each ultramafic rock re-
ported here, the K/Rb increases from
olivine to enstatite to diopside, which
is the normal trend of differentiation.
Ordinarily Rb, because it has a larger
ionic radius than K, is enriched in
magmatic liquids relative to K as dif-
ferentiation proceeds. The apparently
opposite behavior for the ultramafic
minerals may represent an equilib-
rium distribution for K and Rb. On
the other hand, it may be due to the
effects of partial fusion, since alpine
ultramafic rocks can be regarded as
residual. Partial fusion of a rock or
mineral results in a decreased K/Rb
in the liquid and an increased K/Rb
in the residue. Selective partial fu-
sion of diopside or enstatite to pro-
duce basalt might result in a residual
ultramafic rock in which the K/Rb
are higher in the diopside and the
enstatite than in the olivine.
The SrS7/Sr86 determined thus far
fall in the same general range as
those previously reported for ultra-
mafic rocks. There is no indication of
strict proportionality between Sr87/
SrS6 and Rb/Sr. In the four analyses
from Webster and Addie, which have
very similar and low Rb/Sr, there is
a suggestion of proportionality, and
this is now being investigated. It is
significant that the Dark Ridge sam-
ples, with distinctly higher Rb/Sr,
have distinctly higher Sr87/Sr86 than
the Addie-Webster samples. The sam-
ples from Dark Ridge, however, show
a very close similarity in Sr87/Sr86 in
spite of significant differences in
Rb/Sr. The results obtained thus far
are too few to permit final conclu-
sion, but they are tantalizing, and
further work is in progress.
Geological History of the
Grenville Province
T. E. Krogh and G. L. Davis
Age measurements this year pro-
vide the basis for a major revision
of the geological time sequences in the
Grenville province. The real age of
the rocks in this area is uncertain be-
cause age determinations on minerals
yield values of about 900 :± 100 m.y.
(Stockwell, 1964), 31 whereas certain
other geological arguments suggest
that the rocks are much older. In
earlier work we have determined age
values between 1500 and 2000 m.y. for
rocks from this area using the whole-
rock rubidium-strontium method.
Now a major proportion of our effort
is directed toward evaluating the
effect of the 900-m.y. event on the
whole-rock system. Certain innova-
tions in sampling techniques have
made it possible to draw definite con-
clusions about the volume of the rock
that remained a closed system with
respect to migrations of dating ele-
DEPARTMENT OF TERRESTRIAL MAGNETISM
45
ments during the 900-m.y. event.
Most of the rocks in the region are
paragneisses and have definite char-
acteristics indicative of having under-
gone deep burial, intense heating, and
plastic deformation. It has been
commonly assumed that these char-
acteristics were established during
the 900-m.y. event. Through applica-
tion of our sampling techniques we
have been able to show that some of
these characteristics were established
during an earlier metamorphism
about 1500-1800 m.y. ago.
Our investigation has been two-
fold. In certain cases we have studied
smaller and smaller details in order
to evaluate the dating technique. In
others it has been necessary to meas-
ure widely spaced samples to learn
about regional geological patterns.
In detailed studies of two granite
bodies a good fit of points to isochron
plots for 1500 and 1700 m.y. was
accomplished. This establishes the oc-
currence of two discrete events and
eliminates the possibility that these
ages were the result of a younger,
say 900-m.y., event acting on rocks
with a much older true age. Our
studies of widely spaced samples com-
plement the more detailed ones, as
we have found evidence that two
granites in other parts of the region
were emplaced about 1500 m.y. ago
and that another major granite body
formed about 1700 m.y. ago. Simi-
larly, studies of paragneiss samples
collected from a single outcrop yield
age values for the most part between
1500 and 1800 m.y., as do samples
collected over much of the northwest
Grenville area, and over the eastern
Grenville area along the St. Lawrence
River.
The French River area as an ex-
ample of the application of the
method. Measurements on whole-rock
samples from a single granite body
in the French River area of Ontario
yield a whole-rock isochron age of
1725 m.y. (Fig. 27). Circled points
shown on the diagram, representing
a special rock type, will be discussed
later. The relatively small standard
deviation of 16 m.y. in the slope of
the line through this series of plotted
points has considerable significance
both in the evaluation of the assump-
tions of the whole-rock method and
in a geological sense. An explanation
of the significance and of the mean-
ing of the mineral data also shown
in Fig. 27 will require a digression
into certain fundamental concepts of
Rb-Sr dating. In this figure unnum-
bered points represent data for
whole-rock samples. Sample Ont 4-0,
whole rock (WR), is joined with a
broken line to the data point for
apatite in this rock. The data point
for biotite would lie far to the right
of the diagram if it were shown. In
the whole-rock dating method, rep-
resentative samples split from 5 to 50
kg of rock are analyzed for total Rb
and Sr as well as for strontium iso-
topic composition. The age is deter-
mined by comparing the relatively
small amount of radiogenic Sr87 gen-
erated in the rock with the amount
of radioactive parental Rb87 present
in the rock. Because we have no way
of directly measuring the amount of
Sr87 present when the rock was
formed, we assume that at that time
the abundance of Sr87 relative to one
of the stable Sr isotopes (Sr86 by con-
vention) was the same in all the sam-
ples. The assumption is verified if the
analytical data, expressed on a plot
of Sr87/Sr86 versus Rb/Sr, lie on a
straight line.
The geological significance of a
high degree of fit to a straight line
lies in the implication that all samples
did in fact contain isotopically iden-
tical Sr at one time. The isotopic
composition of the initial strontium is
shown at the intercept of the line
with the vertical axis. Any argument
suggesting that a particular granite
46
CARNEGIE INSTITUTION
• I
0.790 -
e~s:
0.770
0.760
co 0.750
CO
in
0.740
0.730
0.720
0.710
0.700
/
ONT 4-0, WR ■
APPARENT ISOCHRON 1725 ± I6m.y. /V _
70-3
SOW! 4-0 APATITE
170-1
02
0.4 0.6
Rb/Sr
0.8
Fig. 27. Isochron plot for the French River, Ontario, granite. Solid line is the whole-rock iso-
chron. Broken line is the mineral isochron.
is older than the indicated isochron
age must include some proposed
mechanism by which the isotopic
composition of Sr can be made iden-
tical in all parts of the sampled body.
This requirement is most severe when
vast volumes of granite act as a
buffer to external chemical and phys-
ical effects. Thus, if a major granite
mass can be shown to have a high
Rb-Sr ratio over an extensive region,
this entire region must be purged of
radiogenic strontium in order to elim-
inate traces of an earlier history. In
contrast to major volumes of whole
rock, minerals need only to exchange
their strontium over a distance of
millimeters in order to lose all specific
trace of a past history.
Data points for apatite, whole rock,
and biotite (Fig. 27) lie approxi-
mately on a straight line, and, if an-
alyzed, other minerals should lie on
or near the same line because apatite
DEPARTMENT OF TERRESTRIAL MAGNETISM 47
and biotite are probably the least and ical interactions occurred at the time
the most reactive minerals in the of last mineral equilibration or dur-
rock. This relationship establishes ing an earlier event,
that the minerals and whole rock all In the French River granite, nar-
achieved an identical isotopic compo- row (5-10 cm) amphibolite dikes
sition at a time indicated by the slope occur at one outcrop area. The vari-
of the line through the data points, ation of Rb and Sr in a suite of
The Sr87/Sr86 of the strontium com- samples cut from a section taken per-
mon to all minerals was 0.738 (Fig. pendicular to a dike shows that the
27), and the time of establishment of Rb content drops to half and the Sr
this isotopic composition was about content doubles along the edge of the
1000 m.y. ago. The slight difference dike (0-2 cm) over that present in
between the apatite whole-rock age granite 7-15 cm from the dike. Iso-
value and the biotite whole-rock age topic data for these samples (Fig. 27,
value probably indicates that biotite circled points) demonstrate that corn-
continued to lose radiogenic stron- plete isotopic mixing did not take
tium after the apatite had ceased to place between the samples at the time
gain radiogenic strontium. Any min- of late mineral recrystallization
eral bias introduced in sample split- (1000 m.y.). These Rb and Sr vari-
ting will move the data points up or ations developed either during an
down this mineral isochron and will earlier metamorphic event or during
thus introduce scatter into the whole- the intrusion of the dike,
rock data. French River paragneiss. Chemi-
The occurrence of mineral iso- cally layered rocks that are thought
chrons demonstrates intergranular to be of sedimentary origin but now
migration of strontium but tells are intensely recrystallized and de-
nothing about the distances involved formed into paragneisses are com-
during the migrations. There is no mon in the French River area. One
requirement, for example, that the outcrop exposed approximately 2
isotopic composition of strontium was miles south of the granite was suit-
ever identical in all grains of a single ably layered for studying isotopic
mineral in one rock. migration. Specifically, we hoped to
Evaluation of the size of the closed determine the time at which the
system. To evaluate the distances in- gneiss formed from its sedimentary
volved in isotope migration we have precursor. Block samples represent-
obtained samples in which the min- ing a continuous section of approxi-
erals, rather than being in a three- mately 18 inches of rock perpendicu-
dimensional mosaic, are concentrated lar to the layering were obtained,
into discrete layers. Monomineralic Within each layer of granitic gneiss
layers are not easily obtained; how- the Rb concentration drops by about
ever, only layers with finite differ- a factor of 3, whereas the Sr content
ences in their Rb/Sr are required, increases by about 50%, as the edge
We have often found that chemical of each biotite-amphibolite layer is
interactions have occurred between approached. The abrupt drop in Rb
rocks with different chemical compo- and increase in Sr correspond to the
sitions to form layers with interme- occurrence of a plagioclase-rich mi-
diate or contrasting characteristics, crocline-free layer adjacent to each
Insofar as the layers can be shown amphibolite layer,
to be reaction zones formed during Further studies of these variations
intense heating of the region, we are are in progress, but at this point it
able to determine whether the chem- is reasonable to conclude that we are
48
CARNEGIE INSTITUTION
examining the products of major
chemical interactions that occurred
at the time of formation of this
gneiss. With this degree of chemical
interaction it is a reasonable assump-
tion that the isotopic composition of
Sr would be the same in all parts of
the section at the time of the re-
actions. If the reactions occurred dur-
ing the 1000-m.y. event, data points
for all parts of the gneiss would lie on
a line with a 1000-m.y. slope. If, how-
ever, the gneiss formed during an
earlier event, further modification of
the isotopic relationships during the
1000-m.y. event are to be expected.
As shown in Fig. 28, samples 1A
and 2B, each approximately 1 cm
wide, are located at an abrupt Rb/Sr
discontinuity; in Fig. 29 data points
MI'ii'iI.'i'i'i'Mi'IiIhMii
Fig. 28. French River, Ontario, paragneiss. Diagram showing changes in concentration of
rubidium and strontium along a section normal to the layering. The vertically dashed layers
are granitic gneiss; the crosshatched layers are biotite-amphibolite.
DEPARTMENT OF TERRESTRIAL MAGNETISM
49
0.760
0.750 -
0.740
CO
CO
to
CO
c/5
0.730 -
APPARENT IS0CHR0N
1622 m.y. ± 70 m.y.
0.720
0.710
0.700
COMPOSITE ^2AB
31
0.2
0.4 0.6
Rb/Sr
0.8
Fig. 29. Isochron plot for the French River paragneiss.
for these two samples define a line
with a slope comparable to that for
minerals in the rock. The other
samples of gneiss do not lie on a
single line with this slope but rather
lie approximately along the isochron
transferred from Fig. 27 for the
French River granite. An arithmetic
composite sample (1A, 2AB, SI) also
appears to lie on this line, suggesting
that the enclosed volume may have
been closed to migrations of the dat-
ing elements during the 1000-m.y.
event. We conclude that the gneiss
formed during an earlier metamor-
phism and did not form during the
late heating event demonstrated by
the apatite in sample 66-88. If we
propose limited isotopic mixing dur-
ing a time when the major phases
remained stable we should expect low
points on the Rb/Sr profile of Fig. 28
to gain, and high points to lose, ra-
diogenic Sr87. Thus samples 2AB and
1C do lie below samples with compa-
rable Rb/Sr in Fig. 29. Similarly, the
50 CARNEGIE INSTITUTION
samples occupying- the valleys in the (Year Book 65, pp. 380-383). Whole-
profile of Fig*. 28 do show an increase rock data for three locations from
in radiogenic Sr over that which the each of the two dominant granite
higher samples would have had 1725 types present are shown in Fig. 30.
m.y. ago. Also, a mineral isochron Samples from three separate out-
could be passed through samples 3Z> crops of the felsic granite (Ont-1,
and 3D as if they had locally ob- 65-165, and 65-162-0) define the same
tained an identical isotopic composi- line as the three samples from differ-
tion approximately 1000 m.y. ago. ent parts of the biotite-hornblende
A least-squares line through all the granite, 65-199, 201, and 202. Again,
granitic gneiss samples yields an age the minerals have undergone isotopic
value of 1622 ± 70 m.y.; another exchange long after the time of
isochron between three samples of formation of the rock. As noted on
gneiss (not plotted on Fig. 29) from the diagram, sample 65-165 is a corn-
other parts of the outcrop defines a posite; it was made in the field by
line for 1430 ± 26 m.y. The three taking samples approximately 5x5
high points on the isochron yield age X 2 cm at 1-meter centers over ap-
values between 1750 and 1900 m.y., proximately 5 square meters of
assuming an initial SrS7/Sr86 of 0.704. freshly blasted road cut. Results for
Clearly, the metasediments must be a 5-kg sample from this outcrop an-
older than 1725 m.y., the age of the alyzed in duplicate last year lie near
granite that presumably intrudes but not on the line defined by the
these gneisses. The data demonstrate other whole rocks. The implication of
that a paragneiss subjected to re- this test of sampling technique is that
gional metamorphism after the time an average of many possibly open
of its formation may yield a variety subsystems collected over a large
of age values. Our best estimate of area of homogeneous material may
the time of formation of this gneiss is approximate a closed system better
obtained by considering the samples than a single block sample,
farthest removed from variations in Another composite, 162-0-9, shown
the Rb/Sr profile shown in Fig. 29 in Fig. 30, was made in the labora-
and by eliminating from considera- tory by combining hand-specimen-
tion such samples as 3/ and 1A, sized samples collected at 2-meter in-
which are susceptible to additions of tervals across a freshly blasted road
radiogenic strontium. Sample 66-88 cut. Results for this composite, and
represents about 30 kg of rock from the 10-kg whole-rock sample from the
the thick layer 87-3; hence this sam- same outcrop, are essentially identi-
ple and sample 3Z) probably define cal. A contact between the felsic
our best approximate age at about granite (65-162) and a biotite-horn-
1750-1900 m.y. for an assumed ini- blende granite occurs on this outcrop,
tial ratio of 0.704. The spread in ages A study of the Rb and Sr concentra-
possible for a single outcrop is in- tions across the contact indicates a
structive in interpreting single whole- gradual drop in Rb from 170 to 95
rock paragneiss age values reported ppm and an increase in Sr from 140
later in this section. to 220 ppm within a distance of about
Lake Muskoka granite and other 10 cm. Such a gradual variation
coeval granites. Additional work has could occur for a magmatic contact
been completed on a granite body that or, alternatively, might have devel-
occurs on the eastern shore of Lake oped during the metamorphism of
Muskoka, approximately 150 km the granites. The rocks themselves
southeast of the Grenville front show the effects of having recrystal-
DEPARTMENT OF TERRESTRIAL MAGNETISM
51
0.820
0.800
0.780
CD
CO
CO
co 0.760
C/5
0.740
0.720
0.700
1 1 1 1 1 1 — r
i — i — r
COMPOSITE 162
65-162-0 WR
APPARENT ISOCHRON 1497 ± 10 m.y.
0-9/
^
&
&
<(0
65-165 COMPOSITE
CONTACT (E)
35-199
■^ APATITE
(65-162-0
/
■ ONT 1-2
" ONT I
/
.©
/
i 65-202
CONTACT (A)
65-201
J I L
J I L
J I
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Rb/Sr
Fig. 30. Isochron plot for a granite at Lake Muskoka, Ontario.
lized under directed stress with the
development of lenticular pods, a few
millimeters thick, of quartz and feld-
spar or of mafic minerals. Here the
foliation is parallel to the contact.
Data for two samples on the estab-
lished Rb/Sr gradient are shown in
Fig. 30. They were sawed from a
block containing the contact, and
were 10 cm apart. Because their
points do not define a line whose slope
is parallel to that of the mineral iso-
chron the rock was not isotopically
mixed during the last recrystalliza-
tion (700 to 800 m.y. ago). They do
lie within the error of our measure-
ments, on the line denned by widely
spaced samples from other parts of
the body. Therefore, the Rb/Sr gra-
dient of which the samples are a part
must have been established approxi-
mately 1500 m.y. ago.
It is a reasonable assumption that
the reerystallization exhibited by
these rocks would be accompanied by
isotopic mixing of strontium, over a
scale of at least a few centimeters.
If so, then the metamorphism of the
granites occurred approximately 1500
m.y. ago.
In two other localities ages within
1% or 2% of those obtained in this
study have been found. Two whole
rocks from a single large outcrop
52 CARNEGIE INSTITUTION
near Round Lake approximately 35 provide the most comprehensive pic-
km southwest of Pembroke, Ontario, ture of local and regional variation
yield two-point isochrons of 1510 yet obtained for a continental area.
m.y. A single sample from an outcrop
south of Sandridge, Ontario, yields Technique and Results
a whole-rock age of 1475 m.y., when
an initial ratio 0.703 is used. In addition to the determinations
Pa ragncissc >,s. Age values for para- on the 15-km grid, to be discussed
gneisses and granitic rocks in On- later, a number of special observa-
tario and Quebec can be calculated tions were made to assess the accu-
from single whole-rock analysis if we racy and precision of the method
assume an initial SrS7/SrS6. For ex- more objectively than had previously
ample, for the French River para- been possible. The repeatability of the
gneiss, age values between 1500 and measurements was tested by making
1900 m.y. can be calculated. Five out two successive drops at a location off
of 6 paragneiss samples collected in the Keweenaw Peninsula; seven days
a region approximately 300 X 150 km later approximately the same site
lying east of Georgian Bay yield age was reoccupied and a third value ob-
values in that range. Three samples tained. The first two observations
of paragneiss from the area between produced identical heat flow values of
the Saguenay River and Sept Isles, 1.11 hfu (1 hfu = 1 microcalorie of
Quebec, on the southeast region of flux per square centimeter per see-
the Grenville, yield age values be- ond), and the third yielded 1.06,
tween 1550 and 1650 m.y. Three indicating with little doubt that the
granitic rocks with age values be- method gives results replicable with-
tween 1000 and 1300 m.y. occur in in 5%.
this area, as well as three others with It has been suggested that a rea-
age values between 1500 and 1600 sonable evaluation of the absolute
m.y. These values indicate that pre- accuracy of the deep-lake technique
Grenville rocks occur in the eastern could be obtained by direct compari-
as well as the northwestern parts of son with results from conventional
the Grenville Province, and that the borehole determinations. To this end,
age patterns in the two areas may a profile of 5 observations was con-
be similar. structed. This profile extended into
the lake from a point near the bore-
HEAT FLOW hole gites of Birch (1954)32 and Roy
S. R. Hart, J. S. Steinhart, and T. J. Smith (1963) 33 on the Keweenaw Peninsula.
The program to use modified As may be seen from Fig. 31, all the
oceanic techniques for the measure- values lie in the range from 1.06 to
ment of terrestrial heat flow in deep 1.17 hfu and, taking into account the
lakes was continued during the year, observed local variation, are readily
and roughly half of the proposed re- compatible with the Birch and Roy
gional survey of Lake Superior was values of 0.93 and 0.95 hfu. There is
completed. More than 90 measure- some indication that the observation
ments of the bottom temperature gra- nearest the shore (5 km) is being
dient were made from the U.S. Coast disturbed by lake edge effects. The
Guard Cutter Woodrush with the only other land observation in the
Department's 6-meter thermistor area is in the Porcupine Mountains,
probe, largely in the western and cen- where the reported heat flow of 1.07
tral part of the lake. The heat flow hfu (Roy, 1963) agrees very well
values calculated for 83 of these sites with the nearest lake result of 1.02
DEPARTMENT OF TERRESTRIAL MAGNETISM
53
D
U
0)
• D
O >
-o
M
3®
a
a
a
CO
o
O
c
1—
a;
</»
O
a
>
o
■
o
.n
•u
<D
w
O
o
U
CO
54
CARNEGIE INSTITUTION
hfu obtained about 20 km to the
north of the Roy station.
Note has been made (Year Book
6.4) of the necessity for correcting
the observed temperatures to account
for the seasonal variation in the bot-
tom lake water. A daily bottom tem-
perature recorder located between the
Bayfield and Keweenaw Peninsulas
in 640 feet of water was operated
from October 1965 to October 1966 to
obtain the Lake Superior annual tem-
perature cycle displayed in Fig. 32.
Perturbations to the sediment tem-
peratures caused by this external flux
have been calculated both for a ho-
mogeneous half-space model and for
a model consisting of a single layer
overlying a half -space. (For details
of the latter calculation, see Lachen-
bruch, 1959.34) As one might expect,
corrections to the observed tempera-
tures derived from the single-layer
model come closer to linearizing the
gradients than those derived from
the simpler half-space model. For this
reason the heat flow values in this
report have been computed from the
bottommost portions of the gradients
corrected under the single-layer as-
sumptions. Conductivities for both
the model and the heat flow calcula-
tions are based on measurements
made on the individual cores taken
with each observation. It is hoped
that a model, now under study, which
allows the continuous variation of
conductivity with depth, will permit
a still more realistic correction for the
annual cycle to be applied, thereby
making available for use those values
of the gradients at shallower depths.
The validity of the annual correc-
tion does not, however, depend solely
upon an appropriate choice of a
model relating thermal conductivity
to depth in the mud. Figure 33 dis-
plays temperature measurements for
four of the cores. Also shown are the
measurements after correction for
the annual temperature cycle. The
single layer over a half-space model
is used. The four cores cover a range
of water depths, a range of time dur-
ing the experiment, and include one
of the most successful corrections and
one of the least successful. It seems
there is a depth effect on the annual
temperature cycle of the bottom water
such that measurements in water
shallower than that in which the an-
nual cycle was measured received
Time connecting observations
Smoothed cycle
Nov
1965
Jan Feb
1966
Mar Apr May Jun July Aug
Fig. 32. Daily variation of bottom water temperature in Lake Superior. AAeasured offshore
of Porcupine Mountains in water depth of 640 feet.
DEPARTMENT OF TERRESTRIAL MAGNETISM
55
1.0
2.0
f 3.0
E
:£ 4.0
Q.
d>
Q
5.0
6.0
7.0
• Observed
Corrected for
annual change
3.4 3.5 3.6 3.7 3.8 3.5 3.6 3.7 3.8 3.5 3.6 3.7 3.6 3.7 3.8 3.9
Temperature, °C
Fig. 33. Corrected and observed temperatures at four sites in Lake Superior.
insufficient correction and those from
very deep water were overcorrected.
It is not immediately clear whether
only the amplitude of the bottom an-
nual cycle changes with water depth
or whether the phase changes as well.
Review of a large number of the
measurements suggests that there is
no simple, precise answer to this
question, but a first-order correction
for depth may be possible.
The thermal conductivity and wa-
ter content of each of the cores were
measured at selected points with
spacing varying from 100 to 10 cm
to provide reliable interpolation to
any point of interest in the sediment
column. Conductivities were deter-
mined with a needle probe by a
method similar to that of von Herzen
and Maxwell (1959) 35; water content
was analyzed by conventional wet
and dry weighing. The resulting plot
of conductivity versus water content
based on about 140 pairs of data
bears a qualitative resemblance to the
graph presented by Ratcliffe (I960)36
for ocean sediments, but there is a
measurable absolute difference be-
tween our derived relationship and
that of Ratcliffe. Moreover, the scat-
ter of the data about a best-fitting
curve is several times as great as
would be expected from estimates of
the largest uncertainties in the meas-
urements. The absolute discrepancies
can be explained by differences in the
sediments used in the two studies,
but the significant scatter of the data
carries the more disturbing sugges-
tion that any average relationship
between conductivity and water con-
tent may, for a specific sample, be in
error by as much as 4% or 5%.
Additional inquiries into this problem
are planned for the coming field
season.
As a note of interest it has been
observed that the bottom muds of
Lake Superior lose their moisture at
a truly astonishing rate. A sample
from core 10 (Fig. 34) with water
content of about 50 % lost almost 1 %
of its initial weight each 20 minutes
when placed in an open dish at room
temperature. The clear implication is
that care must be given wet cores
when their water content is a matter
of quantitative importance.
Discussion. The range of published
continental heat flow values is from
0.7 to about 2 hfu; rarely, in areas
56
CARNEGIE INSTITUTION
rc:
15.4
°> 15.3
15.2-
0
S
'0>
CD
* 15.1-
g-15.0
D
T4# 9- in open dish
*k
X %
xx
Core no. 10-2.5
Air drying at 25 CC
x-
14.8
0 10 20 30 40 50 60 70 80
Time, min
Fig. 34. Weight loss due to water loss as
a function of time for a sample of Lake Supe-
rior mud. Note very high loss rate.
of recent tectonic activity, flux of 2.5
hfu has been observed. The small
number of published values for the
continents has made it difficult to
determine how the range of conti-
nent-wide variation compares with
variation in a small geographic area.
In the oceans variations in heat flow
of a factor of 2 or more have been
observed in traverses of only a few
hundreds of kilometers, and there is
a suggestion of changes of less dra-
matic amplitude in a profile of conti-
nental deep-hole measurements made
in New England by the Harvard
group. From the 1966 measurements
(Fig. 31) we are able for the first
time to get a clear idea of the extent
of regional variation of heat flow in
a continental area. It has sometimes
been stated in the past that heat flow
in the stable shield areas, of which
Lake Superior province is a part, ex-
hibits remarkably uniform heat flow,
and for the few measurements pre-
viously published the variation was
not large. We now see that there is
a variation of more than a factor of
2 even within Lake Superior. Because
of the unusual crustal structure
found in the Lake Superior region
(Year Book 6U\ see also The Earth
beneath the Continents, Steinhart,
Smith, eds., 1966), this is a place
where variation would be likely. The
significant finding is that for the first
time on the continents large and sys-
tematic heat flow variations have
been exhibited on a regional scale.
In the range of observed values
from about 0.5 to 1.2 hfu, or a little
higher, the most interesting are the
low ones, since the lowest values pre-
viously reported anywhere on the
continents were about 0.7 hfu. The
values on the high end of the range
encompass all the values usually
found on the stable continental shield
areas, although those between 1.2 and
1.3 are a little higher than the typical
shield average, which is about 0.8 or
0.9. The low values occur along the
western end of the lake in the same
region that one very low value was
found and thought to be anomalous
in the preliminary measurements of
1963 (Year Book 63). It may be
noted that, if average radioactive iso-
tope concentrations are taken for
many of the typical crystalline rocks
of the Canadian Shield, and their
heat flow generation is considered, a
layer 15-20 km thick would be suffi-
cient to produce a measured heat flow
of 1 hfu.
It is known, however, both from
the seismic velocity structure and
from other considerations, that ra-
dioactive elements are concentrated
upward within the crust, so that,
although the bulk of the heat flow is
the result of radioactivity within the
crust, as much as 20% of the heat
may be generated below the Mohoro-
vicic discontinuity. The values near
0.5 could then be explained by simply
requiring that the normal shield
rocks (granites, gneisses, etc.) form
only a thin layer in this area, if they
are present at all, and that rocks
containing about one half the usual
amount of radioactivity comprise the
principal thickness of the crust. The
commonly occurring rocks of low
radioactive concentrations are the
DEPARTMENT OF TERRESTRIAL MAGNETISM 57
mafic rocks, such as basalt and gab- be rather uniform over the area, we
bro, and they could be expected also might expect that heat flow could be
to exhibit high seismic velocities. In directly correlated with the thickness
the area of the low values such high of the crust. Comparing the heat flow
seismic velocities were found in the map (Fig. 31) with the seismic pro-
1963 seismic experiment (Year Book file (Year Book 6U) , it is clear that
6A) . On the shore to the west of this in a general way there is a correla-
area the Duluth gabbro and Kewee- tion. The thickest parts of the crust
nawan basalts are the principal sur- off the Keweenaw Peninsula produce
face rocks and extend to considerable, the highest values, and the compara-
though unknown, depths. Thus, it tively thin crust at the western end
appears that there is no great diffi- of the lake produces the lowest
culty in providing a mechanism that values. In detail, however, the agree-
accounts for low values, although it ment does not appear as convincing,
is harder to explain the extent of the and the more detailed comparison
local variation in heat flow. Materials will, of necessity, await the additional
with very high seismic velocities are data to be obtained in 1967. Perhaps
found underlying the whole region the revisions in the model of the
of Lake Superior at comparatively crustal structure in the Lake Supe-
shallow depths. If the concentrations rior region reported this year wil1
of radioactivity are really so low, provide for a better fit to the heat
then the variation in values must be flow anomalies. This possibility will
explained by lateral structural be examined along with other geo-
changes that are not altogether clear physical comparisons when the heat
in the seismic results. If we assume flow of the eastern part of the lake
the concentrations of radioactivity to has been measured.
ASTROPHYSICS
L. Brown, W. K. Ford, Jr., C. Petitjean, Vera C. Rubin, W. Trachslin,
K. C. Turner, M. A. Tuve, and C. M. Varsavsky
The astrophysics program at the foil excitation. The results of these
Department is a combination of in- diverse investigations are reported
OPTICAL ASTRONOMY
teracting interests and activities, below.
Basic problems of galactic dynamics
are being studied with observations
from the 21-cm hydrogen line re- W. K. Ford, Jr., and Vera C.Rubin
ceiver and from the image tube spec- Optical astronomy at the Depart-
trograph. Thus, although the tech- ment during the report year has
niques involved are quite different, centered about the use of the DTM
many of the same underlying ques- image tube spectrograph. Early in
tions provide motivation for both the report year a semisolid //2.25
investigations. On the other hand, the schmidt spectrograph camera, de-
DTM Van de Graaff generator has signed by Dr. I. S. Bowen and fabri-
been used for extensive and detailed cated by Davidson Optronics, was de-
experiments on scattering of polar- livered. This camera has produced
ized protons in order that the nature spectra of substantially better quality
of the nuclear force may be better than cameras used earlier, with no
understood, as well as for exploratory significant loss in speed. The spectro-
work in atomic spectra produced by graph was used by Ford and Rubin
58
CARNEGIE INSTITUTION
on about 20 nights, at the 84-inch
Kitt Peak telescope, the 72-inch Per-
kins telescope of the Ohio State and
Ohio Wesleyan Universities at Lowell
Observatory, the 31-inch Fan Moun-
tain telescope of the University of
Virginia, and the 36-inch Belts ville,
Maryland, telescope of the Goddard
Space Flight Center, NASA. Obser-
vations were started on three contin-
uing programs: (1) radial velocities
and line intensities of emission re-
gions in the Andromeda galaxy; (2)
radial velocities of galaxies and com-
pact blue objects in the Virgo cluster
of galaxies; and (3) studies of veloc-
ity fields in selected galaxies of inter-
est. In addition, spectra of suspected
M dwarfs in the Pleiades cluster were
obtained in collaboration with Rev.
M. F. McCarthy, S.J., a visitor from
the Vatican Observatory. Most of the
results discussed below are based on
very few observations and are there-
fore preliminary.
M31. To study the detailed dynam-
ics of a galaxy, it is necessary to have
accurate radial velocities from re-
gions all across the galaxy. Recent
21 -cm hydrogen line observations by
Burke, Turner, and Tuve (Year Book
63, p. 322) have stimulated renewed
interest in comparing optical and ra-
dial velocities in M31. As a start on
a program to obtain velocities of
higher accuracy than previously
available, four emission regions in
the south-preceding end of M31
(Baade's numbers 4, 7, 8, and 22)
were observed. For each region one
plate was obtained with the region
held fixed on the slit. A velocity was
obtained from several measures along
H«, and the probable error listed be-
low measures the internal agreement
on a single plate. The observed ve-
locities of —461 ± 13, —538 ± 6,
—450 ± 78, and —588 ± 8 km/sec,
correspond to velocities of — 314,
—291, —200, and —295 km/sec in
the plane of that galaxy, in good
agreement with the rotational veloci-
ties determined from the 21-cm
neutral hydrogen observations by
Burke, Turner, and Tuve (1964). 37
Observations of several globular
clusters indicate that the absence of
any sharp features in the red spectral
region makes it impossible to obtain
accurate radial velocities from plates
centered in the red. An attempt will
be made in future work, therefore,
to include observations of globular
clusters in the region containing the
H and K lines of Ca II.
Virgo cluster. Although the Virgo
cluster of galaxies contains several
thousand galaxies whose velocities
could be obtained, fewer than 50 ve-
locities are known. Hence all conclu-
sions about velocity dispersion, mass,
and equipartition of energy are based
on a biased sample, because to date
only the brightest members have been
examined. The speed offered by the
image tube system can be used effec-
tively on telescopes of moderate size
to substantially increase the size of
the sample. Radial velocities of Virgo
cluster galaxies may thus be deter-
mined. We also wish to sample the
spectra of compact blue objects in the
cluster region to determine whether
quasi-stellar objects occur with high
spatial density in regions of high
galaxy density.
Of 10 galaxies observed, 5 have
strong emission lines in their spectra,
and velocities have been determined:
NGC 4298, V = 1110 km/sec; NGC
4388, V = 2652 km/sec; NGC 4639,
V = 962 km/sec; NGC 4654, V =
1048 km/sec; and IC 3453, V =
2555 km/sec. The other galaxies have
only weak absorption features, and
all will be reobserved. From the list of
compact blue objects in the Virgo
cluster region (Rubin, Moore, and
Bertiau)38 no. 197 and TON 97 have
been chosen for observation. Both
objects have spectra without emission
lines, and are probably stars in our
DEPARTMENT OF TERRESTRIAL MAGNETISM
59
galaxy. Last year we reported obser-
vations of nos. 56 and 138. Only no.
56 has emission lines and is probably
extragalactic, even though its meas-
ured velocity is V — —225 km/sec.
Studies of selected galaxies. NGC
3389: A supernova was discovered
in NGC 3389, an S + galaxy in Leo,
on February 28, 1967, by A. D.
Chuadze at the Abastumani Astro-
physical Observatory, U.S.S.R. We
obtained two spectra of the super-
nova on March 7 and March 8. The
supernova is of Type I, with broad
unidentified emission features resem-
bling features in supernova studied
earlier by Minkowski. No hydrogen
emission is observed. A third spec-
trum, with both the supernova and
the brightest regions of the galaxy
on the slit, reveals Ha emission with-
in the galaxy. For this position angle,
the velocity variation with the galaxy
has been measured.
NGC 4038: The peculiar galaxy
NGC 4038 has been observed on 11
plates. For each of these, the slit of
the spectrograph was oriented along
a particular axis of interest. Spectra
from three position angles are shown
in the frontispiece. At the right of
each the position of the slit is indi-
cated. The most notable features are
the wide variation in the intensity
ratio H/V(0 III) 5007 A in different
regions of the galaxy, and the small
velocity variations across the galaxy,
amounting to only a few hundred
kilometers per second, thus eliminat-
ing any possibility that NGC 4038 is
an exploding galaxy.
Within 0.7° of NGC 4038 on the
sky is a somewhat similar peculiar
galaxy, NGC 4027. Velocities meas-
ured on a single spectrum are ap-
proximately the same as those in
NGC 4038. A small wisp of material
near NGC 4027, but not connected to
it, is also moving with a velocity sim-
ilar to that of NGC 4038 and NGC
4027.
NGC 3227: NGC 3227 is a Seyfert
galaxy with an elliptical companion,
NGC 3226. We have obtained a num-
ber of spectra of both objects and
have used a wide range of exposure
times.
Short exposures show that the
broad nuclear emission is due to
discrete clouds with velocities which
differ by several hundred kilometers
per second from the central velocity.
On longer exposures narrow emission
lines extend 500 pc outside the
nucleus. Bordering the nucleus, emis-
sion clouds are seen to the blue of
the narrow emission. Emission re-
gions far from the nucleus (3500 pc) ,
with relative velocities of 150 km/sec,
will enable an accurate mass to be
determined for NGC 3227.
RADIO ASTRONOMY
K. C. Turner, C. M. Varsavsky, and
M. A. Tuve
Northern Hemisphere
K. C. Turner
Considerable emphasis was placed
on equipment development during the
report year. Receiver improvements
have been devised which keep the
gain of our hydrogen spectrometer
from changing more than a few per
cent over several hours. The receiver
"front end" exhibiting this stability
is awaiting shipment to the joint
CIW-IARA observatory at Peyreyra.
This new instrument, tested for sev-
eral weeks on our Derwood telescope,
shows a system noise temperature of
about 250°K.
A new digital recording system
offering improved flexibility and ac-
curacy has been installed at the Der-
wood observatory, as well as six
extremely narrow-band filters, only
0.5 km/sec wide.
The surface of the 30-meter tele-
scope at the Avery Road station is
now complete, and interferometric
observations are planned for the com-
ing year.
60
CARNEGIE INSTITUTION
-20 -10 0 +10 +20
Radial velocity in the local standard of rest (km/sec)
Fig. 35. Velocity-galactic latitude plot at / = 0.
Northern hemisphere observations
have been made in the region north
of the galactic center (I = 0, 5 < b
< 50) to investigate the nature of
the ''central bulge" of the galaxy.
This region is also visible at the
CIW-IARA Observatory, and so will
permit measurements to be made to
a common scale in both hemispheres.
Figure 35 shows velocity-latitude
contours for I — 0. It can be seen
that the principal feature is very
narrow. A measurement with the new
narrow-band channels shows that the
peak has a full width at half maxi-
mum of only 5.5 km/sec. This feature
is so large (about 30° in I) and yet
has so low a velocity dispersion as to
suggest that it is quite nearby. (Its
angular size corresponds to a dimen-
sion of about 8.6 kpc if the feature
is located above the galactic center.)
This object extends from about I =
356° to at least I = 12° at b = 30°,
but careful study of the profiles re-
veals the existence of at least two
main peaks about 4 km/sec apart.
Many more observations will be re-
quired to find out what is going on
in this region.
Some preliminary observations
have been made following the edge
DEPARTMENT OF TERRESTRIAL MAGNETISM
61
of the moon as it passes through the
galactic plane. With this occultation
technique, 5 minutes of arc H I clouds
of 100° brightness should be detect-
able at the 5<r level with a 60-foot
dish and a 250 °K receiver.
Southern Hemisphere
K. C. Turner and C. M. Varsavsky
Although work in the southern
hemisphere has been mainly observa-
tional, the second dish has continued
to progress steadily, and should be
usable for observations in the coming
year.
After a month's work on the crys-
tal mixer receiver in June, a noise
temperature of about 900°K was ob-
tained, and observations were begun
on the southern rotation curve, em-
phasizing the improved accuracy ob-
tainable with our narrow-bandwidth
filters (2 km/sec wide). These obser-
vations generally confirmed the ear-
lier Australian work, although finer
detail was observed with our greater
resolution.
The most exciting work of the year
has certainly been that of F. Strauss,
a student of Varsavsky. The purpose
of his research program is to search
for an intergalactic bridge of neutral
hydrogen between our galaxy and the
Magellanic Clouds. So far he has ob-
tained 80 line profiles taken over an
area of about 250 square degrees.
Near the galactic plane the profiles
show 3 to 5 separate peaks that cor-
respond to different clouds or spiral
arms. The peak with the highest ve-
locity corresponds to the outer spiral
arm of our galaxy. This arm has
been traced without interruptions
through the whole range of galactic
longitudes observed, i.e. from I =
225° to I = 285°.
Figure 36 is a plot of the velocity
of this arm, with respect to the local
standard of rest, versus galactic lon-
gitude. Figure 37 shows the position
Fig. 36. Velocity of outer arm versus galactic
longitude.
of this spiral arm in the galactic co-
ordinate system I, b. In both figures
the approximate position and velocity
of the center of the Small Magellanic
Cloud are shown. Figure 38 is a pro-
jection on the galactic plane of both
the spiral arm and the Small Magel-
lanic Cloud. To calculate the distances
of the observed points to the galactic
center, Schmidt's model was used
with the parameters recommended by
the International Astronomical Un-
ion. For distance to the Small Magel-
lanic Cloud we adopted the value of
46 kpc used by the Australians. It
can be seen that the three plots show
a remarkable tendency of the spiral
Fig. 37. Position of outer arm in galactic
coordinates.
62
CARNEGIE INSTITUTION
CO
1
/ = 2-0"
^
Sun
o
o_
o
x Galacric
center
o
• SMC
II
Fig. 38. Projection of outer arm and Smal
Magellanic Cloud on the galactic plane.
arm toward the Small Magellanic
Cloud, both in velocity and in spatial
position. The brightness temperature
of the peak varies from about 80°K
at / = 225" to 5°K at I =285°. The
search will be continued when the in-
stallation of the parametric amplifier
is completed. It is not yet possible to
know whether this spiral arm reaches
the Small Magellanic Cloud.
Other observations at the IARA
include a search, by W. G. Poppel,
for intergalactic hydrogen in the con-
stellation of Microscopium, where a
region of reduced incidence of distant
galaxies has been observed, and an
investigation by S. Garzoli of the
region near the galactic plane from
298° < I < 312°. This region con-
tains the coal sack, a well-known dark
cloud. Both these investigations are
in the preliminary stages and are
representative of the richness of op-
portunity offered by the southern sky
and of the enthusiasm of our south-
ern colleagues.
NUCLEAR PHYSICS
L. Drown, W. Trdchslin, and C. Petitjean
We have continued the study of
the elastic scattering of polarized
protons, using the University of
Basel's polarized ion source; but the
emphasis of the study was changed
from target nuclei of spin zero to
those of spin one, specifically deuter-
ons. The completed analysis of our
measurements for spin-zero nuclei
r-C and 1(;0 has yielded a unique set
of accurate phase shifts for each re-
action. Dr. R. G. Seyler of Ohio State
University has derived the formulas
describing proton-deuteron scatter-
ing, and we have programmed and
checked them with the DTM com-
puter. Analysis will take place this
summer at Ohio State University. To
clear up an ambiguity in the struc-
ture of 41K (significant because of its
role in the theory of isobaric ana-
logue resonances) , we have measured
polarization effects in 40Ar(p,p)40Ar.
This was undertaken as an interest-
ing sidelight at the request of Pro-
fessor D. Robson of Florida State
University.
Elastic Scattering of Polarized
Protons on Nuclei of Spin Zero
The characteristics of the elastic
scattering of protons from nuclei of
spin zero were explained in Year
Book 65 (p. 73), and the results of
proton-alpha scattering were dis-
cussed in some detail. During the
year we analyzed our polarization
measurements of carbon and oxygen
in the same way in order to obtain
more accurate phase shifts. Taking
note of the complications of the
phase-shift analysis of proton-deu-
teron scattering, we made a careful
study of ambiguities in the analysis
of p-12C and p-1G0. By ambiguity we
mean that more than one set of phase
shifts can reproduce the experimental
results. In these two examples there
is none, providing that angular distri-
butions of both cross section and po-
larization are used, but if only one
kind of data is used there is consider-
able ambiguity. Specifically, 9 sets of
phase shifts were found that repro-
duce well the cross-section data for
oxygen alone, and three of them fit
better than the set that fits both
DEPARTMENT OF TERRESTRIAL MAGNETISM
63
cross-section and polarization meas-
urements. The analysis of our polar-
ization data alone is also ambiguous
but has fewer sets of phase shifts
that satisfy the experimental results
than the analysis of cross-section
data alone.
A contour map of polarization in
p-12C scattering (Fig. 39) is illustra-
tive of the nature of the measure-
ments. This map, which is the same
kind as used last year for proton-
alpha scattering (Year Book 65, Fig.
36), shows the polarization expected
for any given scattering angle and
energy. Two resonances, which cause
very rapid changes in polarization as
a function of energy, are not shown,
as the map is intended primarily for
use by experimenters in analyzing
polarized protons, and regions of
rapid variation are not useful for
that purpose.
Professor Robson's suggested ex-
periment was (as mentioned) the
measurement of polarization as a
function of energy for a fixed scatter-
ing angle in 40Ar(p,p)40Ar. The re-
sults, shown in Fig. 40, established
the spin of a closely spaced group of
resonances in 41K, which appear as
a single state when observed with the
low energy resolution of our equip-
ment. The high resolution studies at
Duke University,39 which show the
structure of this group of resonances
in great detail, succeeded in deter-
mining the orbital angular momen-
tum to be I = 1, but the total angu-
lar momentum could be either 1/2 or
2.0
2.5
Ep (MeV)
3.0
Fig. 39. Contour map of polarization in p~12C scattering. The polarization is given in units
of 0.01, and the contour interval is 0.05. The ordinate is laboratory scattering angle in de-
grees, and the abscissa is laboratory proton energy in MeV.
64
CARNEGIE INSTITUTION
-0.2 —
1.85
1.90
Ep (MeV)
1.95
Fig. 40. Polarization as a function of energy for 40Ar(p/p)40Ar at laboratory scattering
angle of 135°. The location of resonances found by a high resolution experiment39 are shown
as a series of vertical marks just above the bottom edge of the graph, and the location of the
single level used for the calculated curves is shown by an arrow. Points with error bars give
the experimental values measured with an energy resolution of 0.015 MeV. The full curve gives
the theoretical value for angular momentum 3/2, the dashed one for 1/2.
3/2. The angular momentum of each
state in the group should be the same
according to the theory of isobaric
analogue resonances. On this assump-
tion we could establish the total an-
gular momentum to be 3/2 and not
1/2, as is clear from reference to the
two theoretical curves in Fig. 40,
calculated from the single-level for-
mula by Professor Robson and J. L.
Adams of Florida State University.
Elastic Scattering of Polarized
Protons on Deuterons
Except for the scattering of nu-
cleons on nucleons, the elastic scatter-
ing of nucleons on deuterons is the
simplest interaction, in terms of the
number of particles involved, open to
experimental investigation with our
Van de Graaff machine. There is
strong experimental evidence that the
polarization effects of nucleon-nu-
cleon scattering, e.g. Year Book 6U
(p. 312), are two orders of magni-
tude smaller than those we can ob-
serve with the polarized ion source,
but measurements reported in Year
Book 65 (p. 77) show that this is not
true for proton-deuteron scattering.
Collisions of protons and nuclei with
spin zero show the nuclear force to
have central and spin-orbit depend-
ence. By using deuterons as targets
we hope to observe in addition a de-
pendence on the relative orientation
of the two spins.
The analysis follows the form used
for spin-zero nuclei as reported in
Year Book 65 (p. 73), which in turn
parallels the previous proton-deu-
teron phase-shift work of Christian
and Gammel,40 who assumed the po-
larization of the interaction to be
identically zero. The scattering by
spin-zero nuclei is described with a
two-by-two matrix (the M matrix)
and by spin-one nuclei with a six-by-
six matrix. The details of these calcu-
lations are not suitable for this re-
port, but the basic ideas are. The
appalling complication of the calcula-
DEPARTMENT OF TERRESTRIAL MAGNETISM
65
tions — about nine times more labo-
rious than for spin-zero nuclei — can
be done by the computer, leaving* the
physicist free to study the effect that
altering one or more phase shifts has
on experimentally measurable quanti-
ties. These calculations are exact in
that they assume only the validity of
quantum mechanics and the short
range of nuclear forces. The accuracy
with which one determines the phases
depends on the accuracy of the ex-
periments.
The phase shifts may be considered
to be a convenient form of nature's
parameters, and they have a simpler
relationship to nuclear forces than
other quantities, frequently in a form
like
TABLE 5. Set of Phases for S, P, and D Waves
Doublet
*- 2S,/2
Quadruplet
4S ,/,--
2p
2P
1/2
3/2
3/2
4P,/2
4P3/2
4,
P5/2
- 4D
1/2
'3/2
'5/2
03/2^"
4D,
sin Sz =
/ 00
JoU
fc /o£/(r)^^1(r)'^;i(r)dr
4,
'5/2
7/2
where Si is the phase shift, k the
wave number, U(r) the potential de-
scribing the nuclear force, and Ui (r)
the incident and Vi(r) the scattered
wave functions. The phase shifts
have their simplest form in the scat-
tering of spin-zero particles, e.g.
alpha particles, by nuclei of spin-zero.
A single phase shift is required for
each partial wave; the familiar
spectroscopic notation of S, P, D, etc.,
denotes orbital angular momenta of
I = 0, 1, 2. . . . The scattering of
protons by nuclei of spin zero re-
quires phases above S to be split ac-
cording to the possible values of the
total angular momentum /. They then
become S, P1/2, P3/2, D3/2, D5/2, etc.
For the scattering of protons by
nuclei of spin one the set of phases
given in Table 5 is necessary for S,
P, and D waves. The superscript de-
notes the multiplicity, i.e. the maxi-
mum number of substates possible,
which is here only quadruplet or
doublet, and the subscript denotes the
total angular momentum. Mixing be-
tween certain states is allowed (a
phenomenon forbidden for spin-zero
target nuclei), as denoted by the
arrows. The analysis of Christian and
Gammel has neither splitting of quad-
ruplet or doublet phases nor mixing
parameters; they describe the inter-
action with only 2S, 4S, 2P, 4P, 2D, *D,
etc.
The observed effects of polariza-
tion are small at low energies, sug-
gesting that they may be produced by
small changes in the phases of Chris-
tian and Gammel. We can in fact
explain the polarization measure-
ments in this manner. The polariza-
tion resulting from a splitting of the
4P phases by 1.4° is shown in Fig. 41
together with measured values. An
equally good fit can also be had with
the parameters mixing quadruplet
and doublet states, but this predicts
a distribution for polarized deuterons
scattered by protons that disagrees
in sign with measurements made
at the University of Wisconsin by
P. Extermann,41 whereas quadruplet
splitting is in good agreement. No
other simple variation of splitting or
mixing parameters gives a reason-
able fit to experiment.
66
CARNEGIE INSTITUTION
0.07
0.C6
0.05
0.04
0.03
0.02
0.0'
D (p , p) D
Ep= 3.00 MeV
B,
CM
Fig. 41. Polarization as a function of scattering angle for D[p,p)D at laboratory energy
of 3.00 MeV. The full curve shows the polarization resulting from splitting the 4P-phases by 1.2°,
viz. 4Pi 2 = 21. 6°, A?z/2 = 22. 8°, and 4P$/2 = 24.0°; the other phases, which are un-
split, have values from Christian and Gammel.40
Mixing between S and D states can-
not be determined by simple polari-
zation experiments, but can be found
by measuring the change in polari-
zation of a polarized beam by scatter-
ing. The experiment should be accom-
plished this summer and fall. It
requires two successive scatterings of
the protons — the first by deuterium,
and the second by helium. Helium
analyzes the polarization of the scat-
tered protons and thereby determines
the change in polarization caused by
the first scattering. Preliminary tests
of equipment have been made for this
experiment.
ATOMIC PHYSICS
L. Brown, W. K. Ford, Jr., V. C. Rubin,
and W. Trachslin
Atomic and Ionic Spectroscopy
with Foil Excitation
Our interest in foil excitation was
reported in Year Book 65 (pp. 76-
77) with a description of the impor-
tant ideas behind the method. We
have since tested the apparatus by
observing the foil-excited spectrum
of sodium that had been accelerated
to a speed of 2.75 X 108 cm/sec. The
experiment was successful, if evalu-
ated by the amount of information
obtained and the performance of the
equipment. Seventy-seven spectral
lines were observed whose wave-
lengths could be measured with an
accuracy of about 0.5 A. The in-
crease in quantity and quality from
earlier experiments comes from the
speed of the image tube spectrograph.
The results puzzle us, however, in
that only 5 or 6 of these lines can yet
be assigned to known transitions in
sodium. Because of this limitation the
present goal of the work — the meas-
urement of lifetimes of excited states
for use in determining abundances of
the elements in astronomical objects
DEPARTMENT OF TERRESTRIAL MAGNETISM
67
— may need modification. Whether
this goal can be reached and whether
there are other uses for these exper-
iments will not be clear for a while.
In the experiment a beam of so-
dium ions passes through a very thin
carbon foil. On emerging from the
foil an atom of sodium may be in any
charge state from neutral to com-
pletely ionized and, if excited, will
radiate under conditions approxi-
mating those of a free atom. The
beam glows after passing through the
foil, and its intensity decreases down-
stream from the foil. A lens images
the glowing beam onto the slit of a
spectrograph containing an image
tube, as shown in Fig. 42. In our
experiments the spectral range 3900
A to 6700 A is observed with a dis-
persion of 22 A/mm. The spectrum
so obtained differs from conventional
laboratory spectra in two ways. The
lines, each of which is a picture of
the beam in a single wavelength, de-
crease in intensity at different rates
with increasing distance from the
foil; the lines are not perpendicular
to the base line, but slant. Both these
characteristics are illustrated in
Plate 5, a portion of the sodium
spectrum.
The light incident on a given point
A in the line on the plate originates
at a specific point B in the beam, and,
since the light intensity of the beam
decreases according to the distance
from the foil, the variation will be
recorded as an apparent wedge-
shaped spectral line. The slant of the
spectral lines results from Doppler
shift. The light will be shifted in
wavelength depending on the velocity
of the radiating atom relative to the
lens. The changing angle between the
direction of the beam and the direc-
tion to the lens causes a different
Doppler shift at each point along the
line. This effect is used to determine
the charge state of the ions in a pro-
cedure invented by Trachslin. An
electric field of 15 kV/cm is applied
perpendicular to the direction of the
beam and parallel to the line of sight.
Neutral atoms are unaffected by it,
but ions are accelerated in the direc-
tion of the field to a speed propor-
tional to their charge and the dis-
tance from the foil. This velocity
component, which is perpendicular to
Image of
beam at
spectro-
graph slit
Fig. 42. The imaging (not to scale) of the glowing sodium beam onto the spectrograph slit.
Light passing from point 8 in the beam to point A in the plate is shifted in wavelength because
the radiating atom at B has a velocity component away from the lens. Because of the finite
aperture of the lens, the light at A results from a range of angles A#, and hence of relative
velocities. The width of the lines in the recorded spectrum is approximately proportional to the
diameter of the lens.
68
CARNEGIE INSTITUTION
the original direction of the beam, is and hence the slant of the line on the
small compared with the total speed, plate, enough to identify the charge
but. since it is in the direction of the state,
line of sight, it shifts the wavelength,
BIOPHYSICS
E. T. Bolton, D. J. Brenner, R. J. Britten, D. B. Cowie, S. Falkow,
D. E. Kohnc, A. Rake, and R. B. Roberts
A central theme runs through this
report, as it has in past years: the
enormous potential for understanding
the relationships and history of crea-
tures through the similarity of nu-
cleotide sequences in their DNA.
Some of the work — studies of brain
function and memory — does indeed
lie outside of areas guided by this
theme. Even there, measurement of
gene expression by means of messen-
ger RNA homology to DNA has a
plausible role. A second theme is also
present, derived from the measure-
ment of the temperature at which
nucleic acid strand pairs dissociate.
This method has been used both for
the classification of DNA and as an
index of the degree of divergence in
viral, bacterial, and animal systems.
As a subject moves from the period
of discovery toward development a
vocabulary grows with it, usually
confused in its early stages. Nucleic
acid sequence pairing is no exception.
The following glossary is an attempt
to relieve some of the growing pains.
It has several purposes: One is to
supply an introduction to the non-
specialist; a second is to present some
personal views on the usages of cer-
tain words. A few current results are
summarized, and there is, naturally,
an attempt to clarify the meaning of
some words. It is hoped that those
who feel strongly will criticize con-
structively and will correct as well
as contribute additional words or
phrases, with definitions.
Apologies are offered for stylistic
license, for theft of words from
other specialties, and for the narrow-
ing down of meanings and usages.
We proceed on the ground of at-
tempted clarity.
AN INSTRUCTIVE GLOSSARY
R. J. Britten
binding — 1. The reassociation of DNA
fragments with immobilized DNA.
2. The adsorption of nucleic acids on
hydroxyapatite or cellulose nitrate fil-
ters.
cation concentration — The concentra-
tion of cations determines the dimen-
sions of the charge cloud around the
negatively charged nucleic-acid mole-
cules and the repulsive Coulomb forces.
The cation concentration and the tem-
perature are used to control the con-
formation or secondary structure of
nucleic-acid molecules. Symbolized by
k representing the molarity of mono-
valent cations. Divalent ions are im-
mensely more effective and must be
considered separately.
Cerenkov counting — P32 is assayed in a
scintillation counter at high gain by
the light emitted by passage of j3 par-
ticles through water. Samples can be
counted without drying, they are not
adulterated, and valuable materials
can be reprocessed.
complementarity — The fraction of a set
of nucleic acid strands which will pair
with a second set, at a given criterion.
(Apologies to Niels Bohr.) See crite-
rion ; pair.
complementary — Said of a pair of nu-
cleic acid strands when all the nucleo-
tides in sequence can be paired in the
Watson-Crick sense — A:T and G:C.
Partially or imperfectly complemen-
tary are acceptable usages. See strand.
concatenation — The formation of chains
of fragments by the reassociation of
DNA strands that typically terminate
DEPARTMENT OF TERRESTRIAL MAGNETISM
69
at various places in the nucleotide se-
quence. Thus reassociated pairs termi-
nate in two single-stranded ends that
can reassociate with other complemen-
tary fragments to continue the chain.
This situation occurs for all DNA ex-
cept whole linear molecules from a
nonpermuted virus. See network; re-
associate.
C0t (kot) — The product of the con-
centration of DNA or RNA and the
time of incubation (moles nucleotides
X seconds per liter). The effectiveness
of the C0t in inducing reassociation
depends on the genome size, the pres-
ence of repeated sequences, the frag-
ment size, the cation concentration,
and the temperature of incubation. See
cation concentration.
criterion — The standard of precision of
pairing set by the cation concentration
in combination with the temperature
during incubation. Now expressed in
terms of these two parameters be-
cause the fraction of unpaired nucleo-
tides implied is not known. See im-
precise; precise.
CsCl density — In an ultracentrifugally
established equilibrium density gradi-
ent, the density of the cesium chloride
solution at the position of a band of
nucleic acid. Aside from minor effects
of temperature and pressure it is con-
trolled by the GC content of DNA and
by its secondary structure. Single- and
double-stranded DNA are separable.
denature — This word should be re-
stricted to the disruption of secondary
structure whether of paired or single-
stranded nucleic acids. Local or tem-
porary denaturation may occur, and
the process may be reversed. See sec-
ondary structure; dissociate.
discrimination — The capability of a
measurement of pair formation to dis-
tinguish between nucleic acids from
two species. This index of their re-
latedness is dependent on the criterion.
See criterion.
dissociate — To separate completely the
two strands of a pair so that one may
diffuse away from the other. See de-
nature.
divergence — 1. The increasing degree
of mismatch among a set of nucleic
acid sequences that occurs on an evo-
lutionary time scale ; example : The
divergence of the DNA of Salmonella
and E. coli has lowered the average
melting temperature of interspecies
pairs compared with that of intra-
species pairs. 2. The act of speciation
which leads to two distinct species
lines; example: The paleontological
record indicates that about 300 million
years have elapsed since divergence of
the lines leading to the amphibians
and the reptiles.
DNA-agar — A system in which large-
fragment-sized single-stranded DNA
is mixed with melted agar and rapidly
cooled. The DNA is immobilized. At
reasonable criterion sheared fragments
will diffuse through the agar and re-
associate with the immobilized DNA.
See criterion.
double stranded — Indicates that two
strands are held together by means of
complementary base pairs. May refer
to local or imperfect regions.
duplex — A base-paired structure be-
tween two complementary or partially
complementary nucleic acid strands. A
nucleic acid strand pair.
ERRF (urf) — An acronym used origi-
nally to describe extremely rapidly re-
associating fractions of higher-orga-
nism DNA. Now it is used to describe
the total fraction of DNA that behaves
as repeated sequences at a given cri-
terion. Here we risk this convenient
"word" for the first time in print. In
the jargon, one says "slow, fast, pre-
cise, or imprecise ERRF" and there
are other such unsavory but accurate
usages. See slow fraction.
family of repeated sequences — The set
of related sequences in the genome of
a given higher organism which will
reassociate with one another. The con-
cept of family is useful. At the mo-
ment, however, different families with
equal numbers of members and equal
degrees of divergence are not resolv-
able. See higher organism; related.
filter methods — DNA is immobilized
by binding to cellulose nitrate mem-
brane filters. Hybridization with RNA
may be conveniently assayed in this
way. In a recent modification further
adsorption of DNA is blocked by treat-
ment and assay of reassociation of
70
CARNEGIE INSTITUTION
fragments with the immobilized DNA
may be made.
fragment — A length of nucleotide se-
quence broken out of a longer mole-
cule by backbone strand scission. See
segment.
GC content — The fraction of the base
pairs in double-stranded DNA which
is made up of guanine and cytosine
(GC) as opposed to adenine and thy-
mine (AT).
genome size — 1. The haploid DNA con-
tent ^measured by chemical assay) of
a cell or virus particle. In rapidly
growing cells partially completed new
chromosome sets are not to be counted.
2. The amount of DNA per typical
unique sequence in the genome (meas-
ured by rate of reassociation). The
comparison of these results indicates
that chromosome pairs of a diploid cell
are nearly identical in sequence. It is
open season for difficulties with these
definitions, as in sex chromosomes;
polyploid plant hybrids ; significantly
different genes in a diploid heterozy-
gote.
higher organism — Here used in a some-
what vague sense to include all orga-
nisms with highly repetitive DNA se-
quences. To date, only viruses, bac-
teria, and a blue-green alga have been
found not to contain ERRF. It thus
appears possible that the line between
organisms that do and those that do
not contain ERRF will follow that
between eukaryotic and prokaryotic
creatures. See ERRF.
homologous, heterologous — Homologous
has been used to represent DNAs of
the same species, and heterologous, to
represent DNAs of a different species,
when they are added to a system con-
taining immobilized DNAs, as in:
Sheared heterologous carrier and
homologous tracer DNA were added
to DNA-agar. This usage of homolo-
gous is bad, because of confusion with
homology.
homology — The degree of similarity
between the nucleic acid sequences of
different species, as in : The homology
between two species is measured by
the capability of their DNA to form
interspecies pairs at a given criterion
and by the thermal stability of the re-
sulting pairs. Numerical specification
is difficult or impossible because more
than one parameter enters.
hybrid — 1. A nucleic acid strand pair
between RNA and DNA. 2. Pairs made
from DNA strands from two species.
The second usage should be avoided.
hydroxyapatite — Calcium phosphate
crystals produced in a controlled pre-
cipitation. It has the extraordinary
property of adsorbing double-stranded
but not single-stranded DNA at prop-
erly controlled salt and temperature
conditions. Imperfect strand pairs of
DNA bind well.
hyperchromicity — The increase in ultra-
violet absorbancy as the secondary
structure is disrupted. Best expressed
as percentage of the maximum ab-
sorbancy when the process is essen-
tially complete.
hypochromicity — The fall in ultraviolet
absorbancy as secondary structure re-
turns after its disruption. Best ex-
pressed as percentage of the maximum
absorbancy.
identical — Proposed substitute for
homologous; example: A competition
experiment with DNA-agar requires
unlabeled sheared fragments of DNA
from a competing species and labeled
sheared fragments identical to the
DNA immobilized in the agar. See
homologous.
imprecise — Imprecise pairing between
nucleic acids implies that a certain
fraction of the nucleotides is not
complementary to those on the opposite
strand, thus interrupting the second-
ary structure and reducing the thermal
stability. As a round number, a 1%
mismatch lowers the melting tempera-
ture 1°C. See melting temperature;
mismatch ; secondary structure ; ther-
mal stability.
incubation— The maintenance of sam-
ples of nucleic acid at elevated temper-
atures and controlled cation concentra-
tion in order to promote reassociation.
intrasequence heterogeneity — The vari-
ation of some parameter such as local
GC content along the length of a
nucleic acid sequence. The implication
is that averages taken over a series of
short lengths will show an interesting
distribution, such as intermixed blocks
DEPARTMENT OF TERRESTRIAL MAGNETISM
71
of high and low values of the param-
eter. Of interest are parameters such
as interspecies homology, purine/ py-
rimidine ratio, and membership in re-
peated families. See GC content.
kinetics — Shorthand for the time
course of a process, as in : The re-
association kinetics measures the con-
centration of complementary sequences.
log C0t plot — A method for presenta-
tion of reassociation kinetics in which
the fraction of DNA reassociated is
plotted against log C0t. It has the ad-
vantage that measurements over a
wide range of concentrations and times
can be combined to give a more com-
plete picture of the reassociation of
higher-organism DNA. See kinetics.
melting curve — The result of an assay
of the degree of dissociation or second-
ary structure disruption as a function
of temperature. Assay may be the in-
crease in ultraviolet absorbancy at one
or several wavelengths, displacement
of strands from immobilized DNA or
displacement from hydroxyapatite. See
hydroxyapatite.
melting temperature — The temperature
at which half of the total change in a
melting curve has occurred.
mismatch — The fraction of nucleotides
in a strand pair which is not comple-
mentary to the nucleotides on the op-
posite strand. Avoid usage that implies
complete lack of complementary se-
quences.
native — Signifies for DNA that the
Watson-Crick helix present in vivo
remains in a preparation. There may
be strand scissions and the DNA may
be fragmented but the strands have
not been separated. May be applied to
other nucleic acids with increasing in-
definiteness, as the secondary structure
is less well known in vivo and less
stable in preparations.
network — A multibranched structure of
reassociated DNA, typically seen with
DNA from higher organisms. As a
result of past translocations of re-
peated sequences, paired regions can
terminate in 4 single-stranded ends.
These ends reassociate with those of
other pairs and form branched struc-
tures. Branching could also occur with
a mixture of DNA strands in which
local regions have diverged to a suffi-
cient extent to create unpaired regions
that can reassociate with other strands.
See concatenation ; divergence.
nonrepeated sequences — Nucleotide se-
quences that occur only once in the
(haploid) DNA complement of an or-
ganism, as might have been expected
for most moderate length DNA se-
quences before repeated sequences be-
came known. See nucleotide sequence.
NTP — Abbreviation for nucleotide
pairs used in stating the length of
double-stranded DNA.
nucleotide sequence — A specifically
ordered row of nucleotides in the DNA
(or RNA) of an organism. In the
DNA, the complementary sequence is
always present (except for single-
stranded viruses) and the two are
often considered together, as in: The
rate of reassociation is a measure of
the degree of repetition of a nucleo-
tide sequence. See complementary.
order of reaction — The reassociation of
complementary strands is controlled by
a bimolecular collision and thus ini-
tially follows second-order kinetics.
Empirically, the reaction actually fol-
lows the second-order equation until it
is nearly completed. Thus the fraction
of unpaired DNA is 1/(1 + K C0t)
where K is a reaction rate constant
controlled by the variables listed
under C0t. In some cases one class of
nucleic acid is present in a minority
and the major component is immobi-
lized so that the number of sites avail-
able for pairing remains constant.
Then a pseudo first-order reaction is
observed and the fraction of the
minor component which remains un-
paired is exp ( — K'C'0t). K' is numer-
ically equal to K if C'0t is the effective
concentration of the major immobilized
component. See C0t.
pair — To form a base-paired structure
between two nucleic acid strands be-
cause of complementary or partially
complementary nucleotide sequences.
To reassociate. Also the product of
this process.
precise — Precise pairing between
strands implies that virtually all of the
bases are complementary and the ther-
mal stability will be close to that of
72
CARNEGIE INSTITUTION
native DXA if the fragments are long
enough (>200 nucleotide pairs).
reanneal — To incubate DXA in order
to induce reassociation, originally over
a range of temperatures since the op-
timum temperature was not known.
S e incubation; reassociate.
reassociate — To incubate nucleic acids
under conditions such that dissociated
complementary strands may collide
with each other and form base-paired
double strands. Also the act of form-
ing such a double strand. Reassociation
does not imply that the particular
strands were paired in vivo. The opti-
mum temperature is 20° to 30° below
the melting temperature and the rate
of the reaction increases with ionic
strength. Sec C0t; dissociate; incuba-
tion.
registration — When a complementary
pair of DXA strands slide (metaphor-
ically speaking) one along the other,
a large number of base pairs are com-
plementary in only one registration.
The strands are then in register and
can pair. Homopolymer pairs have as
many registrations as nucleotides.
DXA fragments containing several
members of one family of repeated
sequences can have several registra-
tions that lead to reassociation. See
family of repeated sequences.
related — Signifies that nucleic acid
fragments contain complementary nu-
cleotide sequences and thus will pair.
It is presumed that complementary
sequences of any significant length —
greater than, say, 20 or 25 pairs —
cannot arise independently. Thus re-
latedness in this sense implies com-
mon origin. The boundary between
relationship and chance similarity is
not yet clearly defined experimentally.
renature — Originally and widely used
to describe complementary double-
strand formation from single strands.
The implication that a nativelike struc-
ture is re-formed is broadly inappli-
cable. Pairs commonly form between
partially complementary strands, as in
the families of repeated sequences of
higher organisms, or with DNA from
different species.
repeated sequence — A segment of
nucleotide sequence which occurs many
times in the DNA of one cell of a
higher organism. A coined word would
be preferable here, since precise repe-
tition appears to be the exception
within the large, greatly divergent
families of repeated sequences.
repetition frequency spectrogram — A
representation of the quantity of DNA
as a function of its frequency of nu-
cleotide sequence repetition. Usually
graphed against the log of the fre-
quency, since the range is greater
than 106 and the resolution, as yet, is
not much better than a factor of 10.
saltatory replications — The hypotheti-
cal events by which families of hun-
dreds of thousands of similar nucleo-
tide sequences are produced in the
DNA of an organism. Large, closely
similar (young), as well as greatly
divergent (old), families are observed.
Thus families are produced in a time
short compared to the time required
for their loss by divergence (a few
hundred million years).
satellite DNA — A minor component re-
solved from the principal band of DNA
in a CsCl equilibrium density gradient.
Firmly engraved on the jargon by a
coincidence in time with Sputnik.
secondary structure — The intramolecu-
lar forces in addition to the phospho-
diester backbone bonds establish a
more or less stable conformation of a
nucleic-acid molecule in solution. The
resulting set of relationships of the
parts is bagged together as the sec-
ondary structure. It is thought to be
determined principally by hydrogen
bonds and hydrophobic forces. In the
case of double-stranded DNA the rela-
tively simple and stable Watson-Crick
helix is dominant while for single-
stranded molecules, less stable, more
transient (and less understood), rela-
tionships predominate.
segment — A length of nucleotide se-
quence distinct in some way from
neighboring stretches, e.g. a member
of a repeated family. Usually not
physically terminated by phosphodi-
ester backbone scission. See fragment.
sequence similarity by chance — A given
short sequence may recur in the DNA
of an organism, even though not of
common origin, simply because there
DEPARTMENT OF TERRESTRIAL MAGNETISM
73
is a limited number of permutations
of a sequence. Any sequence n long
will be expected to occur once in a
random sequence 4TO long. Imperfect
repetitions must be considered and it
is not known that DNA sequences even
approach randomness. However, short
sequences have reduced thermal stabil-
ity, and experimental test indicates
that with a reasonable criterion (50°,
k = 0.18) such sequence similarity is
undetectable in higher organism or
bacterial DNA. See criterion.
shearing — The process of fragmenta-
tion of DNA by creating shear forces
in a suspension. Controlled fragment
sizes have been produced from 105
NTP by stirring at a few hundred
rpm down to 5 X 102 NTP by passing
through an orifice at 50,000 psi.
similar — Broadly used to indicate that
two sequences are partially or imper-
fectly complementary.
slow fraction — A phrase of convenience
to describe the fraction of the DNA
which, at a given criterion, reassoci-
ates at the slow rate expected for
unique sequences, and which there-
fore does not exhibit repeated se-
quences. See unique.
species specific — Strictly, this phrase
refers to elements particular to a sin-
gle species, such as the mouse satel-
lite DNA. A phrase like species de-
pendent applies better to typical DNA
sequences since some homology is ex-
hibited between the DNA of creatures
even from different orders. See ho-
mology.
strand — Conveniently used to represent
fragments of single-stranded DNA or
RNA.
strand scission — The breaking of, or a
break in, the phosphodiester backbone
of a nucleic acid molecule.
thermal chromatogram — The set of data
resulting from the assay of thermally
eluted fractions. See thermal elution.
thermal elution — The process of dis-
sociating strand pairs with steps of
increasing temperature and eluting the
product from hydroxyapatite, DNA-
agar, or cellulose nitrate filters.
thermal stability — A convenient phrase
to indicate that strand pairs are dis-
sociated at a particular temperature
(with a given cation concentration;,
as in : The thermal stability is influ-
enced by the GC content and the
precision of matching of a nucleotide
sequence pair.
translocation — The displacement of a
segment of nucleotide sequence to
another part of the chromosomal DNA
through a series of genetic events.
unique— Designates that a particular
sequence occurs only once in a genome ;
for example, the concentration of a
unique sequence is known from the
genome size and the total DNA con-
centration. The typical situation for
viral and bacterial DNA. See genome
size ; slow fraction.
REPEATED NUCLEOTIDE
SEQUENCES
R. J. Britten and D. Kohne
Our knowledge of the repeated se-
quences in DNA has expanded con-
siderably since their discovery 3
years ago (Year Book 6U) . We do not
yet know the function or origin of
these sequences, but feel certain that
they must play a major role in evolu-
tion as the generality of their occur-
rence in large quantities in the
genomes of higher organisms has be-
come clear.
In the past year the list of species
containing repeated DNA has grown
and no exceptions have been found
among eukaryotic creatures. Individ-
ual organisms (onion, salmon, am-
phiuma) were examined which have
a majority of their DNA in this form.
Also progress was made toward an-
swering the question whether or not
saltatory replications have been the
principal source of the DNA of
higher organisms.
Further studies were made of the
degree of relationship among the
families of repeated sequences within
a given individual. It is now possible
to estimate the average rate of di-
vergence among the members of re-
peated families.
Although repeated DNA indeed ap-
74
CARNEGIE INSTITUTION
pears to occur universally among
higher organisms, there is great va-
riety in the quantity, the number of
members in individual families, and
the degree of divergence among the
members. A survey of the repeated
DXA of many species now appears
necessary to learn the significance of
this variety in evolutionary history.
Considerable effort has been in-
vested in increasing the certainty and
reproducibility of our methods of
measuring the rate of reassociation
and the properties of the reassociated
DXA molecules.
Mouse satellite DNA. A 10% frac-
tion of the DNA of the mouse — called
the mouse satellite — has played an
important role in the development of
our study of repeated DNA se-
quences. In the early summer of 1964
the success of the DNA-agar system
for the study of animal DNA sug-
gested to Britten the existence of
repeated DNA. As a result Britten
and Waring (Year Book 6U) started
studying mouse DNA and discovered
in the fall of 1964 that the satellite
reassociated very rapidly. It was in-
tuitively recognized that this was an
example of repetition. Subsequently,
measurement of the actual rate of
reassociation and its variation with
DNA concentration showed that this
was true, and that a short nucleotide
sequence was repeated about a mil-
lion times in the mouse DNA. Having
this extreme example in our pocket,
so to speak, made the more modest
degrees of repetition (tens and hun-
dreds of thousands) more believable
and gave momentum to the studies
which led to the recognition of their
general occurrence. We did not, how-
ever, choose mouse DNA because it
contained the satellite. In fact, we
were unaware of the existence of the
satellite until it appeared in the ultra-
centrifuge, although it had previously
been observed and reported.
We have not further studied the
mouse satellite but some recent work
must be mentioned because it has im-
portant implications for repeated se-
quences in general. Maio and Schild-
kraut42 have shown that the location
of the mouse satellite DNA within
the cell is not only nuclear but domi-
nantly, if not entirely, chromosomal.
This fact draws closer the relation-
ship between this satellite and other
repetitive DNA which must be chro-
mosomal because its great quantity
admits no other possibility. Peter
Walker and Ann McLaren43 in the
winter of 1964-1965 had prepared
with hydroxyapatite a fraction of
mouse DNA which was subsequently
shown to contain the satellite and
probably other repetitive DNA. Re-
cently Flamm, McCallum, and Wal-
ker44 showed that the satellite may be
separated into two fractions by cen-
trifugation in alkaline CsCl. These
fractions do not reassociate by them-
selves but do when mixed and are
therefore complementary strands.
Their composition is striking. Al-
though both contain 34% guanine
plus cytosine, one strand contains %
of the pyrimidines and the other, %
of the purines.
Occurrence of repetitious DNA,
Below is a list of those life forms
which have been examined for
repetitious DNA: (a) Rate of re-
association measured directly by
hydroxyapatite fractionation and/or
measurement of optical hypochromic-
ity as a function of time, (b) Labeled
sheared fragments bind to DNA from
the same species imbedded in agar
at a C0t so low that repetition must
be present, (c) Sheared nonradio-
active fragments of DNA from the
listed organism compete with the
DNA-agar reaction (b) of a related
species, reducing the amount of la-
beled DNA which binds to the im-
bedded DNA.
DEPARTMENT OF TERRESTRIAL MAGNETISM
75
Protozoans
dinoflagellate (Gyrodinium Cohnii) (a)
Euglena gracilis (a)
Porifera
sponge (Microciona) (a)
Coelenterates
sea anemone (Metridium) (tentacles)
(a)
Echinoderms
sea urchin (Strongylocentrotus)
sperm (a) (b) (c)
sea urchin (Arbacia) (sperm) (a) (b)
(c)
starfish (Asterias) (gonads) (a)
sand dollar (E chinarachnis) (c)
Arthropods
crab (Cancer borealis) (gonads) (a)
horseshoe crab (Limulus) (hepatopan-
creas) (a)
Mollusks
squid (Loligo pealii) (sperm) (a)
Elasmobranchs
dogfish shark (liver) (a)
Osteichthyes
salmon (sperm) (a) (b) (c)
lungfish (a) (b)
Amphibians
amphiuma (liver, red blood cells, mus-
cle) (a)
frog (Rana pipiens) (b)
frog (Rana sylvatica) (c)
toad (Xenopus laevis) (heart, liver,
red blood cell)
axolotl (Amby stoma tigrinum) (c)
salamander (Trituris viridescens) (c)
Birds
chicken (liver, blood) (a) (b) (c)
Mammals
tree shrew (c)
armadillo (c)
hedgehog (c)
guinea pig (c)
rabbit '(c)
rat (liver) (a) (b) (c)
mouse (liver, brain, thymus, spleen,
kidney) (a) (b) (c)
hamster (c)
calf (thymus, liver, kidney) (a) (b)
(c)
Primates
tarsier (c)
slow loris (c)
potto (c)
capuchin (c)
galago (c)
vervet (c)
owl monkey (c)
green monkey (c)
gibbon (c)
rhesus (b) (c)
baboon (c)
chimpanzee (a) (c)
human (a) (b) (c)
Plants
rye (Secale) (c)
tobacco (Nicotiana glauca) (c)
bean (Phaseolus vulgaris) (c)
vetch (Vicia villosa) (c)
barley (Hordeum vulgare) (a) (b)
pea (Pisum sativum, var. Alaska) (a)
(b)
wheat (Triticum aestivum) (a) (c)
onion (Allium sp.) (a)
This expanded list supports our
conclusion (Year Book 65, p. 94)
that repetitious DNA occurs in most
organisms that are more complex
than bacteria. DNA from viral, bac-
terial (E. coli, Clostridium perfrin-
gens), and blue-green algae genomes
examined by our methods have re-
vealed no repetitious DNA. No fungi
or higher algae have been examined.
These forms are intermediate be-
tween bacteria and higher organisms
and are therefore of great interest.
Thus far the line between orga-
nisms which do and those which do
not contain repetitious DNA follows
that between eukaryotic and prokar-
yotic forms. Eukaryotes contain repe-
titious DNA; prokaryotes do not.
Future work will expand the list of
organisms examined and should clar-
ify the boundary between those life
forms which do and those which do
not possess repetitious DNA.
The nonrepeated sequences of am-
phiuma. Amphiuma, sometimes called
the Congo eel, is a curious amphibian
related to the mud puppy. It has been
used as a laboratory specimen be-
cause of its large cells. It contains
the largest amount of DNA per hap-
loid cell of any known creature. A
haploid cell contains 8.6 X 1010 nu-
cleotide pairs, or about 30 times that
of human sperm.
76
CARNEGIE INSTITUTION
We have made an attempt, using
hydroxyapatite, to measure the rate
of reassociation of the slow fraction
of amphiuma DNA. The DNA was
sheared, dissociated, incubated (C0t
= 1000, 50°, k = 0.21), and frac-
tionated to remove ERRF. The slow
fraction was incubated (60°, k =
0.36) at a DXA concentration of 6
nig ml, and samples taken for a 2-
week period and assayed on hydroxy-
apatite. A slow reassociation was ob-
served in the fraction which eluted
at high temperature. Only the initial
part of the reassociation occurred in
the 2-week period and the rate con-
stant had the expected value for
unique sequences in a genome of this
size. Thus it appears that, while most
of amphiuma DNA consists of re-
peated sequences, a small fraction of
even this enormous genome appears
to consist of a set of unique se-
quences.
Time Course of DNA Reassociation
More complete curves tvith the use
of hydroxyapatite. When the DNA of
higher organisms reassociates, the
first part of the reaction is due to
collision of sequences that are present
in perhaps a million more or less
similar copies. Nonrepeated se-
quences collide with their comple-
ments much less often, and only very
much later — a factor of perhaps one
million in time — does the reassocia-
tion reaction go to completion. Pre-
viously (Year Book 65, p. 91) we
presented the data separately for the
fast and slow parts of this reaction.
Different techniques were used be-
cause of the widely separated reac-
tion rates. When hydroxyapatite is
used to assay the degree of reassoci-
ation, it is possible to measure the
complete time course under uniform
conditions by carrying out the re-
action at several widely different con-
centrations.
Reassociation of calf DNA. Figure
43 shows the results of such a series
of measurements for calf thymus
DNA sheared at 50,000 psi.
The measurements at the different
concentrations are concordant. This
establishes beyond doubt that the re-
association process results from a bi-
molecular collision. In turn, the early
part of the reaction is rapid only be-
cause of the high concentration (due
to sequence repetition) of the react-
ing species.
In the case of calf DNA there is
a clear separation between the ERRF
and the slow fraction. At this cri-
terion (60°C, K = 0.18) the rapid
fraction is a somewhat heterogeneous
collection with an average C0t for
half reaction of about 10~2. Therefore
the average frequency of repetition
is about 100,000. The slow fraction
is — again, at this criterion — a rela-
tively homogeneous population with
C0t for half reaction about 700 times
greater than that for E. coli DNA.
To measure this ratio accurately, la-
beled E. coli DNA (also sheared at
50,000 psi) was present at low con-
centration in the experiment done at
the highest concentration of calf
DNA. There will be further comment
on this experiment in the section
dealing with the similarity of the
DNA of diploid chromosome pairs
(P. 84).
Figure 44 shows a less extensive
set of measurements for salmon
sperm DNA. The striking differences
between these data and those for calf
are in the salmon's high quantity of
ERRF (about 80% of the DNA) and
the broad range of its repetition fre-
quency. The criterion for the salmon
measurements is more relaxed at 50°
than at 60°. Some measurements
done at the lower criterion with calf
show only 55% ERRF. Thus there
is a real difference.
The average repetition frequency
of the families is much less in the
DEPARTMENT OF TERRESTRIAL MAGNETISM
77
C0t (M x sec/liter)
Fig. 43. The kinetics of reassociation of calf thymus DNA measured with hydroxyapatite.
The DNA was sheared at 50,000 psi and incubated at 60° in 0.12 M PB. At various times
samples were diluted, if necessary (in 0.12 M PB at 60°), and passed over a hydroxyapatite
column at 60°. The DNA concentrations during the reaction were: triangle, 2; closed circle, 10;
open circle, 600; closed triangle, 8600 micrograms per milliliter; plus sign, radioactively la-
beled E. coli DNA at 43 y/ml present in the reaction containing calf thymus DNA at 8600 y/ml.
0.01
0.1 1 10
Cjjt (M x sec/liter)
100
1,000
Fig. 44. The kinetics of reassociation of salmon sperm DNA measured with hydroxyapatite.
The DNA was sheared at 50,000 psi and incubated at 50°C in 0.14 M PB. The samples were
diluted into 0.14 M PB at 50°C, if necessary, and passed over hydroxyapatite at 50°C. The
DNA concentrations during the incubation were: closed circle, 8; open circle, 1600 ju,g/ml.
78
CARNEGIE INSTITUTION
salmon than in the calf. Some se-
quences are present only in a few
thousand copies (C,t = 1). but a
minority is extremely repetitive, re-
acting at a Cot of les's than 0.01. The
few plants that have been examined
exhibit a pattern more closely re-
sembling; the pattern of salmon DNA
than that of calf DNA.
The contrast between the calf and
salmon ERRF suggests that repeated
sequences result from rare events —
perhaps immensely important to evo-
lution but nevertheless stochastic. If
the repeated sequences had a function
in chromosome structure, for ex-
ample, or in basic cell biochemistry
(genes for a protein class needed in
excess), one would expect great simi-
larity in the pattern from species to
species, at least in the vertebrates.
Here it appears that a carefully
planned survey is needed, and it
might produce surprises.
Precision of Sequence Matching
and Fractionation
A wide range of precision of
matching occurs among reassociated
pairs formed among the repeated se-
quences of DNA from a given species.
The divergence among the members
of a family of repeated sequences re-
sults from nucleotide sequence
changes that occur during the evolu-
tion of the species. The degree of
divergence can be taken to be a meas-
ure of the age of the family and is
of interest in relation to the history
of the evolution of species, and evolu-
tionary processes in general. There-
fore, we have explored hydroxyapa-
tite fractionation of repeated se-
quences on the basis of their thermal
stability.
Hydroxy apatite fractionation.
When the temperature of a hydroxy-
apatite column is raised, adsorbed
double-stranded DNA dissociates and
is eluted in the single-stranded form.
For precisely matched sequences the
temperature of dissociation measured
in this way is quite close to that
measured by optical hypochromicity
(Year Book 65) . Figure 45 shows the
results of fractionation of reassoci-
ated salmon sperm DNA. Sheared
(50,000 psi) DNA was dissociated
0.15
<
Z
Q
lo.io
o
_Q
u
c
c0.05
' 1 ' 1
1 ' 1 -
A
/ \ /H
/ \
/ \ /
\ / \
/ \ '
V v
/ \ 1
\
1 \ /
^ t
/ \ /
/ \ +
a /q
/ \ \
+— \
/ \ v
/ \ \
/ \
/ \ \
/ \
\/ \ \
/
X A \ <
/ ^
r \St^s^ x \ +
/ , "•"^
l , 1 T^-l
f 1 t 1
50
60
70
80
90
100
Elution temperature, °C
Fig. 45. Hydroxyapatite thermal fractionation of reassociated salmon sperm DNA. DNA sheared
at 50,000 psi v/as incubated at 50° in 0.14 M PB [CQt = 370) and passed over hydroxyapatite
at 50' in 0.14 M PB. The adsorbed DNA was eluted by exhaustive washing (0.14 M PB) at
each temperature (plus signs). Four fractions [65° , 70° , 85°, 90°) were separately dena-
tured [100°, 5 min) and reincubated (50°, 0.14 M PB, C0t about 10) and readsorbed on
hydroxyapatite under the same conditions. Two of these were again thermally eluted from a
column: open circles, 65° fraction; closed circles, 85° fraction. The other two were eluted with
0.4 M PB and melted in the spectrophotometer as shown in Fig. 46. The fraction bound and the
details of the shape of the elution curves depend on the degree of incubation.
DEPARTMENT OF TERRESTRIAL MAGNETISM
79
and incubated (criterion 50° C, k =
0.21) for a C0t of 270, and adsorbed
on hydroxyapatite. The resulting
thermal chromatogram (Fig. 45,
dashed line) shows the expected
broad range of thermal stability. To
establish the specificity of the frac-
tionation, samples eluted at 65° and
85 °C were reincubated. They were
then readsorbed and reanalyzed in
the same way. Specificity of the frac-
tionation is shown although the peaks
are quite broad.
The strand pairs formed during the
second incubation are not necessarily
the ones that were originally eluted.
Instead they are almost certainly new
duplexes formed by randomly as-
sorted pairings among this selected
set of strands. In each case, however,
the same average degree of precision
of relationship results. The 65° cut
peaks again at 65°,. and the 85° cut,
at 85°. The degrees of divergence are
thus characteristic of these popula-
tions. The members of the families
presumably have been randomly mu-
tated and all differ from each other
to about the same extent. Similar
studies have been done with calf
thymus DNA with entirely compa-
rable results. In addition, experi-
ments with labeled calf DNA frac-
tions indicate that very little sequence
homology exists between precisely
and imprecisely reassociating sets of
repetitive DNA. Thus these are ap-
parently quite distinct young and old
sets of families. It seems reasonable
to think of families being formed and
then diverging. A model (Plate 6)
schematizing this view appears at the
end of this report.
Length of repeated sequences. Are
reassociated pairs complementary
only in short regions or are they
complementary over most of their
length? The thermal stability of a
pair does not by itself answer this
question, since it appears that com-
plementary sequences that are 100
nucleotide pairs long will have a
thermal stability approaching that of
very long complementary sequences.
However, the ultraviolet hyperchro-
micity is a measure of the extent of
sequence matching. Therefore, the
hyperchromicity of a fraction known
to contain only strand pairs is a
measure of the average length of the
complementary regions present. Re-
sults for two such preparations are
shown in Fig. 46. Native, completely
complementary salmon DNA has a
hyperchromicity of about 0.25 (cal-
culated as 1.00 — 0.75 on the scale
of the figure) . Single-stranded DNA
has a hyperchromicity of 0.06 and
melts mostly at lower temperatures,
as shown by the upper curve.
The 70° and 90° fractions each
have a little better than half the hy-
perchromicity of native DNA. From
this we may conclude in each case
that the potentially complementary
regions are as long as half the frag-
ments and perhaps nearly as long as
the whole fragments.
50
60
70 80
Temperature, °C
Fig. 46. Optical melting curves in 0.14 M
PB of fractions of salmon sperm DNA. Frac-
tions prepared as described in Fig. 45: closed
circles, fraction eluted at 90°; open circles,
fraction eluted at 70°. The upper curve (tri-
angles) is for DNA which did not bind to hy-
droxyapatite (50°, 0.14 M PB) in the first incu-
bation and is therefore purely single stranded.
80
CARNEGIE INSTITUTION
Several complicating factors inter-
fere with drawing a more firm con-
clusion. Reassociated sheared frag-
ments are bound to have single-
stranded regions, since the two
strands do not ordinarily terminate
at the same points in the sequence.
All degrees of overlap will occur, and
for first collisions (no concatenation)
the expected hyperchromicity for
perfect pairing is perhaps only %
that for native DNA. We do not know
the extent of concatenation in these
preparations. Finally, of course, the
sequences in these families have di-
verged from each other, and the un-
matched nucleotides occurring within
the paired sequences reduce the hy-
perchromicity.
These measurements are corrobo-
rated by the hyperchromicity of re-
associated ERRF from many animals,
which is commonly between 0.16 and
0.20. A few CsCl density measure-
ments show a marked decrease in
density upon reassociation, which
also implies a good extent of comple-
mentary pairing.
It appears then that on the average
repeated sequences are not extremely
short (not less than 300 nucleotides)
and may be much longer than our
fragments, which average perhaps
500 nucleotides. One imagines that a
wide distribution of lengths will be
present after a family diverges and
that translocations as well as nucleo-
tide substitutions have occurred.
Rate of Divergence of Nucleotide
Sequences during Evolution
Some studies on the divergence of
nucleotide sequences during verte-
brate evolution were reported in
Year Book 68 (p. 394). This year's
investigations on repeated nucleotide
sequences throw additional light on
these studies.
General ehar act eristics of repeated
DXA from individual organisms. Re-
peated nucleotide sequences comprise
a large fraction of the DNA of all
higher organisms thus far examined
(see list, p. 73) . The repetitious frac-
tion of the DNA is made up of many
families of repeated nucleotide se-
quences. A family of repeated nucleo-
tide sequences can only arise from
the manyfold duplication of an exist-
ing nucleotide sequence. Current evi-
dence suggests that the production
of a family is a saltatory event and
occurs relatively rapidly. (These sal-
tatory events will be discussed in a
later section.) At the time of a salta-
tion all of the members of a family
are identical. With the passage of time
and the occurrence of mutations the
members of a family diverge from
one another to produce a family of
similar but not identical sequences.
The degree of similarity of the mem-
bers of a family can be used as an
index of the age of that family : the
less similar the family members, the
older the family.
The size of the fraction of DNA
which behaves as repetitious DNA
varies with the criterion set by the
conditions of the experiments for de-
tecting such DNA. At a constant salt
concentration, higher temperatures
of incubation require a greater preci-
sion of base-pair matching between
the interacting sequences in order
that a stable duplex be formed. Only
families with relatively similar mem-
ber sequences can form the more pre-
cisely paired duplexes needed for
stability at the higher temperatures.
At lower incubation temperatures
duplexes with more base-pair mis-
matching are stable and therefore
family members that are less similar
can form duplexes. Data of Martin
and Hoyer45 obtained by the DNA-
agar technique illustrate this point
well (Fig. 47). As the temperature
of incubation is lowered (at a con-
stant salt concentration) there is a
rise in the quantity of mouse DNA
fragments which reacts with the
DEPARTMENT OF TERRESTRIAL MAGNETISM
81
40 50 60 70
Temperature of incubation, °C
Fig. 47. These data, from Martin and
Hoyer,45 were obtained with the DNA-agar
technique by reacting chicken, rhesus, and
mouse DNA fragments with mouse DNA-agar
in 2 X SSC at different incubation tempera-
tures. Triangles, mouse DNA fragments on
mouse DNA-agar; squares, rhesus monkey DNA
fragments on mouse DNA-agar; circles, chicken
DNA fragments on mouse DNA-agar.
mouse DNA-agar.* Reassociation
rate data clearly demonstrate that the
standard DNA-agar method meas-
ures only the repeated sequences of
animal DNAs (Year Book 65, p. 94).
Since only repeated DNA was meas-
ured, it is clear that the increased
* The significance of the quantity of bind-
ing at 40 °C is not known. Random inter-
actions of nucleotide sequences may occur at
this temperature and influence the extent of
binding.
binding at lower temperatures was
due to an increase in the number of
families whose members were able
to interact to form stable duplexes.
Lowering the incubation temperature
to examine families with members
that are less and less similar is anal-
ogous to looking back in time. These
families represent a record of histor-
ical events that have contributed to
the composition of the present-day
animal genome.
Related nucleotide sequences in dif-
ferent animals. Figure 47 also pre-
sents data from Martin and Hoyer45
concerning the interaction of chicken
and rhesus monkey DNA fragments
with mouse DNA-agar. The extent of
observed DNA interaction between
two related animal DNAs varies with
the temperature of incubation. The
higher the temperature of incubation,
the lower the measured relatedness
between the two animal DNAs. Since
the standard DNA-agar technique
measures only the repeated DNA se-
quences, the variation in the meas-
ured relatedness with temperature
can only be due to changes in the
number of families held in common
which can form stable duplexes. Fur-
thermore, these studies indicate that
a large fraction of the present-day
chicken and rhesus DNA existed as
repeated DNA sequences at the time
the respective species lines diverged
hundreds of millions of years ago.
The lower average thermal stabil-
ity of the families held in common
between two species (Fig. 47), as
compared to the thermal stability of
families within a species, suggests
that family formation has occurred
many times since the divergence of
these species.
Loss of DNA relatedness during
vertebrate evolution. Measurements
reported previously (Year Book 63,
p. 394) and the more recent measure-
ments of Martin and Hoyer45 share
a surprising feature (Fig. 48) . There
82
CARNEGIE INSTITUTION
100.0
0
jhO.O
c
o
Chicken \
\
\
1.0
100 200 300
Time since divergence (millions of years)
Fig. 48. Reaction of different vertebrate DNA
fragments with mouse DNA-agar under
standard DNA-agar conditions. Each related-
ness percentage is the ratio of the extent of
interspecies binding to the extent of intra-
species binding (mouse on mouse) X 100.
Data were taken from several sources to con-
struct this figure. Values for bovine, man, and
hamster are from Year Book 63, p. 369; chick
and rhesus, from Martin and Hoyer45; the rat
from McLaren and Walker.46 Values for time
since divergence for the various animals were
obtained from paleontological evidence.
is no detectable "threshold" period
for accumulation of differences be-
tween the DNA of different species.
A threshold was expected since the
measurements were carried out at
60°C (about 30°C below the native
DNA Tm) and very many individual
nucleotide changes per fragment are
required at this criterion to prevent
pairing. The explanation is that the
relatedness was measured only be-
tween repeated families of DNA se-
quences. The new families produced
in one species will not reassociate
with the DNA of the other species,
since it is highly improbable that the
same sequence would undergo salta-
tory replication in both species. The
new families increase the total quan-
tity of repeated DNA and effectively
dilute the related sequences, thereby
reducing the measured relatedness
between the two species. The phe-
nomenon of new family production is
known to occur and is probably the
dominant mechanism for the initial
loss of relatedness between species.
It is also possible that many members
of repeated families are actually lost
by a process comparable to genetic
deletion. This would also result in
loss of relatedness without the ex-
pected "threshold. " It is not known,
however, whether deletions of this
type occur.
Figure 49 illustrates an inherent
aspect of studies of DNA relatedness.
The curves represent the loss of DNA
relatedness with time measured at
different temperatures of incubation
(at a constant salt concentration).
The lower temperature gives a larger
value for the time necessary to lose
50% relatedness. This reflects the
greater number of nucleotide se-
quence changes (longer time after
species line divergence) necessary to
render a pair unstable at lower tem-
peratures. The apparent half-time of
loss of DNA relatedness depends
upon the incubation conditions em-
ployed and must be interpreted with
care.
Rates of nucleotide sequence
change. To calculate the rate of nu-
cleotide sequence change it is neces-
sary to determine the number of
nucleotide changes that have oc-
curred during a known time interval.
The slow fraction of the DNA (see
glossary) is useful for this purpose,
since at the time of divergence these
sequences can be presumed to have
been identical and would have exhib-
DEPARTMENT OF TERRESTRIAL MAGNETISM
83
100
100 200 300
Time since divergence, millions of years
Fig. 49. The data are from Martin and
Hoyer.45 The relatedness values were obtained
with the DNA-agar technique by reacting
chicken (open circles) and rhesus (closed cir-
cles) DNA fragments with mouse DNA-agar at
different temperatures of incubation, 50°C and
60°C in 2 X SSC. Each relatedness percentage
is the ratio of the extent of interspecies binding
to intraspecies binding X 1 00.
ited perfect base-pair matching in
the reassociated pairs. During diver-
gence base-pair mismatches appear
in the interspecies DNA duplexes and
the thermal stability of the duplexes
decreases. The temperature at which
only 50% of the sequences of the slow
fractions of the two animals still pair
is a useful point for determining the
average number of base changes
since the time of divergence. The
measured relatedness between mouse
and rat slow fraction DNA is about
50% at 60°C and K = 0.18. Thus
60 °C is the estimated Tm of the pop-
ulation of reassociated related se-
quences. Precisely matched sequences
would melt at about 85° at k = 0.18.
From the paleontological record it
may be estimated that the time since
divergence of rat and mouse is be-
tween 10 and 25 million years. From
this one may estimate that the rate
of divergence reduces the average Tm
between species pairs 1° to 2°C per
million years. Only approximate in-
formation exists that relates the
number of base-pair mismatches to
the change in thermal stability of the
reassociated complex. The best esti-
mate is that somewhat less than 1%
base-pair mismatch reduces the Tm
about 1°C (Year Book 64, p. 322).
The rate of divergence between the
rat and mouse is therefore between
1% and 2% base-pair changes per
1 million years. The rate of change
of individual sequences is, however,
one half of this, since the sum of the
changes in the rat and the mouse has
been measured.
The data of Fig. 47 can be used
to make a rough estimate for the rate
of change of repeated sequences with
time. The curve for mouse reassocia-
tion extrapolates to a point on the
temperature scale 17° above that to
which the rhesus-mouse curve ex-
trapolates. This estimate (weighted
heavily by the more slowly changing
sequences) yields about one sixth of
a degree change per million years.
It seems that the average rate of
change of the repeated sequences is
less than that of the nonrepeated se-
quences. While the rates calculated
here are probably relatively inaccu-
rate, they are nevertheless the best
available and add another dimension
to our ideas about the evolution of
nucleotide sequences and evolution in
general.
Saltatory Events of Replication
Two major classes of evidence
show that repetitive DNA very likely
originated in rather sudden events of
84
CARNEGIE INSTITUTION
excessive replication of particular se-
quences. These classes are: the rela-
tive quantity of different degrees of
repetition, and the relationship of
repetition frequency to the age of
families.
Relationship between quantity of
DXA and repetition frequency. A
moderate fraction of calf or mouse
DXA has a repetition frequency be-
tween 10,000 and a million. There is
little or no DXA with a repetition
frequency between 10 and 1000.
There is a large slow fraction (see
glossary) which shows no repetition.
Such a grossly separated bimodal dis-
tribution could not result from the
production of a small number of ex-
tra copies of randomly selected se-
quences. It could result from the
recurrent copying of particular
"marked" sequences, or from events
in which very many copies are made
in a short time interval (saltatory
replication). Evidence is now avail-
able which clearly indicates saltatory
replication.
Relationship between age of fami-
lies and repetition frequency. For
those cases that have been examined,
young families of repeated sequences
which have not diverged very much
have as many or more than the old
families. The reassociation kinetics of
the two fractions of salmon DNA
shown on Fig. 46 were measured at
50° C. Both fractions showed heter-
ogeneity, but the 90° fraction started
to reassociate at one fourth of the
C0t required for the 70° fraction.
Thus there were present in some
young families even more members
than in any of the older families.
This observation is substantiated for
both calf and mouse DNA. Subse-
quent and more precise observations
of the correlation between number of
members and age of families will be
very informative regarding the his-
torical pattern of saltatory events.
Suppose family formation were a
continuous process in which members
were added in small numbers. Since
divergence occurs we presume that
ancient members of any family have
diverged from each other. Newer
members must be copied from pre-
existing members and must reflect
their divergence. In a continuously
produced family, then, recent mem-
bers are each closely related to only
a small fraction of the whole. Thus
sequences selected (as, for example,
the 90° fraction in Fig. 45) for close
relationship should show a lower fre-
quency of repetition, but as men-
tioned above the opposite seems to
be the case. With saltatory replica-
tions a family would be produced in
toto in a time short compared to the
period required for the divergence of
its members from each other. Thus
young families could have just as
many or more members as old ones.
A schematic diagram of family
history. Since families of many de-
grees of divergence are observed, we
must assume that a long series of
such events has occurred in each of
the species lines leading to present
higher organisms. Plate 6 shows a
three-dimensional schematic model
of the history of families produced
by saltatory replications. The scale
on the left represents the period of
time since each of the families of
repeated sequences was produced.
The lower scale is a measure of
the divergence within each family.
The cutouts represent a prediction of
the differential melting curves which
would be observed for the families if
the DNA were reassociated under
conditions such that only repeated
sequences paired. Thus all the mem-
bers of a young family are closely
related and all pairs melt at a tem-
perature near that of native DNA.
In ancient families a great deal of
divergence has occurred and the av-
erage melting temperature is far be-
low that of native DNA. By chance
DEPARTMENT OF TERRESTRIAL MAGNETISM 85
some members will have suffered a favorably selected phenotype or be
more changes than others and the associated with a favorable genetic
melting curve thus broadens with element. (4) Sufficient time must
time. Melting curves are somewhat pass for its dissemination throughout
broadened due to initial spread in the population. (5) If the set of
GC content. The range of quantities events can be described as a saltatory
of DNA in the various families re- replication, the growth of the family
fleets the wide range of repetition must be terminated within a reason-
frequencies observed. ably short time.
To bring this model to concrete No known processes combine all
form, a number of arbitrary deci- five of these events. Virus infection,
sions were made which are not cen- however, does involve some of them,
tral to the underlying ideas. For ex- In virulent infection very many
ample, the average rate of divergence copies are made of the virus genome,
is not well known, and the degree of In a lysogenic state a chromosomally
spreading is hardly known at all. It integrated segment is transmitted to
is not known how often the saltatory progeny. Further, it is known that
events occur; there is some suspicion segments of DNA that are capable of
that they are quite rare. The model genetic activity in the host may, in
suggests 107-year intervals but as far some living systems, be carried in the
as definite evidence goes, they could virus's life cycle. Cases are not known
occur 10 or 100 times more often. If where large numbers of viral DNA
about 10 families are observable copies are integrated into the host ge-
( criterion 60° C) each would contain nome; nor are any cases clearly es-
(e.g., in calf) about 5% of the DNA tablished in higher organisms where
and would contain, on the average, new host-type genes are carried in
105 copies of a sequence several thou- a virus's life cycle,
sand nucleotide pairs long. We are not proposing that salta-
Divergence betiveen species. It is tory replications are the result of
possible to visualize, on this model, virus infection, but the relationship
the divergence of the sequences may suggest new experimental ap-
shared between species. Suppose, for proaches. There are several instances
example, that the species lines to in which a quantity of DNA results
modern horse and cattle diverged from the production of many copies
about 70 million years ago. If this of particular segments of DNA: the
model represents the DNA of cattle, tremendous quantity of DNA (mainly
horse DNA would contain all the mitochondrial) in certain eggs; the
families older than 70 million years, mitochondrial and chloroplast DNA
For the younger families horse DNA in more typical cells. In none of these
would have a quite distinct set of its . cases is it known that the copies may
own. become integrated in such a way that
What is a saltatory replication? their descendants are transmitted to
The appearance in the genome of a the progeny. Such an event is neces-
f amily of repeated sequences involves sary in order that the members of
a number of events: (1) Many copies a repeated family may diverge from
must be made of a segment of DNA. each other. Only the small fraction
(2) A number of these copies must of the repetitive DNA which does not
be integrated into the genome in such show divergence (having high ther-
a way that they are duplicated and mal stability after reassociation) can
transmitted to progeny. (3) The re- be DNA which has been recently
suiting family must either determine replicated from a master copy.
86 CARNEGIE INSTITUTION
Chromosome Pairs also shows the reassociation of P32-
The nucleotide sequences of the labeled E. coli DNA which was pres-
DXA of the two members of a ho- ent at y200 of the concentration of
mologous pair of chromosomes are the calf thymus DNA. This DNA
not ordinarily completely identical, served as an internal standard, since
Genetic diversity derived from the both rates of reassociation are af-
two parents is normally present in fected to the same extent by the
a diploid organism. However, a close salt concentration, temperature, vis-
similarity is expected, since there is cosity (sheared calf DNA at 7 mg/
only a limited degree of individual ml) or possible nonspecific molecular
variation within a species. The meas- interactions. The result is that the
urements reported here show similar- half period for the reassociation of
ity on the nucleotide sequence level, the calf slow fraction is 690 times
In fact, no differences are detectable greater than that for the E. coli
with a limit of 1 % or 2 % nucleotide DNA. The ratio of the length of
substitution in one member of the DNA in bull sperm (3.2 X 109 NTP)
pair as compared with the other. Our to the length of the E. coli chromo-
present estimate, reported in a pre- some (4.5 X 106 NTP) is just 710.
vious section, of the rate of diver- The maximum error resulting from
gence of nucleotide sequences be- combining all of these data is ap-
tween the vertebrate species is about parently much less than a factor
1 % nucleotide substitution in a mil- of 2. We may conclude that the num-
lion years. Improved measurements ber of different nucleotide sequences
in the future may somewhat change in the diploid calf thymus cell is
this number but its exact magnitude about the same as that in the haploid
does not enter into the argument, sperm cell.
Thus the nucleotide sequences of the The divergence between homolo-
DNA of vertebrate species in general gous chromosomes. Therefore, upon
are slowly changing with time. A reassociation of the slow fraction of
mechanism must be present which calf thymus DNA, about half of the
preserves the similarity of the DNA duplexes will be formed with DNA
sequences of the homologous chromo- derived from the two different mem-
somal pairs while they slowly change bers of homologous pairs of chromo-
together. somes. The thermal stability of the
Hoiv much different DNA in the reassociated DNA is thus a measure
diploid cell? Figure 43 shows the ki- of the average degree of divergence
netics of reassociation of calf thymus between the sequences of homologous
DNA. Our interest here is the right- chromosome pairs. No reduction in
hand part of the curve, which results thermal stability due to this source
entirely from the reassociation of is detectable (Year Book 65, Fig. 50,
what we have termed nonrepeated p. 99) and the limit of accuracy is
DNA. The rate of reassociation of at present about 2°C, which implies
such DNA was reported in Year that no more than 2% of the nucleo-
Book 65 to be about that expected for tides in the average sequences of one
DNA occurring as single copies in member of a chromosomal pair differ
the haploid genome. There does not from those of the other member,
appear to be more than one compo- A direct comparison can be made
nent in the kinetic curve, and the between this limit (lowering of the
resulting reassociated DNA appears melting temperature of reassociated
to have a thermal stability very close strands from the two homologous
to that of native DNA. Figure 43 chromosomes) and the average low-
DEPARTMENT OF TERRESTRIAL MAGNETISM 87
ering of the melting temperature of this conservative mechanism clearly
reassociated strands from two differ- does not operate to such an extent
ent species. The melting temperature that all variation is eliminated from
of the nonrepeated fraction of labeled the population.
mouse DNA when reassociated with There are several levels at which
rat DNA was reported (Year Book steps (1) and (2) could occur: A
65, Fig. 51, p. 101, and Fig. 53, p. haploid set of chromosomes could be
102) to be at least 20° and possibly duplicated to produce a diploid cell
as much as 30° below that for the line as may occur in parthenogenesis,
same tracer on mouse DNA. A failure could occasionally occur in
It must be pointed out that the the process which supplies nonidenti-
commercial cattle from which this cal pairs of chromosomes to daughter
DNA was obtained are relatively in- cells. Crossing over could occur dur-
bred. Similar experiments are obvi- ing mitotic division, which would
ously necessary with DNA extracted create homozygous regions in the
from both haploid and diploid cells chromosomes of daughter cells. If
of wild species lines. such a mechanism operated on any of
Implications of the similarity. A these levels it would act to eliminate
high degree of homozygosity is main- genetic variation within the popula-
tained within any species while the tion and be quantitatively important
various species drift apart. The ge- in population genetics,
netic evidence indicates that this is
true on the level of expressible genes, A Source of Divergence?
whereas the evidence described here An extraordinary variety of forms
indicates that similarity is main- has appeared in the course of evolu-
tained in the actual nucleotide se- tion, reflecting a restless source of
quences of the DNA. Some definite originality. The nature of the forces
statements can be made about the behind the acts of speciation and di-
mechanisms that must operate to vergence remains a central problem in
maintain the similarity. The normal the study of evolution. We may look
events of mating and cell division do forward to surprises that are likely
not supply a mechanism for the main- to turn up before a full understand-
tenance of a close degree of similarity ing of these forces and processes is
between chromosome pairs. One must achieved. In these paragraphs we
look to relatively rare events to pre- propose a general idea suggested by
serve similarity between the members the presence of repetitious DNA, and
of a pair while the nucleotide se- indicate a role that the families of
quences of both change on an evolu- repeated sequences might play. We
tionary time scale. take as a starting point the view that
The necessary elements of these momentous genetic events capable of
rare events are as follows: (1) the inducing grand deviations in the
duplication of nucleotide sequences ; course of evolution must occur. Such
(2) the introduction of homozygous issues are raised with extraordinary
regions into the chromosomes of the clarity in the following quotation
germ line; (3) selection in the popu- from G. C. Simpson47:
lation that favors homozygosity. ml , . . » ,.» . , ., ,,
m, , , jsi. i The history of life is decidedly non-
These events must occur of ten enough mndom TMs fa evident in many fea.
and the selection for homozygosity tures of the record> induding such
must be sufficiently effective that a points already discussed as the phe-
great similarity is maintained be- nomena of relays and of major replace-
tween chromosome pairs. However, ments at defined times. It is, however,
88
CARNEGIE INSTITUTION
still more striking in two other phe-
nomena copiously documented by fos-
sils. Both have to do with evolutionary
trends : first, that the direction of
morphological whence also functional
and behavioral) change in a given
lineage often continues without sig-
nificant deviation for long periods of
time and. second, that similar or paral-
lel trends often appear either simul-
taneously or successively in numerous
different, usually related, lineages.
These phenomena are far from uni-
versal : they are not "laws" of evolu-
tion ; but they are so common and so
thoroughly established by concrete
evidence that they demand a definite,
effective directional force among the
evolutionary processes. They rule out
any theory of purely random evolution,
such as the rather naive mutationism
that had considerable support earlier
in the twentieth century. What direc-
tional forces the data do demand, or
permit, is one of the most important
questions to be asked of the fossil
record.
Suppose that an event occurs in
which a large number of unexpressed
genes, or better, a class of genetic
potentiality, is produced in the ge-
nome but only a small part appears
phenotypically. The small, expressed
part, "the top of the iceberg," if
favorably selected will introduce the
as yet unexpressed class into the pop-
ulation. In effect a hidden package
of potential genetic effects would
have been selected which after
spreading through the population
could lead to radically new features
that could not result from a detailed
balance of mutation and selection.
A saltatory replication producing
100,000 copies of the right sort of
gene is a candidate for a genetic event
with immense potentiality. Not all
the copies would be initially ex-
pressed. Perhaps not many ever
would be. But mutation, transloca-
tion, and recombination with other
genes would yield a whole range of
potential genetic activity which
would perhaps turn up long after the
appearance of the first effects that
resulted from the saltation. If the
early effects were selectively advan-
tageous the whole set of products of
the saltatory replication would be
introduced into the population. The
dynamics of selection would be fun-
damentally altered. The species might
then appear to take a surprising and
even an apparently purposeful course
of evolutionary change. Owing to the
great multiplicity of copies, their se-
lective elimination might be impos-
sible short of eradicating the species.
Potentialities important in the longer
term would then have an opportu-
nity to survive and reach fruitful ex-
pression.
Events of this general type would
be capable of causing a striking di-
vergence of two genetically isolated
populations even though the environ-
ments in which the two populations
existed were identical.
The wide occurrence of families of
repeated sequences and the implica-
tion of saltatory events in their pro-
duction as well as in evolutionary
processes indicates that sudden
events may be far more important to
evolution than we have heretofore
suspected.
INTRAMOLECULAR HETEROGENEITY
OF THE DNA OF TEMPERATE
BACTERIOPHAGES
S. Falkow and D. B. Cowie
The DNA of the bacteriophage X
can be differentiated into three dis-
similar regions in composition by
density gradient centrifugation
(Year Book 65, p. 559). In contrast
to these findings the virulent bac-
teriophages of the T series do not
exhibit a marked intramolecular het-
erogeneity. In addition, the distribu-
tion of guanine + cytosine (G + C)
pairs in the DNA molecules isolated
from a large variety of bacteria is
unimodal and approximately gaus-
DEPARTMENT OF TERRESTRIAL MAGNETISM
89
sian — even after shearing of the
DNA molecule.
The DNA isolated from several sex
factors may also be separated into
compositionally distinct regions (Fal-
kow et al, 1964,48 196649). Sex
factors, A, and virtually all other
temperate phages fit the definition of
episomic element; therefore each is
an independent genetic element of an
accessory nature, which is infectious
and which may be established in two
distinct intracellular, usually mutu-
ally exclusive states: the extrachro-
mosomal and the chromosomally at-
tached state. On the basis of these
similarities we considered that the
clustering of G + C pairs might be
a general feature of episomic ele-
ments. We have tested this possibility
with the DNA of several temperate
phages (genetically all episomic ele-
ments) by an optical method. Table 6
lists the phages employed and the
overall composition of their DNA.
The usual method for studying the
thermal denaturation of DNA has
been to follow the relative increase
in absorbance at 260 m/x as a function
of temperature. The midpoint of the
total increase in relative absorbance
(Tm) is used to obtain the overall
base composition while the hetero-
geneity of the DNA is calculated as
2o-, assuming that the compositional
TABLE 6. DNA Base Composition of DNA
Extracted from Temperate Phages and
Their Bacterial Hosts
DNA
Source
Mole Fraction
G + C%
from Optical
Method
Mole Fraction
G + C% from
Buoyant Den-
sity in CsCI
A
49.7
49.2
Adg
48.8
49.0
Ab2
50.7
50.5
<£80
52.1
53.0
P22
46.9
46.0
E. co/i K-12
50.5
50.2
S. typhimurium
50.9
51.0
distribution of DNA molecules is
normal. The heterogeneity of distri-
bution is more strikingly displayed
by treating the absorbance data at
any wavelength as a differential plot
in terms of the increment of absorb-
ance per unit of temperature (A
O.D./degrees C) as a function of
temperature or of mole fraction G +
C where 1°C is equivalent to 2.1%
G + C. Figure 50 shows that, in fact,
the melting data for a variety of bac-
terial DNA and the DNA of the
phage T3 do approximate a normal
distribution. On the other hand, as
shown in Figs. 51 and 52 the DNA
isolated from the temperate phages A,
$80, and P22 all exhibit a marked
heterogeneity.
Collecting these data requires con-
siderable care and a large number of
points must be measured during the
denaturation process of the highly
purified DNA. The points obtained
very early and very late in the de-
naturation were, of course, the most
subject to error, since these were
based on relatively small increases in
O.D. Small "peaks" were often ob-
served in the melting of all DNA
preparations. We have considered
these as real only if they could be
reproduced in the majority of prepa-
rations, as for example, in the case
of A and <£80 DNA in Figs. 51 and
52. The reproducibility of the method
is shown in Fig. 53 in which the
data from four separate experiments
with two different </>80 DNA prepa-
rations are plotted.
Examination of these melting pro-
files provides a simple means of de-
termining the composition of partic-
ular regions of a phage genome. In
Fig. 52 the melting profiles of A and
two A deletion variants, Ab2 and Adg,
are shown. A detailed comparison of
the profiles of A and Ab2 indicates that
the DNA region missing in the Ab2
deletion mutant is of high A + T
composition. The Adg variant is
90
CARNEGIE INSTITUTION
0.06
0.05
a.
E
o
-o
u
0.04
8 0.03
CL
o
<
0.02
0.01
P. mira bills
P. aeruginosa
40 50 60
Mole fraction (G + C)
70
Fig. 50. Differential plot of the increment of absorbance at 260 m/x per unit of temperature
(triangle O.D./°C) as a function of mole fraction (G + C) for four bacterial DNAs and for
phage T3 DNA.
known to arise by a chromosomal re-
arrangement consisting of a deletion
of high G + C (56%) phage DNA
and a concomitant substitution of the
bacterial galactose genes. Although
the base composition of the galactose
region is unknown, the overall base
composition of the bacterial chromo-
some is 50% G + C. A comparison
of the melting profiles of A and Adg
DNA is consistent with the genetic
data.
Phage P22 was the only generalized
transducing phage studied. The DNA
DEPARTMENT OF TERRESTRIAL MAGNETISM
91
— 0.04
— 0.03
u
0.02 5.
O
— 0.01
34
44 54
Mole fraction (G + C)
Fig. 51. Thermal denaturation of the DNAs of phage <j>B0 or P22 and for Salmonella typhi-
murium. Data plotted as in Fig. 50.
92
CARNEGIE INSTITUTION
34 39 44 49 54 59
Mole fraction (G + 0
Fig. 52. Thermal denaturation of the DNAs of A, Adg, and Ab2 phages. Data plotted
as in Fig. 50.
DEPARTMENT OF TERRESTRIAL MAGNETISM
93
44 54
Mole fraction (G + C)
64
Fig. 53. Thermal denaturation of the DNA of phage <£80. Data plotted as in Fig. 50.
94
CARNEGIE INSTITUTION
of this phage possesses an overall
base composition of 46.5 cc G + C.
This is significantly lower than that
of its host S. typhimurium, which
possesses 50.9% G + C (Fig. 51).
In the past there has generally been
a correlation of near identity of base
composition between a temperate
phage and its host, although it is rec-
ognized that identity is not necessar-
ily a prerequisite for transduction.
Since phage P:2 is a collection of cir-
cularly permuted gene sequences, the
spectral technique provides informa-
tion that could not be obtained by
hydrodynamic shear experiments
coupled with density gradient cen-
trifugation as employed by Hershey
{Year Book 63, p. 581) for A DNA.
It is possible to obtain a better
quantitative estimate of the hetero-
geneity of the melting at 260 m/x by
plotting the denaturation data on
probability graph paper which con-
verts a normal distribution into a
straight line. The points do approxi-
mate a straight line for the DNA of
bacteria and virulent bacteriophages,
whereas a line with three different
slopes was generated for the DNA of
the temperate phages. Figure 54
shows the data for E. coli and X as an
example of the method. Table 7 sum-
marizes the overall composition and
the fraction of the total DNA as de-
termined from the separate lines for
several of the temperate phage DNA
preparations we examined.
The implications of the differentia-
tion of the DNA of temperate phages
into regions of unique composition
are not clear. Skalka ( Year Book 65,
p. 565) has shown that the temporal
differentiation into early and late
genes appears to mimic the compo-
sitional differentiation of the X ge-
nome. These data are not available
for any other phage. As noted by
Hershey et al. (Year Book 63, p.
589), the molecular differentiation of
X could be physiologic or genetic. The
yy
97.5
1 1
1 1 M
Mill
1 U
95
L —
90
—
Ecoli 11
—
80
—
—
70
c
o
'£ 60
1 5°
1 40
—
1 30
—
\-^tf
—
J? 20
—
jt . i
—
10
— y
—
5
—
—
2.5
—
1.0
1 1
MM
Mill
1 1 1
34 40 46 52 58
Mole fraction (G + C)
64
Fig. 54. Percentage of total denaturation
plotted on normal probability graph paper.
The percentages correspond to one hundred
times the area under the normal distribution
curve.
genetic elements that we have cited
exhibit some sort of interaction with
the genome of their host. As we shall
point out in the following sections,
they all share nucleotide sequence
homology with the DNA of their host
and with each other in more than one
section along their molecular length.
The accumulation of bacteria-like
segments into the genome of a phage
does not imply that bacterial viruses
are simply a potpourri of bacterial
genes, but it does seem likely that
phages carry more of the bacterial
chromosome than simply a single rec-
ognition site. It is unreasonable to
think that these bacteria-like seg-
ments are present with no function;
DEPARTMENT OF TERRESTRIAL MAGNETISM
95
TABLE 7. Heterogeneity of Bacteriophage DNA from Data at a Single Wavelength
Overall Mole
Source Mole Fraction Fraction
of DNA (G + C) % Total (G + C)
Mole Mole
Fraction Fraction
% Total (G + C) % Total (G + C)
A
0.492
11.0
0.38
47.0
0.45
42.0
0.58
Adg
0.490
11.0
0.37
56.0
0.46
33.0
0.58
080
0.521
7.0
0.40
35.0
0.465
58.0
0.57
P22
0.467
9.0
0.36
37.0
0.44
54.0
0.50
selection undoubtedly demands that
any gene preserved must play some
role in the survival and maintenance
of the phage genome. We report be-
low that phages usually genetically
isolated, such as A and P22, nonethe-
less share a significant degree of ho-
mology. Does this mean that there
had been previous interaction be-
tween them or that certain nucleotide
sequences are conserved from a com-
mon ancestor? To what extent exist-
ing phages are really recombinational
assemblages remains to be seen. At
present we are simply trying to find
some evolutionary thread preserved
in their polynucleotide chains. It may
be useful to think of the possibility
that the molecular differentiation of
the temperate phages represents a
type of "historical differentiation"
and reflects their evolution.
NUCLEOTIDE SEQUENCE
RELATIONSHIPS AMONG PHAGES
080 AND X, AND E. COLI
D. B. Cowie and S. Falkow
The attachment sites for most of
the temperate bacteriophages of E.
coli are located at specific regions on
the bacterial chromosome. In general,
each region for prophage attachment
is characteristic for each of the dif-
ferent viruses. It has been postulated
that attachment specificity is a con-
sequence of homology between the
prophage and bacterial DNAs, and a
number of temperate phages have
been shown to be genetically related
(i.e., their genomes share nucleotide
sequences) to their bacterial hosts
(Year Book 65, p. 106).
The demonstration that DNA-DNA
homologies play a direct role in at-
tachment specificity has not been
made. The problem is complicated,
since more than one region of DNA
homology is found between A DNA
and the DNA of E. coli, and these
regions are distributed throughout
the A genome (Year Book 63, p. 385) .
Furthermore, A also contains several
regions of DNA homologous to P22
DNA and to the DNAs of other tem-
perate phages (Year Book 63, p. 381) .
The temperate phages A and ^80
have different attachment sites and
transduce different gene segments —
gafr and try+, respectively. Genetic re-
combinants between these two phages
have been obtained by Signer50 al-
though these phages are serologically
unrelated. An investigation of DNA-
DNA reactions among A, Adg, <£80,
and E. coli has been carried out to
examine, in part, the nature of the
homologies existing among them. The
thermal chromatographic method of
investigating DNA-DNA reactions
was used. The resulting elution pro-
files obtained are presented and dis-
cussed in terms of some of the unique
96
CARNEGIE INSTITUTION
characteristics of the DNA of these
temperate phages.
DNA-DNA Reactions Involving
Identical DNAs
Thermal elution profiles, each char-
acteristic of specific DNA-DNA re-
actions, have been obtained from
studies involving labeled DNA frag-
ments reacting with identical or het-
erologous DNA-agar preparations.
Figure do presents the results of re-
acting labeled E. coli DNA fragments
with E. coli DNA trapped in agar
(solid curve). The broken curve
shows the data obtained by reacting
680 DNA fragments with $80 DNA-
agar. The elution profile for the E.
coli DNA /DNA-agar reaction has a
maximum at 72 °C. Two maxima
are observed in the #80 DNA/DNA-
agar reaction, one at 69 °C, the other
at 73°C.
If it is assumed that the release of
labeled fragments from the DNA in
the agar represents the melting out
of fragments bound to the agar-em-
bedded DNA by the matching of nu-
merous base pairs, then the temper-
ature of elution may be attributed to
some function of the guanine plus
cytosine (G + C) content of the re-
acting DNAs. The elution profiles
shown in Fig. 55 would therefore in-
dicate that the overall G + C content
of these two DNAs is somewhat sim-
ilar, although that of the $80 DNA
is a little higher than that of the E.
coli DNA.
The usual method for studying the
thermal denaturation of DNA is to
follow the relative increase in ab-
sorbance at 260 m/x as a function of
temperature. Figure 56 represents
the differential plot of this change in
optical density observed with the
heating of native DNAs from $80
phage and E. coli in SSC/30. A com-
parison of these data with those ob-
tained in SSC/30 by the thermal
chromatographic method (Fig. 55)
indicates that the two procedures
provide similar results. The thermal
chromatographic data for the $80
DNA show some intrasequence heter-
ogeneity not resolved by the optical
method.
In the preceding section (Falkow
and Cowie) another analytic method
was employed which demonstrated
that large-scale intramolecular heter-
ogeneities not observed with bacterial
or virulent phage DNAs are found
among the temperate phage DNAs.
Results obtained with the thermal
chromatographic method are quite
comparable to those obtained with
the modified optical method and show
the same molecular heterogeneity
among the temperate phage DNAs.
For example, Fig. 57 presents the
elution profiles resulting from studies
of the reaction of P22 DNA fragments
with P22 DNA-agar (upper left-hand
curve) . The elution profile is complex,
showing two elution maxima. When
the data for P22 DNA from the optical
studies (Falkow and Cowie) are nor-
malized to the conditions used for the
DNA-agar thermal elutions and
plotted (Fig. 57, lower left-hand
curve), the results prove to be quite
comparable. A similar comparison is
shown for T3 DNA in Fig. 57, where
there is no large-scale heterogeneity
of distribution and the profiles are
unimodal.
Heterologous DNA-DNA
Reactions
These studies of identical DNA re-
actions serve as reference material
essential for studies of the reactions
involving heterologous DNAs.
A portion of $80 phage DNA is
homologous to the DNA of its bac-
terial host. Figure 58 (broken curve)
shows that $80 DNA fragments react
with E. coli DNA-agar. An elution
maximum is seen at a temperature
DEPARTMENT OF TERRESTRIAL MAGNETISM
97
200
0)
C
*£
a.
c
E
D>
D
1_
<
z
Q
-o
c
3
o
tilOO
D
O
-5
o
eg
</> 80 Phage DNA fragments / \
<t> 80 DNA - agar
E. COli DNA fragments
E. COli DNA - agar
Temperature, °C
Fig. 55. Elution profiles obtained from a study of the reaction of E. co// DNA fragments
with E. co// DNA-agar (solid circle), and <j>S0 DNA fragments with </>80 DNA-agar (open circle).
of 76°-77°C and another at 58°- coli DNA fragments with <j>S0 DNA-
59°C. agar (solid curve).
An elution profile was also obtained The temperature required to elute
from the results of a reciprocal ex- some of the heterologous DNA frag-
periment to study the reaction of E. ments from the DNA-agar prepara-
9S
CARNEGIE INSTITUTION
tions (76°— 77°C) is considerably
higher than that required for the
thermal elution of either </>80 or E.
coli fragments from their identical
DNA-agar preparations (Fig. 55).
This finding indicates that authentic
matching- of numerous base pairs
must have occurred between the het-
erologous DNAs. Furthermore, these
matching segments will be contained
in regions in each of the reacting
genomes where the G + C content is
higher than their individual overall
G + C content.
The elution profiles shown in Fig.
58 have another interesting feature.
About 13% of the </>80 DNA frag-
ments that were bound to the E. coli
DNA are eluted at temperatures
between 57° and 60°C. An elution
maximum at these temperatures is
observed when: (1) E. coli DNA
fragments react with ^80 DNA-agar,
(2) </>80 DNA fragments react with
30
t
i
>? 20
4> 80 DNA
O.
O
O
o
o
E. coli DNA
16
12 ^
70 80
Temperature, °C
c
"o
-*—
Q.
O
CD
Q-
Fig. 56. Differential plot of the change in optical density observed with heating native DNAs
in SSC/30; </>80 DNA (open circle); E. co///DNA (solid circle).
DEPARTMENT OF TERRESTRIAL MAGNETISM
99
(j>S0 DNA-agar, (3) A DNA fragments
react with A. DNA-agar, (4) A DNA
fragments react with ^80 DNA-agar.
This region is of interest because
of its extremely low elution tempera-
ture and, as will be seen, it is in-
volved in reactions between the DNA
of phages A and </>80.
Figure 59 demonstrates that <£80
is genetically related to phage A. The
solid curve represents the elution pro-
file obtained from a study of the re-
action of A fragments with <£80 DNA-
agar. The elution data show two
prominent maxima, one at 58 °C, the
other at 74 °C. Twenty-two per cent
of the A DNA reacts with the </,80
DNA. The major elution peak (74°C)
is above the 72 °C peak observed for
the reaction of E. coli fragments with
E. coli DNA-agar (50% G + C) . The
higher temperature required for elu-
tion of the A fragments from the E.
coli DNA indicates that this binding
is probably mostly concerned with the
left end of A DNA (56% G 4- C).
60 65
Temperature, °C
70 75 65 70
0.04
0.03
0.02
£ 0.01-
Q-
o
j f^2 DNA fragments
I P22 DNA - agar
I T3 DNA fragments
agar
DNA
500
400
300
200
100
t
i
c
0
E
CD
D
<
Z
Q
"D
C
D
O
_Q
o
>s
o
D
O
40 45 50 45 50
Mole fraction G + C
55
Fig. 57. Elution profiles obtained from studies of the reaction of P22 DNA-agar fragments with
P22 DNA-agar (upper curves, solid circles) compared with optical data obtained by the thermal
denaturation of native P22 and T3 DNAs (lower curves, open circles).
100
CARNEGIE INSTITUTION
CL
C
0)
E
o
<
Z
Q
C
O
—
O
>
u
o
o
'•n
o
E.
1 1
£■<?// DNA fragments
T
700
- ?
80 phage DNA - agar *
—
Cp 80 phage DNA fragments
E. coli DNA-agar
sc r»
600
t i
i
i
i
\
\
\
\
\
\
500
i i
i
i
i
V
\ \
400
i
\\
1
k
\
/
300
1
i
i
i
1
w
\
/
1
\
200
- l
\
\
I
/
/
/
/
100
\
\
i
i
60 70 80
Temperature, °C
Fig. 58. Elution profiles resulting from a study of the reaction of £. co/i DNA fragments with
080 DNA-agar (solid circles) and 080 fragments with E. coli DNA-agar (open circles).
DEPARTMENT OF TERRESTRIAL MAGNETISM 101
These data agree with the results of agar-bound $80 DNA. When the dif-
Ingraham, Ehring, and Hershey ferences between the two elution dia-
(Year Book 65, p. 563), who used A grams (shown in Fig. 65) are plotted,
DNA fractions (41% G + C, 46% however, two prominent maxima are
G + C, and 56% G + C) to react observed at 58°C and 72°C, as shown
with $80 DNA-agar. Ingraham et at. in Fig. 60. These peaks represent A
found that most of the binding oc- DNA fragments having homology
curred with the 56 % G + C fraction with E. coli DNA as well as with $80
and was twice that observed for the DNA. Other A fragments not affected
46% fraction and almost eight times by the presence of the E. coli com-
the binding observed with the 41% petitor fragments are homologous to
G + C fraction. </>80 DNA but not to E. coli DNA.
The material eluted at 58 °C in The homologous region eluting at
SSC/30 represents a much lower 58° C is believed to be contained in
G + C content, and is possibly con- the low G + C segments of the A
tained in the 41 % G + C fraction of DNA — that is, in the central portion
the A genome. On the other hand, or the right side of the X genome.
imperfect base sequence homology be- A portion of the Adg DNA consists
tween the reacting portions of these of E. coli DNA substituted for the A
heterologous DNAs would also cause DNA contained in the 56% G + C
a reduction in the temperature re- fraction (left side of the A genome)
quired to elute the bound fragments as shown by Kaiser and Hogness.51
from the DNA-agar, compared to a When $80 DNA fragments are sub-
situation where perfect homology jected to reaction with Adg (Fig. 61)
exists. The fact, however, that this no significant differences are ob-
region is observed with reactions in- served in the elution profile from that
volving identical DNA reactions (A obtained when A DNA fragments
DNA fragments with A DNA-agar or were reacted with $80 DNA-agar
<£80 DNA fragments with $80 DNA- (solid curve, Fig. 59). This result
agar) probably means that almost suggests that the deleted portion of
perfect matching of many nucleotide the Adg molecule does not contain the
sequences in this fraction had also A region homologous to $80 and elut-
occurred when the heterologous ing at 58°C.
DNAs were incubated together. The results presented above may
A portion of the A DNA which re- be summarized as follows: A sizable
acts with the $80 DNA is homologous portion of the DNA of lysogenic (or
to E. coli DNA. Figure 59 (broken semilysogenic) viruses has been
curve) shows the effect of adding a shown to be homologous to the DNA
large excess of cold E. coli DNA frag- of their bacterial hosts. Nucleotide
ments to the radioactive A DNA frag- sequence relationships among four
ments and then incubating both at viral-host systems (A phage/E1. coli),
60°C with the $80 DNA-agar. A 15 (TAU)~ phage/E. coli, P22 phage/
43% reduction in the binding of the Salmonella, and T3 phage /E. coli
X fragments to the $80 DNA-agar have been studied by means of an
was observed. A reduction in the extension of the DNA-agar technique
binding of the A fragments over the of Bolton and McCarthy.52
entire elution diagram (shown in the The results presented here thus
figure) is to be expected, since the provide another example demonstrat-
competitor, cold E. coli DNA frag- ing genetic relationships between the
ments, reduces the quantity of A frag- lysogenic virus $80 and its bacterial
ments left available to react with the host E. coli. Furthermore, a portion
102
CARNEGIE INSTITUTION
500
XDNA fragments
080 DNA - agar
A + £". coli fragments
^80 DNA- agar
Temperature, °C
Fig. 59. Elufion profiles obtained from a study of A DNA fragments (.075 fig DNA) re-
acting with cS80 DNA-agar (37 fig DNA, open circles) and a similar experiment (closed circles)
with unlabeled competitor E. coli DNA fragments (460 /xg DNA, closed circles).
of the <£80 DNA which is homologous
to E. coli DNA is also homologous to
X DNA; other </>80 DNA segments are
capable of reacting with A. DNA but
not with E. coli DNA. DNA homolo-
gies have been shown to exist among
all of the temperate phages investi-
gated (with the exception of the de-
fective phage lysogenic to strain 15
E. coli); see Year Book 65 (p. 107).
The viral-viral or viral-host homol-
ogies are contained in numerous seg-
ments throughout the viral DNAs.
In the <£80-A system two DNA frac-
tions eluting at temperature differ-
ences of 16°C (58°C and 74°C, Fig.
59) are found to be homologous. This
extreme temperature difference prob-
ably represents an extreme variation
in G + C content of the reacting
DEPARTMENT OF TERRESTRIAL MAGNETISM
103
segments and therefore separate re-
gions of the A genome, and of the </^80
DNA as well.
One component of the A DNA, ho-
mologous to both E. coli and $80, is
contained in a nucleotide cluster elut-
ing at a temperature of 58 °C. If the
specificity of prophage attachment to
host chromosome is dependent upon
unique DNA homologies with host
cell DNA — and since the attachment
sites for A and $80 are presumably
300
200 —
4)
Q.
c
3
O
u
•*-
[>
*-
u
8
<3
60 70 80
Temperature, °C
Fig. 60. Plot of the difference between the
two elution profiles shown in Fig. 59.
significantly different — the nucleotide
sequences contained in this one com-
ponent probably do not enter into the
attachment process. Another fraction
of the A genome, that eluting around
72°C (Fig. 60), might be excluded
for the same reason. It must be em-
phasized, however, that the results
presented are only qualitative, and
the broad elution maximum at 72 °C
4>80 Phage DNA fragments
X dg DNA - agar
£100
Q_
60 70
Temperature, CC
80
Fig. 61. Elution profile obtained from a study
of the reaction of (£80 DNA fragments with
Adg DNA-agar.
104
CARNEGIE INSTITUTION
12
10
c
o
£
a
<
Z
Q
o
-£2
c 6
o
DNA fragments
£". Ctf// DNA - agar
\
80 DNA fragments
CO// DNA - agar
/
50
60
70
Temperature, °C
80
Fig. 62. Elution profiles obtained from a study of the reaction of A DNA fragments and E. coli
DNA-agar (solid circles) and <£80 DNA fragments with E. coli DNA-agar (open circles).
DEPARTMENT OF TERRESTRIAL MAGNETISM
105
might mask a small fraction of re-
acting DNA that is involved in the
attachment process.
Studies of the reactions of A or $80
DNA fragments with E. coli DNA-
agar, however, provided significantly
different elution profiles, which are
compared in Fig. 62. It is apparent
that the G + C content of the X-E.
coli homologies are in general lower
than that observed for the <j>80-E. coli
system.
The differences between these two
elution profiles can be graphically
presented, as shown in Fig. 63. In
60 70
Temperature, °C
80
Fig. 63. Plot of the difference between the
two elution profiles shown in Fig. 62.
the figure the percentages of the
bound $80 fragments are subtracted
from the corresponding values of the
A fragments, which are bound to the
E. coli DNA-agar. These difference
data show two major peaks, one with
a maximum at 66 °C, the other at
78°C. A third peak at 58°C is also
seen, but as noted, this material is
not believed to be involved in the at-
tachment process.
The 66°C region represents the
percentage of excess binding of the
A DNA fragments over that observed
with </>80 DNA. The 78°C peak cor-
responds to the percentage of excess
binding of the $80 fragments to E.
coli over that observed with the A
DNA.
If the requirement for attachment
of viral DNA to host chromosome for
A and (£80 is dependent upon differ-
ences in nucleotide composition in a
viral-host DNA homology, these two
regions represent the areas having
the maximal nucleotide differences.
The results described serve as
guidelines for future investigations.
Lambda mutants having altered or
deleted attachment sites might be ex-
pected to show alterations in the elu-
tion profiles in the 66° C region, and
$80 mutants, in the 78 °C region. The
thermal chromatographic method
provides a means of isolating and ex-
amining specific DNA segments in
terms of their genetic interrelation-
ships and functions, and it is hoped
that further elucidation regarding
prophage attachment to host DNA
may result.
In the preceding section (pages
94-97) it was demonstrated that
large-scale intramolecular heteroge-
neity exists among the temperate
phage DNAs. The thermal chromat-
ographic results presented above
confirm this conclusion. The elution
profiles obtained from studies of iden-
tical DNA-DNA reactions involving
$80 DNA (Fig. 55), P22 DNA (Fig.
106
CARNEGIE INSTITUTION
57), and A DXA {Year Book 65, p.
114) are all complex. Furthermore,
studies involving DNA-DNA reac-
tions among- these temperate phages
show a large number of different
elution maxima having a spread in
elution temperatures ranging from
5S C to 78 T. The specificity of these
reactions and the wide differences in
their temperature characteristics im-
ply that large-scale heterogeneities
exist among these DNAs.
QUALITATIVE ASPECTS OF
MICROBIAL DNA DUPLEXES
D. J. Brenner and D. B. Cowie
The percentage of DNA fragments
bound to DNA-agar yields little in-
formation concerning the precision of
complementarity of the bound nucleo-
tide sequences. Martin and Hoyer,53
using animal DNAs, and Brenner,
Martin, and Hoyer,54 using bacterial
DXAs, have shown that the amount
of interspecies DNA binding in agar
decreases as the incubation tempera-
ture is raised. A logical explanation
for the increasing discrimination at
higher temperatures between intra-
species and interspecies duplexes is
that in the complementary sequences
held in common between species (or
at least a significant proportion of
them) are less precisely paired than
the intraspecies duplexes. If this is
true the thermal stability of a paired
sequence should decrease with an in-
crease in unpaired bases within the
sequence. When unique animal DNA
sequences are reassociated in free so-
lution the interspecies pairs do dis-
sociate at lower temperatures than
the intraspecies duplexes (Britten
and Kohne, Yeo/r Book 65, p. 78).
Animal DNAs contain large num-
bers of highly repetitive nucleotide
sequences that fall into families with
imprecise sequence complementarity.
Since the rate of reassociation is de-
pendent upon the concentration of
related sequences and the animal ge-
nome is so large, only the sequences
contained in the DNA families re-
associate within a 16-hour incubation
period.
Bacterial and phage systems were
chosen to investigate further the pre-
cision of interspecies pairs. There is
no evidence for repetitious DNA se-
quences in bacteria at incubation
temperatures above 45°, and no mul-
tiple DNA sequences have been dem-
onstrated in bacteriophage, except
for redundancy in the terminal ends
of several coliphage DNAs.
The use of the term "homology"
has been purposely avoided in the
preceding paragraphs. Homology was
used previously to depict intraspecies
and interspecies binding regardless
of incubation conditions. In this con-
text homology has been misinter-
preted to imply the existence of
identical stretches of DNA between
species; for example, if 35% ho-
mology were reported between E.
coli and Salmonella typhimurium,
these organisms might be assumed
to have 35% identical DNA. It is well
known that binding percentages in
bacteria and animals are completely
dependent upon the stringency of
incubation conditions. For example,
Martin and Hoyer found 90% com-
plementarity between the DNAs of
mouse and rhesus monkey at 40°, and
20% complementarity at 64°.
Restricting our discussion to phage
and bacterial systems, where DNA
sequence repetitions do not occur,
complementary pairs will be termed
"precise" if they exhibit the thermal
stability of reassociated intraspecies
pairs, and "imprecise" if their ther-
mal stability is less than that seen in
the intraspecies pairs.
E. coli and S. typhimurium were
selected as the organisms with which
to assess the precision of pairing in
interspecies bacterial DNA duplexes.
DEPARTMENT OF TERRESTRIAL MAGNETISM 107
In agar these organisms show 36% droxyapatite column in E. coli*/E. coli
nucleotide sequence complementarity and E. coli*/S. typhimurium reac-
under standard incubation conditions tions (where the asterisk denotes
(60° in 2 X SSC) and 28% comple- radioactive DNA fragments) incu-
mentarity when the incubation tern- bated 4 hours at 60° and 66° is the
perature is raised to 66°. The pre- result of duplexes formed between
cision of DNA sequences in these or- labeled and unlabeled fragments,
ganisms was determined by thermally Eighty to 90 per cent of the unlabeled
eluting reassociated duplexes from hy- fragments reassociate at the concen-
droxyapatite {Year Book 65, p. 78). tration and period of incubation em-
At 0.14 M PB hy droxyapatite retains ployed (Britten and Kohne, Year
double-stranded, but not single- Book 65, p. 78) .
stranded DNA. Double-stranded DNA The E. coli*/E. coli and E. coli*/
is eluted at a 0.4 M PB. S. typhimurium samples were loaded
Denatured P32-labeled E. coli frag- on hydroxyapatite columns held at
ments were incubated alone (as a the 60° or 66° incubation tempera-
control experiment) at 60° and 66°. ture. The bound (double-stranded)
The data in Table 8 show that at both DNA was then denatured by raising
incubation temperatures little or no the temperature of successive 0.14 M
reassociation of the labeled frag- PB washes by 5° increments to 100°
ments with one another, in excess of and finally washing the column with
that seen in a zero time control sam- 0.4 M PB to remove any remaining
pie, occurred after 4 hours. One may double-stranded material. The eluents
assume, therefore, that virtually all were assayed for total DNA and
the radioactivity bound to the hy- radioactivity to obtain the differen-
TABLE 8. Elution of P32-Labeled E. coli Fragments from Hydroxyapatite
Zero Time F32-
Labeled E. coli
Fragments
P32-Labeled E. coli
Fragments,
60°, 4 hours
P32-Labeled E. coli
Fragments,*
66° , 4 hours
Total C0f
0.0
0.02
0.02
Cpm unbound
1026
1030
893
Cpm bound
16
14
18
% bound
1.5
1.4
2.0
* The total counts are lower in this sample because it was assayed two days after the other
samples.
Three 0.4-/i,g samples of P32-labeled E. coli DNA fragments, each contained in 1 ml of 0.14
M PB, were thermally denatured and quickly cooled. The zero time sample was immediately
passed through a 5.5-ml hydroxyapatite column at 60°. The two other samples were incubated
at 60° or 66° for 4 hours and then passed through hydroxyapatite at the temperature of
incubation. The column was then washed with 0.14 M PB at 60° or 66° until the eluted counts
were down to a background level of 7 cpm. The column was then consecutively washed
with 0.14 M PB at 100° and 0.4 M PB to elute all counts initially bound to the column. The
samples were assayed for radioactivity in the eluting solution by Cerenkov counting. The P32-
DNA fragments contained little acid-soluble material.
108
CARNEGIE INSTITUTION
tial elation profiles. Elution profiles
of the unlabeled material (Fig. 64)
show the melting characteristics of
E. coli and 5. typhi 'murium DNA af-
ter incubation at 60° and 66°. The
slightly higher G + C content of S.
typhi murium DNA is reflected in its
higher T .
Thermal elution profiles obtained
from radioactivity assays are pre-
sented in Fig. 65. E. coli* /E. coli
elution profiles from 60° and 66° in-
cubations are quite similar to the
elution profiles obtained from un-
labeled fragments (Fig. 64), although
their Tms are 3° lower. The E. coli*/
S. typhi murium DNA profile from
the 60° incubation is displaced 9.5°
below that of an E. coli* JE.coli pro-
file from a 60° incubation and the
E. coli* /S. typhimurium profile from
a 66° incubation is displaced 7.5° be-
low the corresponding E. coli* /E. coli
profile. Therefore the complementary
nucleotide sequences in E. coli*/S. ty-
phimurium duplexes are imprecise.
While one cannot say how much, if
any, of the interspecies nucleotide se-
12
11
10
o
c
o
•Z 5
Q.
o
E. coli 60
60
E.coli 66
S. typhimurium 66
S. typhimurium 60
100? U4M PB
Temperature, °C
Fig. 64. Thermal elution profiles of unlabeled reassociated bacterial DNA from hydroxy-
apatite. Fragments of E. coli or S. typhimurium DNA were incubated at 60° or 66° and then
passed through a hydroxyapatite column. The bound fragments were removed from the column
by a series of thermal elutions.
DEPARTMENT OF TERRESTRIAL MAGNETISM
109
quences are precisely paired, this per-
centage is necessarily small.
The Tm of the E. coli*/S. typhi-
murium duplex from a 66° incubation
is 2° higher than that of the duplex
formed at 60°. This difference is to
be expected because there should be
a class of imprecisely paired se-
quences that can reassociate at 60°
but are unable to reassociate at the
more stringent 66° incubation tem-
perature. Normalized binding per-
centages obtained from the 60° and
66° reactions confirm the existence
of this class of imprecise relatives
(Table 9). When the binding per-
centage obtained with E. coli*/E. coli
on hydroxyapatite is arbitrarily des-
ignated 100% E. coli* and S. typhi-
murium exhibit 42% complementarity
at 60° and 35% complementarity at
66°. Under these conditions 7% of
the paired nucleotide sequences
formed at 60° are absent at 66°. An
8% difference occurs in the binding
of E. coli* fragments to S. typhimu-
rium DNA-agar in 60° and 66° incu-
bations. The good agreement between
36
32
c
<a
E
D)
D
!_
<
z
Q
C
D
O
C
o
_
28
24
20
16
12
E coli*/ S. typhi murium
66°(Tm = 75.5°)
E. coli*/ S. typhimurium
-60°(Tm = 73.5°)
E. coli VE. coli 60 c
83°)
60
8 —
80 90
Temperature, °C
Fig. 65. Thermal elution profiles of E. co//*/E. coli and E. co//*/S. typhimurium DNA reac-
tions. Samples of P32 labeled E. coli DNA fragments were incubated with unlabeled E. coli or
S. typhimurium DNA fragments at 60° or 66°. The samples were then passed through hydroxy-
apatite and the bound fragments were removed by a series of thermal elutions.
110 CARNEGIE INSTITUTION
TABLE 9. Extent of Duplex Formation between E. coli and S. typhimurium at 60° and 66°
Hydroxyapatite, %
(Total C0f, 43) DNA-Agar,' %
60° 66° 60° 66°
E. coli* E. coli binding 78 76 23-30 18-23
E. coli* S. typhimurium binding
78
76
33
27
42
35
Normalized E. coli* S. typhimurium binding 42 35 34 26
[E. coli/E. coli = 100 per cent)
1 Taken from the results of Brenner, Martin, and Hoyer.54
Four 0.4-ag samples of P32-labeled £. coli DNA fragments were denatured and quickly cooled.
Two samples were mixed with denatured, unlabeled E. coli fragments in 1 ml of 0.14 M PB and
two samples were mixed with denatured, unlabeled S. typhimurium fragments in 1 ml of 0.14
M PB. The unlabeled fragments were at approximate concentrations of 1 mg. One member of
each pair of samples was incubated at 60° for 4 hours and the other member of each pair was
incubated at 66° for 4 hours. Subsequent treatment of the samples was identical to that de-
scribed in Table 8.
the agar and hydroxyapatite data, tions will occupy a part of our efforts
despite the differences in absolute in the coming year,
binding percentages is of general sig- Phage-host and phage-phage sig-
nificance because data obtained by terns. Specific profiles are obtained
either of these methods may now be following thermal elution of phage
compared with greater confidence. and bacterial DNA-agar duplexes
Having established that the major- (Year Book 65, p. 106). Further-
ity of nucleotide sequences held in more, the elution peaks and Tm values
common between E. coli and S. typhi- observed in thermal elution profiles
murium are imprecisely paired, we from DNA-agar duplexes closely
may now consider several questions mimic the hyperchromicity and Tm
concerning relatedness among bacte- values obtained when DNAs in free
ria: solution are thermally denatured and
1. Can one determine the percent- assayed optically (see Cowie and
age (if any) of precisely paired E. Falkow, preceding sections) .
coli/S. typhimurium nucleotide se- Cowie and Szafranski (Year Book
quences? 65, p. 106), and Cowie and Falkow
2. Will precision of sequence com- (preceding sections) have excellent
plementarity be higher in DNA du- evidence for precise or very nearly
plexes formed between more closely precise binding between the DNA se-
related bacteria, e.g., Shigella and E. quences of phage and their bacterial
colli whose DNAs show 70% binding, hosts and also between the DNAs of
or among some Proteus species whose different temperate phages. In fact
DNAs bind as high as 90% ? the DNA-agar elution profiles ob-
3. Conversely, will less precise tained from the supposedly distant
DNA duplexes be formed between bacteriophage pairs A and </>80 or A
bacteria that are less closely related and P22 showed about 20% precise
than E. coli and S. typhimurium? binding and an apparent absence of
Answering these and related ques- imprecise binding. Only precise pair-
DEPARTMENT OF TERRESTRIAL MAGNETISM 111
ing was observed in the elution pro- relationships between the DNAs of
files from A or </>80 with E. coli DNA- these phages and their E. coli host
agar. These experiments with A, ^80 were investigated. Lambda, 434, and
and P22 employed incubation temper- 434 hy were selected for test because
atures of 60° in 2 X SSC. The agar they are closely related on the bases
was washed in SSC/30 at 50°, which of recombination and susceptibility to
is equivalent to approximately 70° in anti-A antiserum.
2 X SSC. Therefore the DNA frag- The extension of Bolton and Mc-
ments which bound in 2 X SSC and Carthy's DNA-agar technique that
remained stable between 60° and 70° employs thermal elution has been pre-
were not included in the elution pro- viously described (Year Book 65, p.
files. This washing procedure is quite 106) ; however, sufficient modifica-
stringent, thus giving increased ere- tions warrant a restatement of the
dence to the precise reassociation techniques.1
that occurs between phage, and be- Phage-bacterial reactions. Elution
tween phage and bacterial DNAs. profiles from 4M*/E. coli and 434
On the other hand, distantly related hy*/E. coli DNA reactions are shown
duplexes with poor complementarity in Figs. 66 and 67. The similarity of
would not be detected with this wash- these profiles indicates that the nu-
ing procedure. In the present experi- cleotide sequences in 434 and 434 hy
ments the SSC/30 washes were capable of pairing with E. coli DNA
carried out at 30° in order to allow are either very precise or at least
resolution at the low temperature end have the same average G + C dis-
of the elution profile. tribution. The precise complementar-
Quantitative and qualitative DNA ity indicated by elution profiles ob-
relationships in more closely related tained from reactions between these
temperate coliphages as well as the phage DNAs and the DNA of their
1 Preparation of phage DNA. Cultures of RNase and pronase, phenol extracted, etha-
E. coli K12 lysogenic for X, 434, or 434 hy nol precipitated, and finally resuspended in
were grown to late logarithmic phase in SSC/100. Specific activities ranging from
nutrient broth, sedimented, and resuspended 5 x 104 to 1 X 106 cpm//ig DNA have been
in 0.01 M MgS04, induced to produce phage obtained in various preparations,
by ultraviolet light, diluted 1 to 2 in fresh Incubation and elution conditions. Two
nutrient broth, and incubated at 37° in a tenths gram of agar containing 15-80 fig
water bath shaker until lysis was complete. of DNA was incubated with 0.2 ml of de-
Labeled phage lysates were obtained by natured P32-labeled DNA fragments in
resuspending the lysogenic bacteria in a 2 x SSC at a ratio of DNA in agar to
tris-glucose buffer lacking phosphate, incu- DNA fragments ranging from 500/1 to
bating 20 minutes at 37°, then adding P32 100,000/1 for 15-16 hours at 60°. The DNA-
and allowing lysis to proceed. Chloroform agar was then transferred to a tube fitted
was added to the lysates, which were then with a saran-screen bottom and washed with
freed of bacterial debris by differential seven 15-ml portions of 2 x SSC at 60°,
centrifugation. The resulting phage pellets followed by nine 15-ml portions of SSC/30
were exposed to DNase and RNase, then at 30°. Each wash required 5 minutes and
extracted with phenol. Additional phenol the DNA-agar was agitated several times
and ether extractions followed by ethanol during each wash to ensure efficient removal
precipitation were carried out to purify the of unhybridized DNA fragments. At this
DNA further. When pure, the DNA, re- point the DNA-agar received one 10-ml
suspended in SSC/100, was thermally de- SSC/30 wash of 5 minutes duration at 2°
natured and immobilized in agar at a final increments up to 80°. All wash fluids were
concentration of 75-400 fig/g agar. The measured for radioactivity. The percentage
labeled DNA was similarly purified, then of bound DNA fragments was calculated as
sheared at 12,000 psi in a French pressure total counts/bound counts x 100.
cell. The fragments were then exposed to
112
CARNEGIE INSTITUTION
bacterial host, as compared to the
intraspecies phage and E. coli*/E.
coli DXA profiles, is evident, and
similar to the precision indicated by
a A* E. coli DXA elution profile.
An additional feature of these
phage ^/bacterial DNA elution pro-
files is of interest. Structure is pres-
ent in the low-temperature end of
these profiles that is not seen in re-
actions involving identical DNAs.
The low temperature at which these
M
12
* 10
£
o
i_
<
Z
Q
-o
c
z>
o
-£>
rS 6
E.colWE.coh o
/
434hy*/434hy /
434hy*/Eco//
40
50 60
Temperature, °C
Fig. 66. Thermal elution profiles characteristic of reactions between E. coli DNA fragments
and E. coli DNA-agar (open circles, clashed curve); 434 DNA fragments, and 434 DNA-agar
(closed circles, solid curve); and 434 DNA fragments and E. coli DNA-agar (X, solid curve).
The data from these and all succeeding reactions have been converted to percentages of bound
fragments in order to make direct comparisons between experiments.
DEPARTMENT OF TERRESTRIAL MAGNETISM
113
nucleotide sequences are eluted is approximation of relatedness be-
indicative of mismatched sequences, tween the DNAs of phages A, 434,
This point is being more closely in- and 434 hy. The data in Table 10 in-
vestigated, dicate that \ is much more closely
Phage-phage reactions. Binding related to 434 hy and 434 than to </>80
percentages were obtained as a first or P22. Specificity of the method is
14
12
10
c
o
E
<
z
Q
-a
c
O
O
c
° 6
Ecoli*/Ecoti
40
50 60
Temperature, °C
Fig. 67. Thermal elution profiles characteristic of reactions between E. coli DNA fragments
and E. coli DNA-agar (open circles, dashed curve); 434 hy DNA fragments, and 434 hy DNA-
agar (closed circles, solid curve) and 434 hy DNA fragments and E. coli DNA-agar (X, solid
curve).
114
CARNEGIE INSTITUTION
TABLE 1 0. Percentage of DNA Binding among Temperate Coliphages
Labeled DNA
Fragments
Aga r-l mmobilized DNA
434 hy 434 080
'22
Phage 15
(TAU)- T4
Plectonema
boryanum
A
69
67
22
20
0.1
0
0.4
434
62
71
0.9
' The data for c/>80, P22/ and T4 are from Cowie, Year Book 65, pp. 106—123.
14
A*/A —ft
50 60
Temperature, °C
80
Fig. 68. Elution profiles obtained from intraspecies temperate coliphage DNA-agar reac-
tions: A*/A (open circles, dashed curve); 434 hy*/434 hy (closed circles, solid curve); 434*/
434 (X, solid curve).
DEPARTMENT OF TERRESTRIAL MAGNETISM
115
shown by the fact that neither A nor
434 DNA can pair with DNA from
the blue-green alga Plectonema bor-
yanum. Lambda DNA cannot pair
with DNA from phage 15(TAU)_ or
from T4. Direct quantitative compar-
isons of the A, 434, and 434 hy data
were precluded because of differences
in ratios of labeled DNA fragments
to DNA in the agar. It is clear, how-
ever, that A and 434 hy DNA bind to
a large extent as would be expected,
since Kaiser and Jacob55 isolated 434
hy after a series of crosses designed
to exclude all but l%-2% of the 434
genome. The close relatedness be-
tween A and 434 is borne out by the
high incidence of pairing between the
DNAs of these phages.
Elution profiles from 434*/434 and
434 hy*/434 hy DNA reactions are
seen in Fig. 68. The A*/A DNA pro-
file shown for comparison was ob-
tained by using the more restrictive
Temperature, °C
Fig. 69. Reproducibility of thermal elution profiles obtained from reactions between 434
DNA fragments and 434 DNA-agar. The three experiments were carried out with two different
DNA-agar preparations and two different labeled DNA fragment preparations.
116
CARNEGIE INSTITUTION
elution schedule starting: at 50°. The
434 hy and 434 DNA profiles are
quite similar; both of these DNAs
elute at a higher temperature than A
DXA. There is little or no fragment
elution at low temperatures, indicat-
ing that under these incubation con-
ditions there is no imprecise multiple
copy DXA present in the genome of
these phages. There is in fact no in-
dependent evidence for the existence
of imprecise multiple copy DNA in
a phage genome.
Phage 434 DNA has approximately
51.5 c'c G + C as determined from
Falkow's detailed optical analysis of
thermal denaturation data. Since 434
hy DXA has a higher maximum elu-
tion temperature than that of A, it is
possible that 434 hy has either picked
up some high G + C DNA from its
434 parent or has had some low G +
C A DNA deleted (assuming its ge-
nome size is comparable to the 3 X
10: Dalton genome size of A DNA) .
Sensitivity and reproducibility of
elution profiles from DNA-agar du-
plexes are dependent, to a large de-
gree, upon the thoroughness of wash-
ing at each temperature increment
and upon the accuracy with which
temperature is controlled. The repro-
ducibility observed in these experi-
ments is shown in Fig. 69. The major
434*/434 DNA peaks are quite re-
producible; however, the peak at 64°
is distinct in one curve, is present as
a shoulder in a second, and has been
obscured in a third. Reproducibility
in these experiments appears to com-
pare favorably with that of detailed
optical determinations of DNA melt-
ing curves (Fig. 53).
Interspecies phage reactions give
no indication of imprecise DNA pair-
ing in either the main areas of re-
action or in DNA that elutes at low
temperatures. The A*/434 hy DNA
profile (Fig. 70<x) shows a distinct
A-like peak at 70° and a 74° peak
characteristic of 434 hy DNA. A 434
hy*/434 hy DNA reaction is pre-
40
50 60 70 80 40
50 60 70
Temperature, °C
80 40
50 60 70
Fig. 70. Thermal elution profiles characteristic of interspecies phage DNA-agar reactions,
a. 434 hy DNA-agar reacted with DNA fragments (open circles, dashed curve) and with 434 hy
DNA fragments (closed circles, solid curve), b. 434 DNA-agar reacted with A DNA fragments
(open circles, dashed curve) and with 434 DNA fragments (closed circles, solid curve), c. 434
DNA fragments and 434 hy DNA-agar (open circles, dashed curve); 434 hy DNA fragments and
434 DNA-agar (closed circles, solid curve).
DEPARTMENT OF TERRESTRIAL MAGNETISM
117
sented for comparison. Figure 706
shows a A*/434 DNA profile which
also contains the characteristic 70°
A-like peak. It is compared with a
434 */434 DNA profile. The reciprocal
434 hy*/434 and 434/434 hy DNA
profiles that peak at 72° are shown
in Fig. 70c. The temperatures at
which major elution peaks were ob-
tained in the reactions studied thus
far are summarized in Table 11.
A more sensitive means of deter-
mining the precision of nucleotide
sequence pairing is to reisolate the
duplexed fragments from an inter-
species phage reaction and react them
in the intraspecies system. The exper-
iment was carried out with labeled
14
12
10
c
<u
E
o>
o
<
Z
Q
-O
c
D
o
434 *e luted from
434 hy agar / 434
434^434.
i/rr-y^
40
50 60
Temperature, °C
80
Fig. 71. Thermal elution profiles of a typical intraspecies 434 DNA-agar reaction and an
intraspecies 434 DNA-agar reaction in which the 434 DNA fragments had been previously bound
to and reisolated from 434 hy DNA-agar.
118
CARNEGIE INSTITUTION
fragments eluted from a 434*/434 hy
DXA reaction. The fragments were
ethanol precipitated, concentrated,
thermally denatured, and then incu-
bated with 434 DNA agar. The re-
sulting elution profile, shown in Fig.
71, closely parallels the profile ob-
tained from elution of phage 434
DXA duplexes. Therefore the preci-
sion of the reassociated complemen-
tary nucleotide sequences between
434 and 434 hy mimics that of the
reassociated 434 sequences.
It should be noted that a small de-
gree of mismatched base pairs in nu-
cleotide strands would not be distin-
guishable from the data presented in
thermal elution profiles. The high de-
gree of pairing between X and 434
DNA should have made this an ideal
system in which to detect divergence
among nucleotide sequences. It is
clear, however, that no great impre-
cisions exist within the duplexed
portion of X and 434 DNA.
The data describing the precise
complementarity in interspecies du-
plexes may be summarized:
1. E. coli*/S. typhimurium DNA
duplexes contain mainly or com-
pletely imprecise pairs.
2. Phage-host DNA reactions ex-
hibit mainly precisely paired nucleo-
tide sequences; however, there are
imprecisely paired sequences, as evi-
denced by those pairs that dissociate
at low temperatures.
3. The interspecies phage DNA re-
actions examined to date showed no
evidence of imprecise nucleotide se-
quence pairing.
MEMORY AND LEARNING
MECHANISMS
A. V. Rake and R. B. Roberts
For the past 10 years the Biophys-
ics Section has maintained close con-
tact with the studies of Dr. L. B.
Flexner and Dr. J. B. Flexner, car-
ried out at the University of Pennsyl-
vania. During this period they have
determined rates of protein synthesis
in newborn and adult mice and more
recently have uncovered the remark-
able effect of puromycin in blocking
the memory of maze training. During
the past year this continuing collabo-
ration has been supplemented by the-
oretical and experimental studies car-
ried out at this laboratory.
Theoretical. The arrival of the IBM
1130 computer evoked the writing of
a program that might simulate the
behavior of a network of 100 inter-
connected neurones. The object of this
exercise was to begin with very simple
rules of connection and add elabora-
tions as required to approach perfor-
mance simulating adaptive behavior.
In brief, the computer reads the
initial conditions from punched cards
and then proceeds to compute the
state of the network at one instant
from the external signals applied and
the state of the network at the pre-
vious instant (Fig. 72). If a cell has
fired on the previous cycle it is con-
TABLE 1 1 . Thermal Elution Peaks from Phage DNA-Agar Reactions
Labeled
DNA-Ac
jar
Fragments
X
434 hy
434
(£80
P22
E. coli
A
64, 70
70, 74
72
58,74
68
66, 70
434 hy
.
70, 74
72
. . .
. . .
66,70
434
72
72
.
. . .
66,70
£80
58, 74
. . .
. . .
58,
70,74
76
P22
60,
67,74
62,
68-70
DEPARTMENT OF TERRESTRIAL MAGNETISM
119
0 0 0 0 0
0 0 0
0
0
0 0 0 0 0
0 0 0
0
0
0 0 0 0 0
0 0 0
0
0
0 0 0 0 0
0 0 0
0
0
0 0 0 0 0
0 0 0
0
0
0 0 0 0 0
0 0 0
0
0
0 0 0 0 0
0 0 0
0
0
0 0 0 0 0
0 0 0
0
0
0 0 0 0 0
0 0 0
0
0
0 0 0 0 0
0 0 0
0
0
0
0
0
0
0
0
0
0 0 0
0
0
0
0
0
0
0
0 0 0
0
0
0
0
0
0
0
0 0 0
2
0
0
0
0
0
0
0 0 0
2
0
0
0
0
0
0
0 0 0
0
0
0
0
0
0
0
0 0 0
0
0
0
0
0
0
0
0 0 0
0
0
0
0
0
0
0
0 0 0
0
0
0
0
0
0
0
0 0 0
0
0
0
0
0
0
0
0 0 0
2
0
0
0
0
0
0 0 0 0
0
0
0
0
0
0
0 0 0 0
0
1
0
0
0
0
0 0 0 0
0
1
1
0
0
0
0 0 0 0
0
1
1
0
0
0
0 0 0 0
0
1
0
0
0
0
0 0 0 0
0
0
0
0
0
0
0 0 0 0
0
0
0
0
0
0
0 0 0 0
0
0
0
0
0
0
0 0 0 0
0
0
0
0
0
0
0 0 0 0
CYCLE 0
CYCLE 1
CYCLE 2
Fig. 72. Initially all the cells of the network are considered inactive as indicated in cycle 0.
On the first cycle the computer reads a card indicating what input stimulation has been ap-
plied; the cells receiving inputs are considered firing and indicated by the numeral 2. On the
second and succeeding cycles the computer reads additional cards indicating further stimulation,
and calculates which cells fire as a result of signals transmitted to them by cells fired on the
previous cycle. These are shown by the 1 in cycle 2. After 10 cycles the results are printed out.
sidered refractory and will not fire
on the present cycle; if a cell receives
an external input (as read from a
card) it fires unless it is refractory;
if neither of these conditions applies,
the signal delivered by each of the
cells that did fire on the previous
cycle is calculated and summed. If
this sum exceeds a designated firing
level the cell is noted as fired and the
computer proceeds to the next cell.
After completing the calculations for
10 cycles the results are printed out.
The key element of the program
lies in the rules determining the in-
terconnections of one cell and an-
other, i.e., the synapses. With simple
geometrical rules (Fig. 73) the net
is highly unstable. An external input
dies away rapidly if the firing level
is high or expands to fill the net when
the firing level is below a critical
level. This inherent instability was
controlled by introducing a variable
firing level (increasing if too many
cells fire) and by adding inhibitor
cells. With these additional features
in the program the propagation of
diffuse waves through the network
could be observed. Such a variable
firing level might occur in the brain
if excessive firing reduced the avail-
able energy supply or released inhib-
itors. Furthermore, the fluctuating
sensitivity due to oscillations of the
control system might possibly be re-
lated to the a waves of the brain.
0 I 00000000
0
0 11111110
|0| 12222210
oz
301 12333210
00 1 123x32101
u0 1 23 3 3 2 1 0 '
Oh-
=>0 J 1 222 2 2 1 0
Z0 | 1 1 1 1 1 1 1 0
oS
1—
0=)
"~0 1 00000 0 0 0
ou
ojoooooooo
0
Fig. 73. A simple geometrical rule for inter-
cell connections. A cell (indicated by X) is
connected to each of its nearest neighbors by
a strength of 3, to its next nearest neighbors
by a strength of 2, etc. The column of cells at
the left is considered reserved for input signals
and does not receive signals from other cells of
the network. The output column does not de-
liver signals to other cells of the net. This iso-
tropic network is highly unstable and confusing
because it allows propagation in all directions.
120
CARNEGIE INSTITUTION
These diffuse waves of activity
might give some primitive organism
a repertoire of different response pat-
terns corresponding to different stim-
uli and might be representative of
one characteristic feature of the
brain. Another, possibly more inter-
esting, property is propagation along
more clearly defined channels. These
channels suggest strongly that cer-
tain sequences of connections are
much more effective than others. In
adding such a feature to the program
a choice must be made whether to
specify the strength of each inter-
connection or to specify properties of
the cells which then define the
strength of the connections.
YVe have chosen the second alter-
native from considerations of inher-
ited behavior patterns. Certain se-
quences of synapses appear to be
effective and these are designated by
the genes of the organism. As the
genes' best understood action is the
determination of the distribution
(i.e., quantity and types) of proteins
within a cell, we have postulated that
synapses are effective when they have
similar proteins on both faces. Thus
the DNA would not need to carry
information concerning each synapse
but only enough to designate the pro-
tein distribution. In applying this
principle to the computer, each cell
is assigned a complement of proteins
as shown in Fig. 74, and the strength
of the synapse between two cells is
made proportional to the number of
similar proteins that the two cells
hold in common. Thus the distribu-
tion in the figure gives strong chan-
nels horizontally as these cells have
80 units in common.
The same basic mechanism allows
the introduction of learning into the
program. A modification is made
whereby the connections are made
stronger by use. If one cell fires and
a second fires in response, then on
CELL
PROTEIN NUMBER
NUMBER
1
2
3
4 5 6 7
8
9
10
1
100
0
0
0 0 0 0
0
0
0
2
20
80
0
0 0 0 0
0
0
0
3
20
0
80
0 0 0 0
0
0
0
4
20
0
0
80 0 0 0
0
0
0
5
20
0
0
0 80 0 0
0
0
0
6
20
0
0
0 0 80 0
0
0
0
7
20
0
0
0 0 0 80
0
0
0
8
20
0
0
0 0 0 0
80
0
0
9
20
0
0
0 0 0 0
0
80
0
10
20
0
0
0 0 0 0
0
0
80
11
80
20
0
0 0 0 0
0
0
0
12
0
100
0
0 0 0 0
0
0
0
13
0
20
80
0 0 0 0
0
0
0
14
0
20
0
80 0 0 0
0
0
0
Fig. 74. Each cell is initially assigned a certain complement of proteins. The strength of the
connection from one cell to another is then measured by the number of protein units held in
common. The distribution illustrated here provides strong channels for propagation horizontally,
as all cells lying in one row (for example, cells 2, 1 2, 22) have 80 units in common. Cells in
the same column (1, 2, 3, 4, etc.) have 20 units in common, giving a weaker channel in the ver-
tical direction. The distribution changes as a result of firing as described in the text.
DEPARTMENT OF TERRESTRIAL MAGNETISM
121
the next cycle a fraction of the pro-
tein complement of the firing* cell is
transferred to the responding cell.
The protein distributions thus become
more similar and the connection is
accordingly strengthened.
With these features in operation
almost any desired channels of prop-
agation can be developed in the net-
work by a short period of "training."
Reverberating rings are of particular
interest, as they allow a single stimu-
lus to initiate a prolonged response.
In addition, the formation of a con-
ditioned reflex has been simulated by
firing first one and then another of
two adjacent channels.
The operation of the network to
date suggests other developments,
first, in a better simulation of learn-
ing and memory and, second, in mov-
ing further toward adaptive behavior.
Connections between, cells in which
the strength is determined by the
number of proteins in common at the
two faces of the synapse provide a
satisfying simulation of inherited be-
havior and possibly of long-term mem-
ory. The stability of such a connec-
tion does not depend on the stability
of any particular molecule but only
on the stability of the state of differ-
entiation of the cells. Presumably this
stability in turn depends on the sta-
bility of a self-sustaining system.
Furthermore, a slight modification of
this concept is adequate to include
the behavior patterns that are evoked
by the administration of hormones.
Another class of proteins can be
added to the cell with the proviso
that common proteins of this class
will cause a strong connection be-
tween the cells if and only if a hor-
mone (or its products) is present.
Thus the DNA could designate a se-
quence of synapses that were inactive
until made effective by a hormone.
The device needed to simulate
evoked behavior suggests a similar
mechanism capable of simulating the
short-term learning (and memory)
that can occur during inhibition of
protein and RNA synthesis. Let us
postulate a third class of proteins
that requires a factor X if the pro-
teins are to facilitate a synapse. X is
assumed to be present at all times
but enters the synapse only when the
synapse has transmitted a pulse. The
short duration of such a memory im-
plies that X has a short half-life in
the synapse. Furthermore, the en-
hanced ability to learn "significant"
events can be attributed to an in-
creased concentration of X released
as a consequence of events inter-
preted as "significant." Unfortu-
nately, this type of memory is difficult
to include in the program, as it re-
quires the computer to store data
describing the state of each individual
synapse.
A different direction of evolution
would lead toward adaptive behavior.
With the present program the net-
work learns to repeat whatever it has
done whether the outcome is good or
bad. The addition of a "pain center"
could serve to inhibit activity and
thus prevent any further learning of
inappropriate responses.
If the outcome is not evident until
the action is complete, adaptive be-
havior is more difficult to achieve. As
a minimum it seems necessary to
store the stimulus and the response
taken until the outcome becomes
known. At that time favorable stimu-
lus-response patterns would be trans-
ferred to a long-term memory and
the unfavorable ones discarded or in-
hibited. Such a mechanism suggests
the need for short- and long-term
memories in the brain but is difficult
to incorporate in a small program.
The principal result of these exer-
cises with the computer has been a
sharpening of our concepts of mem-
ory mechanisms. Farley has sum-
marized this value: "To use the
computer is to require precision of
122
CARNEGIE INSTITUTION
statement." We concur, and add a
corollary that writing the program is
sufficient for most purposes; its ac-
tual execution on the computer fre-
quently adds little.
Experimental work. The objective
of our experimental work has been
to supplement the investigations of
the Flexners. Several years ago the
assumption was made that the long-
term stability of a well-established
memory might be similar to the sta-
bility of the pattern of protein syn-
thesis in a differentiated cell. A
learning experience might trigger the
synthesis of new proteins and RNA.
If these products included an inducer
(or derepressor) then the synthesis
of the new products would be self-
sustaining and the cell would assume
a slightly altered state of differentia-
tion. Early experiments, in which the
Flexners observed that injections of
puromycin (an inhibitor of protein
synthesis) blocked the expression of
memory, were consistent with this
hypothesis. Furthermore, increased
synthesis of RNA after a learning
experience was reported by others.
Accordingly, we hoped to observe the
properties of this RNA by using the
hybridizing techniques that had
proved so useful in other studies.
As an initial step the RNA of the
brain was to be compared with RNA
of other organs. Exploratory experi-
ments carried out several years ago
had indicated that the mRNA of the
brain had little in common with liver
mRNA. If these experiments were
promising then RNAs from different
regions of the brain and RNAs pro-
duced after a learning experience
could be characterized.
Unfortunately, these efforts have
encountered a number of difficulties.
RNA precursors (including formate,
orotic acid, and uridine) when in-
jected intraperitoneally (i.p.) contrib-
ute very little to the RNA of the brain.
The resulting radioactivity is barely
sufficient as a measure of RNA syn-
thesis and far below the level needed
for hybridization experiments. For
this reason most experimenters have
used intracranial injections (i.e.) of
the RNA precursors. With this pro-
cedure the resulting specific radioac-
tivities are higher but still marginal
for use in hybridization experiments.
An alternative technique (of lim-
ited application) is to use unlabeled
RNA from the brain as a competitor
to labeled RNA from other sources.
As shown in Fig. 75 competition does
occur, indicating that brain RNA
shares common sequences with RNA
of liver. The fraction of RNA that
does hybridize in experiments is al-
ways low and comprises not a ran-
dom sample of the RNA but a special
fraction. Therefore newly formed
RNA made as a result of learning
would be detectable only if it hap-
pened to fall in this restricted class.
In summary, these experiences with
hybridization give no encouragement
to the belief that this technique will
be of much value in studies of the
formation of memory.
In the course of making numerous
i.e. injections of radioactive uridine
we noted that the region labeled was
critically dependent on the depth of
injection (Table 12) . It seemed there-
fore that measurements of RNA syn-
thesis during a learning experience
could be subject to error if the in-
jection were made i.e. Furthermore,
it did not seem that the mice which
had just undergone such a drastic
experience could be considered "nor-
mal" in their reaction to the training.
Other labeling procedures will have
to be investigated.
In the year since these experiments
started several findings have altered
our views of the molecular events in-
volved in memory formation. Mice
can learn and remember (for a short
time at least) during inhibition of
protein and RNA synthesis by puro-
DEPARTMENT OF TERRESTRIAL MAGNETISM
123
TABLE 12. Incorporation of C14 Uridine, Injected Intracranial^, into Mouse Brain Subfractions
2 mm
4 mm
First
Second
First
Second
Specimen
Specimen
Specimen
Specimen
2.9
2.7
85.0
89.2
1.8
1.2
76.6
59.0
1.2
3.3
71.8
83.4
10.9
7.1
137.5
124.0
9.4
6.9
152.0
121.0
2.0
3.9
135.5
101.2
248.0
211.0
34.2
26.0
164.5
109.5
21.3
33.5
28.5
8.5
24.1
34.1
Nerve cord
Pons
Cerebellum
Hippocampus
Thalamus
Corpus striatum
Frontal cortex
Temporal cortex
Parietal cortex
In this experiment 1 0 A of C14 uridine, 2 /xc, was injected bilaterally at the indicated depth
beneath the surface of the frontal lobe of the brain. Ninety minutes later the brain was collected,
the tissue dissected, pooled when bilateral, weighed, and frozen. The fragments were extracted
by the usual Schmidt-Thannhauser method and the radioactivity in the nucleic acid fraction
determined. The results are expressed as cpm/mg of wet tissue for two animals.
100
300 500
Competitor RNA
Fig. 75. Competition of unlabeled normal mouse brain RNA with P32 pulse-labeled liver RNA.
80y of 1.5 hours labeled P32 liver RNA (2000 cts/min/O.D.26o) was incubated with 0.25 g of
agar containing 1 OOy mouse DNA for 18 hours at 60°C in the presence of unlabeled RNA from
normal mouse brain (two preparations) and with unlabeled liver and kidney RNA for comparison.
The relative radioactivity bound to the DNA was determined as percentage compared to no com-
peting unlabeled RNA. The P32 bound to agar without DNA is also indicated.
124
CARNEGIE INSTITUTION
myein and actinomycin D. Acetoxy-
cycloheximide, a stronger inhibitor
of protein synthesis than puromycin,
differs from puromycin, as it does
not block memory and in fact pre-
vents blocking by puromycin when
both are injected together. The action
of puromycin is not to erase a mem-
ory as was originally thought, but to
block its expression ; the memory can
be restored by a subsequent injection
of saline. For short-term memory at
least, synthesis of the macromolecules
does not seem to be required. Fur-
thermore, other reports have indi-
cated drastic effects of hormones on
the retention of memory. Possibly the
search for the molecules involved in
memory formation should be shifted
toward peptides.
In summary, the past year has
given us some first-hand experience
in a new field and we can only hope
that the next year will prove more
fruitful. If new molecules are pro-
duced by numerous cells of the brain
as a result of learning, they should
be detectable; if only a few cells are
involved or if existing materials are
simply released, the search will be
very much more difficult.
REFERENCES CITED
1. Forbush, S. E., and L. Beach, Carnegie
Inst. Wash. Year Book 65, 28-36, 1966.
2. Babcock, H. W., Sci. Am., 202, 53-62,
1960.
3. Ness, N. F., and J. M. Wilcox, Phys.
Rev. Letters, 13, 461-464, 1964.
4. Parker, E. N., Astrophys. J., 128, 664-
676, 1958.
5. Wilcox, J. M., and N. F. Ness, J.
Geophys. Res., 70, 5793-5805, 1965.
6. Venkatesan, D., and A. Dattner, Tellus,
11, 116-129, 1959.
7. McCracken, K. G., V. R. Rao, B. C.
Fowler, M. A. Shea, and D. F. Smart,
IQSY Instruction Manual No. 10, 1965.
8. Herrin, E. T., and J. Taggart, Bull.
Seis. Soc. Am., 52, 1037-1046, 1962.
9. Tatel, H. E., L. H. Adams, and M. A.
Tuve, Proc. Am. Phil. Soc., 97, 658-669,
1953.
10. Hart, P. J., Variation of velocity near
the Mohorovicic discontinuity under
Maryland and northeastern Virginia,
Ph. D. thesis, Harvard University, 1954.
11. Press, F., J. Geophys. Res., 65, 1039-
1051, 1960.
12. Press, F., and S. Biehler, J. Geophys.
Res., 69, 2979-2995, 1964.
13. Smith, T. J., J. S. Steinhart, and L. T.
Aldrich, ./. Geophys. Res., 71, 1141-
1181, 1966.
14. Steinhart, J. S., and R. P. Meyer,
Carnegie Inst. Wash. Publ. 622, 1961.
15. Carnegie Inst. Wash. Year Book 62,
280, 1963.
16. Borcherdt, R. D., and J. C. Roller,
Technical Letter No. 43, U. S. Geologi-
cal Survey, 1966.
17. Lewis, B. T. R., and R. P. Meyer, Contr.
167, Geophys. and Polar Res. Center,
Dept. of Geo!., Univ. of Wise, 1966.
18. Knopoff, L., Publ. 383, Inst. Geophys.
and Planetary Phys., Univ. of Calif.,
Los Angeles, 1966.
19. Meyer, R. P., J. S. Steinhart, W. E.
Bonini, B. F. Howell, Jr., D. E. Willis,
and F. Novacheck, in Abstracts of
Papers, vol. Ill, International Associa-
tion of Seismology and Physics of the
Earth's Interior (27), International
Union of Geodesy and Geophysics, 13th
Gen. Assembly, Berkeley, Calif., 1963.
20. O'Brien, P. N. S., Boll. Geofisica T. ed
A., 9, 22-65, 1967.
21. Smith, T. J., J. S. Steinhart, and L. T.
Aldrich, J. Geophys. Res., 71, 1141-
1172, 1966.
22. Berry, M. J., and G. F. West, Bull Seis.
Soc. Am., 56, 141-171, 1966.
23. Bancroft, A. M., Seismic spectra and
detection probabilities from explosions
in Lake Superior, in The Earth Beneath
the Continents, J. S. Steinhart and T. J.
DEPARTMENT OF TERRESTRIAL MAGNETISM
125
Smith, eds. Amer. Geophys. Union
monograph 10, 234-240, 1966.
24. O'Brien, P. N. S., J. Geophys. Res., 72,
2569-2575, 1967.
25. Kanamori, H., J. Geophys. Res., 72,
2181-2186, 1967.
26. Sumner, R. D., Bull. Sets. Soc. Am., 57,
173-190, 1967.
27. Anderson, D. L., A. Ben-Menahem, and
C. B. Archambeau, J. Geophys. Res., 70,
1441-1448, 1965.
28. McCamy, K., R. P. Meyer, and T. J.
Smith, Bull. Seis. Soc. Am., 52, 923-955,
1962.
29. Stueber, A. M., and V. R. Murthy,
Science, 153, 740-741, 1966.
30. Hart, S. R., Program annual meetings,
Geol. Soc. Am., 86-87, 1964.
31. Stockwell, C. H., Geol. Surv. Canada,
Paper 64-17, Part II, 1964.
32. Birch, F., Am. J. Sci., 252, 1-25, 1954.
33. Roy, R. F., Heat Flow Measurements
in the United States, Ph.D. Thesis,
Harvard Univ., 1963.
34. Lachenbruch, A. H., U. S. Geol. Surv.
Bull., 1083-A, 36 pp., 1959
35. Von Herzen, R., and A. E. Maxwell, J.
Geophys. Res., 64, 1557-1563, 1959.
36. Ratcliffe, E. H., J. Geophys. Res., 65,
1535-1541, 1960.
37. Burke, B. F., K. C. Turner, and M. A.
Tuve, A high-resolution study of M31,
in The Galaxy and the Magellanic
Clouds, IAU-URSI Symp. 20, 99-102,
1964.
38. Rubin, V. C, S. Moore, and F. C.
Bertiau, S.J., Astron. J., 72, 59-64,
1967.
39. Keyworth, G. A., G. C. Kyker, Jr., E. G.
Bilpuch, and H. W. Newson, Nucl.
Phys., 89, 590, 1966.
40. Christian, R. S., and J. L. Gammel,
Phys. Rev., 91, 100, 1953.
41. Extermann, P., Nucl. Phys., A95, 615,
1967.
42. Maio, J. J., and C. L. Schildkraut, J.
Mol. Biol, 24, 29, 1967.
43. Walker, P.M.B., and A. McLaren,
Nature, 206, 1175, 1965.
44. Flamm, W. G., M. McCallum, and
P. M. B. Walker, Proc. Natl. Acad. Sci.,
57, 1729-1734, 1967.
45. Martin, M., and B. H. Hoyer, Biochem.,
5, 2706, 1966.
46. McClaren, A., and P. M. B. Walker,
Genet. Res. Camb., 6, 1965.
47. Simpson, G. C, in This View of Life,
Harcourt, Brace and World Inc., N. Y.,
pp. 164-165, 1964.
48. Falkow, S., J. A. Wholhieter, R. V.
Citarella, and L. S. Baron, J. Bacteriol,
87, 209, 1964.
49. Falkow, S., R. V. Citarella, A. J. Whol-
hieter, and T. Watanabe, J. Mol. Biol.,
17, 102, 1966.
50. Signer, E. R., Virology, 22, 650, 1964.
51. Kaiser, A. D., and D. S. Hogness, J.
Mol. Biol., 2, 392, 1960.
52. Bolton, E. T., and B. J. McCarthy, Proc.
Natl. Acad. Sci., U.S., 48, 1390, 1962.
53. Martin, M. A., and B. H. Hoyer, Bio-
chem., 5, 2706, 1966.
54. Brenner, D. J., M. A. Martin, and B. H.
Hoyer, J. Bacteriol, 94, 486-487, 1967.
55. Kaiser, A. D., and F. Jacob, Virology,
4, 509, 1957.
BIBLIOGRAPHY
Adams, J., see Brown, L.
Aldrich, L. T., see Asada, T., Hart, S. R.,
and Smith, T. J.
Asada T., and L. T. Aldrich, Seismic obser-
vations of explosions in Montana, in The
Earth beneath the Continents, Geophysical
Monograph 10, 382-390, J. S. Steinhart
and T. J. Smith, eds., Wash., D. C, Amer-
ican Geophysical Union, 1966.
Asada, T., see also Hales, A. L.
Baum, W. A., J. S. Hall, L. L. Marton, and
M. A. Tuve, Committee on image tubes
for telescopes, Carnegie Inst. Wash. Year
Book 65, 191-194, 1966.
Bendich, A. J., and E. T. Bolton, Related-
ness among plants as measured by the
DNA-agar technique, Plant Physiol., 42,
959-967, 1967.
Bertiau, F. C, S.J., see Rubin, Vera C.
126
CARNEGIE INSTITUTION
Bolton, E. T., Nucleic acid interactions: a
molecular approach to the study of genes
and their products, Cancer Res., 26, 1964-
1970, 1966.
Bolton, E. T., see also Bendich, A. J.
Boyce, P. B., see Ford, W. K., Jr.
Brenner, D. J., S. Falkow, and D. B. Cowie,
DXA homologies among temperate coli-
phages (abstract), Science, 156, 535-536,
1967.
Britten, R. J., and D. E. Kohne, Repeated
nucleotide sequences of DNA (abstract),
Science, 156, 536, 1967.
Britten, R. J., sec also Waring, M. J.
Brown, L., W. Haeberli, and W. Trachslin,
Phase shifts and polarization in proton-
alpha scattering from 940 to 3200 keV, in
Proceedings of the 2nd International Sym-
posium on Polarization Phenomena of
Nuclcons, Karlsruhe, Germany, Sept. 6-
10, 1965, pp. 349-351, P. Huber and H.
Schopper, eds., Basel and Stuttgart, Birk-
hauser, Verlag, 1966.
Brown, L., W. Haeberli and W. Trachslin,
Phase shifts and polarization in proton-
alpha scattering from 0.9 to 3.2 MeV,
Nucl Plujs., A90, 339-352, 1967.
Brown, L., W. Trachslin, D. Robson, and
J. Adams, Polarization excitation of the
1.85-MeV resonances in 40Ar (p,p)40Ar
(abstract), Bull. Am. Phys. Soc., 11, 840,
1966.
Brown, L., and W. Trachslin, Polarization
excitation of the 1.87 MeV resonances in
40Ar(p,p) *oAr, Nucl. Phys., A96, 238-240,
1967.
Brown, L., and W. Trachslin, Scattering in
polarized protons by helium from 0.9 to
3.2 MeV, Nucl. Phys., A90, 334-338, 1967.
Brown, L., see also Trachslin, W.
Cowie, D. B., Genetic relationships among
viruses and bacteria (abstract), Science,
156, 537, 1967.
Cowie, D. B., see also Brenner, D. J.
Duggal, S. P., M. A. Pomerantz, and S. E.
Forbush, Long-term variation in the mag-
nitude of the diurnal anisotropy of cosmic
rays, Nature, 211*, 154-155, 1967.
Falkow, S., see Brenner, D. J.
Flexner, J. B., see Flexner, L. B.
Flexner, L. B., J. B. Flexner, and R. B.
Roberts, Memory in mice analyzed with
antibiotics, Science, 155, 1377-1383, 1967.
Flexner, L. B., J. B. Flexner, and R. B.
Roberts, Stages of memory in mice
treated with acetoxycycloheximide before
or immediately after learning, Proc. Natl.
Acad. Sci. U. S., 56, 730-735, 1966.
Forbush, S. E., Solar influences on the
cosmic-ray diurnal variation (abstract),
Science, 156, 537, 1967.
Forbush, S. E., Variations from 1937 to
1965 in the annual mean diurnal variation
of cosmic-ray intensity from ionization
chambers at three stations (abstract),
Trans. Am. Geophys. Union, 48, 171, 1967.
Forbush, S. E., see also Duggal, S. P.
Ford, W. K., Jr., Astronomical uses of cas-
cade intensifiers, in Advances in Elec-
tronics and Electron Physics, Photo-
Electronic Image Devices, vol. 22B, 697-
704, J. D. McGee, D. McMullen, and E.
Kalian, eds., New York, Academic Press,
1966.
Ford, W. K., Jr., and P. B. Boyce, Infrared
spectroscopy with a cascaded image in-
tensifier (abstract), Astron. J., 71, 385,
1966.
Ford, W. K., Jr., and Vera C. Rubin, Quasi-
stellar objects with small redshifts
1217+02, 3C 249.1, and 3C 263, Astro-
phys. J., U5, 357-360, 1966.
Ford, W. K., Jr., see also Rubin, Vera C.
Haeberli, W., see Brown, L.
Hales, A. L., and T. Asada, Crustal struc-
ture in coastal Alaska, in The Earth be-
neath the Continents, Geophysical Mono-
graph 10, 420-432, J. S. Steinhart and
T. J. Smith, eds., Wash., D. C, American
Geophysical Union, 1966.
Hall, J. S., see Baum, W. A.
Hart, S. R., Radiometric ages in Uruguay
and Argentina and their implications
concerning continental drift (abstract) ,
Geol. Soc. Am. Program, 79th Ann. Meet-
ing, San Francisco, Calif., p. 86, Nov.
14-16, 1966.
Hart, S. R., and L. T. Aldrich, Fractiona-
tion of potassium/rubidium by amphi-
boles: implications regarding mantle com-
position, Science, 155, 325-327, 1967.
Hart, S. R., J. S. Steinhart, and T. J.
Smith, A regional heat flow survey in
Lake Superior (abstract), Trans. Am.
Geophys. Union, 48, 211, 1967.
Hart, S. R., and G. R. Tilton, The isotope
geochemistry of strontium and lead in
Lake Superior sediments and water, in
The Earth beneath the Continents, Geo-
DEPARTMENT OF TERRESTRIAL MAGNETISM
127
physical Monograph 10, 127-138, J. S.
Steinhart and T. J. Smith, eds., Wash.,
D. C, American Geophysical Union, 1966.
Hart, S. R., see also Ohmoto, H.
Holland, H. D., see Ohmoto, H.
James, D. E., Crustal structure of the Mid-
dle Atlantic States (abstract), Earth-
quake Notes, Eastern Section Seis. Soc.
Am., 27, 16, 1966.
James, D. E., T. J. Smith, and J. S. Stein-
hart, Crustal structure of the Middle
Atlantic States (abstract), Trans. Am.
Geophys. Union, 48, 198, 1967.
James, D. E., and J. S. Steinhart, Structure
beneath continents: a critical review of
explosion studies 1960-1965, in The Earth
beneath the Continents, Geophysical
Monograph 10, 293-333, J. S. Steinhart
and T. J. Smith, eds., Wash., D. C,
American Geophysical Union, 1966.
Kohne, D. E., see Britten, R. J.
Kouvo, O., and G. R. Tilton, Mineral ages
from the Finnish Precambrian, J. Geol.,
74, 421-442, 1966.
Marton, L. L., see Baum, W. A.
Moore, Sandra, see Rubin, Vera C.
O'Brien, P. N. S., Analysis of a small num-
ber of seismic records along an E-W
Alpine profile, Boll. Geofis. Teorica ed
Appl, IX, 22-65, 1967.
O'Brien, P. N. S., Quantitative discussion on
seismic amplitudes produced by explosions
in Lake Superior, J. Geophys. Res., 72,
2569-2575, 1967.
Ohmoto, H., S. R. Hart, and H. D. Holland,
K-Ar and Rb-Sr ages of intrusive rocks
and hydrothermal minerals in the Provi-
dencia area, Mexico (abstract) , Geol. Soc.
Am. Program, 79th Ann. Meeting, San
Francisco, Calif., pp. 152-153, Nov. 14-16,
1966.
Otsuka, M., Analysis of P-wave travel-time
residuals in Peru (abstract), Earthquake
Notes, Eastern Section Seis. Soc. Am., 37,
24, 1966.
Pomerantz, M.A., see Duggal, S. P.
Roberts, R. B., see Flexner, L. B.
Robson, D., see Brown, L.
Rubin, Vera C, and W. K. Ford, Jr., Image
tube spectra of quasi-stellar objects (ab-
stract), Astron. J., 71, 396, 1966.
Bertiau, S. J., Faint blue objects in the
Virgo cluster region, Astron. J., 72, 59-
123, 1967.
Rubin, Vera C, see also Ford, W. K., Jr.
Sacks, I. S., A broad-band large dynamic
range seismograph, in The Earth beneath
the Continents, Geophysical Monograph
10, 543-553, J. S. Steinhart and T. J.
Smith, eds., Wash., D. C, American Geo-
physical Union, 1966.
Sacks, I. S., Diffracted P-wave studies of
the earth's core, 2, Lower mantle velocity,
core size, lower mantle structure, J. Geo-
phys. Res., 72, 2589-2594, 1967.
Smith, T. J., Translation from Russian,
Problems in Deep Seismic Sounding, 166
pp., S. M. Zverev, ed., N. Y., Consultants
Bureau, 1967.
Smith, T. J., J. S. Steinhart, and L. T. Aid-
rich, Crustal structure under Lake Supe-
rior, in The Earth beneath the Conti-
nents, Geophysical Monograph 10, 181-
197, J. S. Steinhart and T. J. Smith, eds.,
Wash., D. C, American Geophysical
Union, 1966.
Smith, T. J., see also Hart, S. R., James,
D. E., and Steinhart, J. S.
Steinhart, J. S., and T. J. Smith, Foreword,
in Problems in Deep Seismic Sounding ,
p. v, S. M. Zverev, ed., N. Y., Consultants
Bureau, 1967.
Steinhart, J. S., and T. J. Smith, Time
terms and structure in western Lake
Superior, in The Earth beneath the Con-
tinents, Geophysical Monograph 10, 198-
204, J. S. Steinhart and T. J. Smith, eds.,
Wash., D. C, American Geophysical
Union, 1966.
Steinhart, J. S., see also Hart, S. R., and
James, D. E.
Sumner, R. D., Attenuation of earthquake
generated P waves along the western
flank of the Andes, Bull. Seis. Soc. Am.,
57, 173-190, 1967.
Suyehiro, S., A search for small, deep earth-
quakes using quadripartite stations in the
Andes, Bull. Seis. Soc. Am., 57, 447-461,
1967.
Suyehiro, S., Quadripartite observations in
the Andes (abstract), Earthquake Notes,
Eastern Section Seis. Soc. Am., 37, 29,
1966.
Rubin, Vera C, Sandra Moore, and F. C. Tilton, G. R., see Hart, S. R., and Kouvo, O.
12S
CARNEGIE INSTITUTION
Trachslin, W., and L. Brown, Polarization
in p-"ii;0-scattering from 2.5 to 3.0 MeV
(abstract), Bull. Am. Phys. Soc, 12, 86,
19C7.
Trachslin. W., and L. Brown, Scattering of
polarized protons by deuterium from 1 to
3 MeV, Nuclear Phys., A90, 593-596,
1967.
Trachslin, W., see also Brown, L.
Tuve. M. A.. Physics and the humanities —
the verification of complementarity, The
Third Cosmos Club Award, pp. 5-20,
Cosmos Club, Washington, D. C, May 9,
1966. Reprinted in The Search for Under-
standing, pp. 41-55, C. P. Haskins, ed.,
Wash., D. C, Carnegie Institution of
Washington, 1967.
Tuve, M. A., Radio echoes (the origin of
radar), in The Search for Understanding,
pp. 73-76, C. P. Haskins, ed., Wash.,
D. C, Carnegie Institution of Washing-
ton, 1967.
Tuve, M. A., see also Baum, W. A.
Waring, M. J., and R. J. Britten, Nucleotide
sequence repetition: a rapidly reassociat-
ing fraction of mouse DNA, Science, 154,
791-794, 1966.
DEPARTMENT OF TERRESTRIAL MAGNETISM
PERSONNEL
129
Director
Ellis T. Bolton
Associate Director
L. Thomas Aldrich
Roy J. Britten
Dean B. Cowie
Scott E. Forbush
W. Kent Ford, Jr.
Stanley R. Hart 1
Staff Members
Richard B. Roberts
I. Selwyn Sacks
T. Jefferson Smith
John S. Steinhart
Louis Brown
David E. Kohne
Staff Associates
Vera C. Rubin
Kenneth C. Turner
Section Chairmen
Astrophysics: W. Kent Ford, Jr. Biophysics: Dean B. Cowie
Geophysics : L. Thomas Aldrich
Felloivs
Don J. Brenner, Fellow of the U. S. Pub-
lic Health Service
David E. James, Fellow of National
Science Foundation and Carnegie In-
stitution 2
Akira Kamitsuki, Kansai University,
Osaka, Japan 3
Thomas E. Krogh, Massachusetts Insti-
tute of Technology 4
Fernando Munizaga, University of Chile,
Santiago, Chile 5
Peter N. S. O'Brien, British Petroleum
Company, Ltd., Sunbury-on-Thames,
Middlesex, England 6
Michio Otsuka, Kumamoto University,
Kumamoto, Japan 7
Claude Petitjean, University of Basel,
Basel, Switzerland 8
Adrian V. Rake, University of British
Columbia, Vancouver, B. C.
German Saa, S. J., Universidad del Norte,
Antofagasta, Chile
Alan M. Stueber, Washington University
Shigeji Suyehiro, Meteorological Re-
search Institute, Tokyo, Japan9
Walter Trachslin, University of Basel,
Basel, Switzerland
Carlos Varsavsky, Instituto Argentino de
Radioastronomia, Villa Elisa, Argen-
tina
Junior Felloiv
Rodolfo Anzoleaga, Universidad Mayor de San Andres, La Paz, Bolivia 10
1 On leave of absence from September 1,
1966.
2 National Science Foundation to Novem-
ber 15, 1966. Carnegie Institution from
November 16, 1966.
3 Deceased August 21, 1966.
4 Through July 31, 1966.
5 Through January 31, 1967.
6 From September 1, 1966.
7 Through October 31, 1966.
8 From April 8, 1967.
9 Through December 31, 1966.
10 Through December 31, 1966.
130
CARNEGIE INSTITUTION
Collaborators and Visiting Investigators
J. L. Adams, Florida State University
Pablo Aparicio, Washington University
David Axelrod, National Institutes of
Health
Arnold Bendich, University of Wash-
ington
Bernard F. Burke, Massachusetts Insti-
tute of Technology
Ramon Cabre, S.J., Observatorio San
Calixto. La Paz, Bolivia
Bayne Carew, University of Wyoming
Mateo Casaverde, Instituto Geofisico del
Peru, Lima, Peru
Salvador del Pozo, Instituto Geofisico
Boliviano, La Paz, Bolivia
Luis Fernandez, S.J., Observatorio San
Calixto, La Paz, Bolivia
Louis B. Flexner, University of Penn-
sylvania
J. Flexner, University of Pennsylvania
Enrique Gajardo, University of Chile,
Santiago, Chile
Albert Gelderman, National Institutes of
Health
Alberto A. Giesecke, Instituto Geofisico
del Peru, Lima, Peru
Willy Haeberli, University of Wisconsin
Bill Hoyer, National Institutes of Health
Brian Lewis, University of Wisconsin
Malcolm Martin, National Institutes of
Health
Martin F. McCarthy, S.J., Observatorio
Astronomico, Specola Vaticana, Italy
Leonard Miller, U. S. Naval Medical Re-
search Center
Alois Th. Purgathofer, Universitat
Sternwarte, Vienna, Austria
D. Robson, Florida State University
Reynaldo Salgueiro, Instituto Geofisico
Boliviano, La Paz, Bolivia
Richard G. Seyler, Ohio State University
Fernando Volponi, Universidad Nacional
de Cuyo, San Juan, Argentina
Engineer and Research Assistant
Everett T. Ecklund
John B. Doak
Paul A. Johnson
Research Assistants
Charles A. Little
Liselotte Beach
H. Lowell Belin 1X
Kenneth D. Burrhus 12
Stephen J. Buynitzky
Laboratory Assistants
Margaret E. Chamberlin
Louise E. Magruder 13
Glenn R. Poe
Neltje W. van de Velde
Office
Chief, Fiscal Section: Helen E. Russell
Office Manager: William N. Dove
Librarian : Lelah J. Prothro (part time)
Secretary: Claudine C. Ator
Stenographers: Dorothy B. Dillin, E.
Kathleen Hill
Typist: Mary T. Sheahan (part time)
Accounting Assistant: Glenda J. Johns-
ton
Shop
Chief Instrument Maker and Shop Man-
ager: John G. Lorz
Instrument Makers : Robert Hoffmaster,
Michael Seemann
Machinist-Instrument Makers
Mossor,14 Carl M. Rinehart
Machinist: Francis J. Caherty
Doy E.
11 From January 16, 1967.
12 From February 1, 1967.
is Through February 28, 1967.
14 Through October 31, 1966.
DEPARTMENT OF TERRESTRIAL MAGNETISM
131
Buildings and Grounds
Carpenter and Maintenance Foreman : Caretaker : Stanley Gawrys
Leo J. Haber Assistant Caretakers : Raymond L. Bat-
Assistant Maintenance Foreman: Elliott tie,15 Bennie Harris,10 Willis Kilgore,
M. Quade Jr.17
Carol Anfinsen
Robert Cadmus
Dorothy Canter
Allen Forsbacka
John Harvey
Samuel Humphrey
Sandra Moore
Part-Time and Temporary Employees
Milan Pavich
John Roddy
Martin Roddy
Paul Roddy
Robert Singleton
Doris Titus
Anne Unger
15 July 19, 1966-February 28, 1967.
16 From March 1, 1967.
17 From December 19, 1966.
Plate 1
Department of Terr en trial M aynfJAxm
■«//,-
" PMG20 Aug 1965
h= 129Km,A = 134.5% m = 6.0, 19.03 s, 69.07 w *68, T0 09h 42m 48.5%
PP 10h 04m 27s (T-JB = -3), SKP 10h 05m 108 (T-JB = -3), sSKP 10h 06m 13s
(T - JB = 0)
Plate 1. SKP and sSKP arrivals at PMG standard station. The upper trace is the long-
period vertical component; the lower trace, the short-period vertical. Note (upper trace) that
sSKP is a mirror image of SKP. The bars on the lower trace show the position and duration of
the longest of the time windows used. The center bar indicates the section of the record used
to determine the noise. Both record sections show minute marks.
Plate :
Department of Terrestrial Magnetism
Plate 3
Department of Terrestrial Magnetism
Deep focus of Santiago del Estero
Zonda station
20th December 1966
"MS
^1
P 12 h 28 m 46s GMT (corrected)
ill
mm
/
Interruption
of energy
isupply
:?3
<■
W
P (400)
Recording velocity 44.5 mm/min
Mechanical seismograph T<> = 3 sec
Plate 3. A clear example of the S-P converted phase.
Plat, - A
Department of Terrestrial Magnetism
^tt|^ V>^;N tz z
pp
*>, ,. k .„^s
E
^ o
o
Plate 4. Seismograms obtained at Zanda (Z), Pie de Palo [PP)f and Hualilan (H). March 6
and 7, above; February 21 and 23, opposite.
Plate UB
Department of Terrell/rial, MofjueUyrn.
»*,/■■„.. X""**S
%^M^M
'■■""C'fi" ■
fy-ff'/fy* ■ #" -%;■*&
l^.*-*--^_|t^— ^~^ --"»■' -wf4/>^"" --■-^'-|4*"-iw*',i^!*4
PP
H
W^°'iSfM Jfcy*^
fcUMfai'
■^^^^•>8Jh*^"*4&^
PP
H
Plate 5
Department of Terrestrial Magnetism
S E
5562
54 01
l_
D
D
0)
k>
c
oo
O
C
5
o
c
"D
o
h_
00
•""
<v
c
c
D
^~
<u
E
_c
o
■*"
*-
M-
M-
o
00
<D
>.
C
oo
C
c
0)
o
c
00
D
a
01
c
E
'oo
D
o
<U
u
i—
u
0)
(1J
i_
"D
D
<D
E
_c
=>
i—
<D
u
i—
<D
D
a
oo
oo
CD
£
c
3
T3
E
O
D
oo
T3
0
O
oo
<D
£
_C
_o
M-
<u
O
_Q
oo
T3
u
c
D
oo
5167
4958
o
D
q-
<
D
k.
_c
0
u
E
Q.
a
o
'~o
.9"^
o
c
oo
3
w-
O
<L>
-C
^
E
(—
_Q
d
Im
C
"D
0)
>
<L>
oo
D
Q.
'a>
D
oo
U
■a
<
oo
<D
g
0
.-^
C
u
:
X
_c
a;
O)
—
c
"o
M-
O
Q)
>
D
,
_c
c
■*"
£
_D
M-
o
0
oo
_C
<D
c
JZ
o
<u
-*-
*■
>
o
■a
o
_c
c
Q.
o
_c
<
u
o
In
M-
S
0
uS
M~
_c
o
"*~
a;
l_
E
a
D
o
o
Q_
•-»-
M-
Plate 0
Department of Terrestrial Magnetism
300
o
in
c
o
I
<
200
lift*. %^.. • » . ■ .
f"r*r >3j, ,,
' ■:■■:..> ■
liifflr
. ' '■
iifii3iii:|iBi||^sSS|
,v-;:
sMii
• ^~ . - ' *-
He . ' -,: -
' s
•■■■ ■ ■:' ' ■ ■ '
'v. ■ : -■ -
#-,-■■<::;
60 70 80
Melting temperature
Plate 6. Schematic diagram suggesting the history of families of repeated DNA sequences.
The left-hand scale indicates the period of time since the saltatory replication that produced
each family. The lower scale — as a measure of divergence — represents the thermal stability
of the reassociated pairs of DNA strands formed by the members of a family. The height of
the peaks indicates the amount of DNA of a given age and thermal stability.
Committee on Image Tubes
for Telescopes
Cooperative Project of Mount Wilson and Palomar Observatories
Department of Terrestrial Magnetism, Lowell Observatory
National Bureau of Standards, and United States Naval Observatory
W. A. Baum
Lowell Observatory
John S. Hall (Chairman)
Director, Lowell Observatory
Flagstaff, Arizona
L. L. Marton
National Bureau of Standards
M. A. Tuve
Department of Terrestrial Magnetism
Contents
Review of Activities 135
Allocation of Intensifier Systems 135
Spectrograph Cameras for Image Tubes 136
Radial Velocities from Image Tube Spectra 137
Acknowledgments 140
Carnegie Institution Year Book GO, 1966-1967
Image Tubes for Telescopes
^3"
in
(M
a
<d-
if)
o^
C\l
co
CD
h~
CD
LO
X
N
in
in
o
P^O cp
r-cr>cn, <r>
NNCO CO
tO^OiD to
X
o
Z
o *•
ro ro
CO U>
o
co
oo
CO
X
o
A night-sky spectrogram obtained with an RCA cascaded image tube with the Bowen cassegrain
schmidt camera on the DTM spectrograph. For accurate radial velocity work it has been
necessary to allow for the distortions introduced by the magnetically focused tube. These dis-
tortions are shown in an exaggerated form in the schematic diagram.
Schematic diagram showing distortions greatly
exaggerated in image tube spectrum.
REVIEW OF ACTIVITIES
The Carnegie Image Tube Commit- several groups have constructed fast,
tee has continued its successful dis- high-resolution spectrograph cameras
tribution program of image tube sys- designed specifically for image tube
terns to astronomical observatories, use. One such camera, designed by
using special funds granted by the Dr. I. S. Bowen, has been used by
National Science Foundation for this Baum, Ford, and Rubin as part of the
purpose. These systems are built Committee's program of tests and
around the cascade image intensi- demonstrations of image tube tech-
fiers developed for the Committee niques. This camera, a cassegrain
over a period of years by the Electron schmidt system, provides much better
Tube Division of the Radio Corpora- image tube spectra than those ob-
tion of America. In addition to the tained previously with ordinary
intensifying tube, the system uses a lenses.
permanent magnet for focusing, a It is becoming increasingly impor-
mechanical mounting with sufficient tant, as more astronomical observa-
insulation for the required 20 kV, a tions are being made with the aid of
voltage divider, and a transfer lens image tubes, to know in detail the
and focusing device. Hence all of the photometric properties and the di-
basic parts needed to operate the in- mensional characteristics of the in-
tensifying system are furnished to tensifier systems. Low-dispersion
each observatory. With the intensify- spectrograms obtained by Ford and
ing system, astronomical observations Rubin with the DTM image tube spec-
can be made that are beyond the ca- trograph have been measured, and
pability of ordinary photographic Dr. Rubin has worked out a radial
techniques used on the same telescope, velocity reduction program that cor-
At present the most important as- rects for the small distortions intro-
tronomical application of image tubes duced by the image tube system. The
is in spectroscopic observations. Un- results indicate that accurate radial
fortunately, image tube systems are velocities can be obtained from these
too bulky to be substituted directly plates if a calibration plate is used
for photographic plate holders in to determine the distortions. It has
spectroscopic cameras. Many of the been demonstrated that, for exposures
cameras in question are fast schmidt of several hours, in most applications
systems with the focus located inter- no magnetic shielding is required
nally and hence inaccessible for bulky around the RCA cascaded tube,
image tube equipment. Therefore,
ALLOCATION OF INTENSIFIER SYSTEMS
With the support of the National are manufactured to the rather high
Science Foundation, the Carnegie performance specifications set by the
Committee has been able to purchase Committee, are individually tested
a number of cascaded image intensi- and each one is fitted into a complete
fiers (RCA type C33011) and to have intensifier system at DTM by Dr.
the basic hardware required for op- Ford. During the report year eight
erating these tubes fabricated in va- such systems were distributed and in-
rious small shops. The tubes, which stalled for the Committee by Dr.
135
136
CARNEGIE INSTITUTION
Ford and Dr. Purgathofer. Four
other tubes, with some auxiliary
parts, were made available on tempo-
rary loan for special testing programs
for the Committee.
Allocation of the Carnegie-NSF
systems is handled by a joint NSF-
Carnegie Allocations Committee un-
der the chairmanship of Dr. John
Hall. Lowell Observatory. This year
sets were lent to Dr. Thornton Page,
Wesleyan University, for use at the
Astronomical Observatory of the
Universidad Nacional de Cordoba,
Cordoba, Argentina; to Dr. F. Ber-
tola, Asiago Astrophysical Observa-
tory of the University of Padova,
Padova, Italy; to The Astronomer
Royal, Sir Richard Woolley, Royal
Greenwich Observatory, Herstmon-
ceux Castle, Hailsham, Sussex, Eng-
land ; to Dr. K. Osawa, Tokyo Astro-
nomical Observatory, Okayama,
Japan; to Dr. A. Purgathofer, Uni-
versitats-Sternwarte, Vienna, Aus-
tria ; to Dr. C. R. O'Dell, Yerkes Ob-
servatory; to Dr. Harlan J. Smith,
University of Texas (SI system) ;
and to Dr. Arne Sletteback and Dr.
Paul L. Byard, Perkins Observatory,
Ohio State and Ohio Wesleyan Uni-
versities. Previous allocations have
been to Yerkes Observatory, Kitt
Peak National Observatory, Lick Ob-
servatory, Lowell Observatory, Mount
Wilson Observatory, Flagstaff Sta-
tion of the U.S. Naval Observatory,
McDonald Observatory, Mount
Stromlo Observatory, Steward Ob-
servatory, Mount Wilson and Palo-
mar Observatories, High Altitude
Observatory, and the Leander McCor-
mick Observatory. The sets sent out
this year bring the number of com-
plete cascaded tube systems that have
been distributed to 21.
Astronomers at these observatories
are using image tubes for observa-
tions in a wide variety of spectro-
scopic programs at both high and
low dispersions. Problems under
study include high-dispersion radial
velocities of interstellar lines and
stars, planetary spectroscopy, classi-
fication and line identification in late-
type stars, moderate- and low-disper-
sion observations of barred spiral
galaxies, peculiar galaxies, radiogal-
axies, and quasi-stellar objects. One
laboratory application of astrophysi-
cal interest is a quantitative study of
high-dispersion emission spectra in a
luminous shock tube of atoms.
SPECTROGRAPH CAMERAS FOR IMAGE TUBES
The performance of an image-
intensifying system depends on the
quality of the input image. It has
proved quite feasible with coude
spectrographs to intercept the con-
verging beam from the camera with
a diagonal mirror and thus to bring a
high-quality image out to an acces-
sible location for the image tube
system. However, because of the rela-
tively small size (38 mm in diameter)
of the photocathode in the RCA cas-
caded tube, the intensifier competes
unfavorably with photography for
many problems in which very long
photographic spectra can be recorded
at high dispersion. On the other hand,
because of the extended red sensitiv-
ity of the S20 photocathode, low-dis-
persion image tube spectra tend to
be longer and hence give more infor-
mation than ordinary photographic
spectra. However, the best of these
low-dispersion systems uses fast
schmidt cameras that are poorly
suited for image tube work because
the focus is internal and therefore
inaccessible.
There have been several successful
approaches to this problem. One of
COMMITTEE ON IMAGE TUBES FOR TELESCOPES 137
these, an //l system involving two a schmidt-type corrector of 100-mm
concentric reflecting spheres, has aperture ; and a simple plano-concave
been designed by Bowen for the Palo- field-flattening lens. The fused quartz
mar cassegrain spectrograph. Suffi- block is 12.5 cm in diameter and 7 cm
cient working distance is provided at thick. The dispersed light from the
the focus so that a fairly conventional spectrograph grating passes through
permanent magnet can be used for the corrector plate and enters the
focusing the image tube. Another sys- block through a flat, annular surface,
tern, that of Meinel and Hiltner, is a After traversing the block it strikes a
folded schmidt camera that, because spherical reflecting surface, is re-
of the short working distance, re- fleeted to the smaller secondary re-
quires a solenoid with magnetic field- fleeting surface located in the center
shaping shims. An //1.5 cassegrain of the front of the block, and is then
schmidt, designed by Lynds at Kitt reflected out through the center of
Peak, is sufficiently small in diameter the second surface. Since the two
(4.2 inches) to fit inside a conven- spherical surfaces are concentric, the
tional focusing solenoid or permanent system has no off -axis aberrations,
magnet so that working distance is and the placing of the schmidt cor-
not a problem. Finally, an //2.25 cas- rector plate at the virtual center of
segrain schmidt, designed by Bowen, curvature of the spheres serves to
has a 6-cm working distance, and eliminate spherical aberration. Resid-
therefore can be operated with the ual chromatism due to refraction at
simple permanent magnet that is in- the last surface causes the spectrum
eluded in the standard cascaded tube to be tilted a few degrees with re-
system. The optical parts of this spect to a plane perpendicular to the
Bowen camera have been fabricated optical axis. The slight curvature of
for the Committee by Davidson Op- the focal plane is removed by the thin
tronics and mounted with an intensi- concave lens mounted just in front of
fier system on the DTM spectrograph, the photocathode. Well-corrected and
This cassegrain schmidt consists well-flattened images are obtained
of three elements : a solid quartz block over a 20-mm diameter. This system
having an annular flat surface and has tremendously improved the qual-
two spherical surfaces forming a pri- ity of the spectra obtained with the
mary and secondary mirror system; DTM spectrograph.
RADIAL VELOCITIES FROM IMAGE TUBE SPECTRA
The DTM image tube spectrograph The plates are measured by Dr.
has been used by Ford and Rubin to Rubin on a two-coordinate Mann
obtain plates from which accurate measuring machine; both x and y are
radial velocities are being deter- recorded for each setting. For each
mined. With the schmidt camera and line on the plate, settings are made at
a Bausch and Lomb 300-line/mm intervals of 0.03 mm along the line,
grating blazed at 7200 A, the disper- Measurements are made with the
sion is 132 A/mm. The slit width is crosswires set at 45° to the #-direc-
150 fx, which corresponds at the 84- tion of travel, so that only the portion
inch telescope to 50 n on the plate or of the spectrum at the point of in-
a resolution of 20 line pairs/mm. A tersection of the wires is considered
slit length of 12.5 mm is generally in each measurement. All plates are
used, corresponding to 160 arc sec on measured twice, with the plate ro-
the sky. tated 180° for the second measure-
138 CARNEGIE INSTITUTION
ment. From the two-dimensional tral range 4500 A to 7000 A. Over
measure of several dozen plates, it is this longer interval, the distortions
concluded that accurate radial veloci- are increased correspondingly. How-
ties can be obtained from image tube ever, they are still sufficiently small
plates with no greater difficulty than that the reduction procedure can
that encountered in determining ve- adequately correct for their effects,
locities from conventional spectro- We show in the frontispiece a 176-
scopic plates. As an illustration, for minute exposure on the night sky,
plates centered near Ha, the following covering the spectral range 4500 to
results are obtained : 7000 A. It can be seen here that the
1. For 10 neon comparison lines, distortions are still small.
5881.895 A to 6929.468 A, the for- For each plate, the radial velocity
ward and reverse measures agree to reduction program written by Dr.
within 0.1 A (2/*). Rubin for the IBM 1130 computer
2. A cubic polynomial to relate corrects for the S-distortion and the
measured distance on the plate to variation in the tilt of the lines from
wavelength will reproduce each of the the measurements of the comparison
10 laboratory wavelengths to within lines on that plate. The curvature cor-
=0.09 A (4 km/sec). rections are predetermined from
3. The distortions due to the cam- measurements made on test plates
era-image tube system are a mini- which are taken with the comparison
mum for spectra centered at the cen- source illuminating the entire slit.
ter of the image tube, and can be The night sky [OI] 6300-A line is
corrected satisfactorily in the reduc- measured at intervals along the line
tion procedure. Three distortions are as a check on the curvature correction
present: S-distortion, nonparallel and the reduction procedure. This
lines, and curvature across lines, line is close to Ha 6563 A on the
These are illustrated schematically plate and in a typical exposure is of
enlarged in the frontispiece. For the comparable density with Ha. Table 1
1000- A region discussed above, the gives results from three such meas-
variation of the y coordinate of the urements of this line.
lower edges of the spectral lines, Ay, It may be seen that there is no
is 12 p. The height of the lines in- systematic trend from one end of the
creases 60 ft for lines 7 mm high. The line to the other. This is important
variation in the tilt of the lines, Ax, for studies of velocity variations
amounts to about 25 fx over this along a line. The mean of the abso-
range. The curvature across each lute values of the residuals is about
spectral line depends principally on 0.2 A, or less than 2 ^. The mean
the height of the line on the image wavelength of the [OI] line differs
tube. With a long slit (12.5 mm) at from the laboratory wavelengths by
the Perkins 72-inch //18 telescope, about the same amount. This is sig-
the spectral lines plus comparison are nificantly less than corresponding
3 mm high on the plate; the curva- values generally found in galactic
ture of each line is less than 3 \u For studies at lower dispersion. From
the same slit length at the Kitt Peak these results we conclude that a
84-inch //7.6 telescope, the length of single measurement of a point on a
lines on the plate is over 7 mm, and line is accurate to ±2 ^ or ±12 km/
the maximum value of the curvature sec at 132 A/mm.
is 7 /a. A single plate covers the spec-
COMMITTEE ON IMAGE TUBES FOR TELESCOPES
139
CO
CM
6
o
CO
o
>.
oo
O
K
CN
O
CN
a.
'?5
-« £
^
£
0
L>
1
CO
<-<
0
CO
_J
E
o
n
i_
u
</)
^
O)
I* 2
["< o
■5 2
U ts
^ c
a
1^
-Q
CO
O
I<<
E
o
CN
l
'_
r— ■
1
CO
o
<<
c
a
u
a.
CO
<J
E
o
G
W
^<
c
D
r-Or-cNioTroTfcooo^r-'OCNio'O^Tr^eorrK
6 ^ r-' 6 d d 6 6 6 cn co d ^ -' 6 d -V -' r-' <oV
+I++++ +I+++II++I+++II
r-"-1tC0,0i0O"0OKl0'-|-,0'O00C>0000TfioCN
OCNr-OOOOOr-C0^qCNr-OO'-<-r-(N00O(
+ I ++++
+ I + + + I I
o o o o o o
+ + I + + + I I
C0'OCNOC0CNNCN>0^J-C000l0OC0iO00i0«OO^^
0;Nr-oqoaowntO;lscoqqKr-i-cNr-K
o^c^dddd(sdaddaf>f>ddoNdddc>t>
oooooooooooooooooooooo
CNCNCOCOCOCOCNCOCNCOCOCNCNCNCOCOCNCOCOCOCNCN
'O>0<0,0^'0,0^,0,0>0,0>0,0>0,0,0,0>0,0>0>0
&■
o
CN
^OTfOv^CO^OOOsiOCO^r^fCNCO
n ci cn "6 6 <- 6 cn 6 f- r-' r-' 6 co
I I ++ I + I + I + I I I +
COIOCXDIOOOOCOONOOCOCO
^ rr co q i- cn - co o >- f- r- q ^
I I ++ I + I + I + I 11 +
OCNOfOCOoOCO'OCNlOOOOOlOr-
ND<)xrr— CNCNOxTOCNCOCOOiO
c>t>ddc>dc>dddc>c>dd
oooooooooooooo
CNCNCOCOCNCOCNCOCOCOCNCNCOCO
>0>0,0,00'0,0>0V0,000>0"0
>O<>CNKC0C0C>00r-CNCOCNCN^-
r-' >- O CO r^ r-' r-' r-' r-' rJ CN 0 O "O
I I + + + + Mill! + +
,— •— CNOKKl0^lO<0 — CNCOCO
CNCNO^rr-r-CNCN^^COOO>0
66006066006006
I I + + + + I I I I I I + +
cocoNTfcocNO-oOTrcNOt
O 0> ; <— ; >0 CO CO «> ; O O 0> CO r— r- CO
c>c>ddddc>c>dc>c>ddd
oooooooooc>oooo
cncncococococncncocncncococo
< 3.-*
CO
CN K r—
6 ^ ^
l*<
*<
p
6
o
CO
<o
< 3.J
CN
CN N CO
<6 ^ d>
\<<
6
o
CO
NO
140 CARNEGIE INSTITUTION
ACKNOWLEDGMENTS
The Committee wishes to express Recently Dr. Ford has been ably as-
its thanks to the National Science sisted by Dr. Vera Rubin. The coop-
Foundation for the support which has eration and interest of many astrono-
been so vital to the success of this mers have resulted in many helpful
program. suggestions and technical improve-
It also wishes to acknowledge the ments in this extensive program. Of
outstanding technical skill, tact, and particular significance is the design
good judgment of Dr. \V. Kent Ford, of the schmidt camera system by Dr.
Jr., in carrying the main burden of I. S. Bowen.
the entire project for so many years.
Department of Plant Biology
Stanford, California
C. Stacy French
Director
Contents
Introduction 143
Biochemical Investigations 149
Role of P700 and cytochrome / in the reaction center of photosystem 1 . . 149
Transfer of energy between reaction centers of photosystem 1 in algae . . 155
Light-induced shifts in the absorption spectrum of carotenoids in red,
brown, and yellow-green algae and in a barley mutant 160
Quenching by quinones of chlorophyll fluorescence in vivo 165
Some essential considerations in the measurement and interpretation of
absorption spectra of heterogeneous samples 171
A spectrophotometer primarily for light-scattering samples at low tem-
perature 175
Absorption spectra of chlorophyll a in algae 177
Elect rophoretic study of the chlorophyll-lipoprotein complexes of Euglena . 186
The chlorophylls extracted from plants by organic solvents 189
Chlorophyll fluorescence in algae and chloroplasts 192
Absorption and fluorescence spectra of Ochromonas danica 196
The action spectrum for blue-light-stimulated oxygen uptake in Chlorella . 197
Intermediates in phytochrome transformation in vivo and in vitro . . . 203
Experimental Taxonomy Investigations 208
Growth responses of Mimulus races and Fx hybrids at the Stanford, Mather,
and Timberline transplant stations 208
Photosynthetic rates of Mimulus races and hybrid derivatives .... 214
Carboxydismutase activity in sun and shade ecotypes of Solidago . . . 216
Further studies on the effect of oxygen concentration on photosynthetic
C02 uptake in higher plants 220
Effect of oxygen concentration on dry matter production in higher plants . 228
Hybridizations in Solidago 233
Comparative physiological studies on Solanum dulcamara 233
Clusters of tree species on both sides of the Pacific 234
Staff Activities 243
Bibliography 243
Speeches 244
Personnel 245
Carnegie Institute Year Book 66, 1966-1967
INTRODUCTION
As an introduction to reports of rate photochemical steps rather than
the photosynthesis group during the a single light-driven reaction are now
past year it may be useful to mention recognized. This discovery of Emer-
first a few of the critical discoveries son and of Blinks provided an en-
that have had a particularly strong tirely new concept about the mechan-
influence on the Department's line of ism of the process. Furthermore, this
interest. During the past twenty-five led to thinking about the efficiency of
years many people from very differ- the overall process as a resultant of
ent fields of science have discovered the quantum yields of two separate
the fascination of studying the pho- photochemical reactions and of the
tosynthetic process, and their efforts losses in the electron transport sys-
have greatly changed the basic con- tern through which the light-activated
cepts of the subject. Warburg's intro- steps are correlated,
duction of Chlorella as experimental The discovery by Duysens — that it
material for quantitative investiga- was possible to measure changes in-
tions had been widely taken up before duced by light in the absorption spec-
that time. At the beginning of this tra of components of the photosyn-
period the evolution of oxygen by thetic mechanism — initiated many
isolated chloroplasts had recently detailed investigations of the nature
been discovered by Hill; the long and sequence of individual steps in
series of investigations with isolated the electron transport chain. The
chloroplasts had begun, and today steps of this chain are now being ar-
continues with increasing understand- ranged into a sequence that is logi-
ing and complexity. Some of the other cal both thermodynamically and ki-
subjects of concern at the start of netically, an enterprise to which Dr.
this period were the chemistry of the Fork and his collaborators at the De-
intermediates in the path of carbon, partment are making numerous basic
the maximum efficiency of light con- contributions. Absorption change
version by photosynthesis, and the measurements have not only been
possibility of deriving from rate used for following various steps in
measurements a comparatively simple photosynthesis but, in the hands cf
physicochemical mechanism for the Kok, have also led to the discovery
process. Another problem of that of a new form of chlorophyll into
time, which is still with us, is the which the energy absorbed by one of
clarification of the chemical nature the photochemical systems is concen-
and mode of action of the pigment trated. This pigment called "P700"
complexes that capture the energy of provides a reasonable interpretation
sunlight for driving photosynthesis in of the so-called photosynthetic unit,
plants. Some of the forms in which the existence of which was deduced
chlorophyll a occurs have been identi- long ago by Gaffron and by Emerson
fied and their complementary func- and Arnold from studies of photo-
tions in photosynthesis have become synthesis rates in flashing light.
evident. This problem remains of Another of the striking discoveries
major concern to the Department. of this period was the finding of
Studies on the participation of var- Krasnovsky that, under appropriate
ious pigments in photosynthesis have conditions, chlorophyll can be reduced
changed radically because two sepa- by light to a pink substance. Just
143
144
CARNEGIE INSTITUTION
what the functional significance of
this pink chlorophyll may be in photo-
synthesis still remains to be clarified.
Why this simple reaction of a sub-
stance, so widely and thoroughly
studied for a hundred years by in-
numerable chemists, was not found
long ago is astonishing. From this ex-
ample we must suspect that there
may well be equally simple, basic
phenomena that have not yet been
stumbled upon.
Another highlight in the under-
standing of chlorophyll chemistry
was Dr. Smith's elucidation of the
nature of the precursor of chloro-
phyll, protochlorophyll, and its trans-
formation into chlorophyll by light, a
reaction which takes place only when
the protochlorophyll is attached to
its particular protein.
During this twenty-five-year period
the path of carbon was described in
considerable detail by Calvin and his
collaborators. Another significant dis-
covery was made by Frenkel with
purple bacteria and by Arnon with
green plants: the formation of adeno-
sine triphosphate through photo-
chemical action. Just how this pro-
duction of a high-energy storage
compound is linked with the electron
transport chain is now one of the
more vigorously pursued aspects of
our subject. These and other impor-
tant findings have built a framework
within which the new discoveries
may, at least temporarily, be placed.
When plants are exposed to light
a dynamic equilibrium is eventually
reached between the reduced and the
oxidized forms of each electron-trans-
ferring component. For any one com-
ponent the ratio of the oxidized to the
reduced form at equilibrium may
vaiy with the light intensity. This
variation with the intensity of the
proportion of the oxidized forms of
cytochrome / and of the pigment
P700 has been used this year by Dr.
Fork and Dr. Amesz to study the
energy-transfer system within and
between the submicroscopic func-
tional packages of pigments and en-
zymes that constitute the photo-
synthetic units.
Within a single unit the light en-
ergy absorbed by any of the hundreds
of chlorophyll molecules migrates
around among them until it is finally
caught by the reduced form of a par-
ticular trapping molecule. There has
been some discussion as to whether
cytochrome / or P700 acts as this en-
ergy trap for system 1. These inves-
tigators found that photosynthesis can
proceed at its maximum rate when a
part of the cytochrome is oxidized but
not when any appreciable fraction of
the P700 is oxidized. This means
that P700 but not cytochrome / is the
active trapping center of system 1.
They compared the rates of photo-
synthesis with the equilibrium ratios
of the oxidized and reduced forms of
both cytochrome / and of P700. From
the relation between the oxidation
ratio of P700 and the rate of photo-
synthesis they concluded furthermore
that in certain red algae there was
transfer of migrating exciton energy
from one photosynthetic unit to an-
other. The probability of such inter-
unit transfer from a unit where P700
is nonoperational to a unit containing
an active reduced P700 molecule was
about 0.5.
Spectral absorption changes caused
by light can be used to follow the par-
ticipation of certain substances in
photosynthesis only when the sub-
stances responsible for the changes
at particular wavelengths are known.
Not all of these measurable changes
can be definitely attributed to known
substances. This year Dr. Fork and
Dr. Amesz found that increased ab-
sorption at certain wavelengths and
corresponding decreases at other
wavelengths, in a variety of algae,
and in a leaf lacking chlorophyll b,
were due to carotenoids. The changes
DEPARTMENT OF PLANT BIOLOGY 145
show that the spectra of some carote- usable at liquid-nitrogen tempera-
noids shift to longer wavelengths ture. At this low temperature the ab-
when the live plants are given light sorption bands of the forms of chloro-
that is absorbed by chlorophylls. phyll are appreciably more distinct
Some quinones found in plants are than at room temperature,
believed to be an integral part of the With the new spectrophotometer
photosynthetic system. Furthermore, we have investigated the absorption
"unnatural" quinones added to chloro- spectra of particular mutants of
plast suspensions can be reduced by Chlorella that, when grown in the
light acting on the chloroplasts. Var- dark, contain very small amounts of
ious quinones quench the fluorescence chlorophyll a and are completely free
of chlorophyll in algae or chloroplasts of chlorophyll b. One of these mutants
and also in pure solutions. The also lacks carotenoid pigments and
quenching effect of added quinones, has a preponderance of one form of
in algae or in chloroplasts, has been chlorophyll a. With this alga we have
attributed to the resulting oxidation come closer than ever before to de-
of natural quenchers that are in close termining the absorption spectrum of
contact with the chlorophyll. Further- a single in vivo form of chlorophyll
more, quenching of fluorescence has a over the entire visible spectrum,
been taken as an indication of the in- By curve analysis of spectra for
fluence of quinones on the photosyn- these mutants, and for a variety of
thetic electron-transport system. Re- different algae grown under appro-
cently, however, Dr. Amesz and Dr. priate conditions, we hope to be able
Fork have found that many quinones to establish the shapes of the absorp-
added to live algae act directly tion spectra of the natural chlorophyll
on the chlorophyll rather than a complexes. Such information is par-
quench the fluorescence indirectly ticularly important because the dif-
through oxidation of components in ferent forms of chlorophyll a are
the electron-transport chain. The known to be associated with two dif-
effect of a given quinone concentra- ferent photochemical reactions in
tion on chlorophyll fluorescence is photosynthesis. At present we can
greater in live algae than it would be give only a very rough estimate of
in solution. This result is attributed the proportions of light of a particu-
to the accumulation of the quinone lar wavelength that are absorbed by
in fats surrounding the chlorophyll, the different forms of chlorophyll.
For detecting the presence of the However, by knowing the shapes of
different forms of chlorophyll, we the spectra of the different forms, it
have been using for many years a should be possible to calculate the dis-
specialized kind of spectrophotometer, tribution between the different chlo-
which gives the first derivative of rophyll forms of absorbed light at
absorbance with respect to wave- any wavelength,
length. The curves it produces are A basic unanswered question is
excellent for the detection of chloro- whether there are only a few specific
phyll components but are not easy to forms of chlorophyll a, identical in
analyze in quantitative terms. We their spectra, occurring in various
therefore modified the instrument to proportions in different plants, or
plot also conventional absorbance each of the recognized "forms" is
spectra. We have modified the sample actually a class of pigments within
holder, which is now particularly which the individual components may
suitable for studying highly scatter- have appreciable variation of wave-
ing material such as intact algae, and length peak position.
146
CARNEGIE INSTITUTION
This question is being investigated
by comparison of absorption curves
of intact algae and by attempts to
isolate the different forms of chloro-
phyll from each other by using dis-
integrated chloroplast material from
appropriate algae. In this isolation
work the experimental difficulty is to
find effective methods for disintegra-
tion and separation that do not them-
selves modify the spectral character-
istics of the chlorophyll complexes.
Most of the past attempts to sep-
arate the forms of chlorophyll from
each other have depended mainly on
differential centrifugation. Mr. Jean-
Marie Michel has been using electric
fields in polyacrylamide gels to sep-
arate particles of disintegrated chlo-
roplasts that differ in their electric
charges. The particles used for these
electrophoretic separations, already
selected by differential centrifuga-
tion, gave preparations differing in
their absorption spectra only after
treatment with a detergent. A par-
ticularly significant preparation from
Euglena was free from carotenoids.
Various methods of further chloro-
plast disintegration are being tried to
avoid the need for detergents while
other separation procedures are being
studied by Mr. Michel and Dr. Michel-
Wolwertz.
Plants grown in the dark, then ex-
posed to light, form chlorophyll a rap-
idly. The newly formed chlorophyll
changes its absorption peak from 684
m/x to 672 mix in a short time. During
this transition period chlorophyll ac-
quires its phytol tail. It was therefore
thought that the difference between
the forms of chlorophyll in fully
greened plants might correspond to
differences in the chlorophyll extrac-
table by organic solvents. Such ex-
tracts had been found by Dr. Sironval
and Dr. Michel-Wolwertz in Belgium
to give several modified chlorophylls
separable by paper chromatography,
in addition to the common chloro-
phylls a and b. It was thought that
these "satellite chlorophylls'" might
be related to the forms of chlorophyll
in live plants. This year, however, Dr.
Michel-Wolwertz found that they
were formed by oxidation during the
extraction and chromatographic sepa-
ration.
The detection of certain pigments
in the presence of others can often be
done by fluorescence spectroscopy,
since the wavelength distribution of
light from a fluorescing pigment is
as characteristic a property as is its
absorption spectrum. Dr. Brown has
investigated the fluorescence spectra
of various algae to correlate, insofar
as possible, the absorption and fluo-
rescence spectra of the forms of chlo-
rophyll in living algae.
The relative contribution of each
pigment to the observed total may be
very different in absorption and in
fluorescence spectra. This is because
the fluorescent efficiency of the indi-
vidual pigments can be very different
and, furthermore, the efficiency of
energy transfer from one pigment to
another may be greatly influenced by
the spatial relations between them.
Many investigations have been
made of the variations in rate of
photosynthesis with time when plants
are illuminated after a dark period.
The time-course curves from such ex-
periments show all sorts of complex
induction and outburst effects that
have been widely used to deduce the
nature of various steps in the mech-
anism of photosynthesis. Not only the
rates of oxygen and of carbon dioxide
exchange resulting from illumination,
but also the converse effect, the re-
adjustment of the rate of gas ex-
change in the beginning of a dark
period following a light exposure,
have been the subject of many
studies.
Such research has usually shown
big differences in the results when
blue light is compared with other
DEPARTMENT OF PLANT BIOLOGY
147
parts of the spectrum. Blue light also
has other specific effects, such as con-
trolling the shape of higher plants,
and influencing the types of chemical
compounds formed by photosynthesis.
For blue light to do these things there
must be a particular pigment, or sev-
eral of them, that absorbs blue light
preferentially. This year Dr. James
M. Pickett measured the influence of
short exposures of blue light on the
oxygen uptake of Chlorella in the suc-
ceeding dark period. The results show
that some sort of a flavin type of com-
pound is the pigment responsible for
this blue light effect. Simultaneous
work by Dr. Kowallik in Professor
Gaffron's laboratory in Florida and
by Professor Ried of Frankfurt, a
former Visiting Investigator at the
Department, gave identical results.
Because of the different plants and
procedures used in the three labora-
tories the agreement greatly strength-
ens the theory that this pigment may
be of basic significance and may be
widely distributed in various plants.
Phytochrome is another pigment
that is probably not directly a part
of the mechanism of photosynthesis,
but is very important in controlling
the pathways of biochemical metabo-
lism. By contrast with the flavin re-
sponsible for blue light effects, phyto-
chrome has been isolated in pure form
and is far better known as a chemical
entity. When phytochrome is illumi-
nated it changes from one form to
another. The direction of the change
depends on the color of the light to
which it is exposed. This year Pro-
fessor Briggs of Stanford and Dr.
Fork continued the collaborative
work on phytochrome, discussed in
last year's report, in which some in-
termediate substances in the trans-
formation reaction of the purified
pigment were described. The recent
results show that similar intermedi-
ate products are formed when phyto-
chrome is transformed by light in
live plants, as well as when it is in
pure form. The rate constants for the
opposing reactions, whose balance
determines the relative concentration
of the two forms for an exposure to
a particular wavelength, were, how-
ever, found to differ in the plant and
in solution. They also determined the
absorption spectra for some of the
intermediate forms in the photo-
chemical transformation of phyto-
chrome.
In recent years many of the investi-
gators of photosynthesis have come
together annually or more often from
all over the world. Furthermore, sev-
eral comprehensive books on the sub-
ject appear each year and a new
international journal, Photosynthet-
ica, has been started. Such close com-
munication between different workers
is essential in planning experiments
and in modifying theories so that the
current work is relevant to present
thought on the subject. There is,
however, a danger inherent in too
much unity of thought in science, as
well as in politics and religion. When
most of the people in any field think
along the same lines, there may be
little likelihood that they will accept
new concepts. No doubt there have
been occasions when the investigators
of photosynthesis have all been simul-
taneously following major misconcep-
tions. However, the volume of special-
ized journal articles has become so
great that no one person can keep in
mind the detailed findings of various
laboratories even within a narrow
part of the field of photosynthesis.
This fact itself may help to diversify
thinking on the subject, thus increas-
ing the probability of hitting upon
important new interpretations.
In spite of all the intense research
efforts on different aspects of the
mechanism of photosynthesis, many
basic questions are still far from
clarification and some significant
areas are practically neglected in the
148
CARNEGIE INSTITUTION
rush to make new contributions to
the detailed understanding of some
of the more popular aspects of the
subject.
Experimental taxonomy. Much of
the current year's effort of the Ex-
perimental Taxonomy group has
been directed toward continued re-
searches on the comparative physiol-
ogy and biochemistry of plants
originating from contrasting environ-
ments. These studies are aimed at
discovering basic physiological mech-
anisms that operate in natural selec-
tion and evolution of higher plants.
The discovery last year that the
amount of oxygen present in normal
air depresses the photosynthetic rate
in most higher plants to about 30%
below that in oxygen-free air has led
to experiments to determine whether
or not growth of such plants is also
inhibited in normal air. Initial results
from two unrelated plants, monkey
flower and beans (Mimulus cardinalis
and Phaseolus vulgaris), do show that
dry weight yield may be twice as
great during a less than 2-week pe-
riod in an atmosphere containing only
2.5% or 5% 02 as compared with
21%. In contrast, a third plant, corn
(Zea, mays) , whose photosynthetic
C02 uptake is unaffected by 02 con-
centration in the range 0%— 21%,
failed to show a significant weight
increase under the same conditions.
These results open a new field of in-
quiry into basic differences that have
evolved in the photosynthetic mech-
anism of higher plants.
Using improved quantitative tech-
niques, Dr. and Mrs. Bjorkman have
demonstrated that the light-saturated
photosynthetic rate ingoldenrod (Sol-
idago virgaurea) is closely correlated
with the activity of the enzyme
carboxydismutase in the same leaves
of these plants. Furthermore, genet-
ically determined differences in the
light-saturated photosynthetic rates
in sun and shade races of Solidago are
closely linked with their capacity to
produce this enzyme.
A race of Solidago originally from
a sunny habitat in northern Norway
attains a high photosynthetic rate
under saturating light intensities
when the plants are previously grown
at high light intensities. The amount
of the enzyme carboxydismutase of
the leaves likewise is high. Another
race of the same species but from a
shaded habitat in southern Sweden
is capable of producing only about
half as much of the enzyme. It also
lacks the ability to attain a high light-
saturated photosynthetic rate.
The marked differences in the ca-
pacity of lowland and alpine races of
the Mimulus cardinalis-M. lewisii
complex to fix C02 under light-satu-
rating conditions reported last year
by Dr. Hiesey, Dr. Bjorkman, and Dr.
Nobs have been followed in first- and
second-generation progeny of crosses
between alpine and lowland forms.
The differences in photosynthetic ca-
pacity under light-saturating condi-
tions are inherited in much the same
way as morphological differences such
as flower color and leaf structure. The
differences in photosynthetic capacity
are partially linked through genetic
coherence with the morphological
characters. The chlorophyll content
of leaves also differs markedly in con-
trasting races, but is inherited inde-
pendently of photosynthetic capacity,
and bears no relation to light-satu-
rated photosynthetic rates.
This year Dr. Nobs and Dr. Hiesey
completed an extensive five-year
study of the performance of first-gen-
eration hybrids between ecological
races of Mimulus of various degrees
of relationship. The study was carried
out at the Stanford, Mather, and Tim-
berline transplant stations. There are
marked differences in the degree to
which hybrid vigor is expressed in
these three contrasting environments.
The expression of these differences is
DEPARTMENT OF PLANT BIOLOGY 149
highly dependent both upon the ge- tudinal vegetational belts in Japan in
netic composition of the parental comparison with comparable altitu-
races used in the crosses and upon dinal transects in western North
the environment in which they are America, with particular emphasis on
being observed. An important finding tree species. The belts of tree vegeta-
is that the survival capacity of first- tion are closely related on both sides
generation hybrids between distinct of the Pacific. This^ concept is
ecological races within this species- achieved when one studies clusters of
complex in widely different climates morphologically closely related spe-
is inherited in an intricate and largely cies that circle the earth in various
unpredictable way. The results are latitudinal belts,
being incorporated in a monographic Modern tendencies to split species
study of the Erythranthe section of and a into smaller and smaller
Mimulus that includes the results of ^ them ag . have
biosystematic, transplant, and physio- , \ & , . , ,, ,
logical investigations. obscured our perception of the rela-
Dr. Jens Clausen, on field excur- tionships that exist between gross
sions in conjunction with attendance morphological characters and broad
at the Eleventh Pacific Science Con- evolutionary clusters of species occu-
gress at Tokyo, made a study of alti- pying different continents.
BIOCHEMICAL INVESTIGATIONS
Role of P700 and Cytochrome / sorption. One light-induced signal
in the Reaction Center showed a delay upon onset of illumi-
OF Photosystem 1 nation and a relatively fast (^0.5
sec) decay upon darkening. The nega-
Jan Amesz and David C. Fork tive maxima in the difference spec-
trum of this signal were at 435 and
In spite of extensive studies rela- 705 m/x, indicating an oxidation of
tively little is known of the nature of P700 (Kok and Hoch, 1961) in the
the primary photochemical reactions light. The second signal showed an
that bring about photosynthesis. Al- immediate response to illumination
though it may be stated that more is and a slow (^6-10 sec) decay in
known about photosystem 1 than the dark. The difference spectrum
about photosystem 2, our knowledge with minima at 553 and 420 m/x and
even of the photochemistry of system a maximum at about 402 m/x indicated
1 is limited. For this reason we stud- the oxidation of an /-type cytochrome,
ied light-induced reactions of P700, probably similar to that isolated by
the presumed primary reactions of Katoh (1960) from Porphyra tenera.
photosystem 1 and cytochrome / in the Figure 1 shows some typical kinetics
red algae Iridaea splendens, Schizy- of light-induced absorbancy changes.
menla pacifica, and Porphyra per- Except for a lowering of the rate of
forata. dark decay upon lowering the tem-
Illumination of Iridaea, which had perature, the absorbancy changes in
been in the dark for a few seconds, response to light were essentially the
with red or far-red (>680 m/x) or same at room temperature and 1°C,
blue light mainly absorbed by system as well as in the presence of DCMU
1, gave absorbancy changes in the [3- (3, 4-dichlorophenyl)-l, 1-dimeth-
regions of P700 and cytochrome ab- ylurea]. No evidence was found for
150
CARNEGIE INSTITUTION
<
<
On
I
420mjj
On
!
435rn/j '
1
Off
^-y;*^w;/.*W^V^V%«,*,vr'4*'v*'
!
705m/ii \
b T
AA=2xlO"3
1
- Fluorescence
T
AA = 5xlCT3
1
ime,sec
Fig. 1 . Kinetics of light-induced absorbance changes AA in Iridaea splendens at 420, 435, and
705 ma due to oxidation of the f-type cytochrome and P700. For traces a and fo the wavelength
of the actinic light was 708 m/x (6.1 nanoeinstein cm-2 sec-1). For traces c and d the actinic light
was a band around 414 m/x, 0.7 nanoeinstein cm-2 sec-1. Trace d shows the absorbance change,
superimposed on a fluorescence signal, which caused the rapid deflections upon illumination and
darkening; trace c shows the fluorescence signal alone. The preceding dark time was 6 seconds
for all traces and the temperature was 20°C.
reactions of &-type cytochromes under
the conditions applied.
The amounts of oxidizable P700
and cytochrome / relative to chloro-
phyll were calculated from the max-
imum light- induced absorbancy
changes to be 1/4.1/310 in Iridaea.
Calculations were based on estimated
specific extinction coefficients of 70
m^I'1 cm-1 at 420 m^ for the oxidation
of cytochrome and of 73.3 mM-1 cm-1
for bleaching of P700 at 705 m**,
which is the same as that of chloro-
phyll a in 80% acetone at the red
maximum. Comparison of the size of
the absorbancy changes at 435 and
705 m/x indicated a specific extinction
coefficient of 25.5 mMr1 cm-1 at 435 hi/a
for P700. The amount of cytochrome
is about the same as found by Nishi-
mura (1967) in Porphyra, but high
compared to that present in blue-
green algae, where cytochrome and
P700 appear to be present in about
equal amounts of one per several hun-
dred chlorophyll molecules.
Quantum efficiency for cytochrome
oxidation. As illustrated in Table 1,
DEPARTMENT OF PLANT BIOLOGY
151
TABLE 1. Quantum Yields for Cytochrome Oxidation
Actinic
DCMU
Sample
Wavelength,
Intensity
Temperature,
Concentration,
Quantum
Species
No.
m//,
(nE cm_2sec_1)
°C
M
Yield
1
692
5.0
21
5X10-5
0.54
1
692
2.6
21
5X10-5
0.54
1
692
1.1
21
5x10-5
0.54
1
692
0.47
21
5X10-5
0.64
Iridaea
.
2
692
2.8
21
none
0.47
2
692
2.8
21
5X10-5
0.49
3
708
6.0
20
none
0.49
3
708
6.0
0
none
0.50
4
708
1.7
4
none
0.67
Schizymenia
1
692
2.8
20
none
0.50
Porphyra
| 1
708
6.5
20
none
0.11
11
684
5.0
20
none
0.12
Quantum yields were measured from the initial decrease in absorbance at 420 m/x upon illumi-
nation. Preceding dark time was 6 seconds for Iridaea, 5 seconds for Schizymenia and 3 sec-
onds for Porphyra. Thallus absorption was about 60%-70% at 692 m/x and 20%-30% at
708 m/x.
quantum efficiencies for the oxidation
of the /-type cytochrome in Iridaea
and Schizymenia were about 0.5-0.65
molecules of cytochrome oxidized per
quantum of red or far-red light ab-
sorbed. These quantum efficiencies
were calculated from the initial rate
of the absorbancy decrease at 420 m^
after a preceding dark time of a few
seconds. Calculation of the efficiency
for oxidation of P700 from the steep-
est part of the absorbancy curves
at 435 m^ gave considerably lower
yields of about 0.2. In Porphyra per-
forata the highest yield for cyto-
chrome oxidation observed was about
0.15. Similarly low or lower values
have been reported by Nishimura
(1967) for other Porphyra species.
Relation between the rates and oxi-
dation levels of P700 and cytochrome
f. An explanation for the delay in
P700 oxidation upon illumination
could be that photooxidized P700
reacts rapidly with reduced cyto-
chrome / and that accumulation of
oxidized P700 can occur only when all
cytochromes in the same reaction
center have been oxidized. It has been
shown (Beugeling and Duysens, 1966 ;
Duysens 1966) that this hypothesis
can explain satisfactorily the kinetics
of P890 (the presumed analogue of
P700) and of a cytochrome in the
purple bacterium Chromatium.
Figure 2 illustrates an experiment
with Iridaea, in which the alga was
illuminated for 3 seconds with an in-
tensity of light sufficient to oxidize
P700 and cytochrome nearly com-
pletely. After various dark periods a
second illumination of short duration
was given and the oxidation of these
compounds caused by the second illu-
mination was recorded. At the begin-
ning of the second illumination P700
was almost completely reduced in all
experiments, but the level of oxida-
tion of cytochrome varied widely for
different dark times applied. If it is
assumed that the oxidation of cyto-
chrome by oxidized P700 proceeds
very rapidly, then the relative initial
rate of P700 oxidation in the light
will be proportional to the fraction of
reaction centers that contain only
oxidized cytochrome molecules. When
it is further assumed that the chance
152
CARNEGIE INSTITUTION
<
<3
On On P700(434mjj)
0.5 sec dark -*/ / ~ c r\
\ ' \ a ~J O.osec Un
""l * AA = I0'3 dark
Cytochrome(420mju)
T
1 ♦.
Off
Off
1
AA=5xl0"3
1
On
t
Off
On
I
3 sec dark \J*
b Off
V,r<v^.-*».«'o*f
On
I
,•"" 6 sec dark f
Off
-fW c
Off
0
Time, sec
Fig. 2. Absorbance changes AA in Iridaea at4°C at 434 m/x (P700) and 420 m//, (cytochrome)
induced by a 1 75-msec exposure to 708 m//, (4.2 nanoeinstein cm-2 sec~T) after preceding illu-
mination with 707 m/x. Dark intervals of various lengths were applied as indicated.
of being reduced in a given time in
the dark is the same for every cyto-
chrome molecule, then
X = xn (1)
where X is the fraction of reaction
centers in which all cytochromes are
oxidized, x is the fraction of cyto-
chromes that is oxidized, and n is the
number of cytochromes per reaction
center. Thus
<fix — fCXn
(2)
where fa is the quantum efficiency for
P700 oxidation; k is a constant, equal
to fa=1. As Fig. 3 shows, the experi-
mental values for the initial rate of
P700 oxidation fit the calculated rela-
tion for n = 4 rather well, in good
agreement with the earlier mentioned
finding that the amount of cytochrome
in Iridaea is 4.1 times larger than
that of P700.
As Fig. 3 also shows, the sum of the
rates of P700 and cytochrome oxida-
tion, which might be taken as the rate
of the primary photochemical reac-
tion, is independent of the oxidation
level of cytochrome. Experiments in
which we measured the sum of these
rates as a function of the oxidation-
reduction level of P700 gave the re-
sults shown in Fig. 4.
As discussed in this year's report
(pp. 155-160), if P700 is a primary
reactant, photochemistry should be-
come less efficient when part of P700
is in the oxidized state, because
part of the reaction centers then are
not operative and light quanta will
be wasted, unless they are trans-
ferred to another reaction center.
The results of Fig. 4, in agree-
ment with those obtained on photo-
synthesis of Cryptopleura violacea
(pp. 155-160, this volume), suggest
that P700 is a primary reactant. How-
ever, the decline in photosynthetic
efficiency appears to be less than pro-
portional to the fraction of P700 that
DEPARTMENT OF PLANT BIOLOGY
153
0.6-
33
.2 0.4
E
=j
~c
o
=3
a
0.2 h
0
1 1 1 1 1 I 1 1 1
D
n O
d P700 + Cyt oxid / /
o P700 oxid / ////
-
a°X / ///
$y / ///
.ov o/ / / /
^/ Aoy / / /
/ \<?7 a/ //
/ y(yWM
/ yyyyyy
/ //yy*
-^ir^^^^x^-^r^yC-- ^yy^\ 1 1 1 1
0
0.5
Cy^0x/Cyt Total
1.0
Fig. 3. Quanfum yield for the initial rate of P700 oxidation in Iridaea (circles) and for the
sum of the initial rates of P700 and cytochrome oxidation (squares) as a function of the level of
oxidation of cytochromes. The oxidation rates were measured from the rate of absorbance
change at 435 and 420 m/x, in experiments as shown in Fig. 2. The solid lines give calculated
curves for P700 oxidation for reaction centers containing one P700 molecule and up to 5 cyto-
chrome molecules (as described in the text).
is oxidized. This suggests that energy
transfer between the reaction centers
occurs. A calculated curve (A. and P.
Joliot, 1964) for 50% probability of
energy transfer to a second reaction
center when the first one is inopera-
tive seems to give the best fit for the
experimental data.
Conclusion. The results are in quan-
titative agreement with the hypothe-
sis that a reaction center in Iridaea
contains one P700 molecule and four
cytochrome / molecules. The observa-
tion that the efficiency of the photo-
chemical reaction is dependent on the
oxidation level of P700 but not on
that of cytochrome indicates that
P700 but not cytochrome / is a pri-
mary reactant. The kinetics and effi-
ciencies of cytochrome and P700 oxi-
dation in the light can be explained
by the hypothesis that both P700 and
cytochrome are in the main path of
photosynthetic electron transport and
that P700, when oxidized photochem-
ically, oxidizes cytochrome / in a rapid
reaction. Accumulation of oxidized
P700 occurred only when all cyto-
chromes in a reaction center were
oxidized.
The results do not prove that cyto-
chrome / is oxidized via P700,
154
CARNEGIE INSTITUTION
1.0
0)
m
E 0.5
=5
O
o
1 1 1 f
0.5
P700red/P700TofQ,
.0
Fig. 4. Sum of the rates of P700 and cytochrome oxidation in Iridaea as a function of the
level of reduction of P700. The dashed lines give the expected relations for different proba-
bilities (p) of transfer of excitation energy between reaction centers (see text). Open and solid
circles represent two different sets of measurements, one done at 0°C and the other at 4°C, with
different samples and actinic light of 708 mfi.
although this is the simplest hypoth-
esis that explains the results quan-
titatively. Direct information about
this could possibly be obtained by the
use of intense flashes of short dura-
tion to oxidize P700. In this respect
it is interesting" to note a recent ex-
periment of Nishimura (1967) who
reported the oxidation of a relatively
large amount of cytochrome / in
Porphyridium cruentum by a single
laser flash. This could argue against
the hypothesis that the cytochrome is
oxidized via P700, which is probably
present in smaller amounts than the
amount of cytochrome that was oxi-
dized.
References
Beugeling, T., and L. N. M. Duysens, in
Currents in Photosynthesis, edited by
J. B. Thomas and J. C. Goedheer, Ad.
Donker, Rotterdam, p. 49, 1966.
Duysens, L. N. M., in Currents in Photo-
synthesis, edited by J. B. Thomas and
J. C. Goedheer, Ad. Donker, Rotter-
dam, p. 263, 1966.
Joliot, A., and P. Joliot, Compt. Rend.,
258, 4622, 1964.
Katoh, S., Plant and Cell Physiol., 1, 91,
1960.
Kok, B., and G. Hoch, in Light and Life,
edited by W. D. McElroy and B. Glass,
Johns Hopkins Press, Baltimore, p.
397, 1961.
DEPARTMENT OP PLANT BIOLOGY 155
Nishimura, M., in Energy Conversion by transfer between system 1 units in
the Photosynthetic Apparatus, Brook- algae we therefore compared, in a
haven Symposia in Biology, No. 19, number of algal species, the relative
Brookhaven National Laboratory, efficiency of Q2 evolution with the
Upton, N. Y., 1967. oxidation level of P700 and of cyto-
Transfer of Energy between chrome / as a function of light in-
Reaction Centers of Photosystem tensity. ^
1 in Algae ^ various intensities of light we
measured the fraction of P700 and
David C. Fork and Jan Amesz cytochrome / that was in the reduced
The concept of the photosynthetic state. Under the same conditions the
unit stemming from the early work relative rate of 02 evolution was
of Emerson and Arnold (1932, a, b) measured as a function of light in-
has been substantiated considerably tensity with a Teflon-covered elec-
during recent years. A photosyn- trode (Year Book 61, p. 343). We
thetic unit can be defined as a pri- used adjacent parts of the thalli or,
mary reaction center with associated with Schizothrix, samples of the same
pigment molecules (chlorophylls, ca- culture. In all experiments discussed
rotenoids, biliproteins) having a below, the actinic light was of a
higher probability of transferring wavelength band such that absorption
their excitation energy to this reac- by system 2 exceeded that by system
tion center than to another center. 1 (blue light for the green alga Viva
Experiments with purple bacteria lobata ; green light for the red algae) .
( Vredenberg and Duysens, 1963 ; and Therefore, we can assume that photo-
Clayton, 1966) and with the green chemistry in system 1 occurred at
alga Chlorella pyrenoidosa (A. and maximum efficiency in the linear part
P. Joliot, 1964) indicate that the pho- of the light curve of photosynthesis,
tosynthetic units in purple bacteria, Figure 5 shows the rate of 02 evo-
as well as the photosynthetic units of lution and the absorption changes
system 2 (the 02 evolving system) in corresponding to the oxidation level
algae are not separated, but that ex- of P700 and of cytochrome as a func-
citation energy can move more or less tion of light intensity in the red alga
freely from one unit to another. When Cryptopleura violacea. The rate of 02
one reaction center is inoperative, by evolution was the steady-state rate
just having trapped an exciton, a after several minutes of illumination,
second exciton may be transferred to The oxidation levels of P700 and cyto-
a different reaction center. chrome were measured with the ap-
The extent of energy transfer from paratus described earlier (Year Book
one photosynthetic unit to another is 63, p. 435) by the increase of absorp-
an important factor in determining tion at 706 for P700 and 420 m^ for
the amount of light energy that even- cytochrome, upon turning off the ac-
tually reaches a functional reaction tinic light. The exposure was long
center when some of the reaction cen- enough to give a steady-state rate of
ters are inoperative. Therefore, it is 02 evolution. To minimize signals
possible in principle to obtain inf or- caused by chlorophyll fluorescence the
mation about this transfer by meas- photomultiplier was placed about 30
uring the efficiency of conversion of cm from the sample cuvette. The
light energy into photochemical prod- measurements were corrected for flu-
ucts as a function of the fraction of orescence by subtraction of the signal
reaction centers that is operational, obtained under the same conditions
To obtain information about energy but without the measuring beam.
156
CARNEGIE INSTITUTION
iq
b 3
<
P O) «
c c
o o
o
>
o
o
c
o
o
CO
o
u
o
0
] 1
Cryptc
1 1 J 1 1 1 1
pleura
1
—
^-"~ ~^- t o-
/
^°*^ 02 evolution
/
/ /
/ /
//
^ 5^ ^706*^700)
~
If
L^^T ^420nvj(Cyt.)
//
-
1 'i
/ //
^7 #/
/ / /
/ / /
-/ //
T (f /
1 / J
I s y3
/ / _/
1 1 1 1 1 1 1
1
0
10
Light intensity, ergs cm-2 sec"1 xlO
-4
Fig. 5. Steady-state level of oxidation of P700, f-type cytochrome, and steady-state rate of O2
evolution in Cryptopleura v/o/acea as a function of the intensity of green light (540 m/x max), 40
nut half bandwidth.
As Fig. 5 shows, P700 was almost
completely in the reduced state at
light intensities where photosynthesis
proceeded with maximum efficiency;
only above the point where the in-
tensity curve for photosynthesis
started to depart from linearity did
a sizable proportion of P700 become
oxidized. By contrast, the cytochrome
was already partly oxidized at lower
intensities. These results therefore
indicate that P700, but not cyto-
chrome, behaves as a true reaction
center because a maximum rate of
photosynthesis can be expected only
when all reaction centers are opera-
tional (in the reduced state).
For quantitative evaluation, the
data of Fig. 5 were replotted as
shown in Fig. 6, which gives the rela-
tion between the relative efficiency of
02 evolution and the fraction of P700
and of cytochrome that was in the re-
duced state. The relative efficiency for
02 evolution was obtained by dividing
the relative rate by the light inten-
sity. The relative efficiency corre-
sponding to the dashed line in Fig. 5
was normalized at unity. The fraction
of P700 and of cytochrome in the re-
duced state was calculated by sub-
tracting the light-induced signal from
the maximum signal obtainable with
high actinic intensity of red or blue
light (system 1) and with 5 • 10~5 M
DCMU. The shaded areas in Fig. 6
indicate the estimated uncertainty. At
a high level of reduced P700 or cyto-
chrome the uncertainty was mainly
determined by the measurement of
the relative photosynthetic efficiency;
at a low level of reduction, by the
measurement of the total amount of
oxidizable P700 or cytochrome.
Figure 6 shows that even when a
significant proportion (about 20%)
of the cytochrome is oxidized, photo-
synthesis can proceed at maximum
DEPARTMENT OF PLANT BIOLOGY
157
0 0.5 1.0
P700red/P700jota| or Cyirec|/Cyif0fa|
Fig. 6. Relative efficiency of O2 evolution as a function of the level of reduction of P700 and
of the f-type cytochrome in Cryptopleura v/o/acea.
efficiency. By contrast, the oxidation
level of P700 appears to be closely
correlated with photosynthetic ef-
ficiency; a drop in the level of reduced
P700 is accompanied by a decline in
relative efficiency of photosynthesis.
If no energy transfer between pho-
tosynthetic units occurred, a relation
between photosynthetic efficiency and
level of reduction of P700 should exist
as represented by the diagonal line in
Fig. 6. However, the curve for P700
lies above this line, indicating that
energy transfer does occur. The
dashed curves have been computed by
A. and P. Joliot (1964) from a sim-
plified model of system 2 for different
probabilities for energy transfer. The
experimental data agree rather well
with a curve corresponding to a prob-
ability of about 50% to 60% for
transfer of energy to a second reac-
tion center when the first one is in-
operative.
Figure 7 shows results obtained
with the blue-green alga Schizothrix
calcicola. With this alga the results
were more complicated than with
Cryptopleura. Here also the left part
of the band lies above the diagonal,
suggesting transfer between reaction
centers of system 1. However, at the
right side of the figure (correspond-
ing to low light intensities) the meas-
urements drop considerably below
the diagonal line. This effect may be
caused by a reaction of oxidized P700
with reduced products of system 1,
either directly or via intermediate
substances such as cytochrome /.
Such a reaction could keep the level
of reduced P700 high and increase
158
CARNEGIE INSTITUTION
1.0
u
C
o
o
>
O 0.5
o
c
UJ
0
i 1
I
1 « i i i A
-
/ k
/ M
-
/ r-
-
Schizothrix
^0^ £ '
-
\M
/
/
9 \
/
Porphyra
- /
-
/, ,
i
1 i i i i
0
0.5
P700red/P700fofa|
1.0
Fig. 7. Relative efficiency of O2 evolution as a function of the level of reduction of P700 in
Schizothrix ca/cico/a and Porphyra perforata.
the rate of electron transport in sys-
tem 1 to a level higher than com-
puted from Oo evolution, as discussed
below. The same explanation could
apply to the red alga Porphyra perfo-
rata, also shown in Fig. 7. Results
with Ulva lobata were similar to
those obtained for Porphyra.
Discussion. The results obtained
with Cryptopleura support the hy-
pothesis that P700, but not cyto-
chrome /, acts as a primary reactant
of light reaction 1 in photosynthesis.
Accumulation of oxidized P700 was
accompanied by a decline in photosyn-
thetic efficiency, but the cytochrome
was already partly oxidized at light
intensities where no such decline was
observed. The results with the other
algae also agree with the assumption
that P700 is a primary reactant. In no
instance did oxidized P700 accumu-
late when 02 evolution proceeded at
optimum efficiency. From these ex-
periments alone it might be argued
that the oxidation of P700 is a side
reaction that only occurs when photo-
synthesis is saturated or inhibited.
However, this would be in disagree-
ment with the observation of Kok et
al. (1963) that the turnover rate of
P700 in spinach chloroplasts is
strongly stimulated by the addition
of NADP and ferredoxin and is in-
hibited by DCMU.
The results with Cryptopleura and
Schizothrix indicate that energy
transfer occurs between photosyn-
thetic units of system 1. In Crypto-
pleura photosynthesis proceeded with
50% relative efficiency even though
P700 was about 70% oxidized. A sim-
DEPARTMENT OF PLANT BIOLOGY 159
ilar phenomenon was seen in Schizo- between photosynthetic units prior to
thrix at high light intensities. the first visit. A formally different
No evidence for energy transfer model, applied to bacteria, by Vreden-
was obtained by the experiments with berg and Duysens (1963) implies
Porphyra. However, these experi- that the reaction centers are homo-
ments could prove the absence of geneously dispersed between the light-
energy transfer in this alga only if harvesting pigment molecules, and
two conditions were fulfilled : that the chance of finding a second
1. The rate-limiting step in photo- reaction center depends only on the
synthesis is not in the Calvin cycle or number of random walks involved,
in one of the reactions leading to the This model can probably not be ap-
reoxidation of the hypothetical pri- plied to algae, for reasons discussed
mary-reduced product of system 1. below.
If it were, accumulation of primary In Schizothrix no correlation has
reductant could limit the efficiency of been found (Vredenberg and Duy-
light utilization in system 1. It is un- sens, 1965) between fluorescence of
likely that this effect occurred in our system 1 and the oxidation level of
experiments, since the concentration P700. For this reason it has been as-
of C02 was not rate limiting; more- sumed (Duysens, 1965) that energy
over, there is other evidence that the from the bulk chlorophyll is trans-
rate-limiting step is in the electron- ferred to the reaction center via a
transport chain between the two sys- special, weakly fluorescent chloro-
tems. phyll that is present in small amounts
2. The second condition is that the and located near the reaction center,
rate of 02 evolution is equal to the Back transfer of energy from this
rate of electron transfer by system 1. chlorophyll to the bulk chlorophyll
This is not true when a ' 'cyclic' ' reac- would be negligible, so that bleaching
tion occurs between the reduced and of P700 would not affect the fluores-
oxidized products of system 1, be- cence yield of the bulk chlorophyll. A
cause then the rate of photochemistry consequence of this hypothesis is that
in system 1 would be higher than esti- energy transfer between reaction
mated from measurements of 02 evo- centers would occur via the special
lution. There is evidence that cyclic chlorophyll which should form a more
reactions occur in intact algae and in or less continuous structure contain-
isolated chloroplasts, especially in the ing the reaction centers.
presence of redox catalysts. This
would explain the results with For- References
phyra and other divergent results as,
for example, those with Schizotrix at Clayton, R. K., Photochem. Photobiol, 5,
low light intensity. 807, 1966.
It must be noted that the probabil- Duysens, L. N. M., Arch. Biol. (Liege),
ity of energy transfer p of 50% to 76, 251, 1965.
60% as derived from the data of Emerson, R., and W. Arnold, /. Gen.
Fig. 6 probably gives only an approxi- Physiol, 15, 391, 1932a.
mate measure of the extent of energy Emerson, R., and W. Arnold, /. Gen.
transfer between the photosynthetic Physiol, 16, 191, 19326.
units in system 1. The number p j0liot, A., and P. Joliot, Compt. Rend.,
gives the probability for transfer to 258, 4622, 1964.
a second reaction center when the Kok> B#) B. Cooper, and L. Yang, in
first one visited is inoperative, but Studies on Microalgae and Photosyn-
gives no information about transfer thetic Bacteria, Japanese Soc. of Plant
160
CARNEGIE INSTITUTION
Physiologists, The Univ. of Tokyo
Press, p. 373, 1963.
Vredenberg, W. J., and L. N. M.
Duysens, Nature, 197, 355, 1963.
Vredenberg, W. J., and L. N. M. Duy-
sens, Biochcm. Biophys. Acta, 94, 355,
1963.
Light-Induced Shifts in the
Absorption Spectrum of
Carotenoids in Red, Brown, and
Yellow-Green Algae and in a
Barley Mutant
David C. Fork and Jan Amesz
It has been known for a number of
years that certain photosynthetic bac-
teria exhibit light-induced changes of
absorption in the region 450 to 540
niju, which are apparently produced by
a shift toward longer wavelengths of
the absorption of a carotenoid pig-
ment. Characteristic difference spec-
tra are observed which show maxima
and minima separated by about 30
m^t. There is evidence (Amesz and
Vredenberg, 1966) that these changes
are not caused by oxidation-reduction
reactions of carotenoid pigments but
rather that they may be produced by
a change in the environment of the
pigment molecules resulting from
electron transport.
We have found in a number of
variously colored algae and a barley
mutant light-induced absorbance
changes which appear to be caused
by a shift toward longer wavelengths
of the absorption bands of a carote-
noid.
Kinetics and absorption difference
spectra. Figure 8 shows absorbance
changes at 495 m/x seen in the red
alga Iridaea splendens, which we have
attributed to a shift of absorption of
a carotenoid. The absorbance change
produced by low-intensity orange light
(622 mp, absorbed mainly by system
2), were relatively simple (trace d).
Illumination produced a slow decrease
of absorbance followed by a steady
state and a reversal upon darkening.
Red light (684 nn<,, absorbed mainly
by system 1) produced more compli-
cated kinetics. A rapid negative
transient occurred, followed by a
slower increase to a steady state
Iridaea
622mp,I=l0.2
Fig. 8. Light-induced changes of absorbance AA at 495 m/x in the red alga Iridaea splendens
induced by actinic light of 684 and 622 m/x. The intensities, I, are expressed in nanoeinstein cm-2
sec"
— i
DEPARTMENT OF PLANT BIOLOGY
161
above the dark level (traces a, b).
The response induced by light of 622
ni/x of high intensity was rather vari-
able. Here the first decrease was fol-
lowed by a second, slower one (trace
c) , but sometimes it was followed by
a slow and small increase in absorb-
ance. Kinetics at 515 rmt were simi-
lar but opposite in sign, suggesting
that they were at least partly caused
by the same compound. Kinetics at
these wavelengths indicated that the
changes were not caused by P700 or
cytochrome /; moreover, these com-
pounds have little absorption in this
spectral region.
Light-induced absorption difference
spectra were measured in the region
440 to 540 m^ in Iridaea. The differ-
ence spectrum of the steady-state
change produced by orange light is
shown in Fig. 9. A very similar spec-
trum resulted from plotting the rate
of change instead of the steady state.
Figure 10 shows difference spectra
obtained with red actinic light and the
same conditions as for trace a of Fig.
8. The open circles give the deflection
after 20 msec of illumination, which
is the time needed to achieve the
maximum negative deflection. The
curve with the solid circles gives the
difference between the maximum de-
flection and the steady state, corre-
sponding to the slow reversal of the
signal at 495 m/x.
The spectra of Figs. 9 and 10 are
all similar and show maxima and
minima at about 465, 480, and 495
m/x, and a maximum (or shoulder) at
515 m/x, which suggests that they are
due to the same compound. The shape
of the spectrum of Fig. 9, which
seems uncomplicated by other
changes in the region above 510 m/x,
suggests that it is caused by a shift
toward longer wavelength of a com-
pound with three maxima at or some-
what below 445, 473, and 504 m/x. The
location and distance of the maxima
(30 m/x) suggest that this compound
is a carotenoid.
The difference spectrum of the red
alga Schizymenia pacifica induced by
orange light, shown in Fig. 11, was
very similar to that of Iridaea, al-
though the changes were much
smaller, and showed maxima and
CO
O
<
<d
+ 1.0
- r
l i i
Iridaea
i
i 1
i l i
514
i
+0.5
-
451
481
N^ -
0
-0.5
-
466
-
-i n
I
1 I i
i
,495,
l I I
!
450
500
Wavelength.mjU
550
Fig. 9. Light-minus-dark difference spectrum of Iridaea produced by 1 second of illumination
with actinic light of 622 m/x (1.1 nanoeinstein cm-2 sec-1).
162
CARNEGIE INSTITUTION
+1.5
+1.0
O
X
<
<
+ 0.5-
C
-0.5
-1.0
495
450
550
Fig. 10.
500
Wave length, mfj
Difference spectra of Iridaea produced by light of 684 m^t (17.7 nanoeinstein cm-2
sec-1). Open circles: absorbance change, AA, produced during the first 20 msec of illumination.
Solid circles: spectrum for the reversal of the absorbance change during illumination (difference
between maximum deflection produced initially and the steady state; see Fig. 1, trace a).
minima at about 452, 463, 480, 495, gests a similar shift in carotenoid
and 511 m/t. The difference spectrum absorption as in the other red algae,
of Porphyra perforata, also shown in but the spectrum is apparently more
Fig. 11, had maxima and minima at strongly distorted by other absorption
470, 483, 497, and 515 m/x, which sug- changes. This was also true of the
"i r
i i 1 r
•.Porphyra
Schizyrnenia
450
550
500
Waveleng+h,mjL/
Fig. 1 1 . Difference spectra of the red algae Schizyrnenia pacifica and Porphyra perforata pro-
duced by light of 622 m/x (1.5 and 8.7 nanoeinstein cm-2 sec-1, respectively).
DEPARTMENT OF PLANT BIOLOGY
163
brown alga Phaeostrophion irregular e
(Fig. 12).
The shape of the difference spec-
trum (Fig. 13) for Botrydiopsis
alpina (Xanthophyceae) was very
similar to that of Iridaea and had
maxima at 450, 482, and 520 and
minima at 468 and 498 m^.
We were unable to distinguish a
carotenoid shift in green algae (Chlo-
rophyceae), since the presence of
chlorophyll b — which produces the
large negative change at 475 and the
positive change at 515 nut (Year Book
65, p. 473) — obscured other changes
in this region. However, we were
fortunate to have a mutant strain
(Chlorina 2) of barley (Hordeum
vulgar e), which lacks chlorophyll b
(Highkin and Frenkel, 1962) but is
capable of growing to maturity. This
mutant, which was obtained by
Robertson in 1933, was made avail-
able to the Department through the
courtesy of Dr. Harry R. Highkin.
The plant produced a difference spec-
trum (Fig. 14) which was still com-
plicated by other absorbance changes
but nevertheless had maxima and
minima suggesting a shift in absorp-
tion of a carotenoid.
We did not invariably observe the
carotenoid absorbance changes. In
Iridaea, at times, we were unable to
observe them. We have thus far not
seen them in the blue-green algae
examined (Synechococcus cedrorum
and Schizothrix calcicola). In barley
we previously saw ( Year Book 65, p.
474) only another change with a peak
near 525 nut and no negative changes
below 500 mcifx. This change may dis-
tort the spectrum of Fig. 14.
Pigment system responsible for the
absorption shift. The kinetics of the
absorption changes in light mainly
absorbed by system 1 or by system 2
illustrated in Fig. 8, suggest that the
shifts in carotenoid absorption were
driven by both pigment systems; ex-
citation of system 1 causes a rapid
change, followed by a slow reversal,
and system 2 causes only a slow
change of absorption. Additional
evidence that the rapid spike (as in
trace a of Fig. 8) is brought about by
system 1 was obtained by experiments
with DCMU which showed no inhibi-
tion of the spike by this compound.
For Iridaea the relative efficiency of
684- or 622-m/x light for bringing
about the carotenoid absorption shift
o
X
<
<
+1.5
-
| 1 1 1
Phaeostrophion
i
I 1
506
1 1
1 1
+ 1.0
-
476
—o^o
p -
+0.5
-
-
0
>ior
! t i i i ! » i i i 1
450
550
500
Wavelength, mp
Fig. 1 2. Difference spectrum of the brown alga Phaeostrophion irregulare produced by light of
622 m/j, (1.3 nanoeinstein cm-2 sec-1).
164
CARNEGIE INSTITUTION
O
X
<
<
1 i '
i
1
' 520 '
i
Boirydiopsis
482
+ 0.5
450
0
-0.5
468
-
-i n
i ! i i
i i
498
1
i i i
i
450
550
500
Wavelength.n^j
Fig. 13. Difference spectrum for the steady-state absorbance changes produced in Botrydiopsis
alpina by actinic light of 672 m/x (8.5 nanoeinstein cm-2 sec-1).
was measured, and the results (Table
2) indicated that the relative activi-
ties were the same as for cytochrome
oxidation, a well-known system 1 re-
action.
The shifts in carotenoid absorption
are remarkably similar in properties
to those seen at 475, 515, and 650 m/x
in the Chlorophyceae and in other
green organisms containing chloro-
phyll b (Year Book 63, p. 441 and
Year Book 65, p. 473). In both cases
excitation of system 1 gives rise to a
rapid transitory change upon illumi-
nation, followed by a reversal upon
darkening. The initial, fast carote-
noid shift, like the 515-nu/, change, is
dependent upon the preceding dark
time, and increases up to a certain
maximum with increasing dark time.
It has been argued (Fork, Amesz,
and Anderson, 1967; Witt et al.,
1967) that the 515-m/x change in
green algae is not produced by an
oxidation - reduction reaction but
rather by a change in the environ-
ment of a compound, probably chlo-
rophyll b, which gives rise to a rela-
tively small change in its absorp-
tion spectrum. The change in environ-
ment would then be caused in some
way by electron transport in system 1
or 2. The same might be true for the
carotenoid shift. There is evidence
(Amesz and Vredenberg, 1966) that
the shift in carotenoid absorption in
the purple bacterium Rhodopseu-
domonas spheroides is not caused by a
chemical reaction, because the ab-
sorption of one quantum of light by
this organism resulted in a shift of
the absorption of about 3 molecules of
carotenoid. A similar analysis of the
changes observed here has, so far,
not been possible because the location
of the absorption bands of the carote-
noid is not known.
DEPARTMENT OF PLANT BIOLOGY
165
+2.0
+1.0-
O
<
<
450
500
Wavelengfh,m/j
Fig. 14. Difference spectrum for the initial
change of absorbance in attached mutant bar-
ley leaves (lacking chlorophyll b) produced by
excitation with a broad band of red light, from
about 620 to 800 m/x (3.6 X 105 ergs cm"2
sec-1). A schedule of 6 seconds light and 12
seconds dark was used.
References
Amesz, J., and W. J. Vredenberg, in
Currents in Photosynthesis, edited by
J. B. Thomas and J. C. Goedheer, Ad.
Donker, Rotterdam, p. 75, 1966.
Fork, D. C, J. Amesz, and J. M. Ander-
son, in Energy Conversion by the
Photo synthetic Apparatus, Brookhaven
Symposia in Biology, No. 19, Brook-
haven National Laboratory, Upton,
N.Y., p. 81, 1967.
Highkin, H. R., and A. W. Frenkel, Plant
Physiol, 37, 814, 1962.
Robertson, D. W., Genetics, 18, 148, 1933.
Witt, H. T., G. Doring, P. Schmidt-
Mende, U. Siggel, and H. H. Stiehl, in
Energy Conversion by the Photosyn-
thetic Apparatus, Brookhaven Sympo-
sia in Biology, No. 19, Brookhaven
National Laboratory, Upton, N.Y., p.
161, 1967.
Quenching by Quinones of
Chlorophyll Fluorescence in
vivo
Jan Amesz and David C. Fork
During the past years quinones
have attracted interest as possible
intermediates in photosynthesis.
There is evidence that plastoquinone
is an intermediate in the photosyn-
thetic chain, and as early as 1954
Wessels made the hypothesis that a
quinone functions as the primary
photooxidant of photosystem 2. Qui-
nones have been found to act as Hill
oxidants, as cofactors of photosyn-
thetic phosphorylation, and as inhibi-
tors of photosynthesis.
Livingston and Ke (1950) reported
that benzoquinone and substituted
benzoquinones quench the fluores-
cence of chlorophyll in organic solu-
tion, and more recently some observa-
tions have been published about
quenching of fluorescence of chloro-
TABLE 2. Relative Activity of Light of 684 and 622 mp, in Causing the Shift in
Carotenoid Absorption and Cytochrome Oxidation in Iridaea splendens
Reaction
Flash Interval
Inhibitor
Activity Ratio,
684/622 m/x
Carotenoid shift
Cytochrome oxidation
1 min
3 sec
1 min
3 sec
6 sec
none
none
DCMU
DCMU
DCMU
1.47
1.34
1.33
1.34
1.46
The measurements were done at 495 m//, for the carotenoid shift and at 420 mfx for the cy-
tochrome oxidation. Actinic light was given in 1 0-msec flashes with dark intervals as indicated
in the second column. The last column gives the activity of absorbed quanta of 684 m^ relative
to those at 622 m/*. The concentration of DCMU, where used, was 5 X 10-5 M.
166
CARNEGIE INSTITUTION
phyll in isolated chloroplasts by
naphthoquinones.
We have studied quenching of
chlorophyll fluorescence in Swiss
chard chloroplasts and algae. The re-
sults indicate that the quenching proc-
ess is caused by direct interaction
of chlorophyll and quinone molecules
and not by stimulation of photosyn-
thetic electron transport, as has been
postulated.
The effect of DCMU and quinones
on fluorescence kinetics. Figure 15
shows recordings of the time course
of fluorescence for Swiss chard chlo-
roplasts in the absence and in the
presence of 10~5 M DCMU. The initial
level of fluorescence upon illumination
is about the same under both condi-
tions but with DCMU the fluorescence
rises much faster and to a higher level
than in its absence. According to
Duysens and Sweers (1963) this in-
crease is caused by the reduction of
Q, the primary photooxidant of sys-
tem 2. The slower rise in the absence
of DCMU probably reflects the reduc-
tion of a larger pool, which may give
rise to the oxygen burst reported
earlier (Year Book 61, p. 334), and
which tends to keep Q in the reduced
state initially. In the following we
will call the final level of fluorescence
the "total fluorescence" and the
difference between the total and the
initial fluorescence will be called the
"variable fluorescence."
Figure 15 also shows the effect of
2-methyl-l,4-naphthoquinone (mena-
dione, vitamin K3). Menadione
strongly quenched chlorophyll fluores-
cence, especially the variable fluores-
cence, which was already strongly
quenched at 1.8 X 10-5 M, a, concen-
tration that hardly affected the initial
fluorescence. The quenching effect as
a function of concentration is shown
in Fig. 16.
Chloro
Swiss
i 1 i 1
plasis,
chard
i 1 i
' DCMU
No DCMU
c
(~ DCMU
I0"5M Menadione
o
to
o
5xlO"5M Menadione
r
n
DCMU
i i i 1
0
I0~3M Menadione
On
i i t i
0
4 0
Time, sec
Fig. 15. The effect of DCMU (10~5 M) and of DCMU with three different menadione concen-
trations on the kinetics of fluorescence in Swiss chard chloroplasts. The excitation light was a
band centered at 545 m/x for the traces on the left and at 420 m/i for the traces on the right.
The intensities and vertical scales for the traces on the left and right, which were made with
separate samples, are different. A 24-second dark period preceded each illumination.
DEPARTMENT OF PLANT BIOLOGY
167
e-ie-
.00
0.75
0
u
c
u
CO
<D
t-
O
Z5
0.50
0.25
0
Chloroplasts, Swiss chard
+DCMU
Initial fluorescence
Total
fluorescence
0
■ih
10
Variable
fluorescence
I0~4
Menadione cone. (M)
Fig. 16. Quenching of the initial, variable, and total fluorescence of Swiss chard chloroplasts
as a function of menadione concentration in 1 0~5 M DCMU: <£ is the relative fluorescence yield in
the absence of quencher and <£' in the presence of quencher. The excitation light was green (545
mji).
A number of other quinones were
tested for quenching activity. Table 3
summarizes the results. Like mena-
dione, most quenchers acted much
more strongly on the variable than on
the initial fluorescence. Some qui-
nones such as phthiocol and lawsone
showed little quenching activity but
others were equal to or more active
than menadione. Reduced menadi-
one and 2,3,5,6-tetramethyl-benzo-
quinone had little or no quenching
TABLE 3. Quenching of Initial and Variable Fluorescence in Swiss Chard Chloroplasts1
Compound
Initial Fluorescence Variable Fluorescence
Concentration (ftM) for 50% Quenching
p-benzoquinone
1200
2, 3-dimethy!-p-benzoquinone
800
2, 3, 5, 6-tetramethylbenzoquinone
450
1 , 4-naphthoquinone
200
2-methyl-l, 4-naphthoquinone
(menadione)
160
2-hydroxy-l , 4-naphthoquinone
(lawsone)
> 2000
2-hydroxy-3-methyl-l , 4-naphthoquinone
(phthiocoi)
> 1000
5-hydroxy-l , 4-naphthoquinone
(juglone)
70
1, 2-naphthoquinone
580
1, 2-naphthoquinone-4-sulfonic acid
> 2000
Phenanthrenequinone
68
1, 2-dihydroxyanthraquinone
(alizarin)
66
m-dinitrobenzene
930
1500
120
83
63
28
> 2000
500
13
340
>2000
16
21
130
* Fluorescence was excited by green light (545 m/x, 2 to 6 • 1 03 ergs cm-2 sec-1), after a
preceding 24-second dark period. The DCMU concentration was 10~5A4.
168
CARNEGIE INSTITUTION
activity. Several quinones were also
tested for their effect on fluorescence
of intact algae. Menadione and 1,4-
naphthoquinone, which were active
with chloroplasts, were found to be
active with algae too, although gen-
erally at a somewhat higher concen-
tration. Lawsone and phthiocol, in-
active with chloroplasts, were in-
active with algae also. Figure 17
illustrates the effect of 1,4-naphtho-
quinone on the red alga Porphyra
perforata.
Mechanism of the quenching effect.
There are several indications that the
quenching, which occurred both with
DCMU and without it, is not caused
by an oxidation of reduced Q (QH)
by quinones:
1. The E0' of the quencher varied
widely: between + 0.18V (2,3-dimeth-
ylbenzoquinone) and about —0.40V
(alizarin); and many of these com-
pounds should be unable to oxidize
QH.
2. The quenching effect was inde-
pendent of the intensity of the ex-
citing light, and the regeneration of
Q in the dark after an illumination
period was found not to be affected
by menadione.
3. With relatively high concentra-
tions of quencher the fluorescence
yield was considerably lower than
that observed (in the absence of
quencher) under conditions where Q
could be assumed to be fully oxidized
(e.g., during efficient photosynthesis,
or shortly after onset of illumina-
tion) .
4. In several experiments both
with chloroplasts and with algae we
did not see any reversion by quinones
of the inhibition of oxygen evolution
by DCMU.
The above observations indicate
that the quenching effect is not caused
indirectly by stimulation of the rate
of electron transport, but rather by a
direct interaction of quinone and
chlorophyll molecules, which then act
as traps for the excitation energy.
o
c
0)
o
o
Z5
Porphyra perforata
DCMU
DCMU
6xlO"5M
1,4- Naphthoquinone
No DCMU
On
DCMU 4.8xlO"4M
1,4- Naphthoquinone
On
No DCMU
I.2xl0-4M
,4- Naphthoquinone
0
4 0
Time, sec
Fig. 17. Fluorescence kinetics of Porphyra perforata in the presence of DCMU (5 X 1 0-5 A4)
and 1 ,4-naphthoquinone. Fluorescence was excited by green light. The preceding dark intervals
were 1 8 seconds for the traces on the left, and 30 seconds for the traces on the right. The traces
on the right were recorded at about 20 times higher light intensity and lower sensitivity of the
apparatus than those on the left.
DEPARTMENT OF PLANT BIOLOGY
169
Since the fluorescence lifetime of
chlorophyll is longer when Q is in
the reduced state than when it is
oxidized (Miiller and Lumry, 1965),
a stronger quenching can be expected
for the total (and thus for the vari-
able) than for the initial chlorophyll
fluorescence, even if the quencher acts
only on the bulk of the pigment. It is
doubtful, however, if this could ex-
plain the large difference in the ex-
tent of quenching of the initial and
variable fluorescence, as observed at
low concentrations of menadione
(Fig. 16).
Additional evidence that the qui-
nones act at a site close to system 2
was given by experiments shown in
Fig. 18. Measurements of absorption
changes at 420 ni/x in Porphyra in-
dicated that menadione and other
quinones inhibited the reduction of
the /-type cytochrome by light ab-
sorbed mainly by system 2. Quinones
thus exhibited the same effect as well-
known system 2 inhibitors such as
DCMU.
In agreement with the results of
earlier experiments of Wessels
(1954) we found that a number of
quinones inhibited the Hill reaction
of chloroplasts with 2,6-dichloro-
phenol-indophenol (DCPIP). In gen-
eral, strong quenchers were more
powerful inhibitors of cytochrome
reduction and of the Hill reaction
than were weak quenchers. However,
a quantitative relation was not found.
Menadione gave 50% inhibition of
DCPIP reduction at 2 X 10"4 M,
about the same concentration as
needed for 50% quenching of the
initial fluorescence, but with 1,4-
naphthoquinone, lawsone, and phthio-
3.
E
O
C\J
c
o
_Q
i_
O
O
C
O
u
1 i i i i I i — i — i — i i i i — n — | — i 1 — i — i i i | — r~i — I — I — I th r
Porphyra perforata
No Menadione I0~4M Menadione
iRedh- -Dark — — I Red I Dark
AI/t +0.005
1
iRedl Green
^"i»'*'
y >(inw«W>»>^>»
Red!
Green
l i i i i I ' i ' i l i ' i i I '
t »» i ' i i i i i I i i
0
10
15 0
Time, sec
10
Fig. 18. Effect of menadione on the cytochrome absorbance changes in Porphyra perforata.
Red actinic light, to excite mainly system 1, had a half band between 670 and 780 m^t and green
actinic light to excite system 2 from 550 to 570 m/x. A downward trace corresponds to the oxi-
dation of cytochrome.
170
CARNEGIE INSTITUTION
col the effect on cytochrome and
DCP1P reduction was stronger than
that on fluorescence.
Emission spectra. We measured the
emission spectra of Porphyra with
green exciting light in the presence
and in the absence of 1, ^naphtho-
quinone and DCMU, using the appa-
ratus for automatic recording of
fluorescence emission spectra con-
structed by French (Year Book 65,
p. 493). The emission spectra (Fig.
19) showed three maxima at 658,
684, and 731 m/i. The first one is pro-
duced by phycocyanin and the second
by chlorophyll a. The maximum at
731 nijn was first observed by Duysens
(1951). As Fig. 19 shows, naphtho-
quenched the chlorophyll
684 ni/x and, somewhat less
that at 731 m/*, but had no
effect on the intensity of phycocyanin
fluorescence. This indicates that the
quinone does not interact with the
biliprotein chromophore, and that the
quinone
band at
strongly
transfer of energy between phyco-
bilins and chlorophyll is not affected.
DCMU alone caused about a three-
fold increase of the 684-m^ band and
a 2.5-fold increase of the 731-mju
band, but had no effect on the phyco-
cyanin emission.
If it is assumed, according to cur-
rent theories, that DCMU stimulates
only the fluorescence of pigment sys-
tem 2, then these data indicate that
the pigment fluorescing at 731 m/x be-
longs not only to system 1 (as indi-
cated by experiments of Duysens,
1951) but also to system 2. Naphtho-
quinone then quenches system 2
fluorescence, possibly by shortening
the lifetime of the excited state of
chlorophyll a fluorescing at 684 m^,
and thus decreasing both the 684-m/x
fluorescence and the extent of the
transfer of energy to the pigment
fluorescing at 731 m/x.
Conclusion. Our experiments dem-
onstrate that a number of substi-
100
0)
cS 75
u
c
CO
CD
o
_2
50
25
0
i 1 1 r
1 1 1 r
~i r
Porphyra perforata
+DCMU(I)
/
650
700 750
Waveleng+h.mjL/
Fig. 1 9. Fluorescence emission spectra, excited by green light, plotted as relative number of
quanta per frequency interval of Porphyra perforata in DCMU (5 X 1 0~5 M) and DCMU with
1 ,4-naphthoquinone (9 X 1 0-5 M). The broken line gives the difference between the two spectra.
DEPARTMENT OF PLANT BIOLOGY 171
tuted quinones strongly quench the thetic Bacteria, Special Issue of Plant
fluorescence of chlorophyll in vivo. and Cell Physiol, Jap. Soc. of Plant
The quenching in vivo occurs at a Physiologists, The Univ. of Tokyo
much lower concentration of quinone Press, p. 353, 1963.
than when the chlorophyll is dis- Livingston, R., and C. L. Ke, J. Am.
solved in organic solvent (Livingston Chem. Soc, 72, 909, 1950.
and Ke, 1950). The quenching is Muller, A., and R. Lumry, Proc. Natl
probably due to direct interaction be- Acad. Sci., U.S., 54, 1479, 1965.
tween chlorophyll and quinone mole- Seelv> G- R> in The Chlorophylls, edited
cules, rather than to stimulation of ^ L- P. Vernon and G. R. Seely Acad,
electron transport. It is possible that Press> New York-London, p. 523, 1966.
the actual quenching mechanism (see ^ {JL S* C*' Re°' TmV' Ch%m'> 73'
Seely, 1966, for a discussion) is the '
same in vivo as in chlorophyll solu-
tion. The stronger quenching in vivo Some Essential Considerations in
may be caused by a concentration of the Measurement and Interpreta-
the quinone in the lipoid part of the tion of Absorption Spectra of
chloroplast lamellae, and also may be Heterogeneous Samples
explained by the fact that quenching James M> pickett and c & French
of the same fraction of chlorophyll
molecules, by formation of traps, The following discussion of well-
could produce a stronger quenching in known but, of necessity, often neg-
vivo than in a dilute chlorophyll solu- lected factors is given here primarily
tion. The low activity of p'hthiocol as a review of basic principles essen-
and l,2-naphthoquinone-4-sulfonic tial to the study of absorption spectra
acid may be due to salt formation in of photosynthetic cells. These con-
the water phase. siderations are particularly important
There is no direct spectrophoto- for spectra that are to be analyzed in
metric evidence (Amesz, 1964) for terms of their component forms of
the hypothesis that a quinone is the chlorophyll.
primary photooxidant for system 2. Sieve effect. The principal source
However, the strong quenching of of error in measuring the shape of
chlorophyll fluorescence in vivo indi- absorbance curves with thin suspen-
cates that association of a quinone sions is that some light gets through
with a chlorophyll molecule can pro- spaces between the cells. This makes
duce an efficient trap for the excita- the value of transmitted light greater
tion energy. It is possible that a than it would be if the pigment com-
photosynthetic reaction center of plexes were uniformly dispersed in
system 2 consists of a similar trap, solution. The sieve effect makes the
supplemented by the enzymatic and Peaks of the absorbance curve lower
structural arrangement necessary for than they should be, relative to re-
the formation of stable photochemical gions of lower absorbance.
products. The error introduced by the sieve
effect becomes smaller as the number
References of particles increases because the
Amesz, J., Biochim. Biophys. Acta, 79, absorbance of a given wavelength for
257, 1964. various light paths then becomes more
Duysens, L. N. M., Nature, 168, 548, uniform throughout the suspension.
1951. However, at high suspension concen-
Duysens, L. N. M., and H. Sweers, in trations the effective path length is
Studies on Microalgae and Photosyn- reduced for wavelengths of high ab-
172
CARNEGIE INSTITUTION
sorbance relative to that for wave-
lengths of low absorbance (Butler,
1964). Thus peaks in absorbance
spectra of dense algal suspensions are
usually flattened because scattering
(Latimer, Year Book 56, p. 259) and
reflectance (Rabideau ctal, 1946) are
not independent of wavelength, while
the sieve effect is negligible.
The absorbance of a suspension
relative to that of the same pigment
in solution can be calculated for a
single layer of absorbing particles.
The calculations are complex for sus-
pensions of more than one layer. The
case of oriented cubic particles has
been described in detail by Duysens
(1956). The analysis in the case of
spherical particles is more complex
because the absorbance is not uniform
over the projected area of the spheri-
cal particles.
Sieve effect for a single layer of
uniform particles. Let / be the frac-
tion of the light beam that encoun-
ters a particle of transmission Tp. If
mutual shading of the particles is
negligible, the measured absorbance
Es of a dilute suspension of uniform
particles is
E8 = log ( ^ (1)
\a-f) +fTP)
But the apparent absorbance of a
single particle is
have in the same volume of solution
with a uniform path length.
Spherical particles. Consider a
sphere of radius r suspended in a
medium of equal refractive index il-
luminated by a circular beam of light
whose radius is equal to that of the
sphere. Assume that each spherical
particle contains a concentration c of
pigment whose absorption coefficient
is e. The absorbance along a light
path of length x through the sphere is
Ex — eCX (3)
The transmission along the same path
is
TB = 10-
(4)
All paths of length x through the
sphere are in a circle of radius y per-
pendicular to the incident beam of
light where
/ x
y* — r* .__ / —
»2 —
(5)
The absorbance E, which is propor-
tional to the extinction coefficient of
the pigment, and transmission Tp of
a single sphere are obtained by inte-
grating the weighted values of Ex
and Tx over all light paths (Duysens,
1956, and Pickett, 1965)
E-
* 2irydy
&x : —
2/=0
77-7"
Ev = log
(£)
(2)
= — ecr
(6)
The important conclusion is that the
measured absorbance of dilute sus-
pensions is not proportional to the
apparent absorbance of one particle.
Moreover, if all light paths through
the particle are not equal, Ep is not
proportional to the extinction coeffi-
cient of the pigment present in the
particle. Therefore, we propose to re-
view the types of errors generally in-
troduced by absorbance measurements
of suspensions relative to the absorb-
ance E that the same pigments would
Tv = IT
2irydy
-n-r2
lQ2ecr _ 4 606 eCr _ i
10.61 (ecr)2 102£Cr
(7)
from which
Ep = log
10.61 (ecr)2102<
lQ2ecr _ 4>606 ecr — 1
(8)
It is clear from equation 8 that Ep
is not proportional to e, the absorp-
DEPARTMENT OF PLANT BIOLOGY
173
tion coefficient of the pigment within
the sphere. The ratio of the apparent
absorbance of a single sphere, Ep, to
the absorbance of the uniformly dis-
persed pigment E is given in Fig. 20,
for various values of 2ecr, the absorb-
ance through the center of a single
sphere.
The transmission (10_2ecr) for 675
ni/A of single Euglena chloroplasts as
measured by Wolken et al. (Strother
and Wolken, 1959; Wolken and
Strother, 1963) falls in the range of
0.4 to 0.6 (2ecr = 0.6 to 0.8) and for
Chlorella, about 0.6 (2ecr = 0.8).
The corresponding ratio Ep/E is 0.94
to 0.92. Flattening of the absorption
peak may be much less than 6% to
8% in these microspectrophotometric
measurements because the absorption
spectrum was determined by measur-
ing the transmission of a single
chloroplast with a beam of light whose
cross section was small compared to
that of the chloroplast.
In practice it is generally more
convenient to measure the transmis-
sion of a relatively dilute suspension.
The ratio of the measured absorb-
ance Es of a suspension with negli-
gible overlap of particles to the ab-
sorbance E of the uniformly dispersed
pigment is plotted as a function of
the absorbance through the center of
the sphere, 2ecr, in Fig. 21. For par-
ticles with a maximum absorbance
(2ecr) of greater than 0.5, the ab-
sorbance maxima are reduced by more
than 10%. The error becomes greater,
the smaller the fraction / of the light
beam intercepted by particles. Thus
accurate absorbance spectra of dilute
suspensions can only be obtained for
particles of very low maximum ab-
sorbance.
Dense suspensions. The sieve effect
is relatively insignificant in very tur-
bid suspensions, since the effective
path length is much greater than the
thickness of the suspension. However,
the path length is not increased to
the same extent at all wavelengths.
The principal variable influencing
effective optical path length is particle
reflectance (Butler, 1964) . Anomalous
dispersion, which is maximum on the
long-wavelength side of absorption
bands (Latimer, Year Book 56, p.
259), is relatively insignificant except
when a very small solid angle of light
is collected by the spectrophotometer
(Butler, 1964).
c
CD
E
Cn
LjJ
O
1 1 1
1.0
-*— *
CD
CD~
CO
C_>
•~.
c
J_
CO
03
-Q
5—
o
0.8
o
CO
"ca
0
C£
1 T
i r
j L
i i i
1.0 2.0
Absorbance through sphere center, 2ecr
Fig. 20. The apparent absorbance Ep of a single spherical particle divided by the absorbance
E of the pigment in the same cylindrical volume of solution with path length 1 .33 scr, is plotted
as a function of the absorbance through the center of the sphere 2 scr. The spherical particle of
radius r contains a uniform concentration c of pigment with absorption coefficient e.
174
CARNEGIE INSTITUTION
E
1.0 2.0
Absorbance through sphere center, 2ecr
Fig. 21. The measured absorbance E8 of a
thin suspension of spherical particles divided
by the absorbance E of the pigment in solution
of equal cross section and path length, 1.33
ecr[f), as a function of absorbance through the
center of a spherical particle, 2 ecr. Each
curve is characterized by f, the fraction of the
light beam which encounters a particle.
According to Butler (1964) accu-
rate absorption spectra of dense sus-
pensions can be obtained by correct-
ing the measured spectrum for
variation of effective path length with
wavelength. The flattening may also
be greatly reduced by suspending the
pigmented particles in a much greater
concentration of white powder.
In summary, there are three prin-
cipal means of measuring in vivo ab-
sorption spectra which may closely
approximate the absorption spectra
of the pigments : (1) The absorbance
may be measured along a fixed light
path with a light beam whose cross
section is small compared to that of
the pigment complex (microspectro-
photometry). (2) The absorption
spectra of very weakly absorbing cells
such as pale-green mutants may be
measured by means of thin suspen-
sions in opal glass containers. (3)
According to Butler (1964) the ab-
sorption spectra of very dense sus-
pensions can be measured in the
presence of large amounts of white
powder or corrected for variation of
effective path length with wavelength.
The height ratio H. We need a rough
test for the likelihood of serious dis-
tortion in an experimentally measured
curve. For this purpose with green
plants we have used the height ratio
H, defined as the absorbance at the
secondary chlorophyll a maximum at
about 625 ni/x divided by the height
of the red absorbance peak at about
675 m/x.
As a standard of comparison we
have taken the value of H for the
alga showing the smallest ratio.
This is Claes Chlorella mutant 871
grown in the dark, which appears to
have nearly all its chlorophyll as
Ca672 and lacks chlorophyll b. For
that alga H = 0.28 at -196°C.
In an alga containing two forms of
chlorophyll with absorption spectra
of similar shape but shifted in wave-
length, the value of H would be larger
and would depend on the wavelength
separation interval. This variation of
H was obtained graphically for a hy-
pothetical system having two identical
spectra but shifted in wavelength by
various amounts. The basic curve
shape was taken as the sum of the 625
and 671 components that fit the ab-
sorption curve of Fig. 26 for dark-
grown Chlorella mutant 871. This
basic curve is the closest approxima-
tion to the actual shape of the spec-
trum of Cfl670 that we have. In
calculating H for the hypothetical
two-component spectrum, curves of
the same height were used. For any
other proportion the change of H with
the separation interval would be
smaller.
The results in Table 4 show that
the ratio in question would be only
0.36 for the maximum separation in-
terval of 13 rn.fi expected between "Ca
670" and "Ca680." Ratios for experi-
mental spectra of algae containing
only chlorophyll a that are larger
than 0.36 can therefore be taken as
an indication of distortion by the
flattening effects.
DEPARTMENT OF PLANT BIOLOGY
175
TABLE 4. The ratio of the Orange, "625"-
mfji, Peak to the Red, "675" m/x, Peak of a
Hypothetical in vivo Chlorophyll a Spectrum
Composed of Equal Parts of Two Identical
Components at Several Wavelength
Separation Intervals
Separation
Interval
Height
Ratio of Orange
m/j,
to Red Maximum, H
0
0.29
5
0.30
10
0.32
13
0.36
15
0.39
References
Butler, W. L., Ann. Rev. Plant Physiol.,
15, 451, 1964.
Duysens, L. M. N., Biochim. Biophys.
Acta, 19, 1, 1956.
Pickett, J. M., Thesis, U. Texas, Austin,
1965.
Rabideau, G. S., C. S. French, and A. S.
Holt, Am. J. Bot., 33, 769, 1946.
Strother, G. K., and J. J. Wolken, Sci-
ence, 130, 1084, 1959.
Wolken, J. J., and G. K. Strother, Ap-
plied Optics, 2, 899, 1963.
A Spectrophotometer Primarily
for Light-Scattering Samples
at Low Temperature
C. S. French and Mark Lawrence
In Year Booh 65 (p. 498) a plan
was discussed for modifying the de-
rivative spectrophotometer so it could
also be used as a double-beam instru-
ment for absorbance measurements
of highly scattering samples. The
integrating sphere then under dis-
cussion has been abandoned. Instead
of the sphere, a different type of sam-
ple chamber was constructed to be
particularly suitable for absorbance
measurements of cell suspensions.
To reduce scattering errors the ma-
jor requirement in the design of the
sample holder was that it should
gather a representative fraction of
the transmitted light from the sample
over a wide angle. Four other re-
quirements were that it should (1)
accommodate sample holders that
could be used at liquid nitrogen tem-
perature; (2) have a peak height of
the plotted absorbance curve for a
particular sample adjustable to any
desired value and still have known
scales in absorbance or transmission
units; (3) by means of an entirely
transistorized electronic measuring
circuit eliminate maintenance prob-
lems of vacuum tube amplifiers ; and
(4) be simple enough for routine use
by different people. The device, which
meets these requirements reasonably
well, has been in use since April 1967.
A brief description is given here.
Optical system. The sample holders
shown in Fig. 22 use Shibata's opal
glass diffusion principle for both
windows. The aluminum spacers also
serve as heat-conducting elements and
supports. Since the opal glass win-
dows are held on with silicone grease
and oiled modeling clay only, they do
not break when the sample freezes.
A holder for the usual 1-cm-square
liquid absorption cells can be used
in place of the opal glass sandwiches
when the material is not to be frozen.
The sample chamber is shown in
Fig. 23; just behind the sample holder
is another opal glass window. Between
this window and the photocell is a
light-mixing tube with white walls.
This tube helps to give uniform illu-
mination on the photomultiplier tube
for light falling on any part of the
opal glass window at its entrance.
The brass box in the sample chamber
is thermally insulated by 2.5 to 4 cm
of plastic foam. The lower part of
this box can hold 100 cc of liquid
nitrogen or other liquid for operation
at a desired temperature. The en-
trance window is lucite 2.54 cm thick,
preceded by a double-walled glass
chamber to reduce fogging at low
temperature. A turret can be rotated
to bring any one of three multiplier
176
Sample cell
CARNEGIE INSTITUTION
ce
yj \J
i"
Opal glass windows
Plasticene
cone grease seal
Liquid NL level
Aluminum spacer »
Fig. 22. The sample holder — an opal glass sandwich. The aluminum spacer and support is also
a heat conductor.
tubes into position for different spec-
tral regions. A filter-holder attach-
ment can be used in front of the
photomultiplier for special experi-
ments.
The optical system, long in use for
derivative spectrophotometry, has
been kept. This is a tungsten lamp
with a Bausch & Lomb monochro-
mator having a 10 X 10 cm grating
blazed for green. The slits isolate a
beam of 1 rrnx half -width. Glass color
niters are used to reduce the stray
light which is inevitable from a single
monochromator. The large monochro-
mator is particularly valuable in that
plenty of light is available for meas-
urements of scattering samples with
three sheets of opal glass in the beam.
A rotating half mirror alternates
the beam from the sample to the ref-
erence cell about 10 times a second.
Adjustable sectors on the mirror
shaft pass light from separate bulbs
to photocells that activate the appro-
priate electronic circuits while either
the sample or the reference cell is
illuminated.
Electronics. The electronic measur-
ing system uses solid-state opera-
tional amplifiers throughout. The first
stage is a preamplifier with auto-
matic gain control that makes opera-
tion possible over a wide range of
input currents from the photomulti-
plier. In Fig. 24 two block diagrams
show the operating principles and
controls for recording absorbance or
transmittance. The operating zero-
position control, not shown, is ad-
justable to facilitate recording of the
difference spectra between two sam-
ples.
In many spectrophotometers the
logarithm needed for absorbance
plots is generated by a logarithmic
slide-wire in the recorder. Change of
scale can then be done only by chang-
DEPARTMENT OF PLANT BIOLOGY
177
Fig. 23. The sample chamber.
ing slide-wires. In the instrument de-
scribed the logarithm is produced by
a transistor in the feedback circuit
of an operational amplifier. The loga-
rithmic signal appears as an electrical
voltage that can be amplified or at-
tenuated. This makes it possible to
adjust the scale of the recording to
compare curve shapes for different
samples.
When the "curve size" control is at
its extreme position, the 10-inch full-
scale deflection of the recorder cor-
responds to the value of absorbance
or transmission selected by the range
switch. On any scale the actual plot
can be reduced in size by the curve
size control. When this is done cali-
bration marks for either absorbance
or transmission can be put on the
record by setting a "calibrator" dial
to the values desired.
Absorption Spectra of Chloro-
phyll a in Algae
C. S. French
Much of the present-day research
on photosynthesis is concentrated on
two closely related questions. One of
these is to find means by which the
light energy absorbed by photosyn-
thetic pigments is transferred from
one pigment to another, then trapped
and converted to chemical energy as
a change in the oxidation state of a
reaction center. The other problem is
to define the succeeding reactions of
specific substances in the pathways
of electron flow from the primary re-
action center that makes products
useful to plants.
There are two types of primary re-
action centers believed to contribute
energy to the electron flow system at
178
CARNEGIE INSTITUTION
ejiM
•\WW\Af
c
E
o
k_
D
«/>
D
<D
E
<D
£
G
-Q
O
D)
C
o
E
o
»-
D)
O
_o
D)
DEPARTMENT OP PLANT BIOLOGY
179
8JIM 8p||s japjODay
0
O
o>
c
a
c
3
U
L.
u
a
c
k.
D
t/>
o
<D
E
u
c
o
i_
u
o
a
d)
D
-C
U
«4-
u
O
c
E
o
</>
i_
Ol
fc
o
</i
l:
-o
0
1-
_*
+-
o
<1>
o
£
CQ
0
*"
s.
c
CQ
0
CN
a
o
D)
J2
u-
a
180 CARNEGIE INSTITUTION
two points in a series-linked system lated with well-measured absorbance
of redox compounds. These two reac- spectra for the same cultures.
tion centers are activated by different Ideally we would like to be able to
pigment systems each of which con- separate each chlorophyll a form in
tains predominantly, but not exclu- a pure state without change of its
sively, one of the different "forms" spectral properties, measure its ab-
of chlorophyll a and other pigments, sorbance and fluorescence, and also,
Thus system 1 is largely powered by for a well-defined reaction, its action
light absorbed by C(2680 and system spectrum. Work on the difficult prob-
2 by light absorbed by chlorophyll b, lem of separating the chlorophyll
C«670, and the "accessory pigments," forms is progressing in several lab-
carotenoids and phycobilins. oratories and some of it is described
Considering the large amount of elsewhere in this report.
intensive work being done on the The present report will discuss
functioning of these photosynthetic some attempts to derive absorption
systems it is remarkable that so little spectra of the individual forms of
is known about the absorption spec- chlorophyll a by analysis of new meas-
tra of the individual pigments ini- urements of absorption spectra of
tially catching the light used to drive algae, and mainly of those algae lack-
the two systems of photosynthesis, ing chlorophyll b. From past at-
Most of the available information on tempts, largely by derivative spec-
the partition of absorbed light be- troscopy, it seems that each of the
tween the two systems has come, not forms designated as Ca670, Co680,
from absorption spectroscopy, but Ca695, etc., may represent classes of
from measurements of action spectra, pigment complexes rather than defi-
About all we know is that in green nite chemical compounds with identi-
plants significantly more light is ab- cal spectra in different plants. This
sorbed by system 1 than by system 2 conclusion is, however, by no means
at wavelengths beyond 685 m/x. How- certain because small amounts of un-
ever, system 2 still absorbs some recognized components could well
light even at longer wavelengths be- shift the apparent position of the ob-
cause of the overlapping spectra of served curves away from their true
the pigments in the two systems, positions in the spectrum.
Furthermore, at 650 and 480 m/x, the Some absorption spectra of intact
bands of chlorophyll b, system 2 ab- algae. Spectra of some algae useful
sorption is larger than at other wave- for characterizing the forms of chlo-
lengths. Some system 2 action spectra rophyll a are given in Fig. 25; and
show a significant contribution from Table 5 lists values of H, the peak
C,;670 and some others appear to height ratio for comparison with
involve Ca680. Beyond this qualita- chlorophyll a in solution. Figure
tive information only a few attempts 25 (A) gives the spectra of Tribonema
have been made to specify the exact at two temperatures. These curves
proportion of any wavelength that are probably flattened but the sharp-
goes to each functional system. Some ening by low temperature clearly
few measurements of system 1 and shows the doublet structure of the
system 2 action spectra have been main red band, which is not evident
made with precision comparable to at 22 °C. The original record indicates
that attainable with absorbance meas- a possible, but uncertain, doublet
urements. However, action spectra structure of the 625 band, perhaps
for the two systems in various species with broad components at about 620
still have not been precisely corre- and 630 m/t. The bands at 496 and
DEPARTMENT OF PLANT BIOLOGY
181
TABLE 5. Ratios of the Orange Band Height
to that of the Red Band for Chlorophyll a
22°C,
-196°C,
Chlorophyll a
H
H
Ether
0.16
Acetone
0.19
Special Solvent*
0.30
Claes Chlorella
871 D
...
0.28
520 D
0.37
0.35
515 D
0.37
0.30
WT D
0.30 (0.36)
...
7r/bonema
0.38
0.50
Botrydiopsis
0.36
0.47
Highkin barley
0.47
0.66
Ochromonas
...
0.43
Ochromonas refrozen
0.36
* Freed and Sander, 1951.
463, and some of the absorption at
shorter wavelengths, are presumably
due to carotenoids, whereas the other
labeled bands are attributed to forms
of chlorophyll a.
Figure 25(B) gives the spectra of
Claes wild-type Chlorella at two tem-
peratures. The low ratio of the peak
absorbance at about 625 to about 670
mfi (H = 0.31, at 22°) and the very
small absorbance at 530 m/x shows a
surprising lack of flattening in this
spectrum for a green alga. The 486-
and 651-m/x bands are attributed to
chlorophyll b ; those at 492, 467, 460,
and some shorter absorption bands,
to carotenoids. The 433 and 437 peaks
might be due to different forms of
chlorophyll a.
In Fig. 25 (C) , we have spectra for
the alga Botrydiopsis and for a leaf
of Highkin's barley mutant, both
lacking chlorophyll b. The curves ap-
pear strongly flattened but show the
two major forms of chlorophyll a,
and the barley has a band at about
710 m/x.
Claes Chlorella mutants 515, 520,
and 871, when grown in the dark, con-
tain very little chlorophyll and no
chlorophyll b, thus providing excel-
lent material for the study of the
shapes of the chlorophyll a spectra.
Furthermore, mutant 871 appears to
be free of carotenoids absorbing in
the visible spectrum. Figure 25(D)
shows large blue peaks, believed to
be zeta-carotene, which obscure the
blue chlorophyll a spectrum. Figure
25(E) for mutant 520 shows longer
wavelength carotenoid spectra.
In Fig. 25(F) we have a close
approximation to the spectrum of
Ca672 with only a small percentage of
Ca680 and probably very little inter-
ference from carotenoids. Its shape is
discussed in detail below. An ethanol
extract of chlorophyll from mutant
871 was chromatographed by Dr.
Michel- Wolwertz who found two frac-
tions— ordinary chlorophyll a and a
component spectrally similar to, but
chromatographically distinct from,
pheophytin. This suggests that the
spectrum for mutant 871, while it is
the simplest we have yet seen, may be
composed of other components be-
sides Co670. She found the other
Claes mutants grown in the dark to
contain only chlorophyll a.
Analysis of observed spectra. Our
ultimate hope is to derive curve
shapes for spectra of the individual
chlorophyll forms that, when added
together in correct proportions,
would match any observed spectrum
that was not distorted. Until that can
be done it is helpful to compare ex-
perimental spectra by the relative
proportions, peak positions, and half-
widths of the minimum number of
normal probability curves that, when
added together, match the measured
curves.
We have previously used deriva-
tives of probability curves to analyze
derivative spectra in order to ap-
proximate the relative content of
different forms of chlorophyll a
(Brown and French, 1959). The
curves for the major components
used for Chlorella were broader at
673 than at 683 m/x. Cederstrand et al.
182
CARNEGIE INSTITUTION
c
o
<
4C0
450
500 550 600
Wavelength,m/J
650
700
Fig. 25. The absorption spectra of various algae; all except the wild-type Chlorella lack
chlorophyll b. (A) Tribonema Sp. (Starr 639). The curves at two temperatures are directly compa-
rable. The sharpening at low temperature shows the red chlorophyll a band to contain two forms
of chlorophyll a not evident in the room temperature spectrum. (With D. C. Fork.) (B) Claes wild-
DEPARTMENT OF PLANT BIOLOGY 183
(1966) have matched absorbance ure and the table. The peak wave-
spectra of Chlorella, plotted on a length position, the half -widths, and
wave number scale, with probability the proportion of the total area for
curves peaking at 668 and 683 m/x the gaussian curves of Fig. 26 are
and having equal half -widths of 21 given in Table 6.
m/x. For estimating the shape of the Mutant 520 appears to contain
narrow bands here considered we more Ca680 than do 871 and 515. Its
have used wavelength rather than best fit came from component curves
wave number scales. The discussion nearly like those for 871 and 515 but
will be limited to spectra measured slightly different, as shown in Table
at — 196°C that appear to be free of 6 and Fig. 26(D). Whether or not
flattening. these small differences of 1 to 2 m/x
Starting with two spectra having in peak position and bandwidth re-
most of the chlorophyll as Ca670 we quired to match the spectra of differ-
see in Fig. 26 (A) that for mutant 871 ent mutants are significant remains a
a single probability curve does not fit question.
the main peak. With the same prob- These curve analyses have pointed
ability curve at 672 m/x and 22 m^ out the previously disregarded fact
half -width, an attempt was also made that a large fraction of the light inci-
te match the spectrum of mutant 515 dent on a photosynthetic cell is ab-
as shown in Fig. 26(B). The misfit sorbed by the secondary chlorophyll
for both spectra is taken to mean that a bands having broad maxima in the
components other than a single chlo- 625-m/x region. Table 6 shows that
rophyll a form are present. 40 % of the total absorption from
Therefore, enough gaussian curves about 570 nut to the red end of the
were taken to give a close fit for mu- spectrum should be attributed to
tant 871 (disregarding the slight these bands.
error in the sensitive trough region The secondary orange band is pre-
at 650 m.[i) . The curves used are given sumably different in peak position
in Fig. 26(C) and Table 6. Then, with and width for the various forms of
the same peak positions and half- chlorophyll. It is so wide that wave-
widths as for 871, the heights were length peak differences probably
adjusted to come as near as possible amounting to about one fifth of its
to matching mutant 515 with unsat- width are obscured. There certainly
isfactory results that are not shown, are, however, differences in the shape
However, the slight changes in band of the orange bands in different algae
positions and widths recorded in that may be worth further study
Table 6 gave a good fit to 515, as after the variations in the red band
shown in Fig. 26 (D) . For mutant 515 are better understood,
the secondary component had to be A particularly interesting effect,
slightly different, as shown in the fig- susceptible to curve analysis, is re-
type Chloreila, dark grown. The curve heights are not directly comparable. The low value of
H, 0.31 at 22°, indicates very little flattening in this sample. (C) Botrydiopsis afpina and High-
kins mutant barley both lacking chlorophyll b. Both curves appear strongly flattened. (With D. C.
Fork.) (D) Claes Chlorella mutant 515, dark grown. The large blue peaks are probably zeta-
carotene. The red band is shown at about 20 X the scale for the blue. VV3/4 = 15 m/x, H =
0.295. Probably free from flattening. (E) Claes Chlorella mutant 520, dark grown. The 51 5, 482,
and 453 bands are probably carotenoids. W3/4 = 17 m/A, H = 0.345. Probably free from flat-
tening. (F) Claes Chlorella mutant 871 at two temperatures. The red band is nearly all Ca 672.
W3/4 = 14 m/x, H = 0.29. Probably free from flattening.
184
CARNEGIE INSTITUTION
u
oS
u
c
o
_Q
i_
o
I/)
_Q
<
600
650
700 600
Wavelength,m/J
650
700
Fig. 26. Analyses of the red chlorophyll a absorption bands of several Claes Ch/ore//a mu-
tants grown in the dark and of Ochromonas danica, in collaboration with Dr. Brown. The char-
acteristics of the gaussian component curves are listed in Table 1. The sum of these components
is shown as a dotted line where it deviates from the measured curve being matched.
ported in another section by Dr.
Brown. This is in Ochromonas danica,
an alga containing chlorophyll c
rather than chlorophyll b. The three-
peaked spectrum of the frozen sample
changed to a two-peaked spectrum
after thawing and refreezing. At-
tempts to match the three-peaked
curve with a small number of gaus-
sian curves were not fruitful. How-
ever, the spectrum of the ref rozen ma-
terial was fitted successfully as shown
in Fig. 26(F). By changing only the
relative heights, but not the peak
positions or widths of these curves
matching the ref rozen sample, it was
DEPARTMENT OF PLANT BIOLOGY
185
TABLE 6. Probability Curves Fitting Algal Absorbance Spectra
Component
Secondary
band 620
Chlorophyll c
695 680 670
612
Chlorella Mutant 871
Peak, m/A
695 681 672
625
Half-width, m/x
(11) 10 20
51
...
Proportion, %
1 3 55
Chlorella Mutant 51 5
41
Peak, m/jL
695 683 672
622
Half-width, m/x
(11) 11 21
53
...
Proportion, %
1 4 53
Chlorella Mutant 520
42
Peak, mtt
694 681 671
622
Half-width, n\[x
20 10 19.5
62
• • •
Proportion, %
3 12 39
Qchromonas Refrozen
46
Peak, mfji
696 686 671
633
612
Half-width, m/i
16 16 22
60
19
Proportion, %
14 7 39
34
6
not possible to match the three-peaked
curve of the same sample before
freezing. The nearest approach to a
fit is illustrated in Fig. 26(G). This
difference suggests that the spectra of
the components themselves were af-
fected by the treatment. Just from
looking at the curves before and after
ref reezing we had the impression that
the relative proportions of the bands
but not their characteristics had been
influenced. This does seem to be more
nearly true for a repetition of this
experiment with another culture
shown in Fig. 34 of Dr. Brown's
report.
Because mutant 871 seems to be
free of carotenoids we fitted the blue
part of its spectrum with gaussian
curves with the results given in Fig.
27. Whether or not these represent
only components of chlorophyll a re-
mains to be decided when other ca-
rotenoid-free samples are found in
algae or in preparations such as the
Si fraction shown in Fig. 30 of Mr.
Michel's report. Analysis of the blue
and of the red spectra of carotenoid-
less samples are needed to determine
which of the blue bands correspond
to the forms of chlorophyll that are
now identifiable only by their red
bands.
In conclusion we may say that the
present best available approximation
to the red part of the spectra of Ca
' 1 '
Mutant 871 y
i i i l
I i
//
// j
'4I8\ \
<\)
u
^^^ j
C)
\ /44l\\
c
O
/390\ /
O
V)
_Q
<
L-^i l i-
-^r I i >-
Vv467
CiV I """^
400 450
Wavelength,mji/
Fig. 27. The blue spectrum of Claes Chlorella
mutant 871 is approximately matched by the
sum of the following gaussian curves described
by peak position, half-width and percentage of
the total area of their sum: 390 m/x, 27 m/x,
27%; 418 m/x, 29 mti, 47%; 441 mtt, 18 mxt,
24%; 467 m/x (18 m^), 2%.
1S6
CARNEGIE INSTITUTION
"670" is described by the sum of two
gaussian curves, with the following
characteristics: (1) A max = 622 mp,
\\\ „ = 60 imt; (2) A max = 672,
Wa s = 22 m,u, band height ratio (1)/
(2) = 0.27 (because of overlap, the
band-height ratio, (1) / (2) , is not ex-
actly the same as the total curve-
height ratio, H) .
References
Brown. J. S., and C. S. French, Plant
Physiol, 54, 305, 1959.
Cederstand, C. N., E. Rabinowitch, and
Govindjee, Biochim. Biophys. Acta,
126, 1, 1966.
Freed, S., and K. M. Sancier, Science,
114,275, 1951.
Electrophoretic Study of the
Chlorophyll-Lipoprotein
Complexes of Euglena
J. M. Michel
Many investigations, some from
this laboratory, have shown that the
red part of the absorption spectrum
of green organisms results from the
juxtaposition of at least three dis-
crete forms of chlorophyll a, presum-
ably lipoprotein complexes. They are
identifiable only from differences in
their absorption spectra. We know
very little about their actual chemical
properties. We do not know if the
differences between these substances
are due to different arrangements of
chlorophyll bound to the same carrier,
to different carriers for the same
pigment molecule, or to some other
set of conditions.
Most of the attempts to separate
these forms have been by differential
centrifugation using different ways
of solubilizing the chloroplast mate-
rial. Only a few attempts have
used electrophoresis, although zone
electrophoresis methods are now
available with properties particularly
adapted for the study of such com-
ponents. Among these methods, acryl-
amide gel electrophoresis is of par-
ticular interest because it is possible
to tailor the size of the pores inside
the gel to the size of the particles
being studied. (Ogawa et al., 1966;
Thornber et al., 1967). We used ac-
rylamide gel electrophoresis to study
the protein-chlorophyll complexes of
chloroplast fragments of Euglena.
Methods
Euglena gracilis, Indiana Culture
Collection No. 752, grown for 21 days
in low light intensity, was used. When
so grown, Euglena contains an ap-
preciable amount of a longer wave-
length absorbing chlorophyll form,
Ca695 (Brown and French, 1961).
We prepared the chloroplast frag-
ments using the needle-valve disinte-
grator and the anionic detergent Na-
deoxycholate (DOC) as a solubilizing
agent according to the method of
Brown et al. (1965).
This method gives two fractions on
centrifugation for 30 minutes at
30,000 g; a sediment called here Sed
30,000 and a supernatant called Sup
30,000. Prior to electrophoresis, the
fractions in tris-HCl buffer 0.01 M,
pH 8.9 were mixed with an equal vol-
ume of a solution of sucrose 40% in
tris-glycine 0.022 M, pH 8.9.
Disc electrophoresis was performed
on the fractions Sed 30,000 and Sup
30,000, according to the method of
Steward (1965). Before electrophore-
sis the sample was layered on top of
the concentration gel; the buffer in
the reservoir was tris-glycine 0.02 M,
pR 8.9.
Results
The electrophoretic patterns of the
two fractions Sed 30,000 and Sup
30,000 are shown in Fig. 28.
Sed 30,000 did not penetrate the
gel. The stationary band is labeled S0.
The Sup 30,000 fraction, however,
showed two clearly separated bands
DEPARTMENT OF PLANT BIOLOGY
187
Sup 30,000
Sed 30,000
S0
b,. ,
So **
+
No Amido-
stain black
No
slain
Amido-
black
Fig. 28. Electrophoretic patterns of Sed
30,000 and Sup 30,000.
Sx and S2 moving to the anode and
some material which did not move
into the gel. This S0 band could be a
contamination of Sup 30,000 by par-
ticles of Sed 30,000.
The Si band was blue-green and the
S2 band was green with a yellow
front ; under an ultraviolet lamp only
the S2 band showed a visible red fluo-
rescence. Figure 28 shows also the pat-
tern after coloration of the proteins
by the specific dye Amido Black B10.
This shows the absence of other
mobile proteins in the centrifugal
fractions used for electrophoresis.
The red part of the low-tempera-
ture absorption spectra of the bands
Si and S2 as well as the sample stay-
ing at the starting point for Sed 30,-
000, are shown in Fig. 29. The absorp-
tion spectra of the bands Sx and S2 at
room temperature are given in Fig.
30 for the entire visible spectrum.
The absorption spectra of Si and S2
are almost identical in the red, hav-
ing a maximum at 670 mp and lack-
ing the "long-wavelength" absorbing
form. There is, however, a minor
difference in that Si has relatively
higher absorbance than S2 near 660
and 690 m/x and less at about 670 m/ju
Sed 30,000 has the same absorption
spectrum as the original extract; it
contains most of the chlorophyll b.
Figure 30 shows clearly that the
band S2 contains most, if not all, of
the carotenoids, whereas the band Si
appears to be free of carotenoids. This
result is similar to that of Ogawa et
al. (1966) who found a differential
separation of the carotenoids using
sodium laurylsulfate as solubilizing
agent for spinach chloroplasts. Brown
et al. (1965) achieved a similar sepa-
ration by differential centrifugation
of DOC-treated Euglena chloroplasts.
The fluorescence spectra of the
bands Si and S2 are reported in Fig.
31; the maximum of emission is at
680 m^. When compared with the
original extract and Sed 30,000 the
large emission band at 720 m/x dis-
appears in the fluorescence spectra
of Si and S2. Comparison of the ab-
600 650 700
Wavelength, m/j
Fig. 29. Absorption spectra at liquid nitrogen temperature of electrophoretic fractions.
188
CARNEGIE INSTITUTION
1
-1 J 1 1 ! 1 J T— "
-mri i 1
— i J T — r
—j. — , — , — , ,_
i-
oT
u
c
o
o
1
\ N
\ \
\ \
I
! 1 ! ! 1 1 1
24°C
! I t i i
<
i i i i
1 1 1 1 1 1 1 1
450
650
500 550 600
Wavelength.mjti
Fig. 30. Absorption spectra of electrophoretic fractions Si and S2 at room temperature.
700
sorption and fluorescence emission
spectra of Si and S2 gives direct
evidence that the 670-m^ absorption
maximum is correlated with the 680-
m/i emission maximum.
From these data we can conclude
that DOC solubilizes two components
from the chlorolast lamellae of
Euglena. The two components are
different in their electrophoretic mo-
bilities and in their pigments. Ca-
rotenoids are associated with the
S2 band, but the chlorophyll portion
of the pigment system is essentially
chlorophyll a with nearly the same
spectral properties in both cases.
From the different electrophoretic
mobilities it may be inferred that the
protein portions of the complexes are
different, whereas the spectra sug-
_
a>
u
c
u
<J.
o
•--
0
3
1 1
\ 1
I 1 1
\ 24CC
/\\-
°0
\
\
-
ll
1
1
1
ll
1
1
ll
ll
/I
/ /
/ /
\ s
\ N
\ S
!
C
!
V
s
1 1
s
650
700 750
Waveleng+h,m/j
Fig. 31. Fluorescence spectra of electropho-
retic fractions at room temperature.
gest that the chlorophyll a molecules
are bound to the protein in a similar
or identical manner. However, we
cannot yet exclude the possibility that
the S2 band could contain some lipids
dissolved in detergent micelles.
Two more green bands have been
separated by disc electrophoresis
from material treated with sodium
laurylsulfate. They are much less
abundant than Si and S2J which makes
their study more difficult.
To see if the different forms of
chlorophyll are chemically separable
entities or if the complexes exist
only in the highly ordered structure
of the lamellar membrane, we are now
involved in attempts to separate the
Ca672, Ca685 and the Ca695 forms as
such. Preliminary experiments show
that, with appropriate experimental
conditions, it is possible to separate
two different electrophoretic bands
from Euglena chloroplasts without
using any detergent. However, so far
these two bands having identical
spectra each contain all three in vivo
chlorophyll a forms.
References
Brown, J. S., and C. S. French, Biophys.
J., 1, 539-550, 1961.
Brown, J. S., C. Bril, and W. Urbach,
Plant Physiol, hO, 1086-1090, 1965.
Ogawa, T., F. Obata, and K. Shibata,
DEPARTMENT OF PLANT BIOLOGY
189
Biochim. Biophys. Acta, 112, 223-234,
1966.
Steward, F. C, R. F. Lyndon, and J. T.
Barber, Amer. J. Bot., 52, 155, 1965.
Thornber, J. P., R. P. F. Gregory, C. A.
Smith, and J. L. Bailey, Biochemistry,
6, 391-396, 1967.
The Chlorophylls Extracted
from Plants by Organic Solvents
Marie-Rose Michel-Wolwertz
Several studies reported in this
Year Book and elsewhere have dem-
onstrated the existence of different
chlorophyll forms in vivo (for in-
stance, Year Book 58, p. 278; Year
Book 59, p. 330; Year Book 65, p.
483). It is generally considered that
these "in vivo forms" differ in the ar-
rangement and (or) state of aggrega-
tion of one single type of chlorophyll
molecule with the lipoprotein carrier.
Meanwhile, Sironval et al. (1965)
showed that two chemically different
pigments, namely, chlorophyllide a
and chlorophyll a, were responsible
for the two forms C684 and C672 pres-
ent in vivo during the initial period
of greening of etiolated leaves.
Michel-Wolwertz and Sironval
(1965) have isolated several chloro-
phyllous pigments from ethanol ex-
tracts of Chlorella by means of paper
chromatography. These are in addi-
tion to the common chlorophylls a and
b (called ax and &i). They suggested
that these "satellites" might be re-
sponsible for the different absorbing
forms in vivo.
The purpose of the present study
was to test this possibility.
1. We tried to find out whether the
light regime received by a plant
would influence the pigment composi-
tion of its extract.
2. We studied the effect of the
presence of air or of nitrogen during
extraction and chromatography upon
the pigment composition of extracts
of different species.
Method
Chlorella pyrenoidosa, Pringsheim
(211/8b) and Euglena gracilis were
used for extraction of pigments. For
comparison we also used spinach
leaves obtained from local stores.
Chlorella and Euglena were ex-
tracted with boiling 95% ethanol. Im-
mediately after extraction, the pig-
ments were transferred to petroleum
ether (b.p. 45-60 °C) by the addition
of excess water. Spinach leaves were
extracted by grinding with sand and
pure acetone in a mortar. Pigments
in acetone were directly transferred
to diethyl ether by adding excess
water. Extractions were made in dim
white light. Pigments were separated
by two successive paper chromato-
grams. In this study, we considered
pigments of the chlorophyll a type al-
most exclusively, that is, pigments
separated by rechromatography of
the spot of chlorophyll a obtained
from the first chromatogram.
For some experiments, we used a
chamber to do both extraction and
chromatography under nitrogen.
Results
Pigment composition of extracts
from Chlorella pyrenoidosa cultivated
under various light exposures. Table
7 compares the pigment composition
of extracts from Chlorella pyrenoid-
osa cultivated under these conditions :
(1) 11 days in continuous light; (2)
11 days in continuous light, then 2
days in complete darkness. (In this
case the extraction was made in the
dark) ; (3) 11 days in continuous
light, then two days in complete dark-
ness, after which the Chlorella re-
ceived a 4-msec light flash before
extraction. (The extraction was made
in the dark immediately after the
flash.)
The relative proportions of the dif-
ferent pigments are about the same
in all the extracts. These results show
that in Chlorella pyrenoidosa the pig-
190
CARNEGIE INSTITUTION
TABLE 7. Pigment Composition of Extracts from Chlorella pyrenoidosa
Grown Under Various Light Regimes
Pigments, % of Total Chlorophyll a
Light Regime
ai
°2 + 3
a4
as
89.5
5.0
3.6
1.8
90.2
5.2
2.7
1.7
92.7
4.0
2.0
1.1
91.4
4.6
2.6
1.3
88.7
6.1
3.2
1.8
88.2
6.0
3.0
2.8
93.3
3.9
1.8
0.9
89.0
5.8
2.8
2.3
Light 1 1 days
Light 1 1 days + dark 2 days
Light 1 1 days + dark 2 days
+ 1 flash of 4 ms
merit composition does not vary ac-
cording- to the light regime received
by the algae.
Effect of oxygen during extraction
and chromatography on the pigment
composition of extracts of different
species. Table 8 gives the relative
proportions of the different pigments
found in extracts of Chlorella, Eu-
glcna, and spinach, when all the pro-
cedures of extraction and chromato-
graphy were done in air. The results
show that the proportions of the sat-
ellites are not the same for the three
species. For example, the same solvent
extracts of Euglena are rich in satel-
lites compared to those of Chlorella.
The fact that the proportions of the
satellites vary from one organism to
another had been interpreted as in-
dicating either that these satellites
were present in the living cells and
the proportions varied in the different
organisms studied or that the common
chlorophylls a and b were differently
altered from various plants during
extraction and chromatography.
In order to decide between the two
possibilities we extracted and chro-
matographed the extracts from Chlo-
rella, Euglena, and spinach, both in
the presence and in the absence of
oxygen. The results (Table 9) show
that extraction and chromatography
in the absence of oxygen reduces or
prevents formation of satellite chloro-
phylls. This is true also for the chloro-
phyll b type satellites. In the absence
of oxygen we found only bi, but in the
presence of oxygen, satellites b2, b3,
etc., were also present. The absence
of oxygen during extraction was
especially effective in preventing the
formation of satellites (Table 10).
TABLE 8. Proportions of the Different Pigments in Extracts from
Chlorella, Euglena, and Spinach Leaves
Pigments, % of Total Chlorophyll a
Material*
ai
Q2 + 3
a4
as
91.0
4.7
2.6
1.6
90.2
5.2
2.7
1.7
69.6
16.4
7.5
6.7
71.0
14.2
7 A
7.7
70.0
15.0
7.2
7.8
Chlorella pyrenoidosa
Euglena gracilis
Spinach leaves
93.6
3.9
1.2
1.3
* Chlorella and Euglena were extracted with boiling 95% ethanol; spinach leaves, with pure
acetone (extraction and chromatography were done in air).
DEPARTMENT OF PLANT BIOLOGY
191
TABLE 9. Influence of the Presence or Absence of Air During Extraction and
Chromatography on the Proportions of the Satellite Pigments
Pig
ments, °/
9 of Total Chi
orophyll a
Under
Air
Under
Nitrogen
Material*
a-i
C*2 + 3
a4
as
CM
°2 + 3
a4
a5
Chlorella
pyrenoidosa
90.9
4.7
2.3
1.7
100
0
0
0
Euglena
gracilis
71.2
16.2
6.5
6.2
98.2
1.2
0.6
0
Spinach leaves
93.6
3.9
1.2
1.3
95.2
2.8
1.1
0.9
* Chlorella and Euglena were extracted with boiling 95% ethanol; spinach leaves in pure
acetone. For each extraction the same sample was used under air and under nitrogen.
However, satellites were also formed
in the presence of oxygen during
chromatography (Table 11). In this
experiment the same volume of a
chlorophyll solution was deposited on
six paper chromatograms in the
presence of air. The time required for
depositing the solution was varied
from 1 to 10 minutes. It is clear from
the results that longer exposure to
air destroys a± and consequently
creates more satellite chlorophyll,
especially a2+3 and a5.
Conclusions
From these considerations, the con-
clusion is inescapable that all the sat-
ellites may be formed from the parent
compounds by chemical alterations
during extraction and also during
chromatography. This does not ex-
clude the possibility that enzymatic
alteration may occur during extrac-
tion. Such alteration might account
for the different proportions of satel-
lites in extracts from different spe-
cies. However, the extraction of spin-
ach was made by grinding in acetone
and this procedure was slower than
extraction in boiling ethanol.
It appears very unlikely that these
satellite chlorophylls are present in
living cells, since they are almost
completely eliminated if extraction
and chromatography are done in the
absence of oxygen. From this work it
appears that they are not related to
the different in vivo forms of chloro-
phyll a.
References
Michel-Wolwertz, M. R., and C. Sironval,
Biochim. Biophys. Acta, 9b, 330-343,
1965.
Sironval, C, M. R. Michel-Wolwertz, and
A. Madsen, Biochim. Biophys. Acta,
94, 344-354, 1965.
TABLE 10. Influence of Oxygen During Extraction and Chromatography on the
Proportions of the Satellite Pigments from Euglena gracilis
Atmosphere
Pigments, % of Total Chlorophyll a
During
Extraction
During
1 st Chroma-
tography
During
2nd Chroma-
tography
ai
02 + 3
a4
a5
Nitrogen
nitrogen
nitrogen
98.2
1.2
0.6
0
Nitrogen
nitrogen
air
96.2
2.0
1.1
1.0
Nitrogen
air
air
90.5
5.7
2.0
1.7
Air
air
air
71.5
16.0
6.5
6.3
192
CARNEGIE INSTITUTION
TABLE II. Quantities of the Different Pig-
ments Found on the Chromatogram as a Func-
tion of the Time in Air between Application of
the Chlorophyll Solution to the Paper and the
Start of Chromatography, optical density*
Relative Qi
jantities of the Chlorophyll a
Pigments
Time,
Min
ai
Q2+3
04
05
1
0.217
0.033
0.022
0.014
2
0.170
0.040
0.022
0.021
3
0.158
0.051
0.022
0.025
4
0.121
0.056
0.023
0.023
5
0.126
0.058
0.024
0.025
10
0.131
0.054
0.023
0.024
* The values given are the peak optical
densities of the separated pigment bands dis-
solved in 4 ml of ether, 1 -cm light path.
Chlorophyll Fluorescence
in Algae and Chloroplasts
J. S. Brown
Some fluorescence spectra of par-
ticular algae whose spectra show great
contrasts with each other were meas-
ured at 20 °C and at — 190 °C. Very
small amounts of the algae were sus-
pended in a wet paste of BaS04 to
reduce the difference in light scatter-
ing between the room temperature
and the frozen samples. The incident
light was 436 im*. The curves of Fig.
32 were all recorded at essentially
the same size to facilitate comparison.
However, the apparatus sensitivity
adjustments were recorded to allow
comparisons of the relative intensity
of emission at the two temperatures.
The ratios between the yields at
— 190 °C and at 20 °C are given in
Table 12.
Any one of these curves is pre-
sumed to be the sum of a number of
overlapping fluorescence spectra each
of which is characteristic of or
emitted by a particular form of chlo-
rophyll. My objective was to deter-
mine the basic fluorescence spectrum
for each of the pigment types present.
The amount of fluorescent light radi-
ated by any one pigment in compari-
son with that from other pigments de-
pends directly on the relative number
of incident quanta it absorbs or re-
ceives from other pigments, and in-
versely, on the rate at which the pig-
ment transfers quanta to other pig-
ments or photochemical traps in the
chloroplast. The efficiency of energy
transfer and trapping can vary with
temperature. Therefore a direct pro-
portionality between the amount of
any chlorophyll form, as determined
from absorption measurements, and
the intensity of its fluorescence emis-
sion cannot be expected. The correla-
tions sought are between the wave-
TABLE 12. Ratios of the Fluorescence Yield at -190°C to the Yield at 20°C for Selected
Wavelengths and for the Total Fluorescence between 650 and 770 m/x
Organism
680 m/x
691 mju,
720 m/jt,
740 m/x,
Total
Fluorescence
Phaeodactylum
High-light grown
Low-light grown
0.23
0.40
0.53
0.71
4.2
12.
1.5
5.1
1.5
4.3
Euglena
High-light grov/n
Lov/-light grov/n
0.17
0.05
0.21
0.05
2.9
1.6
3.3
2.1
1.2
1.5
Ch/ore//a
0.10
0.19
2.4
1.1
0.82
Ochromonas
0.23
0.64
2.2
1.7
0.93
Spinach
chloroplasts
0.50
0.55
2.9
4.5
1.3
DEPARTMENT OF PLANT BIOLOGY
193
length positions of absorption peaks
and their corresponding fluorescence
maxima.
Last summer Mr. Stephen J.
Fulder attempted to derive the fluo-
rescence spectra of the individual
forms of chlorophyll by comparisons
of a large number of fluorescence
spectra at low temperature measured
in different laboratories. His results
0
u
qS
o
c
CO
(J
co
0)
u
o
Lu
650
700
750 650 700
WaveIength,mjL/
750
Fig. 32. Fluorescence emission spectra of several algae and spinach chloroplasts measured at
20°C and -190°C. Excitation at 436 m/*.
194 CARNEGIE INSTITUTION
showed wide variations in the shapes (1964) ascribes a long wavelength
of the individual bands when they (730 ni/x) emission maximum ob-
were deduced from different sets of served from chlorophyll in solution.
data. The present work should pro- Brody (1962) observed this band at
vide a more coherent group of spectra low temperature and attributed it to
for such curve analysis. a chlorophyll dimer. However, Goed-
Some qualitative conclusions al- heer's experiments demonstrate that
ready have become evident. In the this band occurs in dilute as well as
first place the composite nature of the concentrated solutions and at both
major 685-mp peak (Year Book 58, 20°C and — 196°C. It is relatively
p. 328) as the sum of components at greater in more concentrated solu-
longer and shorter wavelengths ap- tions because of self -absorption of the
pears even more plausible. Secondly, main emission band at shorter wave-
since these measurements were made lengths. Thus a significant proportion
with a very dilute suspension of algae of long wavelength fluorescence in
in a thick paste of BaS04, it is not vivo may arise from the second vibra-
likely that the difference between tional level intrinsic in the chlorophyll
spectra at the two temperatures can a molecule. This hypothesis to account
be attributed entirely to reabsorption for a long wavelength emission band
distortion by greater light scattering requires no corresponding long wave-
at low temperature. Therefore the length absorption band. On the other
actual lowering of the fluorescence at hand, if the emission is from a dimer
680 and 691 m/x shown in Table 12 is or aggregate as Brody suggests, an
believed to show an increased effi- absorption band for the dimer must
ciency in transfer of energy to longer exist. Possibly Ca695 is aggregated
wavelength forms at low tempera- chlorophyll, but the nature of this
ture. absorbing form of chlorophyll has not
Long ivavelength emission. The yet been demonstrated. The wave-
following evidence shows that there length position of the long wavelength
are several different long-wavelength fluorescence band in a particular
emission bands in addition to F-680 sample depends on the proportion of
and F-690. In Euglena gracilis, Phae- the two types of fluorescence.
odactylum tricornutum, and Ochro- Short wavelength emission. Duy-
monas danica, chlorophyll a-695 may sens (1952) first provided evidence
accumulate to a proportion of the that the chlorophyll a absorbing at
total chlorophyll easily detectable in longer wavelengths was largely non-
absorption spectra, but it is also more fluorescent in red and blue-green
labile than the chlorophyll forms ab- algae. More recently Duysens and
sorbing at shorter wavelengths. When Sweers (1963) correlated the major
the amount of Ca695 is relatively portion of fluorescence with the ac-
large, the fluorescence band F-720 is tivity of photosystem 2 in which the
also more evident, and when it is par- shorter wavelength absorbing chloro-
tially destroyed F-720 also decreases, phyll a is considered to be functional.
No one has yet observed an absorp- These findings led to the assumption
tion band that would be a likely that Ca670 is the source of F-685,
source of emission beyond 725 m/x the main fluorescence emission maxi-
which is so evident in spinach chloro- mum at room temperature.
plasts and Euglena. Fluorescence be- The variability in position of the
tween 720 and 740 mti may result emission maximum in different algae
from the second vibrational level of at room temperature shown in Fig.
chlorophyll a to which Goedheer 32 and additional data from experi-
DEPARTMENT OF PLANT BIOLOGY 195
ments on Ochromonas discussed in at 20°C in the beginning. On the other
another section of this report indicate hand, the increases in yield at 720 and
that other forms of chlorophyll may 740, which were originally 2.9 and 3.3
contribute to the main emission band. (Table 12), were reduced to 1.2 and
A discrete fluorescence emission band 1.5 after a second freezing. These
between 690 and 700 m/x at — 190 °C yield changes are reflected in changes
has been observed in several labora- in shape of the emission spectra (as
tories. Goedheer (1964) has sug- illustrated in Fig. 34 for Ochro-
gested that Ca680 is the chlorophyll monas).
form which fluoresces at about 690 The disruptive thawing and freez-
rn.fi (F-690) both at room tempera- ing process has either lowered the
ture and — 190 °C. In Ochromonas the possibility of energy transfer be-
main emission maximum is at 690 m/x, tween pigment molecules or has de-
suggesting that in this case Ca680 stroyed the longer wavelength ab-
may be the major fluorescing chloro- sorbing pigment complex Ca695. Ab-
phyll form. sorption spectra of Euglena, Phaeo-
Thus we now have evidence that all dactylum, and Ochromonas, measured
three forms of chlorophyll a may at — 190 °C and again after thawing
fluoresce under physiological condi- and refreezing, do show a relative de-
tions but that the proportion of fluo- crease in long wavelength absorption,
rescence contributed by each form Figure 33 shows an example for
varies considerably with the algal Ochromonas.
species and growth conditions. The The structural or specially oriented
same conclusion was reached several nature of Ca695 is suggested. Only
years ago concerning the relative ab- those organisms that are rather eas-
sorption by each chlorophyll a form ily disrupted show a large change in
in different algae. shape of the emission spectrum at low
Energy transfer. Although no temperature after a single thawing
clearly consistent hypothesis concern- and refreezing. Chlorella or spinach
ing energy transfer emerges from chloroplasts suspended in isotonic
these studies as yet, some postulations sucrose buffer must be thawed and
can be made. These corroborate and frozen several times before the yield
extend those of Goedheer (1966). changes that are induced by cooling
Upon cooling, the yield of the are different from those observed
shorter wavelength fluorescence after the first measurement at
bands always decreases and that of — 190 °C.
the longer wavelength bands always An explanation for these observa-
increases. After the cells are damaged tions may be that at very low temper-
in some way such as heating to 50 °C ature the possibility of energy trans-
for 10 minutes, placing in 10% fer between Ca670 and Ca695 in-
ethanol overnight, or freezing and creases, resulting in a decreased
thawing, the yield changes upon cool- fluorescence yield at 680 m^ and an
ing are very different. For example, increased yield at 720 m/x. Possibly
the Euglena grown in high light some energy transfer between Ca670
showed slightly more than a fivefold and Ca695 occurs also at room tem-
decrease in fluorescence yield at 680 perature, and Cfl695 acts as an energy
m/x when the cells were first cooled collecting trap in the photochemical
from 20 °C to — 190°C. But when this sense.
sample was thawed and brought Last year we suggested that ac-
again to — 190 °C, the yield at 680 m/x tivity during photosynthesis of both
was twice as great as from the cells Ca670 and Ca695 might be observed by
196
CARNEGIE INSTITUTION
i I — r
First frozen
Thawed, refrozen v.
iii
■ i i
400
450
500 550 600
Wavelength, m/x
Fig. 33. Absorption spectra of Ochromonas danica measured at — 190°C. Solid line curve
measured with initially frozen cells; dashed curve from same sample, thawed and refrozen.
monitoring their respective fluores-
cence emission maxima. This should
be possible with either Eitglena or
Phacodactylum cultured in an appro-
priate manner. Now we can add that
with Ochromonas fluorescence from
Ca680 may also occur as a major band
at room temperature.
References
Brody, S. S., and M. Brody, Trans. Fara-
day Soc, 58, 416, 1962.
Duysens, L. N. M., Dissertation, Utrecht,
1952.
Duysens, L. N. M., and H. E. Sweers, in
Studies on Microalgae and Photosyn-
thetic Bacteria, Japanese Soc. of Plant
Physiologists, University of Tokyo
Press, Tokyo, p. 353, 1963.
Goedheer, J. C, Biochim. Biophys. Acta,
88, 304-317, 1964.
Goedheer, J. C., in Biochemistry of
Chloroplasts, edited by T. W. Goodwin,
Academic Press, London, p. 75, 1966.
Absorption and Fluorescence
Spectra of Ochromonas danica
J. S. Brown
The algal flagellate Ochromonas
danica has a red absorption spectrum
with three major components at 670,
682, and 692 m^. Furthermore, both
its absorption and fluorescence spec-
tra are very sensitive to mild distur-
bances such as freezing or gentle
warming (Allen et al. 1960, and Year
Books 57 and 58) . The older measure-
ments that produced these data were
all done at room temperature and
only derivative absorbance spectra
were recorded.
We have recently extended these
experiments with improved equip-
ment and have measured both absorp-
tion and fluorescence at — 190°C as
well as at 20 °C. The results, qualita-
tively similar to those reported ear-
lier, are more precise and useful for
comparison with absorption and fluo-
rescence spectra of other algae dis-
cussed in other sections of this
report.
Figure 33 shows absorption spectra
of an Ochromonas suspension. The
three adjacent bands evident in a de-
rivative absorptive spectrum are diffi-
cult to distinguish in an absorbance
spectrum at room temperature but
are clearly visible in the low-tempera-
ture spectrum. The experimental de-
rivative spectrum measured at 20 °C
was fitted previously with three
gaussian derivative components with
DEPARTMENT OF PLANT BIOLOGY
197
peaks at 670, 682, and 692 m/x. At-
tempts to fit the low-temperature in-
tegral absorption spectrum with simi-
lar gaussian component bands have
been unsuccessful.
Figure 32 in the preceding section
of this report shows that at 20 °C
Ochromonas has a single fluorescence
emission band at about 690 m^. At
— 190°C, a second broad emission
maximum appears at about 720 m^.
The aliquot used for these measure-
ments was from a culture in which
essentially all of the cells were ac-
tively motile. Occasionally spectra
had a shorter wavelength emission
band between 670 and 680 m^ in ad-
dition to bands at 690 and 720 nut.
The solid curve of Fig. 34 shows an
example with a distinct shoulder on
the short wavelength side of the 690-
m/x band. We found that the relative
height of the 675-imi band was cor-
related with the number of rounded,
nonmotile cells in the culture.
Thus a particular culture may show
all three fluorescence maxima at low
temperature, the relative height of
the 675-m/x band depending on the
number of damaged cells. It is, of
course, possible to damage all the cells
by freezing and thawing or by homog-
enizing with the needle-valve press.
After such treatment the emission is
nearly all at about 675 m^, as shown
in Fig. 34 measured at — 190°C, and
Year Book 58 (p. 329), measured at
20°C.
An absorption spectrum of cells
that have been frozen, thawed, and
cooled again to — 190 °C is shown in
Fig. 33. The shortest wavelength
band near 670 m^ increases, and the
two longer wavelength bands de-
crease after ref reezing. A comparison
of the absorption spectra in Fig. 33
with the fluorescence spectra in Fig.
34 before and after thawing and re-
freezing indicates that the three ab-
sorption bands are the sources of
emission of the three fluorescence
1 L_J L
650 700 750
Wave!engfh,m/j
Fig. 34. Fluorescence emission spectra of
Ochromonas danica at — 190°C, and after
thawing and refreezing. Excitation at 436 mju..
bands. In addition, changes in absorb-
ance throughout the visible spec-
trum are evident. When more com-
parative spectra are available, we
hope, by correlating some of these
changes with those in the red region,
to determine other absorption max-
ima for the forms of chlorophyll a.
Chlorophyll a-680 may be the pri-
mary fluorescent form in Ochro-
monas. Since Ca680 is thought to be
the major pigment in photosystem 1,
the correspondence between fluores-
cence and photosynthetic activity
may be different from that in other
algae where Co670, which functions
in photosystem 2, is the main fluores-
cing form.
Reference
Allen, M. B., C. S. French, and J. S.
Brown, in Comparative Biochemistry
of Photoreactive Systems, edited by
Mary Belle Allen, Academic Press,
New York, p. 33, 1960.
The Action Spectrum for
Blue-Light-Stimulated Oxygen
Uptake in Chlorella
James M. Pickett
Emerson and Lewis {Year Book
U0, p. 158; and 1943) found the rate
198 CARNEGIE INSTITUTION
of oxygen uptake by Chlorella py- of oxygen uptake stimulated by blue
renoidosa to increase after an ex- light in Chlorella pyrenoidosa.
posure to low-intensity blue light. Electrode method. We constructed
Stimulation of oxygen uptake by 480- a Teflon-covered oxygen electrode
m,a light was increased by keeping which has provision for repeated in-
cells in the dark for 90 minutes. Fol- jection of samples from a reservoir of
lowing an exposure to 480 m/a, the dark-adapted cells (Year Book 65, p.
rate of oxygen uptake remained 487). The ability to compare measure-
greater than the former dark respira- ments on aliquots of a single suspen-
tion rate for approximately 1 hour, sion of Chlorella greatly facilitated
Ried (Year Book 64, p. 399) re- determination of the action spectrum.
ported that Chlorella pyrenoidosa Optical setup. The image of a tung-
kept in the dark, then exposed to sten ribbon filament or a mercury arc
flashes of blue light (A < 540 m/x), was focused on the platinum cathode.
gave increased rates of oxygen up- Spectral bands were isolated by
take. The maximum increase in rate Balzers interference filters (half-band
of uptake occurred 6 to 10 minutes width approximately 10 m/x). Inten-
after the flash. This increase was sities were measured with a cali-
not inhibited by 5 X 10~6 M DCMU brated photocell. Far infrared was re-
[3- (3,4-dichlorophenyl) -1,1-dimethyl- moved by 15 cm of water,
urea]. Under optimal conditions the Culture conditions. Chlorella pyre-
extra amount of oxygen consumed noidosa (Pringsheim 211-8b) was
after a short-wavelength flash was grown at 20° C in 1-liter erlenmeyer
more than 500 times greater than the flasks above cool white fluorescent
amount evolved photosynthetically lamps giving 400 foot-candles. The
during the flash. flasks, containing 200 ml of Knops
Kowallik and Gaffron (1966) medium with added microelements,
found no blue-light stimulation of were aerated with 2% C02 in air and
oxygen uptake in the nonphotosyn- shaken. After 2 to 3 days the sus-
thetic alga Prototheca, which has pension containing 3 to 5 /xl cells/ml
little absorption in the visible. How- was centrifuged and the cells resus-
ever, after dark adaptation for more pended in medium without microele-
than 1 hour, the dark yellow mutant ments to give a final concentration of
of Chlorella vulgaris (211-llh, 50 til cells/ml. This suspension pro-
M.20), which also lacks chlorophyll, vided a layer 3 to 4 cells thick in the
did show the blue-light stimulation electrode chamber. A 3-ml aliquot of
of oxygen uptake. Kowallik and this suspension was placed in the
Gaffron found that the rate of dark reservoir and kept dark at 20 °C for
respiration decreased about 60% in at least 15 hours before the experi-
algae kept in a glucose-free medium ment.
for 6 hours. However, in blue light, Exploratory experiments to develop
the initial rate of oxygen uptake was the procedure of measurement. A
stimulated as much as 30 % for this typical response to a 5-second flash of
time. blue light is shown in Fig. 35. The
In the experiments the stimulation maximum rate of oxygen uptake,
of oxygen uptake was limited to reached 10 to 15 minutes after the
wavelengths shorter than 550 m>; flash, was measured by the maximum
therefore, the responsible pigment difference between the extrapolated
could not be a chlorophyll. These ob- dark current before the flash and the
servations led us to attempt to deter- measured current after the flash, as
mine the action spectrum for the rate shown in Fig. 35. The rate of oxygen
DEPARTMENT OF PLANT BIOLOGY
199
o
CM
o
m
Q-
c+_
F
O
0
<y
QO
n
U>
CC
u
O
JZ
u
x
G)
o
x +30
0 -
■30
Photosynthesis-
Biue light
5 sec,469m/j
Stimulation of
oxygen uptake
I
0
4
Time,min
8
12
Fig. 35. The rate of oxygen exchange before and after a 5-sec flash of 469 m/x (0.49 X 1 0 9
ein/cm2/sec = 1250 erg/cm2/sec). The cells were in the dark 47.5 hr before the experiment.
uptake remained high for at least 30
minutes. The maximum rate of oxy-
gen uptake stimulated by a 5-second
flash of 469 m^ was about 1 % of the
light-saturated rate of oxygen evolu-
tion.
The stimulation of oxygen uptake
by blue light developed slowly after
the cells were placed in darkness, as
seen in Fig. 36. Almost all data used
for the action spectrum were meas-
ured with cells which had been dark
for 15 to 48 hours. Figure 36 also
shows that it was necessary to correct
the measured rates of uptake for the
change in response which occurred
during each 6- to 8-hour experiment.
Preliminary experiments indicated
that 469 m/x was near the peak of the
action spectrum. Figure 37 shows
that the response to 469 m/x was
nearly linear up to 0.75 nein/cm2/sec.
These data provided the basis for
selection of approximately 0.5 nein/
30
Time in dark, hr
Fig. 36. Effectiveness of 5-sec exposures to 0.5 X 1 0-9 ein/cm2/sec of 469 m/i as a function of
time in darkness before the flash.
200
CARNEGIE INSTITUTION
CD
CD
CO
o.
Z5
CN
o
CD
1.5
Light intensity,
einsteins cm-2 sec-1 x 10~9
Fig. 37. Light saturation curve for uptake after 5-sec flashes of 469 m/x for cells kept dark for
48 hr. Corrected for change in response with time.
cm2/sec (1300 ergs/cm2/sec) of 469
nin. as the reference beam with which
other wavelengths were compared.
To compare the effectiveness at
several wavelengths it was necessary
to make measurements over a period
of several hours. With the tungsten
lamp the measuring sequence used
was: 469 m/x, Xu A2, A3, 469 m/x, A4,
A5, As, 469 m/x. We attempted to ad-
just the intensities at all other wave-
lengths to give approximately the
same rate of oxygen uptake as that
measured at 469 m/x. The intensities
available with the tungsten lamp were
sufficient from 401 to 496 m/x. The
mercury lamp gave sufficient inten-
sity at 366, 407, and 436 m/x.
Since the effectiveness at 469 m/x
changed with time, an interpolated
value was used to correct the results
at other wavelengths. The corrected
results were then scaled by normaliz-
ing to the average effectiveness at
469 m/x, or at 436 m/x when the mer-
cury lamp was used.
Action spectrum. The resulting
action spectrum is shown in Fig.
38(A) and is in general agreement
with Ried's recent results (personal
communication). The curve was
drawn to fit the mean effectiveness at
each wavelength. The bar at 469 m/x
represents the mean ±3 standard
deviation units (n = 39) . All but two
measurements at 469 m/x were within
two standard deviation units. The
scatter at other wavelengths is gen-
erally greater than at 469 m/x, which
probably indicates that some of the
scatter was introduced by the nor-
malization process. If so, the shape of
the action spectrum for a single sus-
pension must change somewhat with
time. The measurements with the
mercury lines are in good agreement
with the measurements made with
bands (half -band width approxi-
mately 10 m/x) isolated from tungsten
emission. The effectiveness beyond
550 m/x was always less than 5% of
the effectiveness at 460 m/x. The ac-
tion spectrum was not changed by us-
ing suspensions of 25 /xl cells/ml in-
stead of 50 /xl cells/ml.
The spectrum in Fig. 38(A) sug-
DEPARTMENT OF PLANT BIOLOGY
201
80
i r
i 1 r
i 1 1 1 1 1 r
A
350
400
450
Wavelength, m/j
500
550
Fig. 38(A). Action spectrum for blue-light-stimulated oxygen uptake in Chlorella. (B) Comparison
of riboflavin absorption in water with the action spectrum corrected for shading by a 1 0-/x thick
layer of cells.
gests that the pigment responsible
for the blue-light-stimulated increase
in rate of oxygen uptake is a flavin.
The near-ultraviolet peak at 375 nut
distinguishes the action spectrum
from that of a carotenoid. If a cis-
carotenoid were responsible for the
effect, this peak should occur below
330 mil.
Shading by chlorophylls and carot-
enoids would shift the observed peak
of the action spectrum toward the
red. The absorption spectrum of
whole cells was therefore measured in
the Beckman DK-2 spectrophotometer
with opal glass plates. The corrected
action spectrum (Fig. 38B) was esti-
mated by multiplying the observed
effectiveness (Fig. 38 A) by the re-
ciprocal of the transmission of a 10
/x-thick layer of cells (absorbance at
440 m.fx = 0.575). This correction for
shading, by a layer half the estimated
thickness used, shifts the visible ab-
sorption peak from 460 to 440 m^ and
introduces a shoulder at 470 m/x. The
location of the near-UV peak remains
unchanged at about 375 rn.fi. However,
the ratio of the visible to near-UV
peak is increased from 1.15 to 1.42.
The corresponding value for ribo-
flavin (Fig. 38B) is 1.17. Correction
for shading by a thinner layer would
have given smaller shifts. The fact
that the corrected spectrum lies
within the scattering of the measure-
ments explains why dilution of the
cell suspension from 50 /J cells/ml
(20^-thick cell layer) to 25AJ /ml
did not change the action spectrum
202
CARNEGIE INSTITUTION
significantly. It should be noted that
the absorption spectrum of flavopro-
teins is generally more complicated
than that of riboflavin itself. The
peak height ratio in different flavo-
proteins is variable, and shoulders
may occur on the short- or long-wave-
length sides of the visible peak.
Because of the scatter in the data,
one can only conclude that the pig-
ment responsible for the blue-light-
stimulated increase in rate of oxygen
uptake is probably a flavin.
Effect of poisons. After adding
10-5'" M DCMU in ethanol (final
ethanol concentration, 1%) the
steady-state rate of photosynthetic
oxygen evolution from 0.62 nein/cm2/
sec at 469 m^ was inhibited by 94%.
The rate of oxygen uptake caused by
a 5-second flash of 0.62 nein/cm2/sec
was not inhibited under the same
conditions (Fig. 39).
By contrast with DCMU, cyanide
poisons the light-stimulated uptake of
oxvgen much more than photosyn-
thesis. With 2 X 10-4 M KCN the in-
hibition of light-stimulated oxygen
uptake was 97 % , but steady-state oxy-
gen evolution only 28%. A further
indication that the prolonged increase
in oxygen uptake by blue light is re-
lated to respiration is presented in
Fig. 40. Addition of 1% glucose to
starved cells stimulated the blue-light
effect 63 % . In the presence of glucose
the rate of dark respiration was in-
creased by an amount 10 times
greater than the maximum rate of
light-stimulated oxygen uptake. The
ratio of the rate of respiration before
addition of glucose to that after addi-
tion of glucose is unknown. Only 21 %
of the increase in dark respiration
was rapidly inhibited by cyanide.
However, cyanide sensitivity of light-
stimulated oxygen uptake and the
sensitive part of respiration were
identical (Fig. 40).
Discussion. Our results probably
concern the same stimulation of oxy-
gen uptake by blue light measured
by Emerson and Lewis (1943), Ried
(Year Book 6h, p. 399), and also
Kowallik and Gaffron (1966).
Our data indicate that the pro-
longed increase in the rate of oxygen
uptake observed after exposure to
I
CD
+30-
on
Q-
E
ra
CD
CD
s 0
SZ
O
X
CD
CM
O
° -30
CD
ro
i
i
New
i
I
sample
Control X
t i
+DCMU >v
i
i
0
10 0 10
Time, min
Fig. 39. Effect of DCMU on oxygen uptake. Oxygen uptake after a 5-sec flash of 0.62 X 1 0~9
ein/cm2/sec of 469 m/j, light is not inhibited by 1 X 1 0-5 M DCMU, which inhibited oxygen evo-
lution 94%. Reservoir dark 72 hr.
DEPARTMENT OF PLANT BIOLOGY
203
OH
-a
CD
CI)
\._
ca
=3
<xT
F
^i
(U
•+— '
CO
a.
1
Z5
-I—'
SZ
C7>
o
CD
ra
CD
i_
o
cz
o
CD
-5 -4
KCN concentration, log M
Fig. 40. Effect of cyanide on light-stimulated oxygen uptake and respiration. Light-stimulated
oxygen uptake of starved cells (open circles) after 5-sec flash of 1 .2 X 1 0-9 ein/cm2/sec
of 469 mjx. Respiration (open circles) after stimulation by 1 % glucose.
small doses of blue light is activated
by a flavin. The action spectrum (Fig.
38A) is almost identical to that of
Kowallik (1967a) for increased oxy-
gen consumption by a chlorophyll-less
mutant of Chlorella exposed to con-
tinuous blue light. The effect does not
appear to be coupled to system 2 of
photosynthesis because the uptake
is not inhibited by DCMU. Kowallik
(1967&) found that light-stimulated
oxygen uptake in Chlorella had a re-
spiratory quotient of one. These re-
sults are consistent with our data for
cyanide sensitivity.
Although the data suggest that oxy-
gen consumption stimulated by blue
light is probably mediated by mito-
chondrial respiration, there is no evi-
dence that the responsible flavin is
directly involved in respiration.
Summary. The action spectrum for
the stimulation of oxygen uptake by
blue light was determined for Chlo-
rella pyrenoidosa. The action spec-
trum is limited to wavelengths
shorter than 550 nut and has peaks at
460 and 375 m/x. The peaks are
separated by a definite minimum at
400 rn.fi. The action spectrum indi-
cates that the responsible pigment is
a flavin.
The effect is not closely coupled to
system 2 of photosynthesis because
the uptake was not inhibited by
10-5 M DCMU. Light-stimulated oxy-
gen uptake may be closely related
to cyanide-sensitive dark respiration,
since both exhibit the same cyanide
sensitivity.
References
Emerson, R., and C. M. Lewis, Am. J.
Bot., 80, 165, 1943.
Kowallik, W., and H. Gaffron, Planta, 69,
92, 1966.
Kowallik, W., in Energy Conversion by
the Photo synthetic Apparatus, Brook-
haven Symp. Biol., Brookhaven Na-
tional Laboratory, Upton, N.Y., p. 467,
1967a.
Kowallik, W., Plant Physiol, 42, 672,
19676.
Intermediates in Phytochrome
Transformation in vivo and
in vitro
Winslow R. Briggs and David C. Fork
The plant pigment phytochrome, a
chromophoric protein, mediates a
wide range of physiological processes
in green plants from the algae
through the higher angiosperms. It
can exist in either of two spectrally
distinct forms : a red-absorbing form
Pr with maximum absorption be-
tween 650 and 670 im*, and a far-red-
204
CARNEGIE INSTITUTION
absorbing form Pfr with maximum
absorption between 715 and 735 m^.
The two forms are readily intercon-
vertible by red and far-red light. Both
have absorption peaks in the long
ultraviolet or blue, attributable to the
ehromophoric group, a linear tetra-
pyrrole. These latter peaks show spec-
tral shifts comparable to those in the
red and far red upon appropriate
illumination.
Previous work {Year Book 61*, pp.
406-412) has shown that a mixture
of high-intensity red and far-red
light causes rapid cycling of the pig-
ment from one spectral form to the
other, producing measurable steady-
state levels of spectrally detectable
intermediates in partially purified
phytochrome from dark-grown seed-
lings of Arena sativa L. Actinic light,
provided by a Sun-Gun lamp, was
passed through a 5-cm water filter,
and a Corning 2030 cutoff filter to
remove wavelengths below 620 m^.
Absorbancy changes were measured
at 513 m/x by means of previously
described instrumentation (Year
Book 63, pp. 435-441). The inter-
mediates were detected by their in-
crease in absorbancy at this wave-
length when the actinic beam was
turned on, and their subsequent ab-
sorbancy decrease when it was turned
off. Appropriate filters between the
sample and the phototube prevented
actinic light from reaching the photo-
tube, but allowed passage of the low-
intensity 543-m/>t measuring beam.
Except for Q10 experiments, sample
temperature was maintained at ap-
proximately 5°C.
A kinetic analysis of absorbancy
decay after illumination suggested
that two intermediates with different
rate constants were decaying in par-
allel to Pfr. Time course studies for
the formation of the intermediates
in the light showed that the more
rapidly decaying one increased to its
maximum concentration within 2 sec-
onds and then decreased to a steady-
state level approximately one half as
high by the end of 12 seconds of illu-
mination. The more slowly decaying
intermediate appeared less rapidly,
reaching its maximum level at about
12 seconds and then remaining there.
The decay constant r for one inter-
mediate was about 0.2 and for the
other about 1.1 seconds at 5°C. The
Q10 for dark decay (between 5°C and
25 °C) of the intermediates was be-
tween 1.9 and 2.3, although the over-
all Q10 for transformation of the
pigment in either direction was ex-
tremely close to 1.0. Thus, under the
conditions of the experiment, decay
of intermediates could not have been
limiting the rate of light-induced pig-
ment transformation. The two inter-
mediates studied appeared to corre-
spond to the two slowest-decaying
intermediates on the Pr or Pfr path-
way, as described by Linschitz,
Kasche, Butler, and Siegelman (1966)
from flash photolysis experiments.
Phytochrome intermediates in vivo.
It was of some interest to determine
whether the intermediates studied
with partially purified phytochrome
could also be detected for native phy-
tochrome in vivo, and if they could,
whether or not their kinetics were
similar to those in solution. For this
purpose we used approximately 1.5
grams of coleoptile tips between 2 and
4 mm in length, which were excised
from oat seedlings grown for 5 days
in complete darkness. Care was taken
to eliminate any fragments of the
primary leaves, since these contain
substantial amounts of protochloro-
phyll, which would interfere with the
desired spectral measurement. The
coleoptile tips were packed into a
cuvette approximately 2 cm in diam-
eter, forming a layer of tissue about 2
mm thick. The sample was spread
carefully in the cuvette to eliminate
any small light leaks. The cuvette was
then placed over the photomultiplier
DEPARTMENT OF PLANT BIOLOGY
205
tube and kept surrounded with ice
during the measurements. Although
it was not possible to monitor tissue
temperature directly, experience with
liquid samples showed that the aver-
age temperature was probably be-
tween 5°C and 8°C.
Figure 41 shows records of absorb-
ancy changes during and after three
different durations of actinic illumina-
tion. It is obvious that with the short-
est period of illumination the decay
rate is substantially higher than that
found after the two longer periods.
A plot of decay half-time against
concentration of intermediate at the
end of illumination shows a steep
increase in decay half-time with in-
creasing concentration of interme-
diate. These observations are in close
agreement with those previously ob-
tained with partially purified phyto-
chrome. In the experiment shown in
Fig. 41, light intensity was kept
constant, and concentration of inter-
mediates was varied by varying expo-
sure time. When the reciprocal exper-
iment was done (exposure time kept
constant and concentration of inter-
mediates varied by varying light
intensity) decay half-times remained
constant and were independent of
intermediate concentration. These re-
sults are also in agreement with those
previously obtained with the in-vitro
system. Since the signals obtained
with the tissue sample were smaller
and substantially noisier than those
obtained with phytochrome solutions,
it was not possible to do any further
kinetic analysis on the decay patterns
found. However, since the kinetic
data that were available closely
matched those previously obtained
with solutions, it seems reasonable to
conclude that the same two inter-
mediates were being observed in vivo
as had previously been observed in
vitro. Absorption spectra of this and
similar preparations of oat coleoptile
a.
e
en
LO
o
c
o
i_
o
en
O
O
<D
en
c
o
_c
u
n vivo
+ =3 sec
0
'•'II )ll ll
t =.75 sec
T
.004
f =.38 sec
4
10
6 8
Time, sec
Fig. 41. Absorbancy changes at 543 m/A induced in intact oat coleoptile tips by high-intensity
actinic light (620 to 800 m/x, 4 X 10s erg cm"2 sec"1).
206 CARNEGIE INSTITUTION
tips showed almost no protochloro- across the sample is far lower than
phyll (before initial illumination) or that in previous experiments with
chlorophyll (after illumination) but phytochrome solutions.
did have very high absorbancy at 667 A difference spectrum for inter-
niu after far-red irradiation or at 735 mediates in vitro. Partially purified
m,u, after red irradiation. Thus the phytochrome was obtained from
predominant pigment absorbing at dark-grown oat seedlings as de-
the longer wavelengths was clearly scribed earlier (Year Book 6U, pp.
phytochrome. 406-412) for investigation of the
One important difference between spectral properties of the intermedi-
the behavior of the intermediates in ates between 365 and 580 m/x. The
vivo and that in vitro should be source of actinic light was the same
noted : The steady-state level of in- as above, although the intensity was
termediates in vivo, per unit of meas- somewhat lower (2.9 X 105 ergs
urable phytochrome in the sample, is cm-2 sec-1) . Preliminary experi-
approximately five times as high as ments showed that high concentra-
that in vitro, although the measured tions of sucrose or glycerol sub-
light intensity at the sample surface stantially slowed intermediate decay,
in the current experiments was less allowing formation of considerably
than one half that used previously higher steady-state levels than ob-
with phytochrome solutions. One tainable in buffer alone. Figure 42,
could account for this difference in top two tracings, shows absorbance
steady-state levels in two ways: changes at 560 m/x for identical
either slower decay or more rapid amounts and concentrations of phy-
formation in vivo. A comparison of tochrome diluted 50% either with
intermediate behavior in the tissue buffer or with glycerol (7°C). At
and in solution (cf. Fig. 41 and Year this wavelength glycerol almost
Book 6J+, Fig. 31) reveals that both doubles the height of the signal. Thus
explanations obtain. Decay in vivo is the difference spectrum was obtained
somewhat slower, and rise time is both with glycerol, to obtain maxi-
significantly shorter. Since rise time mum signal size, and without, to
is independent of length of exposure determine whether or not glycerol
time for illuminations longer than 1 spectrally altered the intermediates,
second, the comparison for this param- The bottom two tracings of Fig. 42
eter is easily made: steady state is show the absorbancy changes found
reached in vivo in approximately 0.3 at 418 and 385 m/x in the presence of
second, while requiring well over 1 glycerol. A sharp absorbance decrease
second in vitro. Hence for phyto- was found at 418 m/x and increases at
chrome in the intact tissue, forma- 385 and 560 m/x. The complete differ-
tion of intermediates is facilitated ence spectra with and without 50%
and decay is hindered; however, glycerol are shown in Fig. 43. Each
speculation on possible mechanisms point represents the average height
is premature. The differences in rise of a minimum of six signals, obtained
time are actually more dramatic than in every case with a 3-second light
appears on first inspection of the period followed by 6 seconds of dark-
figures, since the total absorbancy of ness (the regime illustrated in Fig.
the tissue sample is far higher than 42). Though the signals with glyc-
that of the phytochrome solution, erol are significantly larger, the two
Thus in the current experiments the curves are otherwise essentially the
average light intensity integrated same, with minima near 418 m/x, max-
DEPARTMENT OF PLANT BIOLOGY
207
<J
c
o
.£}
u
O
(/)
_Q
O
O
0)
c
a
_c
O
30 45
Time,sec
Fig. 42. Absorbancy changes at several wavelengths induced in phytochrome solutions with or
without 50% glycerol by high-intensity actinic light (620 to 800 m/i, 2.9 X 1 05 erg cm-2 sec-1).
ima near 380 ni/x, and isobestic points
near 398 and 485 m/x.
Several points should be made
about these difference spectra. First,
they represent the difference in ab-
sorbancy between the intermediates
and Pfr, to which both decay. Second,
it is not possible to obtain values
above 580 m/x, since the technique re-
stricts one to measurements at wave-
lengths well below the range of the
actinic beam. Third, the spectra are
composite, representing both the fast
and slow intermediates previously de-
scribed. Though it should be possible
to separate these two components by
detailed kinetic analysis of the rec-
ords, such analysis has not yet been
done. Fourth, these difference spectra
are not identical with the difference
spectrum between Pr and Pfr. The
latter spectrum has a minimum at
about 410 m/x, a maximum at about
385 Hi/*, and isobestic points at 398
and 495 m/x. Finally, 3 seconds of il-
lumination were not sufficient to
cause the maximum possible absorb-
ancy changes in the glycerol prepara-
tion. Experiments that allowed the
change to go to completion gave much
larger changes but the spectra were
the same.
Reference
Linschitz, H., V. Kasche, W. L. Butler,
and H. W. Siegelman, /. Biol. Chem.,
2^,3395-3403 (1966).
208
CARNEGIE INSTITUTION
400
450 500
Wavelength,mjL/
550
Fig. 43. Difference in absorbancy between intermediates and Pfr from 365 to 580 m/x. The
absorbancy changes were measured from records similar to those shown in Fig. 42, by means
of the same actinic light.
EXPERIMENTAL TAXONOMY
INVESTIGATIONS
Growth Responses of Mimulus
Races and F1 Hybrids at the
Stanford, Mather, and
tlmberline transplant stations
Malcolm A. Nobs and William M. Hiesey
A segment of the long-range in-
vestigations on Mimulus is concerned
with the comparative performance of
parental and Fa hybrids of a wide
array of crossing combinations at the
three altitudinal transplant stations.
These studies, begun in 1961 and com-
pleted in 1966, will be briefly sum-
marized.
The questions. Earlier inquiries on
the genetic structure of ecological
races of Potentilla (Carnegie Publica-
tion 615) and Achillea (Year Book 51,
pp. 122-124) dealt mainly with
crosses between extreme altitudinal
and latitudinal races. Studies on
crosses between ecological races only
slightly or moderately differentiated
from each other in comparison with
crosses between ecological extremes
have been mostly lacking. Because of
its biosystematic structure, the Mim-
ulus cardinalis-M . lewisii complex is
particularly well suited to investiga-
tions aimed at filling this gap in our
knowledge (Year Book 64, pp. 427-
428). Questions which this work was
designed to clarify include the follow-
ing:
1. How do first-generation hybrids
between closely related yet distin-
guishable ecologic races of the same
species but from distinct habitats
compare with hybrids between highly
diverse races in their capacity to sur-
vive and grow at the altitudinal
transplant stations at Stanford,
Mather, and Timberline ?
DEPARTMENT OF PLANT BIOLOGY 209
2. Is there evidence of hybrid vigor Timberline transplant stations. These
(heterosis) in such crosses and, if so, three combinations are chosen to rep-
to what degree is its expression modi- resent the range of patterns of re-
fied in these contrasting environ- sponse found in a number of intra-
ments? cardinalis racial crosses.
3. Does maternal (i.e., cytoplas- 1. Los Trancos X Yosemite. The
mic) inheritance play a part in the Los Trancos race of M. cardinalis
expression of any of the first-genera- from the coast of central California
tion hybrid combinations in any of has been previously described (Year
these environments ? Book 60, pp. 381-384). It survives
In practice, it is difficult to obtain only at the Stanford station situated
objective answers to these questions near sea level. The individual parent
because of the nature of ecological plant used in the crosses, 6546-5, is
races. Our data are biased because, not a heavy producer of dry matter,
from practical experimental consid- as indicated by the arrow at the lower
erations, we are impelled to assemble left of Fig. 44. The individual of the
and establish our experimental cul- Yosemite race used as a parent in this
tures at Stanford. This mild coastal cross (6694-105), originally from
climate of central California favors 880 m elevation in the Sierra Nevada,
the growth and establishment of low- grows with considerable vigor at
altitude as compared with high- Stanford and produces a higher dry
altitude ecological races. We are weight yield, as shown by the arrow
nevertheless able to overcome this at the lower right of the figure. A
bias to a considerable degree through sample population of 10 Ft individuals
extended observations at the moun- resulting from this cross (which in-
tain stations. eludes reciprocals) shows consistently
Analysis of the available data from greater vigor than either parent when
reciprocal crosses of all hybrid com- grown at Stanford, as reflected by the
binations made in Mimulus, either dry weights shown by the bar dia-
intraspecific or interspecific, has thus gram at the bottom of Fig. 44.
far failed to yield any evidence of The same population sample of
maternal or cytoplasmic inheritance cloned Ft hybrids grown at Mather
at any of the three altitudinal sta- at 1400 m elevation displays a slight
tions. The answer to the third ques- enhancement of vigor as compared
tion appears, therefore, to be clearly with the parents at this same station,
negative. The following discussion but the expression of heterosis is so
will accordingly be concerned only much smaller than at Stanford that
with the first two questions, and is it is barely evident. At Timberline,
based on a study of the pooled data where the Fx hybrids barely survive,
of reciprocally produced hybrids in they may nevertheless be said to dis-
each combination mentioned. play a slight degree of heterosis as
Intra-cardinalis hybrids. A useful compared with the parent clones,
measure of overall growth in Mimulus both of which are nonsurvivors.
is the total dry weight of the above- 2. Los Trancos X San Antonio
ground parts that are produced by a Peak. When the same Los Trancos in-
plant during the course of its normal dividual used in the preceding cross
growing season. The graphs in Fig. is combined with a clone of M. cardi-
44 show the productivity of three nalis from San Antonio Peak, origi-
different combinations of intereco- nally from 2200 m elevation (clone
typic Fi hybrids of Mimulus car- 7120-15), the Fx hybrid population
dinalis at the Stanford, Mather, and when grown at Stanford displays
210
CARNEGIE INSTITUTION
§ 2
■S -5
*2
■£ o
*- <D
^v.
■e
<
¥
->-
>-
O
o
NO
o
o
o
o
E
CO
Cl
o
O jt
O oo
m —
o
o
CM
J o
CD
< 2
<1> £
•♦— o
O
c
_o
CO
<
D CN
(1)
1 -
.?£
to
jc >
*■ o
~0 to
c c
o .2
8 1
u «"
o , — ,
u E
o o
§ s
i- CO
o —'
CL O
c
.** :=
C _Q
■5 £
<o c
"5
I?
^ o
0 2
1 T
J2 <l>
i. o
«/> -rr
.2 E
o o
">. CO
t= c
-£> o
<D O
-- i-
D O
o c
s ^
c ■£
8«
— c
D D
'■E E
> o
■♦—
!•=
■» D
o -a
"Is
£ 13
D C
«/> o
C "to
.2 o
1 -
Q. *
2 -Q
-o o
■£. °
£ 2
J*
o —
I "g
3 °
a) _g
c a
D O
*■ a
■£ -c
■*■* *-
o
u
o
k.
u
a
D)
D
c
O X
D) c
.E D
£ a
o ">
■c E
k.
-0.-0
to
D) 2
I/I C
o o
« '-c
?■ °
D> D
J= C
a> «»
.E J
,E x
= o
Q
a
c
d
0>
C
o
N
O
H ">
c
o
N
\_
o
JBI
(D >-
-C O
I— c
a
>
e
D
<o
E
<i>
DEPARTMENT OF PLANT BIOLOGY 211
about the same degree of hybrid vigor tions. All races of M. lewisii are es-
as the preceding cross, as shown by sentially nonsurvivors when grown in
the bar diagram in Fig. 44 (second the Stanford garden and can only be
graph from the bottom) . At Mather maintained successfully there in con-
the same population of cloned Fx hy- trolled growth chambers. First-gen-
brids grows somewhat more vigor- eration hybrids between contrasting
ously than the Fi hybrids between altitudinal or latitudinal races of M.
Los Trancos X Yosemite, but this is lewisii do show some improvement in
scarcely significant. The great reduc- growth over their parents at Stanford
tion in heterosis expressed at Mather but they are essentially nonsurvivors.
as compared with Stanford is of At Mather the response is only
the same order of magnitude as in slightly better than in the Stanford
the preceding combination. At Tim- environment. There the parents are
berline the influence of the San An- likewise essentially nonsurvivors and
tonio Peak parent in enhancing the although occasionally Fx plants may
vigor of the hybrids as compared with produce well during a single season,
the parents is somewhat more marked none have survived more than two
than in the preceding cross and may seasons. At Timberline all races of M.
be significant. lewisii survive and the Ft hybrids
3. Los Trancos X Baja. The in- show definite heterosis as compared
fluence of the genome of the Baja race with the parents,
of M. cardinalis originally from the When any race of M. lewisii is
San Pedro Martir at 600 m elevation crossed with any race of M. cardinalis
(clone 7119-16) when crossed with the vigor of the Ft progeny at Stan-
the same Los Trancos parent used in ford generally exceeds the perfor-
the two preceding combinations is mance of either parent. Moreover,
striking in the high degree of hetero- unlike interracial Fx hybrids within
sis shown by the resulting Fa progeny M. cardinalis, heterosis in cardinalis-
grown at Stanford. This is shown by lewisii combinations is clearly ex-
the bar diagram in Fig. 44 (third pressed at all three altitudinal trans-
from the bottom), which also por- plant stations. This is shown in Fig.
trays the high degree of individual 45 in which the dry yield performance
variability in this Fx hybrid popula- at Stanford, Mather, and Timberline
tion. The same Baja genome when in three combinations of cardinalis-
combined with other races of M, car- lewisii are shown in comparison with
dinalis consistently yields Fi hybrids the intr n-cardinalis cross Los Trancos
displaying more hybrid vigor in the X Yosemite.
Stanford garden than hybrids pro- 1. M. cardinalis Los Trancos X M.
duced from any other interracial com- lewisii Yosemite. The coastal M. car-
bination within this species. dinalis clone 6546-5 from Los Trancos
At Mather, however, even the effect when crossed with M. leivisii, origi-
of the Baja genome in enhancing het- nally from Yosemite National Park,
erosis is barely expressed. This is at 880 m elevation (clone 7121-5),
consistently shown also by all other yields a population of Fx progeny
Fx combinations between ecological which grows with moderate vigor at
races of M. cardinalis. At Timberline Stanford, well exceeding the Los
also there is essentially the same re- Trancos parent in overall yield, as
sponse. shown in Fig. 45 (the second bar dia-
Mimulus cardinalis X M. lewisii Ft gram from the bottom). The overall
hybrids in comparison with intra- growth of this Fx population at Stan-
car dinalis and intra-lewisii combina- ford nevertheless falls short of that
212
CARNEGIE INSTITUTION
At Timberlme
(Graphed at X10 scale)
Los Trancos X Timber line
M cardmolis X lewisii
Los Trancos X Yosemite
M. cardinal is X lewisii
Los Trancos X Yosemite
M. cardmolis X cordinalis
~\
P#^
T.L.
*>.
t
Y
At Mather
(Los Trancos M. Cardi
is is a nonsurvivor
Yosemite X Semite
M. cardmolis X lewisii
Los Trancos X Timberline
M cardmolis X lewisii
Los Trancos X Yosemite
M cardinal is X lewisii
Los Trancos X Yosemite
M. cordinalis X cardmolis
At Stanford
Yosemite and Timberline races of
M. lewisii are nonsurvivors
Yosemite X Yosemite
M cardmolis X M lewisii
Los Trancos X Timberline
M car di nali s X lewisii
Los Trancos X Yosemite
M. cardmolis X lewisii
Los Trancos X Yosemite
M cardmolis X cardmolis
t t
Y Y
t
T.L.
♦ *
Y
t
L.T.
t
LT.
♦
LT.
>
m
(Los Trancos and Yosemite races of
M. cordinalis are nonsurvivors)
00 200 300
Dry weight of tops, grams
400
Fig. 45. Graphs of dry weight yields of parents and first-generation hybrids between different
races of M. cordinalis and M. lewisii in comparison with the parents and with an \n\ra-cardinalis
hybrid. See legend of Fig. 44 for explanation of the graphs. The arrows with asterisks indicate
parental races of M. lewisii; those without asterisks, parental races of M. cordinalis.
DEPARTMENT OF PLANT BIOLOGY 213
of the Fx population of M. cardinalis lewisii (clone 7121-5) from the same
Los Trancos X M. cardinalis Yosem- area is of particular interest in study-
ite, which is shown in the lowermost ing the effect of combining the ge-
bar diagram of the same figure for nomes of these two distinct but closely
comparison. related taxa originally growing in
At Mather the heterosis of the two essentially the same natural habitats
combinations is drastically reversed in comparison with crosses between
as compared with Stanford: M. car- ecological races native to different
dinalis Los Trancos X M. lewisii altitudes and latitudes.
Yosemite is strikingly more vigorous The most striking feature of the
than M. cardinalis Los Trancos X M. Fx population from this cross is its
cardinalis Yosemite. At Timberline marked similarity to the responses of
the relative performance of the the Fa hybrids between the two car-
two Fi combinations is approximately dinalis X leivisii combinations pre-
the same as at Mather although the viously discussed at the three trans-
actual total dry weight yield at Tim- plant stations (Fig. 45).
berline in both instances is reduced Conclusions. From the available
by a factor of nearly 10 because of data, the following conclusions re-
the much shorter growing season. garding the genetic structure of nat-
2. M. cardinalis Los Trancos X M. ural populations of Mimulus cardi-
lewisii Timberline. When the Los nalis and M. lewisii appear to be
Trancos clone of M. cardinalis (6546- valid:
5) is crossed with the high-Sierran 1. The genetic differences that have
M. lewisii originally from Timberline evolved between M, cardinalis and M.
at 3300 m elevation (7405-4), the re- leivisii are of greater magnitude and
suiting Fx progeny show about the evolutionary significance than the
same degree of vigor as the progeny genetic differences between ecological
of the cross M. cardinalis Los Trancos races within either species.
X M. leivisii Yosemite. The parallel- 2. The genetic differences between
ism in the performance of these two ecological races of either species are,
Fi populations when grown at the nevertheless, of sufficient magnitude
Stanford, Mather, and Timberline to be of ecological significance in nat-
transplant stations is striking (Fig. ural selection.
45) . This result fails to support an 3. Hybrid vigor, or heterosis, is the
earlier hypothesis that the F1 between general rule in crosses within the
M. cardinalis Los Trancos X M. lew- Mimulus cardinalis-M . lewisii com-
isii Timberline would be more sue- plex in both intraspecific and inter-
cessf ul at Timberline than the Fx be- specific crosses. The expression of het-
tween M. cardinalis Los Trancos X erosis in a given Fi population may
M. lewisii Yosemite. The effect of be profoundly influenced by climate
combining the genomes of M. cardi- as, for example, in mtYSL-cardinalishy-
nalis and M. lewisii clearly over- brids grown at Stanford as compared
shadows genetic effects due to recom- with Mather, or in intra- lew is ii hy-
binations of different races of M. lew- brids grown at Timberline as com-
isii on a common parent of M. cardi- pared with Mather. In other instances
nalis. heterosis may be almost equally ex-
3. M. cardinalis Yosemite X M. pressed over the entire range of
lewisii Yosemite. The combination climates at Stanford, Mather, and
between M. cardinalis originally from Timberline, as in combinations be-
Yosemite National Park at 880 m tween M. cardinalis and M. leivisii. In
elevation (clone 6694-105) and M. such combinations heterosis is espe-
214 CARNEGIE INSTITUTION
dally evident at Mather where most at approximately 100,000 ergs/cm2/
of the parental races of both M. car- sec, or about one third the intensity
dinaiis and M. lewisii fail to be sue- of full sunlight. Leaves of the same
cessful. individual plants grown under a weak
4. There is no evidence to support light of 25,000 ergs/cm2/sec intensity
the hypothesis that the survival of Fx have a considerably lower light-satu-
hybrids at contrasting altitudes is re- rated rate, and the differences in
lated in a simple way to the altitudi- light-saturated photosynthetic rates
nal origin of the particular ecological of ecological races and hybrids tend
races that enter into the crosses. to disappear. The ratio between the
light-saturated photosynthetic rate
PHOTOSYNTHETIC KATES OF °f * ^ ?}°™ . ^°WX[ "nder high
MhnulHS RACES AND HYBRID ™raU? loW ^ ht llltenSlty C&n *??
■p. fore be used as one measure of its
performance.
William M. Hicscij, Malcolm A. Nobs, Tab]e 13 showg light-saturated
and OUe Bjorkman photosynthetic rates of M. cardinalis,
The higher rates of light-saturated subalpine M. leivisii, their reciprocal
photosynthesis of coastal Mimulus Fx hybrids, and two F2 individuals
cardinal is as compared with subalpine that differ widely from each other in
M, lewisii described last year (Year their morphology and in their growth
Book 65, pp. 461-468) have been con- responses at the Stanford, Mather,
firmed by means of different precon- and Timberline transplant stations.
ditioning treatments. All of these clones were previously
As previously emphasized, such grown under the same conditions,
differences in photosynthetic capacity under high light intensity in con-
of contrasting races of Mimulus and trolled cabinets, before measurement.
hybrids are revealed only when the The F2 clone 7135-35 resembles M.
experimental plants are previously car dinaiis both morphologically and
conditioned by growing them under a in its survival, and in interstation
high light intensity as, for example, transplant responses. It is a vigorous
TABLE 13. Photosynthetic Rate and Chlorophyll Content of Mimulus Parents and Hybrids
Gross Light- Ratio, P Rate chlorophyll Content of Leaves
Saturated Photo- of Leaves Grown
synthetic Rate under High and mg/dm2 mg/g fresh
mg C02/dm2/hr* Low Light Intensity surface weight
Mimulus cardmalis, Los Trancos,
6546-5
Mimulus lewisii, Timberline,
7405-4
Mimulus cardinalis X lewisii,
Ft Hybrid, 6546-3
Mimulus lewisii X cardinalis,
Ft Hybrid, 65A7A
Mimulus cardinalis X lewisii,
F2, caraVnal/s-like, 71 35-35
Mimulus cardinalis X lewisii,
F2, lewisii-Uke, 71 1 1 -1 6
38.9
1.8
5.07
1.60
26.3
1.4
3.40
0.90
29.0
1.5
4.35
1.41
30.1
1.5
4.00
1.42
33.1
1.9
4.94
1.62
31.3
1.8
3.93
1.05
* Plants previously grown under high light intensity. All values are means of several indepen-
dent measurements.
DEPARTMENT OF PLANT BIOLOGY 215
survivor at Stanford but dies at the different leaf-surface to leaf -volume
Mather and Timberline stations. In ratios.
contrast, the F2 clone 7111-16 The major points of interest shown
strongly resembles the M. lewisii by the data in Table 13 may be sum-
parent. The resemblance is so close marized:
that it would be classified as a mem- 1. Coastal M. cardinalis has a sub-
ber of this species, and is a nonsur- stantially higher light-saturated
vivor at Stanford and Mather. photosynthetic rate than does subal-
The two F2 individuals yield segre- pine M. lewisii. Likewise, the ratio
gating F3 progenies that differ widely between light-saturated rates of
from each other in overall genetic photosynthesis of propagules grown
composition. The growth responses at high light intensities to those
and survival of samples of these F3 grown at low light intensities is
progenies are currently being tested much higher in the coastal clone.
at the altitudinal transplant stations. 2. The two reciprocal Ft hybrids
The ratio between the light-satu- have a light-saturated photosynthetic
rated photosynthetic rates of propa- rate that is intermediate between the
gules of these clones grown under parents, as do also the two very unlike
high light intensity to those grown F2 individuals. The ratio between
under low light intensity is given in high and low light-grown propagules
the second column of figures from the differs markedly between the F1 and
left in Table 13. These ratios differ, the two F2 individuals, being as high
especially in that M. cardinalis has a as in the M. cardinalis parent in both
higher ratio than M. lewisii. The F2 plants.
ratios of the F1 hybrids are inter- 3. Chlorophyll content is markedly
mediate between the parents, but both higher in the coastal M. cardinalis
of the contrasting F2 individuals have parent than in M. lewisii, and inter-
ratios as high as the M. cardinalis mediate in both reciprocal Fx individ-
parent. uals. In the cardinalis-like F2 chloro-
The bulk chlorophyll content of the phyll content is essentially the same
leaves also differs greatly between the as in the M. cardinalis parent, and in
two parent clones and can be re- the leivisii-like F2 it is nearly as low
garded as another segregating char- as in the M. lewisii parent,
acteristic in the hybridization studies These data suggest that differences
independent of the light-saturated in photosynthetic characteristics in
photosynthetic rate. The chlorophyll Mimalus are inherited in much the
content affects the capacity of the same manner as morphological mark-
leaves to absorb light and therefore ers {Year Book 63, pp. 432-435), and
influences the rate at low intensities that one may expect the principle of
of incident light, but not at saturating genetic coherence (Year Book 62, pp.
light intensities {Year Book 65, pp. 387-389) to apply to the inheritance
461-468). The chlorophyll content of of factors controlling various steps of
the leaves is shown in the two col- the photosynthetic processes that
umns at the right of Table 13 and is determine observed measured rates
expressed on both a leaf -area and a under specified controlled conditions,
fresh-weight basis. The differences Studies such as those described
in relative values between these two above and biochemical investigations
methods of computing chlorophyll comparable with those reported below
content are due primarily to heredi- for Solidago are being extended to
tary differences in leaf thickness be- include M. nelsonii from Mexico, Fx
tween the clones, which result in hybrids between nelsonii and subal-
216 CARNEGIE INSTITUTION
pine M. lewisii, and the tetraploid conditions were identical and as de-
amphiploid derived from these spe- scribed earlier (Year Book 64, p.
cies as described in last year's report. 420). In all measurements of photo-
Data from these studies are still too synthesis, single rosette leaves, at-
incomplete to be reported. tached to the intact plants, were used.
Recently matured leaves (about 80%
of full expansion) were used both for
CARBOXYDISMUTASE Activity in Sun photosynthesis measurements and
and Shade Ecotypes of Solidago enzyme determinations, except where
specified
OUc Workman and Monika Bjorkman ■rtl , ' ,7 „. ,
' J Photosynthesis measurements.
Preliminary studies, reported last Light-saturated rates of photosyn-
year. indicated that ecotypes of Soil- thetic C02 uptake were measured
dago uirgaurea originally from open with the apparatus described in Year
as contrasted with shaded habitats Book 63 (pp. 430-431) modified to
may differ in their capacity to pro- meet the requirements of the present
duce the photosynthetic enzyme car- measurements. Light intensities, pro-
boxydismutase (ribulose diphosphate viding up to 1.5 X 10-7 absorbed ein-
carhoxylase) . Of the two Solidago steins cm-2 sec-1 at 665 imi, half-
clones investigated, the clone from an bandwidth 35 m/x, were obtained from
open habitat showed a higher activity a quartz-iodine lamp (DWY) , appro-
of the enzyme than the one from a priate lenses, and water and inter-
shaded habitat, in extracts prepared ference filters. To minimize undesir-
from the leaves when grown under a able gradients in C02 concentration
high light intensity. Unrelated spe- in the leaf chamber its volume was
cies limited in natural distribution to kept small (15 ml), and high flow
open locations, such as Plantago Ian- rates (0.5 to 1.0 1/min) were used.
ceolata, showed a higher activity of The average C02 concentration of the
carboxydismutase than species lim- air in the leaf chamber was 0.030%.
ited strictly to shaded habitats, such Enzyme determinations. Crude en-
as Lamium galeobdolon. zyme extracts were prepared from
This year we have studied this each separate leaf by homogenizing
problem in greater detail. An im- the pre-weighed leaf sample (approx-
proved assay procedure for enzyme imately 200 mg fresh tissue) in 10 ml
determinations has been developed, of a mixture of 0.04 M tris-HCl, 0.01
additional Solidago clones have been M MgCl2, 0.25 mM EDTA, and re-
studied, and photosynthesis measure- duced glutathione (GSH). The final
ments and enzyme determinations pK was 7.8 at 23 °C. Addition of GSH
have been made on the same leaves. to the mixture used for homogeniza-
Plant materials. The recent meas- tion was essential for high and repro-
urements were made with cloned ducible enzyme activities. Although
individuals of two populations of two in most cases 1.0 mM GSH was suffi-
races of Solidago virgaurea, "Bes- cient for maximum activity, higher
kades," native to an exposed alpine- concentrations were required with
arctic heath in northern Norway, and some samples. A concentration of 5.0
"Hallands Vadero," native to a dense mM GSH proved to be sufficient for
oak forest in southern Sweden. Sev- all clones and was used throughout
era] clones of each race were grown the investigation. The supernatant,
in controlled cabinets at two different obtained after spinning the homoge-
light intensities, 25,000 and 110,000 nate at 30,000 g for 20 minutes, was
erg cm-2 sec-1 (400-700 m/x). Other used for the enzyme assays without
DEPARTMENT OF PLANT BIOLOGY
217
further purification. All preparative
procedures were carried out at 0° to
2°C.
The enzyme assays were consist-
ently made 40 to 60 minutes after the
start of the homogenization. As a
standard procedure, the reaction was
started by the addition of 0.1 ml of
the enzyme extract to 0.4 ml of a
freshly prepared mixture of 2.5 /miole
NaHC1403 (0.5 /xcurie//miole) , 0.14
/miole ribulose-l,5-diphosphate, 1.25
/miole GSH, 0.1 /.mole EDTA, 10
/miole tris-HCl, and 2.5 /miole Mgd2.
The pB. was 7.8 at 23 °C. The sodium
salt of ribulose-l,5-diphosphate was
obtained by conversion of the barium
salt. This was either purchased from
Sigma or was kindly supplied by
Dr. Daniel McMahon, University of
Chicago. The latter preparation
yielded somewhat higher activities
with given enzyme preparations in
comparison with the former and was
used as a standard throughout the
investigation.
After two minutes' incubation at
23 °C the reaction was stopped by the
addition of 0.1 ml 6 I acetic acid.
A 0.25-ml aliquot was pipetted into a
liquid scintillation counting vial and
dried at 90° C for 2 hours. Then 0.25
ml of water was added to each vial,
followed by 10 ml of Bray's solution.
The radioactivity measurements were
made with a Packard Tri-Carb
Liquid Scintillation Spectrometer.
The rate of C14 incorporation was
entirely linear with enzyme concen-
tration in the range used, but declined
somewhat with increasing incubation
time. No change in the activity of the
enzyme preparation was observed
during a period of 1 hour if the en-
zyme was kept at 0° to 2°C. At least
four replicate assays were run for
each preparation, the typical varia-
tion among replicates being about
5%. The rate of C14 incorporation in
the absence of ribulose diphosphate in
all cases was less than 1 % of the rate
in the presence of this substrate.
Results. The carboxydismutase ac-
tivities in preparations from selected
clones of the Beskades population
from an open habitat, and of Hal-
lands Vadero from a densely shaded
habitat are compared in Table 14.
The three clones of each population
were selected to include the extremes
of variation with regard to leaf anat-
omy as well as to their photosynthetic
characteristics as found in previous
extensive work.
It is evident that the Beskades
plants grown at a high light intensity
show the highest enzyme activities
regardless of the basis used for ex-
pressing them. The enzyme activities
of the Hallands Vadero clones are
relatively low when preconditioned to
either high or low light intensities.
Two of the Beskades clones (033 and
039) seem to require a high light in-
TABLE 14. Carboxydismutase Activity in Contrasting Races of Solidago virgaurea from
Sunny and Shaded Habitats*
Clone /xmole C02/min/g Tissue /xmole C02/min/mg Protein
Light intensity for growth
high
low
high
low
033
5.0
2.9
0.26
0.16
Beskades (sun race)
039
5.2
3.5
0.24
0.18
076
7.0
6.4
0.26
0.28
124
2.8
2.7
0.14
0.11
Hallands Vadero (shade race)
132
2.2
2.6
0.17
0.14
014
4.6
3.2
0.14
0.12
* Each value represents the mean of independent experiments with two or more different plants
of the same clone.
218
CARNEGIE INSTITUTION
tensity to produce maximum enzyme
activity, whereas in the remaining
clone (076) the activity is high
whether the clone has been previously
grown in strong light or weak light.
The specific activity of the enzyme,
that is. the activity on the basis of
soluble protein, shows considerably
less variation among individual
clones and the differences between
the two races become more distinct
than when the activity is expressed on
the basis of fresh weight. When
grown at high light intensity the
Beskades clones have nearly twice the
specific activity of the Hallands
Vadero clones. This is not associated
with a lower level of soluble protein
in the Beskades clones; the total as
well as the soluble fraction of the leaf
protein is generally higher rather
than lower in the Beskades as com-
pared with the Hallands Vadero
plants.
It is interesting to note that even
though the assays were carried out at
bicarbonate concentrations several
times below saturation, and also at a
suboptimal temperature and ribulose-
diphosphate concentration, the spe-
cific activities of the Beskades plants
are among the highest reported in the
literature for any plant species. It
should be pointed out, however, that
if compared under physiological C02
concentrations the light-saturated
rate of C02 fixation by the intact leaf
is much higher than the rate of C02
fixation catalyzed by the enzyme ex-
tract of the same leaf.
We investigated the possibility that
the apparently lower specific activ-
ities of Hallands Vadero in compar-
ison with Beskades leaves might be
caused by the presence of an inhibitor
in Hallands Vadero not present in
Beskades, or conversely, to an activa-
tor present in Beskades but not in
Hallands Vadero. For this purpose
homogenates of leaves from the two
populations were mixed, and the re-
sultant activity compared with the
activities of the separate extracts.
The results presented in Table 15
show that the activity of the com-
bined homogenates is very close to
the calculated mean of the activities
of the separate extracts. This is evi-
dence for the absence of any such
inhibitor or activator. Also, the pres-
ence of a competitive inhibitor would
be expected to shift the apparent
value of the Km. No differences of Km
have been found among extracts from
the different clones.
As has been reported earlier
(Bjorkman and Holmgren, 1963),
pigment bleaching may result when
shade habitat plants of Solidago are
grown under light of high intensities.
This effect increases with leaf age
and, at advanced stages of bleaching,
deterioration of the chloroplasts is
often apparent. Precautions were
therefore taken to avoid complica-
tions that might result from such sec-
ondary effects. The light intensity
was kept at a somewhat lower level
than in previous investigations, and
only recently matured leaves showing
no visible evidence of chlorophyll
bleaching or chloroplast disturbances
were used for enzyme determinations.
To test to what extent a low car-
boxydismutase activity may be the
result of light damage of the chloro-
plasts, enzyme activity was deter-
mined in extracts of leaves in which
TABLE 1 5. Test for Presence of Inhibitors and
Activators of Carboxyclisrnutase in Extracts of
Leaves Grown in Strong Light
Preparation
Activity,
cpm/vial
(1)
Hallands Vadero 132
3839
± 56*
(2)
Beskades 076
7676
± 140
(3)
Mean of (1 ) and (2)
5758
(4)
Hallands Vadero 132 +
Beskades 076 1 :1
5865
± 175
* Mean and standard deviation, four inde-
pendent determinations.
DEPARTMENT OF PLANT BIOLOGY 219
increasing degrees of bleaching had an attempt to gain some information
been induced. As shown in Table 16, on this problem we determined the
the specific enzyme activity remains relationship between enzyme activity
the same even to a stage where the and photosynthesis in Solidago
chlorophyll content is less than one clones. Figure 46 shows the results
third of the normal, and severe struc- obtained when the rate of light-
tural disturbances of the chloroplasts saturated photosynthesis of leaves
have become apparent. from a number of different Solidago
It thus seems highly probable that clones of both races is plotted against
the lower enzyme activity found in the specific activity of the enzyme in
the Hallands Vadero as compared extracts prepared from the same in-
with the Beskades plants when grown dividual leaves. The two variables
at high light intensities is caused by a show a very high correlation. The
genetically determined lower capacity results of a detailed study of two
of Hallands Vadero to produce the clones of each population when grown
enzyme. It is tempting to speculate under high and low light intensities
that these differences have arisen as a (Table 17) further support the con-
result of an adaptation of the photo- elusion that there is a strong relation-
synthetic machinery to the light en- ship between the light-saturated rate
vironment of the plants in their of photosynthesis and the specific
native habitats. Under the low light activity of the enzyme. Both show
intensities prevailing on the floor of a a parallel variation regardless of
dense forest, efficient utilization of whether the differences are due to
light for photosynthesis would re- genetically distinct clones or to pre-
quire a high capacity of those steps conditioning by light,
concerned with the absorption and If the assumption is valid that the
utilization of the light in primary enzyme concentrations in vivo are
photoacts, whereas the capacity of proportional to the enzyme activities
enzymic steps could be relatively low. measured in vitro, these findings
With abundant light, on the other strongly suggest that the failure of
hand, the capacity of nonphotochem- the shade clones to adjust to the effi-
ical steps must be high for efficient cient use of strong light for photo-
photosynthesis, synthesis may be caused, at least in
However, the extent, if any, to part, by a limited capacity to produce
which the comparatively low level of the carboxylation enzyme,
carboxydismutase of the Hallands Whether Solidago clones from
Vadero leaves actually does limit the shaded and exposed habitats differ in
light-saturated rate of photosynthesis their capacity to produce photosyn-
under conditions of normal — and thus thetic enzymes other than carboxy-
limiting — carbon dioxide concentra- dismutase is not known. In consider-
tions is a question yet unanswered. In ing the question of what internal
TABLE
16.
Carboxydismutase Activity in Relation to Chlorophyll Bleaching by Exposure to
Strong Light. Clone, Hallands Vadero 132
Leaf No.
Carboxydismutase
Total Chlorophyll, Activity,
mg/mg protein fimole C02/min/mg protein
1
2
3
0.17 0.14
0.12 0.17
0.05 0.17
220
CARNEGIE INSTITUTION
1
c 0.2
3
o
in
Q-
.
<u
_c
e
c
■^
.*
F 0.1
c
o
Q.
u
:
0.1 0.2 0.3
Specific activity, pmol C02/min/mg protein
Fig. 46. Rate of light-saturated photosynthe-
sis in vivo in relation to the activity of carboxy-
dismutase in extracts from the same leaves.
factors may limit the rate of light-
saturated photosynthesis, it should
also be kept in mind that with nor-
mal CO j concentrations these factors
include not only the capacity of en-
zymic steps but also the barriers to
the diffusion of C02. Of great interest
in this connection are the findings by
Dr. Holmgren in Uppsala, Sweden,
that shaded habitat clones of Solidago
when grown in strong light show a
greater stomatal resistance to diffu-
sion than do exposed habitat clones.
Perhaps a simple explanation for
these differences can be found in the
physical structure of the leaves. How-
ever, the stomatal aperture is under
metabolic control and may very well
be strongly influenced by the rate at
which biochemical steps of photosyn-
thesis operates. The possibility can-
not be ruled out, therefore, that the
higher stomatal diffusion resistance
found in the shaded habitat clones is
associated with a lower capacity of
one or several enzymic steps in
photosynthesis.
Further Studies of the Effect
of Oxygen Concentration on
Photosynthetic C02 Uptake in
Higher Plants
Olle Bjorkman
Last year we reported that under
atmospheric C02 concentrations the
rate of photosynthetic C02 uptake
was about 30% inhibited by the oxy-
gen in the air in a number of species
of higher plants. The degree of inhi-
bition was remarkably constant over
a wide range of light intensities and
among various species differing
widely in their maximum photosyn-
thetic rate. However, two species of
green algae tested showed no inhibit-
ing effect of 21 % oxygen when photo-
synthesis was measured under con-
ditions similar to those used for the
higher plants. Although no conclusive
evidence was obtained with regard
to the mechanism of the inhibitory
TABLE 17. Light-S
aturated Rate of
and
Photosynthesis and Carboxydismutase
Sun Clones of Solidago
Activity in Shade
Grown in
Strong Light
Grown in
Weak Light
Origin
of Clone
Photosynthesis,
/xmole
C02/min/mg
protein*
Carboxydismutase
Activity, fxmole
C02/min/mg
protein*
Photosynthesis,
/imole
C02/min/mg
protein*
Carboxydismutase
Activity, [imole
C02/min/mg
protein*
Beskades 1 jJJ*
1 076
Hollands J 132
Vcidero 124
0.20
0.24
0.12
0.11
0.26
0.26
0.17
0.14
0.13
0.20
0.10
0.08
0.16
0.28
0.10
0.11
' All values represent means obtained from several ramets of each clone. Photosynthesis and
enzyme determinations were carried out on comparable but not necessarily the same leaves.
DEPARTMENT OF PLANT BIOLOGY 221
effect, some of our results suggested effective inhibitors of the enzyme
that the inhibition might be caused glycolic acid oxidase. Recently he re-
by a leakage of electrons from highly ported (Zelitch, 1966) that the addi-
reduced photosynthetic intermediates tion of this inhibitor to leaf discs of
back to molecular oxygen, possibly at tobacco stimulated the rate of net
a site between the two photosystems. C02 uptake in air, at least temporar-
That an enhancement of the rate of ily. Moreover, if the leaves were fed
the normal respiratory process by an with C14-labeled glycolate, the rate of
increased oxygen concentration C02 evolution in light and C02-free
would be responsible for the apparent air increased, as did the radioactivity
inhibition of C02 uptake was not con- of the C02 evolved. Thus, there are
sidered as a likely explanation, since several lines of evidence that strongly
it is well established that respiration indicate a close similarity between
is saturated at a low partial pressure the effects on C02 exchange obtained
of 02. when the oxidation of glycolate is in-
As part of their extensive work hibited, and those obtained when oxy-
directed toward uncovering the rela- gen concentration is reduced. Also, it
tionship between photosynthesis and has been shown by Hess and Tolbert
respiration in higher plants, Krotkov (1967) that the green alga Chlorella,
and co-workers investigated the de- in which the C02 exchange was found
pendence of the C02 compensation not to be affected by 21% oxygen,
point on oxygen concentration. They lacks glycolate oxidase. Some of our
found that this point (i.e., the C02 studies this year have therefore been
concentration at which there is no concerned with the question of
net uptake or release of C02) was whether a stimulation of glycolate
close to zero at very low oxygen con- oxidation is, in fact, the primary
centrations and that it increased cause of the inhibitory effect of oxy-
linearly with concentration up to gen on photosynthetic C02 uptake in
100% 02 (Tregunna, Krotkov, and higher plants.
Nelson, 1966; Forrester, Krotkov, Quantitative measurements of C02
and Nelson, 1966 a, h) . One exception exchange and glycolic acid content
to this rule was corn (Zea mays) in under air and low-oxygen concentra-
which the C02 compensation point is tion. If the increase in the rate of
zero even at an 02 concentration of C02 uptake, which takes place when
normal air. From these and other oxygen concentration is reduced, is
data they concluded that the main ef- caused by inhibition of the rate at
feet of increasing oxygen on C02 ex- which glycolate is oxidized to glyoxy-
change is a stimulation of the rate of late, glycolate should accumulate in
C02 evolution in light. They also con- the same way as it does when the
eluded that the process underlying enzyme mediating this step is inhib-
this evolution of C02 in light operates ited. If the rate of glycolate formation
through a pathway different from is not strongly affected by oxygen
that of the normal respiratory pro- concentration, the rate of glycolate
cess. Fock and Egle (1966), working accumulation would be expected to be
with bean plants and the liverwort proportional to the difference between
Conocephalum, arrived at similar the rates of C02 uptake in low oxygen
conclusions. and in air.
Several years ago Zelitch showed To test whether such an accumula-
that glycolic acid accumulates in the tion actually takes place, the follow-
leaves of higher plants in the pres- ing experiment was carried out : The
ence of hydroxysulfonates, which are rate of steady-state C02 uptake in
222
CARNEGIE INSTITUTION
air at 22 °C was determined on a de-
tached leaf of Solidago mtUtiradiata
whose petiole was placed in a small
volume of water, and the effect of
changing the oxygen concentration
between 21 % and 1.5% was meas-
ured repeatedly. The leaf was then
kept at 1.5% 0„ and the rate of C02
uptake recorded continuously for 1
hour. After this time the leaf was
killed by rapid immersion (<2 sec)
in boiling 1% NaHS03 solution. The
water in which the petiole was in-
serted during the experiment was
added to the solution, which was then
kept at 95 °C for 5 minutes. The leaf
was homogenized, and the superna-
tant obtained after centrifugation
was assayed for glycolate, by the
technique described by Zelitch(1958).
The controls were matched leaves
kept in air during the experiments.
The results in Table 18 show that no
increase in glycolate content took
place when the leaf was kept under
low 02 concentration, even though the
rate of C02 uptake was 41% higher
than in air.
Experiments in which Mimulus
plants were grown under 4% and
21% oxygen concentration over a 10-
day period (this report, p. 232) like-
wise yielded negative results. No sig-
nificant difference in glycolate content
of the leaves was found between
plants grown under 4% and 21% 02.
Effect of temperature on the inhi-
bition of CO z uptake by oxygen. With
saturating light and normal CO- pres-
sure most higher plants have a flat
temperature optimum, usually be-
tween 15° and 30 °C. An attractive
hypothesis to explain why the rate of
C02 uptake declines at a relatively
low temperature is that the rate of
CO- evolution in the light increases
more strongly with increasing tem-
perature than does photosynthesis.
Zelitch's findings that the stimulation
of C02 uptake by the addition of a
glycolate oxidase inhibitor increases
strongly with temperature appears to
be in good agreement with this ex-
planation. In Zelitch's experiment
with tobacco leaf discs, C02 uptake
was stimulated several fold by the in-
hibitor at 35° C but was not affected
at 25°C.
If such a strongly temperature-de-
pendent photorespiratory process un-
derlies the observed inhibitory effect
of oxygen on C02 uptake, this effect
should be small or nonexistent at a
low temperature (in tobacco leaves at
25° C) and should increase strongly
with temperature. The data shown
for Solidago and Mimulus in Fig. 47
indicate that this is not the case as
far as the steady-state rate of C02
uptake is concerned. The degree of in-
hibition caused by 21% 02 is almost
independent of temperature.
The results thus obtained both in
the experiments on glycolate accumu-
TABLE 18. Glycolate Content in Detached Leaves of Solidago multiradiata, T. L. 7621-2,
in Relation to CO2 Uptake in Low O2 and in Air
jiimole/leaf
(1) C02 fixed during 1 hour in 1.5% 02
(2) C02 fixed during 1 hour in 21 % 02
(3) Increase in CO2 fixation due to low O2
concentration (1) minus (2)
(4) Glycolate content expected if 1 mole CO2 is
formed for each mole of glycolate oxidized
(5) Glycolate content found in 1.5% O2
(6) Glycolate content found in 21 % O2 (controls)
48
34
14
> 14
< 1
< 1
DEPARTMENT OF PLANT BIOLOGY
223
pj'aBuDipxa ZQ^) JO 3-p^j
■SX.i
"* °? 5
,D> io CO
U. CO O
224
CARNEGIE INSTITUTION
lation in low 02 concentration and in
those on temperature dependence of
CO_> uptake fail to support the suppo-
sition that the primary cause of the
observed inhibition of C02 uptake by
oxygen in normal air is stimulation
of a "photorespiratory process'' medi-
ated by the enzyme glycolate oxidase.
It is noteworthy, however, that the
time course of the change in C02
uptake that results when C02 concen-
tration is altered is markedly affected
by temperature. The curves shown in
Fig. 48 typify the time course ob-
tained with Mimulus and Solidago
leaves, but it should be pointed out
that the magnitude of the transient
changes varies considerably among
different leaves, and also varies with
the detailed experimental conditions.
Effect of wavelength on the inhibi-
tory effect of C02 uptake by oxygen.
The steady-state rate of C02 uptake
in normal air seems to be a remark-
ably constant fraction of the rate in
low 02. The degree of inhibition
caused by 21% 02 was found to be
about the same among diverse species
of higher plants even though their
light-saturated rate of C02 uptake
varied greatly. Furthermore, the de-
gree of the inhibition is constant
over a wide range of light intensities
and, as already mentioned, also of
temperatures. The results shown in
Tables 19 and 20 indicate that the in-
hibition is also about the same at
different wavelengths of the light
over the range investigated. Thus
there is no indication that the ob-
served inhibition of C02 uptake, at
least under rate-limiting light inten-
sities, is due to a photorespiratory
process sensitized by pigments other
than those operating in photosynthe-
sis. Whether or not the spectral dis-
tribution of the light influences the
degree of inhibition at or beyond sat-
urating light intensities is not yet
known.
Absence of an inhibition of C02
uptake by 21% oxygen in certain
higher plant species. The thermo-
philic grasses corn and sugar cane,
which exhibit unusually high light-
saturated rates of photosynthesis, are
capable of reducing the C02 content
in a closed system to zero. The results
of Forrester et al. (1966b) also indi-
cated that the rate of C02 uptake in
corn was little affected when 02 con-
centration was changed from 1% to
21% 02. The absence in corn of an
inhibitory effect of 02 in this range
was confirmed in this laboratory. No
significant inhibition was found at
temperatures ranging from 10° to
40 °C, light intensities from 2 X 103
to 1.2 X 106 erg cm 2 sec-1, and wave-
lengths from 430 to 700 niju. Simi-
larly, the dicotyledonous species
Amaranthus edulis, which, like corn
and sugar cane, is capable of high
rates of C02 fixation and has a C02
compensation point close to zero in air
(El-Sharkawy et al., 1967), was
found to lack an apparent inhibition
of C02 uptake by 21% 02.
Figure 49 shows the different re-
sponse pattern in the rate of C02 ex-
change to changes between light and
dark in the three species Mimulus
cardinalis, corn (Zea mays), and
Amaranthus edulis. Mimulus was
chosen to represent species whose C02
uptake is inhibited by 21% 02.
Particularly noteworthy are the
differences among the species with re-
gard to the presence of a marked
transient high rate of C02 evolution
during the first minute after the light
is turned off. The presence of such a
post-illumination burst of C02 has
been taken as evidence for a stimula-
tion of the rate of C02 production in
the light (e.g., Tregunna et al., 1964).
The fact that the extent of the burst
decreases with decreasing 02 concen-
tration (Fock and Egle, 1966, and
Tregunna et al., 1966) and that it is
absent in corn, even in the presence
of air, at first sight seems to be con-
DEPARTMENT OF PLANT BIOLOGY
2^0
sistent with the hypothesis that there
is a close connection between the C02
burst and the inhibitory effect of 02
on net C02 uptake. It is known, how-
ever, that glycolysis is inhibited in the
light. Darkening relieves this inhibi-
tion and may lead to an overshoot of
oxidation, which may, in part, explain
the observed C02 burst. Moreover,
the present finding that Amaranthus
CD
>
CO
CD
CN
O
o
CD
CD
cn
c
cu
SZ
Time, minutes
Fig. 48. Time course of the rate of CO2 exchange in response to changes in oxygen con-
centration at different temperatures. Same leaf and conditions as in Fig. 47(A).
226 CARNEGIE INSTITUTION
TABLE 19. Ph<
Dtosynthetic CO2
Uptake
under 0.2%
02 c
ind under
Air at
Different Wave-
lengths in
So
lid a go
virgaurea, H<
ulland
s Vdderb*
124
Rate of CO
'2 Upt<
ake,*
Wave-
Half band-
nano-mol
cm-2 :
sec-1
length, m/i
width,
m/x
In low O2
In Air
Inhibition, %
440
37
0.39
0.24
38
493
15
0.38
0.26
32
535
12
0.36
0.25
32
554
12
0.37
0.25
32
575
13
0.36
0.24
33
605
13
0.36
0.24
33
633
14
0.37
0.24
33
665
13
0.38
0.25
33
* Measurements were made at 22 °C and 0.03% CO2. Light intensity at 665 m/x was 6 X
1 03 erg cm-2 sec-1. At other wavelengths the intensity was adjusted to give the same rates of
CO2 uptake as at 665 m/x. The rates were linearly related to light intensity.
edtrtis leaves exhibit a pronounced inhibitory effect of 02 on C02 uptake
burst of C02 immediately following il- show a temperature dependence con-
lumination, despite the fact that 21 % sistent with the finding that inhibi-
02 does not inhibit the net C02 uptake tion by glycolate oxidase inhibitors of
in these same leaves, shows that such the rate of C02 production in light is
a burst does not necessarily indicate strongly increased with temperature.
an inhibitory effect of 02. Measurements of the inhibition of
Conclusions. Experiments designed net C02 uptake caused by 21% 02 at
to test the hypothesis that the inhi- different wavelengths of light give no
bition of photosynthetic C02 uptake indication that a "photorespiratory
by 02 in normal air can be explained process" whose action spectrum dif-
simply by an enhancement of the rate fers markedly from that of photosyn-
of glycolate oxidation mediated by thesis causes the inhibition.
the enzyme glycolate oxidase have Although no new evidence support-
yielded negative results. An increased ing such a conclusion has been ob-
rate of C02 fixation in low 02 concen- tained, it seems reasonable that the
tration was not correlated with ac- inhibition is caused primarily by a
cumulation of glycolate. Nor does the back-reaction between a highly re-
TABLE 20. Photosynthetic C02 Uptake under 0.2% 02 and under Air at Different
Wavelengths in Mimulus cardinalis 71 20-7
Rate of C02
Upt
ake,*
Wave-
Half band-
nano-mol cm-2 i
sec-1
length, i
Tl/X
width,
m/x
In low O2
In Air
Inhibition, %
440
37
1.39
0.93
33
540
48
1.33
0.89
33
590
30
1.38
0.94
32
650
49
1.38
0.93
32
700
55
1.40
0.91
35
* Measurements were made at 23 °C and 0.03% CO2. Light intensity at 650 m/x was 4 X
1 04 erg cm-2 sec-1. At other wavelengths the intensity was adjusted so that the rates of CO2
uptake were approximately the same as those at 650 m/x. The rates were partially light saturated.
DEPARTMENT OF PLANT BIOLOGY
227
Light Dark Light Light Dark Light
3 C
6 8 0
Time, minutes
Fig. 49. Time course of the rate of CO2 exchange in response to light and dark in Mimulus
cardinalis, 7211-4; Zea mays, Ferry Morse hybrid 901; and Amaranthus edulis. Seeds of Ama-
ranthus, native to Argentina, were kindly supplied by Dr. R. S. Loomis and Dr. W. A. Williams
of the University of California at Davis. Measurements were made at 24 °C and 0.03% CO2.
White light of an intensity of 3 X 1 05 and 5 X 1 05 erg cm-2 sec-1 (400—700 mfx) was used
with Mimulus and the two other species, respectively.
22S
CARNEGIE INSTITUTION
dueed photosynthetic intermediate
and molecular oxygen that increases
with 02 concentration. Hydrogen pe-
roxide resulting from such a back-
reaction could have several secondary
effects. For example, it could increase
the rate of oxidation of glyoxylate ; it
also could stimulate the production
of glycolate from intermediates of the
Calvin cycle as proposed by Coombs
and Whittingham (1966).
Comparative studies of photosyn-
thesis among species differing in their
response to oxygen concentration ap-
pear to be a promising approach to
gaining new insight into the mecha-
nism of inhibition by 02. Of particular
interest in this connection are the re-
sults of Hatch and Slack (1967),
which indicate that certain tropical
grasses, including corn and sugar
cane, which lack an apparent inhibi-
tion of C02 uptake by 21% 02, utilize
a different pathway for photosyn-
thetic C02 fixation than most other
plants.
References
Coombs, J., and C. P. Whittingham, Proc.
Roy. Soc, B., 164, 511, 1966.
El-Sharkawy, M. A., R. S. Loomis, and
W. A. Williams, Physiol. Plant., 20,
171, 1967.
Fock, H., and K. Egle, Beitr. Biol. Pflan-
zen, 42, 213, 1966.
Forrester, M. L., G. Krotkov, and C. D.
Nelson, Plant Physiol, 1*1, 422, 1966a.
Forrester, M. L., G. Krotkov, and C. D.
Nelson, Plant Physiol., 41, 428, 1966&.
Hatch, M. D., C. R. Slack, and H. S.
Johnson, Biochem. J., 102, All, 1967.
Hess, J. L., and N. E. Tolbert, Plant
Physiol, 42, 371, 1967.
Tregunna, E. B., G. Krotkov, and C. D.
Nelson, Can. J. Bot., 42, 989, 1964.
Tregunna, E. B., G. Krotkov, and C. D.
Nelson, Physiol Plant., 19, 723, 1966.
Zelitch, I., J. Biol. Chem., 233, 1299,
1958.
Zelitch, I., Plant Physiol, 41, 1623, 1966.
Effect of Oxygen Concentration
on Dry Matter Production in
Higher Plants
OUc Bjorkman, William M. Hiesey,
Malcolm Nobs, Frank Nicholson, and
Richard W. Hart
The fact that the rate of photosyn-
thetic C02 uptake in most higher
plants is strongly enhanced by low 02
content in the atmosphere propounds
the question of whether or not the
rate of dry matter production is also
enhanced. An increase would be ex-
pected if (1) the enhancement of C02
uptake is due to a net increase in the
yield, and not merely an effect caused
by a greater portion of the reducing
power and phosphate bond energy
formed in photosynthesis being used
for the reduction of C02 at the ex-
pense of other vital endergonic proc-
esses, and (2) if low 02 concentration
does not adversely affect secondary
growth processes.
Although a great deal is known
about the effects of 02 partial pres-
sure in the root medium, few investi-
gations have been concerned with the
dependence of growth on the 02 con-
centration of the atmosphere. In re-
cent years Siegel et al. (1963) have
shown that a number of plant proc-
esses such as germination, root, and
coleoptile elongation are essentially
unaffected or sometimes even en-
hanced, by subatmospheric 02 levels,
whereas senescence is suppressed. In
some cases early seedling growth was
found to be somewhat greater at 10%
as compared with 21 % 02, and young
seedlings were able to grow at con-
centrations as low as 5%. Before any
conclusions can be drawn as to
whether the enhancement of C02 up-
take by low 02 is matched by an en-
hancement of dry matter production,
however, further experimentation
under precise control not only of 02
concentration but also of other ex-
ternal variables, particularly the con-
centration of C02, is needed.
DEPARTMENT OF PLANT BIOLOGY
229
To make such studies feasible, new
growth chambers were developed
during the year. A brief description
of their design and the results of ex-
ploratory growth experiments under
different 02 concentrations are given
below.
Growth cabinets with controlled 02
and C02 concentration. The basic
construction of the cabinets is the
same as described previously (Hiesey
and Milner, 1962) , except that the air
circulation system is closed to the ex-
ternal atmosphere. In addition, the
cabinets are connected with large col-
lapsible plastic bags (approximately
500 1) to equalize the internal and
external pressures. This prevents
leakage of external air into the cab-
inets, which would otherwise take
place when the external pressure in-
creases or when the internal pressure
decreases due to decreases in temper-
ature.
A schematic diagram of the control
system is given in Fig. 50. Thermis-
tor-operated controllers provide pre-
cise temperature control. Gas from
the cabinets is continuously pumped
through a paramagnetic 02-analyzer
(Beckman, Model F3) and an infra-
red C02-analyzer (Beckman, Model
15 A or LIRA, Model 300) . Both the
02-analyzer and the C02-analyzer are
equipped with electronic controllers,
specifically designed for this purpose
Light Bank
fl
O
o
o
Q
Q
O
O
3
a.
Cabinet A
>U th $k
Cabinet B-
n^ m w
—
c
13
N2or02in
C02in
Fig. 50. Block diagram of gas control system. The valve positions are shown for the control
cycle of cabinet A.
230
CARNEGIE INSTITUTION
by Mark Lawrence. Each consists of
a solid-state operational amplifier used
in a voltage comparator circuit which
compares the output of the gas ana-
lyzer with an internal reference.
When the gas concentration deviates
from the preset level, a transistor
switch activates a solenoid in the con-
trol ling gas lines to correct this
condition.
To avoid undesirable differences in
COj or Oa concentration between the
two cabinets, the same analyzers are
used for both cabinets. An electronic
timer switches the gas sampling and
controlling circuits between the two
cabinets at preset intervals (usually
1 min) . When a cabinet is being op-
erated at or above atmospheric 02
concentrations, C02-free air is slowly
fed into the cabinet to prevent the
CO 2 concentration from increasing
beyond the preset value ; at subatmos-
pheric 02 concentrations, the C02-
free air is replaced with C02-free N2.
The absolute accuracy of C02 control
is about ±5 ppm in the range 200 to
400 ppm. The maximum difference
between cabinets (if both are set to
the same value) is only about ±1
ppm. The accuracy of the 02 control
is about ±0.5% in the range 0% to
25% 02.
Light is provided by 96-inch Syl-
vania VHO fluorescent tubes (2.4 m
long) , supplemented with incandes-
cent lamps. Since a single light bank
is used for the illumination of both
cabinets, differences in light intensity
and quality between the two cabinets
can be kept very small.
The root medium can be aerated
with gas whose composition is inde-
pendent of that of the cabinet atmos-
phere. Gas-tight seals between root
and shoot are obtained with foam
neoprene gaskets (4 mm thick).
Prior to an experiment, cuttings or
seedlings are inserted through a tiny
hole in each gasket at an early stage
of development. As the stems expand,
tight seals are obtained by the neo-
prene gaskets without undesirable
effects on the plants.
Transpiration water is condensed
on the cooling coils of the refrigera-
tion system and drained through a
trap. The cabinets work well, and
several successful runs have been
completed with Phaseolus, Mimulus,
and Zea.
Effect of 02 concentration on
growth rate. In all experiments de-
scribed below, the plants were grown
under continuous light and 0.03%
C02. Except where specified, the roots
were immersed in a large volume of
nutrient solution which was aerated
with normal air (21% 02, 0.03%
C02). Within each experiment all
factors but 02 concentration were
kept constant.
The short-term response of growth
to different 02 concentration was
studied on bean seedlings (Phaseolus
vulgaris, var. Mexican Red). Care-
fully matched 5-day-old seedlings
were grown under 2.5% and 21% 02
at 29 °C and a light intensity of 5 X 104
erg cm-2 sec-1. Plate 1 illustrates the
response obtained in a 6-day period.
Clearly the growth was strongly en-
hanced at the low 02 concentration.
As shown in Table 21, the net in-
crease in the total dry weight of the
seedlings was about twice as high
with 2.5% as compared with 21%
02. All organs showed an enhanced
growth in low 02; there were no
major differences in their relative pro-
portions. Root growth was strongly
increased, even though the roots
themselves were aerated with normal
air.
The effect of low 02 concentration
on the growth of bean seedlings over
an extended period was studied in
another experiment. In this case 12
pairs of 6-day-old seedlings (Phase-
olus vulgaris, var. Ferry Morse Bush
869), sampled at random, were
grown in perlite with nutrient solu-
DEPARTMENT OF PLANT BIOLOGY
231
TABLE 21.
Effect of O2 Concentration on Dry Weights of the Bean Seedlings,
6 days' growth (Plate 1 )
Ratio
21%02, 2.5% O,. 2.5%
grams grams 21 %
(1) Leaves, final weight
(2) Roots, final weight
(3) Stem, final weight
(4) Entire plant, final weight
(5) Entire plant, original weight
0.288
0.118
0.130
0.536
0.203
0.520
0.237
0.164
0.921
0.221
1.80
2.00
1.26
1.72
(6)
Growth in 6 days, (4) minus (5)
0.333 0.700
2.10
TABLE 22. Effect of 02
Concentration on Dry Weights of Mimulus
21%02, 5%02,
grams grams
Ratio,
5% 02
21 % 02
(1)
(2)
(3)
Entire plant, final weight*
Entire plant, original weight
Growth in 1 0 days, (1 ) minus (2)
0.675* ± 0.029 1.186* ± 0.043
0.110 0.110
0.565 1 .076
1.76
1.90
* Means are significantly different at P = 0.01.
tion fed automatically. The roots
were kept at the same 02 concentra-
tion as the shoots. Light intensity was
8 X 104 erg cm-2 sec-1, and tempera-
ture was 27 °C. Plate 2 illustrates the
typical appearance of the plants after
17 days under 5% and 21% 02. No
adverse effects of growing the plants
under low 02 concentration were
found. The plants grew very rapidly
under these conditions, and produced
an abundance of flowers. The net in-
crease in shoot dry weight was about
twice as great under 5% 02 as under
air. No quantitative measurements of
root weights were made, but it was ob-
vious that root growth also was gen-
erally greater in the plants grown in
low 02. With beans chlorosis became
increasingly apparent in developing
shoots after about 10 days in normal
air. Presumably the effect is caused
by growing the bean plants under
continuous light. Interestingly, no
signs of chlorosis could be detected in
the plants grown under 5% 02.
Another noteworthy effect of grow-
ing the bean plants under low 02 is
that the rhythmic movement of pri-
mary leaves was arrested after about
two days. In air this movement con-
tinued for at least five more days.
Mimulus cardinalis, Jacksonville
1211-4;, provides an experimental ma-
terial much better suited for the pres-
ent studies than Phaseolus. Uniform
cuttings of a single clone can easily
be obtained, and the problem of ge-
netically variable seedling material
can be avoided. Moreover, continuous
light has no adverse effects on Mimu-
lus; on the contrary, it results in a
very rapid growth.
The results obtained with Mimulus
thus far indicate that the yield in dry
matter production is strongly en-
hanced by low 02. As shown in Table
23 the net gain in dry weight was
much higher in 5% and 2.5% 02 than
232
CARNEGIE INSTITUTION
TABLE 23. Effect of 02 Concentration on Mimulus, dry weights
21% 02,
grams
2.5% 02,
grams
Ratio,
2.5% Q2
21%02
(1) Entire plant, final weight*
(2) Entire plant, original weight
(3) Growth in 10 days, (1) minus (2)
0.589 ± 0.017
0.070
0.519
0.887 ± 0.036 1.51
0.070
0.817 1.58
* Means are significantly different at P = 0.01.
in air. under the conditions tested
(7.2 X 104 erg cnr2 seer1, 24°C).
It appears that 2.5% 02 might be too
low for maximum enhancement of
dry matter production.
Growing Mimulus under 2.5% and
dcc Oo resulted in a greater root-to-
shoot ratio, and a lower water content
of the shoots, than in air (9.8% dry
matter in low 02 versus 8.4% in air).
The plants grown under low 02 had
thicker leaves and a more compact
growth habit than those grown in air.
Photosynthesis measurements on
Mimulus and Phaseolus leaves
showed that the degree of inhibition
of net C02 uptake by 21% 02 was
unaffected by growing the plants un-
der different 02 concentrations. The
inhibition was about 30% regardless
of preconditioning. The absolute rate
of light-saturated C02 uptake on a
leaf area basis was higher in leaves
developed in low 02 than in air. This
was associated with a greater leaf
thickness of the former.
As was mentioned in the preceding
section of this report, Zea mays
(corn) differs from most other
higher plant species in that it lacks
an apparent inhibition of C02 uptake
by 21% 02. It seemed of interest
therefore to compare its response to
different 02 concentrations during
growth with that of Mimulus and
Phaseolus. In a preliminary experi-
ment, five matched pairs of corn seed-
lings, var. Ferry Morse Hybrid 901,
were grown under 5% and 21% 02,
with other conditions the same as
those given for Mimulus. As shown in
Table 24, the enhancement of dry
matter production in low 02, if any,
was much less than in Mimulus and
Phaseolus. Interestingly, however,
the root-to-shoot ratio and the per-
centage of dry matter in the shoots
(10.8% in low 02, 6.3% in air) were
higher in the plants grown in low 02.
This was also true of the other two
species.
References
Hiesey, William M., and Harold W. Mil-
ner, Bot. Gazette, 124, 103-118, 1962.
Siegel, S. M., L. A. Rosen, and C. Giu-
marro, Nature, 198, 1288, 1963.
TABLE 24. Effect of O2 Concentration on Dry Matter Production of Corn Seedlings
21 % 02,
grams
5% 02
grams
Ratio,
5% 02
21%02
(1)
(2)
(3)
Entire plant, final weight*
Entire plant, original weight
Grov/th in 1 0 days, (1 ) minus (2)
1.479 ± 0.123
0.210
1.269
1.683 ± 0.129
0.210
1.473
1.14
1.16
* Difference between means is scarcely significant: 0.3 < P < 0.4.
DEPARTMENT OF PLANT BIOLOGY 233
Hybridizations in Solidago California, and between these two
** i 7 a -KT t. sMi t>- i i and lowland counterparts from the
Malcolm A. Nobs, Olle Bjorkman, and , . . , „ ~ , ,
William M.Hiesey same tw0 continents, all from about
39°N. latitude. In addition, races
The exploratory crossings reported from approximately 70 °N in Scan-
in Year Book 65, p. 471, between dif- dinavia and in Alaska are being
ferent forms of the Solidago virgau- crossed with each other, and also with
rea-S. multiradiata complex have the southern races just mentioned,
yielded a number of Ft hybrids now Observations on the responses of pa-
growing in the Stanford garden that rental races and their Fa and F2 prog-
show interesting and complex char- enies at the Stanford, Mather, and
acteristics. Among these, for ex- Timberline transplant stations are
ample, is a cross between the shade essential to interpret the genetic corn-
form of Solidago virgaurea from Hal- position and potential of the parental
lands Vadero in southern Sweden, races.
mentioned in earlier sections of this We are indebted to Dr. John Ko-
report, and a subalpine from S. randa and Dr. William Mitchell of the
multiradiata from near our Timber- Alaska Experiment Station at Palmer
line transplant station in California, for collections of living material of
The success of this and other hybrid- Alaskan forms,
izations has prompted us to follow a
more extensive program this year
utilizing other intercontinental com- Comparative Physiological Studies
binations featuring interaltitudinal on Solarium dulcamara
and interlatitudinal crosses between
contrasting races of what appear to
be members of the Solidago virgau- The recent developments in plant
rea-S. multiradiata species-complex ecological research aimed at uncover-
that is widely distributed over the ing internal physiological mecha-
northern hemisphere. nisms that have evolved in plants
Our first objective is to determine native to different climates excite in-
the degrees of biosystematic relation- quiries concerning species groups
ship between key ecological forms other than Solidago and Mimulus. The
that have been referred to various combined results from independent
named taxa, yet appear to be closely studies on a diversity of species are
enough related to be capable of gene more likely to lead to the discovery of
exchange. general principles than the study of
A second objective is to compare only one or two species-complexes,
the genetic constitution of parallel The choice of material to be used
ecological races that are found in for this kind of work is of utmost
corresponding kinds of habitats in importance because of the exacting
widely separated continents. Such practical experimental requirements
races apparently were evolved inde- for such investigations. Essential re-
pendently over long periods of geo- quirements are (1) that the species
logical time from a common ancestral occur over a wide geographical range
stock. having a great diversity of habitats,
The hybridizations that are being (2) that its races be easy to grow and
made include combinations between be propagated as clones in order to
high-altitude forms from the Sierra facilitate experiments with geneti-
Nevada of Spain and high-altitude cally identical material in various con-
forms from the Sierra Nevada of trolled environments, and (3) that its
Eckard Gauhl
234
CARNEGIE INSTITUTION
leaf and stem structure be such as to
facilitate measurements of gas ex-
change in intact leaves while still
attached to the living plant.
Solatium dulcamara L., the bitter-
sweet, meets these requirements
well. It is distributed in Europe from
the Atlantic coast in France north-
ward to central Norway, and south
to northern Africa. It extends to
India and Asia in the east and occurs
in scattered areas throughout China
and Japan. In central Europe it grows
in shady swamps with alder (Alnus
glutinosa), or in deep shade within
stands of reed grass (Phragmites
communis) along borders of lakes.
Other races occupy exposed habitats
on coastal sand dunes near sea level,
or on rocky slopes of southern ex-
posure in the alps up to altitudes of
1,700 m. The most contrasting eco-
logical races are readily distinguish-
able morphologically.
A number of select clones have
been collected in central and southern
Europe, some of which have been
brought to Stanford under a U.S. De-
partment of Agriculture import per-
mit. Current work is being confined
mainly to three clones that represent
extremes of habitat origin and of
growth responses to controlled con-
ditions. One clone originally from a
reed grass marsh near Frankfurt
grows better under low light intensity
than under high. Another from a
sand dune along the Baltic Sea in
northern Germany that has thick,
fleshy leaves and short internodes be-
comes prostrate when grown under
strong light, and semi-erect under
weak light. A third clone, originally
from a dry, exposed rocky habitat
near Rovinj along the northern Dal-
matian coast of Yugoslavia, resembles
in growth habit the inland marsh
form, but grows better under strong
than under weak light.
Comparative studies on rates of
C02 uptake of intact leaves when
grown under high and low light in-
tensities have been started along lines
described above for clones of Solidago
and Mimulus. It is planned to pursue
these studies in other directions also,
with special consideration of inter-
relations between water metabolism
and photosynthetic performance in
contrasting climatic races of Sola-
tium.
Clusters of Tree Species on Both
Sides of the Pacific
Jens Clausen
Trees are suitable for use in gross
surveys of the vegetations of the
world. They are conspicuous enough
to be easily seen and fairly good rec-
ords of the limits of tree species are
available.
Records and experimental data in-
dicate that the greater number of the
world's tree species are arranged in
clusters of species, which experimen-
tal taxonomists call "cenospecies."
Certain taxonomic sections of genera
are on the size-order of species clus-
ters. The species of such a cluster are
closely enough related to permit occa-
sional interchange between their
heredities, either at present or in the
past. The evolutionary significance of
the species cluster has generally been
overlooked.
Species clusters of this kind circle
the earth within definite latitudinal
belts. Modern taxonomists, however,
constantly look for minor differences
between the taxa within each cluster
to apply different species names to
them in various parts of the world. By
splitting the species clusters, some-
times even into separate genera, sense
has been lost of the deep-seated phys-
iological and ecological character-
istics that are built into the heredities
of each cluster and that can be ex-
pressed in morphological characters.
Heredity holds the natural clusters
DEPARTMENT OF PLANT BIOLOGY 235
subunits. The principle here discussed species of the larch cluster are now
characterizes also herbaceous species being commercially intercrossed,
complexes around the world. Two species clusters belong to the
Such facts suggest that genes that genus Pinus, namely a two-needle and
govern morphology and those govern- a five-needle complex of species. Spe-
ing physiology are correlated in their cies belonging to different clusters do
inheritance. Such correlations are not interbreed. This hereditary sepa-
based on the principle of interlocking ration enables the species of two clus-
heredity (coherence) that we have ters to maintain their identity when
shown exists between climatic races growing together in the same local
of plants. Similar although much area.
stronger correlations probably limit The two-needle cluster of species
families and orders to their climatic includes the North American lodge
zones. pole pines, P. contorta Doug., P. mur-
Developing a sense of close rela- rayana Grev. et Balf., and P. bank-
tionship among the members of a siana Lamb., with the Japanese P.
species cluster brings a sense of per- thunbergiana Franco, P. densi flora
spective that is vital to the study of Sieb. et Succh., and with the scotch
taxonomy. A study of the tree species pines of Siberia to Europe, P. silves-
clusters on both sides of the Pacific tris L., P. montana Mill., and the
illustrates the principles involved in dwarf P. mugho Turra. Pinus silves-
this kind of modern plant taxonomy, tris is the type species of the cluster.
Based on worldwide field experi- Many natural hybrids have been sus-
ences, Misao Tatewaki's comprehen- pected among the latter three and
sive report (1958), supplemented by among the Japanese species. Righter
personal discussions with him on field and Stockwell (1949) described the
tours in Hokkaido in 1966, have fertile species hybrid P. murrayana
deeply influenced the report to follow. X P- banksiana, and intermediates
Trees of high latitudes. On both are known where the two species meet
sides of the Pacific approximately 12 in the wild. Like the larches, the two-
clusters of tree species follow a wide needle pine cluster circles the earth
belt, extending from about 70° to 35° and apparently shares a common gene
North latitude (Table 25). The indi- pool.
vidual species of each of the 12 clus- The five-needle pine cluster in west-
ters are species of high latitudes and era North America is represented by
high altitudes. Interestingly enough, the hardy whitebark pine P. albi-
6 of the 12 clusters belong to the caulis Engelm. On the Asiatic side of
pine family Pinaceae. Carolus Lin- the Pacific are the species P. pumila
naeus (1753) included all of the 6 Regel, P. koraiensis Sibth. et Succh.
clusters in the genus Pinus. of Northeastern Asia to Japan, and
The top cluster in Table 25 in- P. sibirica du Tour, from the Altai
eludes all species of the larch genus to the Ural Mountains and Peninsula
Larix. There are North American L. Kola. The type species of the cluster
laricina Koch, the Japanese L. lep- P. cembra L., occurring in the Car-
tolepis Gord., the East Asiatic L. pathian Mountains to the Swiss Alps
kamtschatica Carr., L. gmelini Gord., (Critchfield and Little, 1966), com-
and L. sibirica Ledeb., and the type pletes the circle. P. cembra, P. sibi-
species of the cluster, the Euro- rica, and P. koraiensis are generally
pean L. decidua Mill. C. Syrach Lar- erect trees; P. pumila is always an
sen (1937) found that all these spe- elfinwood bush, but P. albicaulis is a
cies are able to interbreed, and the tree in subalpine environments and
236 CARNEGIE INSTITUTION
TABLE 25. Trees of Highest Tolerance, Northern Hemisphere, High Latitudes, 35°— 70°N
Eastern Hemisphere
Pacific
Ocean
Western
Hemisphere
1 . Larches
2. Scots pines
3. Cembras
4. Spruces
5. Firs
6. Hemlocks
7. Junipers
8. Aspens
9. Willows
10. Birches
11. Alders
1 2. Mountain
ashes
PINACEAE, n = 12
LARIX decidua sibirica-kamtshatica
leptolepis
PINUS silvestris silvestris-densiflora
montana
PINUS cembra sibirica-koreensis-pumila
PICEA excelsa .... obovata-jezoensis-glehnii
smithii
ABIES alba . . . .sibirica-sachalinensis-veitchii
pindrow
TSUGA (extinct) diversifolia-sieboldii
brunoniana
CUPRESSACEAE, n = 1 1
JUNIPERUS communis
SALICACEAE, x = 19
POPULUS tremula tremula-jezoensls
davidiana
SALIX phylicifolia pyrifolia-pulchra
BETULACEAE, x = 14
BETULA tortuosa cajanderi-ermani
ALNUS incana maximovicsii-inokumae-crispa
ROSACEAE, n = 17
SORBUS aucuparia amurensis-scopulina
commixta
laricina
contorta-banksiana
albicaulis
sitchensis-glauca
lasiocarpa
mertensiana canadensis
communis
tremuloides
pulchra-arbuscula
kenaica-papyrifera
incana-crispa ....
sitchensis-canadensis
decora
Dots indicate that the particular species is continuous over the region. The second lines of the
clusters indicate the geographic location of certain southern species.
an el fin wood cushion type above the
tree line (Year Book 62, pp. 394-398;
and Clausen, 1965). Members of this
cluster of species extend to and be-
yond the tree line in North America
and Eurasia. Intermediates are re-
corded where the species meet. This
cluster of species and the members of
the two-needle cluster occur together
in many populations but do not inter-
cro
As far as is known at present, the
spruces, firs, and hemlocks constitute
three other high latitude circumpolar
clusters of trees, as shown in Table
26. The spruces (Picea), firs (Abies),
and hemlocks (Tsuga) are composed
of several geographic species that
have many intermediates. Species
within the Picea and Abies clusters
are known to interbreed (Wright,
1953) . Like the other members of the
pine family, all appear to have 12
pairs of chromosomes, and they were
included by Linnaeus as species of the
genus Pinus. The presently recog-
nized species represent geographic,
climatic, and edaphic adaptations to
many environmental niches within
the belt covered by the three genera
around the world. The species of a
cluster share a common gene pool.
DEPARTMENT OF PLANT BIOLOGY 237
Juniperus communis L. is still found a diploid Japanese species,
recognized as one species around the Alnus inokumae Mur. et Kus., having
world by conservative taxonomists. n = 7. Chiba was able to cross it with
Its lowland and inland forms are of tetraploid forms of A. hirsuta Turcz.,
erect columnar shape, whereas forms A. glutinosa Gaertn., and A. japonica
from windy coasts and alpine habi- Call., n = 14, and the hybrids have
tats are low elfinwood cushions. In 2n = 21 chromosomes (Chiba, 1966:
transplant experiments in Denmark, 1-16). The genus is commonly sup-
C. Syrach Larsen has shown that the posed to have many hybrids around
differences in growth form are hered- the earth.
itary. The junipers of this species The birches, Betula, also intercross
have 11 pairs of chromosomes and irrespective of differences in chromo-
cover approximately the same wide some number and of differences in the
belt as the six clusters of conifers appended species name. Helms and
previously discussed. Jorgensen (1925) found that Betula
Five nonconif erous high-latitude alba, n = 28, growing in a wet Danish
clusters of tree species include the moor, crossed with B. verrucosa, Ehr.,
aspens, willows, birches, alders, and n = 14, along drier, sandy edges of the
mountain ashes (Table 25). Each of moor. Partially fertile hybrid prog-
these is composed of closely related enies ranging from approximately
species that together circle the earth 21 pairs to intermediate numbers be-
at high to medium North latitudes, tween 21 and 27 were found along the
All are well known for their inter- borderline of the two populations,
specific crossability ; the willows, suggesting that possibly 7 is the basic
birches, alders, and mountain ashes chromosome number also in Betula.
cross irrespective of differences in The variation in birch is so great that
number of chromosomes. classification is very difficult, but
The quaking aspens constitute a Table 25 lists some of the names ap-
circumboreal natural group of spe- plied to its species around the high
cies, differing from other poplars by latitudes.
their ability to spread vegetatively An equal number of species names
through root shoots. A stand of 20 have been applied to the Sorbus aucu-
to 50 trees may consist of one indi- paria cluster, the mountain ashes of
vidual. The aspens have 19 pairs of the rose family, having 17 pairs of
chromosomes, and the North Amer- chromosomes. In North America are
ican Populus tremuloides Michx. is Sorbus americana March., S. decora
being commercially intercrossed with Sarg., S. scopulina Greene, and S.
the Asiatic-European P. tremula L. sitchensis Roem; in Japan are S.
The hybrids are fertile and possess sambucifolia Roem., S. matsumurana
hybrid vigor. The Japanese aspens, Koehne, and S. commixta Hedl. ; in
P. jezoensis Nakai and P. davidiana northeastern Asia is S. amurensis
Dode, have rhizomes like the other Koehne, and in Europe, the type spe-
quaking aspens, and P. davidiana has cies S. aucuparia L. The growth
been crossed with P. tremula (Chiba, forms vary greatly among the mem-
1966). ber species, but a member of this
The North American-Eurasiatic highly variable cluster can always be
alders, Alnus, were thought to follow expected to occur at high latitudes
a 14-series in chromosome number and high altitudes.
(Gram, Muhle Larsen, C. Syrach The 12 cold-tolerant species clus-
Larsen, and Westergaard, 1941). ters discussed here are on both sides
Chiba, however (1966: 155-156) , of the Pacific, but the member species
238 CARNEGIE INSTITUTION
within each cluster are difficult to dis- monticola Dough, and possibly the
tinguish because they share common sugar pine, P. lambertiana Dougl. On
gene pools. Members of distinct clus- the western side of the Pacific are the
tors, on the other hand, are clearly Armand pine, P. armandi Franch,
separated from each other. native to Taiwan, South China, and
Within the southern hemisphere Vietnam, the Indian blue pine, P.
there are no species clusters that in griffithi McClelland of the Himalayas
tolerance to extremely cold climates to Kashmir, and the Balkan pine,
match the 12 uniquely cold-tolerant P. pence Griseb., native to southeast-
northern hemisphere species clusters, ern Europe. Little and Righter(1965)
Clusters of tree species of medium intercrossed these species in all direc-
cold tolerance. A different and much tions, producing successful hybrids,
larger group of tree species occupies The white pines accordingly form an
medium latitudes from about 55° to evolutionary cluster that spans the
25 °N. on both sides of the Pacific, earth at medium latitudes.
Some of these species and genus clus- Several genera of the Taxodiaceae-
ters grow at comparable latitudes on Cupressaceae group circle the earth,
both sides of the Pacific in both the and they follow an 11-series in chro-
northern and southern hemispheres, mosome numbers both north and
The several thousand medium tolerant south of the equator and across the
tree species are less tolerant to cold oceans (Table 26). These genera con-
than the species native to the high tain the cypresses, the incense cedars,
latitudes, but they are much more thujas, sequoias, and swamp cy-
tolerant than species native to the presses. Specialists have split each of
low latitudes on both sides of the these clusters into distinct genera by
equator. At latitudes where species calling attention to small differences,
clusters of both high and medium but have neglected the gross simi-
tolerance overlap, those of medium larities and adaptations and chromo-
tolerance grow at lower altitudes than somal heredity that unite them across
the species of the highly cold-tolerant the oceans and across the equator,
clusters. One of these clusters is the cy-
Like the high latitude species, presses, Cupressus-Chamaecyparis.
those of medium tolerance are orga- Along the west coast of North Amer-
nized in evolutionary clusters of ica five to six Cupressus and two
closely related species. Sufficient ex- Chamaecyparis species are recog-
perimental evidence is on hand to nized. Osborn (1941) described, how-
indicate that the species of several of ever, a hybrid between the Monte-
these clusters share common gene rey cypress, Cupressus macrocarpa
pools on both sides of the oceans. Hartw., native to the central Cali-
Table 26 lists examples of tree species fornia coast, and the Alaska cedar,
clusters of medium tolerance on both Chamaecyparis nootkatensis Spach.,
sides of the Pacific. a mountain species growing from
The top cluster is a belt of white northern California to Alaska. This
pines of medium tolerance having five "intergeneric" hybrid is fertile, and
needles and elongated cones that con- its progeny segregates the parental
stitute the taxonomic section Strobus. characters, suggesting that at least
On the North American side this clus- these two species belong to one genus
ter is marked by the type species, the and that in this group the prevailing
eastern white pine, P. strobus L., the taxonomy is misleading. In Japan
Great Basin limber pine, P. flexilis Chamaecyparis pisifera Lindl. was
James, the western white pine, P. previously considered a Thuja, and C.
DEPARTMENT OF PLANT BIOLOGY
239
TABLE 26. Examples of Trees of Medium Tolerance, Latitudes 20°-55°, Both Hemispheres
(several thousand species)
Pacific Western
Eastern Hemisphere Ocean Hemisphere
PINACEAE, n = 12
White pines: PINUS peuce-griffithii-armandi monticola-strobus
TAXODIACEAE-CUPRESSACEAE, x = 11
,.~*^r-<r.n..r. I Taiwan Sierra Nevada
Incense cedars: UBOCEDRUS ... j New ZeaIand ai|e
Sequoias: (extinct) . . METASEQUOIA . . CRYPTOMERIA SEQUOIA-TAXODIUM
Cypresses: CHAMAECYPARIS obtusa nootkatensis
CUPRESSUS sempervirens macrocarpa
Thujas: THUJA orientaiis stondlshii plicata
FAGACEAE, Northern Hemisphere, n = 1 2
White oaks: QUERCUS robur crispula lobafa-alba
Chestnuts: CASTANEA sativa crenata . . .americana
CASTANOPSIS cuspidata chrysophylla
Beeches: FAGUS silvatica crenata grandifolia
NOTHOFAGUS, n = 13, Tasmania, New Zealand Chile, Patagonia
MAGNOLIACEAE, x = 1 9
MAGNOLIA obovafa . . . grandiflora
TILIACEAE, x = 41
Basswood: Till A cordata amurensis-japonica . . .americana
ACERACEAE, x = 1 3
Maples: ACER platanoides mono macrophylla-
saccharum
obtusa was shifted from Cupressus to
Chamaecyparis. The European spe-
cies is referred to as Cupressus sem-
pervirens L.
Excessively detailed taxonomic
studies have been even more mislead-
ing in presenting relationships be-
tween the incense cedars, the old
genus Libocedrus, which spans the
Pacific both north and south of the
equator (Table 26). Based on differ-
ences in the arrangements of stomata
and leaves, and disregarding the re-
markable gross similarities, the genus
has been split into five. The chromo-
some number is 11 in L. bidwilli
Hook, f ., and L. plumosa Sarg. of New
Zealand (Hair and Beuzenberg,
1958) , and also n = 11 in the Chilean
L. chilensis Endl. (Hunziker, 1961)
on the other side of the Pacific, evi-
dence that this odd number of chro-
mosomes persisted through epochs of
time and wide spatial isolation.
The Sequoia-Taxodium group
(Table 26) represents another 11-
chromosome conifer cluster. It is now
split into the genera Cryptomeria and
Metasequoia on the west side of the
Pacific, and Sequoia, Sequoiaden-
dron, and Taxodium on the east side.
Miki (1965) showed that counterparts
of Sequoia and Sequoiadendron are
present in Tertiary beds in Japan. In
past ages Sequoia relatives covered
north latitudes between 34° and 58°
around the earth (Florin, 1963:201
and maps 28-30) . Embryologies differ
in this primitive conifer group but
one senses great similarity between
the assumed genera of this cluster al-
though experimental evidence is
lacking.
240
CARNEGIE INSTITUTION
Among the broad-leaved trees of
the medium latitude and altitude the
members of the beech family, Faga-
ceae, stand out (Table 26). The oak
genus. Que reus, circles the earth at
medium northern latitudes, although
a few species spill over to the vicinity
of the equator and beyond at high
altitudes in Central America and
Indonesia (Oersted, 1871). Applying
conservative classification, between
300 and 500 species are known in the
genus, which is divided into six to
eight subgenera, each having several
subsections.
Hybridization is common between
oak species around the world, but the
genus contains many genetically sep-
arate species clusters. The clusters
are deciduous in the northern parts
of the oak belt, and evergreen in the
south.
The chromosome number is uni-
formly n = 12 in about 58 species of
Qua reus, Castanea, Castanopsis, and
Fagus, a situation similar to that of
the pine species.
The white oaks in Table 26 ex-
emplify one widely distributed de-
ciduous cluster of oak species, namely
Que reus alba L. of eastern North
America, Q. lobata Nee of California,
Q. crispula Nee of Japan, and Q.
robur L. of Europe. Members of dis-
tinct oak clusters may grow adja-
cently without intercrossing.
The beeches, Fagus, have probably
only one species cluster around the
earth, but hybrids between the Euro-
pean Fagus silvatica L. and the North
American F. grandifolia Ehrh. are be-
ing produced by shipped pollen at the
Horsholm Arboretum in Denmark,
which suggests a close genetic rela-
tionship.
The southern hemisphere genus
Nothofagus is morphologically a close
counterpart to Fagus. It crosses the
Pacific between South America, New
Zealand, and Australia. The southern
hemisphere beeches differ from Fagus
in having 13 instead of 12 pairs of
chromosomes as seen in the geneti-
cally closely related red and mountain
beeches N. menziesii (Armstrong and
Wylie, 1963). Nothofagus possibly
differentiated chromosomally from
the northern beeches very long ago,
but their evolutionary history is still
unknown.
Other genera that cross the Pacific
and have closely related species on
both sides are the magnolias, the bass-
woods, and the maples. Their chromo-
some numbers follow unusual odd
basic series that remain the same on
both sides of the Pacific (Table 26).
Such innate constancy, correlated
with the generic morphologies and
persistence within the latitudinal belt
on both sides, testifies to coherence in
heredity that persisted through geo-
logic ages as plant families moved
around the world.
Low latitude tree species. The
greatest number of the world's recog-
nized tree species, possibly some
60,000 to 80,000, have remained
within a belt of some 50° in latitude
on both sides of the equator. They
are here because most of them are
tropical and unable to tolerate even
occasional mild frosts. This charac-
teristic contrasts sharply with the
trees of the high latitude group that
exist in regions that are subjected to
frosts eight to nine months of the
year.
The low latitude tree species occupy
the forests on both sides of the Pacific
and the oceanic islands, extending
through southeastern Asia to Africa
and to both of the Americas, includ-
ing the Caribbean Islands. Many fam-
ilies and orders are restricted to these
latitudes. Those families that extend
into the temperate zones have evolved
genera of different kinds within the
temperate zones.
Within the low latitudes the tree
lines are at remarkably low altitudes
(Clausen, 1963), and species native
DEPARTMENT OF PLANT BIOLOGY 241
TABLE 27. Examples of Trees of Low Tolerance, Low Latitudes, 25 'N to 25 °S
(Major Portion of World's Trees, Possibly 50,000 to 70,000 Species)
Eastern Hemisphere Pacific Ocean Western Hemisphere
ARAUCARIACEAE, n = 1 3
ARAUCARIA species in Australia, New Zealand Peru, Brazil
PODOCARPACEAE
Podocarps: PODOCARPUS-DACRYDIUM,
New Guinea to 3600 m, New Zealand Peru, Brazil
RANALES, buttercup-magnolia relatives;
DRIMYS, Tasmania, New Guinea to 4000 m Peru, Brazil
ROSALES
Cunonias: WEINMANNIA, New Zealand, Hawaii Andes, Brazil
Rosaceae, POLYLEPIS leptophylla dry Andean, 3000 m
Legumes, SOPHORA, New Zealand, Hawaii, 3000 m high Andes
PARIETALES, camellia relatives
TERNSTROEMIA, CLUSIA Brazil, tree line 1900 m
MYRTALES, myrtles, melastomes
EUGENIA, ca. 1 000 species
Africa, India, Indonesia, Australia, Pacific Islands Mexico, Andes, Brazil
RUBSALES, coffee and camphor relatives
COFFEA, Congo, Ethiopia
CINCHONA Peruvian Andes, 3700 m
Missing at low latitudes: willows, walnuts, birches, alders, oaks, beeches, chestnuts.
there may show apparent damage terials from which trees capable of
even before the freezing point of surviving at high altitude might have
water has been reached. evolved. Characters used in plant
At the equator only very few tree classification are generally morpho-
species occur at altitudes beyond 3000 logical. One must therefore conclude
meters (Table 27). Such trees include that physiological adjustments to cli-
certain Araucarias and Podocarps, matic zones are somehow tied to the
Drimys, which is remotely related to morphology of the major plant orders
the buttercups and magnolias ; the so- and families. These gross ad just-
called pepper tree, Schinus molle, of ments are retained as members of
the Anacardiaceae ; the camphor tree, such families migrate across oceans.
Cinchona, of the Peruvian cloud for- The genus Eugenia of the myrtle
ests ; and Polylepis leptophylla, a tree- family provides an illuminating ex-
like plant of the rose family, through- ample. It has evolved about 1,000 spe-
out the drier Andes. The base of cies that cover both of the Americas
Table 27 indicates also that typical at low latitudes — besides 100 in the
mid-latitude tree families miss the Caribbean Islands, 45 in the Pacific
low latitudes completely. Islands, about 34 in Australia and
The many thousands of tree species Southeast Asia, 150 in the Indies, and
that occur only below the frost line 45 in Africa. It has been able to move
suggest that there is no lack of ma- across oceans, but has not extended
242
CARNEGIE INSTITUTION
into cooler latitudes. The large num-
ber of species indicates a high degree
of morphological variability, but this
and many other genera appear to be
trapped climatically.
In the discussion above it has been
necessary to amplify somewhat the
concept of vegetation in relation to
latitudinal and altitudinal zones. Veg-
• ions overlap, permitting mem-
bers o( the hardiest group to adjust
to habitats within the medium lati-
tude zone by occupying the higher
altitudes or cooler local habitats. Sim-
ilar overlapping is found between the
medium and low latitude plants, so
that certain species of a cluster may
spill over to a different belt.
It is necessary in studies of this
kind to apply a relatively conserva-
tive concept of species and genera.
Modern tendencies to split species,
genera, and families obscure the
broad, deep-seated physiological and
ecological relationships that are built
into the various heredity patterns
and hold the natural entities together
even on their worldwide migrations.
References
Armstrong, Jean M., and Ann P. Wylie,
A new basic chromosome number in
the family Fagaceae, Nature, 205,
1340-1341, 1963.
Iba, Shigeru, Studies on tree improve-
ment by means of artificial hybridiza-
tion and polyploidy in Alnus and Pop-
. Hull. Oji Inst. Forest Tree
Improvement, l, 70-165, 1966. Oji
Paper Co., Kuriyama, Hokkaido, Ja-
^n.
n, Jens, Tree lines and ^erm plasm
-a study in evolutionary limitations,
Proc. Natl Acad. Sci., 50, 860-868,
1963.
Clausen, Jens, Population studies of al-
pine and subalpine races of conifers
and willows in the California hiprh
Sierra Nevada, Evolution, 19, 56-58,
1965.
Critchfield, William B., and Elbert L.
Little, Geographic distribution of the
pines of the world, U.S. Dep. Agr.
Forest Ser., Misc. Pub. 991, iii + 93
pp., maps 1-51, 1966.
Florin, Rudolf, The distribution of coni-
fer and taxad genera in time and
space, Acta Hort. Berg., 20, 122-312,
1963.
Gram, K., C. Muhle Larsen, C. Syrach
Larsen, and M. Westergaard, Contri-
butions to the cytogenetics of forest
trees, Roy. Vet. Agr. Coll. Yearbook,
Copenhagen, 44-58, 1941.
Hair, J. B., and E. J. Beuzenberg, Con-
tributions to a chromosome atlas of
the New Zealand flora, New Zealand J.
Sci., 1, 617-628, 1958.
Helms, Anna, and C. A. Jorgensen, Bir-
kene paa Maglemose, Botan. Tidsskr.,
39, 57-134, 1925.
Hunziker, Juan H., Estudios cromosomi-
cos en Cupressus y Libocedrus, Rev.
Invest. Agr. 15, 169-185, Buenos Aires,
1961.
Larsen, C. Syrach, The employment of
species, types and individuals in for-
estry, Roy. Vet. Agr. Coll. Yearbook,
Copenhagen, vii + 151 pp., PI. I-VI,
1937.
Little, Elbert J., and Francis I. Righter,
Botanical descriptions of forty artifi-
cial pine hybrids, U.S. Dept. Agr.
Forest Service Tech. Bull., 1345, 47 pp.,
Washington, B.C., 1965.
Miki, Shigeru, Sequoiadendron prima-
rium Miki n. sp. and Sequoia couttisiae
Heer from Tertiary beds in Japan,
Bull. Mokugawa Women's Univ., 13,
1-7, 1965.
Oersted, A. S., Bidrag til Kundskab om
Egefamilien., Vidensk. Selsk., Skr. 5
Raekke, Naturv. Matem. Afd., 9, 335-
506, Tab. I-VIII + 1 map, 1871.
Osborn, A., An interesting hybrid coni-
fer, Cupresso-Cyparis leylandii, J. Roy.
Hort. Soc, 66, 54-55, 1941.
Righter, Francis I., and Palmer Stock-
well, The fertile species hybrid, Pinus
murraybanksiana, Madrono, 10, 65-69,
1949.
DEPARTMENT OF PLANT BIOLOGY 243
Tatewaki, Misao, Forest ecology of the Wright, Jonathan W., Summary of tree-
islands of the North Pacific Ocean, breeding experiments, U.S. Dep. Agr.,
J. Fac. Agr., Hokkaido Univ., 50, 371- Forest Serv., Northeast. Forest Exp.
484, PL I-XXX, 1958. Sta., Sta. Paper No. 56, 47 pp., 1953.
STAFF ACTIVITIES
Several botany classes from the visiting" with friends and colleagues
University of California at Davis and of up to 40 years' standing, Clausen
at Santa Cruz, and from Stanford, was introduced to Japanese investi-
visited the laboratory and the moun- gators in cytogenetics, ecology, taxon-
tain stations during the year. Demon- omy, and plant breeding at the
strations and talks were given by the universities in Tokyo, Kyoto, and
Experimental Taxonomy group. The Sapporo, and the National Institute
Bay Area Biosystematists group met of Genetics at Misima.
at the laboratory in January. Studies of the Japanese forest veg-
The greenhouses and the garden at etation were made possible by a spe-
the laboratory were used by Dr. Peter cial traveling symposium September
Raven of Stanford for his studies of 4 to 9 through the Japan Alps. The
the evening primrose family. Mrs. symposium subject was "The Ecologi-
Mary Mantuani, a graduate student cal Basis of Nature Conservation in
at Duke University, started a com- Alpine and Subalpine Zones." About
parative study of water relations of 70 biologists from different countries
ecological races of Solidago at participated in the symposium orga-
Mather. nized by Professor Makoto Numata of
The Northern California Photobi- Chiba University. After the tour, Dr.
ology and Photochemistry Group vis- Clausen and Professor Misao Tate-
ited the Department in January and waki of the University of Hokkaido
heard short talks from each of the studied the forests of Hokkaido for
Photosynthesis group. Dr. Brown about a week.
served as vice president of the Group On September 17 and 18 Dr. Shi-
this year. geru Chiba, Director of the Kuriyama
Dr. Jens Clausen attended the Institute for Improvement in Forests,
Eleventh Pacific Science Congress in and Dr. Clausen discussed principles
Tokyo. He also attended the Interna- of forest genetics. Dr. Chiba showed
tional Symposium on Plant Biosys- the modern Institute and the exten-
tematics at Tokyo University, sive experiment grounds owned and
arranged by the International Organi- operated by the Oji Paper Company,
zation of Plant Biosystematics. While
BIBLIOGRAPHY
Amesz, Jan, and David C. Fork, Quenching Bjorkman, Olle, and Paul Holmgren, Photo-
of chlorophyll fluorescence by quinones in synthetic adaptation to light intensity in
algae and chloroplasts, Biochim. Biophys. plants native to shaded and exposed
Acta, 1US, 97-107, 1967. habitats, Physiol. Plantarum, 19, 854-859,
Amesz, J., see also Fork, David C. 1966.
Anderson, Jan M., see also Fork, David C. Clausen, Jens, Stability of genetic charac-
Bjorkman, Olle, Comparative studies of ters in Tragopogon species through 200
photosynthesis and respiration in ecologi- years, Trans. Proc. Botan. Soc. Edinburgh,
cal races, Brittonia, 18, 214-224, 1966. UO, 148-158, 1966.
244
CARNEGIE INSTITUTION
C, David C., and Jan Amesz, Energy
ansfer between photosynthetic units of
pigment system 1 in algae, Biocliim.
phys. Acta. k:. 366-368, 1967.
Fork. David C, Jan Amesz. and Jan M. An-
derson, Light-induced reactions of chloro-
phyll /> and P700 in intact plants and
chloroplasi fragments, Brookhaven Syvnp,
\. 19, 81-94, 1966.
Fork, David C, and Yaroslav de Kouchkov-
sky. The 518-m/i change and its relation
the photosynthetic process, Photochem.
PhotobioL, 5,609-619, 1966.
Fork. David C, sec also Amesz, Jan.
':.. C. S., Review of Molecular PJiysics
in Photosynthesis, by Roderick K. Clayton,
A . S ntist, 54, 332A-334A, 1966.
French. C. S. (Inventor), and Charlton M.
Lewis (Registered Patent Agent), Plane
Table Plotter, a map making device, U.S.
Patent =3,302,293, 1967.
rich, C. S.. Chloroplast pigments, in
Biochemistry of Cliloroplasts, Vol. I,
edited by T. W. Goodwin, Academic Press,
London, pp. 377-386, 1966.
French, C. S., Review of The Cldorophylls,
edited by Leo P. Vernon and Gilbert R.
Seely, Science, 154, 1318, 1966.
French, C. S., Die Photochemische Nutzung
der sichbaren Strahlung durch zwei Licht-
reaktionen der Photosynthese, Novo Acta
Lcopoldina, N.F., 31, 169-187, 1966.
French, C. S., see also Pickett, J. M.
Hiesey, William M., Review of The Genetics
of Colonizing Species, edited by H. G.
Baker and Ledyard Stebbins, Quart. Rev.
Biol., 41, 418-419, 1966.
Holmgren, Paul, see also Bjorkman, Olle.
de Kouchkovsky, Yaroslav, see also Fork,
David C.
Lewis, Charlton M., see also French, C. S.
Pickett, J. M., and C. S. French, The action
spectrum for blue-light-stimulated oxygen
uptake in Chlorella, Proc. Natl. Acad. Sci.
U.S., 57, 1587-1593, 1967.
SPEECHES
Amesz, Jan, Spectrophotometry measure-
ments of reactions of photosynthetic inter-
mediates, Seminar, Department of Bot-
any, University of Illinois, Urbana,
Illinois, October 11, 1966.
Bjorkman, Olle, Photosynthetic adaptability
to light in plants native to shaded and
exposed environments, Seminar, Botany
Department, University of California,
Davis, California, December 6, 1966.
Bjorkman, Olle, Adaptation of photosynthe-
to light intensity in sun and shade
ecotypes of Solidago virgaurea, Seminar,
Department of Biological Sciences, Simon
Praser University, Burnaby, B. C, Can-
ada, January 26, 1967.
Bjorkman, Olle, Ecological adaptation of
photosynthesis, Seminar, Department of
Biology, University of California, Santa
Cruz, California, May 17, 1967.
Brown, Jeanette S., Chlorophyll fluorescence
emission of a diatom, Pacific Slope Bio-
chemical Conference, University of Cali-
fornia, Davis, California, June 17, 1967.
Clausen, Jens, Biosystematic consequences
of ecotypic and chromosomal differentia-
tion, International Symposium on Plant
Biosystematics, Tokyo, Japan, August
25, 1966.
Clausen, Jens, Species clusters in tree vege-
tations around the Pacific, Eleventh Pa-
cific Science Congress, Tokyo, Japan,
August 27, 1966.
Fork, David C, Studies on components in
the electron-transport chain of photosyn-
thesis, Seminar, Botany Department, Uni-
versity of California, Davis, California,
November 22, 1966.
French, C. S., Forms of chlorophyll in algae
and mutants, Pacific Slope Biochemical
Conference, University of California,
Davis, California, June 17, 1967.
Hiesey, William M., Genetic vs. environ-
mental effects on photosynthetic rate of
clones of contrasting climatic races of
Mimulus, American Society of Plant
Physiologists, AIBS Meeting, College
Park, Maryland, August 17, 1966.
Hiesey, William M., Cytogenetic and eco-
physiological studies in Mimulus, Seminar,
Department of Biology, California Insti-
tute of Technology, Pasadena, California,
May 16, 1967.
Michel-Wolwertz, M. R., The different
chlorophylls extracted from plants, Pacific
Slope Biochemical Conference, University
of California, Davis, California, June 17,
1967.
Nobs, Malcolm, Cytogenetics and climatic
selection in Mimulus, Seminar, Depart-
ment of Genetics, University of Califor-
nia, Davis, California, October 27, 1966.
DEPARTMENT OF PLANT BIOLOGY
245
PERSONNEL
Biochemical Investigations
Staff: C. Stacy French, Director; Jean-
ette S. Brown, David C. Fork; James
H. C. Smith, Emeritus
Carnegie Corporation Fellows: Jan
Amesz,1 Ulrich W. Heber2
Institution Research Fellows: Marie-
Rose Michel-Wolwertz,3 Jean-Marie
Michel,4 James M. Pickett5
Technical Assistants: Alan de Schwein-
itz,6 Marion A. Koerper,7 Mark C.
Lawrence, Suzanne Parmelee,8 William
T. Rhodes9
Part-time Laboratory Helpers: Christine
M. Anderson,10 Stephen J. Fulder11
Experimental Taxonomy
Staff: Jens C. Clausen, Emeritus; Olle
Bjorkman, William M. Hiesey, Mal-
colm A. Nobs
Institution Research Fellow: Eckard W.
Gauhl12
Technical Assistants: Frank Nicholson,
Pamela Radford13
Summer Research Assistants: Oakley
Shields,14 Stephen G. Wood15
Clerical Assistant: Marylee Eldredge
Part-time Garden Helpers: Robert M.
Baker,16 James M. Barnes,17 John M.
Emmel,18 Lawrence W. Hart,19 Ralph
R. Liddle,20 John C. Nobes21
Accountant-Secretaries : Richard F. Gill,22
Clara K. Baker23
General Department Secretary: Lena R.
Barton
Mechanical Engineer: Richard W. Hart
Custodian: Jan Kowalik
1 Through April 21, 1967. From Biophysical
Laboratory, State University, Leiden,
The Netherlands.
2 From May 3, 1967. From Botanisches In-
stitute, Universitat Dusseldorf, Bonn, Ger-
many.
3 From Jan. 4, 1967. From Centre de Re-
cherches de Gorsem, Belgium.
4 From Jan. 4, 1967. From Centre de Re-
cherches de Gorsem, Belgium.
5 From Oct. 1, 1965. From Department of
Zoology, University of Texas, Austin.
6 From Oct. 12, 1966, through May 30, 1967.
7 To September 28, 1966.
8 Aug. 22, 1966, through June 16, 1967.
9 To September 30, 1966.
10 Through August 19, 1966.
11 July 5, 1966, through Aug. 26, 1966.
12 From April 11, 1967. From Botanisches
Institut, Universitat Frankfurt, Ger-
many.
13 October 24, 1966, through June 30, 1967.
14 Through September 10, 1966.
15 From June 15, 1967.
16 To September 1, 1966.
17 From June 8, 1967.
18 Through July 29, 1966.
19 Through August 15, 1966.
20 May 9, 1967, through June 5, 1967.
21 Through October 21, 1966.
22 Through July 31, 1966.
23 From August 6, 1966.
PLATES
Plate 1 . Growth of bean seedlings in air containing low versus normal oxygen concentration.
Above. Plants at the beginning of a growth experiment (top row) and after 6 days in normal
air with 21 % O2 (second row from the top). Be/ow.- Corresponding matched plants before and
after growth in 2.5% 02-
Plate 1
Department of Plant Biology
O
o
o
o
o
O
CN
to
X
o
o
Plate 2
Department of Plant Biology
Plate 2. Bean seedlings after growing for 17 days in normal air (left) and corresponding
plants grown in 5% O2 (right).
Mount Wilson and Palomar
Observatories
Operated by Carnegie Institution of Washington
and California Institute of Technology
Pasadena, California
Horace W. Babcock
Director
OBSERVATORY COMMITTEE
Horace W. Babcock
Chairman
Carl D. Anderson
Jesse L. Greenstein
Eobert B. Leighton
Allan E. Sandage
Olin C. Wilson
Carnegie Institution Year Book 66, 1966-1967
Contents
Introduction 253
Observing Conditions 254
Solar Physics 255
Solar magnetograms .... 255
Oscillatory motions in the solar
atmosphere 256
High-resolution magnetograms . 258
Cinematography 258
Eclipse observations .... 259
Coronal physics 259
Continuous observations . . . 259
Spectroscopic study of sunspots . 259
Planets, the Moon, and Comets . . 260
Mars 260
Infrared photometry and colorim-
etry of planets 260
Differential infrared spectrometry
of the Moon 261
Daytime infrared . colorimetry
program 261
Comets 261
Stellar Spectroscopy and Photom-
etry 261
White dwarfs ,261
Dwarf K and M stars .... 262
Photoelectric flux measurements
at the center of stellar H and
K lines 262
Dwarf Me stars 263
Dependence of rotation on age for
stars of solar type . . . 263
Stellar evolution and the conserva-
tion of angular momentum . 264
Structure and evolution of W
Ursae Majoris stars . . . 264
Faint blue stars 264
A Bootis stars 265
Peculiar A stars 265
Stellar envelopes 265
Binaries 266
Helium abundance 267
Stellar composition 268
Variable stars 271
Infrared stellar spectroscopy . . 272
Infrared photometry .... 272
Temperature scale 272
Absolute Spectrophotometry . . . 273
Star Clusters 273
Color-magnitude diagrams . . 273
Helium deficiency 274
"Blue stragglers" 275
NGC 6171 275
Other globular clusters . . . 275
Praesepe cluster 276
Cluster Stars with Anomalous
Spectra 276
Interstellar Gas and Gaseous Nebu-
lae 276
Interstellar absorption lines . . 276
Nova envelopes 277
Crab nebula 277
Extinction in emission nebulae . 278
High-latitude planetary nebulae . 278
Infrared Sky Survey 279
Galactic Structure 279
Galaxies 280
Distance modulus of NGC 2403 . 280
Variables in dwarf galaxies . . 281
Seyfert galaxies 282
Compact galaxies 282
Counter-jet in M 87 . . . .284
Ultrafaint blue stars .... 284
Catalogue of galaxies and clusters
of galaxies 285
Statistical analysis 285
Classification of clusters of gal-
axies 286
Special cluster 286
Supernovae 286
Radio Galaxies 287
Cosmic-ray sources 287
Redshifts of radio galaxies . . 287
Energy distribution .... 287
Direct photography .... 287
X-Ray Sources 288
Scorpius X-l 288
Sources in Cygnus 289
Quasi-Stellar Sources .... 289
Spectroscopy 289
Parkes source 0237-23 .... 289
Spectral energy distribution . .291
Optical survey for radio-quiet
quasars or quasi-stellar ob-
jects 291
Space distribution 293
Optical variability 293
Infrared emission 294
Theoretical Studies 294
Heating of the solar atmosphere 294
QSO spectra 294
Observational cosmology . . . 294
Star formation 295
Guest Investigators 295
00-Inch Photometric Telescope . . 302
Optical design 302
Optical work 302
Mechanical parts 302
Site 303
Instrumentation 303
Image-tube spectrograph . . . 303
Multichannel spectrometer . . 303
Prime-focus corrector .
Astro-electronics laboratory
Other instrumentation for
large reflectors
Solar equatorial telescope .
Photographic laboratory .
Site Investigation ....
Image, motion at sites in Chi
Microthermal data .
Bibliography
Staff and Organization .
Le
. 303
. 304
the
. 305
. 306
. 306
. 306
306
307
308
313
INTRODUCTION
It is worthy of note that an increas- achieved during the past year. The
ing convergence is apparent between central region of our Galaxy, seen
the fields of stellar physics and solar from our position far out toward the
physics. The study of the sun, the edge, is heavily obscured by clouds of
only star that can be observed in de- dust and gas that are concentrated
tail, has provided facts about such near the plane of the spiral system. In
phenomena as sunspots, flares, the visible light it has been impossible to
chromosphere and corona, rotation, penetrate these clouds and to discern
internal circulation and mixing, mag- anything of the structure of the cen-
netic fields, the solar magnetic cycle, tral region. But in longer wave-
and transfer of angular momentum, lengths the Galaxy is far more trans-
both internal and external. As im- parent, and some years ago radio
proved tools and techniques are ap- astronomers detected a strong radio
plied, and as knowledge of the evolu- source, Sagittarius A, in the general
tion of stars accumulates, we find that direction of the center. Recently, Eric
stellar analogues of solar phenomena Becklin, a graduate student at Cal-
are becoming the subjects of specific tech, using an optical telescope with
researches. The example of flare a special detector, discovered that the
stars is well known. The spectro- galactic center is an emitter of far
scopic study of chromospheres of infrared radiation. Subsequently, he
dwarf M and K stars has been fruit- and Dr. G. Neugebauer, working in
ful, and efforts are under way to de- the wavelength region 1 to 3 microns
tect the stellar analogues of the solar (some 2 to 6 times the wavelength of
magnetic cycle. Convection zones, in- visible light), have mapped an area
ternal circulation, and mixing are ac- about equal in apparent size to that
tive subjects. It was the study of the of the moon in which they have
ever-expanding magnetic patterns on plotted the observed infrared radia-
the surface of the sun, scarcely a tion-intensity pattern including some
decade ago, that gave proof of the six intensity peaks of differing shape
continuous outflow of plasma (ion- and sharpness; the most intense peak
ized gas with associated magnetic is coincident with Sagittarius A. This
fields) from the sun — an effect that is the first evidence on the structure
has since acquired the convenient de- and brightness of the nucleus of the
scriptive name of solar wind. Loss Galaxy, a region that one suspects
or transfer of angular momentum by may have been the site of extraordi-
a star, with its resultant effects on nary physical activity in the early
stellar rotation and "equatorial ac- stages of the development of our
celeration," is a subject near the fore- stellar system.
front of research. The origin of the At various observatories, spectra
planets of the solar system has long have now been obtained for approxi-
been a subject for speculation or for mately 120 quasars, including both
elaborately structured theories re- radio-emissive (QSS) and radio-quiet
mote from the opportunity for deci- (QSO) types. Data on photometry
sive observational tests, and this is a and color exist for a comparable num-
situation that is slowly but surely ber. QSO candidates can be distin-
changing. guished by their colors, and new ones
In the field of galactic structure, can be discovered almost at will. Data
a significant new result has been by Luyten and Sandage suggest that
253
254
CARNEGIE INSTITUTION
the total number of QSOs over the
entire sky as faint as magnitude
19.7 (blue) is about 100,000. Quasars
are receiving the best efforts of a
growing number of observational as-
tronomers and theoreticians at many
places, and several significant find-
ings have emerged during the past
year from the work of the staff and
visitors at Mount Wilson and Palo-
mar. Among these findings, a few
have been selected for brief mention.
Schmidt has investigated the space
distribution of quasars, and he esti-
mates that their "local" space den-
sity, including those observed as radio
sources, is 1 per 10s cubic megapar-
secs. From a study of 33 brighter
sources from the 3C Catalog for
which redshifts are known, he finds
that the space density of quasars
must be increasing with redshift. The
excess density factor has been placed
at 15 for redshift of 0.5 and at about
100 for redshift of 1. Such a result
has great implications for cosmology.
New redshifts have been observed
for a number of QSSs and QSOs; the
largest yet found is for the Parkes
radio source 0237-23, a candidate
suggested by Bolton and first ob-
served optically by Arp. Later spec-
troscopic studies by Greenstein and
Schmidt show that there are, in ad-
dition to the emission lines for which
AX/X0 = z = 2.22, two sets of ab-
sorption lines giving z = 2.202 and
1.956, respectively. Lines from ex-
cited states occur, suggesting rela-
tively high density and some colli-
sional excitation. Interpretive studies
based on spectroscopy of other qua-
sars showing absorption features
have provided some preliminary indi-
cation of temperature and density in
these remarkable objects.
Practically all quasars vary in lu-
minosity, and two that are highly
variable have been investigated in
detail by Oke, who measured absolute
energy distributions in their spectra.
The source 3C 279 varied as much as
0.25 magnitude per day and by nearly
2 magnitudes during the year. Oke's
new measurements confirm a recent
finding by Sandage, Westphal, and
Strittmatter that it is the continuous
spectrum that varies, while the emis-
sion lines remain constant.
These and other observational find-
ings are gradually providing the con-
ditions that circumscribe the de-
velopment of a physical model for
these bizarre objects that were un-
known only seven years ago.
OBSERVING CONDITIONS
Rainfall at Mount Wilson was 55.59
inches, as compared with the 63-year
mean of 35.33 inches. Total snowfall
was 68 inches. The temperature was
highest (90°F) on August 5 and low-
est (19°F) on April 12 and 20. The
heaviest snowfall in one day was 10
inches on April 11, and precipitation
was registered on 17 days of that
month.
At Palomar Mountain 45.53 inches
of rain fell, of which 24 inches oc-
curred during a 5-day period begin-
ning on December 3, 1966. Total
snowfall was 37 inches.
Observations with the 100-inch and
200-inch telescopes were scheduled on
363 nights of the year, and no lost
TABLE. 1.
Observations
Telescope
Complete
Nights
Partial
Nights
Total
Hours
Worked
60-inch
100-inch
200-inch
205
250
210
66
53
87
2213
2607
2585
MOUNT WILSON AND PALOMAR OBSERVATORIES
255
time because of mechanical failures scopes, as limited by sky conditions,
was reported. The time actually used is shown in Table 1. The total number
for observations with the major tele- of nighttime hours was 3800.
SOLAR PHYSICS
Routine solar observations were
made by Thomas Cragg, Merwyn
Utter, and Robert Howard on 307
days. The records of various kinds
made between July 1, 1966, and June
30, 1967, were as follows:
Direct photographs
288
Ha spectroheliograms, 30-foot
focus
578
K2 spectroheliograms, 30-foot
focus
592
Magnetograms
299
Sunspot drawings
277
Magnetic classifications of sunspot
groups were made visually on 226
days during the year.
Alan Title assembled and arranged
for publication a series of the best
spectroheliograms that had been ob-
tained in the course of the six-year
program of the study of solar velocity
fields initiated by Leighton.
Solar Magnetograms
The new magnetograph, image
scanning, and data systems are in
operation at the 150-foot tower. Dur-
ing a magnetogram scan, data are re-
corded on magnetic tape. The data
are analyzed by means of the IBM
7094 computer at the California In-
stitute of Technology. Partial support
of this program is provided by the
Office of Naval Research.
From the velocity data the differ-
ential rotation of the sun is deter-
mined along with the limb redshift.
This is done by means of a least-
squares solution for the coefficients of
a sine-squared, sine-fourth-power ex-
pansion of the solar latitude. For the
standard magnetogram with 17 arc-
second resolution, there are about
11,000 data points on the solar disk,
each with a very accurate line-of-
sight velocity measurement. Thus the
differential rotation of the scan can
be determined with precision. It is
planned to continue observing the
solar differential rotation throughout
the current solar cycle. J. W. Harvey
of the High Altitude Observatory is
collaborating with Howard on this
project.
Another quantity that is analyzed
is the integrated magnetic field over
various parts of the solar disk. Dur-
ing this year the magnetic fields in
both polar regions have been meas-
ured and prove to be quite weak —
about 1 gauss negative. The magnetic
flux in each polar region is about 1020
maxwells. It is expected that contin-
ued observations of these quantities
and the magnetic flux in other lati-
tude zones will provide valuable data
about large-scale solar activity.
Solar magnetograms are computer-
plotted for each day's observation.
Figure 1 shows an example of a full-
disk magnetogram. These magneto-
grams are being published monthly
along with Mount Wilson sunspot
magnetic classifications and solar
data from other observatories in the
publication Solar -Geophysical Data
issued by the Environmental Science
Services Administration of the U. S.
Department of Commerce, and ob-
tainable through the Superintendent
of Documents, Government Printing
Office.
"Dopplergrams" are plotted from
the velocity data after subtracting
the background solar rotation. An ad-
ditional Dopplergram is plotted after
the data have been averaged over
area in order to smooth out some of
the small-scale features and to dis-
256
CARNEGIE INSTITUTION
MOUNT WILSON OBSERVATORY MAGNETOGRAM
DATE (RS.T.) 6/18/67 21
TIME(UT) 22.29-23.78
w
Fig. 1 . This full disk magnetogram was drawn by a computer-controlled mechanical plotter from
data obtained at the 1 50-fool solar tower. Solid contour lines enclose regions of positive field and
dotted contour lines enclose regions of negative field. This map shows many bipolar regions in
various stages of development. All magnetic activity on the surface of the sun first appears as
compact, bipolar regions. As these magnetic regions age, they expand in characteristically in-
clined patterns. The remnants of the trailing parts drift slowly toward the poles to give the sun its
apparent dipolar magnetic field.
close large-scale low-amplitude mo-
tions on the solar surface. Both of
these Dopplergram programs were
started near the end of the year and
no results are yet available.
Oscillatory Motions in the Solar
Atmosphere
Howard has completed an exten-
sive study of oscillatory motions in
the solar atmosphere. He used the
solar magnetograph, which can be
adapted as a very sensitive detector
of Doppler shifts of a spectrum line.
Using the line A5250 (Fe I) and ob-
serving at fixed points on the sun's
image, he and the other solar observ-
ers obtained nearly 250 hours of ob-
servations with apertures of various
sizes. The amplitudes of the well-
known 300-second oscillations are
about 25% weaker in regions where
the magnetic field is greater than 80
gauss as compared to regions where
the field is less than 10 gauss. No dif-
ference in the frequency spectrum
MOUNT WILSON AND PALOMAR OBSERVATORIES
257
of the oscillations could be found
between strong-field and field-free re-
gions, thus indicating that the oscilla-
tions are not basically of hydromag-
netic origin. Howard suggests that
the oscillating motion occurs only
where the magnetic field is effectively
absent, and that the reduction in
amplitude in a strong-field region is
the effect seen when a certain frac-
tion of the area within the aperture
does not contribute to the oscillation.
There is evidence to support the view
that at least a large part of the mag-
netic flux observed on the solar sur-
face exists as small isolated bundles
of lines of force within which the field
is high — perhaps 500 gauss.
One of the features of the new
magnetograph exit-slit assembly is
provision for observing two spectrum
lines simultaneously and for obtain-
ing magnetic and velocity signals
from each. Although the correspond-
ing portion of the data system has
not yet been built, it is possible to re-
cord in an analogue fashion the ve-
locity signals from two lines simul-
taneously. This was done as a part of
the oscillation program, and signifi-
cant phase differences amounting at
times to as much as 30 seconds were
found upon comparison of the signals
from two lines formed at depths dif-
fering by perhaps 100 km in the solar
atmosphere. The higher-level line of-
ten showed oscillations of a slightly
higher frequency.
A short-period oscillation is found
superposed on the 300-second oscilla-
tion. These "SPOs" come in bursts
that last for a minute or two and have
peak-to-peak amplitudes of about
0.05-0.10 km/sec. Their periods fall
in the range 1-5 seconds. All at-
tempts to explain them as seeing ef-
fects or instrument effects have
failed. The horizontal scale of the
SPOs is smaller than that of the 300-
second oscillations, and they are more
nearly isotropic. They do not repre-
sent a high-frequency tail of the 300-
second oscillation, but appear to be a
completely separate phenomenon and
are not markedly dependent in their
occurrence or strength upon phase or
amplitude of the longer-period oscil-
lation. These observations were re-
corded with a digitizing interval too
long to provide power spectra of the
SPOs, but it is clear from their ap-
pearance that the SPOs have a much
wider range of frequency than do the
slower (300-second) oscillations. The
amplitude of the SPOs in the solar
atmosphere must be large, and they
must contribute greatly to the non-
radiative energy flux. It is suggested
that they represent a large "micro-
turbulence" line-broadening effect.
In an attempt to establish that the
SPOs were a solar phenomenon and
not some subtle effect of daytime see-
ing, Dr. William C. Livingston of the
Kitt Peak National Observatory col-
laborated with Howard in a joint ef-
fort to observe the bursts simultane-
ously at Kitt Peak and Mount Wilson.
On June 2 both observers set the
aperture of their magnetographs on
the same spot in the solar atmosphere
and observed the variation in line-of-
sight velocity for several hours. A
comparison of the two observations
indicated that the SPO bursts came
always at the same time at the two
observatories. Even some individual
SPOs could be identified on both
records. This is strong evidence that
the oscillations are solar in origin.
Dr. J. M. Wilcox of the Space Sci-
ences Laboratory of the University
of California at Berkeley has been
cooperating with Howard in a study
of velocity and magnetic fields on the
solar surface, using the new scanning
and data system of the 150-foot tower
telescope. The observations consisted
of repeated scans over a straight line
on the solar disk. The data have been
analyzed to reconstruct the structure
and time dependence of the velocity
258
CARNEGIE INSTITUTION
fields, magnetic fields, and brightness
at many points along the scan line.
First results indicate that the dis-
tances over which 5-minute oscilla-
tions are in phase may be as great as
30.000 km, contrary to previous esti-
mates of oOOO- or 4000-km cell sizes
for such oscillations.
Wilcox and Howard are also co-
operating on a study of the differen-
tial solar rotation as determined from
autocorrelation analysis of magnetic-
field data from 1959 to the present.
When the magnetic data are digitized
and an autocorrelation is run for one
latitude,, the correlation peak due to
rotation is. in general, easily identi-
fied. Thus the latitude dependence of
the rotation of magnetic features may
be studied. The long interval over
which data are available will provide
a clue to possible time dependence
of solar differential rotation. It is
planned that the autocorrelation tech-
nique applied in this way will also
be used in a study of the large-scale
features of solar activity and of the
solar magnetic-field distribution.
High-Resolution Magnetograms
Rust has developed computer pro-
grams for analysis of high-resolution
(2-10 arc-second) magnetograms of
small regions on the solar disk. These
"fine-scan" programs extend the
rapid data-gathering capacity of the
magnetograph, guider, and tape-re-
cording system in the 150-foot tower
to a wider range of solar studies than
was possible with programs designed
for whole-disk scanning only. De-
tailed maps of the velocity and mag-
netic fields near Ha filaments, and in
active regions, will be produced en-
tirely by computer techniques from
data obtained during the summer of
1967. Some sequential magnetic scans
have already been obtained in an ac-
tive region during a small flare. The
scans have a 10-minute time resolu-
tion and a 10-arc-sec spatial resolu-
tion. Rust hopes to coordinate obser-
vations of this kind with flare patrol
sequences made by Lockheed observ-
ers in order to examine the connec-
tion between changes in the photo-
spheric magnetic fields and the
sudden energy releases occurring in
flares. In another application of the
fine-scan programs, Rust and Howard
are studying the magnetic fields in
the sun's polar regions.
Cinematography
A new Ha- filter was installed at the
150-foot tower and used in periods of
good seeing. Assisted by Lawrence
Anderson and Terence Allen (Caltech
undergraduates), Zirin obtained a
number of interesting films of various
solar phenomena. One of the principal
goals was to study oscillatory motion
in Ha. Another goal has been to un-
derstand the nature of the chromo-
sphere as disclosed by limb observa-
tions. By using a narrow-band
suppressor plate, Zirin showed that
he had been wrong in the interpreta-
tion of the "second limb" on filter-
grams, and that Simon and White had
been correct in their contention that
the inner limb is the photosphere seen
in filter side bands. But he was able
to confirm another "inner limb" —
that of the low chromosphere — ex-
tending from zero to 1500 km above
the photosphere. This appears as a
dark band crossing active regions at
the limb. Because the chromospheric
emission profile broadens with height,
the dark band is most apparent in
pictures in the wing of Ha. The
broadening of Ha in the upper chro-
mosphere may be connected with the
high velocities of the oscillating ele-
ments.
In addition to the cinematography
of quiet sun features, a number of
films of solar flares were obtained.
Outstanding among these were high-
resolution records of the class 3 flare
of August 28, 1966. One of the inter-
MOUNT WILSON AND PALOMAR OBSERVATORIES
259
esting features is the upward erup-
tion of a filament preceding the flare,
a type of phenomenon reported in
other instances by the Lockheed Solar
Observatory. The X-ray and radio
emissions do not begin until the flare
brightening occurs, some time after
the prominence erupts.
Eclipse Observations
The predicted infrared coronal
lines Al.43 p [Si X] and A3.03 /x [Mg
VIII] were successfully observed by
Munch, Neugebauer, and McCammon
during the total solar eclipse of No-
vember 12, 1966, from the airplane
"Galileo" of the National Aeronautics
and Space Administration. For this
purpose, the 50-cm Ebert spectrom-
eter used in the past for observa-
tions of infrared stellar spectra was
fitted with two PbS detectors at suit-
able exit slits and attached to a fixed
F5 parabolic reflector. Solar light was
delivered to the instrument by a gyro-
stabilized heliostat through a special
5.5-inch CaF2 window. On the basis
of the measured wavelengths and in-
tensities of the lines, it appears that
they should be observable from high-
altitude ground stations. The absolute
intensity of the [Si X] line is about
one tenth that of the strongest coro-
nal lines in the photographic region
and does not correspond to the
strength of the permitted silicon spec-
trum in the rocket ultraviolet.
Coronal Physics
Zirin worked on a new interpreta-
tion of the excitation of coronal lines.
It has been known that coronal lines
may be excited by photospheric radi-
ation or by electronic collisions, but
the significance of the variation in
this process from line to line has not
been clear. There are a number of
coronal lines (for example, Fe XV
A7060, Fe XIII A10798) for which
photospheric excitation is impossible
or very small, either because the
transition from the ground state is
completely forbidden or because the
required energy is too great. These
lines will therefore be excited by col-
lisions only, and their intensity will
vary as the square of the electron
density Ne. They will thus have very
low intensity at low coronal density.
Ordinary coronal lines may be excited
by photospheric radiation and will
vary as Ne at low densities, so they
will be relatively stronger. The effect
is marked for the ratio of Fe XIV
A5303 to Fe XV A7060. The latter
has always appeared anomalously
weak compared to the strong Fe XV
resonance line at 283 A; we see now
that it can have appreciable density
only for Ne > 2 X 109.
Continuous Observations
For continuous study of the chro-
mospheric network and of active re-
gions, Leighton and Zirin proposed
some years ago the establishment of
one or two temporary observing sta-
tions well spaced in longitude around
the world. With a grant from the Na-
tional Science Foundation, Zirin has
now set up such a station at Tel
Aviv University, Israel, with a small
telescopic pulse camera and Ha filter.
Its operation will be the responsibility
of the Tel Aviv staff.
Spectroscopic Study of Sunspots
Makita analyzed the green system
of a TiO band in the spectrograms of
a large sunspot which Dr. Olson of
the University of Illinois obtained
with the 150-foot tower telescope in
September 1966. Only approximate
equivalent widths were measured,
since the lines were weak and most
of them were disturbed by blending.
Coarse analysis, plotting the line
strength as a function of the rota-
tional quantum number, gave the
loci of the maximum and minimum
widths. The most probable rota-
tional temperature thus obtained was
3060°K.
■260
CARNEGIE INSTITUTION
PLANETS, THE MOON, AND COMETS
Ma rs
For the 19G7 opposition of Mars
an extensive series of spectroscopic
observations had been planned. Un-
usually poor weather conditions,
however, permitted the completion of
only about one quarter of the pro-
gram. Using the 100-inch coude scan-
ner and a Fabry-Perot interferom-
eter with a free spectral range of
1.96 A and a finesse of 75, Munch has
measured photoelectrical^ the inten-
sities of lines in the weak C02 band
5r3 at AS690 A with a precision higher
than has been previously available by
photographic means. A complete re-
duction of the data is now in progress.
In the spectral range between 2.0
and 2.4 a, Munch and Neugebauer at-
tempted to obtain spectra of different
areas on the surface of the planet.
For this purpose there was con-
structed a chopping system that com-
pares the spectra of two areas within
a scanning spectral channel and also
in a broadband monitor channel.
Since the defining apertures are
fixed at the entrance plane of the
100-inch coude spectrograph, an all-
reflecting field rotator had to be made
also. The spectral resolution used was
about 40 A, and the angular resolu-
tion could be as small as 1 arc-sec,
although poor seeing conditions did
not allow full advantage to be taken
of it. The data obtained have not yet
been processed, but preliminary ex-
amination shows the existence of con-
spicuous broadband differences in
planetary surface brightness. Using
the same chopping scheme, difference
spectra between the center and the
limb of the planet were attempted;
these should effectively cancel the ef-
fects of telluric absorption.
Beginning in January 1967, inter-
ference-filter observations in the re-
gion of the 2-ji bands of C02 have
been made periodically by T. Hilge-
man, a graduate student at the Cali-
fornia Institute of Technology, with
the 24-inch telescope at Mount Wil-
son. The effects of telluric absorption
are calibrated by lunar observations
at various air masses. The purpose of
the observations is to determine the
degree of constancy of the C02 con-
tent of the Martian atmosphere. The
work will be continued as late in the
year as possible.
Infrared Photometry and Colorimetry
of Planets
Murray observed Mars successfully
in the evening twilight with the 200-
inch on May 2 and 3, 1967. Other
scheduled attempts in April and May
were unsuccessful due to weather
conditions. A maximum spatial-reso-
lution map of the 8-13-/X brightness
temperature was obtained on May 2,
and similar maps in the 9-, 11-, and
13-/X intervals were acquired the fol-
lowing night. Venus was observed at
similar air mass for comparison in
each case. These results constitute the
most detailed information on the
surface temperatures of Mars yet
acquired.
Murray observed the 8-13-/* emis-
sion of the shadow of Jupiter's satel-
lite, Io, as it crossed the planet's disk
during the evening twilight of May
2. No anomalous emission was ob-
served, the upper limit being esti-
mated at a flux not greater than twice
that of the disk itself over an area
equal to the geometric size of the
shadow. The origin of the earlier ob-
served shadow enhancements remains
unexplained.
A maximum nighttime surface
temperature of Mercury of 155°K
was estimated by Murray from his
earlier observations.
MOUNT WILSON AND PALOMAR OBSERVATORIES
261
Differential Infrared Spectrometry
of the Moon
A. F. H. Goetz completed his doc-
toral thesis in which he achieved
greater than an order-of-magnitude
improvement in the detection of
spectral variations of emissivity in
the 8-13-/A region between different
localities on the lunar surface. The
24-inch telescope was used. A detec-
tivity approaching 0.2% in relative
emissivity was obtained over most of
the wavelength interval and two new
anomalous areas were discovered.
Previously reported anomalies were
not confirmed. Such emissivity varia-
tions probably are indicative of dif-
ferences in age and possibly also in
composition of lunar surface mate-
rials; they constitute an important
new source of information about
selenology.
An attempt by Goetz to extend the
method of differential infrared spec-
trometry to the observations of Mars
with the 200-inch telescope was un-
successful because of the strong sur-
face variations of intensity charac-
teristic of planetary images.
Daytime Infrared Colorimetry
Program
Because the terrestrial planets are
such strong infrared emitters, and be-
cause whole-disk colorimetry at 9, 11,
and 13 ji potentially can provide im-
portant information not otherwise
available about these objects, an ef-
fort has been under way for over a
year by Murray and students to de-
velop a suitable colorimetric tech-
nique involving direct comparison
with the sun. One major problem has
been solved — that of rapid intercom-
parisons with the sun in order to min-
imize the effects of rapidly varying
daytime infrared transmission. The
solution involves the use of a 1-inch
aperture coelostat system mounted on
the 24-inch telescope. With this sys-
tem, the solar image can be rapidly
brought into the photometer field of
view for signal comparison without
moving the telescope off the planetary
object being observed.
Comets
A comet, 1966e, was discovered
with the 48-inch schmidt telescope by
Rudnicki. The periodic comet Grigg-
Skjellerup 1966f was rediscovered,
also with the 48-inch schmidt by
Rudnicki, upon the request of Dr. G.
Sitarski of the Polish Academy of
Science.
STELLAR SPECTROSCOPY AND PHOTOMETRY
White Dwarfs
A third list of confirmed white
dwarfs, with colors and spectra, has
been prepared by Eggen and Green-
stein, bringing to 202 the number of
those with relatively complete data.
The space motions show that relics of
all population types are present. They
have U, V, W components for 132
stars; one half of the stars have
tangential velocities less than 33 km/
sec. The tangential velocities of 22%
are less than 20 km/sec, and 10%
are greater than 100 km/sec. In spite
of the selection effect arising from
proper motion, halo stars make up
only 10%. The peculiar spectra occur
very often in the higher-velocity
group.
Special efforts have been made to
find red degenerate stars, types DG,
DK, or DM, or red DC stars. The
success of this effort is still small;
only 7 very red stars are known with
confirmed peculiar spectra, of which
three have known parallax. The anom-,
262
CARNEGIE INSTITUTION
alously high luminosities of this
group suggest that our discovery
technique is still poor; the faintest
new object, LP 658-2, has about the
same luminosity as do Wolf 489 and
van Biesbroeck 11 (about +15.5 =
My). If the colors are interpreted as
from a blackbody continuum, the
radii derived are large, ranging up-
ward from 0.01 Rq for G38-27 and
W489, to 0.10 Kg for G5-28 and
H67-138. A possibility exists that
these large red stars are objects of
small mass, contracting past the main
sequence (although no spectral anom-
alies, such as emission lines, are
noted). The paucity of red degener-
ate stars might become understand-
able if the theory of freezing of the
nucleons into a solid lattice (Vila,
Van Horn) should predict a low spe-
cific heat at the low interior tempera-
ture of these highly evolved stars.
Greenstein and Miss Trimble have
measured the Einstein gravitational
redshift of a group of 53 white
dwarfs. Lack of information on the
true radial velocity makes the result
statistical, as does the low accuracy
of the velocities. Thirty-seven stars
with hydrogen lines gave a mean
systematic redshift (K-term) of
— 66 km/sec. The helium-rich stars
have a lower K-term and systematic
differences from line to line, caused
by pressure shifts. If an approximate
temperature scale is derived from
UBV colors, the median radius is
0.0107 Ro. The mass derived is about
0.98 uto; this surprisingly high value
is reduced to 0.86 n\o for a helium
core and 0.71 ttl© for an iron core.
These masses are appreciably higher
than those derived heretofore.
Dwarf K and M Stars
Spectroscopic observation of nu-
merous dwarf K and M stars, mostly
from the catalogs of Vyssotsky et al.,
has been completed by O. C. Wilson.
The plates have all been measured
for radial velocity and this material is
now in press. It turns out that the
probable error of a single observa-
tion, defined as one plate measured
by two individuals, is ±1.8 km/sec,
and the overall agreement with the
system of the General Catalogue is
excellent.
Careful eye estimates of the in-
tensity of the chromospheric Ca II
emissions have been made with a
comparator and certain standard
stars. Internal consistency checks in-
dicate that these estimates are suffi-
ciently accurate to yield a reliable
frequency distribution of the line
strengths. This work is not yet com-
pleted, however, and is being con-
tinued.
Photoelectric Flux Measurements at
the Center of Stellar H and K Lines
The procedure involved in this
work with the coude scanner of the
100 -inch telescope was described
briefly in Year Book 65 (p. 141). As-
sessment of the first year's accumula-
tion of data is now nearly complete.
During this period, observation was
restricted almost entirely to a sample
of about 140 F2-G2 stars from the
catalog of Stromgren and Perry, in-
cluding all of those objects in which
bright H and K components have been
seen on spectrograms of 10 A/mm.
The average probable error of a
single measurement of flux in either
the H or the K line is found to be
±1.8%. This is fairly satisfactory
but is still more than twice as large
as the probable error expected on the
basis of statistics alone. It is hoped
that the observational accuracy can
be improved. In any event, the ac-
curacy is quite adequate to show the
general characteristics of chromo-
spheric emission throughout the
main-sequence band in the portion of
the IT-R diagram under investigation.
One result of considerable interest
is the ratio of the chromospheric H
MOUNT WILSON AND PALOMAR OBSERVATORIES 263
and K components. It turns out that, period. It is tentatively concluded,
for the stronger emissions, the K therefore, that (1) periodic or quasi-
intensity is about 25% greater than periodic light variations are found in
that of H. Presumably this means some dMe but not dM stars, and (2)
that the gas masses in the stellar these variations are not related to
chromospheres are not optically thick binary characteristics. Rejecting vol-
in the Ca II lines. ume pulsation and eclipses as mech-
The first year's observations pro- anisms to explain the light variabil-
vide first-epoch values of chromo- ity, it seems that the most promising
spheric flux for the emission-line model appears to be the rotational
objects. It is proposed to continue modulation of a star with a nonuni-
observing these in order to see form distribution of surface bright-
whether long-term variations (ana- ness.
logues of the solar cycle) can be de- ^ 7 - „ , ,. A ,
tected. In addition, it is hoped to ex- Dependence of Rotation on Age for
tend this investigation soon to giants btars of boiar l ype
and supergiants. The rotational velocities of a large
sample of field stars with spectral
Dwarf Me Stars types between F2 IV, V, and G3 IV,
Following P. F. Chugainov's dis- V have been obtained by Kraft from
covery of periodic light variability in coude spectrograms of resolution
Popper's flare star, HDE 234677, about 6 km/sec. It is shown that the
Krzeminski and Kraft have searched average rotational velocity is higher
for similar variations in brightness among those with Ca II emission than
among emission-line and nonemis- among those without. Since there is
sion-line stars in the range of spectral strong evidence from the work of
type dK7 to dM3.5. Six emission-line Wilson that stars with Ca II emission
and three nonemission-line objects are younger on the average than most
were observed extensively. It was solar-type stars, a picture is advanced
found that: (1) Three emission-line in which rotation declines with ad-
objects are certainly variable; Chu- vancing age. Kraft proposes that
gainov's period (P = 3<?83) for HDE magnetically coupled winds, similar
234677 is confirmed but the light am- to the solar wind, are responsible for
plitude is now smaller. The other two this deceleration even after the star
objects are probably periodic or takes up a position on the main se-
quasi-periodic variables; BD + 34° 106 quence. The winds are presumably
has a period of 3^92 and AC + 31° generated only in stars that have sub-
70565 has a period of 4<?65. The re- surface hydrogen convection zones
maining emission-line stars showed and concomitant chromospheres.
no variability in excess of 0^03. (2) Estimates of rotational velocity
Nonemission-line stars had no varia- among solar-type stars in the Hyades
bility in excess of the same limit. and Pleiades that confirm this general
Spectroscopic observations indicate picture were also obtained. The mean
that BD+34°106 and HDE 234677 rotation in the Pleiades turns out to
are both double-line spectroscopic bi- be the highest of all groups considered
naries; AC + 31°70565 shows no ra- and averages about 40 km/sec for
dial velocity variation. The orbital stars of mass 1.20 nio. The time scale
period of BD + 34°106 has not been for the reduction by a factor of 2 of
found, but the orbital period of the rotational velocities of Pleiades
HDE 234677 is 5<?981, a number in- stars is equal to the age of the Hyades,
commensurate with the photometric about 4 X 10s years.
264
CARNEGIE INSTITUTION
The results do not bear directly on
Dicke's suggestion that the solar wind
decelerates only the outer convection
zone of the sun and, thus, that the in-
terior retains a large fraction of its
"primordial" angular momentum.
This is because the derived time scale
refers to the conditions of wind den-
sity and magnetic-held strength that
were presumably appropriate when
the sun was as young as the Pleiades
stars, not to the present values of
these quantities. The results do indi-
cate, however, that solar-type stars
rotate slowly, in comparison to more
massive main-sequence stars, as a
result of the deceleration produced
by winds.
Stellar Evolution and the Conserva-
tion of Angular Momentum
Kraft has given observational evi-
dence that whenever an evolving star
encounters regions of the H-R dia-
gram where subsurface convection
zones play a role in the internal struc-
ture, rotational deceleration takes
place as a result of magnetically
coupled winds. These findings apply
to contracting T Tauri variables,
main-sequence stars below F5 V, and
for post-main-sequence red giants be-
tween about GO III and G5 III. The
rate of this deceleration, dJ/dt, can
be estimated from the difference in
rotational velocity of two points on
an appropriate evolutionary track,
and the time elapsed between occupa-
tion of the two points. Using the re-
cent evolutionary calculations of
Iben, Kraft is able to show that, very
roughly, dJ/dt = k (LK)17, where LK
is the absolute rate of emission of
energy in the K2 emission. Thus the
rate of deceleration increases ap-
proximately as the 1.7 power of the
strength of the chromospheric emis-
sion. Qualitatively this means that
the torque exerted by the wind in-
creases as the activity of the chro-
mospheric regions increases ; this, in-
deed, is almost self-evident when we
consider the sun's magnetic activity
and the solar wind.
Structure and Evolution of W Ursae
Majoris Stars
Current ideas on the formation,
structure, and evolution of the main-
sequence contact-binaries of type W
UMa were reviewed by Kraft. It was
shown from relative star counts that
W UMa stars cannot now be trans-
ferring mass from one component to
the other, and cannot, therefore, have
evolved from close but not contact bi-
naries satisfying the mass-luminosity
relation. The possible formation and
evolution of Lucy's contact configura-
tion were considered. Kraft suggests
that such a configuration evolves on
the nuclear time scale of the primary
into a U Geminorum binary; losses
of angular momentum in magneti-
cally coupled winds are probably too
small to be important on this time
scale. Some consequences of this evo-
lutionary identification were explored
for U Gem variables. In particular, if
W UMa stars last on the average
about 3 X 10° years, then star counts
indicate that U Gem stars last about
1 X 10° years, and it is not easy to
reconcile this result with the lifetime
of 106 years appropriate to the blue
components in these systems. Accre-
tion heating of the blue star is sug-
gested as a possible way out of this
dilemma, and some mass-loss mech-
anisms appropriate to the red com-
panion are considered.
Faint Blue Stars
Krzeminski carried out UBV pho-
tometry of about 30 faint blue stars
in the regions of the north and south
galactic poles, chosen from the To-
nantzintla catalogs. One of the ob-
jects, Ton S 120, was found to be an
eclipsing binary with a period of
3h28m. Its light curve resembles that
of Nova DQ Herculis, but with much
MOUNT WILSON AND PALOMAR OBSERVATORIES 265
deeper eclipses. The eclipse is partial, In particular, the "manganese" stars
its depth is 2.5 mag, and its duration occupy a very small range in this
is about 30 minutes. Secondary eclipse diagram.
has not been detected. The continu- Deutsch, with Abt, Conti, and Wal-
ously trailed spectra obtained by lerstein, has discussed the double-line
Kraft showed large radial-velocity spectroscopic binary HD 98088. The
variations. primary star is known to be a periodic
In order to improve the period and magnetic variable star. It is also a
features of the light curve, Krze- spectrum variable, with the lines of
minski observed photoelectrically HZ Sr II exhibiting two intensity maxima
29, a peculiar subdwarf with very per cycle. As has been known for
wide and complex He I lines recently some years, the periods of orbital mo-
discovered by J. Smak to be a vari- tion, magnetic reversal, and spectrum
able star with period 17.5 minutes; variation are identical, indicating
it is supposed that this object is a that the primary star is a rigid ro-
close binary system. tator. The spectroscopic and photo-
. metric observations are satisfied by a
X Boons btars model in which the primary has nor-
Oke has used photoelectric spectro- mal Balmer lines and mass (2.2 ttl©)
photometric observations and Hy pro- for an A3 V star. The secondary star
files to determine effective tempera- lies near A8 V, but its spectrum is
tures and gravities of the known X probably abnormal at the K line and
Boo stars. These stars are near spec- at D.
tral type A0, are rapid rotators, and Dr. Stephen Strom of the Harvard
have abnormally weak metal lines, and Smithsonian Observatories is
Two of the stars, y Aquarii and 9 Hy- collaborating with Sargent on an
drae, have energy distributions indi- abundance analysis of the sharp-lined
eating that they are binaries. All peculiar A star, HD 204411. Equiva-
stars, apart from 2 Andromedae, are lent widths were measured for about
near or slightly above the initial main 1300 lines. The star appears to have
sequence, and the values of log g are a normal composition,
similar to those of stars on the initial
main sequence or the brightest stars Stellar Envelopes
in the Hyades. This class of star is Mrs. Boesgaard has made a study
therefore unevolved or only slightly of the line profiles of Ha and the so-
evolved. The star 2 And has a low dium D lines on direct-intensity
value of log g and also appears to be microphotometer tracings of 4 -
below the main sequence. 8-A/mm spectrograms of over 40 late-
type stars. These stars may be divided
Peculiar A Stars m^-0 three groups: (1) Both the
Jugaku and Sargent completed an sodium D lines and the Ha- line
analysis of photoelectric scans of the are asymmetric, with the short-
continuous energy distributions of 17 wavelength side being the steeper;
peculiar A stars and 5 A- and B-type (2) the Ha line only, and not the so-
normal stars. Temperatures and dium lines, shows asymmetry; and
gravities were derived for all the (3) both the hydrogen and sodium
stars. It was found that most of the lines are completely symmetric.
Ap stars have atmospheres character- (There are no cases so far of a star
istic of dwarf B stars. The Ap stars showing a symmetric Ha profile but
are segregated in (log g, Teff) dia- asymmetric sodium lines.) The three
gram according to their peculiarities, groups are about equally populated
266
CARNEGIE INSTITUTION
with stars in this random sample. On
the basis o£ measurements made of
half widths, equivalent widths, the
velocity difference between the core
and the center of the line (defined at
the half-intensity point), the stellar
absolute magnitude measured by the
Wilson-Bappu method, and previously
known spectral types and R-I color
indices, the groups can be interpreted
in the following way. The asymmetry
of the almost double-line structure of
the sodium D lines probably arises
from the presence of a circumstellar
envelope, and the asymmetry in Ha
may be due to the chromospheric be-
ginnings of this envelope. The second
group of stars may be in an earlier,
more rudimentary stage of mass loss
and development of a circumstellar
envelope. The stars in group 2 and
the "normal" stars in group 3 seem
similar in almost all respects except
for the Ha asymmetry. The stars in
group 1 tend to have higher absolute
magnitudes and later spectral types.
These stars also have larger equiva-
lent widths for a given half-width
than the stars in the other two
groups. The velocity differences be-
tween the core and the line center
range from —5 to —15 km/sec, com-
mensurate with ejection velocities as-
sociated with circumstellar envelopes.
The doubling of absorption lines
noted in the photographic infrared
spectra of S stars by Merrill and
Greenstein is being extensively stud-
ied by Tsuji. Available are Palomar
spectrograms of R Andromedae
(postmaximum in 1956 and 1957) or
R Cygni (postmaximum, 1964), and
Okayama 74-inch telescope spectra of
R Cyg (maximum, 1964) and R And
(1965 maximum and postmaximum).
All postmaximum spectra clearly
show line doubling, while the violet-
displaced component is absent or very
weak at maximum, except in Rb I and
K I lines, which seem to be produced
in an outermost, very low-tempera-
ture expanding shell. In other ele-
ments, the violet lines are at about
— 35 km/sec. High excitation lines
have the strongest violet components,
with an indicated temperature near
3000°K, as against 2500°K for the
red component. The excitation tem-
perature of the red component de-
creases after maximum, so that the
decrease of H- opacity makes this
shell transparent, revealing the hotter
layers exemplified by the violet com-
ponent. The strong emission of the
infrared Ca II triplet is located at
— 35 km/sec, the same velocity of ex-
pansion as the violet absorption com-
ponent. As this expansion is highly
supersonic, it will produce a shock
wave, after whose passage the kinetic
temperature rises to 20,000-30,000°
K, sufficient to ionize Ca II and H.
The emission lines of H and Ca II,
therefore, can be produced in an ex-
panding shell. A tentative model of an
S-type long-period variable consists
of (1) an outermost, very low-tem-
perature expanding shell (violet-dis-
placed resonance lines), (2) a normal
extended atmosphere (red-displaced
absorption component) , (3) high- tem-
perature, expanding emitting-layer
immediately behind the shock front
(emission lines), and (4) an expand-
ing layer behind this emitting-layer
(violet-displaced absorption compo-
nent) .
Binaries
Heintze has obtained 20 A/mm
spectra of the early-type eclipsing
variables U Coronae Borealis and
V 448 Cygni. In cooperating with J.
Grygar of the Ondrejov Observatory,
he found the limb-darkening coeffi-
cient of the 09.5 component of SZ
Camelopardalis to be near 0.45, and
that of the B2(?) component to be
between 0.3 and 0.6.
Conti has completed his study of
HR5110 (HD 118216). This F star,
which has emission lines from the sec-
MOUNT WILSON AND PALOMAR OBSERVATORIES 267
ondary at H and K and H« filled in consistent with a spectral type of B3,
by emission, is interpreted as a semi- (2) that the Si/H abundance ratio is
detached close binary, viewed at low normal, and (3) that the He/C
inclination. From an anomalous in- abundance ratio is lower than normal,
f rared color, the secondary is deduced An attempt was made by Stephen
to be a subgiant K star. This binary Strom to determine the helium con-
is much like the well-known eclipsing tent of subdwarfs by using the loca-
systems, Z Herculis and S Velorum. tions of 62 field subdwarfs in the
There should be about five times as theoretical H-R diagram. The posi-
many semidetached systems in exist- tions of the subdwarfs in the H-R
ence as are known through eclipses, diagram relative to the Hyades main
It may be possible to discover more sequence permits the estimation of a
of these from infrared color anoma- range of possible helium abundances.
lies. These estimates are based upon ho-
mology relations that follow from re-
Helium Abundance cent interior calculations. Accurate
positions of the subdwarfs on the
Sargent has obtained coude spec- H-R diagram were determined
trograms (18 A/mm or 9 A/mm) and through a combination of the best
continuum scans of four blue halo available parallax data and a calibra-
stars (Feige 56 86, and 92, and tion of the relation between corrected
AC + 9°6-12) which earlier lower- (B — V) and effective temperature,
dispersion studies have shown to be The direct determination of the effec-
deficient in helium and heavier ele- tive temperature for the subdwarfs
ments. This material, together with was accomplished by comparing spec-
spectrograms of similar southern trum scans covering the wavelength
stars, is being analyzed in coopera- region 5000-7500 A, with fluxes pre-
tion with Dr. Leonard Searle of the dieted from model atmospheres.
Mount Stromlo Observatory. The These scans were obtained with Oke's
main aim is to set an upper limit on scanner at the Cassegrain focus of
the helium abundance with which the the Mount Wilson 60-inch reflector.
Galaxy originated. This has impor- Values of effective temperature were
tant cosmological implications. determined for stars having a wide
Sargent obtained 4.5 A/mm spec- range of ultraviolet excesses, and
trograms of 38 Draconis B9p, which these values allowed a calibration of
Eggen has described as one of the the effective temperature-corrected
brightest horizontal-branch stars in (B — V) relation. A discussion by
the old disk population. The star has Strom and Strom (Astrophys. J., in
a very low v sin i; an analysis of its press) of the permissible values of
spectrum is being made with the helium abundance all but rules out
assistance of Mr. S. Adelman. zero as a choice for the initial galactic
Dr. Peter Strittmatter of the Insti- helium content,
tute for Theoretical Astrophysics, Spectral scans at selected 50 A
Cambridge, England, is collaborating bands from AA3300-10800 A have
with Sargent on a study of the spec- been obtained by Visvanathan for the
tra and compositions of the five B study of the continuum of helium
stars with weak helium lines in the stars BD + 10°2179, HD 160641, and
Orion I Association. Preliminary re- BD + 13°3224. The construction of
suits for i Ori B are (1) that the Si models of pure helium atmosphere in
Ill/Si II lines ratios and the Hy the temperature range of the ob-
profiles indicate a temperature that is served stars is in progress.
268
CARNEGIE INSTITUTION
Stellar Composition
Conti has completed his studies of
the lithium-beryllium ratio in main-
sequence stars. The new observations
of the Be line essentially doubled the
number of stars discussed previously.
The image tube on the 73-inch camera
was employed for about 7 of the
faintest stars in the program. Using
the parameter Li Be rather than the
abundance of either element, one can
atangle the depletion mechanism
from the effect of different initial
light-element content. The change of
lithium-beryllium among main-se-
quence stars can be interpreted as an
age effect due to depletion of the
lithium. The depletion mechanism for
lithium is more effective in G stars
than in F stars, i.e., the rate is more
rapid. The variations in the beryllium
abundance from star to star can be
interpreted as different initial
amounts of this element (and lith-
ium). Some otherwise normal stars
contain no detectable Be or Li; pre-
sumably this is the result of initial
stellar conditions unfavorable to the
production of light elements.
Mrs. Boesgaard has found a very
strong line due to Li I on a 6.8 A/mm
spectrogram of T Sagittarii taken by
Greenstein with the 200-inch coude
shortly after maximum light in Au-
1966. This was independently
discovered by P. C. Keenan on a 13.5
a mm Palomar spectrogram. The
lithium line A6707 has a measured
equivalent width of 1.46 A; a rough
curve-of-growth analysis yields a
Li/Ca abundance ratio that is about
two orders of magnitude larger than
the Greenstein-Richardson solar
value. Several other spectrograms of
M-S stars and S stars are being an-
alyzed for lithium abundance.
Equivalent-width measures and
model-atmosphere studies on 8 sharp-
lined Pleiades A stars have been com-
pleted by Conti and Strom. Two turn
out to be normal in abundance, like
Vega, and have microturbulence
amounting to a few kilometers per
second. The other 6 stars all have a
higher turbulence, up to 7 km/sec.
These stars are also all overabundant
in the s-process elements (Sr, Zr, Y,
and Ba) by about a factor of 4 with
respect to Vega. Three of the stars
are additionally deficient in Sc, Ca
(and C, with less certainty) and, in
their composition, appear to be the
early-type analogues of Am stars.
It is difficult to understand these
anomalous abundances in a young
cluster in terms of current astrophys-
ical ideas.
The isotopic abundances of mag-
nesium can be studied by the MgH
molecular bands. Mrs. Boesgaard has
taken spectrograms of 10 bright late-
type stars at 3 A/mm in the region
of the 1,0 (A4844) and 0,0 (X5211)
bands for this purpose. The terres-
trial isotope ratios are Mg24/Mg25/
Mg26 : 78.6/10.1/11.3. The stellar
amounts of Mg25 and Mg2G are at
least 10%, and in many cases it ap-
pears that 20% for both MgJ25 and
Mg26 fits the observations better.
In the photographic infrared the
crowding of lines in late-type stars is
not serious and metallic lines can be
found that are free of blends. The
relative abundance of zirconium-
titanium can then be determined by a
method that should be insensitive to
the model. Tsuji finds that Zr/Ti in
R Cygni is 30 times larger than in
the sun, and 20 times larger in R
Andromedae. These are lower limits,
since zirconium is more strongly de-
pleted by molecule formation than is
titanium. For example, at 2500 °K
and a gas pressure of 103 dynes/cm2,
with solar composition, the fraction
of Zr in the form of neutral atoms is
only 5 X 10"8, the rest being ZrO. For
titanium, nearly half of the atoms are
neutral. Thus even when Zr I/Ti I is
10, the Zr/Ti ratio may be 103. Mrs.
MOUNT WILSON AND PALOMAR OBSERVATORIES 269
Boesgaard has collected spectra of K, variable S star, HR 1105. A Palomar
M, M-S, S, and Ba II stars to deter- spectrogram at 2A/mm, obtained by
mine the Zr/Ti ratio, and finds the Greenstein. has been elaborately
expected increase from M to M-S to studied. After improvement of the
S-type stars. She is also studying HR molecular constants, machine compu-
2028 (classified as M2+ IV by Yama- tation of the detailed band structure
shita), which is a high-velocity, over about 20 A was made for a wide
lithium-rich, and apparently metal- variety of isotopic abundances of
weak star. zirconium. Limits to the reliability of
On a well-exposed postmaximum the isotopic composition arise from
spectrogram of R And, taken in 1956, plate noise and from blending of ZrO
the red CN system is fairly strong in rotational lines with weak TiO fea-
the violet-displaced velocity system, tures, lines of CaH, and a very few
Tsuji finds the lines to be similar to atomic features. A few definite con-
those of cool carbon stars, and that elusions may be drawn: (1) The
the C13N14 features suggested by abundance of O18 is not large, in that
Wyller can be used. The C12/C13 ratio no trace of the displaced ZrO18 band
of R And is smaller than the terres- heads can be found. (2) It seems un-
trial value, but not as low as the likely that the terrestrial ratios of
equilibrium value suggested by the Zr isotopes hold. (3) It is impossible
CNO cycle. Thus incomplete proc- to exclude a wide variety of Zr iso-
essing of the C12 suggests that the tope ratios, but mixtures in which
C13 («, n) O16 neutron source is oper- either Zr93 (an unstable isotope) or
ating in the S stars. those in which the heavier, even iso-
Coude spectra of M stars have been topes exist seem most likely to be
obtained in the photographic infra- present in HR 1105. It will be neces-
red by Tsuji; HD 95735, a high- sary to provide even greater spectral
velocity dwarf, is of particular inter- resolution to settle this important
est. Its lines are much weaker than question; either pure (and fainter)
those of giants. The only strong mul- S-type variables should be observed,
tiplets are K I (l),NaI (4),theCa or more precise spectrophotometry
II triplet, Fe I (60), and Ti I (33, profiles of HR 1105 will be needed.
68). Tsuji plans to compute realistic There are no wavelength shifts to be
M-star model atmospheres, and is expected in ZrO rotational lines, but
estimating the line opacity from rather changes of details in the line
molecular bands, such as TiO and profiles of individual, unblended lines.
MgH. The isotope splitting is small com-
In collaboration with L. H. Aller pared to the line width,
and K. Hunger, Jugaku is continuing Greenstein has obtained high-
the study of the early-type helium- dispersion spectra of several K stars,
rich star HD 160641. Between A3270 following a discovery by Spinrad at
and A4920, most of the lines identified the Lick Observatory that line
are of He I, He II, C II, C III, N II, strengths measured photoelectrical^
N III, O II O III, Ne II, Si III, Si were variable from star to star
IV, S II, S III, and Mg II. Equivalent among both giants and dwarfs. Spec-
widths of about 190 lines have been tra confirmed variable strength of the
measured and will be used for coarse Na I D lines and, most interestingly,
and fine analyses of this star. of the blue CN band. Two K dwarfs
Schadee and Mrs. Locanthi have have CN as strong as a giant. The
attempted to measure the zirconium other conspicuous peculiarity is the
isotope abundances in the bright non- increased strength of the strong
270 CARNEGIE INSTITUTION
pressure-broadened wings of Cr I, gram on G- and K-type stars by tak-
Mg I, and Xa I lines in both sub- ing spectra at 2 and 7 A/mm with
giants and dwarfs. The possible ex- the coude spectrograph of the 100-
planations are increased abundance inch telescope. The approximate
of metals, a high gas-to-electron pres- wavelength region is AA4300-6700.
sure, and. for CX, a high value of the The abundance analysis will be exe-
C 0 ratio. The high C/O ratio is most cured at the Astronomical Institute
surprising in unevolved K dwarfs. of the University of Amsterdam. He
The hydrogen-to-helium ratio in .is also studying spectrograms of S
horizontal-branch blue stars, in globu- Cephei at nine different phases, taken
lar clusters, and in halo stars re- at his request by Greenstein and
mains a major puzzle. Greenstein has Kraft. The plates cover the region
obtained long-exposure 90 A/mm AA3900-4900 and have a dispersion of
spectrograms of blue stars in M 13, 4.5 A/mm.
M 15, and M 92, for which Oke has After Dr. Bodo Baschek pointed
spectrophotometry scans. Except in out an error in their original formu-
M 13, the He I lines are very weak lation, Conti and Deutsch corrected
or absent in stars whose UBV colors their discussion of last year concern-
would demand their presence. The low ing the significance of photometric
He H ratio found by Sargent and line-strength indices in solar-type
Searle, by Sargent, and by Green- dwarfs of the disk population. The
stein and Munch seems to be con- new results still indicate that these
firmed. This direct contradiction of indices are highly sensitive to micro-
the big-bang prediction of a He/H turbulence and that the weak-line
ratio near 30 % remains as a serious characteristic may not always be
anomaly. One puzzling feature of the taken to indicate metal deficiency.
globular-cluster spectra was the ap- The authors now agree, however, that
pearance of the K line of Ca II in abundance-blanketing can have an
stars of negative B — V color. Radial- appreciable or dominant effect on
velocity measurements suggested that line-strength indices. In an individual
the K line was interstellar rather case, spectroscopic analysis may be
than stellar, and that interstellar Ca required to establish the relative im-
II lines probably are produced far out portance of turbulence-blanketing and
in the halo of our Galaxy, above the abundance-blanketing,
thin-layer gas model normally as- Conti and Deutsch have completed
sumed. such analyses for two pairs of stars:
If the globular-cluster blue stars A Serpentis and ft Canum Venatico-
are as hot as their colors indicate, rum at GO V, and r Bootis and y Ser-
therefore, the weakness of the helium pentis at F6 IV. In both pairs the
lines suggests a low ratio for He/H second star differs from the first
at the surface. A suggestion has been chiefly by showing weaker lines, as
made by G. Greenstein, Truran, and indicated by the Stromgren indices
Cameron that in nonrotating, non- mr and the ultraviolet excesses
convective B stars the helium sepa- S(U — B) . From spectrograms at
rates by diffusion downward from the 6.7 A/mm in the yellow, over 200 lines
hydrogen, and that the surface com- of common metals were measured for
position is not a reliable indicator of equivalent width. The curves of
the true He/H ratio in the gas out of growth for the stars in each pair were
which the globular clusters were directly compared. It was found that
formed. the stars of these pairs differ little in
In July, Koelbloed finished his pro- microturbulence but are distinctly
MOUNT WILSON AND PALOMAR OBSERVATORIES 271
different in metal content. In these vestigated appear to have signifi-
cases, then, the differences in turbu- cantly smaller masses than those to
lence-blanketing are very small and be expected for stars in this region of
the differences in abundance-blanket- the color-magnitude diagram, which
ing are appreciable. On the basis of would be evolving from left to right,
the revised calculations of Conti and Values of the pulsation constant com-
Deutsch, a quantitative agreement puted for these stars indicate that
that is reasonable is found between they are probably pulsating in a har-
the predicted index differences and monic mode higher than the funda-
the observed ones. The evidence no mental.
longer appears to support the earlier A member of a wide binary system,
conclusion of Conti and Deutsch that ADS 2849A = HD 24550, was found
microturbulence decays in solar-type by Dickens to be a variable with an
dwarfs on the same time scale as extremely small amplitude of 0.02-
chromospheric activity. 0.03 mag and a period of about 110
min. The absolute magnitude is de-
Variable Stars rived from that of the companion
Danziger and Oke have completed stars; the gravity and temperature
a study of the short-period variable inferred from the UBV colors show
VZ Cancri. Line-blanketing correc- that the star has an extremely small
tions were applied to the scan obser- pulsation constant, suggesting a high
vations and effective temperatures harmonic mode of pulsation,
and gravities were derived around A number of stars in the galactic
the cycle. The minimum temperature cluster NGC 6940 have been investi-
and the phase relations between light, gated photoelectrically by Dickens in
temperature, and velocity are similar the search for short-period variabil-
to those of cluster-type variables such ity. Four of the stars investigated so
as RR Lyrae, SU Draconis, and X far show some variability, and, in
Arietis. The surface gravity and the particular, VR 77 varies with an am-
period-density law suggest an ab- plitude of about 0.05 mag and a period
normally low value of Q and a low of about 0<?2. This star has a spectral
mass. VZ Cnc is slightly metal deft- type of F0, an absolute magnitude of
cient, but lithium is overabundant by +0.5, and lies near the blue end of
a factor of 25 relative to the sun. the Hertzsprung gap. It could well be
Danziger and Dickens have derived an example of a 8 Scuti variable, a
effective temperatures, surface gravi- type which heretofore has not been
ties, and rotational velocities for a identified in clusters,
number of recently discovered short- Dickens is obtaining spectrum
period variables of the 8 Scuti type, scans in the wavelength region 3500-
using photoelectric spectrum scans 8000 A for a number of bright stars
and coude spectra. Some of the new between A3 and F7 in order to inves-
variables have higher temperatures tigate the statistics of gravity and
and gravities than the previously temperature in the region of the
known variables of this type. Larger color-magnitude diagram occupied by
values of the projected rotational the 8 Scuti stars. Spectra of the same
velocity, v sin i, also occur among the stars are being obtained at the Royal
new variables and indicate that, in Greenwich Observatory in order to
the region of the color-magnitude provide accurate rotational velocities
diagram occupied by these stars, pul- and line profiles for the stars,
sation is not incompatible with rapid R. F. Garrison has been taking
rotation. About 50% of the stars in- classification dispersion spectra of
272 CARNEGIE INSTITUTION
most of the long-period variable stars lines. Among early- type stars (B8-
north of declination —30° with max- F), A10830 is found only in super-
ima brighter than 9th magnitude, giants so far, in particular /? Orionis
Spectrograms have been obtained cov- (absorption) and e Orionis (emis-
ering as much of their cycles as prac- sion) . The image tube that was used
ticable, with preference for the pre- formerly at the 144-inch camera has
maximum and maximum phases. This now been adapted to the 72-inch cam-
program is being carried out in col- era of the 200-inch coude spectro-
laboration with Deutsch and Keenan graph. This has made possible a sub-
who are continuing to obtain spectro- stantial increase in speed, so that
grams at 20 A mm in blue and violet observation of fainter stars is now
for studying systematic anomalies in proceeding.
various absorption lines and bands.
Infrared Photometry
Infrared Stellar Spectroscopy The search for 10-/* circumstellar
Using the 100-inch coude spectro- emission near a Ori has been contin-
graph and the detecting system em- ued by Westphal. Observations on
ployed for observing Mars, Munch three nights have shown no emission
and Xeugebauer have obtained addi- with a brightness temperature
tional spectra of a Bootis, a Herculis, greater than 100°K within 2 arc-min
p Pegasi, and x Cygni in the 2.0-2.4 of the star.
fi range. These spectra are being
studied for the purpose of identifying Temperature Scale
absorption features suspected in Heintze has emphasized a problem
lower-resolution spectra obtained already recognized by Oke and Searle.
earlier with the Cassegrain spectrom- The B — V colors of bright, nearly
eter. unreddened B stars and spectral
Vaughan and Zirin have continued scans suggest a temperature anomaly
their study of the He I A10830 absorp- between BO and AO. Heintze feels
tion in late-type stars as a key to the that the temperature scale must be
nature of chromospheres in these lowered, and that the surface gravi-
stars. The line has now been found in ties also must be lowered. He suggests
emission in 5 stars (e Geminorum, 12 that at BO, B3, and B5 the effective
Pegasi, c Herculis, y Draconis, and i temperatures are near 27,000, 16,500,
Aurigae), and temporal changes have and 14,500°K, respectively. The
been found in 2 stars (a Aquarii and y sharp-lined B stars that he has
Dra) . The data suggest that the ab- scanned suggest that for the main
sorption or emission originates in sequence log g < +4. This lower
discrete clouds or streams. The prin- temperature scale disagrees with that
cipal conclusions of the study remain given by the Balmer discontinuity
as stated last year. It has now been and the Hy profiles. There has been a
noted that a considerable fraction of similar discrepancy, known for some
the stars showing strong A10830 are time, in the sense that the surface
close binaries. The A10830 absorption gravity as deduced from observations
also appears relatively weaker in using even the most advanced model
dwarf stars. In many of the stars atmospheres has been too low for the
with A10830, this line shows a sizable mass of the unevolved main-sequence
velocity shift with respect to other stars.
MOUNT WILSON AND PALOMAR OBSERVATORIES 273
ABSOLUTE SPECTROPHOTOMETRY
A program of absolute spectro- guiding platform. A two-channel
photometry that will include a re- pulse-counting system has been con-
calibration of the flux of a Lyrae and structed for the program. This system
other bright standards has been ini- will continuously monitor the sky
tiated by Oke and Schild. The absolute background and make possible obser-
calibration is needed to make pos- vations in moonlight, when the 18-
sible the comparison of continuum inch schmidt telescope cannot be used
fluxes predicted by stellar models with effectively for photography,
spectrophotometric measurements of Standard sources to be compared
stellar fluxes. The need for an im- with stars include a standard lamp
proved calibration arises in the su- whose calibration has been referred to
perior accuracy with which stellar fundamental standards maintained
fluxes can be measured for compari- by the National Bureau of Standards
son with increasingly sophisticated and to a blackbody furnace operating
and accurate stellar models. at the melting point of platinum. The
During the report year, instrumen- furnace, still under construction, will
tation for the recalibration has pro- employ as a radiation source a cylin-
gressed under the direction of Schild. drical cavity 25 mm long with an
A 4-inch F3.3 Newtonian telescope aperture of 1 mm. The cavity will be
has been constructed and mated to the surrounded by platinum heated to its
existing prime-focus scanner of the melting point in an induction furnace.
200-inch telescope. Because of the For observations at Palomar, the fur-
short focus of the calibration tele- nace will be installed on the roof of
scope, lamp bulbs and other standard the machine shop where it will be
sources located as close as a half mile visible from the dome of the 18-inch
are readily compared with stars. To telescope.
ensure stability of optical properties Visvanathan and Oke have taken
and to minimize scattering by the ac- up a study of 25 stars in the Pleiades
cumulation of foreign matter on alu- cluster. The absolute fluxes in selected
minized optical surfaces, the calibra- 50-A bands from 3450 to 8000 A, and
tion telescope is hermetically sealed Hy line strengths, have been obtained
in a nitrogen-filled tube having quartz for 16 stars. These are being corn-
entrance and exit windows. The cali- pared with fluxes computed from
bration telescope and scanner can be model atmospheres to determine ef-
mounted on the Palomar 18-inch fective temperatures,
schmidt telescope, which serves as a
STAR CLUSTERS
Color-Magnitude Diagrams V ~ 17.5 mag to And the character of
The photometric survey of the the density distribution along the
bright parts of color-magnitude (C- horizontal branch and the steepness
M) diagrams of northern globular of the giant branch. Both features
clusters was continued by Sandage are valuable in the morphological
with the assistance of Katem, using description of clusters because they
photoelectric standards and photo- have previously appeared to be cor-
graphic interpolation techniques. Re- related with metal abundance,
connaissance studies of NGC 5897 The C-M diagram of NGC 5897 re-
and NGC 5053 were completed to sembles M 3, M 92, and w Centauri,
274
CARNEGIE INSTITUTION
with most of the stars along- the
horizontal branch confined blueward
of the RR Lyrae gap. The gap ap-
pears at V = 16.2, which indicates a
distance modulus of m — M ~ 15.7 if
Mv = =±0.5 for the RR Lyrae stars.
The reddening- appears to be very
small and is here neglected. The
cluster appears to be moderately
metal poor because AV of the giant
branch, read at B—V = +1.4, is 2.7
mag brighter than the horizontal
branch.
XGC 5053 is a sparse cluster in
which photometry can be carried
close to the center. Early work in col-
laboration with Dr. H. L. Johnson of
the Lunar and Planetary Laboratory,
University of Arizona, combined with
recent data, shows that the horizontal
branch occurs at V=16.9, which gives
m—M ~ 16.4, again assuming Mv =
— 0.5 for the RR Lyrae stars. The
morphological characteristics are sim-
ilar to NGC 5897.
Helium Deficiency
Sandage and Wildey completed a
study of the remote globular cluster
XGC 7006. The horizontal branch
occurs at V = 18.8, which indicates
the cluster is at the large distance of
45 kpc from the sun. Its distance from
the galactic plane is 15 kpc. The C-M
diagram is unique. The density dis-
tribution along the horizontal branch
would indicate high metal abundance
if the correlations, heretofore ac-
cepted, were to hold, because the
branch is very heavily concentrated
toward the red end, as in NGC 6171,
NGC 6356, NGC 6712, and 47
Tucanae. However, the giant branch
has &V = 2.6, which is 0.5 mag
larger than the value for strong-lined
clusters. Furthermore, spectrographic
data by Morgan show NGC 7006 to
be very metal poor.
The foregoing is a clear-cut viola-
tion of the correlation between metal
abundance and density gradient along
the horizontal branch, and shows that
a second parameter, in addition to
metal abundance, controls the charac-
ter of this branch. Theoretical calcu-
lations by Faulkner suggest that this
parameter is the variation of the he-
lium abundance. The sense of the
change is that NGC 7006 is deficient
in helium compared to the majority
of globular clusters in the halo.
Evidence from other remote clus-
ters, such as the Draco system with
data by Miss Swope, the Leo II sys-
tem by the same author, and all clus-
ters in the Small Magellanic Cloud by
Gascoigne, suggests that the NGC
7006 anomaly occurs from many ag-
gregates that have not participated
in the chemical evolution of the galac-
tic system. In this regard, NGC 7006
may not be gravitationally bound to
the galaxy, as evidenced by its ex-
tremely high radial velocity. If the
result is true that the helium abun-
dance differs inside and outside the
galaxy, this may be a clue to some of
the early events of the nucleosyn-
thesis.
Other clusters are known which
are certainly not bound gravitation-
ally to the galactic system, as shown
by an excess of kinetic energy in the
radial coordinate over the potential
energy required for negative total
energy via the viral theorem. A test
of the hypothesis of different helium
abundance inside and outside the gal-
axy can be made by determining
whether or not the NGC 7006 anom-
aly of the horizontal branch is present
in these aggregates.
Completing work begun at the Lick
Observatory, Sargent has discussed
low dispersion spectra of eleven blue
horizontal-branch stars in M 13, M
15, and M 92. The main conclusions
were: (1) In each cluster, helium is
deficient by a factor greater than
5-10, and (2) there are abnormally
strong interstellar K lines of Ca II
in the spectrum of stars in these
MOUNT WILSON AND PALOMAR OBSERVATORIES 275
clusters, which are all relatively un- its present rotation against the torque
reddened. The observation of helium exerted by the solar wind. The argu-
weakness in a star in M 15 is in con- ment tends to support Dicke's con-
flict with the observation by O'Dell, tention that the sun's interior rotates
Peimbert, and Kinman (Astrophys. much more rapidly than does the sur-
/., HO, 119) that the He/H ratio is face,
normal in a planetary nebula in the
cluster. NGC 6171
"Blue Stragglers" Th,e R,R ,L7 rfe ™riablra in the
_ , , , . , , , metal-rich globular cluster NGG 6171
Sargent has continued worn, begun are being studied by Dickens in the
at the Lick Observatory on stars in UBV col usi direct Mes ot>
M 67, which Deutsch described as tained at the Newtonian focus of the
blue stragglers in Year Book 65 100_inch teieScope. A total of 60 V,
(pp. 148-150). Photoelectric meas- 56 £, and 22 ?7 plates were obtained
urements were made of the continuous in 1966 and further lateg are being
energy distribution for ten of these taken in 1%7> Measurement of the
stars These were combined with Hy lateg hag -ugt commenced using the
profiles measured on Lick spectra in Sartorius iris_diaphragm photometer
order to obtain gravities and tern- that hag recent] been modified to
peratures and hence masses for these ide digitized output of the X and
stars Although there are difficulties y coordinates and the iris-diaphragm
in fitting the observed energy distn- readi on punched cards. Computer
butions to models, it appears that the programs have been written to proc-
stars have masses near 1 mo, and egs the punched-card output,
hence that the blue stragglers in this The reddening in the field of NGC
cluster should be described as hori- 1(m hag been determined from photo-
zontai-branch stars. electric UBV observations of nearby
Deutsch has continued to observe Md gtars> A value of E = Q 29
the so-called blue stragglers— the wag obtained) which confirms the
anomalous A stars that occur in sev- previous value f ound by Sandage but
eral old open clusters. In developing which ig somewhat less than the
the hypothesis that these objects are value of E = Q 37 ma derived by
the metamorphs of solar-type stars sturch baged on UBV observations of
that have evolved through the giant four variables near minimum light.
stage of evolution, he now proposes
to explain their positions relative to ^,7 ^7 , 7 ^7 ,
the main sequence by taking them to 0ther Globular Clusters
be helium - core horizontal - branch Clusters being studied by Dickens
stars of the kind for which Dr. John are NGC 5466, NGC 6402, NGC 6981,
Faulkner has given models appropri- and NGC 7099. A number of direct
ate to Population II objects. He has plates, together with preliminary
also shown that if there now resides photoelectric sequences, have been ob-
in the solar interior as much angular tained with the 100-inch and 60-inch
momentum as was contained in the telescopes. Most attention has been
protomorphs of the blue stragglers given to the metal-rich cluster NGC
when they were solar-type dwarfs, 6981, which contains about 40 RR
then the radiation entering the base Lyrae variables. The observations of
of the solar hydrogen convection this cluster will be combined with
zone will supply angular momentum those of NGC 6171 in order to provide
at a rate that is sufficient to maintain more statistical weight to the analy-
276
CARNEGIE INSTITUTION
ses of the relatively few variables
that occur in that cluster.
Praesepe Cluster
Rotational velocities have been de-
termined by McGee, Khogali, Baum,
and Kraft using- an electronographic
image tube as well as conventional
spectroscopy for virtually all the
stars in Praesepe brighter than V —
9.6 mag. Xew UBV photoelectric
observations have been made by
Dickens and Krzeminski for all the
stars for which rotations are avail-
able. The photometry is combined
with that obtained much earlier by
Johnson. The brighter stars in Prae-
sepe appear to rotate on the average
about 15rc faster than those in the
Hyades at the same absolute magni-
tude.
The effects of rotation on the tem-
perature and surface gravity have
been investigated in the U — B, B —
V diagram. A weak correlation of
ultraviolet excess with departure
from mean rotation (at a given color)
is obtained, although it is rather less
pronounced than that found earlier
in the Hyades by Kraft and M. Wru-
bel. The effect of rotation on absolute
magnitude has been investigated in
the color-magnitude diagram in a
manner similar to that followed pre-
viously by Strittmatter; he analyzed
the same cluster but used less accu-
rate rotations by Treanor and in-
cluded fewer stars in the analysis. A
much weaker correlation of magni-
tude excess with rotation excess (from
the mean at a given color) is found
by means of the new data. This result
is more nearly in accord with the
smaller effects predicted by models in
uniform rotation than those in non-
uniform rotation.
Cluster Stars with Anomalous
Spectra
In his classification of the Upper
Scorpius stars, Garrison noted a num-
ber of stars which have the colors of
B3-B5 stars but the spectra of B7-
B9 giants, with some peculiarities.
He is now measuring 20 A/mm plates
of these stars taken by Kraft and
Deutsch. In a survey of the Orion
belt region several more stars have
been found with these characteristics.
Since one of them is only a few min-
utes of arc away from o- Orionis E, a
"helium-rich" star, it is interesting to
examine their fit on an H-R diagram.
For this reason, spectra have been
obtained and classified for most of
the stars in the cluster surrounding
a Ori.
INTERSTELLAR GAS AND GASEOUS NEBULAE
Interstellar Absorption Lines
Yaughan and Munch continued
their study of interstellar sodium and
calcium lines with interferometric
equipment attached to the coude
spectrograph of the 100-inch tele-
scope. The newly completed coude
scanner of the 200-inch telescope,
equipped with a Fabry-Perot inter-
ferometer, is also being employed.
Both installations give a velocity res-
olution of the order of 1 km/sec,
comparable with the resolution now
reached on radio-absorption observa-
tions at 21 cm.
Present work confirms that the
line-of-sight velocities of distinct in-
terstellar sodium (Na I) and calcium
(Ca II) components are identical to
within a measuring accuracy of ±0.5
km/sec (p.e.). The relative strengths
of many sodium and calcium com-
ponents differ considerably from one
interstellar "cloud" to another, prob-
ably implying local excitation dif-
ferences. Not all such differences are
MOUNT WILSON AND PALOMAR OBSERVATORIES
277
consistent with an earlier correlation
by Routly and Spitzer (Astrophys.
J., 115, 227, 1952) between Na I/Ca
II and the peculiar velocities of the
clouds. A detailed study of this prob-
lem is in progress.
In the Orion region, the Na I spec-
tra obtained by Vaughan and Munch
contain some 35 real and reasonably
distinct absorption features in the
directions of five bright stars within
11° of NGC 1976 (the stars i, f, 9\
e, o) Orionis) . A similar distribution
of Ca II features is found. In the best
spectra the weakest features whose
reality is certain show equivalent
widths of 1.5 ml, although a few
slightly weaker features seem to be
detected. Component velocities range
from —5 to +40 km/sec heliocentric,
or from —22 to +23 km/sec in the
local standard of rest (the velocity
associated with differential galactic
rotation is about +7.5 km/sec at
500 pc). Intercomparison of velocities
within the line complexes in dif-
ferent directions indicates that of
the 35 features two or three are
surely common to four directions at
least, and two are common to three
directions. It appears that the region
is crossed by at least seven individ-
ually coherent "clouds," "wisps,"
or "sheets" of interstellar gas, some
of which subtend angles of 11° or
more. Since a feature associated with
NGC 1976 (the Orion nebula) is one
of these, its transverse size can be
taken to exceed 100 pc.
Nova Envelopes
With the purpose of studying the
interaction of matter ejected in nova
explosions with surrounding inter-
stellar material, a number of old
novae are being photographed anew
with the 200-inch by Munch and com-
pared with earlier images obtained by
Baade and others.
The envelope of Nova Herculis 1934
has been found to be definitely decel-
erating. Between 1956 and 1966 the
expansion rate of the semi major
axis was 0.19 arc sec/yr, compared
with the mean rates 0.30, 0.25, 0.23,
and 0.22 arc-sec/yr derived between
the time of explosion and 1940, 1942,
1956, and 1966, respectively. For the
semi major axis, the rate between
1956 and 1966 was 0.10 arc-sec/yr,
while the mean rates from zero age
are 0.22 arc-sec/yr for 1940; 0.19 for
1942; 0.17 for 1956; and 0.15 for
1966. The most remarkable change
observed in 1966 in the envelope of
Nova Her 1934 is that the bar run-
ning across the central star in the
direction of the minor axis, which
in 1966 appeared brighter than the
elliptical ring, was relatively much
fainter.
The envelope of Nova Persei 1901
also shows indications of deceleration.
Between 1949 and 1964, the expan-
sion rate of the fastest filaments was
0.35 arc-sec/yr, while the mean rate
between initial explosion and those
dates are 0.55 and 0.50 arc-sec/yr,
respectively. Not all the filaments in
this complex envelope, however, show
the same behavior, and indeed a few
of them appear to be accelerating.
Crab Nebula
For a doctoral thesis, Virginia
Trimble has undertaken the measure-
ment of a series of Ha -[N II] plates
of the Crab Nebula taken with the
100-inch and 200-inch telescopes by
Baade and Munch, with the purpose
of determining the proper motions of
the filaments. On the whole, the fila-
mentary mass is in a state of expan-
sion, with an apparent center that is
8" east and 15" south of the suspected
central star. The deviation is several
times the probable error of the deter-
mination of the convergence point.
The date of convergence, assuming
uniform velocities, is about 90 years
after zero age (1054 A.D.), indicat-
ing some acceleration, but somewhat
278 CARNEGIE INSTITUTION
smaller than found earlier by Duncan. The flux density from the nebula in
On the basis of this determination the two wavelength bands 1.5-1.8 /x
and published radial velocities, the and 2.0-2.4 fx has been measured with
distance to the nebula is 1.1 or 1.7 an accuracy of 10%. The results,
kpc, depending on whether it is a when combined with O'Dell's (1963)
prolate or an oblate spheroid. Spec- near-infrared photometry, show that
trograms of various portions of the the origin of the infrared radiation
nebula taken by Munch have been re- is nonthermal with a spectral index
measured for radial velocities, which, of about —0.8.
combined with the proper motions,
will provide a three-dimensional Extinction in Emission Nebulae
picture of the structure and kinemat- Spectrophotometric data for 25
ics oi the nebula, as well as a unique early-type stars imbedded in 14 emis-
determination of its size and distance, sion nebulae have been obtained by
J. Scargle has completed a photo- Christopher M. Anderson utilizing the
electric and photographic study of the photoelectric scanner at the 60-inch
continuum radiation of the Crab and 100-inch telescopes. The apparent
Nebula. The plates taken with the spectral energy distributions have
100-inch and 60-inch Mount Wilson been compared with the predictions
reflectors and the 48-inch schmidt at of model-atmosphere calculations in
Palomar, are being used in connec- order to obtain the wavelength de-
tion with 200-inch direct plates by pendence of the interstellar extinc-
Munch, Baade, and Arp to study tion. It has been found that the stars
rapid motions in the ''amorphous" HD 164492 in NGC 6514 and Her-
mans. The spectral energy distribu- schel 36 in NGC 6523 show the same
tion was measured through inter- reddening anomaly as does the star
ference filters at the Newtonian focus 61 Orionis. A theoretical study of the
of the 100-inch. In agreement with anomaly is under way by Anderson,
earlier work by Oke, the characteris- since the previously accepted expla-
tics of the radiation in the ultraviolet nation, based on the removal of small
appear to indicate departures from a particles by radiation pressure, can-
single power law in the relativistic not operate because of viscous drag,
electron-energy spectrum, or the ex- as Krishna Swamy and C. R. O'Dell
istence of bremsstrahlung from a hot have recently shown,
gas also producing the observed X
rays. A second result is that the slope High-Latitude Planetary Nebula
of the continuum is the same all over A blue star imbedded in very faint
the nebula — a fact that places restric- nebulosity has been discovered by Arp
tions on possible models for injection and investigated by Arp and Scargle.
of fresh relativistic electrons by the Spectra of the nebulosity and central
stellar remnant of the supernova, star show the object to be an old
Scargle has also studied the physics planetary nebula, but to have a higher
of a relativistic gas in a magnetic galactic latitude (&11 — —47°) than
field in order to understand the prop- any previously listed by Abel (Astro-
agation of disturbances in a gas such phys. J., lUUy 259, 1966). The central
as the high-energy component of the star is identified as No. 932 in the
Crab Nebula. Haro-Luyten Palomar Catalogue
Eric Becklin, a graduate student in (Bol. Obs. Tonantzintla y Tacubaya,
physics, has made infrared observa- 3, 37, 1962). A distance of 150 pc has
tions of the Crab Nebula using the been computed for the star, with the
60-inch telescope at Mount Wilson, use of an accurate proper motion
MOUNT WILSON AND PALOMAR OBSERVATORIES 279
furnished by Luyten together with parent diameter is about 4". A spec-
the measured radial velocity. Inter- trum at 400 A/mm on 103a-D emul-
ference-filter photographs of the neb- sion shows a typical planetary nebula
ulosity in the light of [0 II] and Ha emission-line pattern and a faint cen-
show an unexpected comet shape for tral-star continuum. Since the appar-
the outer material and elongation ent brightness of the object is only
along a different axis for the inner V — 15 mag, B = 16 mag, it is ap-
material. parently at a great distance above the
An object near the north galactic galactic plane. Further analysis is in
pole, at R.A. — 12h57^5 and Dec progress.
= +27° 50', was noticed by Haro Several 200-inch direct photographs
(Publ. Astron. Soc. Pacific, 63, 144, and spectra of Sharpless 71, a galac-
1951) to have emission features tic nebulosity, and its central star
characteristic of a planetary nebula, were obtained by Arp. Joseph Miller
Although the object is stellar in ap- of the University of Wisconsin is
pearance on Sky Survey plates, a using these observations in his study
photograph in good seeing by Arp of the object,
with the 200-inch shows that its ap-
INFRARED SKY SURVEY
The survey of the sky for objects data for the first coverage and for
emitting in the spectral ranges 0.68- about five sixths of the second cover-
0.92 \l and 2.01-2.41 p, sponsored by age have gone through preliminary
the National Aeronautics and Space processing. Approximately 6000
Administration, has been continued sources have been found brighter than
at Mount Wilson by Leighton and K = 3, of which about 450 have / —
Neugebauer and several graduate stu- K > 5. The sources show a marked
dents. A description of the survey concentration along the galactic
instrumentation was given in Year plane ; the mean I — K of sources ob-
Book 6Jf> (pp. 25-26) . served in the galactic plane is larger
Although only about one third of by about one magnitude than that
the nights during the last year were for sources out of the plane. A statis-
used, a second coverage of the sky tical study of the data is now in prog-
has been essentially completed. The ress.
GALACTIC STRUCTURE
In August 1966, E. Becklin ob- 110 arc-sec resolution. There is one
served radiation from the vicinity of main source of radiation which is
the center of the Galaxy in the wave- about 3-4 arc-min (10 pc) in diameter
length region 2-2.5 p. Further obser- and which is extended along the ga-
vations were subsequently made at lactic plane. Within observational
0.9, 1.65, 2.2, and 3.4 /x by Becklin error, this main source agrees both in
and Neugebauer using the 24-inch, position and size with the radio source
60-inch, and 200-inch telescopes. Sagittarius A. Scans of high spatial
The general infrared radiating resolution show that the source is non-
structure of the central region of the gaussian in shape. Instead, its radia-
Galaxy is shown on the contour map tion is highly concentrated toward
in Fig. 2. This represents data ob- the center, where there appears to be
tained with the 24-inch telescope with a core of radiation 30 arc-sec in diam-
280
CARNEGIE INSTITUTION
■28' 40' -
-::--:;
c
in
r
c
-I -29° 00 -
O
UJ
-29° IC' -
RIGHT ASCENSION (1950)
Fig. 2. A contour map of the galactic center region at 2.2 fx taken with an aperture of 1.8 arc-
min diameter is presented. Contour lines are separated by 8.5 X 1 0~20 W m~2 Hz-1 ster-1 (5.2
X 10~10 W cm-2 jut1 ster-1) and are uncertain by about the same amount. The X represents a
pointlike source.
eter (1.5 pc linear size). There is
also a pointlike source just slightly
offset from the position of maximum
brightness. Besides this main source,
Fig. 2 shows that there is general
background radiation lying predom-
inantly along the galactic plane and
that there are also several weaker
extended sources of radiation. None
of the weaker infrared sources is
coincident with a known radio source.
A comparison of the infrared
structure of the galactic center has
also been made with the infrared
structure of the nucleus of M 31 as
determined from observations with
the 200-inch telescope. When a resolu-
tion of 14 pc is used for comparison,
the two galactic centers look similar
in both shape and brightness.
GALAXIES
Distance Modules of NGC 2/+03
As a first step for the redetermina-
tion of the Hubble constant, Tam-
mann and Sandage have completed
measurements and analysis of the
variable .stars in NGC 2403. This late
Sc galaxy closely resembles M 33 in
stellar content and in the character
ing member of the M 81 group and is
the first galaxy beyond the Local
Group in which cepheid variables
were found soon after the 200-inch
telescope went into operation in 1949.
Hubble began the work in 1950.
Plates continued to be taken by San-
dage until 1963 and a photoelectric
of the spiral pattern. It is an outly- sequence was determined in four ob-
MOUNT WILSON AND PALOMAR OBSERVATORIES 281
serving seasons from 1962 to 1965 types in M 31 and M 33; and (5) the
ranging from B = 10 to B = 22.8. angular size of the largest and the
Altogether, 160 blue plates and 22 average of the five largest H II re-
yellow plates are available covering gions, calibrated from linear sizes
the time interval from 1910 to 1963. determined in M 33 and the Magel-
By blinking the complete material, 56 lanic Clouds.
variable stars have been located, of The five methods agree remarkably
which 17 are cepheids with periods well (±0.2 mag), and although the
from 20 to 87 days, 8 are very bright final analysis is not quite complete,
blue variables with irregular light an interim value of (m — M)0 =
curves, 14 are supergiant red varia- 27.56 ± 0.2 is suggested as of June
bles of the SRc or Ic class, 1 is an 1967. Even after final reduction is
eclipsing binary, and 16 are faint un- complete, an improvement of this
classified variables among which are value can eventually be made from a
cepheids of undetermined period. recalibration of photometric data in
The 17 cepheids are all extremely the LMC and SMC for the bright red
faint. None becomes brighter than variables that appear near F = 11.
B = 21.2, and most are fainter than A start on this program is expected
B = 22.0 at maximum light. Light next year during a visit by Sandage
curves have been determined for all to Australia.
17 variables in both B and V wave- The distance to NGC 2403 is im-
lengths. All but one of the cepheids portant because it permits the cali-
disappear below the plate limit (B ~ bration study of linear sizes of H II
23.5) at minimum, and all become regions to be expanded to the M 81
fainter than the limit of the magni- and M 101 groups. With distances to
tude sequence (B = 23.0) at some these groups available, the number of
phase of their light curve. calibration galaxies of Sc and Irr
The distance modulus of NGC 2403 types is enlarged to nearly 20 sys-
has been determined by five methods, terns, and the mean value of the linear
which depend on (1) the cepheids at diameters, D (largest) and D (5),
maximum light in both B and V, to- w[\\ have high weight. To obtain the
gether with a new composite period- Hubble constant the angular diame-
luminosity relation at maximum, ob- ters 0f h II regions and the magni-
tained by combining recent data for tudes of the bright red stars can then
cepheids in the LMC, SMC, M 31, and ^e used m the range 31 > (m - M)
NGC 6822, calibrated by nine ceph- > 27 for determining distances to
eids of known MB (max) and Mv Sc and Irr galaxies that have meas-
(max) from the galactic system; (2) ure(j redshifts. Kristian has begun
the apparent magnitude of the red the measurement of angular sizes of
supergiant variables, assuming Mv the H II regions of such galaxies for
(max) = -8.10 for the first-ranked which 200-inch H« interference-filter
member of the class as derived from piates are available, and a first value
similar stars in NGC 6822 and the of H0 is expected next year.
Magellanic Clouds; (3) the apparent
magnitude of the brightest resolved Variables in Divarf Galaxies
stars as obtained by counts across the Miss Swope has now found periods
face of the galaxy, and calibrated as- photographically for 76 RR Lyrae
suming MB = —9.5 as determined variables in the Leo II system. Their
from other members of the Local mean period of 0.589 day is shorter
Group; (4) the bright irregular blue than the 0.611 day found for the
variables calibrated from their proto- Draco-system variables and the am-
2S2 CARNEGIE INSTITUTION
plitudes appear to be smaller than pletely dominated by a nonthermal
those found in Draco. The magnitudes component that, conjecturally, arises
are not yet based on a photoelectric from the center of the nucleus. Oke
sequence. However, the two variables and Sargent's model is similar in its
with period longer than 1 day corre- essentials to one that I. S. Shklovsky
spond in absolute magnitude to those has derived from considerations of
found in Draco and Sculptor and are the radio-frequency spectra of Sey-
brighter than the variables of like fert galaxies.
period found in galactic globular A spectroscopic comparison has
clusters. The Leo II system also has been made by Arp of the compact
six irregular variables that lie on the galaxy III Zw 2 [R.A. = 0h8^0, Dec
bright end of the giant branch in the = + 10°42' (1950)] and the radio
color-magnitude diagram. There are galaxy 3C 120, also designated Parkes
also six with a blue color index that 0439 + 05. Both have typical Seyfert
vary slowly and that are bluer and spectra. The radio source in 3C 120
slightly brighter than RR Lyrae var- has a small angular diameter and a
iables. Xo such stars were found in flat radio spectrum.
Draco. The nuclei of ten Seyfert galaxies
have been studied further by Kurt S.
Seyfert Galaxies Anderson as a thesis project. Photo-
Oke and Sargent have completed electric-scanner observations of the
their work on the nucleus of the Sey- continua and emission features have
fert galaxy NGC 4151. Their model been obtained in the region AX3400-
of the nucleus, deduced from photo- 5800, and are currently being ex-
electric scans and photographic spec- tended to approximately A8400. Spec-
trograms, consists of an object 50 pc trograms are also being obtained to
in diameter, filled with 105 ttlo of assist in the determination of line
gas at 106OK and with an elec- strengths and profiles to permit the
tron density Ne = 100 cm-3. This gas estimation of electron densities and
gives rise to the coronal-line spectrum temperatures. Theoretical calculations
first observed by O. C. Wilson. In this have been undertaken in order to as-
are embedded cooler clouds or fila- certain the nature of the processes
ments of total mass 3 X 104 tit© responsible for the observed optical
that have random motions of about and radio continuous spectra.
1000 km /sec, giving rise to the main
forbidden-line spectrum of the object. Compact Galaxies
The cool component fills only 0.004 of Sargent began observations with
the volume of the nucleus and is sup- the 200-inch prime-focus scanner of
posed to be in pressure equilibrium the continuous energy distributions
with the hot component, having Te = and emission-line strengths in eleven
20,000 °K and Ne = 5000 cm-3. It is compact galaxies discovered by
conjectured that the broad wings in Zwicky (four lists privately circu-
the H lines and He II A4686 are pro- lated) and by Arp (Astrophys. J.,
duced in a dense gas having Ne > 14,2, 402). The resolution was 50 A
10 cm-3 and Te <~> 20,000°K. This and the wavelength range covered
last component contains only a few was AA3300-6000. Four of the galax-
tens of solar masses of gas and oc- ies studied (I Zw 0246 — 03, I Zw
cupies a volume, perhaps at the center 0955 + 51, II Zw 0431 — 02, and III Zw
of the nucleus, only about 0.01 pc in 0824 + 21) have energy distributions
diameter. It is concluded that the op- that span the small range exhibited
tical continuum of NGC 4151 is com- by normal dwarf and giant ellipticals.
MOUNT WILSON AND PALOMAR OBSERVATORIES 283
Four of the remaining objects (I Zw nucleus of NGC 5934) has a continu-
0930 + 55, I Zw 1439 + 53, II Zw ous spectrum intermediate between
0430 + 05, and II Zw 1622+41) have the compacts with and without emis-
spectra with broad emission and sion lines. Greenstein has found that
much flatter continua than the non- the spectrum of this object is dom-
emission-line objects. There is an in- mated by A-F stars in the blue and
dication that there are two compo- by a cooler population near A.5000.
nents in the continuous spectra of the The two Arp compacts have fairly flat
emission-line compacts ; namely, a spectra with emission lines similar to
flat component that dominates the those of the Zwicky emission-line
blue end of the spectrum and a com- compacts. Sargent intends to observe
ponent rising toward longer wave- many of these objects in the near in-
lengths (although less steeply than frared during the next season. He is
the energy distributions of normal also analyzing the emission-line data
ellipticals) that dominates the red in an attempt to understand the phys-
end of the spectrum. It is not likely ical conditions in the compact gal-
that either of these components is axies.
primarily due to recombination radi- A large family of compact galaxies
ation from a hot gas because Oke has with a great variety of spectra con-
found that I Zw 1727 + 50 has a two- taining very blue, normal visual, and
component continuous spectrum simi- infrared objects, some with weak
lar to that of the emission-line com- radio emission, can now be regarded
pacts; however, this object shows no as firmly established. Zwicky has corn-
emission lines either on the scans or piled a list containing 446 such ob-
in spectrograms obtained by Zwicky jects. From observations of differen-
(Astrophys. /., 1^3, 192). tial redshifts, the percentage of sky
At the telescope the compact gal- coverage by clusters of galaxies, and
axies in Zwicky's lists appear to be the theory of characteristic limiting
sharply bounded and 5-15 seconds of bodies, Zwicky inclines to the view
arc in diameter. So far, any pro- that the quasi-stellar radio sources
nounced central condensation (which belong to the family of compact gal-
is always in the form of a star like axies and that Einstein shifts are
nucleus) is shown only by the emis- responsible for a large part of the
sion-line objects. observed redshifts.
Spectrograms of three of the emis- Compact galaxies remarkably often
sion-line compacts have been de- occur in small groups, some members
scribed by Zwicky (op. cit.). The of which may be interconnected by
additional galaxy II Zw 0430 + 05 was luminous intergalactic formations,
observed spectroscopically by B. Among the latter, the widely sepa-
Peterson and by Sargent with the rated but interconnected compact gal-
Newtonian spectrograph of the 100- axies at R.A. = 19h26m, Dec = +61°
inch telescope. It has broad emission 50' (1950) may be mentioned,
lines of H and [O III] and appears Only about six clusters of compact
to be a Seyfert galaxy. The redshift galaxies among 8000 clusters have
is z = 0.0323 and the absolute mag- been found so far. Examples are the
nitude of the nucleus is Mv = —20.8. medium-distant cluster 0658 + 6321
On this basis, II Zw 0430 + 05 is in- with a population of 189 semicompact
trinsically the most luminous known galaxies of very uniform types and
Seyfert nucleus. It is also a radio the distant cluster 0152.0 + 3337 with
source. a population of 144 medium-compact
The object I Zw 1356 + 37 (the to compact galaxies. Observations of
284 CARNEGIE INSTITUTION
differential redshifts among: these and show that it is an almost pure
may. in Zwicky 's opinion, make it emission feature, observable in the
possible to decide if some fraction of light of both [0 II] and H«. Forbid-
these shifts is due to the Einstein den emission lines of [0 II], [N II],
gravitational effect, and thus throw and [S II] have been found in the
new light on the distances of the nucleus of M 87. Other lines are seen
quasi-stellar objects. which seem to be asymmetric with
In 1965 Arp reported the discovery respect to the nucleus. The possible
of a small double galaxy not much connection of the latter lines with the
larger than a globular cluster (Astro- jet and counter-jet will be studied in
phys. /., 142, 402, 1965). Since then the coming season and a detailed
he has obtained a spectrogram at 190 analysis of the redshifts of all these
A mm in very good seeing on which lines will be attempted. All the ellip-
it is possible to measure the differen- tical (E) galaxies in this region of
tial velocity of the emission lines of the Virgo cluster are distributed in
the two components. O'Connell and a narrow band on either side of M 87.
Arp now report a derived mass of Arp considers it significant that the
10s m,©, on the assumption that the axis of this distribution of companion
two components are mutually bound galaxies coincides with the position
in orbit about each other. The mass- angle of the jet and counter-jet.
to-luminosity ratio agrees well with
ratios previously found by Zwicky for Ultrafamt Blue Stars
blue compact galaxies. The mass den- In continuation of earlier investi-
sity in the present object is about one gations, a survey of the location and
hundred times that in the average the magnitudes of individual blue
galaxy. Triple-image studies by Arp stars on 200-inch blue and yellow
and O'Connell with the 48-inch plates has been undertaken by
schmidt demonstrate that this object Zwicky in collaboration with Luyten.
has a large ultraviolet excess. Spec- Fifteen pairs of 5" X 7" plates (with
trum scans by Sargent confirm the a scale of 11"2 arc/mm) were used.
ultraviolet excess. Arp and O'Connell The blue exposures, generally of 27
have analyzed other compact galaxies minutes, were on unfiltered 103a-O
with ultraviolet excesses and have plates; the yellow exposures, gener-
pointed out some of their connections ally of 45 minutes, were made on a
with Haro blue galaxies, compact 103a-D plate through a GGII filter.
galaxies, and quasi-stellar galaxies. The observations, unfortunately, were
mostly made under inferior conditions
Counter-Jet in M 87 0f seeing. The plates were blinked by
It has long been known that the Luyten, who also determined the star
bright elliptical galaxy in the Virgo positions — usually relative to a single
cluster, M 87, has a jet with a con- AGK-2 or BD star on the plate. Luy-
tinuous spectrum extending about 22" ten is also responsible for the magni-
from the nucleus of the galaxy. Arp tude and color estimates. The former
has now discovered luminous streaks were derived by comparing the blue
extending in the opposite direction, plates with the blue Sky Survey
The markings are brightest between plates taken with the 48-inch tele-
30" and 40" from the nucleus, but scope, on the assumption that their
extend out to more than 120" from limiting magnitude is 21.2. Colors
the center of the galaxy. Spectro- were estimated relative to some six
grams and interference-filter direct comparison stars of about the same
photographs register the counter-jet magnitude, and subsequently cor-
MOUNT WILSON AND PALOMAR OBSERVATORIES
285
rected for the expected average color
of these stars. It is hoped that the
accidental errors in the magnitudes
and colors do not exceed 0r"5 and
0^25, respectively
Data for 217 ultrafaint blue ob-
jects will be published as a separate
list. These objects are generally
fainter than mp — 19.0 and are stellar
in appearance. The number of stars
is as yet too small and the material
not sufficiently homogeneous to allow
one to draw significant conclusions
concerning the rate of increase with
fainter magnitudes. The intention is
to determine proper motions for ob-
jects of 17 mag and brighter from
the 48-inch Survey plates, and then
to combine these measurements with
more accurate photoelectric colors in
order to derive the relative frequency
of white dwarfs, subdwarfs, horizon-
tal-branch stars, and compact galax-
ies or even quasars among these
objects.
Catalogue of Galaxies and of
Clusters of Galaxies
Volumes I, II, and III, which cover
the unobscured areas from Dec —3°
to +57°, and Vol. V, Dec -3° to
+ 21°, on the southern galactic cap,
have now been published. These four
volumes contain, respectively, data on
9500, 6700, 5300, and 4200 galaxies
as well as on 1300, 2350, 2700, and
1200 clusters of galaxies.
About half of the work necessary
on the clusters of galaxies and on the
magnitudes and positions of the gal-
axies brighter than rnp = 15.7 to be
included in Vol. IV, covering the un-
obscured areas north of Dec +57°,
has been completed by Zwicky and E.
Herzog.
Work has been completed on the
clusters of galaxies that are to be in-
cluded in Vol. VI, which covers the
unobscured areas between Dec +21°
and +51° of the southern galactic
cap. Positions and magnitudes for
about two thirds of the galaxies to
be included in this volume have been
determined by C. Kowal.
This work has been supported in
part by a grant from the National
Science Foundation.
Statistical Analysis
The distribution over the least ob-
scured parts of the sky of 2350 clus-
ters of galaxies listed in Vol. II of the
Catalogue has been analyzed by M.
Karpowicz and Zwicky by the method
of dispersion-subdivision curves. It
was found, as from the previous anal-
ysis of data in Vol. I, that this dis-
tribution is random except for some
slight deviations that are possibly
ascribable to the absorption of light
by intergalactic dust clouds in the
clusters in Virgo, Coma, Hercules,
and Leo A and B.
The distribution of the 1200 clus-
ters of galaxies listed in Vol. V of the
Catalog has also been analyzed by
Karpowicz and Zwicky by means of
the dispersion-subdivision method.
They report that this distribution is
the most random found so far in any
region of the sky and that, as for the
distribution of the clusters in Vols. I
and II, no clustering of clusters of
galaxies is found.
The analysis by Zwicky of the dis-
tribution of clusters of galaxies over
the northern galactic cap, in right
ascensions from Milky Way to Milky
Way and declinations —3° to +57°,
shows the remarkable feature that the
total numbers of clusters of galaxies
in the half areas east and west of
R.A. 13h0m are, respectively, in the
ratio of about 2 to 3, while the nearer
clusters are more evenly distributed.
The relatively high depletion of the
numbers of medium distant (MD),
distant (D), very distant (VD), and
exceedingly distant (ED) clusters in
the eastern half of the area mentioned
is ascribed to the greater depth of
interstellar dust on that side of the
2S6
CARNEGIE INSTITUTION
Milky Way, and, conjecturally, to
intergalactic dust in the Virgo, Coma,
Hercules, and other near clusters,
and in the Ursa Major and the Shane
Clouds. A total relative excess of
about 0.2-0.3 mag in dimming over
the eastern half of the area could
account for the difference in the total
counts of clusters. The total area of
the sky covered by the clusters in the
two halves, significantly, is much less
affected. The distribution over the
part of the southern galactic cap an-
alyzed in Vol. V does not show any
analo go u s a symmet r y .
Using the contours as they are
plotted in the Catalogue, Zwicky es-
timates that rich clusters with sym-
bolic velocities of recession in the
range 5000 < Vs < 100,000 km/sec
in projection cover three fifths of the
whole celestial sphere. He calculates
that, to redshifts of the order of z =
2.0, the sky is therefore covered
three times by the central parts of
clusters, even if we assume that the
number density of clusters is uni-
form.
Classification of Clusters of Galaxies
In the first four volumes of the
Catalogue of Galaxies and of Clusters
of Galaxies, 7505 rich clusters are
listed. They have been classified as
compact, medium compact, and open.
A first survey concerning the types
of galaxies to be found in clusters
has now also been made. Zwicky
points out that clusters are of differ-
ent morphological types and that total
populations of galaxies in clusters
vary widely. The material content
of the clusters varies inasmuch as
the following preponderances of
galaxies among their brightest mem-
bers have been found: (1) Clusters
containing only irregular galaxies.
These are very rare. (2) Clusters con-
taining a great variety of spirals and
irregular systems. (3) Clusters of all
types of galaxies, except that few
very compact systems are in evidence.
(4) Clusters like type 3 that also con-
tain many compact objects. (5) Clus-
ters composed mainly of elliptical
galaxies but with few compact ob-
jects. (6) Clusters like Type 5 that
also contain many compact galaxies
as well as ellipticals with giant com-
pact cores. (7) Clusters of compact
galaxies. Although numerous groups
of compact galaxies have been found,
rich clusters consisting of only com-
pact galaxies are rare, at least at dis-
tances at which the symbolic velocity
or recession is less than 50,000 km/
sec.
The Local Group, although it is not
a rich cluster, would be described as
of type 3.
From the data given in the Cata-
logue, the volume r, within which, on
the average, one rich cluster of gal-
axies may be expected, has been re-
calculated assuming a Hubble con-
stant of 100 km/sec per megapc. It
was found that t is equivalent to a
cube with an edge of 40 megapc. This
closely parallels the value derived by
Zwicky in 1938 from the study of only
about one hundred of the nearest clus-
ters of galaxies, which had been pho-
tographed with the 18-inch schmidt
telescope.
Special Cluster
As a curiosity, Zwicky mentions
the exceedingly distant cluster CI
1101 + 7042, which shows a ringlike
arrangement of 39 galaxies on an
ellipse about 5' in diameter, and
which might possibly be influenced by
a gravitational lens effect.
Supernovae
On the search program directed by
Zwicky, 15 supernovae were discov-
ered at Palomar: 7 by Kowal, 7 by
Rudnicki, and 1 by Reaves.
Extended statistical studies on the
frequency and distribution of super-
MOUNT WILSON AND PALOMAR OBSERVATORIES
287
novae were started by Zwicky and
A. P. Fairall (working at the Mc-
Donald Observatory of the University
of Texas).
Light curves have been constructed
and spectral characteristics have been
studied of the supernovae SN 1962e
(by Zwicky and Rudnicki) and SN
1966b (by Zwicky, Gates, Bertola,
and Ciatti). The light curve of SN
1966e in NGC 4189 is being deter-
mined by Kowal.
Zwicky, as chairman of Committee
28 of the International Astronomical
Union, has continued to coordinate
the search for supernovae at 14 ob-
servatories in both hemispheres and
has, in particular, proposed the con-
sideration of new instrumentation
and procedures to accelerate the rate
of discovery and subsequent detailed
observations and evaluations. Special
attention is being given to procedures
of searching for supernovae within
the bright cores and disks of ellipti-
cal So, Sa, and Sb galaxies, many of
which have presumably been missed
by the methods used in the past.
RADIO GALAXIES
Cosmic-Ray Sources
The origin of cosmic rays is usually
associated with supernovae in our
Galaxy. The alternate hypothesis that
cosmic rays are universal has been
based on estimates of the contribu-
tion made by radio galaxies (Bur-
bidge and Hoyle) . The long harmonic-
mean life (109 years) for radio
galaxies derived by Schmidt (see
Year Book 65) reduces their contri-
bution to the cosmic rays by a factor
of 103. As a consequence, most prob-
ably only a small part of the locally
observed cosmic rays can be attrib-
uted to radio galaxies.
Redshifts of Radio Galaxies
As time permitted, Sandage con-
tinued the program of redshift deter-
mination of radio galaxies identified
from the 3CR catalog. The project
is coordinated with parallel work by
Schmidt with the aim of eventually
determining redshifts of all identified
radio galaxies from the homogeneous
Third Cambridge Catalogue, The re-
sults will eventually be combined with
existing photoelectric magnitudes in
the BVR system which Sandage has
been assembling for several years.
During the report year, redshifts
of 9 new radio galaxies, corrected for
galactic rotation, were obtained as
follows: 3C 29 (AA/A0 = z =
0.0450), 3C 76.1 (0.0328), 3C 135
(0.1270), 3C 227 (0.0855), 3C 236
(0.0988), 3C 285 (0.0797), 3C 381
(0.1614), 3C 449 (0.0181), and 3C
455 (0.0331).
Energy Distribution
Oke has begun a study of absolute
energy distributions in the continua
of radio galaxies and related objects.
Radio source 3C 371 has an energy
distribution that is very similar to
that of Seyf ert galaxies and to some
of the Zwicky compact galaxies that
are being studied by Sargent.
Direct Photography
Direct photography of radio
sources with the 200-inch telescope is
being continued by Arp. The fields
include quasi-stellar radio sources
and some that are blank upon visual
inspection, but most of the objects be-
ing investigated are radio galaxies.
The more important radio-galaxy
identifications with 3C radio sources
have been and are being photo-
graphed, and some 4C sources are
also being inspected. John Bolton has
furnished a list of the most important
Parkes sources for which optical in-
formation is needed. One preliminary
288 CARNEGIE INSTITUTION
result is that a number of radio gal- giant characteristics) turn out, under
axies that have been classified "D" on high resolution, to be various kinds of
small-scale photographs (implying spirals.
X-RAY SOURCES
Scorpius X-I began a series of photoelectric meas-
A cooperative effort was begun urements of the object with the 200-
with the American Science and Engi- inch telescope immediately after
neering - Massachusetts Institute of Tokyo identification results were re-
Technology X-ray astronomy group ceived by telephone from Giacconi.
for the optical identification of X-ray The object was found to vary in opti-
sources. These workers, under the cal brightness by ±0.03 mag in time
leadership of R. Giacconi, performed intervals as short as 1 minute.
a series of spectacular rocket experi- Sandage then obtained spectro-
ments in 1966 and early 1967 from grams on six nights in July at the
which relatively precise celestial co- same time that Patrick Osmer, a stu-
ordinates were obtained for the dent at the California Institute of
brightest point X-ray source Sco X-l, Technology, monitored the object
and for the fainter sources Cyg X-l, photoelectrically with the Palomar
Cyg X-2, and Cyg X-3. 20-inch reflector. Variations of 0.8
The first flight concentrated on mag were observed from night to
Sco X-l. By use of an X-ray modu- night, and on these the very short-
lation-collimator, together with op- period fluctuations were superposed,
tical photographs of the sky taken at Large night-to-night variations in
1-sec time intervals through the col- the spectral-line intensities were also
limator on board the rocket, the ASE- found during this period.
MIT team succeeded in obtaining a The spectra show all hydrogen
set of positions that were accurate to lines in emission down to at least
= 1 minute of arc in both coordinates, H 10. The high excitation He II line
with two equally probable positions at A4686 is present. A complex blend
differing by 5 minutes of arc. of high excitation C III, N III, and
Predictions of the nature of the O II is present from AA4630 to 4655,
optical counterpart of the X-ray but the intensity of this feature
source are based on the hypotheses changes radically from night to night,
that the X rays are produced either as does the absolute intensity of the
by free-free transitions of hot gas hydrogen lines.
(bremsstrahlung), or by synchrotron The interstellar H and K lines of
emission. These predictions had sug- Ca II are seen in absorption with
gested that the optical object should equivalent widths of 0.2 A. Analysis
be very blue with B — V ~ +0.1, by Munch suggests that the path
U — B ~ -0.9, and should appear length of Ca II is NCa n ' L = 5 X 1012
at V ~ 12 mag. atoms cm-2. This, combined with esti-
Communication of the X-ray posi- mates of average ionization equilib-
tion by Giacconi to both the Tokyo rium conditions and abundance ratios
Observatory and to Mount Wilson and of carbon and calcium to hydrogen,
Palomar permitted the identification gives N2H • L — 2 X 10'20 atoms cnr5.
and confirmation in June 1966 of a By combining the N2 • L values with
starlike object of the correct color the measured X-ray absorption due to
and magnitude in one of the error galactic hydrogen at 1 kev, Sandage
boxes of the X-ray positions. Sandage estimates that the distance is about
MOUNT WILSON AND PALOMAR OBSERVATORIES 289
400 psc. The X-ray absorption data each of six nights. Jugaku suspects
were obtained from rocket flights by that radial-velocity changes also oc-
Steward, Mark, Chodel, and Grader cur on a short time scale of about 60
at the University of California. This minutes, but the measurements are
distance, combined with the observed exceedingly difficult and the results
X-ray flux, shows that the source are marginal. Westphal and Kristian
emits about 1037 ergs/sec in X-rays in are continuing the work by remeasur-
the range 1-10 A. ing the plates with digital techniques,
In April 1967, Westphal and San- using a microphotometer connected
dage began an experimental monitor- to the Caltech 7094 computer. Analy-
ing of the optical flux of Sco X-l sis is planned with the use of West-
with time resolutions as small as 2 phaFs cross-correlation method that
seconds. Observations with the 60-, is capable of very high precision for
100-, and 200-inch telescopes, using wavelength determination,
the pulse-counting equipment operat- Neugebauer is making a series of
ing in an automatic mode as designed measurements of X-l Sco at a wave-
and built by Dennison's group, length of 2.2 /*.
showed that the source was highly
variable in time intervals as short as Sources in Cygnus
1 minute. Continuous monitoring of A second rocket flight by the ASE
the source for intervals of 2 to 4 group produced positions for three
hours on five separate nights shows X-ray sources in Cygnus with an ac-
evidence for nonrandom variations curacy of about ±11 minutes of arc
with amplitudes of 0.04 mag in a time in both coordinates. Three-color im-
interval of 4 minutes, superposed on age plates of the Cygnus region were
longer-term variations with periods obtained by Osmer, and several ultra-
of about 30 and 60 minutes. Fourier violet candidate objects were marked
analysis of the records gives partial before the X-ray positions were
evidence that some of the character- available. One of these objects was
istic periods repeat from night to subsequently found to lie within the
night, but the work is still in a pre- X-ray error box and is considered as
liminary state. a tentative identification with Cyg
Jugaku has measured Sandage's X-2. Position measurements by Peach
spectrographic plates for radial ve- and photoelectric photometry by San-
locity changes and finds evidence for dage are consistent with this possi-
small, but apparently significant, var- bility on the basis of the X-ray po-
iations from night to night. The sition and the predictions of the color
plates were taken by the single-trail and magnitude of the optical object
technique with a long slit over time from the bremsstrahlung hypothesis.
intervals of from 2 to 3 hours on
QUASI-STELLAR SOURCES
Spectroscopy 454.3 (0.860), 4C -04.6 (1.445), AO
Spectroscopic observations of 0952 + 18 (1.472), 3C 205 (1.533),
quasi-stellar sources are being con- 4C 29.50 (1.927), and PKS 0229 + 13
tinued by Schmidt. Further redshifts (2.067) .
have been determined for the QSSs Sargent and Schmidt have collab-
4C 01.4 (z = 0.261), 4C 15.1 (0.451), orated with Dr. J. N. Bahcall of the
3C 57 (0.670), 4C -00.6 (0.720), 3C California Institute of Technology
290
CARNEGIE INSTITUTION
in an analysis of the absorption spec-
trum of 3C 191 using a 200 A/mm
spectrogram obtained by Schmidt.
The main part of the analysis was a
study of the relative populations of
the ground-state fine-structure levels
of various ions, which Bahcall has
shown may be used to obtain limits
on the electron density and photon
flux in the region where the absorp-
tion lines are formed. Fourteen ab-
sorption lines were studied, typically
having equivalent widths of about 10
A. A study of the central intensities
of different absorption lines showed
that most of them are unsaturated.
The main conclusions of the analysis
are: (1) Most of the absorption lines
are resolved and have half -widths of
the order of 3 A in the rest frame of
the source. (2) Either the electron
density is of order 10+3 cm-3 or the
distance between the continuum
source and the absorbing regions is
of order 10+2 sl pc. (3) The value of
the fine structure constant at z = 2
equals the laboratory value to within
measuring errors (about 5%). (4)
The C/Si abundance ratio by number
is 2.5/1, with a certainty that is prob-
ably less than a factor of 3. Thus
the composition of 3C 191 is signifi-
cantly different from Population I
stars, where the C/Si ratio is about
13. (5) There is no evidence on this
spectrogram for the dependence of
Zabs on ionization potential, which
Bahcall found earlier in the data of
Burbidge, Lynds and Burbidge, and
Stockton and Lynds. (6) There is no
evidence for absorption lines arising
from metastable states of S II and C
III, as other observers have claimed.
The emission-line spectrum of 3C
273 B was obtained by Greenstein at
38 A/mm. No trace of any sharp lines
was found, and no evidence of A3727
of [0 II] . The latter is suppressed at
high density. Its absence indicates
that no cool, low-density gas is pres-
ent, for example, in a cool shell or in
the cool interfilamentary spaces. The
relatively cool shells seen in absorp-
tion in more highly redshifted quasi-
stellar sources have small internal
velocity dispersion and electron den-
sities near 104 cm-3. Such shells
might have produced sharp forbidden
emission lines of [O II] if present in
3C 273B. Apparently the dense fila-
ments that produce the H and [0 III]
lines of 3C 273B are imbedded in a
hotter region.
Parkes Source 0237-23
In the fall of 1966, John Bolton,
director of the Australian Radio
Observatory at Parkes and, at that
time, guest investigator at Mount
Wilson and Palomar Observatories,
pointed out the unusually flat radio
spectrum and sharp, synchrotron self-
absorption cutoff of the radio source
Parkes 0237-23. Two spectra at 400
A/mm and one at 190 A/mm were
obtained with the 200-inch by Arp.
Later, spectrograms further into the
red, and photometry were obtained by
Kinman of the Lick Observatory.
This relatively bright quasi-stellar
source has the largest redshift meas-
ured to date. The emission-line red-
shift is z = 2.22. Arp, Bolton, and
Kinman reported only a slightly
smaller redshift, z = 2.21, for the
numerous sharp absorption lines that
are present in the spectrum. Later,
E. M. Burbidge reported an absorp-
tion-line redshift of z = 1.95. This
led to an attempt by Greenstein to
observe it at high dispersion. Two
spectrograms at 90 A/mm permitted
a resolution near 1.5 A on the better
plate. Although a larger number of
probably real absorption lines were
measured, a selected list of about 50
was published by Greenstein and
Schmidt. The existence of two sepa-
rate sets of lines with shifts of 2.202
and 1.956 was established. The lines
are sharp and deep. There is little
doubt that lines from excited states
MOUNT WILSON AND PALOMAR OBSERVATORIES
291
exist, suggesting relatively high den-
sity and some collisional excitation.
In addition to the usual strong ab-
sorption lines such as Lyman a, C IV,
Si III, and Si IV, the ionized metals
are very strong, notably Ti III, Fe III,
and Mn II. Continued observations of
this extraordinary double shell are
planned.
Spectral Energy Distribution
Oke has continued to measure ab-
solute spectral energy distributions
of the continua of quasi-stellar
sources. Wavelengths are now chosen
in such a way that some of the
stronger emission lines also are meas-
ured. Data now exist for several
sources in addition to the seven for
which energy distributions have al-
ready been published. For 3C 345
several sets of observations exist, in-
cluding detailed observations of the
Mg II line at A2798. No obvious vari-
ations have been seen. The object 3C
208 has an energy distribution simi-
lar to 3C 245, and the A1550 line of C
IV must be very strong. Source 3C
334 has a spectrum in which the flux,
Fv, increases only slowly as v de-
creases; the Mg II line A2798 is
strong. Source 3C 249.1 has a nearly
flat spectrum in terms of Fv, and a
strong Mg II line. The radio-quiet
quasi-stellar object PHL 938 has an
energy distribution and line intensi-
ties almost identical with those in
3C9.
The two highly variable quasi-
stellar sources 3C 279 and 3C 446
have been studied in detail. Changes
in magnitude in 3C 279 of 0.25 mag
per day have been observed. The vis-
ual magnitude has varied by nearly 2
mag during the past year. Following
the outburst of 3C 446 discovered by
Sandage in June 1966, scanner obser-
vations showed a decrease in bright-
ness of the continuum at the rate of
0.1 mag per day for 5 days. In both
objects the continuum is somewhat
redder when the object is fainter. The
absolute strengths of A2798 of Mg II
in 3C 279 and A1550 of C IV in 3C
446 have remained constant, inde-
pendent of any change in the con-
tinuum. This confirms an earlier re-
port by Sandage, Westphal, and
Strittmatter. The rapid light varia-
tions of the continuum imply that
most of the continuum comes from a
region with a diameter of the order
of 3 X 1016 cm. The region produc-
ing the nonvariable emission lines
must be much larger, and conven-
tional models with dimensions of 1
pc or more apply.
Wampler and Oke have studied 3C
273 in great detail using the Lick
and Palomar photoelectric scanners.
They have found several broad com-
plex emission features; the evidence
is strong that these are lines of Fe II.
A considerable contribution to the
[0 III] line, A5007, is in fact due to
Fe II. One feature is probably the
sodium D lines. No evidence for He
I A4471 or He II A4686 is found. The
presence of permitted Fe II lines and
the absence or extreme weakness of
[Fe II] lines indicates that the elec-
tron density in the emitting region is
at least 106 cm-3.
Optical Survey for Radio-Quiet
Quasars or Quasi-Stellar Objects
Work has continued on the nature
of the many faint blue objects found
in high galactic latitudes. Sandage
and Dr. W. J. Luyten of the Univer-
sity of Minnesota completed the first
phase of their investigation of the
Haro-Luyten blue star field centered
at lh36m, +6° (1855). As reported
last year, UBV photometry of 69 ob-
jects, divided equally among the three
Haro-Luyten color classes, showed
that a large number of the whiter
stars of color class III (PHL catalog,
Table IV), are F and G subdwarfs
292 CARNEGIE INSTITUTION
of the globular-cluster main-sequence density to B = 18.1 is at least 1 per
population. These stars populate the square degree.
galactic halo and occur in all magni- In a separate program, UBV pho-
tude intervals of the PHL catalog to tometry of 28 objects in the blue-star
its limit (B ~ 18.1). field of Richter near the globular
For the bluer stars of color classes cluster M 3 indicated that the surface
I and II (PHL catalog, Tables II density of radio-quiet QSOs is be-
am! Ill), objects of two types dom- tween 1 and 3 per square degree to
inate the list. Sandage obtained spec- B ~ 19.7, and, therefore, that the
trograms of 28 objects in the lh36m, number of QSOs increases rapidly
+6° held with the prime-focus spec- with increasing limiting magnitude.
trograph of the 200-inch. These were In view of these encouraging re-
selected from a candidate list of pos- suits, Luyten and Sandage began a
sible radio-quiet quasi-stellar galaxies more extensive search and analysis
in this field. The list was composed of program for blue objects in seven
objects from the 69-object photo- fields for which early first-epoch Sky
metric sample which had negligible Survey plates existed. These fields
proper motions as measured by Luy- are: 8h28m, +18°; 9h32m, +24°;
ten on 48-inch schmidt plates with 10h24m, +18°; llh16m, +30°; 12h34m,
an 11-year interval. Additional ob- +30°; 14h18m, +24°; and 15h10m,
jects were included from the bluest +24°. Three-color image plates were
PHL color class, again on the basis obtained with the 48-inch schmidt and
of negligible proper motion, even catalogs of the 8h48m, +18° and
though no UBV photometry existed. 15h10m, +24° fields have been pre-
The spectra showed that the can- pared and published. At the end of
didate list was dominated by radio- the report year, Sandage had obtained
quiet quasars (formerly called QSG UBV photometry for 73 faint objects
but now referred to as QSO, denoting in the 8h field and 15 objects in the
"quasi-stellar objects") and white 15h field with the 200-inch, and Luy-
dwarfs. Eleven definite QSOs were ten had initiated a good start on the
found, three of which showed several proper-motion measurements.
emission lines so that unambiguous Preliminary analysis of the photo-
redshifts could be determined. The metric data shows that the color dis-
redshifts were z = 1.990 for PHL tribution is of the same character as
1127, 0.298 for PHL 1194, and 1.847 in the lh36m field. The separation into
for PHL 3424. Although the remain- two types of distributions in the U —
ing eight objects had only one broad B, B — V diagram at V ~ 15 mag is
emission line somewhere in the region striking. Brighter than this limit,
AA3150-5000 A, these lines were of most objects follow the standard U —
such a character that the QSO iden- B,B — V line for main-sequence OB
tification was positive. stars, nuclei of planetary nebulae, hot
Sandage and Luyten estimate that, subdwarfs, and blue horizontal-
to B = 18.1, the surface density of branch stars of globular clusters.
QSOs in this particular field is at Fainter than this limit, few, if any,
least 0.3 per square degree, and prob- such stars are present, but rather the
ably 0.5 per square degree. This is distribution is almost entirely con-
100 times larger than the surface fined to the neighborhood of the
density of radio quasars in the SCR blackbody line, which is the domain
catalog and agrees with the esti- of the QSOs, white dwarfs, U Gemi-
mate given last year before spectra norum variables, and composite stars,
had been obtained. The white-dwarf These data suggest that the density
MOUNT WILSON AND PALOMAR OBSERVATORIES 29
o
gradient perpendicular to the galac- tion of QSSs must be increasing with
tic plane for normal OB stars, hot redshift, the excess density factor
subdwarfs, etc., is so steep that few being about 15 at redshift 0.5, and
exist at halo distances corresponding about 100 at redshift 1. These esti-
to V c=l 15. On the other hand, white mates depend on the cosmological
dwarfs with Mv > +10 do exist to model; those given are for a value of
distances of at least z = 500 pc and q0 between 0 and 1.
contribute to the sample fainter than This density increase at earlier
V = 15. There are also many objects cosmic times is of the same order of
in the photometric sample that fall magnitude as that found by Cam-
above the blackbody line and are, bridge radio astronomers from source
therefore, good QSO candidates. On counts. An alternate interpretation of
the basis of the data now available, the source counts considered by the
Sandage and Luyten estimate that radio astronomers is that the mean
there are at least 100,000 QSOs to an luminosity rather than the density of
optical limit of B ~ 19.7 over the the sources was higher at earlier
entire sky, and that this number is cosmic times. The present investiga-
an increasing function of limiting tion of 3C sources shows that the
magnitude. density increase is independent of
Braccesi has proposed and tested a luminosity, and thus excludes a varia-
technique for discovering QSOs photo- tion of mean luminosity with cosmic
graphically through their infrared time.
excess, which in three-color photome- The distribution of absolute optical
try should permit them to be distin- magnitudes per unit volume is fairly
guished from stars. In trials using constant over the observed range of
1-N plates at the 48-inch schimdt some four or five magnitudes. The
telescope, he produced a list of four radio luminosity function shows a
QSO candidates that were later considerable gradient, the number
confirmed to have large redshifts. per magnitude interval increasing by
Three of the spectrograms were ob- a factor of 3 or 4 per magnitude. If
tained by R. Lynds of the Kitt Peak the quasi-stellar objects found by
National Observatory and one by Sandage are relatively weak radio
Schmidt. sources, then their surface density on
... the sky can be explained by extend-
Space Distribution mg. j-j^ fair]y steep radio-luminosity
Schmidt has studied the space dis- function only a few magnitudes be-
tribution of QSSs. For this purpose yond the observed range of radio
he used the 33 sources in the re- absolute magnitudes. If this is cor-
vised 3C catalog that are optically rect, then the local (2 = 0) space
brighter than visual magnitude 18.4 density of all quasi-stellar objects, in-
and for which the redshift has been eluding those observed as radio
determined. For each source a sources, is about 10s Mpc-3.
"maximum" redshift, zm, can be com-
puted at which it would still be con- Optical Variability
tained in the revised 3C catalog Peach has started a program to
(which has a lower limit of 9 flux monitor the optical variability of the
units at 178 Mc/s) and would be known quasi-stellar sources. Using
optically not fainter than magnitude the 48-inch schmidt, blue plates are
18.4. A comparison of the actually taken monthly of all objects observ-
observed redshift with zm for each able at the time. Despite the poor ob-
source shows that the space distribu- serving season, over 300 plates of 83
294 CARNEGIE INSTITUTION
objects have been obtained within the of the brighter QSSs has been de-
interval from November 1966 to June layed by poor weather.
1967. This program is intended both
to provide information on optical Infrared Emission
variability throughout this class of Neugebauer and Becklin have
object and to detect sudden changes measured the infrared emission from
of brightness that may be correlated eight quasi-stellar sources at 2.2 /x
with spectroscopic or other changes, with the 200-inch telescope; (V — K)
A further program to observe is about 3. Further observations are
shorter-term variations among some now in progress.
THEORETICAL STUDIES
Heating of the Solar Atmosphere QSO Spectra
Stein proposed a model for the Dr. J. N. Bahcall of the California
heating of the solar atmosphere in Institute of Technology has collabo-
which the magnetic field controls the rated with Sargent on a study of the
heating through its effect on wave emission lines to be expected in the
propagation. The solar atmosphere is spectra of quasi-stellar objects be-
heated mainly where the magnetic tween Lyman a (A0 1215) and the Ly-
field is large. Where the field is weak, man limit (A0 912). Estimates were
the acoustic wave flux, produced by made of the strength of emission lines
the turbulence associated with the of the abundant elements, based on
convective zone, is dissipated in the the behavior of these elements in
form of shocks in low-lying high- QSO spectra that have already been
density layers, and produces little observed at longer wavelengths. This
heating. On the other hand, where work was intended partly to facili-
the field is large and acoustic waves tate the discovery of QSOs having
become fast-mode waves, the dissipa- redshifts greater than z = 2. It was
tion decreases at great heights, where concluded that only one very strong
the ratio of the Alfven speed to the line, A0 977 of C III, is to be expected
sound speed becomes large. Thus, in the wavelength region studied.
where the magnetic field is large, the Bahcall and Sargent also pointed out
wave energy can be carried up to that two effects, the presence of Ly-
great heights. To pursue the model man a in the B filter and the existence
in detail, the propagation properties of a large Lyman discontinuity in
of magnetohydrodynamic waves were absorption, will cause QSOs having
investigated. A study of the reflec- redshifts greater than z = 2.2 to be
tion and refraction of fast-mode and redder in U — B than objects having
slow-mode waves from density and z < 2.2. Bahcall and Sargent thus
magnetic-field jumps was nearly com- suggest that the search for optical
pleted, and an investigation of their identifications of radio sources should
propagation in a stratified atmos- not be confined to objects having a
phere in the Wentzel-Kramers-Bril- pronounced ultraviolet excess,
louin approximation was begun. A
crude model of the magnetic field in Observational Cosmology
the network was also considered, and A series of computer programs has
it was found that only along lines of been written by Peach for the analy-
force reasonably close to the vertical sis of the redshift-magnitude obser-
does the Alfven velocity increase up- vations by Sandage of first-ranked
ward. cluster galaxies and radio galaxies.
MOUNT WILSON AND PALOMAR OBSERVATORIES 295
The data are analyzed in terms of the dent, has completed calculations of
homogeneous isotropic cosmological the dynamics of a collapsing proto-
models of general relativity, both star for various initial conditions. In
with and without the inclusion of a all cases the collapse is found to be
cosmological constant in the redshift- extremely nonhomologous, in such a
magnitude relation. These programs way that a very small part of the
determine the best values of the den- mass at the center reaches stellar
sity and deceleration parameters in a densities and stops collapsing before
least-squares sense; the errors of most of the cloud has had time to
these quantities are investigated by collapse very far. The stellar core
Monte Carlo techniques. The neces- thus formed subsequently grows in
sary evolutionary corrections to the mass as material falls into it. During
apparent magnitudes of distant ob- most of this time, the forming star is
jects can be included in the compu- completely obscured by dust in the
tations. As the observational data are inf ailing cloud, the absorbed radia-
not yet completely reduced, a final tion reappearing in the infrared as
value of the deceleration parameter thermal emission from the dust. The
is not available, but preliminary cal- resulting star is almost a conven-
culations indicate a higher value than tional Hayashi pre -main -sequence
those previously obtained. These cal- model, but it appears rather low on
culations further indicate that the the Hayashi track. It seems that some
cosmological constant is zero to the of the observed properties of T Tauri
precision of the determination. stars may be explained on the basis
of the infall of material and the in-
Star Formation f rared emission from dust, as found
Richard B. Larson, a graduate stu- in this work.
GUEST INVESTIGATORS
Under the guest investigator pro- object have been submitted for publi-
gram, which has been in effect at the cation while others are in prepara-
Observatories for many years, 25 as- tion. Photographic and photoelectric
tronomers from other institutions spectrophotometry observations of
made use of the observing facilities planetaries are complementary in the
at Mount Wilson or Palomar. sense that strong lines can be meas-
Dr. Lawrence H. Aller of the Uni- ured best by the photoelectric tech-
versity of California at Los Angeles nique, and weak lines by the
and Dr. Stanley J. Czyzak of The photographic method. However, the
Ohio State University observed plane- photographic measurements must be
tary nebulae with the 100-inch and calibrated photoelectrically. Advances
60-inch telescopes at Mount Wilson, in the theory of atomic line strengths
At the 100-inch coude spectrograph and collisional cross sections permit
they secured plates of NGC 6818, a degree of precision in the studies of
6886, and IC 4997. With the photo- the physical processes in gaseous
electric spectrum scanner of the 60- nebulae that puts severe requirements
inch they observed NGC 40, 650, on the observations. It is now evident
1514, 2440, 4361, 6537, 6751, 6790, that much more attention will have to
6818, 6826, 7026, 6905, 7354; IC 1747, be paid to photoelectric intensity
4846, 5217; and VV 286. Particular measurements. A fundamental limi-
emphasis was placed on NGC 2440, tation in photoelectric work is set by
and results on the spectrum of this the spectral resolution, which re-
296 CARNEGIE INSTITUTION
quires in all instances a judicious South Wales, Australia, worked as a
combination of photographic and guest investigator at the Mount Wil-
photoelectric photometry. son and Palomar Observatories. He
With the 48-inch schmidt telescope obtained approximately one hundred
Dr. G. A. Tammann of the University 5 X 7-inch plates with the 48-inch
of Basel, working on Professor W. schmidt telescope to determine the
Becker's program, obtained R, G, U- precise optical positions of the coun-
plates for two fields centered on the terparts of identified radio sources of
globular clusters M 13 and NGC 4147. small diameter. The plates were sub-
The availability of faint photoelectric sequently measured on the new
sequences in these helds will permit proper-motion machine at the Lick
the ready determination of the lumi- Observatory. All reductions have been
nosity and density function for field completed and the material is being
stars down to the plate limit, while prepared for publication. The posi-
the delay continues in the investiga- tions will be of particular value in
tion of some other fields taken earlier future calibration of position meas-
with the same telescope because urements on other radio sources and
faint photoelectric standards are lack- were made primarily for use at the
ing. Parkes 210-foot telescope.
Dr. Roberto Barbon of Asiago About 20 of the foregoing plates
Astrophysical Observatory recorded were taken in two colors to confirm
spectra of five bright meteors with suspected quasi-stellar objects. Fur-
the 18-inch schmidt telescope and ob- ther photoelectric or spectroscopic
jective prism while working on a observations were made by Kinman
spectroscopic survey of regions near and E. M. Burbidge at the Lick Ob-
the galactic pole. All of the spectra servatory and by Arp, Sandage, and
are of low excitation with the usual Schmidt at Palomar. Included is the
Xa I, Mg I, and Fe I emission lines. QSO with the largest known redshift
The interpretation of some bandlike (PKS 0237-22, z = 2.223) and two
features in the red portion of the others with the second and fifth larg-
spectrum of three of the meteors is est redshifts (PKS 0056-17, z =
rather uncertain; explanation in 2.125, and PKS 0229 + 13, z = 2.065).
terms of N2 and 02 bands is tenta- These results have been published
tively suggested. The work of inter- (Arp, H. C., J. G. Bolton, and T. D.
pretation is in collaboration with Dr. Kinman, Astrophys. J., 1U7, 840,
J. H. Russell of the University of 1967; and Kinman, T. D., J. G. Bolton,
Southern California. R. W. Clarke, and A. Sandage, Astro-
On a 48-inch schmidt plate, taken phys. J., 47, 848, 1967).
on August 15 to search for faint Twelve 14 X 14-inch plates were
blue, high proper-motion stars, Bar- taken by Bolton in two colors (GG-13
bon found a new comet, 1966c. filter, 8-min exposure; UG-1 filter,
With the collaboration of Dr. A. 60-min exposure) on centers of the
Braccesi, who provided the radio Palomar Sky Survey in the declina-
data, and by inspection of three-color tion 0° zone. Five of these plates are
plates, Barbon identified the follow- of excellent quality. They are being
ing 4C radio sources with blue stellar examined for objects with marked
objects: + 03.02, +02.04, +01.02, UV excess and a deep survey at 11-
-01.06, -01.61, -02.04, and -04.89 cm is being made with the Parkes
Between August 1966 and Febru- 210-foot telescope in the areas cov-
ary 1967, John G. Bolton, director of ered by these plates for radio-optical
the Parkes Radio Observatory, New comparison.
MOUNT WILSON AND PALOMAR OBSERVATORIES 297
Danziger of the Harvard College A/mm; the wavelength interval cov-
Observatory has obtained as guest ered was from 3400 to 7400 A. Special
investigator with the 200-inch tele- efforts were made to ensure the
scope lithium abundances in main- highest possible photometric accu-
sequence stars down to spectral types racy; the interagreement of overlap-
K0 in the Pleiades and Ursa Major ping spectrograms suggests acci-
clusters. The results for the Pleiades dental errors of 1% to 2% of the
support previous, less secure, obser- continuum. The plates were traced
vations that suggested less depletion with one of the Sinclair Smith micro-
in the Pleiades stars than in Hyades photometers of the California Insti-
stars of the same type. This would be tute of Technology, a linear scale of
consistent with depletion of lithium ordinates being obtained with an ana-
in stars on the main sequence. logue output system brought from
Danziger and Dickens are studying Cambridge. An investigation of the
spectral characteristics of new short- instrumental profile at A6328 in the
period variable stars and are accumu- second order was made, using a
lating rotational velocities, gravities, helium-neon laser source,
and effective temperatures. An at- Dr. Hugh M. Johnson of the Lock-
tempt is being made to explore the heed Palo Alto Research Laboratory
variable star gap in the region of observed with the 48-inch schmidt
short-period variability. telescope at Palomar. Exposures were
A program is under way to investi- made on the fields of X-ray sources
gate the physical properties of the en- Sco X-l, Cyg XR-1, Leo XR-1, Lyr
velopes of shell stars and Be stars — XR-1, and EX Hya. Plates of Cas A
using low-resolution scans and spec- were exposed in an attempt to detect
tra obtained at Mount Wilson by the yellow continuum that is to be
Danziger. expected between the radio-frequency
In a study of the hydrogen-poor and X-ray extremes of the spectrum,
star HD 30353, Danziger, Waller- and which may be revealed by high-
stein, and E. Bohm-Vitense have contrast techniques. Tests for polari-
used interstellar polarization meas- zation were made of NGC 6302, a
urements, reddening of neighboring candidate for identification as Sco
B stars, continuum scans, and models XR-2. One of the 14 t/V-excess stars
of hydrogen-poor stellar atmospheres in the field of Sco X-l was independ-
to estimate the most probable values ently found by C. B. Stephenson to
of reddening, E (B — V) = 0.35, show a peculiar O-type spectrum on
effective temperature (10,000°K) and Case objective-prism plates, and John-
absolute magnitude (—3.2). son obtained a slit spectrogram of it
Dr. and Mrs. Roger Griffin of the at the Kitt Peak National Observa-
Cambridge Observatories, England, tory, identifying it as the X-ray
spent four months in Pasadena to ob- source. The Palomar plates gave the
tain material for a high-resolution first evidence of null polarization for
high-accuracy photometric atlas of Sco X-l.
the spectrum of Procyon. The 100- With the 200-inch Hale telescope,
inch coude spectrograph (114-inch Dr. Philip C. Keenan of Perkins Ob-
camera and 133B grating) was used servatory obtained 22 coude spectro-
on 10 nights, and a total of 17 very grams of variable stars and two
wide spectrograms was secured, using plates of comparison stars on 12
the second, third, and fourth orders nights during the summer of 1966.
of the grating to obtain reciprocal This program was undertaken in co-
dispersions ranging from 0.7 to 1.6 operation with Deutsch and Garrison.
298
CARNEGIE INSTITUTION
One of the Mira variables, T Sgr,
observed on September 3, showed the
strongest absorption of the line Li
I A 670$ yet observed in an S-type
star. The measured total absorption
of 1.2 A for X6708 in T Sgr was about
three times as great as that observed
by Miss Bretz in HD 212667 or in
AA Cyg, which have been considered
as the S-type stars in which lithium
was most abundant. T Sgr has a
nearly pure S-type spectrum, and is
spectroscopically similar to R Gem
and R Cyg except for the great sur-
face abundance of lithium. The only
stars with even stronger Li lines are
such carbon stars as WZ Cas, which
have been referred to as "lithium
stars."
Dr. Jerome Kristian of the Uni-
versity of Wisconsin attempted to de-
tect the cosmological distortion effect
by measuring the apparent distribu-
tion of position angles for galaxies in
clusters within 300 Mpc. The meas-
urements were made on 200-inch di-
rect photographs, mostly taken by
Sandage. While the results are not of
high weight, no evidence for the dis-
tortion effect was observed. It should,
indeed, be absent in Friedman and
steady-state universes. The results
were used to set an upper limit on
the size of the local incident gravi-
tational field. Preliminary measure-
ments were also made on the Palomar
Sky Survey plates of the projected
spatial distribution of galaxies in
more distant clusters. In addition to
an improved limit on the distortion
effect, it is hoped that such measure-
ments will yield a value for q0. This,
if it can be found, will be of lower
weight than the value obtained from
redshift-magnitude measurements,
but it may provide an independent
check.
Kristian has begun a program for
the measurement of the sizes of H II
regions on direct photographs of gal-
axies. In this connection, he has de-
vised a method of determining the
intensity response of a plate by mi-
crophotometric measurements of
star-image profiles. This method,
when calibrated by photoelectric
measurement of two or three rela-
tively bright stars, may be applicable
to the determination of magnitudes of
stars on a plate, over a wide range,
to within a few tenths of a magni-
tude; it offers the possibility of ob-
taining improved magnitude data
from the large number of existing un-
calibrated direct photographs.
Dr. Barry M. Lasker, holding a
National Science Foundation Fellow-
ship, continued the program of inter-
mediate bandwidth photometry of
elliptical galaxies that was described
in last year's Annual Report. Part of
the effort was directed toward re-
observing a sufficient number of
standard stars to provide adequate
checks for consistency on the photo-
metric system used during these
years. A multivariate regression pro-
gram was written for the IBM 7094
computer to obtain the best set of
values for the standard stars; pre-
liminary inspection of the results in-
dicates that the photometric system
is defined to about 0.015 magnitude
in all bands except in the reddest and
bluest, where the accuracy is some-
what reduced. In addition to reobserv-
ing several galaxies that were first
done in the previous year, the sample
of galaxies observed was increased,
and additional data for members of
the small groups and of certain clus-
ters were obtained.
Professor W. J. Luyten of the Uni-
versity of Minnesota has continued
his program of taking new plates to
be blinked against the Palomar Sky
Survey plates for the Proper Motion
Program. To date nearly 600 regions
have been rephotographed with the
48-inch schmidt telescope; of these,
110 have been examined under the
blink microscope, and some 50,000
MOUNT WILSON AND PALOMAR OBSERVATORIES 299
proper-motion objects have been 2, 3, and 4 have similar colors while
marked. In particular, the entire re- satellite 1 has a striking irregularity
gion north of declination +69° has in the blue. It is hoped to continue
been examined and some 21,000 this program during the coming win-
proper-motion stars marked in this ter and to include some asteroids as
region. Luyten has also used the 48- well as other satellites,
inch schmidt telescope for taking The McMath-Hulbert spectrometer
parallax plates of the large proper- attached to the Snow telescope was
motion star LP 9-231 at 17^6^8, used by Dr. Walter E. Mitchell, Jr.,
+ 82° 44', and from 37 plates with of the Perkins Observatory during
three exposures each derives a paral- September 1966 in the double-pass
lax of 0" 104 ± 0"025- mode to obtain high-resolution pho-
Thomas B. McCord of the Division toelectric records at the center of the
of Geological Sciences at Caltech solar disk, covering the wavelength
used the 60-inch telescope during range 4337-6150 A. This map is com-
the winter and spring of 1967 to parable in quality and coverage with
determine differential reflectivity var- the earlier double-pass map recorded
iation across the surface of the planet in 1962. It is intended to supply for
Mars, with emphasis on differences a different phase of the solar cycle
between the mare and desert areas, the data for central intensities free
A double-beam photoelectric pho- of scattered light. Some effects of
tometer was used; this allowed simul- spectrograph seeing were noted,
taneous measurement of two areas namely in the occasional smearing of
of Mars through the same aperture, finely spaced iodine doublets and in
filter, and detector system. Twenty- the introduction of variation in the
one narrow-band interference filters intensity in a line profile of up to 1 %
spaced evenly between 4000 A and or 2 % of the continuum. Two-channel
8000 A were used. Throughout the radio recording was carried out suc-
wavelength region indicated, the sys- cessfully in several instances to com-
tem permitted the examination of the pensate for varying sky transparency,
ratio of the reflectivity of various Louis E. Simpson assisted in this pro-
regions about one arc second in di- gram.
ameter with an accuracy of about Dr. Guy C. Omer of the University
2%. Some of the results to date are: of Florida used the 48-inch schmidt
(1) The reflectivity ratio of desert telescope at Palomar on two nights
to mare areas follows an almost lin- in August 1966 to obtain red and blue
ear increase from 1.00 at 4000 A to plates centered on the following clus-
about 2.10 at 8000 A. There is, how- ters of galaxies: Abell 168, 2065,
ever, an enhancement of the mare 2151, 2199, 2255, and 2256. The
reflectivity of about 5% between 5600 plates of A 2199 have been analyzed
A and 7600 A that appears to be and the cluster has been found to have
seasonally dependent. (2) The reflec- essentially the same spatial distribu-
tivity of Mare Acidalium appears to tion of density as the Coma Cluster,
differ by about 5% from that of including the theoretically interesting
Septis Major in the region 4400- point of inflection at about two thirds
7600 A. The same system was used of the cluster radius. The peculiar
to determine precise (^0.004 mag) galaxy Arp 125 is not a member of
partial light curves for the four Gali- this cluster. However, Arp 125 ap-
lean satellites of Jupiter. These ob- pears to be a member of the cluster
servations have not yet been fully A 2197, which is just north of A 2199.
reduced but it appears that satellites These two clusters appear to be at
.">
00 CARNEGIE INSTITUTION
about the same distance and may be coude spectrograms of the B5 eclips-
interrelated. Radio surveys of all of ing binary, U Oph, was obtained in
the foregoing clusters are planned for order to resolve the discrepancy in
the near future in collaboration with mass determinations of Plaskett
T. D. Carr. (masses 5.3 and 4.6 nio) and of
Dr. D. E. Osterbrock and Mr. D. Abrami (4.0 and 3.9 v^o) . Since this
\Y. Weedman of the University of is one of the most favorable B-type
Wisconsin obtained spectrograms systems for the purpose, the discrep-
with the 100-inch coude spectrograph ancy is not tolerable. Plaskett's re-
to continue the study of line profiles suits have turned out to be essentially
in planetary nebulae. Fifteen spectro- correct. (2) Preliminary results are
grams of the nebulae IC 418, NGC available for the masses of the sub-
2392. and NGC 3242 were obtained giant systems WW Dra and SS Boo.
with the same instrumental setup For the former, Joy's old values of
used for the previously published 3.9 and 2.3 Vi\o have required re-
work on summer-sky planetaries vision to values of 1.3 and 1.3, more
(Osterbrock, Miller, and Weedman, in line with expectation from current
Astrophys. J., 1^5, 697, 1966). All ideas of stellar evolution. For SS Boo,
the spectrograms were traced on the however, it appears that the masses
Caltech microphotometer, with a ve- of both components are probably less
locity resolution of approximately 5 than 1 nio, a result difficult to inter-
km sec. Of the three nebulae, NGC pret in terms of current theory unless
3242 is fairly typical of other plane- there has been appreciable mass loss
taries, while IC 418 has an unusually from the system as a whole. Popper
small range in velocity of expansion, has also made a valuable study of the
so that H and [0 III] lines are not systematic and accidental errors in
resolved even at high dispersion, radial velocity determination with the
though the [0 II] and [N II] lines Mount Wilson spectrographs,
are, and XGC 2392 has an unusually The 48-inch telescope was used on
complicated velocity structure, show- four nights in January by Dr. Sid-
ing a range of almost 100 km/sec. ney van den Bergh of the David Dun-
Dr. J. E. Dyson of the University lap Observatory, University of To-
of Wisconsin has traced on the Cal- ronto. He obtained plates of a number
tech microphotometer multislit spec- of H II regions in the winter Milky
trograms of the Orion Nebula ob- Way, giving special attention to areas
tained by Munch and Wilson. Four containing globules, "elephant trunk
plates were traced at [O II] 3727 A, structures," and "fossil H II regions."
[O III] 5007 A, and Hy. Regions A number of plates of M 31 and M 33
where the oxygen lines are split were were obtained as part of a 10-year
examined in detail. These observa- study of the brightest variables in
tional results will be compared with a these galaxies. A number of new
theoretical model of the supersonic bright cepheids in M 31 have been
turbulence within the nebula. discovered on these plates by A. F. J.
Dr. Daniel M. Popper of the Uni- Moffat, and a 17th magnitude RR
versity of California at Los Angeles Lyrae variable with a period of
continued his spectroscopic inves- 0.5788 days has been found projected
tigations with the 100-inch and 60- on one of the spiral arms of M 31.
inch telescopes. His special interest in Using 200-inch photographs, van den
the improvement of stellar mass Bergh is also investigating changes
evaluations has led to the following in the expansion of the filaments as-
new results: (1) A new series of sociated the supernova Cas A.
MOUNT WILSON AND PALOMAR OBSERVATORIES 301
Dr. George Wallerstein of the Uni- the star was varying. Routine spec-
versity of Wisconsin used five nights trograms of the peculiar stars, RS
with the 200-inch telescope to com- Oph and VY CMa, show no substan-
plete a survey for lithium in 28 lb tial changes as compared with plates
supergiants of types F and G, and taken during the past few years.
18 cepheid variables. Visual inspec- The Astronomer Royal and D. R.
tion shows that none of these has a Palmer of the Royal Greenwich Ob-
strong lithium line. The spectrograms servatory carried out a short pro-
are being microphotometered for line gram with the Newtonian spectro-
identifications and measurement of graph of the 100-inch telescope. In
equivalent widths. The data will be August and September 1966, they
discussed in cooperation with Conti determined the radial velocities of
of the Lick Observatory, who has six cepheid variable stars classified
completed a similar survey for beryl- as W Virginis in the General Cata-
lium in A and F supergiants. If the logue of Kukarkin et at. (CW), or as
apparently low lithium content is C II cepheid variables by Petit, with
confirmed by measurements, it indi- the object of searching for fast-mo v-
cates either that all the observed lb ing cepheid variable stars. These have
supergiants and cepheids have mixed been supposed to exist, on account of
their surface material to a sufficient the presence of variables with periods
depth to deplete their lithium content, longer than one day in globular clus-
or that stars of sufficient mass to ters ; it might be supposed that there
evolve into the F-G lb and cepheid are similar stars in the galactic field,
regions of the H-R diagram never had that is to say, variable stars with
as large a surface lithium content as the same kinematic properties as the
main-sequence stars of 1-2 ttl©. RR Lyrae stars, and with period
Wallerstein has searched unsuc- greater than 1<?0. Some writers have
cessfully for interstellar D line in identified the CW and C II stars with
stars located in the direction of high- this class of object, but an examina-
velocity clouds discovered by their 21- tion of the kinematics of the CW and
cm emission. The data have been com- C II stars with known radial veloci-
bined with a similar negative ties had shown that they were in fact
observation of the same stars at the quite different from those of the RR
Ca II K-line obtained two years ago Lyrae stars generally. Hence the
at the Lick Observatory. By combin- search. On the present occasion, two
ing the density of H I as derived from of the six stars had velocities in
the 21-cm data with the upper limits excess of 100 km/sec, namely V 716
on the equivalent widths of the inter- Oph (-334 km/sec) and FM Del
stellar absorption lines it can be (-154 km/sec) , the rest having kine-
shown that the H I clouds lie beyond matic properties rather similar to the
the stars. Distances as large as 1500 known CW and C II stars. These two
parsecs are derived. fast-moving stars can be compared
One Palomar spectrogram by Wal- to UY Eri, discovered by Greenwich
lerstein of RU Cam, the 22-day observers working in Egypt to have
former variable that ceased variation, a radial velocity of +162 km/sec. All
has been combined with three Lick three have periods less than 2% days
Observatory plates and five David and seem to be an extension of the
Dunlap Observatory spectrograms to RR Lyrae stars beyond the conven-
show that the velocity is currently tional limit of a period of 1^0.
constant and that it is equal to the Using the 60-inch telescope, Mr.
mean velocity found by Sanf ord when Robert L. Younkin of the Jet Pro-
302
CARNEGIE INSTITUTION
pulsion Laboratory of the California
Institute of Technology measured the
narrow-band color of the disk of Sat-
urn from X3300 to 11,000 A when
the rings were closed and the color
of the rings when narrowly open; he
also made equatorial limb-darkening
scans at visible and ultraviolet wave-
lengths with the rings closed.
Younkin has measured the limb dark-
ening of Uranus within and adjacent
to strong methane absorption bands
and has determined that there is in-
creased limb darkening within the
band. Younkin and Munch have de-
termined the radiometric albedo of
Uranus to be 0.32 and the effective
temperature to be 56 °K. The mono-
chromatic geometrical albedo attains
a maximum of 0.75, indicating a plan-
etary phase factor of 1.33.
60-INCH PHOTOMETRIC TELESCOPE
Optical Design
Construction of the new 60-inch
photometric telescope for the Palo-
mar Observatory is proceeding under
the grant made by the National Sci-
ence Foundation to the Carnegie
Institution. Several major parts have
been completed or are being fabri-
cated.
The Ritchey-Chretien design for
the telescope has been investigated by
Bo wen and procedures have been set
up for testing the primary and sec-
ondary mirrors. The primary will be
tested with the aid of the 60-inch flat
that was figured at the Observatory
60 years ago for tests of the Mount
Wilson 60-inch and 100-inch mirrors.
With such a flat, the additional cor-
rection required to give a clean knife-
edge cutoff of the Ritchey-Chretien
primary is very small. This correc-
tion can be supplied by introducing
behind the first focus a 5-inch spher-
ical concave mirror of 19-inch radius.
Both the Cassegrain and coude sec-
ondaries of the Ritchey-Chretien sys-
tem are such close approximations to
hyperbolas with displaced conjugate
foci that they can be tested by the
Hindle sphere procedure. Two 36-
inch diameter mirror blanks are on
hand for the required Hindle sphere
tests of the 60-inch primary and sec-
ondary optics. These two blanks will
be finished to a radius of curvature
suitable for the Hindle tests; they
will later be used as camera mirrors
in coude spectrographs for the 60-
inch telescopes at Mount Wilson and
Palomar.
Optical Work
The primary 60-inch fused silica
mirror has now been shaped on the
new 60-inch grinding machine. Dia-
mond-wheel trimming is proceeding
interchangeably with work on the
secondary mirrors, one of which is
now ready for final figuring. All of
the accessory handling equipment ap-
pears to be operating satisfactorily.
Mechanical Parts
The cell for the primary mirror is
complete in the shop, as are all parts
of the lever support system for the
mirror; the support system will not
be required until optical tests are
undertaken.
All major components of the tele-
scope mounting have been detailed
except the base pedestal. Fabrication
of major components was started in
October 1966, and such work is pro-
ceeding with some schedule rear-
rangement owing to delays in pro-
curement of certain manufactured
items such as bearings and gears.
Rule has analyzed the specifications
and tolerances for the telescope drive,
and the detailed design of the drive
MOUNT WILSON AND PALOMAR OBSERVATORIES 303
system is well advanced. Fabrication carried out by Westphal with 100-foot
of the mechanical drive subassemblies steel towers, one at the selected site
has been scheduled. and one on the ridge some 200 yards
northwest of the 200-inch telescope.
^l^e Brush clearing" was completed at
The location for the photometric the site of the 60-inch telescope, as
telescope on a knoll approximately well as at the locations for two new
% mile southeast of the 200-inch cottages, one for a night assistant's
dome was approved upon recommen- residence and the other for staff and
dation by the Observatory Committee, guest observers. Acceptance of pre-
The choice was made in part on the liminary floor plans for the three new
basis of terrain and environmental structures has been recommended by
factors and in part on the basis of ex- the Observatory Committee and the
tended tests of microthermal activity plans have been updated by Rule.
INSTRUMENTATION
Image-Tube Spectrograph proved magnetic-shield system that
The design and construction of a reduces these variations to a permis-
nebular spectrograph utilizing a Car- sible amount, without appreciably
negie-RCA image tube has been re- changing the magnetic field required
ported in the last three Year Books. to focus the electron beam, has been
This instrument, with its optical de- designed by Bowen.
sign effected by Bowen and its me- Tne spectrograph has been fitted
chanical design by Rule, has been in with comparison light sources and a
use on the 200-inch telescope since preliminary guiding system to be
August 1967. Tests and adjustments use^ until the larger offset guider is
have been carried out in the labora- ready. This spectrograph is the first
tory recently by Bowen and Schmidt. of several major interchangeable
The concentric-mirror spectrograph Cassegrain instruments for the 200-
camera, which operates at a focal inch telescope to fit the common trun-
ratio of //0.8, yields images in which nion and offset guider system.
practically all of the light over the
designed field of 12° falls in a band Multichannel Spectrometer
8-10^ wide. Over a wider field of 16° The multichannel photoelectric
small amounts of astigmatism and spectrometer described in Year Book
field curvature are present, but not 65 (pp> 174_175) is in an advanced
enough to degrade appreciably the stage of constrUction and should be
images formed by the electronic tube. in operation on the 200-inch telescope
With a first-order 600-lme/mm grat- before the end of 196? TWg instru.
ing, the spectrograph covers the ment alsQ win fit the trunnion and
wavelength range 3200-7100 a in one common guider system at the Casse.
exposure with a dispersion of 220 in focus>
A/mm. Ihe 1/1.2 reimagmg camera
operates at a focal ratio f/1.1 on the
phosphor side; it yields its predicted
definition. A new 4-element corrector for the
The effect on the definition of varia- 200-inch, working at //3.52, has been
tions in the magnetic field perpen- designed by Dr. C. G. Wynne of Im-
dicular to the electron beam of the perial College, London. The required
image tube has been studied. An im- blanks of UBK-7 optical glass have
Prime-Focus Corrector
304 CARNEGIE INSTITUTION
been ordered and a lens barrel adap- in July 1967. High-speed binary
table to the prime-focus pedestal has scalers permit the use of these ampli-
been devised. This lens is designed tiers with the existing counters while
to yield excellent images in a field 25 high-speed counters are being tested.
arc-minutes in diameter over a wave- The general success of pulse-count-
length range from 3650 to 10,140 A. ing photometry has resulted in the
All surfaces are spherical except for installation of a temporary pulse-data
the last, which is flat. system on the 60-inch telescope at
Mount Wilson. Although this ar-
Astro-Electronics Laboratory rangement lacks the sophistication of
The coordinated instrument-devel- the other data systems, it has been
opment program under Dennison's very useful and has permitted the
supervision is making significant ad- abandonment of dc amplifiers. A new
vances toward achieving the most data system will be designed and
effective operation level possible for built as part of the 60-inch moderni-
Mount Wilson and Palomar Observa- zation program.
tories' telescopes. Observers report A digital control system for the
that the use of the digital data sys- two-channel interferometric spectrum
terns in photoelectric work has nearly scanner, used in conjunction with the
doubled the number of observations coude spectrograph of the 200-inch
obtained per night. Coude spectro- telescope and described in Year Book
photometric measurements impracti- 65 (p. 175), has been completed, and
cable with previously available tech- preliminary observations have been
niques can now be made routinely, made with it by Munch and Vaughan.
Solar magnetograph plots of high in- The digital system controls the spec-
formation content, generated by the trograph-scanner drive as well as a
7094 computer from tape-recorded motor on the interferometer gas-pres-
data, are now a daily accomplish- sure controller. This unit is basically
ment. It is expected that faster re- similar to the one that has been in
duction programs will permit the as- use on the 100-inch telescope for sev-
tronomer to make more meaningful eral years (see Year Book 6U, pp. 48-
use of the data. Past experience has 49). Results show that the instrument
demonstrated repeatedly that auto- is highly satisfactory in spectral re-
matic data-reduction equipment in- gions that can be reached in the third
creases not only accuracy but also and second orders of the existing 12-
operator efficiency. inch composite coude grating used as
The high-speed pulse amplifiers de- a predisperser. The Ca K and Na D
scribed in Year Book 65 (p. 173) have lines are included in these regions.
been developed, constructed, and de- For longer wavelengths a grating of
livered by Solid State Radiations, higher intrinsic dispersion would be
Inc., of Santa Monica, Calif. These desirable.
amplifiers have been tested in the A new data system has been in-
Astro-Electronics Laboratory and are stalled in the Santa Barbara Street
found to give paired pulse-resolution building to serve both the iris pho-
times of less than 30 nanoseconds and tometer for star magnitude measure-
to be of exceptionally low noise. The ments on direct photographs and the
shielding appears to be adequate to X-Y measuring engine for determin-
reject all man-made interference, and ing star positions to 1 n on plates up
there is every reason to believe that to 20 inches square. This data system
these units will fulfill expectations, was designed to work with either in-
Routine operation of the units began strument. It has sufficient flexibility
MOUNT WILSON AND PALOMAR OBSERVATORIES
305
to be used in the future with any de-
vice having up to 3 coordinates that
can be mechanically coupled to an
optical - mechanical hybrid encoder.
Each encoder provides a resolution of
1 part in 106.
When used with the iris photom-
eter, the encoders are connected to
the plate-coordinate positioners and
to the servo-driven gear that deter-
mines the iris diameter. The plate
carriage is manually positioned in
two coordinates. The mechanical load-
ing by the encoders is extremely light.
Recording on printed punch cards
for computer reduction, the system
handles also supplementary data such
as the object name, the plate number,
the date, individual star identifica-
tion, and the operator's estimate of
image quality.
The laboratory now has in opera-
tion apparatus developed by Oke that
allows multiplier phototubes to be
tested under simulated telescope con-
ditions. At controlled temperatures
ranging down to that of dry ice and
lower, relative and approximately ab-
solute quantum efficiencies are rou-
tinely determined. Pulse - counting
equipment can also be tested.
Several experimenters have shown
that quantum efficiencies of some pho-
tomultiplier tubes can be improved
by oblique illumination. Using par-
tially aluminized hemispheres, Oke
and Schild have found a practicable
way by which oblique illumination
can be used at the telescope. For a
group of red-sensitive 520 photomul-
tipliers, they have found gains in
speed of from 1.5 to 2.3 at X5000 and
2.4-3.8 at A8000. For the infrared SI
photomultipliers tested, gains range
from 1.5 to 1.7 and are almost inde-
pendent of wave length. The tech-
nique is being applied to the multi-
channel spectrometer and can be
used for single-tube operations as
well. The reasons for these gains are
being investigated theoretically.
Other Instrumentation for the
Large Reflectors
A single-channel sky-compensated
polarimeter specially designed accord-
ing to specifications by Visvanathan
to work at the Cassegrain focus of
the 100-inch and 60-inch telescopes
is under construction. The special
feature of this polarimeter is that it
eliminates the spurious polarization
introduced by the third (oblique)
mirror that is a part of the Casse-
grain system of these telescopes. This
was achieved by placing the analyzer
in front of the third mirror and fol-
lowing it by a Lyot-type quartz de-
polarizer that depolarizes the light
falling on the third mirror. The ana-
lyzer and depolarizer are rotated
around the axis of the telescope and
are indexed by remote control. A 5.5-
inch-diameter polaroid analyzer and
quartz depolarizer are used, followed
by a photoelectric photometer at the
focus.
A skylight-eliminator accessory for
the prime-focus spectrograph of the
200-inch telescope is being made to
designs by Greenstein and Rule. This
skylight and moonlight eliminator
will fit in place of the existing decker
plate to provide indexed setting and
automatic scanning at slit-length
ranges of 1, 2.5, and 5 mm, with
controlled single- or multiple-cycle
scanning rates from 3 to 125 minutes.
A new air-conditioned data room
in the dome of the 100-inch telescope,
providing approximately 250 square
feet of working area, has been con-
structed on the mezzanine floor west
of the coude observing room. It offers
controlled environment for the elec-
tronic data-acquisition systems and
working space for testing and as-
sembly of related components. A win-
dow projecting into the observing
floor has been provided for viewing
telescope operations.
As part of the project for mod-
ernization of the Mount Wilson 60-
;06
CARNEGIE INSTITUTION
inch telescope, the ground floor of
the dome has been cleared of extrane-
ous equipment and some substruc-
tures have been removed. The front
face of the coude constant-tempera-
ture room has been altered to provide
more working area for the projected
coude spectrograph. New footings
have been poured and steel pipe
columns erected for the new mezza-
nine floor structure. The mezzanine
floor will provide an air-conditioned
data room on the east, a larger clock
room, and a coude room. The difficult
work of modification, to be completed
later this year, is proceeding under
the supervision of Couch.
Solar Equatorial Telescope
It is expected that the solar equa-
torial telescope, funded by a grant
from the National Aeronautics and
Space Administration, will be ready
early in 1968. Figuring is proceeding
on the optics of the main off-axis
Cassegrain system. A Littrow spec-
trograph for the telescope has been
designed under contract by the firm
of Boiler and Chivens. The spectro-
graph, having numerous automatic
features, is to be housed in a rotating
tank. It is designed principally for
rapid medium-dispersion spectros-
copy. A grant has been received from
the National Science Foundation to
defray part of the cost of the labora-
tory building and tower to house the
new telescope. As a result of the sur-
vey carried out by Leighton, two sites
are now under consideration — Big
Bear Lake and Lake Elsinore.
Photographic Laboratory
Miller and Difley continued the
standard tests of all shipments of
photographic plates destined for use
at the telescopes on Mount Wilson and
Palomar Mountain, measuring rela-
tive speed and chemical fog as well
as checking for surface flaws. Tabu-
lated results of the tests are routinely
posted on both mountains for use by
the observers. The photographic lab-
oratory has carried on other projects
related to the perfection of a more
efficient developer formula, tests of
procedures for baking plates to im-
prove response for long exposures,
and testing of techniques for expos-
ing plates at low temperatures to
minimize reciprocity failure. An im-
proved, absolute system for testing
emulsion sensitivity is under de-
velopment.
SITE INVESTIGATION
Image Motion at Sites in Chile
One astronomical seeing monitor
(ASM) has been continued in opera-
tion on Cerro Morado (30°12'S) in
Chile under the direction of Irwin. A
second ASM has been operated by
Carnegie Institution personnel at the
European Southern Observatory on
Cerro La Silla (29°15'S) since Sep-
tember 1966. Analysis of all simul-
taneous (265) Morado-La Silla quar-
ter-night means on 78 nights give
median zenith seeing of 0"68 for Mo-
rado and 0-69 for La Silla. Excellent
seeing (<0"55) occurred 22% of the
time on Morado and 28 % of the time
on La Silla, according to the same
data, which did not include the winter
months of June, July, and August.
The two sites may be considered es-
sentially indistinguishable insofar as
seeing is concerned. Earlier lengthy
ASM comparisons between Morado-
Tololo and Morado-Pachon indicated
equality of the median seeing to
O'.'Ol, suggesting that this entire area
of Chile, at least from altitudes of
7100 to 8900 feet and close to the
main cordillera, has equally good see-
ing with a high percentage of nights
MOUNT WILSON AND PALOMAR OBSERVATORIES 307
that are truly excellent. Although nights of the 8-month period, Sep-
every experienced astronomer who tember 1966 to April 1967, inclusive,
has observed for significant lengths were considered to be useless.
of time on Tololo has verbally testi-
fied to the extraordinary quality of Microthermal Data
the seeing, it should be remembered It is generally conceded that "high-
that no definitive comparison of the altitude seeing" originating at 20,000-
ASM seeing has been made with any 40,000 feet in the atmosphere, is
other well-established scale of seeing, usually the main contributor to op-
^ The combined Morado-La Silla data tical image degradation in telescopic
give a median ASM zenith seeing of observing, but that "low-level seeing,"
0"73 for the 12-month period, with up to perhaps 100 feet above ground
excellent seeing (<"55) 19% of the level, may be a significant contributor
time. This compares with 0"79 and for some fraction of the time at good
20%, obtained in the previous 21/2- sites. On the reasonable assumption
year period. Almost all of the 1966- that low-level seeing is correlated
1967 excellent seeing occurred in the with microthermal fluctuations in the
months of January, February, March, air, Westphal, following earlier ex-
and April. The monthly data are periments by R. Lynds at Kitt Peak
given in Table 2. National Observatory, has developed
Tololo cloud data were kindly sup- field equipment for sensing and re-
plied by Dr. N. Sanduleak of the cording air-temperature fluctuations
Cerro Tololo Inter-American Observ- in the frequency range 5 Hz to 0.2
atory. For the year, 67% of the Hz. The fluctuations are sensed by
nights were photometric (completely small-bead thermistors supported at
free of clouds for 6 hours or more) four levels by 100-foot steel towers,
and 83% were useful nights. These and the mean power at each level is
figures may be compared with 67% recorded on a strip chart,
and 79%, respectively, for the 4-year To compare the microthermal ac-
period preceding. Only 18 of the 242 tivity at various locations as an aid
TABLE 2.
ASM Zenith Seeing
Image Motion, seconds of arc
Date
No. of
Nights
Median Zenith Seeing
1966-1967 1963-1966
1966
June 11 0"84 1"Q1
July 9 1"00 1"20
August 14 0"90 1"04
September 18 0"95 0"83
October 15 1"02 0"86
November 17 0"82 0"66
December 17 0"72 0"79
1967
January 21 0"57 0"63
February 25 0"64 0"65
March 19 0"76 0/.,56
April 22 0"60 0"95
May 13 0"81 0"72
0"73 0"79
Year 201
SOS
CARNEGIE INSTITUTION
to telescope site selection, Westphal
erected two mierothermal-sensing
towers on Palomar Mountain and
four on Cent) Morado. Reduction of
selected parts of the Palomar data
shows that the nighttime thermal
activity of the air is significantly
lower at the site chosen for the 60-
inch photometric telescope than on
the ridge west of the 200-inch. Com-
parison of several hundred samples
of half-hourly seeing estimates by
observers at the 200-inch telescope
with the thermal activity at the 95-
foot level on the nearby tower shows
a positive correlation of 0.45. On this
basis it is felt that the thermal
activity is a useful criterion for
choice among local sites where the
high-altitude seeing is likely to be
uniform.
Preliminary reduction of micro-
thermal records has been accom-
plished by Westphal and Irwin for
the four towers erected on Cerro
Morado. Comparison of results from
one tower with the optical data from
the adjacent ASM shows better cor-
relation with optical seeing than was
obtained at Palomar. The absolute
thermal activity at Morado is con-
sistently from 5 to 10 times lower
than that measured at Palomar.
A recording thermo-hygrograph
was installed on the south peak of
Papilones (6300 feet) in October
1966; this peak is 24 miles north-
northeast of La Serena. In April an
anemometer was installed on the
northwest ridge of Campanita (8100
feet), 16 miles north of La Silla. A
visit was made to Cerro Morado by
Babcock in September 1966, with
trips to La Silla, Cerro Papilones, and
Cerro Campanita. Donald L. Buck
was appointed Project Supervisor and
arrived in Chile in early June 1967.
The continued cooperation and as-
sistance of personnel of the Cerro
Tololo Inter-American Observatory,
of the European Southern Observa-
tory, and of the University of Chile
are gratefully acknowledged.
BIBLIOGRAPHY
Abt, Helmut A., Patrick S. Osmer, and
Robert P. Kraft, Atmospheric characteris-
tics of a Cepheid of unusually large light
amplitude: TV Camelopardalis. Astro-
phys. J., 145, 479-486, 1966.
Acton, Loren W., see Zirin, Harold.
Anderson, Jean H., see Luyten, Willem J.
Arp, Halton, Atlas of peculiar galaxies.
Astrophys. J., Suppl. Ser., 14, No. 123,
1-20, 1900.
Arp, Halton, Peculiar galaxies and radio
sources. Astrophys. J., 148, 321-365, 1967.
Arp, Halton, J. G. Bolton, and T. D. Kin-
man, A quasi-stellar object with a red-
shift of 2.22 and an unusual absorption
spectrum. Ibid., 147, 840-845, 1967.
Babcock, Horace W., The Zeeman effect in
astrophysics. Physica, 33, 102-121, 1967.
Babcock, Horace W., Optical astronomy in
perspective. Science, 156, 1317-1322, 1967.
Bahcall, John N., and Wallace L. W. Sar-
gent, Predicted emission spectra of quasi-
stellar sources in the far ultraviolet.
Astrophys. J., 148, Pt. 2, L65-L68, 1967.
Bahcall, John N., Bruce A. Peterson, and
Maarten Schmidt, On the absorption
spectrum of 1116+12. Ibid., 145, 369-371,
1966.
Barbon, R., C. E. Kearns, and K. Rudnicki,
Photographs of Comet Barbon (1966c)
and Comet Rudnicki (1966e). Publ.
Astron. Soc. Pacific, 79, 298-270, 1967.
Baum, William A., Magnetic focusing of
image tubes, in Advances in Electronics
and Electron Physics, Vol. 22A, Photo-
Electronic Image Devices, Third Sym-
posium, pp. 617-628, J. D. McGee, D.
McMullan, and E. Kahan, eds. Academic
Press, London and New York, 1966.
Baum, William A., Multicolor photometry of
spiral arms, in Intern. Astron. Union
Symp. No. 24, Spectral Classification and
Multicolor Photometry, pp. 288-290, K.
Loden, L. O. Loden, and U. Sinnerstad,
eds. Academic Press, London and New
York, 1966.
MOUNT WILSON AND PALOMAR OBSERVATORIES
309
Baum, William A., see also McGee, J. D.
Becklin E., see Ulrich, B. T.
Bolton, J. G., see Arp, Halton; Kinman,
T. D.
Bowen, Ira S., Statement of aims and lim-
itations of program, Intern. Astron.
Union Symp. No. 27, Construction of
Large Telescopes, pp. 4-7, D. L. Craw-
ford, ed. Academic Press, London and
New York, 1966.
Bowen, Ira S., Control of thermal effects.
Ibid., pp. 60-65.
Bowen, Ira S., Optimum thermal effects
for large domes. Ibid., pp. 170-174.
Bowen, Ira S., Future tools of the astrono-
mer. Quart. J. Roy. Astron. Soc, 8, No.
1, 9-22, 1967.
Bowen, Ira S., and Bruce H. Rule, Palomar
60-inch photometric reflector. Sky and
Telescope, 32, 185-187, 1966.
Bumba, V., Robert Howard, and Sara F.
Smith A note on the distribution of
magnetic fields in the solar atmosphere, in
Proceedings of the International Sympo-
sium on Solar Magnetic Fields, pp. 3-6,
G. Barbera, Florence, 1966.
Clarke, R. W., see Kinman, T. D.
Conti, Peter S., and I. J. Danziger, The
abundances of lithium and beryllium in
some F dwarfs and K giants. Astrophys.
J., U6, 383-391, 1966.
Conti, Peter S., and Armin J. Deutsch, Color
anomalies and metal deficiencies in solar-
type disk-population stars. Ibid., 14-5, 742-
745, 1966.
Conti, Peter S., and Armin J. Deutsch,
Color anomalies and metal deficiencies in
solar-type disk-population stars, II. Ibid.,
U7, 368-369, 1967.
Conti, Peter S., Jesse L. Greenstein, Hyron
Spinrad, George Wallerstein, and M. S.
Vardya, Neutral oxygen in late-type
stars. Ibid., 148, 105-128, 1967.
Conti, Peter S., see also Danziger, I. J.
Danziger, I. J., and Peter S. Conti, Lithium
observations in the Pleiades and in F
stars in the field. Astrophys. J., 146, 392-
398, 1966.
Danziger, I. J., and L. V. Kuhi, The vari-
ability of Rho Puppis. Ibid., 146, 743-747,
1966.
Danziger, I. J., and J. B. Oke, Spectropho-
tometry of VZ Cancri. Ibid., 147, 151-157,
1967.
Danziger, I. J., see also Conti, Peter S.
Dennison, E. W., A microphotometer for
use with photographic and electrono-
graphic image devices, in Advances in
Electronics and Electron Physics, Vol.
23, pp. 435-440, L. Marton, ed. Academic
Press, London and New York, 1967.
Dennison, E. W., Data systems, in Intern.
Astron. Union Symp. No. 27, Construction
of Large Telescopes, pp. 154-157, D. L.
Crawford, ed. Academic Press, London
and New York, 1966.
Deutsch, Armin J., Even simpler methods of
abundance determination from stellar
spectra, in Intern. Astron. Union Symp.
No. 26, Abundance Determinations in
Stellar Spectra, pp. 112-117, H. Hubenet,
ed. Academic Press, London and New
York, 1966.
Deutsch, Armin J., Interpretation of
anomalous line strengths in Ap and Am
stars. Ibid., pp. 268-271.
Deutsch, Armin J., Rapid rotation of the
solar interior. Science, 156, 236-237, 1967.
Deutsch, Armin J., see also Conti, Peter S.
Dickens, R. J., HD 24550: A new 8 Scuti
variable. Astrophys. J., 148, Pt. 2, L33-
L36, 1967.
Eggen, Olin J., Contact binaries, II. Mem.
Roy. Astron. Soc, 70, 111-164, 1966.
Eggen, Olin J., and Allan Sandage, Ex-
amination of the evidence for the exist-
ence of pygmy stars. Astrophys. J., 11+8,
911-917, 1967.
Eggen, Olin J., and Allan Sandage, Reply
to Zwicky's comments. Ibid., 921-922.
Faulkner, John, Olin C. Wilson, and Icko
Iben, Age-helium content correlations for
two low-mass field binary systems. Astro-
phys. J., 14.6, 271-274, 1966.
Faulkner, John, see also Gunn, James E.
Fernie, J. D., W. A. Hiltner, and Robert P.
Kraft, The association II Pup and the
classical Cepheid AQ Pup. Astron. J., 71,
999-1002, 1966.
Ganson, A., see McGee, J. D.
Giaconni, R., P. Gorenstein, H. Gursky,
P. D. Usher, J. R. Waters, Allan Sandage,
P. Osmer, and J. Peach, On the optical
identification of the X-ray sources Cyg
X-l and Cyg X-2. Astrophys. J., 148,
L129-L132, 1967.
Gorenstein, P., see Giaconni, R.
Greenstein, Jesse L., Astronomy, Americana
Annual, pp. 90-95, Americana Corpora-
tion, New York, 1967.
Greenstein, Jesse L., Radio astronomy
beyond the solar system, in Americana
Sci. Suppl. 1965, p. 303, Americana Cor-
poration, New York, 1966.
10
CARNEGIE INSTITUTION
Greenstein. Jesse L.. The comparison of
observation and prediction in nucleosyn-
thesis. Intern. Astron. Union Symp. No.
26. Abundance Determinations in Stellar
- Bctra, pp. 348-357, H. Hubenet, ed.
Academic Press. London and New York,
1966.
Greenstein. Jesse L., Science and the de-
struction of value. J. Blaisdel Inst., 2,
5-18, 1967.
Greenstein. Jesse L., Man in the universe.
Medical Opinion and Rev.. S, 78-87, 1967.
Greenstein. Jesse L., Observing1 the universe.
Science J. (England), 2, 71-76, 1966.
Greenstein. Jesse L., The question of the
quasars, in Science Year 1966, World
Book Sci. Ann., pp. 76-89, Field Enter-
prises Educational Corp., Chicago, 1966.
Greenstein, Jesse L., and Guido Munch, The
weakness of helium lines in globular-
cluster and halo B stars. Astrophys. J.,
U6, 618-620, 1966.
Greenstein, Jesse L., and Maarten Schmidt,
The two absorption-line redshifts in
Parkes 0237-23. Ibid, 148, Pt. 2, L13-L15,
1967.
Greenstein, Jesse L., see also Conti, Peter S.
Gunn, James E., A fundamental limitation
on the accuracy of angular measurements
in observational cosmology. Astrophys. J.,
U7, 61-72, 1967.
Gunn, James E., Bruce A. Peterson, and
John Faulkner, Expected number of blue-
shifts and redshifts of ejected sources.
Nature, 211, 502, 1966.
Gursky, H., see Giaconni, R.
Hill, Byron, Construction of initial site
facilities, in Intern. Astron. Union Symp.
No. 27, Construction of Large Telescopes,
pp. 193-196, D. L. Crawford, ed. Academic
Press, London and New York, 1966.
Hiltner, W. A., see Fernie, J. D.
Howard, Robert, Magnetic fields and the
solar cycle. Astron. Soc. Pacific Leaflet
No. 454, 8 pp., April 1967.
Howard, Robert, On the measurement of
small-scale fields on the sun. Observatory,
86, 160-162, 1966.
Howard, Robert, see also Bumba, V.; Zirin,
Harold.
Hughes, E. E., see Ulrich, B. T.
Iben, Icko, see Faulkner, John.
Irwin, John B., Clearer views for astron-
omy: another view. Science, 154, 1275-
1276, 1966.
Joy, Alfred H., Stellar flares. Astron. Soc.
Pacific Leaflet No. 456, 8 pp., June 1967.
Karpowicz, M., On the non-existence of
clusters of clusters of galaxies. II, Z.
Astrophys., 66, 301-307, 1967.
Karpowicz, M., sec Zwicky, Fritz.
Kayser, Susan E., Photometry of the near-
by irregular galaxy, NGC 6822. Astron.
J., 72, 134-148, 1967.
Kearns, C. E., and K. Rudnicki, Twenty-
eight photographs of Comet Humason
(1961e). Z. Astrophys., 64, 337-361, 1966.
Kearns, C. E., see also Barbon, R.
Khogali, A., see McGee, J. D.
Kinman, T. D., J. G. Bolton, R. W. Clarke,
and Allan Sandage, Radio and optical
data on sixteen quasi-stellar objects.
Astrophys. J., 147, 848-850, 1967.
Kinman, T. D., see also Arp, Halton.
Kraft, Robert P., Studies of stellar rotation,
IV, A comparison of rotational velocities
in the a Persei cluster and the Pleiades.
Astrophys. J., 148, 129-140, 1967.
Kraft, Robert P., The line spectrum of
pulsating variable stars, in Intern.
Astron. Union Symp. No. 28, Aerodynamic
Phenomena in Stellar Atmospheres, pp.
207-238, Richard N. Thomas, ed. Aca-
demic Press, London and New York, 1967.
Kraft, Robert P., see also Abt, Helmut A.;
Fernie, J. D.
Krzeminski, W., Ton S 120: A new very
short-period eclipsing binary. Commission
27, Intern. Astron. Union Inform. Bull.
No. 160, October 6, 1966.
Krzeminski, W., and K. Serkowski, Photo-
metric and polarimetric observations of
the nearby strongly reddened open cluster
Stock 2. Astrophys. J., 147, 988-1002,
1967.
Kuhi, L. V., see Danziger, I. J.
Lasker, Barry M., A liquid-filter cell for
telescope use. Publ. Astron. Soc. Pacific,
78, 329-332, 1966.
Leighton, R. B., and B. C. Murray, The
behavior of carbon dioxide and other
volatiles on Mars. Science, 153, 136-144,
1966.
Leighton, R. B., see also Ulrich, B. T.
Lungershausen, W. T., see Searle, Leonard.
Luyten, Willem J., and Allan Sandage, On
the nature of faint blue objects in high
galactic latitudes, I, Photometry and
proper motions in PHL Field 1:36 + 6°
and Richter Field M3, II. Astrophys. J.,
148, 767-780, 1967.
Luyten, Willem J., and Allan R. Sandage,
Eight hundred faint blue stars in a field
centered at 8:48+8°. Search for Faint
MOUNT WILSON AND PALOMAR OBSERVATORIES
311
Blue Stars, A, Pt. XL, 4 pp., published
by The Observatory, University of
Minnesota, Minneapolis, Minn., 1966.
Luyten, Willem J., Jean H. Anderson, and
Allan Sandage, Faint blue stars in a
field centered at 15:10+24°. Ibid., Pt.
XLIII, 4 pp., 1967.
Luyten, Willem J., Jean H. Anderson, and
Allan Sandage, Faint blue stars in a
field centered at 11:16+30°. Ibid., Pt.
XLVI, 4 pp., 1967.
McCammon, D., G. Munch, and G. Neuge-
bauer, Infrared spectra of low-tempera-
ture stars. Astrophys. J., 147, 575-586,
1967.
McCammon, D., see also Ulrich, B. T.
McGee, J. D., A. Khogali, A. Ganson, and
W. A. Baum, The spectracon — an elec-
tronographic image-recording tube, in Ad-
vances in Electronics and Electron Phys-
ics, Vol. 22A, Photo-Electronic Image
Devices, Third Symp., pp. 11-30, J. D.
McGee, D. McMullan, and E. Kahan, eds.
Academic Press, London and New York,
1967.
Munch, Guido, Infrared lines of the solar
corona, I, Prediction. Astrophys. J., 145,
237-241, 1966.
Munch Guido, Absorption lines from inter-
mediate-latitude interstellar complexes.
Publ. Astron. Soc. Pacific, 78, 305-311,
1966.
Munch, Guido, see also Greenstein, Jesse L.;
McCammon, D.
Murray, B. C, see Leighton, R. B.
Neugebauer, G., see McCammon, D.; Ulrich,
B. T.
Oke, J. B., A spectrophotometric study of
X Arietis. Astrophys. J., 145, 468-478,
1966.
Oke, J. B., Photoelectric spectrophotometry
of quasi-stellar sources. Ibid., 668-683.
Oke, J. B., Spectrophotometric observations
of rapid variability in 3C 279 and 3C 446.
Ibid., 147, 901-907, 1967.
Oke, J. B., The use of low-resolution and
moderate-resolution photoelectric spectro-
photometry for the analysis of stellar
spectra, in Intern. Astron. Union Symp.
No. 26, Abundance Determinations in
Stellar Spectra, pp. 3-14, H. Hubenet, ed.
Academic Press, London and New York,
1966.
Oke, J. B., The continuous spectrum of
pulsating variable stars, in Intern.
Astron. Union Symp. No. 28, Aerody-
namic Phenomena in Stellar Atmospheres,
pp. 177-198, R. N. Thomas, ed. Academic
Press, London and New York, 1967.
Oke, J. B., see also Danziger, I. J.; Wamp-
ler, E. J.
Osmer, Patrick S., see Abt, Helmut A.;
Giacconi, R. ; Sandage, Allan.
Peach, J., see Giaconni, R.
Peterson, Bruce, A., The spectrum varia-
tions of HD 124224 and C6 Arietis. Astro-
phys. J., 145, 735-741, 1966.
Peterson, Bruce A., see also Bahcall, John
N. ; Gunn, James E.
Rudnicki, K., SN 1965n (=Z 175) a super-
nova of type I in NGC 3074. Astron.
Nachr., 289, 247-249, 1967.
Rudnicki, K., General features of clusters
of galaxies. Astron. Zh. (USSR), 44,
77-81, 1967.
Rudnicki, K., and F. Zwicky, A supernova
on an intergalactic bridge. Astron. J., 72,
407-409, 1967.
Rudnicki, K., see also Barbon, R.; Kearns,
C. E.; Zwicky, Fritz.
Rule, Bruce H., Possible flexible mirror sup-
ports and collimation servo-control, in
Intern. Astron. Union Symp. No. 27, Con-
struction of Large Telescopes, pp. 71-75,
D. L. Crawford, ed. Academic Press,
London and New York, 1966.
Rule, Bruce H., Telescope design, General
tolerance requirements: c. Mechanical.
Ibid., pp. 101-106.
Rule, Bruce H., Types of mountings. Ibid.,
106-123.
Rule, Bruce H., Development of engineering
and fabrication techniques. Ibid., pp. 123-
125.
Rule, Bruce H., Critical architectural and
engineering requirements (for large
domes) . Ibid., pp. 175-179.
Rule, Bruce H., Large telescope engineering
operations. Ibid., pp. 200-209.
Rule, Bruce H., Problems of organization.
Ibid., pp. 213-218.
Rule, Bruce H., and G. M. Sisson, Types of
mounting. Ibid., pp. 106-123.
Rule, Bruce H., see also Bowen, Ira S.
Sandage, Allan, Cosmology: The "big bang"
theory. America Illustrated (in Russian),
No. 120, 21-22, 1966.
Sandage, Allan, Redshifts of nine radio
galaxies including the abnormal system
3C 305. Astrophys. J., 145, 1-5, 1966.
Sandage, Allan, The correlation of colors
with redshifts for QSS leading to a
smoothed mean energy distribution and
new values for the i£-correction. Ibid.,
146, 13-24, 1966.
Sandage, Allan, Astrophotography, Encycl.
312
CARNEGIE INSTITUTION
Americana, Vol. 2. pp. 5S6-5S7, Ameri-
cana Corporation, New York, 1967.
Sandage, Allan. QSS 3C 440 twenty-fold
increase in optical brightness. Intern.
Astron. Union Circular Xo. 1961, July 27,
1966.
Sandage, Allan, Observational properties of
radio galaxies and quasi-stellar sources,
in Proceedings Intern. SeJiool of PJiysics,
Enrico Fermi. Course XXXV, High
Energy Astrophysics, pp. 10-42, L. Grat-
ton. ed. Academic Press, London and
New York, 1966.
Sandage. Allan, A time for decision.
Scientific Research, 2, 46, 1967.
Sandage, Allan, P. Osmer, and ten others,
On the optical identfication of Sco X-l.
Astrophys, J., 146, 316-322, 1966.
Sandage, Allan, J. A. Westphal, and P. A.
Strittmatter, The change of intensity,
color, line strength, and line position in
the quasar 3C 446 during the 1966 out-
burst. Ibid., 322-326.
Sandage, Allan, see also Eggen, Olin J.;
Giaconni, R.; Kinman, T. D.; Luyten,
Willem J.
Sargent, W. L. W., The spectra of blue
horizontal-branch stars in three northern
globular clusters. Astrophys. J., 148, Pt. 2,
L129-L132, 1967.
Sargent, W. L. W., see also Bahcall, John
X.; Searle, Leonard.
Schadee, Aert, On the rotational tempera-
ture of a sunspot. Astrophys. J., 14.5, 348-
349, 1966.
Schadee, Aert, The solar X8668 line inter-
preted as CN. Ibid., 147, 363-364, 1967.
Schadee, Aert, The relation between the
electronic oscillator strength and the
wavelength for diatomic molecules. J.
Quant. Spectrosc. Radiat. Transfer, 7,
169-183, 1967.
Schmidt, Maarten, Lifetimes of extragalac-
tic radio sources. Astrophys. J., 146, 7-12,
1966.
Schmidt, Maarten, The gravitational field of
the Galaxy, in Intern. Astron. Union
Symp. No. 25, Theory of Orbits in the
Solar System and in Stellar Systems, pp.
61-64, George Contopoulos, ed. Academic
Press, London and New York, 1966.
Schmidt, Maarten, Quasi-stellar objects.
Science J. (England), 2, 77-83, 1966.
Schmidt, Maarten, Spectroscopic observa-
tions of extragalactic radio sources, in
Solvay Conference on the Structure and
Evolution of Galaxies, Thirteenth, pp.
130-134, Interscience Publishers, New
York, 1965.
Schmidt, Maarten, see also Bahcall, John
N.; Greenstein, Jesse L.
Searle, Leonard, W. T. Lungershausen, and
W. L. W. Sargent, Studies of the peculiar
A stars, IV, The relative abundances of
four iron-peak elements. Astrophys. J.,
145, 141-152, 1966.
Serkowski, K., see Krzeminski, W.
Sisson, G. M., see Rule, Bruce H.
Skumanich, A., see Wilson, Olin C.
Smith, Sara F., see Bumba, V.
Spinrad, Hyron, see Conti, Peter S.
Strittmatter, J. A., see Sandage, Allan.
Takakubo, K., and Hugo van Woerden,
Neutral hydrogen at intermediate galactic
latitudes, II, Gaussian components of the
survey profiles. Bull. Astron. Inst. Nether-
lands, 18, 488-533, 1966.
Title, Alan, Selected Spectroheliograms, 19
pp. + 50 plates, published by the Cali-
fornia Institute of Technology, Pasadena,
1966.
Trafton, L. M., The pressure-induced mono-
chromatic translational absorption co-
efficients for homopolar and nonpolar
gases and gas mixtures with particular
application to H2. Astrophys. J., 146, 558-
571, 1966.
Trafton, L. M., Model atmospheres of the
major planets. Ibid., 147, 765-781, 1967.
Ulrich, B. T., G. Neugebauer, D. McCam-
mon, R. B. Leighton, E. E. Hughes, and
E. Becklin, Further observations of ex-
tremely cool stars. Astrophys. J., 146,
288-289, 1966.
Usher, P. D., see Giaconni, R.
Utter, Merwyn G., The heavens in 1967.
Astron Soc. Pacific, Ann. Ser., 8 pp.,
January 1967.
Vardya, M. S., see Conti, Peter S.
Wallerstein, George, see Abt. Helmut A.;
Conti, Peter S.
Wampler, E. J., and J. B. Oke, The emis-
sion-line spectrum of 3C 273. Astrophys.
J., 148, 695-705, 1967.
Waters, J. R., see Giaconni, R.
Werner, Susan, see Zirin, Harold.
Westphal, J. A., see Sandage, Allan.
Wildey, Robert L., Far-infrared stellar
astronomy, in Intern. Astron. Union
Symp. No. 24, Spectral Classification and
Multicolour Photometry, pp. 267-273, K.
Loden, L. O. Loden, and U. Sinnerstad,
eds. Academic Press, London and New
York, 1966.
MOUNT WILSON AND PALOMAR OBSERVATORIES
010
0L0
Wilson, Olin C, The masses of nine visual
binary systems, in Modern Astrophysics,
pp. 241-249, M. Hack, ed. Gauthier-
Villars, Paris, 1966.
Wilson, Olin C, The distance of the Hyades
and the correlation between luminosity
and Ca II emissionline widths. Publ.
Astron. Soc. Pacific, 79, 46-56, 1967.
Wilson, Olin C, and A. Skumanich, Chro-
mospheric activity as a function of age
in main-sequence stars, in Intern. Astron.
Union Symp. No. 2U, Spectral Classifica-
tion and Multicolour Photometry, pp. 40-
43, K. Loden, L. 0. Loden and U. Sinner-
stad, eds. Academic Press, London and
New York, 1966.
Wilson, Olin C, see also Faulkner, John.
Woerden, Hugo van, see Takakubo, K.
Zirin, Harold, Production of fast drift
bursts. Nature, 212, 909-910, 1966.
Zirin, Harold, Solar Atmosphere, The, 502
pp., Blaisdell Publishing Co., Waltham,
Mass., 1966.
Zirin, Harold, and Loren W. Acton,
Spectrographic observations of the flare
of July 20, 1961. Astrophys. J., U8, 501-
509, 1967.
Zirin, Harold, and Robert Howard, The
structure of the solar chromosphere, II,
Spectroheliograms in X10830A and their
interpretation. Ibid., U6, 367-371, 1966.
Zirin, Harold, and Susan Werner, Detailed
analysis of flares, magnetic fields, and
activity in the sunspot group of Sept.
13-26, 1963. Solar Physics, 1, 66-100,
1967.
Zwicky, Fritz, Pygmy stars at issue. Astro-
phys. J., U8, 919-920, 1967.
Zwicky, Fritz, Die Methode der systemat-
schen Feldueberdeckung I. Bild der Wis-
senschaft, 4, 74-79, 1967.
Zwicky, Fritz, Die Methode der sys-
tematischen Fedlueberdeckung II. Ibid.,
156-159.
Zwicky, Fritz, Deplacements gravitationnels
vers le rouge dus a des corps cosmiques
compacts. C. R. Acad. Sci. Paris, 262,
1566-1569, 1966.
Zwicky, Fritz, Objektivprismen und Ob-
jektivgitter. Die Sterne, US, 89-92, 1967.
Zwicky, Fritz, Systematische Entdeckung
von Kosmischen Objekten. Ibid., 105-112.
Zwicky, Fritz, Entdecken, erfinden, forschen
im morphologischen weltbild, 268 pp., Th.
Knaur, Miinchen, 1966.
Zwicky, Fritz, List of compact galaxies and
compact parts of galaxies, eruptive gal-
axies and post-eruptive galaxies, Fourth
(June 1966) and Fifth (March 1967),
California Institute of Technology, Pasa-
dena, Calif.
Zwicky, Fritz, The 1965 Palomar supernova
search. Publ. Astron. Soc. Pacific, 78, 471-
473, 1966.
Zwicky, Fritz, Pygmy stars: first pair.
Science, 153, 53-54, 1966.
Zwicky, Fritz, and E. Herzog, Catalogue of
Galaxies and of Clusters of Galaxies, Vol.
III, California Institute of Technology,
Pasadena, Calif., 1966.
Zwicky, Fritz, and Maria Karpowicz, Area
of the sky covered by clusters of galaxies,
IV. Astrophys. J., 1U6, 43-50, 1966.
Zwicky, Fritz, and W. J. Luyten, Ultrafaint
blue stars and compact objects observed
with the 200-inch telescope. Search for
Faint Blue Stars, A, Pt. XLV, 4 pp.,
published by The Observatory, University
of Minnesota, Minneapolis, Minn., 1966.
Zwicky, Fritz, and K. Rudnicki, Zur
Nichtexistenz von Haufen von Galaxien-
haufen. Z. Astrophys., 6 A, 246-265, 1966.
Zwicky, Fritz, see also Rudnicki, K.
STAFF AND ORGANIZATION
Dr. Ira S. Bowen, former director of
the Observatories, was designated Dis-
tinguished Service Member of the Car-
negie Institution. He has continued to
devote much time to problems of optical
design, especially to perfection of the
new low -dispersion image-tube spectro-
graph for the Hale telescope.
Dr. Wallace L. W. Sargent, formerly
a Research Fellow and more recently on
the faculty of the University of Califor-
nia, San Diego, joined the staff of the
Observatories on September 16, 1966.
Dr. Arthur H. Vaughan, Jr., formerly
a Carnegie Fellow and later Staff Asso-
ciate, became a Staff Member of the
Observatories. In addition to conducting
research, he has assumed responsibility
for certain problems of Observatory in-
strumentation, particularly in connection
with the modernization program for the
Mount Wilson 60-inch telescope.
Research Division
Distinguished Service Member, Carnegie
Institution
Ira S. Bowen
314
CARNEGIE INSTITUTION
Staff Members
HaltoD C. Arp
Horace W. Babcock, Director
Edwin W. Dennison
Armin J. Deutsch
Jesse L. Greenstein1
Robert F. Howard
Robert P. Kraft
Robert B. Leighton2
Guido Munch3
J. Beverley Oke3
Bruce H. Rule, Chief Engineer
Allan R. Sandage
Wallace L. W. Sargent4
Maarten Schmidt3
Arthur H. Yaughan, Jr.
Olin C. Wilson
Harold Zirin5
Fritz Zwicky5
Staff Member Engaged in Post Retire-
ment Studies
Alfred H. Joy
Staff Associates
John B. Irwin6
Bruce C. Murray7
Gerry Neugebauer8
James A. Westphal9
Senior Research Felloivs
David Koelbloed10
Jun Jugaku
Konrad Rudnicki6
Carnegie Fellows
Robert J. Dickens
Wojciech Krzeminski
John V. Peach
David M. Rust
Natarajan Visvanathan
National Science Foundation Fellow
Barry M. Lasker11
NATO Fellow
Alessandro Braccesi 12
Research Fellows
Ann Merchant Boesgaard
Peter S. Conti10
Ivan J. Danziger10
J. W. R. Heintze
Dora Russo Lackner
Mitsugu Makita
Aert Schadee
Rudolf E. Schild
Robert Stein
Stephen E. Strom13
Henrietta H. Swope
Takashi Tsuji
Senior Research Assistant
Dorothy D. Locanthi
Research Assistants
Frank J. Brueckel
Sylvia Burd
Thomas A. Cragg
Robert F. Garrison
Howard Gates
Emil Herzog
Joyce Humphreys14
Basil Katem
Margaret Katz
Charles T. Kowal
Duk Hee Lee15
A. Louise Lowen
Charles Petit16
Kathleen Reynolds
Malcolm S. Riley
Philip Rust16
Merwyn G. Utter
Grace D. Vess
Student Observers
Saul J. Adelman
Christopher M. Anderson
Kurt S. Anderson
Eric Becklin
John Castor
Solomon Giles
Alexander F. H. Goetz
Thomas B. McCord
Andrew Mackay
Patrick S. Osmer
Bruce A. Peterson
Jeffrey D. Scargle
Virginia L. Trimble
Donna E. Weistrop
Photographic Laboratory
William C. Miller, Photographer
John A. Difley, Photographic Techni-
cian
Librarian
Marline Gerrity
Instrument Design and Construction
Lawrence E. Blakee, Senior Electronic
Technician
Maynard K. Clark, Senior Electronic En-
gineering Assistant
Floyd E. Day, Head Optician
Madeleine Dolley, Draftsman
Raymond Dreiling, Machinist
Loyal Elam, Jr., Draftsman
Robert D. Georgen, Machinist
Fred Idzinga, Electronics Specialist
MOUNT WILSON AND PALOMAR OBSERVATORIES
315
Melvin W. Johnson, Optician
Frederick O'Neill, Machinist
Michael Morrill, Technical Aide
Rudolf E. Ribbens, Designer and Super-
intendent of Instrument Shop
Howard Sachs, Engineer
S. Robert Salow, Senior Electronics En-
gineer17
Benny W. Smith, Electronics Specialist
Robert G. Stiles, Optician
David Thompson, Technical Assistant
Virgal Z. Vaughan, Electronics Specialist
Ralph W. Wilson, Machinist
Felice Woodworth, Draftsman-Illustrator
Clare Worden, Draftsman
Maintenance and Operation
Mount Wilson Observatory and Offices
Wilma J. Berkebile, Secretary
Fern V. Borgen, Stenographer and
Receptionist
Hugh T. Couch, Superintendent of
Construction
Helen S. Czaplicki, Typist-Editor
Fanny G. Gabrielson, Stewardess16
Eugene L. Hancock, Night Assistant
Mark D. Henderson, Custodian
Frank Hernandez, Chauffeur6
Anne Hopper, Accountant
Rienaldo Jacques, Night Assistant
Doris Jeffery, Stewardess18
Sharon McDonell, Accounting Clerk19
Philip J. McManus, Jr., Chauffeur18
Ethel Marzalek, Stewardess
James R. Mosier, Assistant Superin-
tendent6
Bula H. Nation, Head Stewardess6
Alfred H. Olmstead, Laborer
Glen Sanger, Chauffeur
William D. St. John, Custodian
Henry F. Schaefer, Night Assistant
Elizabeth M. Shuey, Secretary
Benjamin B. Traxler, Mountain Super-
intendent
Frederick P. Woodson, Assistant to
Director
Palomar Observatory and Robinson
Laboratory
Fred Anderson, Machinist
Ray L. Ballard, Administrative As-
sistant
Jan A. Bruinsma, Custodian
Maria J. Bruinsma, Stewardess
Eleanor G. Ellison, Secretary and
Librarian
Beulah Greenlee, Stewardess
Frank K. Greenlee, Custodian
Lilo Hauck, Secretary
Victor A. Hett, Night Assistant
Byron Hill, Mountain Superintendent10
Helen D. Holloway, Secretary
John D. Jordan, Night Assistant20
Charles E. Kearns, Assistant Superin-
tendent
J. Luz Lara, Mechanic
Patricia Lynch, Clerk-Typist21
Mildred Newton, Department Clerk22
Dennis Palm, Night Assistant
Marilynne Rice, Secretary
Carol Russell, Secretary
Barrett A. Staples, Mechanic
Gary M. Tuton, Night Assistant
Hendrika E. van Buuren, Stewardess18
John E. van Buuren, Custodian18
William C. Van Hook, Mountain Super-
intendent
Betty A. Wallace, Secretary
Site-Testing Operations, Chile
Donald L. Buck, Project Supervisor
Manuel Casanova, Assistant Observer
Rolando H. Cortez, Assistant Ob-
server23
Herman Rojas, Assistant Observer
Thomas Veliz, Assistant Observer
Manfred Wagner, Observer
1 Professor of Astrophysics and Executive
Officer for Astronomy, California In-
stitute of Technology.
2 Professor of Physics, California Institute
of Technology.
3 Professor of Astronomy, California In-
stitute of Technology.
4 Assistant Professor of Astronomy, Cali-
fornia Institute of Technology.
5 Professor of Astrophysics, California In-
stitute of Technology.
6 Resigned June 30, 1967.
7 Associate Professor of Planetary Science,
California Institute of Technology.
8 Associate Professor of Physics, California
Institute of Technology.
9 Senior Research Fellow in Planetary
Science, California Institute of Tech-
nology.
10 Resigned August 31, 1966.
11 Resigned January 17, 1967.
12 Resigned April 11, 1967.
13 Resigned February 15, 1967.
14 Resigned May 19, 1967.
16
CARNEGIE INSTITUTION
. \ 1906.
l* $ • help.
Resig] :. \ mbex 14. 1966.
: March 31. L9<
« Resigned June 21, 1967.
-1 Resigned January 13, 1967.
-'-* Resigned February 4. 1967.
23 Resigned February 10, 1967.
Geophysical Laboratory
Washington, District of Columbia
Philip H. Abelson
Director
Carnegie Institution Year Book 66, 1966-1967
Contents
Introduction 321
Experimental and Descriptive Min-
eralogy 327
Pyroxenes 327
Quantitative electron-probe
analysis of pyroxenes
(Boyd) 327
Synthesis and stability of ferri-
diopside (Huckenholz,
Schairer, and Yoder) . . 335
Experimental studies on inver-
sion relations in natural
pigeonitic pyroxenes
(Brown) 347
Electron-probe study of exsolu-
tion in pyroxenes (Boyd
and Brown) 353
Hedenbergitess-wollastonitess in-
version in a Skaergaard
pyroxene (Lindsley, Brown,
and Muir) 359
Subsolidus relationships in part
of the hedenbergite-ferro-
silite join at low pressures
(Lindsley and Munoz) . . 363
Mossbauer study of synthetic
Ca-Fe clinopyroxenes (Dun-
don and Lindsley) . . . 366
Effect of shearing on enstatite
polymorphism (Munoz) . 369
High-pressure stability rela-
tions of spodumene (Munoz) 370
X-ray properties of jadeite-
acmite pyroxenes (Gilbert) 374
Other silicates 376
Yoderite : Synthesis, stability,
and interpretation of its nat-
ural occurrence (Schreyer
and Yoder) 376
A reconnaissance study of the
system MgO-Al203-Si02-
H20 at pressures between
10 and 25 kb (Schreyer) . 380
The aluminum silicates (Rich-
ardson, Bell, and Gilbert) . 392
The composition of synthetic
Fe-staurolite (Richardson) 397
The stability of Fe-staurolite +
quartz (Richardson) . . 398
The stability of fayalite (Wones
and Gilbert) 402
Sulfides and related minerals . . 403
High-temperature phase rela-
tions in the Cu-Fe-S sys-
tem (Kullerud) . . . .404
High-temperature phase rela-
tions in the Cu-Ni-S system
(Kullerud and Moh) . . 409
The Cu-Fe-Ni-S system (Craig
and Kullerud) .... 413
Minimum melting of nickelifer-
ous pyrrhotite ores (Craig
and Naldrett) . . . .417
Melting relations over a portion
of the Fe-S-0 system and
their bearing on the tem-
perature of crystallization
of natural sulfide-oxide liq-
uids (Naldrett) .... 419
Reaction between pyrrhotite and
enstatite-ferrosilite solid
solutions (Naldrett and
Brown) 427
Effect of water on the melting
of pyrrhotite-magnetite as-
semblages (Naldrett and
Richardson) 429
Succession of mineral assem-
blages in pyrrhotite-rich
Ni-Cu ores (Craig, Nal-
drett, and Kullerud) . . 431
The Fe-Ni-S system ... 434
Violarite stability relations
(Craig) 434
Partial pressure of sulfur in
the vapor coexisting with
the Fe^S-Ni^S solid so-
lution at 600° and 400 °C
(Naldrett and Craig) . 436
400 °C isothermal diagram
(Craig, Naldrett, and
Kullerud) 441
High-pressure differential ther-
mal analysis of CuS (Kul-
lerud, Bell, and England) . 442
Sulfide-silicate relations (Kul-
lerud and Yoder) . . . 442
Inclusions in diamonds .... 446
Mineral inclusions in diamonds
(Meyer) 446
319
Igneous Petrology 450
Melting relations of igneous rock
series (Tilley, Yoder, and
Schairer) 450
The pyroxenite facies conversion
of volcanic and subvolcanic,
melilite-bearing and other
alkali ultramafic assemblages
(Tilley and Yoder) ... 457
Melting relations of some calcal-
kaline volcanic rocks (Brown
and Schairer) 460
The join nepheline-diopside-anor-
thite and its relation to alkali
basalt fractionation (Schair-
er, Tilley, and Brown) . . 467
Akermanite and related melilite-
bearing assemblages (Yoder) 471
Albite-anorthite-quartz-water at 5
kb (Yoder) 477
Effect of pressure on the boundary
curve in the system diopside-
albite-anorthite (Lindsley and
Emslie) 479
The influence of pressure on the
composition of eutectic liquids
in the binary systems sani-
dine-silica and albite-silica
(Luth) 480
Crystallography 485
The crystal structure of ardennite
(Donnay and Allmann) . . 485
The crystal structures of 2MX
phengite and 2M2 muscovite
(Guven) 487
A mechanism of stacking se-
quences in dioctahedral micas
(Guven) 492
Refinement of the crystal struc-
ture of pigeonite, (Mg039
Fe0 r2Ca009)SiO3 (Morimoto
and Guven) 494
Omission solid solution in mag-
netite (Kullerud, Donnay, and
Donnay) 497
Relative orientations of inter-
grown crystals (Meyer, Don-
nay, and Donnay) .... 498
Cross-twinning as a mechanism of
phase transition (Kullerud,
Donnay, and Donnay) . . 503
Statistical Petrography (Chayes) . 505
Negative variances among theo-
retical open variables formed
from Harker arrays . . . 505
Biogeochemistry 510
Production of hydrocarbons from
the organic matter in a Re-
cent sediment (Mitterer and
Hoering) 510
Organic acids from the oxidation
of Recent sediments (Hoer-
ing) 515
Racemization of amino acids in
fossil shells (Hare and Abel-
son) 526
Geochronology (Krogh, Davis, Al-
drich, Hart, and Stueber) . . 528
Geological history of the Grenville
province 528
Structural Geology 536
Natural slip folds in which the
fold axes nearly parallel the
slip lines (Hansen) . . . 536
Real versus apparent displacement
in slip folds (Hansen and
Scott) 538
Experimental Techniques . . . 539
P-V-T measurements on hydrogen
up to 600 °C and 1800 atmos-
pheres (Presnall) .... 539
High-pressure, high-temperature
X-ray diffraction (Meyer,
Bell, and England) ... 541
Pressure effect on the platinum
versus platinum 10% rhodium
thermocouple (Bell, England,
and Boyd) 545
Staff Activities 547
Journal of Petrology .... 547
Lectures 548
Penologists' Club 551
Summary of Published Work . . 553
Bibliography 560
References Cited 560
Personnel 570
320
INTRODUCTION
Much of the research performed associated with major universities.
at the Geophysical Laboratory is con- Foreign participants have returned
ducted by postdoctoral fellows and to their own countries, where they are
guest investigators. They bring to the continuing activities in earth science.
Laboratory new ideas, enthusiasm, Collaborative research is not 1 inl-
and a desire to gain experience and ited to fellows or staff in residence
new insight. At the beginning of his at the Laboratory. As a result this
stay at the Geophysical Laboratory annual report describes activities in
each fellow is closely associated with which senior staff members, fellows,
a staff member. Every effort is made guest investigators, and collaborative
to make the new man comfortable and investigators participated. This year
to provide him quickly with facilities considerable attention was devoted to
for research. Often fellows are able experiments involving the pyroxenes,
to begin experiments within a day Almost any pyroxene found in na-
after their arrival, using equipment ture contains 5 to 7 cations in
and materials provided by their amounts greater than 1%. The most
sponsors. With time, the relation be- abundant varieties are formed from
tween fellow and sponsor changes. MgSi03, CaSi03, and FeSi03. Through
Often the young men embark on com- ionic substitution and solid solution
pletely independent work, using other cations, such as Al8+, Ca3+, Fe3+,
equipment provided for them, if nee- Na+, and Ti4+, can be incorporated into
essary, by special appropriation. the crystal structure. The extent of
Men who have held fellowships solid solution is dependent on phys-
at the Geophysical Laboratory often ical conditions such as temperature,
remark that an important feature of pressure, and oxygen fugacity. The
their stay has been the educational results of differing conditions are
aspect of the experience; that they seen in the varied compositions and
have learned much, not only from the textures of pyroxenes associated
staff but also from the other fellows, with basalts, layered intrusives, meta-
In part, this education is associated morphic rocks, and ultramafic nod-
with informal discussion. Perhaps ules such as those found in kimber-
more important is the intense work lites.
on joint projects the fellows have en- The wealth of mineral associations
gaged in. Almost all of them partici- involving pyroxenes and their wide
pated in one or more joint efforts, in chemical and structural differences
addition to the research they con- have led to broad interest in labora-
ducted as individuals. tory studies of these minerals. The
Fellows are stationed at the Geo- significance of research on pyroxenes
physical Laboratory for periods of has been enhanced by a growing in-
2 to 4 years. Guest investigators terest in the chemistry of the earth's
remain here from some months to mantle, in which pyroxenes are
several years. During the past 15 known to be a major constituent,
years 56 fellows and 25 guest investi- Petrologists are also hopeful that
gators have been in residence. They quantitative knowledge of equilibria
have represented 15 universities of among the pyroxenes will be an ex-
the United States and 15 foreign traordinarily useful tool in determin-
countries. Most of our alumni are now ing complex petrogenetic history.
321
322 CARNEGIE INSTITUTION
Recognizing these opportunities, staff reached equilibrium compositions.
members at the Geophysical Labora- Pyroxenes of pigeonitic composi-
tory in the past have conducted many tion are common in basic igneous
studies on this group of minerals. The rocks, either in the monoclinic form
classic methods of experimental pe- or as orthorhombic hypersthene with
trology have now been supplemented coarse lamellae of exsolved augite.
by a powerful tool — the electron They are at present the least under-
microprobe. stood, most problematical members of
Boyd has found that an accuracy the common pyroxene quadrilateral,
of lrc to 2r < of the amount of an ele- Brown has studied the inversion re-
ment present in an unknown can be lationships of natural specimens at 2-
achieved in electron-probe analysis if 20 kb and at zero pressure in the
standards are selected so that the ab- presence of andesitic liquids. He be-
sorption correction is kept below lieves that protopyroxene is stable at
about 10%. Analyses for 10 elements high temperatures only for enstatite-
have been made for a group of diop- rich compositions, and that at low
sidic pyroxenes from kimberlite nod- pressures the natural pigeonites crys-
ules. These pyroxenes show a much tallized in the monoclinic form at
wider range of solid solution with en- about 1050 °C, exsolved augite on
statite than do pyroxenes from other cooling to about 1020 °C, and inverted
geologic environments. Comparison to orthorhombic hypersthene at about
with the solvus in the pure system 990 °C. Polymorphic mineral inver-
CaMgSi206-MgSi03 indicates a range sions and exsolution in the presence
of equilibration temperatures for of magmatic fluxes appear to be much
these nodules of about 900°-1300°C. more rapid than in isolated crystals,
Thus they appear to have come from thus permitting studies relevant to
a considerable range of depth in the the crystallization history of many
upper mantle. igneous pyroxenes at the lower pres-
Ca-rich and Mg-rich pyroxenes in sures.
large basic intrusions such as the Clinopyroxenes from many alkaline
Bushveld, Skaergaard, and Stillwater igneous rocks contain a remarkable
commonly show exsolution lamellae amount of ferric iron in excess of so-
that are 1 to 100 jx thick. Boyd and dium. The phase relationships in the
Brown have made an electron-probe system diopside-woilastonite-hematite
study of these pyroxene lamellae, bear directly on the formation of
Semiquantitative analyses of lamellae Fe203-rich clinopyroxenes in igneous
8-30 \l thick show that clinohyper- rocks that have been subjected to
sthene lamellae in augite have the strongly oxidizing conditions. Accord-
same composition as separate grains ingly, Huckenholz, Schairer, and
of host rhombic enstatite. Similarly, Yoder have studied the system diop-
augite lamellae in hypersthene have side-wollastonite-hematite in order to
a composition close to separate grains determine the nature of incorporation
of host augite. The hypersthene hosts of ferric iron in diopside, the prin-
have inverted from pigeonite, but the cipal end member of clinopyroxenes,
Mg-rich lamellae in augite have re- in the absence of sodium at 1 at-
tained their monoclinic symmetry, mosphere pressure. A series of
possibly because of a structural con- compositions were prepared along
trol by the monoclinic host. The la- the joins diopside - ferri-Tschermak's
mellae commonly show compositional molecule, diopside-andradite, and di-
gradients, and thin lamellae in the opside-hematite of this system. The
range 1-5 p have apparently not maximum solid solution of the ferri-
GEOPHYSICAL LABORATORY 323
Tschermak's molecule (CaFe23+Si02) series from Paricutin, Mexico, re-
in diopside is 33% by weight at ported last year by Tilley, Yoder, and
1157°C, and that of hematite in di- Schairer, have been extended to cover
opside is 1% at 1292°C. a larger suite of analyzed lavas.
Lindsley has continued his investi- Strontium-isotope ratios of basalt
gations of the Ca-Fe pyroxenes. Work and andesite in this succession are
with Brown and Muir shows the identical and are characteristic of
pressure-temperature dependence of basalts. These results appear to pre-
the hedenbergite-wollastonite inver- elude the possibility of the origin of
sion in a hedenbergite from the the andesites by assimilation of old
Skaergaard intrusion. These data, sialic crustal material. Melting work
combined with the quartz-tridymite on the unique tholeiitic volcanic
inversion curve, show that the late series of Thingmuli, Iceland — extend-
stages of the Skaergaard crystallized ing from picrite basalt to rhyolite as
at 500 ±100 bars, over a temperature a fractional crystallization trend — is
interval that included 965° ± 15 °C. compared with the earlier reported
The late-stage hedenbergites of both experimental study on the historic
the Skaergaard and Bushveld intru- Kilauea tholeiite succession. Liquidus
sions coexist with both quartz and determinations already reported on
fayalite. Lindsley and Munoz have the border facies or so-called chilled
found that synthetic hedenbergite margins of layered igneous complexes
solid solutions in equilibrium with have been extended to cover the mar-
fayalite + quartz have compositions ginal facies of other intrusions,
close to those of these natural igneous Brown and Schairer have studied
hedenbergites. Dundon and Lindsley melting relations in calcalkaline rocks
report on Mossbauer studies on syn- of circumoceanic island arcs, using
thetic Ca-Fe pyroxenes. examples from the West Indies and
Munoz has investigated the high- the Solomon Islands. Volcanism in
pressure stability relations of spodu- such arcs is particularly explosive and
mene, a lithium pyroxene. This is the catastrophic (e.g., Mount Pelee and
first detailed equilibrium study of a Krakatoa) . The chemical and melting
geologically interesting lithium-bear- relations suggest that these rocks,
ing system at high pressures. Pres- rich in calcic plagioclase and low in
sure drastically alters the phase rela- potassium, represent a calcalkaline
tions of lithium aluminum silicates; group different in type and origin
notably, the low-pressure, high-tem- from those found at continental mar-
perature tetragonal polymorph (/?- gins.
spodumene) is replaced at higher During the past years significant
pressures by a hexagonal polymorph progress has been made toward un-
with a high-quartz structure. derstanding crystallization relations
Jadeite-acmite pyroxenes are im- in simplified synthetic basaltic com-
portant constituents of rocks of the positions, particularly those with a
blue-schist and eclogite facies. Gilbert low anorthite content in the plagio-
has determined the cell parameters clase feldspar crystals. This year
of synthetic jadeite-acmites and has Schairer, Tilley, and Brown studied
found that the substitution of Fe3+ the join nepheline-diopside-anorthite
for Al in these pyroxenes appears to and found the crystallization relations
be random, with no measurable excess in synthetic basaltic compositions
volume of mixing. with a high anorthite content in the
Experimental studies on the melt- feldspar. The study has yielded crit-
ing relations of the calcalkali volcanic ical information on basalts of the al-
324
CARNEGIE INSTITUTION
kali type bearing nepheline with or
without a basic plagioclase.
Experimental studies by Tilley and
Yoder were conducted at high pres-
sures on a suite of volcanic and sub-
volcanic, melilite-bearing, and other
alkali ultramafic rocks. The results
indicate that these assemblages are a
high-level equivalent of a pyroxenite
fades assemblage at deeper zones in
the crust.
The primary component of the
melilites in a major igneous rock
series is akermanite. Its stability
field has been delineated, and the
pressure and temperature restric-
tions on melilite-bearing rocks have
been outlined by Yoder.
The high-pressure hydrous mineral
yoderite has been synthesized by
Schreyer and Yoder, and Schreyer
investigated its relationship to other
phases in the MgO-Al203-Si02-H20
system in the range 10-30 kb.
Schreyer has synthesized Mg-stauro-
lite in the course of these studies and
has outlined series of reactions at
pressures believed to exist near the
base of the average continental crust.
Lindsley and Emslie have shown
that a shift in the boundary curve in
the diopside-albite-anorthite system
at high confining (dry) pressures is
not an adequate model for the gener-
ation of anorthositic magmas.
Oxidation-reduction trends in ig-
neous and metamorphic rocks are
important to an understanding of the
evolution of rocks. One of the key
reactions involved in these trends is
the oxidation of the iron-olivine,
f ayalite. Wones and Gilbert have com-
pleted the first experimental investi-
gation of this reaction at tempera-
tures between 600° and 800°C.
Richardson has investigated the
stability of Fe-staurolite + quartz in
equilibrium with a fluid composition
controlled by the quartz-fayalite-
magnetite oxygen buffer. His results
show that the association of stauro-
lite with quartz in metamorphic rocks
indicates fluid pressures greater than
1.5 kb and that the characteristic ab-
sence of cordierite in some types of
regional metamorphism indicates
fluid pressures greater than 3.5 kb.
Richardson, Bell, and Gilbert have
determined kyanite-sillimanite equi-
librium between 700° and 1500°C.
They have demonstrated that water
has no effect on the position of the
polymorphic transition; an independ-
ent knowledge of the temperature of
the kyanite-sillimanite isograd, pre-
served in metamorphic rocks through-
out the world, therefore permits the
estimation of load pressure and hence
depth of burial during metamor-
phism. Using temperatures measured
from oxygen-isotope fractionation
between minerals, they have shown
that an area of central Connecticut
was metamorphosed under at least
6.5 kb pressure, corresponding to a
depth of burial of about 15 miles
during the metamorphism of this
area. Comparison of the kyanite-sil-
limanite isograd with a dehydration
reaction isograd allows the calcula-
tion of both load and water pressure.
This calculation was made for the
second sillimanite isograd in central
Connecticut and shows that water
pressure was considerably less than
load pressure.
Diamonds are generally conceded to
originate in the earth's mantle, and
thus mineral inclusions that occur in
diamonds are probably samples of
the mantle. Meyer has examined some
of these inclusions and has observed
that there are similarities between
them and the primary minerals in
kimberlite ultrabasic nodules. A dif-
ference is the presence of muscovite
and coesite as inclusions. These
phases have not been found in the
nodules.
Investigation of the phase relations
applicable to massive Ni-Cu pyrrho-
tite ores of the Sudbury type has now
GEOPHYSICAL LABORATORY 325
been extended through a temperature Donnay, and Donnay, who used the
range from above 1000 °C to below method of precession goniometry.
400°C. New data indicate that the Kullerud, Donnay, and Donnay have
ores probably form at or above interpreted experimental evidence
1000 °C. Changes occur on cooling so relative to marcasite-pyrite inter-
that the final sulfide assemblage re- growths and suggest that cross-
fleets only low-temperature phase re- twinning is the mechanism, and inter-
lations. Phase relations at liquidus sections of twinned lamellae are the
temperatures have been studied by locus, of the transformation of mar-
Kullerud and Moh in the Cu-Fe-S and casite to pyrite.
Cu-Ni-S systems and by Craig and In a continuation of his research on
Kullerud in the Cu-Fe-Ni-S system, mica structures Gliven has completed
Naldrett, in a study of a portion of refinements of the crystal structures
the Fe-S-0 system, has found that of 2MX muscovite and 2M2 phengite.
solidus temperatures of pyrrhotite- The results show a decrease of dis-
magnetite mixtures are strongly de- tortions and a shortening of the c
pendent on the sulfur content of the axis in phengite, explaining its oc-
pyrrhotite. The presence of water has currence under specific conditions in
no effect on the pyrrhotite-magnetite metamorphic petrology. By a com-
solidus. Substitution of Ni for Fe has parison of the structures of 2MX and
no effect, but substitution of Cu for 3T muscovite, a mechanism of stack-
Fe depresses solidus temperatures ing sequence is proposed that is based
about 20 °C. Naldrett and Craig have on structural factors arising within
continued their study of the partial the single mica layers,
pressure of sulfur in the vapor coex- Chayes has continued his study of
isting with the Fe^S-Nii^S solid correlation in closed arrays, attempt-
solution, ing to apply to petrographic data the
The structure of the rare manga- techniques described in Year Book 65.
nese arsenic aluminum silicate arden- He has found that if the range of
nite was determined by Donnay and variance is not excessive, it is usually
Allmann. A new silicate building possible to construct the required set
block, Si3Oio, forms part of the struc- of uncorrelated open variables and to
ture. compute from their parameters the
Two metallo-organic crystal struc- appropriate null correlations against
tures, those of tetramethylplatinum which to test the correlations actually
and hexamethyldiplatinum, described observed in sample closed arrays.
in the literature and much discussed When one of the variances is very
in the past 20 years as examples of much larger than the others, how-
nontetrahedral carbon atoms, were ever, as in the suites of analyses of
discredited by Donnay, Krieghoff , and subalkaline volcanics usually por-
Cowan. They were able to prove that trayed in Harker diagrams, negative
the crystals originally used had elements may occur in the vector of
been incorrectly identified; knowing open variances. In Harker arrays, in
the correct chemical compositions, fact, this is the rule rather than the
namely, trimethylplatinum (IV) hy- exception. Variance is by definition
droxide and trimethylplatinum (IV) positive, and, since the computation of
iodide, they were able to determine each expected closed variance (or co-
the structures, which are normal. variance) involves a quantity that is
The problem of determining the the sum of all the open variances, a
orientation of a crystal included in negative element anywhere in the
another crystal was solved by Meyer, vector of open variances terminates
326
CARNEGIE INSTITUTION
the testing procedure. Chaves has
recently been attempting to eliminate
negative elements from the vector of
open variances by transformation to
variables that are substantively jus-
tifiable linear combinations of the
original weight percentages. In the
subalkaline suites so far examined,
the transformation of Na20 to ab
nearly always eliminates the negative
variances ordinarily characteristic
of the open equivalents of Na20 and
A1;0;,
Granitic rocks and gneisses having
whole-rock rubidium-strontium ages
of 1500-1800 million years have been
found over much of the northwest
Grenville province in Ontario, as well
as 500 miles to the east in Quebec
along the southeast margin of the
Grenville. Most of the rocks of this
province were metamorphosed about
1000 m.y. ago. Krogh and Davis have
determined that the radius of the
rock system closed to migration of
the radioactive dating elements dur-
ing this regional metamorphism was
on the order of 5-10 cm. The finding
of older rocks in the 1000-m.y. Gren-
ville province proves that intense
metamorphism preceded the Grenville
orogeny.
Mitterer and Hoering have demon-
strated that a mild temperature treat-
ment of the insoluble organic matter
in a Recent sediment produced a num-
ber of substances commonly found in
petroleum. Normal alkanes, isopre-
noid hydrocarbons, and porphyrins
were produced in a few days at
200 °C. Chemical reduction of the
long, straight carbon chains and
other structural components occurred.
They found that a Recent sediment
can cause such reduction. Olefins and
alcohols are reduced to saturated hy-
drocarbons when heated with a Re-
cent sediment.
Hoering has studied the products
of the oxidation of kerogen. He has
identified short- and long-chained
fatty acids and has devised effective
means of separating and identifying
the products. He found that less than
1 % of the organic carbon in a Recent
marine sediment can be accounted for
as long, straight-chained fatty acids.
The protein of molluscan shells is
made up of L-amino acids. Hare and
Abelson have examined the stereo-
isomeric configuration of amino acids
in fossils and have found that com-
plete racemization has already oc-
curred in lO-m.y.-old specimens. Sub-
stantial amounts of D-ammo acids are
present in 60,000-year-old specimens.
Presnall has continued his investi-
gations on the P-V-T properties of
hydrogen. The data now available
over the range 200°-600°C and 0-
1800 atmospheres provide a basis for
carrying out hydrothermal experi-
ments in which the oxygen fugacity
can be arbitrarily fixed independently
of pressure and temperature.
Hansen spent much of this year
assembling, testing, and calibrating a
rock-deformation apparatus designed
for operation up to 10 kb and 1000 °C.
His continued structural analysis of
folds in rock has resulted in the dis-
covery of a set of natural slip folds
whose axes practically parallel the
slip lines.
Meyer, Bell, and England designed
a new high-pressure, high-tempera-
ture piston-and-cylinder X-ray dif-
fraction device, which utilizes boron
carbide as a pressure vessel.
Effects of pressure on platinum-
platinum 10% rhodium thermocou-
ples have been measured by Bell,
England, and Boyd. At 5 kb, 500 °C,
the effect is 3°C. A significant in-
crease of the effect at higher pres-
sures and temperatures seems to lead
to an error of serious proportions in
the range 10-50 kb and 500o-1500°C.
An apparatus built to measure this
error can be operated to 13 kb,
1000° C, and further work will en-
compass this range.
GEOPHYSICAL LABORATORY
327
EXPERIMENTAL AND DESCRIPTIVE
MINERALOGY
PYROXENES
Quantitative Electron-Probe
Analysis of Pyroxenes
F. R. Boyd
Instrumental developments have
profoundly influenced experimental
petrology in recent years, and it is
now clear that developments in X-ray
analysis, in particular the advent of
the electron microprobe, are having
a similar impact on analytical petrol-
ogy. Use of the electron probe per-
mits nondestructive quantitative anal-
ysis of mineral grains at least as
small as 8-10 /*. The degree of homo-
geneity as well as the composition of
a material can be estimated in probe
analysis, and this capacity will prob-
ably revolutionize our concept of equi-
librium in mineral systems. Use of
the probe to fix the composition of
coexisting phases in the products of
laboratory experiments will provide a
technique that is both more sensitive
and more direct than estimating com-
position from cell parameters or in-
dex-of -refraction measurements.
Quantitative electron-probe analysis
requires either the use of standards
very close in composition to the un-
knowns or the application of complex
corrections for differential emission,
absorption, and fluorescence of X
rays. The complexity and composi-
tional range of minerals such as
pyroxenes often make the use of
analyzed natural standards unsatis-
factory. Various theoretical and em-
pirical approaches to the correction
problem have been made, but as yet
there is no generally established
method of reducing data. Hence, anal-
ysis of any particular suite of speci-
mens requires testing of standards
and correction procedures. In com-
mon rock-forming silicates elements
of low atomic number predominate,
and the principal correction is for
differential absorption. Fluorescence
is often negligible, and atomic num-
ber effects can be avoided by proper
choice of standards.
Correction for Absorption
In silicates X-ray absorption effects
are generally severe for elements be-
low potassium in the periodic table.
For example, about 60 % of the NaK«
radiation generated in a diopsidic py-
roxene by an electron beam at 15 kV
is absorbed before it leaves the sam-
ple. Absorption can be evaluated by
using the expression derived by Phili-
bert (1963) and modified by Dun-
cumb and Shields (1963). This
expression was developed for micro-
probes in which the electron beam is
normal to the specimen surface. With
the Materials Analysis Company
probe* used in the present investiga-
tion the beam is inclined to the sur-
face at an angle of 62.5°. To allow
for this change in geometry, the term
x = (fi/p) cosec 9 in the Philibert
expression is replaced by x — (v>/p)
sin 0! cosec 6>2 (Macres, 1963). In
these equations fi/p is the absorption
coefficient, 61 is the angle of incidence,
and 6 or 02 is the take-off angle. All
absorption factors quoted hereafter
have been calculated with this ex-
pression.
The efficiency of the Philibert ex-
pression in evaluating absorption in
silicates has been tested by analysis
of various silicate glasses prepared
by J. F. Schairer. Table 1 shows a
microprobe analysis of Schairer's
CaMgSi206 glass with his CaSiOs
glass and MgSi03 glass as standards.
The values headed /(xK are the cal-
*We gratefully acknowledge the assist-
ance of the National Science Foundation
under grant GP 4384 in the purchase of
this instrument.
328
CARNEGIE INSTITUTION
TABLE 1 . Analyses of Pure CaMgSi206 Glass, with CaSi03 and MgSi03 Glass as Standards
Standard
f(x)u
f[x)s/f(x)u
Probe
Expected
CaO
MgO
Si02
CaSi03
MgSi03
\CaSi03
^MgSi03
0.909
0.541
0.685
0.685
1.011
1.098
1 .055 I
0.939 J
26.2
18.7
55.3*
25.90
18.62
55.48
Totals
100.2
100.00
* Average of 55.2 and 55.4.
ciliated fractions of generated X rays
that are emitted from the sample, and
the values headed f (x) *//(x)« are the
absorption corrections. For example,
the /(,\)„ value of 0.541 given for
MgO means that 54.1^ of the gen-
erated MgKa radiation is emitted, and
the /(x)«//(x)« value of 1.098 means
that the observed intensity ratio (un-
known/standard) for MgKa radia-
tion is increased 9.8% to arrive at
the listed MgO composition. The f(x)u
values were calculated from the ex-
pected compositions. Either MgSi03
glass or CaSi03 glass can be used as
the standard for silicon, and results
are given for both. These make an in-
teresting test of the Philibert expres-
sion because the correction is greater
than 1.00 with CaSi03 as the standard
but less than 1.00 with MgSi03. The
two Si02 analyses agree well with
each other and with the expected com-
position. Analyses for MgO and CaO
are also in good accord with expected
compositions.
A further test is shown in Table 2.
Here two of Schairer's glass compo-
sitions on the join MgSi03-Al203 are
analyzed, with a third glass on this
join as a standard. The absorption
corrections are relatively small, less
than 4%, and the analytical results
check well.
Nevertheless, if the absorption cor-
rection is large, significant errors
appear. Table 3 shows analyses of the
three MgSiO.-Al203 glasses, with
high-purity MgO, A1203, and Si02 as
standards. For Mg the absorption
correction is about 8%, and the probe
TABLE 2. Analyses of Two Glasses on the
Join MgSi03-AI203, with a Third Glass on this
Join, Containing 5.00% Al203/ as a Standard
Hx)s/f(x)u
Probe
Expected
MgO
B
0.998
36.9
36.14
Si02
1.016
54.0
53.86
Al203
0.985
10.1
10.00
Totals
C
101.0
100.00
MgO
0.998
32.8
32.13
Si02
1.044
48.2
47.87
Al203
0.961
20.0
20.00
Totals
101.0
1 00.00
TABLE 3. Analyses of Three Glasses on the
Join MgSi03-AI203/ with Pure MgO, Al203/ and
Si02 as Standards
f(x)u
f{x)s/fMu
Probe
Expected
A
MgO
0.594
1.079
37.7
38.15
Al203
0.543
1.309
5.1
5.00
Si02
0.635
1.209
58.3
56.85
Totals
101.1
100.00
MgO
0.595
B
1.077
36.4
36.14
Al203
0.551
1.290
10.3
10.00
Si02
0.626
1.227
55.3
53.86
Totals
C
102.0
100.00
MgO
0.595
1.077
32.4
32.13
Al203
0.565
1.258
20.5
20.00
Si02
0.608
1.263
49.5
47.87
Totals
102.4
100.00
GEOPHYSICAL LABORATORY
329
analyses for MgO are in satisfactory
agreement with expected composi-
tions. For Si and Al the correction
is in the range 20% -30%, and the
probe analyses deviate from expected
compositions by about 2% -3% of the
amount of Si02 and A1203 present.
The analyses for A1203 and Si02 are
greatly improved by making the ab-
sorption correction, but the accuracy
is not as good as in cases where the
correction is small. In all cases shown
in Table 3 the Philibert expression
has overcorrected the probe measure-
ments, resulting in totals substan-
tially greater than 100.0. Beaman
(1967), in a comparative study of
correction procedures, has also noted
this tendency.
At best, electron-probe analyses
have an accuracy on the order of db
l%-2% of the amount of an element
present. If this degree of accuracy is
desired it is clearly essential to select
standards so that the absorption cor-
rection will be kept less than about
10%. This restriction does not ordi-
narily cause a severe problem for sili-
cate analysis. Corrections can be
evaluated approximately by computer
in advance of analysis, and suitable
standards can be chosen.
Selection and Preparation
of Standards
A program of analysis of Mg-rich
silicates from kimberlite nodules
initiated last year (Boyd, Year Book
65, p. 252) is being extended. Particu-
lar attention has been directed to the
diopsidic pyroxenes from these nod-
ules, inasmuch as these can provide
information on the equilibration tem-
peratures of the nodules. These diop-
sides contain Si, Ca, Mg, Fe, Al, and
Na as major cations, together with
minor amounts of Ti, K, and Mn. It
is not ordinarily possible to distin-
guish between ferrous and ferric iron
with a microprobe, but analyses can
be recalculated as oxides with an as-
sumed valence for Fe. Uncertainty
over the valence of Fe does not seri-
ously affect the analytical totals for
these minerals because they are rela-
tively poor in Fe.
Synthetic glasses have been found
to be the best standards for this
analytical program. These glasses can
be made homogeneous by repeated
crushing and fusion. In most cases,
their compositions are known pre-
cisely. Their grain size can be made
as coarse as desired, which is a great
convenience in a standard subjected
to repeated analyses. Na glasses have
a limited stability under the electron
beam, but we have found that glass
compositions on the join NaAlSi206-
CaMgSi206 have adequate stability,
provided that the beam is kept larger
than 10 n and the specimen current
is held below 0.1 fta.
Many glasses containing Si, Ca,
Mg, Al, and Na have been prepared
by J. F. Schairer in the course of his
phase-equilibrium studies. Analyti-
cal checks on some of these for Si,
Ca, Mg, and Al have already been
described. Difficulties with Na were
anticipated because of its volatility
and because the absorption coeffi-
cients for Na are less well known
than those for elements of higher
atomic number, but in practice these
difficulties have not proved serious.
Table 4 shows probe analyses for Na
for two glass compositions on the join
CaMgSi206-NaAlSi206 together with
a natural jadeite analyzed by E. G.
TABLE 4. Analyses of Na Standards, with a
Glass of the Composition 65% CaMgSi206—
35% NaAlSi20$ as a Reference
HX)U
Hx)s/f[x)u
Wt
% Na
Probe
Expected
Natural
jadeite*
di95Jds
0.459
0.414
0.408
0.926
1.027
1.042
10.8
1.74
0.58
10.66
1.71
0.57
* USNM 94829; analysis by E. G. Zies
(Yoder, 1950b).
OOA
oo<
CARNEGIE INSTITUTION
Zies. A third glass on this join was
used as a standard. Absorption cor-
rections for these analyses range up
10 7.5%, but the results check very
well.
Standards for Ti and Ni were pre-
pared bv synthesizing glasses on the
joins CaMgSi,0,;-TiO, and CaMgSi*
Oo-XiO, with the techniques devel-
oped by Schairer. The compositions of
these standards are being checked by
C. 0. Ingamells.
Fe, Mn, and Cr have variable va-
lence, and when incorporated in a sili-
cate liquid will reduce and dissolve in
the platinum container. It is virtually
impossible to make a homogeneous
glass of predetermined composition
containing these elements at atmos-
pheric pressure. Nevertheless, reduc-
tion reactions that evolve oxygen are
greatly inhibited by high pressures.
Homogeneous glasses containing
small amounts of Mn and Cr were
synthesized by melting and quenching
from 1700°-l800°C and 10 kb. The
compositions of these glasses are not
known so precisely as those of the
glasses prepared at atmospheric pres-
sure. Unfortunately, it has not yet
been possible to prepare them in suf-
ficient quantity for wet-chemical
analyses because of the small size of
high-pressure reaction capsules.
A check on the Mn glass was made
by using it as a standard for the
analysis of a number of minerals with
a wide variation in Mn content. The
results are shown in Table 5. These
samples are not ideal: The cumming-
tonite is very inhomogcneous, as
shown by the high value for the ratio
<j/\/N ; also, there is a relatively
large difference in the atomic num-
bers of the magnetite and the silicate
standard. Nevertheless, the composi-
tion of the Mn glass is established
well enough for use as a standard
for analysis of the small amounts of
Mn in kimberlite minerals.
The composition of the Cr glass
was checked by analyzing it relative
to certified K2Cr04. This analysis gave
1.287^ Cr, whereas the expected value
was l.olc/o. The agreement is satis-
factory, but there was a large dead-
time correction for this analysis, and
more checks are needed.
A homogeneous glass containing
iron could probably be synthesized at
high pressures. This will be at-
tempted, but for the present a natural
f orsterite from Balsam Gap, analyzed
by E. G. Zies, is being used as an
iron standard. A natural orthoclase,
supplied by C. 0. Ingamells, has
proved to be homogeneous, and has
been a most satisfactory potassium
standard.
Kimberlite Diopside Analyses
About 50 diopsidic pyroxenes from
South African kimberlites have been
analyzed for Ca, Mg, Fe, and Al.
Results for most of these major ele-
ment analyses were given in Year
Book 65 (p. 252) . The Ca/ (Ca + Mg)
ratios were found to peak at 0.47-
0.48 atomic %, with scattered values
trailing out to 0.34. If the solvus in
the pure system CaMgSi206-MgSi03
can be applied quantitatively to these
TABLE 5. Analyses for Mn in Various Natural
Minerals, with a Glass on the Join
Mg3AI2Si30i2-MnSi03 as a Standard*
Probe
Expected
a/VN Wt % tAn Wt % Mn
Spessartite 7 33.0 33.38
Cummingtonite 25 0.86 0.75
Magnetite 3 0.35 0.30
Diopside ... 0.11 0.10
* The glass contains 4.19 wt % Mn.
or = standard deviation.
N = mean count.
Spessartite, Wodgina (Mason and Berggren,
1941).
Cummingtonite, Mikonui (Mason, 1953).
Magnetite, L4-1 75.
Diopside, Thaba Putsoa, E-3 (Nixon, von
Knorring, and Rooke, 1963).
Values for the ratio a/~\JN are italicized.
GEOPHYSICAL LABORATORY 331
natural pyroxenes, most of them are tites. The Monastry and Malibo Matso
found to have equilibration tempera- nodules also contain spinel, and the
tures in the range 900°~1000°C, with Monastry nodule has rather abundant
a few ranging above 1300 °C. More phlogopite. The Lauwrencia nodule
recent analyses emphasize but do not has been described by Nixon, von
change this distribution. It can be Knorring, and Rooke (1963, No. E-
reasonably interpreted as indicating 11), who gave analyses for the whole
that most of these kimberlites came nodule and for the pyrope. A wet-
from a depth in the mantle where the chemical analysis for the Thaba Put-
temperature was 900°-1000°C, with soa diopside (Table 6) has been pub-
progressively fewer coming from in- lished by Nixon and co-workers
creasingly greater depths where the (1963) and is compared with the
temperature was higher. electron-probe analysis in Table 7.
Application of experimental studies There are discrepancies for Fe and
to these rocks is uncertain because we Al, but agreement on other elements
do not sufficiently understand the is reasonably good,
effects of minor elements on the solid The analyses were carried out on
solution equilibria. To provide better polished thin sections or grain
understanding of minor element dis- mounts, under operating conditions
tribution in these minerals the ana- given in Table 8. Standards used are
lytical program has been extended to described above. For most analyses
include complete analyses of five diop- the spot size was kept at about 10 p.
sides selected from the larger group; These minerals are coarse grained,
results are given in Table 6. These and there was no point in using a
diopsides are from widely separated smaller beam. Count rates were cor-
areas in South Africa. Thaba Putsoa rected for drift, background, dead
and Malibo Matso are Basutoland time, absorption, and fluorescence,
pipes, and Lauwrencia is in South Drift, determined by counts on stand-
West Africa. Monastry is in the ards before and after analysis of un-
Southeast Free State and Shinyanga knowns, was held to less than 3% in
is in Tanganyika. The Shinyanga di- virtually all cases,
opside is from a heavy mineral con- Background is trivial in analysis
centra te, and the assemblage from of major elements such as Si and Ca.
which it came is thus unknown. The In such cases a few readings were
four others are from garnet perido- taken with spectrometers set 0.050-
TABLE 6. Electron-Probe Analyses of Diopsides from Kimberlite Nodules
Thaba Putsoa Shinyanga Monastry Farm Lauwrencia Malibo Matso
Si02 55.3 2 54.5 2 54.9 3 54.4 2 55.2 2
Ti02 0.17 ... 0.13 ... 0.07 ... 0.41 ... 0.003
Al203 2.08 8 2.44 2 2.15 20 3.08 5 2.99 6
Cr203 0.86 2 0.79 2 3.17 50 1.47 4 1.82 27
FeO*
3.35
2
4.06
7
2.15
7
2.61
2
2.13
3
MnO
0.11
...
0.14
...
0.09
• • •
0.09
. . .
0.07
9 • •
CaO
16.2
2
15.5
2
20.0
4
19.6
3
20.2
6
MgO
21.2
7
20.9
7
16.2
3
17.4
2
16.7
2
Na20
1.26
7
1.53
7
2.55
7
1.98
2
2.34
5
K20
0.05
. . .
0.04
...
0.01
. . .
0.02
. . .
<0.002
. . .
Totals
100.6
100.0
101.3
101.1
101.5
* Total Fe as FeO. _
Values for the ratio a/\/N are italicized.
g o o
CARNEGIE INSTITUTION
TABLE 7. Electron-Probe and Wet-Chemical
Analyses of the E-3 Diopside, Thaba Putsoa
Probe
Wet Chemical
SiQ2
55.3
54.61
Ti02
0.17
0.23
AI2O3
2.08
1.30
Cr203
0.86
0.92
FeO'
3.35
4.04
MnO
0.11
0.10
CaO
16.2
16.20
MgO
21.2
20.88
Na20
1.26
1.28
K20
0.05
0.12
Totals
100.6
99.68
* Total Fe as FeO.
Wet-chemical analysis by M. H. Kerr in
Nixon, von Knorring, and Rooke (1963).
Original analysis gives FeO, 3.02; Fe203,
1.14; H20-, 0.41; H20-, 0.10.
0.100 A up and down scale from
peak and averaged. Accurate meas-
urement of background is essential,
however, in the analysis of minor
elements, particularly those present
in quantities less than 1%. Initially,
backgrounds were measured at peak
position on blanks with atomic num-
ber close to the unknown, but this
procedure was found to produce large
errors in background determinations.
An example is shown in Table 9.
Background measurements for A1K«
on a number of pure substances of
similar atomic number are shown.
Extreme values of these backgrounds
differ by a factor of 2. Obviously, if
one were attempting to determine the
concentration of a small amount of Al,
measuring background on a blank of
this sort could lead to serious errors.
The alternative procedure of scan-
ning the peak is tedious, but it seems
necessary. This was done for Al, Fe,
Mn, K, Ti, and Cr. Background was
determined for 10 points ranging
from 0.050 A above peak to 0.050 A
below peak on each specimen. Results
were plotted, and graphical extrapo-
lation under the peak gave a ratio
between the true background at peak
position and the background at an
arbitrary position, usually 0.050 A
above peak. During analysis, the
background was measured at this ar-
bitrary point and corrected by the
predetermined factor. In favorable
cases this procedure makes it possible
to determine background with an ac-
curacy on the order of ±l%-2%.
But in many cases where there are
overlapping peaks (e.g., MnK« and
CrK/3) the uncertainty is much
greater.
The range of background factors
obtained is shown in Table 10. The
TABLE 8. Operating Conditions for Analyses in Table 6
Counting
Points
Kv
;•
Counts/Sec
Line/ Background!
Time, sec
Analyzed
Si
15
0.02-0.05
2000-5000
200-300
10-20
40-170
Ti
15
0.1
60-100
1.01-3.0
100
15-20
Al
15
0.1-0.2
350-600
20-30
100
15-30
Cr
20
0.1
600-2000
6-14
30-50
20-40
Fe
20
0.1
350-700
11-25
30-50
23-40
Mn
20
0.2
50-70
1.5-2
100
20-25
Ca
15
0.03-0.06
850-2000
200-300
20-30
40-80
Mg
15
0.03-0.06
500-1000
200-300
20-30
40-80
Na
15
0.1
50-120
5-25
100
17-40
K
15
0.1-0.2
12-33
1.01-2.0
100
15-30
* i — sample current, p.o.
f Si analyses v/ere made with no slit on detector. All others were made with 20-mil slifr and
pulse height analyzer windows set to pass !>95% of signal.
GEOPHYSICAL LABORATORY
333
TABLE 9. Background Determinations for AlKa
Measured on Peak with Various Pure Substances
Z Counts/Sec
MgO
10.0
10.6
Si02
10.0
21.5
MgSi03
10.0
16.5
CaMgSi20$
10.8
21.6
size of the factor is a measure of the
slope of the background and is larg-
est for Al and Cr. Although these
pyroxenes are similar chemically,
background factors vary substan-
tially. For example, K and Na back-
ground factors range over 10%.
Nevertheless, not much accuracy
would be lost in the analysis of a large
suite of specimens if mean back-
ground factors were determined for
a few samples and extrapolated to
the group as a whole.
At high count rates the electronic
counting system is incapable of re-
cording all the pulses received. The
observed count rate and the true
count rate are related by the expres-
sion N = N'/(1-N't), where N and
A/7 are the true and observed count
rates and r is the dead time. Consid-
TABLE 10. Some Factors Used in Making the
Analyses Given in Table 6
A
£
C
Ca
0.905
1 .002-1 .004
Mg
0.525
1.021-1.046
Si
0.664
1.021-1.036
Al
0.563
0.946-0.961
1.25-1.30
Fe
0.951
1.015-1.022
1 .05-1 .09
K
0.874
0.994-0.997
1.04-1.14
Na
0.396
1 .057-1 .079
1.07-1.17
Mn
0.940
1.016-1.021
1 .05-1 .08
Ti
0.921
0.987-0.993
1.09-1.13
Cr
0.923
0.997-1 .002
1.22-1.24
A. f(x) values calculated for analysis 1 .
B. Range of f[x)s/f[x)u values.
C. Range of values for the ratio of true back-
ground at peak position to background 0.050
A above peak (for KKa, 0.040 A above).
erable effort has been expended in
determining dead time, with results
that are not yet satisfactory. This
correction becomes important only
when differences in count rates on
standard and unknown exceed several
thousand counts per second. Dead-
time determinations for various in-
strumental setups were evaluated by
the method outlined by Wittry (1963)
for fixed-time measurement. Initially
we attempted to determine dead time
on the samples being studied with
usual settings for the pulse height
analyzers and other components.
However, plots of N'/i versus N',
where i is the specimen current,
showed a large scatter. Dead-time
values ranging up to 5 ^sec were
found, and the results were not re-
producible. Additional measurements
were made with metallic specimens.
Data from these measurements
showed a greatly reduced scatter and
yielded reproducible values in the
range 2-3 jusec. The problem with the
carbon-coated silicate samples is not
understood, but it may be that
charges that built up in or under the
carbon film at high specimen currents
influenced the count rate. It is safest
to choose standards so that large
dead-time corrections are not en-
countered.
The range of absorption corrections
made for these analyses is given in
Table 10. Most of them are less than
5%, and the largest, for Na, are in
the range 5%-10%. The /(x)« values
were calculated from the intensity
ratios after corrections for drift,
background, and dead time, and were
then iterated. Only 2-3 iterations
were needed. These corrections defi-
nitely improve the analyses, but the
totals suggest that the Philibert ex-
pression overcorrects even with small
corrections. The totals for these diop-
side analyses range between 100.0
and 101.5. The same range covers the
totals for the standard analyses in
oo 1
oo4
CARNEGIE INSTITUTION
Tables 1 and 2 and an unpublished
forsterite analysis by H. 0. A. Meyer.
None are less than 100.0 and the
mean total is 100.8. Overcorrection
with large corrections is obvious from
the data in Table 3. Perhaps a larger
number of analyses will show this
apparent bias to be fortuitous. If not,
it may prove possible to modify the
Philibert expression to eliminate the
overcorrection.
Fluorescence by characteristic X
rays and background radiation can
also cause errors in microprobe anal-
yses. In silicates fluorescence by the
continuum is usually less important
than characteristic fluorescence. Reed
(1965) has provided a simple expres-
sion for identifying significant fluo-
rescence effects by characteristic
radiation and an easily calculated
procedure for evaluating them. Anal-
ysis 1 in Table 6 was checked for
all possible fluorescence of K lines
by K lines. Only the fluorescence of
CrKa: by FeK« was found to be signif-
icant, and this effect is in the range
0.2%-1.0% of the amount of Cr pres-
ent. The correction is thus small but
worth making, and all the Cr analyses
in Table 6 have been corrected by
using Reed's expression.
The diopside analyses in Table 6
show a wide variability in Ca/(Ca +
Mg) ratio, indicating a large range in
solid solution with enstatite. They
also show a substantial variation in
Cr O and A1208. These variations are
to be expected on the basis of phase-
equilibria studies if the diopsides
come from a range of depth in the
mantle and hence a range of equili-
bration temperatures and pressures.
Not enough data are yet available to
relate the variations to differences in
mineral assemblage.
Individual kimberlite pyroxenes are
more homogeneous than the pyroxenes
of gabbroic and basaltic rocks. Ex-
solution lamellae are virtually absent,
even in those diopsides showing a
high degree of solid solution toward
enstatite. One of the advantages of
microprobe analysis is that quantita-
tive estimates of inhomogeneity can
be made. The ratio a/yN will ap-
proach 1 for a large number of counts
on a perfectly homogeneous sub-
stance, where a is the standard devi-
ation and N is the mean count. Be-
cause of inherent counting variance
the value for a/^N in the range 1-3
cannot be taken as evidence of in-
homogeneity. In principle, a value
over 3 provides assurance of inhomo-
geneity at the 99% confidence level,
and its magnitude is a measure of the
inhomogeneity. This approach loses
sensitivity with decreasing line-to-
background ratio, but cr/^N has been
calculated for the major elements in
the analyses in Table 6. The Shin-
yanga diopside, which is the most sub-
calcic diopside known to the author,
is perfectly homogeneous insofar as
the microprobe can determine. Only
the Monastry diopside is inhomo-
geneous for all elements. Cr and Al
tend to be inhomogeneous to a much
greater degree than other elements.
The concentrations of titanium and
potassium are uniformly small in
these diopsides, and in the Malibo
Matso specimen they approach the
limit of detection, with line/back-
ground ratios in the range 1.01-1.02.
Student's t can be used to test the
significance of small differences be-
tween background and peak count
rates and to estimate minimum de-
tectable differences. In this case a
signal-background difference equiva-
lent to a concentration of 18 ppm of
titanium was found to be significant
at the 99% confidence level. The po-
tassium signal was less than signifi-
cant at this level. For the experi-
mental variance of this analysis the
minimum signal-background differ-
ence that would have been significant
corresponds to 20 ppm of potassium.
GEOPHYSICAL LABORATORY
335
Synthesis and Stability of
Ferri-Diopside
H. G. Huckenholz, J. F. Schairer, and
H. S. Yoder, Jr.
The principal "molecules" of the
Ca-rich clinopyroxenes in alkali ba-
salts and their derivatives are diop-
side (CaMgSi206), hedenbergite
(CaFe2+Si206), acmite (NaFe3+Si206),
jadeite (NaAlSi206), and Tscher-
mak's molecules. Compared with the
first four a much larger complexity
in composition is found in Tscher-
mak's molecules, and various combi-
nations of the main divalent
(Ca,Mg,Fe2+) and trivalent (Al,Fe3+)
or even tetravalent (Ti) cations are
possible. Restricted combinations are
seen in nature; CaAl2Si06, CaTiAl206,
and probably MgAl2Si06 are among
the most common that occur in Ca-
rich clinopyroxenes of alkaline rocks.
In most of those pyroxenes bearing
Tschermak's molecules, a remarkable
amount of Fe3+ is also present, espe-
cially when the Na20 content is rela-
tively small. That is, most of the clino-
pyroxenes have much higher amounts
of ferric iron than is required to
form the acmite molecule. This is
illustrated in the Na20 versus Fe203
diagram of Fig. 1. The presence of a
ferri-Tschermak's molecule CaFe3+
Fe3+Si06 (abbreviated FTs), is in-
ferred, in which Fe3+Fe3+ occupies the
sites of MgSi of the standard diopside
structure.
To study the nature of the in-
corporation of ferric iron in the di-
opside in the absence of soda, a series
of compositions in the plane diopside
Di5oAc50
DiiooAco 5 10 15
Weight per cent Fe203 in clinopyroxene
Fig. 1. Plot of NCI2O versus Fe203 of Ca-rich clinopyroxenes from alkali basalts and their de-
rivatives. (Data were taken from 26 references; complete citations will be given in a forthcoming
paper.)
°°,6
oo<
CARNEGIE INSTITUTION
(CaMg^O«) -wollastonite (CaSi03)-
hematite (Fe 03) of the quaternary
system CaO-MgO-Fe2Os-SiOa (Fig.
2) were prepared along the joins di-
opside - f erri-Tschermak's molecule,
diopside-andradite, and diopside-
hematite, as well as some additional
points in this plane. The data in the
present investigation were obtained
at 1 atmosphere pressure from
quenching mixtures that had been
held between 1000° and 1400°C. Re-
sults are presented in T-X diagrams
and in two isothermal sections of the
di-wo-hem plane.
The particular plane chosen for
study is not ternary, because ferrous
iron is present, in amounts depend-
ing on the temperature and composi-
tion ; magnetite solid solution appears
as a phase in the more iron-rich part
of the system at temperatures higher
than 1200°C. The composition of the
phases therefore cannot be expressed
in terms only of di, wo, and hem, and
the system should properly be treated
Fe?0.
MgO
CaO
SiO-
Fig. 2. The quaternary system CaO-MgO-
Fe203-Si02 and the plane wollastonite-diopside-
(enstatite)-hematite (Fe203). Abbreviations for
phases encountered: FTs, ferri-Tschermak's
molecule; Andr, andradite; Wo, wollastonite;
Di, diopside; En, enstatite; and Mg-Andr,
magnesium-andradite.
as quaternary. For this reason the
joins studied are pseudobinary at
higher temperatures.
The Join D lop side-F erri-
Tschermak's Molecule
The stable phases crystallizing at
atmospheric pressure on the join di-
FTs (Fig. 3) are clinopyroxene solid
solution (cpxss), wollastonite solid
solution (woss), pseudowollastonite
(pwo), andradite solid solution
(andrss), magnetite solid solution
(mtss), and hematite (hem). Mag-
netite^ is stable only in the presence
of a liquid, and andrss, only in the
subsolidus region.
The maximum degree of solid solu-
tion of ferri-Tschermak's molecule in
the stable diopside is greater than
31.1 and less than 35.0 wt % and is
believed to be close to 33% at 1157°C.
To set up suitable determinative pro-
cedures for ferri-diopside solid solu-
tions, the unit-cell parameters were
measured along the join di-FTs
within the range of stable solid solu-
tion with the use of materials that
had been held for 1 week at 1155°C.
Least-squares refinement of data ob-
tained from powder X-ray diffraction
charts (CuKa; X = 1.5418) was car-
ried out with a program for the IBM
7094 digital computer which was
written by Charles W. Burnham. The
reflections measured were 150, 510,
132, 240, 041, 421, 331, 330, 311, 22f,
202, 131, 311, 310, 221, 220, and 021,
with silicon as an internal standard.
Table 11 and Fig. 4 show the results,
together with data obtained by Cole-
man (1962), who carried out pre-
vious crystallographic work in order
to determine the change in cell di-
mensions up to 10 mole % FTs. There
is a substantial increase of a, c, and
V, but a decrease of b from diopside
to difl8.oFTs.3i.i. The angle /3 remains
essentially constant. Coleman's data
reveal the same tendencies in cell
constants but show a much wider scat-
GEOPHYSICAL LABORATORY
337
tfl
C3
III
C\J
o
in
fO
CT
n_
^
O
o
U
O .1= o
* c
«/> O <U
° o _
>
_Q
<
O *
<u .- a
- ° i
° sr
» S. s
3
U
_o
o
E
-"*
a
E
£ a
a
<u 0s
lit:
I C o
a o u=
T3 M= '5»
•5; D ft
a o a
O "> £
^ T3 8
.£ "o ^
o "» c
-2a
c
o
"U
0)
c
(D
ui I-
I „
5 if
n O °
2 •= co
5 "o 5
Q. 10
CO
•2 S o
.tz x "
«» o <d
CL >>'Jp,
I o o
8 J-o
3 w a
O
m ^
02 02
X 02
Q. *-
S 2"g
a c
£ 2 c
o
30 'ajnjDjadwaj.
: 8 .^
co a, *° D
.5?-f= 15 c
/*
0.7
0.5
0.3
20(24O)-20(O4I)
A2 0CuKa
0.1 -
2 0(22T)-2 0(2O2)
Qj 10 20 30
Weight per cent Ca Fe23Si06
Fig. 4. Unit-cell parameters and angular separation within the range of stable solid solution
along the join diopside— ferri-Tschermak's molecule. Crosses represent Coleman's data (1962).
GEOPHYSICAL LABORATORY
339
c
o
•*-
3
o
to
-o
o
CO
<L>
73
c
*5>
o
a
o
3
a
O
■
CO
*z
V.
;g
(D
u_
"o
TJ
CO
c
<D
o
IH
o
*«o
T3
a
"io
o
a
5
o
c
o
c
o
U
o
u_
CD
N
_>.
D
c
<
c
D
a>
■*-
a>
E
o
i_
o
o_
U
■
*E
3
UJ
—i
CQ
i<
CD
s !> o
"v» u
Q.CO
O
5* -8
=2-0
D
z
u
-*f CN CN
O O O CO O
o o o o cs
o o d o o
o «o
+1 +1 +1 +1 +1
O i-;
oo •— '<
CO *— 00 O "t
1 —
OO •— O O CN
^ °i °? «o «o
O CO io o tj-
— -"*
CO CN CN
O O O CN CO
o o o o cn
d d d d d
N -*
+1 +1 +! +1 +1
"*. *s
^ d
.— «o co oo -t
r—
00 i— o CO o
^ °: °1 to co
o co >o o -«t
— xr
CO CN CO
O O O CO o
O O O O CO
d d d d d
00 CO
+1 +1 +1 I! +1
IT) "O
r-" 0
'OCN^OOOO
r—
K cn K O u-)
^ °i *1 V> CN
O CO io o ^
,- ^r
CN i— CN
O O O r- CO
o o o q ^~
d d d d d
o —
+1 +1 +1 +1 +1
NO
io d
•— co O N K
O CN K)C0 IO
K O; CM ^ ^
<> co io o N-
— -**
o o
q q
d d
+1 +1
O — K
°. °. °.
odd
+1 +1 II
oo K oo
xrooo
°1»nc>
O CO if) O CO
«- "ST
io U-)
^r cn
K O
O
O CD v£
u_ u_ o^
340
CARNEGIE INSTITUTION
ter than the data obtained in the
present study.
The substitution of Fe;^Fe3+ for
MgSi within the range of stable solid
solution along- the join di-FTs at
1155T expands the cell volume V by
about 14 rr.* Comparison of the aver-
ago tetrahedral distances of Si-0
(1.61 a: taken from International
Tables for X-Ray Crystallography,
Vol. 3. 1962) and of Fe3+-0 (1.86 A)
and the octahedral distances of Mg-0
(2.07 a) and of Fe3-0 (2.01 A)
shows the more marked difference
for tetrahedral substitution by Fe3+.
In short, the effect of substitution
makes the tetrahedra expand con-
siderably and the octahedra shrink
slightly. The substitution of Mg by
Fe3- within the octahedra results in
a minor decrease of b, whereas the
expansion of the tetrahedra is re-
vealed not only in c but also in a. The
increase of c is caused only by the
substitution of Si by Fe3+. On the
other hand, the change of a is effected
by a contraction of the octahedra as
well, and the expansion of the tetra-
hedra in this direction is very much
larger than the contraction of the
octahedra because two expanded
(Si,Fe3+)206 chains of the diopside
structure are involved. The FeO con-
tents in the range of the stable solid
solution are small (Table 11) and
cause no measurable change in the
cell dimension.
The angular separation of certain
reflection pairs can be used to deter-
mine the composition of ferri-diop-
side solid solution along the join di-
FTs, as well as in certain portions of
the di-wo-hem plane. The most sensi-
* The increase in cell volume suggests
that the ferri-Tschermak's molecule is not
likely to form even at high pressures. In-
formative runs were carried out on the pure
FTs composition at 20 kb and 1300 °C
through the courtesy of Professor W.
Schreyer. The stable phases present under
these conditions are andraditeS3 + hematite
and small amounts of magnetite.
tive angular separation is that of
20(24O)-20(O41), but these reflec-
tions are of low intensity and cannot
be applied in polyphase assemblages.
Another angular separation, 29 (221) -
2(9 (202) is useful, except in the range
di10o to di90FTs10, where the deter-
minative curve passes through a mini-
mum and where the hematite content
is higher; the 1120 reflection of hem-
atite at 35.5° 29 coincides with the
221 reflection of the ferri-diopside
solid solution.
At solidus temperatures for com-
positions between 20% (at 1000°C)
and 33% FTS (at 1157°C), the clino-
pyroxeness breaks down to a clino-
pyroxeness + hematite. This particu-
lar clinopyroxeness must lie off the
join di-FTs in the more wollastonite-
rich part of the di-wo-hem plane, as
can be seen from an inspection of the
isothermal sections at 1137° and
1157°C (Figs. 7 and 8). To investi-
gate this assumption, the correspond-
ing clinopyroxeness in coexistence
with hematite was examined by X
rays, with the reflection pairs de-
scribed above. The reflection shifts
are in fairly good agreement with
those for clinopyroxeness crystallizing
in the upper portion of the field of
ferri-diopside solid solution below the
join, but they are different from those
of clinopyroxenes that lie directly on
it.
The cpxss + hem assemblage is
replaced by cpxss + hem + andrss
at compositions greater than 35-42
wt % at 1000° and 1157°C, respec-
tively. The clinopyroxeness occurring
in this three-phase assemblage can-
not lie on the join di-FTs. By X-
ray determinations and inspection of
the corresponding isothermal sections,
the pyroxene compositions can be
determined as di66wo17hem17 and
diesWOighenijg at temperatures of
1137° and 1157°C; but these pyrox-
ene solid solutions may, in addition,
be expressed as di66wo5FTs29 and
GEOPHYSICAL LABORATORY 341
di62wo5FTs33 or as dieeFTsasandrn and wollastoniteHB. The clinopyroxeneBS,
di62FTs27andrii. The coexisting garnet based on X-ray measurement, must
is almost pure andradite and has the lie between the joins di-FTs and di-
same cell edge as the end member on andr within the di-wo-hem plane
the join diopside-andradite (Fig. 5), (Figs. 5 and 6). The assemblage cpx,8
since the very restricted replacement + wo8S is replaced by cpxS8 + woHS
of Ca by Mg does not effect a meas- + andr88, from andr40 to andr49.
urable change in a. At temperatures AndrS8 + woS8, and andr88, are the
above 1157°C, cpxss + hem + andrS8 only stable phases at compositions
is no longer stable and is replaced by greater than andr97 and andr99, re-
cpxss + hem + woss. Wollastonite spectively.
coexisting with diopside on the join The garnet crystallizing as the pure
CaMgSioOs-CaSiOs has the composi- end member at the join di-andr has
tion wo93di7 according to Schairer and a lattice parameter of 12.053 ± 0.003
Bowen (1942) . A, which is higher than that of 12.048
The liquidus phases on the join di- A reported by Skinner (1956) for
opside-FTs molecule are clinopyrox- hydrothermally synthesized andradite
ene, magnetite, and hematite. With and lower than that of 12.06 ± 0.01
X-ray and microprobe techniques it A given by J. Ito (personal communi-
is possible to define the composition cation) for an andradite produced by
of the primary phases that crystallize gel techniques. The lattice parameter
near liquidus temperatures. Primary for a natural garnet of almost pure
hematite contains traces of Mg and andradite composition from Val
Ca; and even the ferrous iron con- Malenco near Sondria, northern Italy,
tent, which cannot be demonstrated was determined recently by Quareni
by microprobe analysis, must be low, and de Pieri (1966) as 12.061 A. The
since Phillips and Muan (1960) re- range of substitution of Ca by Mg
ported a very limited solid solution of and its effect on cell dimension were
magnetite in hematite of 1 % at about investigated by studying two compo-
1400°C (in air). The composition of sitions on the join andradite-Mg-
the primary clinopyroxeness is ap- andradite (the theoretical garnet
proximately the same as that in the composition on the join enstatite-
bulk composition. The primary mag- hematite if the di-wo-hem plane were
netites consist of solid solutions of extended to MgSi03 composition).
Fe2+Fe23+04 and MgFe23+04, including There is a very limited solid solution
a small amount of CaFe23+04. It can be of Mg-andradite in andradite. From
seen from the compositions marked in the data obtained, the solid solution
Fig. 3 (in mole %) that the Mg con- should be in the range of 1% at 1050°
tent of magnetitess increases with in- and 2% at 1137°C. The substitution
crease in the CaMgSi206 component, causes no change in the cell edge of
. . the garnet, and the a parameter of
The Join Diopside-Andradite the C0-existing andraditess of the two-
The stable phases crystallizing on or three-phase assemblages is identi-
the join diopside-andradite (Fig. 5) cal with that of pure andradite with-
are clinopyroxeness, wollastoniteS8, in the limit of errors,
pseudo wollastonite, hematite, and an- At temperatures above 1157°C
dradite8S. At higher solidus tempera- andradite is no longer a stable phase,
tures solid solution of andradite in and there is a change in the mineral
diopside has not been observed thus assemblages to pseudowollastonite
far. The phases present in this part + hematite, pseudowollastonite +
of the diagram are clinopyroxeneS3 and wollastonitess + hematite, wollas-
342
CARNEGIE INSTITUTION
o
D)
U-
c
UJ (\l
«/>
1- O
O
Q»P
«/>
<JC
c
£?N
O
Q <u
•„J^
2 Ll.
O
<r rr>
D
>
O
d)
O
i_
en
M
JQ
<
6
i—
<D
JZ
a.
*n
o
O
CD
E
O
©
o
T3
D
c
o
CM
1
O
T3
rO
*«/>
o
10
CO
Q.
O
ro
+ CM
"D
CD
Li_
#c
to
*o
O
•~~
o
•*-
o
m
C
o
o
h_
T3
a>
<D
CL
E
+.
*5
si
Jq
CD
o
o
~a>
*
^
o
o
o
o
o
ro
c
#o
o
a
E
o
u
O
C\i
3
to
t_
CD
>
©
-*-
O
o
i—
©
a
E
LU
Q
CD
O
0\J
CD
en
tfl
o»
Q_
o
o
»o
Q
o
#
D)
0o '8jn4DJ3duuax
GEOPHYSICAL LABORATORY
1500
1400
O
CD
1300
a.
E
1200
100
LIQUID
mtss+ L
cpxss
^cpxss+ mtss+ L
""^^.cpXgg+mtgg+hem + L
^ ^ ^c px ss + hem 4- L
cpxss + hem
mtss + hem+L_
DIOPSIDE 10
CaMgSi206
20 30 40
Weight per cent Fe2 03-
50
Fig. 6. Temperature versus composition plot of data obtained from Presnall (1 966) and in this
study on the join diopside-hematite at 1 atmosphere. Abbreviations as in Fig. 3.
tonitess + hematite, and hematite +
clinopyroxeness + wollastonitess. Just
as in the case of the di-FTs join, there
is a small field in which an andradite
solid solution is stable in the presence
of hematite and either woss or pwo,
or both, from 1137° to 1157°C. The
cell edge of this particular andraditess
is 12.045 ± 0.002 A, which is smaller
than that of the andradite crystalliz-
ing below 1137°C. Such behavior is
probably caused by ubiquitous ferrous
iron forming a skiagite molecule of
Fe32+Fe23+Si3Oi2 composition, in which
Fe2+ takes the place of Ca. This is
indicated by a shrinkage of a from
12.053 to 12.045 A. Above 1137° ±
5°C the garnet starts to decompose,
but since pseudowollastonite and hem-
atite cannot incorporate much of the
ferrous iron present in the runs, it
must stay in the andradite structure
until the garnet breaks down com-
pletely at 1157° ± 5°C. It should be
noted that clinopyroxeness and wol-
lastonitess, which can incorporate fer-
rous iron, stabilize andradite and its
solid solutions up to 1157°C.
The Join Diop side-Hematite
Presnall (1966) worked out the
liquidus relationships for this join,
and Segnit (1951) studied the incor-
poration of ferric iron in diopside
within the subsolidus region. He re-
ported that a maximum of 10 wt %
Fe203 was present in the diopside.
Unfortunately, no temperature data
are given for the corresponding diop-
side solid solutions. With Presnall's
liquidus data and those obtained from
quenching experiments in this study,
it can be seen in Fig. 6 that the solid
solution of hematite in diopside is
very restricted. It is 5% at 1150° C
and less than 6% at 1200°C. The
344
CARNEGIE INSTITUTION
maximum degree of solid solution is
depicted in Fig. 6 as 6% at about
1292° ± 5°C. The substitution along
the join diopside-hematite can be ex-
pressed in two parts as 2Fe3+ for
MgSi, demonstrated on the join diop-
side-ferri-Tschermak's molecule, as
well as 2Fe8+ for CaSi.
Isothermal Sections at 1137° and
1157° C of the Di-Wo-Hem Plane
The results of the thermal studies
outlined previously are shown graph-
ically in two isothermal sections. To
illustrate the most drastic change in
the phase relationship, the 1137° and
1157°C sections have been chosen.
Isothermal section at 1137° C (Fig.
7). The stable phases at 1137°C are
cpxss, hem, woS8, pwo, and andrss. In
the diopside-rich portion of the plane,
cpxss + hem as well as cpxss + woss
coexist. The phases in the wo-rich
portion are wosg + pwo + andrss,
wosg + andrss, and hem + andrss.
The largest area is covered by the
assemblage hem + cpxss 4- andrss.
It is separated from the woss + cpx9S
+ andrss assemblage by a narrow
two-phase field of cpxss + andrss.
Isothermal section at 1157° C (Fig.
8). At a temperature of 1157°C gar-
net is no longer stable, and all assem-
blages with andradite and its solid
hem (Fe203 )
I37°C
3^./TS/^l2.052i0.004A
(Co Fei3Si06)
(Ca3 Fe|3Si3Ol2) ^205i±O.OO^A
z-- c :-:3-:c3
■ ->•'-'•'*
(CaMgSi206)
Fig. 7. Isothermal section of the diopside-wollastonite-hematite plane at 1137°C. Solid circles
are the compositions studied. Some indicate the cell edge of the andradite in the garnet-rich por-
tion of the plane.
GEOPHYSICAL LABORATORY
345
solutions are replaced by other
phases. The assemblage andrS8 +
cpxsa with wo88 or with hem is
changed to hem + cpxS3 + wo88. The
assemblages hem + andr88, woB8 +
andrss, and woss + pwo + andrss are
replaced by hem + woS3 or pwo, or
both. The breakdown of andradite and
its solid solutions does not effect a
change of the di-rich part of the plane.
Compared with the 1137°C section
(Fig. 7) the clinopyroxene solid solu-
tion field is increased to its maximum
of solid solution, and also the areas
of hem + cpxss as well as wo88 +
cpxss are expanded markedly.
Geologic Discussion
Results of this study indicate the
formation of ferri-diopside solid solu-
tions at magmatic temperatures.
These findings are in accord with ob-
servations that the rocks of the alkali
basalt series are characteristically
enriched in ferric iron. In general,
most of the clinopyroxenes from oli-
vine basalts, hawaiites, mugearites,
basanites, and nephelinites contain
small to moderate amounts of both
the ferri-Tschermak's and acmite
molecules. With increasing sodium in
the bulk composition in excess of
potassium, the acmite content in the
clinopyroxenes increases as well, and
in the syenitic end members the py-
roxene phase is aegirine or aegirine-
augite. The content of ferri-Tscher-
mak's molecule in these clinopyrox-
enes is very small or is lacking, as
may be seen from data of Larsen
hem (Fe203;
1157 °C
(Ca Fe+23Si05)
pwo+woss
(Ca3Fe£3Si30|2) ;£/&/-
>V (CaMgSi206)
Fig. 8. Isothermal section of the diopside-wollastonite-hematite plane at 1 157° C.
346
CARNEGIE INSTITUTION
(1942), Yagi (1953), and Tyler and
Kin^ (1967) . Detailed information on
alkaline rocks with potassium affini-
ties and their related pyroxenes is
very meager. In general, the most
common pyroxene in potassium-rich
syenites, shonkinites. theralites, and
ii elites is augitic in composition, not
acmitic. From the few data available,
clinopyroxenes from potassium-
abundant rocks seem to be richer in
ferri-Tschermak's molecule. Leucitite
lavas from the 1929 eruption of
Mount Vesuvius contain ferrian au-
gites (Alfani, 1934; Miiller, 1936)
with an average amount of about 8
mole rc FTs. Augite found in a mon-
chiquite from Khibina, Kola Penin-
sula (Lupanova, 1934), consists of 12
mole % FTs; and Kashkai (1944)
reported an augite from basaltic tuff
in Azerbaidzhan with 10.4%. The
largest amount of ferri-Tschermak's
molecule present in a natural speci-
men is 18.5 mole %; and no sodium is
reported in the chemical analysis
given by Putman (1942). This par-
ticular clinopyroxene was found near
Cape Tourmente, St. Joachim, Mont-
morency County, Quebec.
Ferrian augite is also found in jacu-
pirangite which contains nepheline
from Hessereau Hill, Oka, Quebec,
close to the border of a carbonatite
complex (Peacor, 1967). It contains
7.5 mole % ferri-Tschermak's mole-
cules and 0.8 mole % acmite mole-
cules. In conclusion, larger amounts
of the ferri-Tschermak's molecule of
the clinopyroxenes are probably asso-
ciated with those igneous rocks that
have undergone strongly oxidizing
conditions and contain small amounts
of either sodium or alkalies or both.
The sodium deficiency prohibits the
formation of larger amounts of the
acmite molecule, and potassium en-
ters the diopside structure in very
small amounts.
The typical occurrence of andradite
is in contact or thermal metamor-
phosed, impure, calcareous sediments,
and the assemblage andradite +
clinopyroxene with either wollasto-
nite or hematite is of particular inter-
est with regard to the skarn deposits.
The skarn mineral assemblages can
be formed in nature from limestones
and marbles similar to those obtained
in the experiments under dry and
oxidizing conditions, if the bulk com-
position of the primary sediment,
which is undergoing thermal meta-
morphism, corresponds closely with
the diopside - wollastonite - hematite
plane.
In addition, andradite, which is
present in alkaline igneous rocks, is
found in nepheline syenites and ijo-
lites and their volcanic equivalents,
the phonolites and nephelinites. The
andradite in this particular case is
generally the titaniferous variety,
melanite or schorlomite, and the co-
existing clinopyroxene is an aegirine-
augite or aegirine. The phase
relationships in the system diopside-
wollastonite-hematite, however, bear
directly on the formation of andradite
under magmatic conditions. The main
feature of these particular igneous
rocks is the presence of the mafic
mineral association of clinopyroxene,
andradite, wollastonite, and an iron
ore (in most cases reported as titano-
magnetite, ilmenite, and occasionally
hematite), in which andradite is
formed mostly late in relation to the
formation of clinopyroxene and wol-
lastonite. This can be deduced from
the petrographic descriptions of the
alkaline rocks from the Iron Hill
stock (Larsen, 1942), the Alno com-
plex, Sweden (von Eckermann, 1948,
1958), the Homa Bay Area, Kenya
(Pulfrey, 1950), the Kaiserstuhl vol-
cano, Germany (Wimmenauer, 1962),
and others. The formation of andra-
dite in natural igneous rocks is in
good agreement with the thermal
data outlined in the previous para-
graphs. No andradite crystallizes on
GEOPHYSICAL LABORATORY 347
the liquidus in the system di-wo-hem; and Schairer, 1964) . Rhombic bronz-
the primary phases are clinopyrox- ites (Pbca) show no evidence of in-
eness, wollastonitess, magnetitess, and version from or to another form, are
hematite. Andradite is formed at commonest in the range Fsiri-Fs2,,
temperatures below 1157°C in the and have more Ca than the inverted
synthetic system. In alkali basalts, protoenstatites (cf. CVO and CVC,
which have higher liquidus tempera- Fig. 9). Monoclinic pigeonites (P21/c)
tures, pure andradite cannot be a often have compositional analogues
primary phase; it is likely to be pri- either as clinohypersthenes or as
mary in mugearites and phonolites, rhombic hypersthenes, both with
which melt below 1157°C. augite lamellae; they are restricted to
approximately Fs28-FsC5, and contain
Experimental Studies on Inversion more Ca than the rhombic bronzites
Relations in Natural Pigeonitic (cf. HA and HB, Fig. 9) .
Pyroxenes The suggestion that protopyroxene
~ ,, „ may be the stable, high-temperature
G. M. Brown i « x. 1 • m. /ir ^
phase for natural pigeonites (Yoder,
The relationships between subsoli- Tilley, and Schairer, Year Book 62,
dus exsolution and inversion in pi- pp. 88-95) is difficult to accept at
geonite, clinohypersthene, and rhom- present because:
bic hypersthene have been studied at 1. The protoenstatite structure
pressures of 2-20 kb and at zero favors low Ca contents (Boyd and
pressure in the presence of andesitic Schairer, 1964; Perrotta and Stephen-
liquid. For hypersthene at zero pres- son, 1965) , as indicated by natural
sure, the rhombic <=± monoclinic in- inversion products (Fig. 9). Pigeon-
version took place at about 990 °C. ites are fairly calcic phases, and
Solution of augite in clinohypersthene higher-temperature phases would be
and exsolution from pigeonite were even more calcic through solid solu-
observed to occur between 990 °C and tion with augite.
1020 °C at zero pressure. Previous ex- 2. Protopyroxenes have been syn-
perimental work on polymorphism in thesized only for Mg-rich composi-
Ca-poor pyroxenes and extrapolation tions (summary in Brown and Smith,
to natural compositions have caused 1963) and are unstable at > 2 kb
perplexity in interpretation, espe- pressure at 1300 °C (Boyd, England,
daily since the effects of composi- and Davis, 1964) and, probably, at 1
tional changes have not been con- atmosphere with moderate Ca con-
sidered in sufficient detail. A proposed tents above about 1350 °C (Boyd and
subdivision of Ca-poor pyroxenes, Schairer, 1964, p. 293) or even
shown in Fig. 9, takes into account 1100°C according to Perrotta and
both compositional differences and Stephenson (1965).
observations on natural inversion re- 3. Indirect evidence for quenching
lationships. Monoclinic, polysynthet- of protopyroxene of nonenstatite com-
ically twinned enstatites in meteo- positions to a metastable monoclinic
rites and in a rare type of terrestrial phase (polysynthetic twinning, relict
rock (Dallwitz, Green, and Thomp- crystal form, and volume-change frac-
son, 1966) are restricted approxi- tures) is unsatisfactory, each feature
mately to Fs0-Fs13, contain less Ca being interpretable also according to
than the bronzites, and may be ex- other phase changes (e.g., Yoder, Til-
plained according to metastable in- ley, and Schairer, Year Book 62, p.
version from a protoenstatite (with 90; Brown and Smith, 1963, p. 192).
Pbcn space-group symmetry; cf. Boyd The evidence that Opx is stable at
348
CARNEGIE INSTITUTION
CoMgSi206
MgS :3
' //
BU/ SlAi/B2A
cvc £l
Ha
S2A
\ VpMC \ MU
S3a s4a
Mb
B4a
CVO y
^"B3a
V
_v_
FeSiO:
20 30 40
Mole per cent FeSiC^
50
Fig. 9. Plot showing the compositional field of natural pigeonitic pyroxenes within the quadri-
lateral CaMgSi206-CaFeSi206-MgSi03-FeSi03. A triple subdivision (mole %) is proposed for the
Ca-poor pyroxenes common to igneous rocks, i.e., the protoenstatites <=^ clinoenstatites (FS0-FS13);
the rhombic bronzites (Fsi3-Fs2s); and the monoclinic pigeonites ^± rhombic hypersthenes (Fs28-
Fs^). The chosen ranges are based only on majority averages in order to illustrate the discussion;
exceptions are likely to have a significance of their own. Coexisting Ca-rich phases (where ana-
lyzed) are shown to illustrate the significant pattern of coexistence between phase pairs. CVC and
CVO: range of clinoenstatites and orthorhombic bronzites, respectively, in Cape Vogel lava, Papua
(Dallwitz, Green, and Thompson, 1966, Tables 5 and 6, microprobe analyses). Bl— B4 and B1A—
B4A: range of Ca-rich and Ca-poor pyroxene phases, respectively, coexisting in Bushveld intru-
sion, South Africa (Atkins, 1965). SI— S4 and SI A— S4A: range of Ca-rich and Ca-poor pyroxene
phases, respectively, coexisting in Skaergaard intrusion, East Greenland (Brown, 1957). ST and
STA: coexisting pair from Stillwater intrusion, Montana (Hess, 1960). HA and HB: pigeonite and
bronzite found together in lava from Hakone, Japan (Kuno and Nagashima, 1952). MC: pigeonite
from Moore County meteorite (Hess and Henderson, 1949). MU: pigeonite from Mull andesite
(Hallimond, 1914). The primary bronzites range from B1A to B3A, and the pigeonites (either
primary or inverted) from HA to MU (HB being exceptional in containing more CaSiOa than where
present as the only Ca-poor phase). Note the consistently longer tie lines, i.e., miscibility gap, for
Bushveld, compared with Skaergaard or Stillwater pairs.
higher temperatures than Cpx (e.g.,
Boyd and England, Year Book 6U,
Fig. 33; Lindsley, Year Booh 6Uy Fig.
51) is complicated by the likely effects
of experimental shearing stress at low
temperatures (Turner, Heard, and
Griggs, 1960; Riecker and Rooney,
1966; Munoz, this report), which
results in extreme contradiction of
natural evidence for the thermal his-
tory of phenocrystic pigeonites (Aki-
moto, Kamada, and Kushiro, 1966).
In regard to pigeonitic pyroxenes,
we are concerned with two types hav-
ing similar composition and geological
environment but contrasting crystal-
lographic properties:
1. A monoclinic phase (P21/c),
such as HA, MC, and MU (Fig. 9),
which occasionally attains clinohyper-
sthene composition by augite exsolu-
tion (Hess and Henderson, 1949;
Preston, 1966) but has no inversion
textures. Cooling histories variable,
occurring as phenocrysts (Kuno and
Nagashima, 1952; Hallimond, 1914);
groundmass crystals in lavas (Lewis
and White, 1967); in sills (Hess,
GEOPHYSICAL LABORATORY
349
1949), plutonic complexes (Brown,
1957), and meteorites (Hess and
Henderson, 1949) ; and as lamellae
within augites (Bown and Gay,
1957).
2. An orthorhombic phase (Pbca)
with abundant augite lamellae ori-
ented usually, but not always, in a
pattern suggesting exsolution from
a previously stable monoclinic phase
(Plate 1B-D), such as ST A, B4A,
and S2A-S4A (Fig. 9). Common in
plutonic and hypabyssal rocks (Hess,
1941; Brown, 1957; Atkins, 1965),
notably absent from lavas, and usu-
ally called "inverted pigeonites.,,
Experiments at 2-20 kb
Specimen B4A, from the Bushveld
intrusion, was selected for study,
being a hypersthene (Pbca) with
broad augite (C2/c) exsolution lamel-
lae. In composition (Table 12) it is
very similar to that of a Mull pigeon-
ite with P21/c space group symmetry
(MU, Fig. 9) . The aim of the experi-
ments was to define the P-T condi-
tions under which the Bushveld spec-
imen (provided by Dr. F. B. Atkins)
inverted to P2±/c symmetry at higher
temperatures and back to Pbca sym-
metry at lower temperatures.
Runs at pressures between 5 and
20 kb were made in a solid-media
piston-and-cylinder pressure appara-
tus, the powdered charges being held
in iron capsules with tightly fitting
lids.
Complete reaction of Opx was de-
fined as the production of a clinopy-
roxene phase (P2t/c), recognized by
optical and X-ray powder diffraction
methods and in most, but not all cases,
by polysynthetic twinning. These
products were used for reversal of
the reaction, in all cases to Opx. Ad-
ditional information was obtained, op-
tically, by observing the behavior of
the augite exsolution lamellae in the
rhombic hypersthene, with increasing
temperatures. Slightly below the in-
version temperatures, the augite
coalesced as blebs in the hypersthene
grains and persisted within the clino-
pyroxene inversion product for about
20 °C above the temperatures of com-
plete reaction. Above those tempera-
tures, homogeneous clinopyroxene
prevailed.
The results are given on Fig. 10,
showing both the P-T curve for com-
plete reaction of rhombic pyroxene to
clinopyroxene, indicating maximum
temperature values, and a band
drawn tentatively to include the re-
sults of incomplete reaction. Since the
reaction was found to be time depend-
ent, the curve for true maxima is
probably lower or has a lower slope
than the full one drawn here. Data
TABLE 1 2. Chemical Analyses of Natural Ca-
Poor Pyroxenes Used for Experimental Studies
on Inversion Relations (see Fig. 9)
B2A
B4A*
S3A*
MU
SiQ2
54.82
48.9
50.9
49.72
AI2O3
1.87
1.5
1.8
0.90
Fe203
1.22
2.1
0.7
1.72
FeO
11.49
28.7
25.1
27.77
MnO
0.28
0.6
0.3
0.98
MgC
28.71
13.8
16.4
12.69
CaO
1.44
3.6
4.2
3.80
Na20
0.07
0.2
0.1
0.23
K20
0.02
0.1
...
0.12
H20+
n.d.
n.d.
n.d.
1.27
H20"
0.02
...
n.d.
0.08
TiQ2
0.21
0.5
0.5
0.85
Cr203
0.14
n.d.
n.d.
n.d.
Totals
100.29
100.0
Atomic %
100.0
100.13
Ca
2.8
7.6
8.9
8.4
Mg
77.7
40.7
48.2
39.2
Fe
19.5
51.7
42.9
52.4
* Recalculated to 100% after subtraction of
coexisting analyzed augite impurity in material
used for pigeonite analysis.
B2A. Bronzite (SA 722), Main Zone gabbro,
Bushveld intrusion (Atkins, 1965, Table 5).
B4A. Inverted ferropigeonite (SA 616), Up-
per Zone ferrodiorite, Bushveld intrusion (Atkins,
1965, Table 5).
S3A. Inverted pigeonite (EG 4341), Middle
Zone gabbro, Skaergaard intrusion (Brown,
1957, Table 1).
MU. Pigeonite from andesite (inninmorite),
Mull, Scotland (Hallimond, 1914).
350
CARNEGIE INSTITUTION
o
1500
1400
1300
: : :
Q.
100
LIQUID
a Liquid
a Liquid + Cpx
■ Cpx [►From Opx
H Cpx + Opx
□ Opx
H Opx + Cpx ]
b Opx f From Cpx
Clino - Px
1000
900 -
800
Pressure, kb
Fig. 10. P-T diagram for the composition Wo8.9Eri48.2 Fs42.9 (mole %), a pigeonitic pyroxene
from the Bushveld intrusion (see Fig. 9 and Table 1 2, No. B4A). Recalculated Ca free, the compo-
sition is En40Fs6o (w* %)• The starting material (labeled Opx) for inversion to the higher-temperature
phase was in the form of orthorhombic hypersthene (Pbca) with augite lamellae. The phase labeled
Cpx is here viewed as a likely high-temperature, stable clinopyroxene phase (see text). The inver-
sion from Cpx to Opx was occasionally achieved, with the use of material from the Opx — > Cpx
inversion, but the difficulties of achieving complete reaction and of overcoming the sluggishness
of the Opx — > Cpx reaction at lower temperatures have caused the drawing, tentatively, of a band
representing rate-controlled reaction below the maxima line. Experiments at zero pressure per-
mitted reaction to proceed easily in the presence of an andesitic liquid, both from Opx — > Cpx
and, with the use of a natural monoclinic pigeonite, from Cpx — > Opx.
from runs made hydrothermally at 2
kb, together with those at zero pres-
sure (see below), are also shown in
Fig-. 10 and indicate that the slope,
rather than the temperature level of
the maxima curve, is most in need of
further confirmation.
Experiments at 2 kb were per-
formed by Dr. D. H. Lindsley in an
externally heated pressure vessel of
molybdenum alloy (Climax Molyb-
denum Corp., alloy TZM), using ar-
gon as the pressure medium. Charges
were sealed with water in silvers-pal-
ladium^ tubes and buffered for f0o
with the f ayalite-quartz-magnetite as-
semblage. An unexplained failure of
the pressure vessel, well within the
published elastic limit, precluded
further experiments at 2 kb.
Experiments at Zero Pressure
In addition to observations on B4A,
studies were made on another "in-
verted pigeonite" (S3 A), a mono-
clinic pigeonite (MU) equivalent in
composition to B4A, and a rhombic
bronzite low in calcium (Fig. 9 and
Table 12). Oxidation was restricted
by containing the sealed Pt tube, with
GEOPHYSICAL LABORATORY 351
iron filings in Pt foil, in an evacuated, rock of the Sandwich Horizon ("last
sealed silica-glass tube; the runs were liquid" EG 4330) .
made at various temperatures by Dr. A melting temperature curve was
J. F. Schairer. drawn between the two end members
Sluggish reaction was confirmed in (1230° and 1020 °C, from Year Book
preliminary runs without a flux, spe- 62, p. 80, and Year Book 63, p. 94),
cimen B4A showing the rhombic -> a general linear relationship being
monoclinic inversion only above verified by obtaining the liquidus
1200 °C, in runs lasting 7 to 17 days, temperature of a 50:50 mixture
Extrapolation from the high-pressure (1120°C) . For the 20 : 80 mixture, the
data (Fig. 10) indicates that changes liquidus temperature obtained by ex-
at zero pressure under these condi- periment (Cpx appearing at 1060 °C)
tions are dependent upon long reae- was very close to the interpolated
tion times, and observations made temperature (1065°C). Exploratory
after short heating are therefore mis- runs were also made on the Mull an-
leading. desite (M 244) containing plagioclase
Under natural conditions, reactions and pigeonite phenocrysts (MU),
must have occurred in the presence of and in these, plagioclase appeared on
magmatic fluxes. What these fluxes the liquidus at 1070 °C and Cpx at
are would be difficult to specify for 1050 °C. From the results it seems
relatively dry magmas, but the fluxes clear that the pigeonites under con-
can be made to participate in the sideration (MU, S3 A, and B4A, Fig.
experiments as molten rock. The Mull 9) crystallized from an andesitic (or
lava containing the pigeonite pheno- ferrodioritic) magma at about 1050°-
crysts (MU) is a useful example, and 1060 °C and that, therefore, inversion
whole-rock powder was used in this studies in this magmatic medium are
study, together with a few milligrams relevant to lower temperatures at
of previously separated pigeonite atmospheric pressure. The Skaer-
crystals added to aid identification gaard and Bushveld inverted pigeon-
and reduce the proportion of liquid ites were added to the 20:80 Skaer-
available for precipitation of extra gaard-mixture glass in both cases, in
phases (materials provided by Dr. view of the closely similar mineralogy
H. S. Yoder, Jr.) . In order to examine and chemistry of the two intrusions
the behavior of the "inverted pigeon- and the absence of estimates for the
ites" (B4A and S3A), whole rocks Bushveld successive magmas. All runs
cannot be used because they are were made for 7 days, with the sealed
crystal accumulates, free from crys- silica-tube technique. Refractive in-
tallized contemporary magma. In each dex and X-ray diffractometer meas-
case, however, the pigeonites (both urements on products for which re-
inverted and unin verted) apparently action was not achieved were com-
crystallized from liquids formed close pared with the starting materials; the
to the Middle Zone stage of f ractiona- absence of significant compositional
tion established for the Skaergaard change indicated lack of extensive re-
intrusion (Wager and Brown, 1967). action with the experimental immer-
This liquid, equivalent to the ferrodi- sion liquid, and therefore further in-
oritic residue after 80 % of the intru- dicated that the chosen mixtures have
sion had crystallized (Wager, 1960, a composition appropriate to the
Table 4), was reproduced by fusing natural conditions of pigeonite equi-
a mixture of 20% of the Skaergaard libria.
chilled rock ("first liquid" EG 4507) For the two "inverted pigeonites"
with 80 % of the last-formed residual (S3 A and B4A), inversion to Cpx
352
CARNEGIE INSTITUTION
took place easily, at much lower tem-
peratures than in the absence of liq-
uid. Both inverted completely at
about the same temperature, slight
differences in Mg:F6*+ ratio (Fig*. 9)
not having a detectable effect. Spec-
imen B4A inverted at 1000 °C but
showed no reaction at 990 °C. This
observation agrees with results at
high pressure (Fig. 10); in contrast,
inversion occurred only at tempera-
tures above 1200°C in the absence of
liquid. Pigeonite MU, already in the
monoclinic form {P21/c) attained by
B4A and S3A on inversion, provided
a ''reversal" pertinent to the question
of pigeonite inversion in the natural
state. The reaction was sluggish: only
a trace of rhombic pyroxene (Pbca)
was produced at 980°C, but at 970°C
more than half the monoclinic phase
had inverted to rhombic pyroxene.
Distinct augite blebs were seen in the
monoclinic as well as in the rhombic
hypersthene phases produced at 980°
and 970 °C (but not in the monoclinic
phase at 1000°C), indicating exsolu-
tion prior to inversion, with lowering
temperatures.
Yoder, Tilley, and Schairer (Year
Book 62, pp. 89-90) suggested inver-
sion of the bronzites (Fig. 9) from
protobronzites on cooling, although
experimental work involving protoen-
statites has shown metastable inver-
sion to clinoenstatite on quenching
(Boyd and Schairer, 1964) or to a
disordered enstatite (Brown and
Smith, 1963), and not to the type of
rhombic pyroxene characterizing the
natural bronzites (Plate 1A). The
magnesian bronzites shown in Fig.
9 (SI A, B2A), which crystallized as
early phases from the Skaergaard and
Bushveld magmas and which are very
similar in composition and coexist-
ence relations with augite, were stud-
ied for inversion relations. In view of
our better knowledge of the Skaer-
gaard liquids, the "first liquid" was
used as an immersion medium by fus-
ing the chilled facies (EG 4507).
Melting data on this rock ( Year Book
62, p. 80) show plagioclase on the
liquidus at 1230 °C, with Px entering
at 1190 °C. Present runs on the bronz-
ite crystals (plus liquid) failed to
produce inversion to a higher temper-
ature form even at 1230 °C, although
similar run times (7 days) were
ample for complete reaction of the
pigeonites. According to optical and
X-ray powder methods of identifica-
tion, no criteria indicated the forma-
tion of protobronzite or a monoclinic
quenching product, and there was no
detectable compositional change.
Hence it is proposed that the bronzite
group shown in Fig. 9, differing in
texture and composition from both
the inverted protoenstatites (Dall-
witz, Green, and Thompson, 1966)
and the inverted pigeonites (Plate 1) ,
crystallized directly in the rhombic
form.
Comments on Pigeonite Stability
At low pressures and high tempera-
tures the enstatites appear to exist
in the proto form (Pbcri), bronzites
in the rhombic form (Pbca)f and
pigeonites in the monoclinic form
(P2t/c). Associated with this subdi-
vision, based on Mg/Fe2+ ratios, is a
steplike increase in the Ca contents
of these three phases where crystal-
lized from natural magmas.
Studies on the pigeonitic pyroxenes
have shown inversion relations such
that a monoclinic phase appears to be
stable at a higher temperature than
the rhombic phase. The evidence
could be interpreted as showing for-
mation of Cpx as a quenching phase,
but pyroxenes of composition almost
identical with those obtained above
the inversion curve (Fig. 10) and
with the same space-group symmetry,
occur naturally in varied environ-
ments and invert stably, in the experi-
ments, to rhombic pyroxenes below
the inversion curve. The observation
GEOPHYSICAL LABORATORY S53
of both exsolution and homogeniza- favoring the higher oxygen coordina-
tion in the experiments is also com- tion around Ca for pigeonitic com-
parable with the observations relating pared with hypersthenic composi-
subsolidus exsolution with inversion tions, but it has not been found in
textures in natural materials. From natural rocks,
experiments at zero pressure, tenta-
tive values suggest that the examined Electron Probe Study of
pigeonites (Fig. 10) crystallized at Exsolution in Pyroxenes
about 1050 C, and that some began
to exsolve augite at about 1020 °C, F.R.Boyd and G.M.Brown
and inverted to rhombic hypersthene Calcium-rich and calcium-poor py-
at about 990 °C. roxenes occurring together in many
The results from high-pressure terrestrial and meteoritic rocks are
runs suggest that the slope of the of great interest to petrologists, since
P-T curve is only slightly pressure their contrasted compositions are re-
dependent but that it is positive in lated to the temperatures and, to a
the same way as the low clinopyrox- lesser exent, the pressures of crystal-
ene ^ rhombic pyroxene curve (Year lization.
Book 61+, Figs. 33 and 51). Since this Large basic igneous intrusions,
implies, in the latter case, that low such as the Bushveld, Skaergaard,
clinopyroxene has a higher density and Stillwater, contain very similar
than rhombic pyroxene (Stephenson, assemblages of coexisting pyroxenes,
Sclar, and Smith, 1966) , then (by although in detail the Bushveld pyrox-
the same argument) the high clino- enes show a slightly wider miscibility
pyroxene has a lower density than gap (Brown, this report, Fig. 9) and
the rhombic phase. the whole of the pyroxene-bearing
For MgSi03 the slope of the proto rocks probably crystallized at slightly
^ rhombic inversion curve shows greater depth than the Skaergaard
that dT/dP equals 80°C/kb (Year (Lindsley, Year Book 65, p. 232). All
Book 6Jf, Fig. 33), whereas for the pyroxenes, within the composi-
FeSi03, dT/dP equals 10°C/kb (Year tional range where two coexist, also
Book 6U, Fig. 51) . This may be used contain exsolution lamellae of another
to argue against the proposal that the pyroxene phase, and these lamellae
high-temperature stable phase of are generally coarser for any particu-
FeSi03 has a proto form. If so, linear lar composition in the Bushveld than
T-X relationships between MgSi03 in the other two intrusions,
and FeSi03 at various pressures Studies on the compositions of some
(extrapolating into metastable re- lamellae and host pyroxenes are now
gions) would indicate much higher practicable, with the electron probe,
temperatures for a proto phase at and are expected to yield valuable
En40Fs6o than is indicated by Fig. 10. information on the development of
The Ca content of the pigeonites is exsolved lamellae of various thick-
likely to have lowered inversion tern- nesses and orientations at different
peratures but not to the extent esti- temperatures. In addition, the infor-
mated here (about 140°C at 5 kb). mation gained from probe traverses
Hence, the possibility of stability of of lamellae and host will be impor-
protopyroxenes with pigeonite com- tant in considering (1) the extent of
positions is believed to be extremely cation migration and ordering on the
slight. A phase with pyroxenoid struc- scale of a few microns; (2) the dis-
ture may possibly be stable at the tribution of a wide range of chemical
high temperatures and low pressures elements between the distinct crystal
Q E
54 CARNEGIE INSTITUTION
structures encountered in augite ex- discussed below, that in relation to
solution from hypersthene (and vice the beam size we can resolve only
versa) ; and (3) the relationship be- those lamellae that are thicker than
tween exsolution and inversion in the about 4-5 fi. Viewing the scans A-C
pigeonites, clinohypersthenes, and together, there is a definite propor-
rhombic hypersthenes. tionality between thickness of lamel-
As a preliminary to quantitative lae and apparent calcium content
studies, we have made a series of (CaKa X-ray intensity) as a func-
qualitative scans across grains show- tion of peak height. Scans B and C,
ing a great variety of exsolution tex- particularly, also provide a visual
tares. These serve as a basis for dis- demonstration of the spacing and fre-
cussion of the limitations imposed on quency of alternating lamellae of
quantitative analysis, according to varying widths.
thickness and internal homogeneity of Specimen A is a bronzite with 1-2 i<,
the lamellae, and of the usefulness of lamellae (Brown, this report, Plate
qualitative and semiquantitative ex- 1A), which are clearly too thin for
animations of both compositions and complete resolution. However, two
periodicity of lamellae within a par- points of particular interest emerge
ticular host phase. from the scan picture: first, that the
Analyses were made on polished lamellae are more calcic than the
thin sections of rocks containing host, and that therefore the texture is
pyroxenes of interest and on grain not of the type attributed to twinning
mounts. It was found essential to use by Henry (1942); and second, that
transmitted light with cross polariza- the proportion of lamellae to host is
tion to locate grains suitable for far too great for the lamellae to be
analysis. Scans were made for CaKa augitic in composition (cf. Hess,
at 30 kV and a specimen current of 1941), because wet-chemical analysis
0.01-0.02 /ia. The electron-beam di- shows that the crystals of host plus
ameter was determined by scanning lamellae contain only about 2% CaO
an 8-IO-/1 layer of silver laminated by weight. The bulk composition is
in steel. The silver-steel sandwich was such that it may lie to the (Mg,Fe)-
prepared by silver-plating two pol- rich side of the two-pyroxene field, in
ished steel blocks. The two plated which case we may be observing Ca
blocks were then clamped together, distribution between phases involved
and a surface was polished normal to in the polymorphic inversion of Ca-
the silver lamination. Scans of this poor pyroxenes.
sandwich, monitoring specimen cur- Specimens B-D are pigeonites, gen-
rent, indicated a beam size of 1.0 ± erally believed to have exsolved the
0.1 //.. The X-ray spot size is some- coarse augite lamellae prior to in-
what larger than this. Scanning with version to the present, orthorhombic
FeK« X rays gave a spot size of hypersthene host-phase. The lamellae
1.8 ± 0.2 p.. For CaK« X rays during are much thinner in the Skaergaard
analyses of the pyroxene lamellae, (B) than in the two Bushveld speci-
the beam was probably about 1.5 /x. mens (C, D), and the scans show that
Scans of the silver-steel sandwich an underestimate of Ca content would
are shown in Fig. ll(E-F). result from accepting any but, prob-
Scans across four orthopyroxene ably, the left-hand peak of scan B.
crystals from different gabbroic rocks Scan C shows good resolution of the
are shown in Fig. 11 (A-D), the ex- central part of the thick lamella
solved lamellae being a more calcic (27 /x). Scan D shows exsolution of
pyroxene in each case. It is clear, as relatively Ca-poor pyroxene within
GEOPHYSICAL LABORATORY
355
MICRONS
Fig. 11(A). Electron-probe scans across pyroxene crystals containing exsolved lamellae of
another pyroxene phase. (A) Bronzite (W80), Stillwater. (B—D) inverted pigeonites (hypersthene
+ coarse augite lamellae) from Skaergaard (EG 4330) and Bushveld (SA 616 and 1019),
respectively. (E) Scan across a silver lamination in steel, monitoring the specimen current. Slopes
of the sides of the peak indicate a beam size of 1.0—1.1 /x. (F) Scan across the same lamination
shown in (E) monitoring the FeKa X rays. X-ray spot size is 1 .6—1 .9 jx.
356
CARNEGIE INSTITUTION
the Ca-rich lamella. Exsolution of within lamellae and within host (see
coarse lamellae was apparently fol- small peaks, scan C).
lowed by more cation ordering in Scans across four augite crystals
response to further cooling, not only are shown in Fig. 11 (G-J) , where
at host lamella contacts but also again the proportionality between
Fig. 1 1 B. [G-J) Augites from Bushveld (SA 1019, 11 39, 61 6, another crystal from 1 01 9). Speci-
men numbers cross-refer to Brown (1957) and Atkins (1965), except for W80 from a bronzitite
collected by Brown from the Ultramafic Zone of the Stillwater intrusion.
GEOPHYSICAL LABORATORY 357
lamellar width and CaKa intensity is Specimen SA 1019 (see scan J,
clearly demonstrated. In this case the Fig. 11) was selected for semiquan-
lamellae are of a Ca-poor pyroxene, titative analysis of Ca, Mg, and Fe
of the monoclinic variety recorded by content in the host and lamellae of
Bown and Gay (1957) in specimens augite crystals having coexisted at
with similar optical properties. Only solidus temperatures presumably
specimens from the Bushveld gabbros with compositions close to augite for
have lamellae of sufficient width to the host and pigeonite for the lamel-
permit complete resolution. Scans G lae. The bulk compositions have not
and H show that the thicker the lamel- yet been established by wet-chemical
lae, the lower the calcium counts, and analysis, but texturally similar pyrox-
both show the presence of thin Ca- enes from the Bushveld, Skaergaard,
poor lamellae, evenly spaced between and Stillwater intrusions have au-
the thick ones. In fact, the pattern for gitic and pigeonitic bulk composi-
the 30-ju zone of augite host in scan tions. Because of the extreme inhomo-
G is very similar to that for the 60-//, geneity of the pyroxenes in SA 1019,
augite lamella in scan D (both from it would be impossible to establish
the same rock) . Scans H and / show bulk compositions by microprobe
a marked regularity in the spacing analysis.
of Ca-poor lamellae along the trav- Semiquantitative analysis of lamel-
erses and, in the case of /, in the thick- lae and hosts for Ca, Mg, and Fe
ness of lamellae as well. Both features were made at 20 kV and 0.05 ju,a.
indicate the future usefulness of About 10 points were counted for
probe traverses in considering the each analysis. Intensities have been
magnitude of cation diffusion within corrected only for background and
pyroxene structures at varying tern- drift. These analyses are intended to
peratures. In scan / the profile of a be a reconnaissance. They will be f ol-
very thick lamella shows a definite lowed by more complete analyses with
compositional gradient at the edges evaluation of fluorescence and absorp-
and heterogeneity within the central tion effects.
part. Scans of the silver-steel sand- The compositional data are shown
wich (Fig. 11, E, F) indicate that the on Fig. 12, not plotted as points be-
resolution of these pyroxene lamellae cause of the semiquantitative nature
should be better than they appear, of the measurements (particularly re-
It is probable that a part of what ap- garding Mg contents) but in areas
pears to be imperfect resolution is denoting the probable limits of error,
due to compositional zoning within The augite host and the augite lamel-
the lamellae. lae in the hypersthene host both fall
It may be noted that the thicker, within boxed area A, whereas the hy-
monoclinic lamellae (>5 //,) in the persthene host and the clinohyper-
augite scans are very poor in calcium, sthene lamellae in the augite host both
Such lamellae, oriented parallel to fall within boxed area B. The zone of
(001) of the augite host, have gen- possible tie lines (dotted) is reason-
erally been assumed to be pigeonitic ably oriented in relation to those
in composition (after Hess, 1941). shown for bulk wet-chemical analyses
Binns, Long, and Reed (1963) of comparable Bushveld and Skaer-
showed from electron-probe analysis gaard pairs. In regard to the more
that these lamellae may approach important Ca contents attained at
clinohypersthene in composition, and subsolidus temperatures, relative to
we have confirmed this finding for those indicated for solidus tempera-
one of the Bushveld gabbros. tures by the now well-established
358
CARNEGIE INSTITUTION
trend lines shown in Fig*. 12, the fol-
lowing conclusions can be drawn.
The augite is unlikely to have ini-
tially exsolved clinohypersthene, since
the separate phase of pigeonitic bulk
composition has changed to clinohy-
persthene composition only after ex-
solution of much augite and after this
augite has attained the high Ca con-
tent of the host material in the sep-
arate augitic phase. Assuming augitic
and pigeonitic initial bulk composi-
tions, it would seem that exsolution
in both phases probably began with
the formation of lamellae that in each
case were close to the other separate
phase in Ca content. Slow cooling
then led to gradual equilibration with
temperature; such that both the la-
mellae in augite and the host in pi-
geonite became clinohypersthene, both
being in equilibrium with a highly
calcic augite. Then the clinohyper-
sthene host inverted to orthorhombic
hyp erst hene. The clinohypersthene
lamellae failed to invert, possibly be-
cause of the stabilizing effect of the
monoclinic augite host structure. Evi-
dence for such a sequence is ex-
tremely important in regard to the
concept that the rhombic hyper-
sthenes have inverted from a mono-
clinic phase, of pigeonite composition
prior to augite exsolution. At even
lower temperatures, in some cases,
further exsolution on a fine scale oc-
curred in all four phases, i.e., in the
rhombic hypersthene (Brown, this
report, Plate 1C), and augite lamel-
lae (Fig. 11, D) , and in the augite
host (Fig. 11, G, H, J) and clinohy-
persthene lamellae (Fig. 11, J).
Estimates of temperatures will be
difficult to make for these complex
series of events in igneous pyroxenes,
particularly since the lower-tempera-
ture part of the range can be obtained
only from the compositions of the
very thin lamellae. It has been esti-
mated experimentally (Brown, this
CaMgSi206
CaFeSigOg
0
20 30 40 50
Mole per cent FeSiC^
Fig. 12. Compositions of augite host and augite lamellae in hypersthene (A), and hypersthene
host and clinohypersthene lamellae in augite (B), by electron-probe analysis of coexisting pyroxene
pair in Main Zone gabbro (SA 1019), Bushveld intrusion. Boxed areas represent probable limits of
error in semiquantitative measurement, and dotted area, the zone of possible tie lines. Composi-
tions of augite and pigeonite prior to exsolution would fall on trend lines within dotted area, by
analogy with collected data from other Bushveld and related pyroxene pairs (Brown, 1957;
Atkins, 1965).
GEOPHYSICAL LABORATORY
359
report), however, that for the Bush-
veld and Skaergaard intrusions, co-
existing pyroxenes probably crystal-
lized between about 1200 °C (augites
and bronzites) and 1050 °C (augites
and pigeonites), from Mg-rich to
more Fe-rich pairs, and that in the
pigeonites the exsolution prior to in-
version took place between about
1020° and 990 °C. Higher pressures
are shown to increase these tempera-
tures very little. Hence it would seem
that for the pair considered here we
are dealing with extensive cation mi-
gration (20-30 /x) within pyroxene
crystals, between about 1050° and
1000 °C, and that migrations over 2-p
lamellar distances and reequilibration
between coarser lamellae and host
took place below 1000 °C. For the
more magnesian bronzites and au-
gites, crystallization temperatures
were higher, and the fine scale of the
exsolution lamellae suggests a greater
temperature interval between crys-
tallization and the initiation of ex-
solution.
Hedenbergites s- Wollastonitess
Inversion in a Skaergaard
Pyroxene
D. H. Lindsley, G. M. Brown, and
I. D. Muir*
The rocks formed during the late
stages of fractionation of the Skaer-
gaard intrusion contain hedenberg-
itic pyroxenes (Hdss), some of whose
textures suggest that they have in-
verted from ferriferous wollastonite
solid solutions (Woss) (Wager and
Deer, 1939, p. Ill; Brown and Vin-
cent, 1963, pp. 186-190). Work on
the hedenbergite-Woss inversion in
synthetic Ca-Fe pyroxenes last year
led to the suggestion that the pres-
sure dependence of this reaction
might yield the pressure at which
the late-stage products of the Skaer-
gaard differentiation had crystallized
* Cambridge University, England.
(Year Book 65, p. 232). In this
section we report experiments on an
analyzed, inverted ferriferous wollas-
tonite (EG 4471) from the Skaer-
gaard intrusion. These results,
combined with data on the quartz-
tridymite inversion (Kennedy, Was-
serburg, Heard, and Newton, 1962),
indicate that some Upper Zone (UZc)
rocks of the Skaergaard intrusion
crystallized at 600 ± 100 bars, over
a temperature interval that spanned
the value 970° ± 20 °C.
The Upper Zone of the Skaergaard
layered series of crystal accumulates
consists of f errodiorites showing pro-
gressive iron enrichment with height.
The analyzed f erroaugites from these
rocks plot on an extension of the reg-
ular trend shown by the augites ana-
lyzed from underlying zones, the com-
plete trend ending very close to the
CaSi03-FeSi03 edge of the common
pyroxene quadrilateral, with a heden-
bergite of composition Wo42.5Eno.4
Fs57.i (mole %). This brown pyrox-
ene occurs in the last demonstrable
differentiate (EG 4330; see Fig. 13)
of the intrusion and shows no tex-
tural evidence of inversion from a
ferriferous wollastonite. In rocks
slightly lower in the Upper Zone
layered series (Fig. 13), occur green
hedenbergites with the mosaic texture
ascribed to inversion from ferrifer-
ous wollastonites. The green heden-
bergite forming the subject of this
study, from Skaergaard specimen
EG 4471, is the richest in ferrosilite
of the pyroxenes reported by Brown
and Vincent (1963, Fig. 3). Between
the levels of EG 4471 and 4330, rocks
such as EG 1881 contain green pyrox-
ene rimmed by or intergrown with
the brown variety, thus providing a
link between the two types. Note that
EG 4471 and 1881 are crystal accu-
mulates, whereas EG 4330 probably
represents a liquid. Chemical analy-
ses, plotted positions within the py-
roxene quadrilateral, and photomicro-
360
CARNEGIE INSTITUTION
"^1 I Specimen Calcic Pyroxene
IS
No.
Ca
Mg Fe
«
2400-
Z3CC
2100-
|| 3021 43.8 — 5.7—50.5
I I
4330 42.5 — 0.4— 57.1 ■
- 1881 42 2 — 1.9— 55.9 —
_ 4471 403 — 1.2— 58.5 —
— 4318 41.5 —10.7—478 —
— 4316 40.0 —21.0 — 390 —
Olivine
Fo
-—5 -•
— 0 —
— 0 —
— 2 —
Pyroxene
Type
Silica
Polymorph
Hdss Trid, Qtz
Hd„ Qtz
^c
-Hdss,Woss--Qtz
— Woss Qtz
— Hdce Qtz
•24
o
-♦—
to
>\
O
a>
Q
Fig. 13. Distribution of specimens in part of the Upper Zone and Upper Border Group of
the Skaergaard intrusion. Adapted from Brown and Vincent (1963, Fig. 1). Compositions of
calcic pyroxenes in atomic %. Hdss, hedenbergite solid solutions; WoSS/ ferriferous wollastonite
solid solutions, now inverted to green Hdss; Qtz, quartz; Trid, quartz pseudomorphs after tridy-
mite.
graphs of these three examples are
given by Brown and Vincent (1963).
Analyses of similar examples of in-
verted ferriferous wollastonites are
also given by Muir (1951), his
EG 1974 (2350 m) and EG 4143
(2375 m) being comparable with
EG 4471. (Structural heights given
for Muir's specimens have been re-
vised by Wager and Brown, 1967.)
Because both phases in the Hdss-
Woss inversion are solid solutions, the
inversion must take place over a tem-
perature interval, except, of course,
for the special case of composition at
a maximum or minimum in the inver-
sion temperature. The departure of
the EG 4471 inverted Woss composi-
tion from the trend line for brown
hedenbergites (Brown and Vincent,
1963, p. 181, Fig. 2) is consistent
with the existence of an inversion in-
terval during crystallization of Upper
Zone c. Furthermore, the texture of
EG 1881 may well reflect concurrent
crystallization within the inversion
interval, rather than the rimming of
early green Wo,, by later brown Hd8S.
Inasmuch as EG 4471 must have
crystallized at temperatures above,
and EG 4330 below, the inversion in-
terval, crystallization within the in-
terval must have been restricted to a
zone of rocks less than 180 m thick.
Rocks from Upper Zone c all con-
tain quartz as an intercumulus phase.
Rocks immediately overlying this
zone (Upper Border Group y, approx-
imately 100 m thick) contain quartz
pseudomorphs after tridymite, to-
gether with primary quartz (for ex-
ample, EG 3021, Brown and Vincent,
1963, Plate 2C). If both coexisting
pairs of minerals — Hdss-Woss and
quartz-tridymite — occurred in the
same layer, then at some time during
its crystallization history that layer
must have been at a pressure and
temperature corresponding to the in-
tersection of the appropriate inver-
sion curves in P-T space. As the co-
existing pairs occur at different
levels, an estimate of the pressure
and temperature differences between
the two levels will indicate how close
the crystallization sequence came to
the region of intersection.
Specimen EG 3021 occurs less
GEOPHYSICAL LABORATORY 361
than 100 m above EG 4330 and 200 m is a crystal accumulate.) We conclude
above EG 1881, corresponding to that the pressure difference between
maximum pressure differences of EG 1881 and EG 3021 was not more
about 30 and 60 bars, respectively, than 60 bars, and the "average" tem-
Crystallization at each level took perature difference was not more
place over a temperature range, of than 30 °C. Thus the intersection in
course, but it is nevertheless mean- P-T space of the quartz-tridymite and
ingful to estimate differences in the Hdss-Woss inversion curves will pro-
temperatures at which two layers vide useful information on the crys-
achieved equivalent proportions of tallization parameters of Upper Zone
crystals and liquid, or at which the c of the Skaergaard intrusion,
clinopyroxene now present in each To locate the appropriate HdS8-
layer first crystallized. We may refer Woss inversion interval for the Skaer-
to these as "average" temperature gaard we have investigated the pres-
differences. On the basis of the com- sure-temperature relations for the
positions of clinopyroxene and of green Hdss from EG 4471 at pres-
olivine, EG 3021 probably crystallized sures from 0 to 5 kb. The starting
at about the same time — and hence material was a split of the mineral
over approximately the same temper- separate analyzed by Brown and Vin-
ature interval — as rocks in the lay- cent; part of the material was in-
ered series between EG 4318 and verted to Woss in vacuum at 1000 °C
EG 4471 (Fig. 13). According to the to permit reversal of the reaction,
estimates of Wager (Wager and The experiments were greatly ham-
Brown, 1967, p. 169, Table 10), that pered by slow reaction rates. Evacu-
part of the layered series lying be- ated tube experiments held at tem-
tween EG 1881 and the equivalent of perature for 1 month show only
EG 3021 represents less than 1% of partial reaction. Experiments at 200-
the total volume of the layered series. 500 bars pressed the limits of the ap-
The crystallization interval for the paratus employed (a "Tuttle press"
entire layered series was probably with a pressure vessel of alloy R-41),
less than 300 °C (from the data of and each point in Fig. 14 represents
Tilley, Yoder, and Schairer, Year a compromise between desirable run
Book 62, p. 80; Year Book 63, p. 94). duration and the time a pressure ves-
Now, although it is unlikely that the sel would operate without failure
temperature difference corresponding (12-72 hours). In most cases it was
to 1 % of volume is that same fraction impossible to let the reductions go to
of the temperature interval (in this completion, with the result that most
case, 3°C), it is even more unlikely experiments yielded the two reaction
that the temperature difference products Hdss + Woss, and the place-
should be greater than 10%, i.e., ment of boundaries for the inversion
30 °C. A similar temperature differ- interval in Fig. 14 is subjective. Two
ence may be obtained by comparing useful but not definitive criteria
the liquidus temperatures of EG 4472 guided the placement of the bound-
(same level as EG 4471), i.e., 1035°C, aries. (1) Textures: Experiments
and of EG 4330, i.e., 1020°C {Year outside the probable inversion inter-
Book 62, p. 80; Year Book 63, p. 94). val show the stable phase as euhedral
(These samples have roughly the crystals replacing anhedral starting
same stratigraphic separation as do material, whereas experiments with-
EG 1881 and the layered equivalent of in the interval show euhedra of both
EG 3021. The 1035°C value for phases. (2) Further reaction of the
EG 4472 is high because that sample charge when returned to pressure and
362
CARNEGIE INSTITUTION
IIOO-
o
o
CD
Z5
"5
CD
Q.
E
|S
1000 -
900
0 12 3
Pressure, kilobars
Fig. 14. Hd88-WoS8 inversion for green pyroxene from sample EG 4471, Skaergaard intrusion.
Slow reaction rates prohibit exact determination of inversion interval, which is probably narrower
than is shown. Amount of shading approximately proportional to amount of Hdss in run products.
Wo.., ferriferous wollastonitess; Hd,., hedenbergitess; Trid, tridymite; Qtz, quartz. Height of sym-
bols reflects uncertainty in temperature of each experiment. Width of symbols reflects pressure un-
certainty at 5 kb but exaggerates the uncertainty at lower pressures. The encircled numbers show
schematically the inferred relations of samples EG 4471, 1881, 4330, and 3021 to the inversion
curves; see Fig. 13 for details of their stratigraphic relations. EG 3021 did not crystallize Woas
because the content of Mg stabilized clinopyroxene.
temperature: If repeated experiments
on one charge showed further forma-
tion of the product phase (e.g., WofiS
formed at the expense of HdKB), the
experimental conditions probably
were outside the inversion interval
and in the field of the product phase.
We suspect, but cannot prove, that
the inversion interval is narrower
than is shown in Fig. 14. The inver-
sion interval at the appropriate com-
position for the synthetic CaSi03-
FeSiOo system at 1 atmosphere is so
narrow that Bowen, Schairer, and
Posnjak (1933, p. 218) were unable
to detect it. Furthermore, the small
deviation of the composition of the
green pyroxene from EG 4471 from
the normal Skaergaard trend line
(Brown and Vincent, 1963, Fig. 2)
suggests that the inversion interval
is narrow in T-X section. The higher
temperatures for the beginning of in-
version in vacuum obtained by Yoder,
Tilley, and Schairer {Year Book 62,
p. 91) are probably to be attributed
to the shorter duration (7 days) of
their experiments. Our reversed point
within the interval at 940 °C is based
on 30-day experiments.
GEOPHYSICAL LABORATORY 363
Several methods were employed to of the experiments on synthetic
maintain the oxidation state of the pyroxenes reported last year. The
iron in the samples for experiments shift of composition and the low /o2,
at different pressures. (1) For zero- combined with fairly large pressure
pressure experiments the charges uncertainties in the solid-media pis-
were wrapped in Ag70Pd30 foil and ton-and-cylinder pressure apparatus
then sealed in evacuated silica-glass at 5 kb, require that the 5-kb results
tubes. The solubility of iron in the be used with caution,
silver-palladium alloy is very low The results of the experiments on
under these conditions (Muan, 1963), pyroxene from EG 4471 are shown
and the foil prevents reaction be- in Fig. 14. The Hd,<;-WoK3 inversion
tween the charge and the silica tube, interval intersects the quartz-tridy-
The oxygen fugacity (/o2) of the ex- mite curve in the region 980°-950°C,
periment is that of the charge. (2) at 600-450 bars. From its texture and
For 200- to 500-bar experiments the mineralogy, EG 1881 must lie within
charges were sealed in silver-palla- the Hdss-Woss inversion interval and
dium capsules with 5% -10% water at pressures above the quartz-tridy-
by weight; these capsules were then mite inversion. EG 3021 must strad-
sealed in larger gold capsules with a die the quartz-tridymite inversion
fayalite-magnetite-quartz (or tridy- curve and lie at temperatures not
mite) buffer plus water. Because the more than 30° above and at a pres-
pyroxene from EG 4471 originally sure not more than 60 bars below EG
coexisted with a silica mineral, faya- 1881. The relations of these samples
lite, and titaniferous magnetite, the to the inversion curves must there-
/o2 imposed by the buffer must closely fore be approximately as shown sche-
approximate that prevailing during matically in Fig. 14. Thus crystalliza-
crystallization of the Upper Zone of tion of Upper Zone c of the Skaer-
the Skaergaard. (If the activity of gaard intrusion must have taken
Fe304 is proportional to its mole frac- place at 600 ± 100 bars and over a
tion in titaniferous magnetite, the re- temperature interval that included
suiting /o2 is less than half an order the range 990°-950°C. We can esti-
of magnitude below that for the FMQ mate a third important parameter by
buffer.) The fugacity of water in the extrapolating data obtained from
experiments is probably higher than magnetite-ilmenite relations parallel
that during crystallization of the to the FMQ buffer curve (Buddington
rock, but as no hydrous minerals and Lindsley, 1964, p. 331) to this
form in the charge, this is not a se- temperature range and obtain an /o2
rious drawback. (3) For experiments value of 10~12 atmospheres. A similar
at 5 kb the charges were confined in value is obtained by correcting the /o2
metallic iron capsules. Some reaction of the fayalite-magnetite-quartz as-
between ferric iron in the charge semblage for the decreased activity
(1.25 wt %) and the iron capsule un- of Fe3Oi in titaniferous magnetite,
doubtedly took place, with a resulting ^ „
increase in the FeO content of the Subsolidus Relationships in Part
charge (from 30.78 to about 32.5 of the Hedenbergite-Ferrosilite
wt % if all the ferric iron was re- JolN AT Low PRESSURES
duced) . The resulting /o2 would ap- D. H. Lindsley and J. L. Munoz
proximate that of the fayalite-quartz- Extreme iron enrichment by frac-
iron buffer, well below that of tional crystallization of basaltic
methods 1 and 2, but similar to that magma yields hedenbergitic clinopy-
364
CARNEGIE INSTITUTION
roxenes (Hd,<) whose compositions
lie near the hedenbergite-ferrosilite
(Hd-Fs) join of the pyroxene quad-
rilateral. In the Skaergaard and
Bushveld intrusions the last crys-
tallizing clinopyroxenes (which co-
exist with fayalite + quartz) have
the compositions ,\Yo.!C.r>En0.(Fs:,:.1 and
Wo _ -Fn -Fs < (mole %) , respec-
tively (Wager and Brown, 1967). In
contrast, Bowen, Schairer, and
Posnjak (1933), in their pioneering
study of the system CaO-FeO-Si02,
reported that the Hdss in equilibrium
with fayalite + tridymite at 980°C
has the composition Wo2SFs:2 (mole
% ) . In view of this discrepancy, and
because techniques of controlling tem-
perature and the oxidation state of
iron have greatly improved since the
original experiments were made, we
have reinvestigated the subsolidus
relations of part of the hedenbergite-
ferrosilite join at pressures from 0
to 2 kb. It appears that a portion of
the Bowen-Schairer-Posnjak meta-
silicate diagram reflects metastable
conditions ; we find that the composi-
tion of synthetic hedenbergitess in
equilibrium with fayalite + quartz is
very close to that from the iron-rich
portions of the Skaergaard and Bush-
veld intrusions.
The reaction pertinent to this prob-
lem is
Fs-rich hedenbergite = Fs-poor
hedenbergite + fayalite + quartz (1)
As a sidelight to his study of ferro-
tremolite, Ernst (1966, p. 48) studied
this reaction, using several oxygen
buffers in hydrothermal experiments.
His experiments, in which he used
fayalite - magnetite - quartz (FMQ) ,
wiistite-magnetite (WM), and iron-
magnetite (IM) buffers, all yielded
the assemblage hedenbergiteHH +
fayalite + quartz. For these buffers,
reaction 1 is independent of oxygen
fugacity (/o2), inasmuch as no
oxygen is consumed or evolved. (For
values of /0., outside the stability
range of fayalite, oxygen takes part
in the reaction, and the reaction is
no longer /Oo-mdependent.) Ernst
pointed out that his were synthesis
experiments, and as equilibrium had
not been demonstrated, the results
were to be viewed with reservation.
With Ca-Fe pyroxenes synthesized
at high pressures, we have demon-
strated equilibrium by reversing re-
action 1 in the range 600°-900°C at
pressures of 1 and 2 kb, using the
FMQ and WM buffers to control /o2.
Durations of the experiments ranged
from several hours at 900 °C to more
than a month at 600° C. Compositions
of the resulting hedenbergitess were
determined from the 221, 310, and
311 spacings. Our results, together
with Ernst's combined FMQ-WM-IM
curve for comparison are superposed
onto the Bowen-Schairer-Posnjak
diagram (Fig. 15). Ernst's curve lies
at more iron-rich compositions, and
we tentatively suggest that it may rep-
resent a metastable solvus between
hedenbergite and ferrosilite. We plan
to test this idea by performing
buffered hydrothermal experiments
at higher pressures where iron-rich
pyroxenes become stable.
The discrepancy between our re-
sults and those of Bowen, Schairer,
and Posnjak might result from
several factors:
1. Differences in /0). Our experi-
ments were performed at the /o2 of
the FMQ and WM buffers, whereas
theirs were in equilibrium with metal-
lic iron and thus probably near the
fayalite-quartz-iron buffer. As the
ferric iron contents of all phases in
both sets of experiments are low,
however, the difference in /o2 should
have little effect.
2. Differences in pressure. The
Bowen-Schairer-Posnjak experiments
were performed at 1 atmosphere total
pressure, and ours, at 1 and 2 kb. Re-
GEOPHYSICAL LABORATORY
365
1400
1200
1000
o
QJ
CD
£
|2
800
600
400
200
Woss+Trid+01
Trid 4- Liq
Trid+01+Liq
\ Hdss+ Trid + 01
\
®
O/ <
Hd + Fay + Qtz-HHd«
A*
of Fs- poor Hdss+ Fay 4- Qtz
FMQ Buffer
ss
<3
Buffer _
\Af
50
CaFeSi20s
40
30 20 10 0
Mole per cent CaSiC>3 FeSi03
Fig. 15. The join hedenbergite-ferrosilite at low pressures. Light lines, 1 -atmosphere data from
Bowen, Schairer, and Posnjak (1933, p. 213), in equilibrium with metallic iron; small amounts of
Fe203 are present in the liquids. Heavy lines, hydrothermal data obtained at 1 and 2 kb. Curves
representing the composition of hedenbergitess coexisting with fayalite + silica (equals reaction 1
of text) are (1) Bowen, Schairer, and Posnjak; (2) hydrothermal synthesis with FMQ, WM, and IW
buffers (Ernst, 1966); (3) reversed equilibrium at 1 and 2 kb, this report. Hd, hedenbergite solid
solutions toward FeSiOs; Fs, FeSiOs; Liq, liquid; Ol, olivine; Trid, tridymite; Woss, Fe-wollastonite
solid solutions. Width of each data symbol (triangle), shows uncertainty in composition based on
X-ray measurements; height of symbol is at least double the uncertainty in temperature.
action 1 is known to be pressure de-
pendent (Year Book 65, p. 232), but
the effect is so small that we are un-
able to see a difference between our
1-kb and 2-kb experiments. It is un-
likely, therefore, that the 1 -atmos-
phere data should be greatly differ-
ent from our results. The use of water
as a pressure medium in our experi-
ments should serve only to catalyze
the reaction and not to shift the
equilibrium, inasmuch as no hydrous
phases are formed.
3. Incomplete reaction due to short
run duration. The Bowen-Schairer-
Posnjak experiments lasted for 48
hours or less. We can duplicate their
results by crystallizing glasses in
vacuum for a few days; however,
pyroxenes whose compositions lie in
the Bowen-Schairer-Posnjak heden-
bergite field but in our hedenbergitess
+ fayalite + Si02 field show incipient
breakdown to fayalite + tridymite
when held for 4 weeks at 940 °C in
vacuum. For this reason we conclude
366
CARNEGIE INSTITUTION
that the Bowen-Schairer-Posnjak re-
sults relevant to reaction 1 represent
incomplete reaction, probably because
limitations of their apparatus —
especially control of temperature and
the atmosphere within the furnace —
precluded experiments of long dura-
tion.
Mossbauer Study of Synthetic
Ca-Fe Clinopyroxenes
Robert Dinido7i* and D. H. Lindsley
An important aspect of crystal
chemistry is the determination of the
environment of an atom in a partic-
ular structural site. Slight energy
differences associated with otherwise
similar sites may result in ordering of
the atoms occupying those sites. The
standard technique applied to this and
similar problems is X-ray crystallog-
raphy; inasmuch as X-ray diffrac-
tion shows the average electron den-
sity of a number of atoms, however, it
may yield the topology but not the de-
tailed environment of a particular
site. Studies employing the Mossbauer
effect can usefully supplement X-ray
data by providing information on the
energy associated with an individual
site.
The Mossbauer effect is "the reso-
nant scattering of gamma rays emitted
without recoil by nuclei at atoms
bound in solids" (Wertheim, 1964fr,
p. 253). For example, Co57 can decay
to an excited state of Fe57, which in
turn decays to the ground state by
emitting recoil-free gamma rays, par-
ticularly at 14.4 keV. A specimen con-
taining Fe:7 placed between the
Co"7 source and a radiation detector
may resonantly absorb these gamma
rays. In general, however, the binding
energies of the Fe"37 in the sample
are slightly different from those in
the source, and thus the resonant ab-
sorption energy of the sample differs
from the energy emitted by the
♦Woodstock College, Woodstock, Maryland.
source. The effective energy of the
gamma rays experienced by the
sample can be varied by moving the
source relative to the sample, and
thus one can plot the degree of ab-
sorption as a function of energy
(commonly expressed as a Doppler
velocity in millimeters per second).
The displacement of the position of
maximum absorption (or center of
split lines) from zero Doppler veloc-
ity is termed the isomer shift and is
a measure of the change in nuclear-
energy levels of the Fe57 atoms in
response to the immediate environ-
ment of the structural site. (For a
detailed review, see Wertheim,
1964&.)
We have applied Mossbauer ab-
sorption techniques to members of
the hedenbergite-ferrosilite series in
an attempt to estimate the relative
Fe2+ populations in the Mi and M2
sites of the crystal structure. A
similar study has already been made
in several laboratories on the en-
statite-ferrosilite series (Bancroft,
Burns, and Howie, 1967; Evans,
Ghose, and Hafner, 1967; Marzolf,
Dehn, and Salmon, 1967) . In contrast
to the enstatite-ferrosilite series,
where both the Ma and M2 sites are
in 6-fold coordination, hedenbergite
has the Mj in 6-fold and the M2 in 8-
fold coordination. One of the inter-
esting but unsolved problems in the
hedenbergite-ferrosilite series is the
manner in which the M2 site changes
from 8-fold to 6-fold coordination.
The energy source in our experiments
was Co57; the absorber was the nat-
urally occurring fraction of Fe57 in
the iron of the samples, which were
synthesized on composition at high
pressures and temperatures (Year
Book 05, p. 232).
When Fe11 is located in an aniso-
tropic environment that gives rise to
a nonvanishing electric field gradient
at the iron nucleus, a doublet absorp-
tion pattern is observed (Wertheim,
GEOPHYSICAL LABORATORY 367
1964a). Pyroxenes exhibit spectra B2 also agrees with a similar assign-
that can be resolved into four lines, ment, but the intensity ratio of Ax to
which we label Alf Blf B2, and A2 in Bx does not agree, for reasons dis-
the order of increasing Doppler veloc- cussed below,
ity, as shown in Fig. 16. The Fe11 in hedenbergite (Fsr,oWo50)
In accord with the known crystal is completely ordered, presumably
structures of the ferromagnesian on the smaller Mt site, as shown
pyroxenes the lines At and A2 are as- by the narrow lines of the two-line
signed to Fe11 on Mx and the Bx and absorption pattern. The positions of
B2 lines are attributed to Fe11 on M2 the hedenbergite absorption lines
by each of the three groups that have (Fig. 16) are closest to the Bt and A2
studied this series. Marzolf, Dehn, lines of the other spectra. As the
and Salmon (1967) found that the amount of iron in the pyroxene is in-
absorption patterns of two synthetic creased, both sites in the crystal must
Fs40En60 samples, one in the clino be populated by Fe11, and we obtain
form and the other in the ortho form, poorly resolved four-line spectra. On
were almost identical. This fact may the basis of ionic size a strong pref er-
be taken as support for the assump- ence of Ca11 for the larger M2 sites is
tion implicit in the following discus- expected and hence a smaller inten-
sion, namely, that the Mossbauer ab- sity for the lines assigned to M2. This
sorption is sensitive mainly to the would lead us to assign B2 to Fe11 on
immediate environment and that cer- M2 in agreement with the assignment
tain changes in second-nearest neigh- for ferromagnesian pyroxenes. But
bors can be ignored as a first approxi- the line Bx is in each case of greater
mation. The assignment of the A and intensity than Al9 and this greater in-
B lines in the ferrosilite-enstatite tensity argues for the assignment of
series implies that the preferred M2 Bx to Fe11 on Mi and Ax to Fe11 on M2.
site has associated with it a smaller Regardless of the assignment of At
electric field gradient at its iron and Blt the values of the isomer shift
nucleus than the Mx site. for all the pyroxenes are greater than
The positions of the corresponding 1.2 mm/sec relative to sodium nitro-
lines in the f errosilite-hedenbergite prusside, and the quadrupole splitting
series given in Table 13 are in agree- is of the order of 2 mm/sec. These
ment with a similar assignment. The values indicate that the Fe11 is in a
intensity ratio of peak A2 to peak high-spin state (Spijkerman, Ruegg,
TABLE 13.
Fe57 Mossbauer Abso
rption 1
Patterns
of (Ca.
;Fe2_.r
)Si206
Ai
Bi
B2
A2
Pos*
Wf
HtJ
Pos
W
Ht
Pos
W
Ht
Pos
W
Ht
Fs50Wo50
-0.11
0.33
20.4
2.20
0.33
20.9
Fs6oWo40
-0.33
0.35
7.9
-0.08
0.40
17.1
2.01
0.43
8.2
2.27
0.46
15.2
Fs7oWo3o
-0.29
0.37
6.6
-0.01
0.34
7.3
1.89
0.43
5.4
2.26
0.41
7.2
Fs90Wo10
-0.30
0.42
9.5
-0.02
0.53
11.4
2.01
0.66
9.2
2.34
0.43
10.4
Fsioo
-0.30
0.37
4.9
-0.04
0.30
5.1
1.93
0.29
4.2
2.30
0.34
6.5
* Pos is the position of the absorption peak relative to our Co57 in Pd source (add 0.44 to
convert to the sodium nitroprusside scale), in mm/sec.
f W is full width at half height in mm/sec.
\ Ht is percentage of absorption. These parameters are the result of a least-squares fit of
Lorentzian lines to the experimental points.
36S
CARNEGIE INSTITUTION
(9|i6j8qu9p3H) Sjunoo |djoj_
o
O
o
o
o
o
O
o
o
o
o
O
o
o
o
K»
to
(D
CM
If)
(T>
00
N
Is-
CO
SJ) SJlinOO |D|0J_
T
O
o
M-
£
"io* "O
o
_o _o
s>
to
u 0)
LL.
.= J3
k.
u _g
o
T3 CD
tO
o o
J*
to
o
— ' to
-*—
<u
O c
Q.
■^ CD
O C
^—
"D O
O
Q,
"o £
t= o
o
£ <J
£
£
<L>
£ "u
>
)r. "-
to
o •-
c
a a>
o
X *£
c
■o"
the
four
o
o
«+-
BQ
o cd
>
"5T ...
C to
CD
~ -*
J*
o
"D CD
xr
a) o.
"*
to O
■r—
1
o
"O O
o
O —.
o
O
o
V)
\ (solid-
t shown
E
LL.
* 2
E
"D
to " — '
CD *-
O
« «=
c
D
o
=> O
O"
>%
*o>
«o C
*+—
l_
J. O
CD
CD
c
0)
leas
nt th
>
0)
< |
'ad
— D)
0) o
C i_
C
O CO
>»
D _Q
O
O
CD U
La
■4—
2 £
CD
£
a
«o
to
M- CD
0 ,>
c
o
T3 5)
■C
£ "»
a
to v
|8
o
to
J2
O
£ a
o
o c
«o .9
— N
D
.... *-
o
•X C
O CD
CD •-
to
Q. O
:o
■J
each
four
to
CD
^_
LL,
o 2
_c
■*- o
"J1
<>
2 o
'"■
oS-a
._ r- *0
LL.
._ LL.
GEOPHYSICAL LABORATORY 369
and May, 1965) and enable us to use garding the relationship between
the diagram of Walker, Wertheim, shear and the chemical potential of
and Jaccarino (1961) to discuss the an elastic solid (Verhoogen, 1951).
isomer shift. (An updated version of Nonetheless, the rhombic enstatite =
their plot is given in Marzolf, Dehn, clinoenstatite inversion supplies the
and Salmon, 1967.) If we assign the first unequivocal example of a large
pair A2-B1 to Fe11 on Mi and A1-B2 to effect of shearing stress on the tem-
Fe11 on M2, we obtain an average perature of a reversible phase trans-
isomer shift of 1.31 mm/sec for iron formation.
on M2 and 1.55 mm/sec for iron on Boyd and England (Year Book 6U,
Mi for the data of Table 13. For the pp. 117-120) reviewed the expen-
ses* configuration on the diagram of mental work bearing on the polymor-
Walker, Wertheim, and Jaccarino, phism of MgSi03, and in confirmation
this indicates almost no 4s contribu- of the findings of Sclar, Carrison, and
tion to the Fe11 on the Mx site and Schwartz (1964), reported an equilib-
about 10% 4s electron configuration rium curve for the rhombic enstatite
on the M2 site. Where the A lines are <=* clinoenstatite inversion using the
assigned to the Mi sites and the B single-stage piston-and-cylinder ap-
lines to the M2 sites, there is very paratus. Clinoenstatite was shown to
little difference in s electron density be the low-temperature, high-pres-
at the nuclei on the two sites but con- sure polymorph. Their curve extrap-
siderable difference in the asymmetry olates to a temperature of 630 °C at
of the sites. 1 atmosphere pressure and shows a
Refinement of the crystal struc- very small positive slope of 2.6°/kb.
tures of the hedenbergite-ferrosilite Subsequently, Riecker and Rooney
series may remove some of the am- (1966) reported the formation of
biguity in the interpretation of these clinoenstatite from rhombic enstatite
spectra. When these data are avail- at temperatures as high as 1000 °C
able we can make crystal field calcu- under shearing stresses. Clinoensta-
lations of the quadrupole splittings on tite has likewise been reported from
the two sites in the crystal structure, highly strained regions of mechani-
These calculations may clarify the as- cally deformed samples of natural en-
signment of the lines. statite (Turner, Heard, and Griggs,
1960). Since these results sharply
Effect of Shearing on contradicted the apparent equilibrium
Enstatite Polymorphism curve determined by Boyd and Eng-
land, similar experiments were per-
J.L.Munoz formed in this laboratory in the
The concept of "stress minerals" shearing squeezer (Bell and England,
(Harker, 1950) has been largely 1967) to investigate the effect fur-
abandoned because of the synthesis ther. Starting materials were an
of presumed stress minerals (e.g., equal mixture of synthetic rhombic
kyanite, chloritoid) under purely enstatite and clinoenstatite. Reaction
hydrostatic conditions and the syn- was detected by observation of
thesis of presumed antistress min- changes in the relative intensities of
erals (e.g., andalusite) under shear- the reflections on an X-ray diffractom-
ing conditions. Moreover, some work- eter trace: (231) reflection for the
ers have argued that stress minerals orthorhombic polymorph and (221)
"in a very special sense" (i.e., capable for the monoclinic polymorph,
of existence only under high shear) At all pressures studied (up to 20
contradict thermodynamic theory re- kb), these results are in complete
370 CARNEGIE INSTITUTION
agreement with those of Riecker and evidence, the AgCl melts under the
Rooney. At 890°C (more than 200°C conditions of the experiment and thus
above the piston-and-cylinder equi- provides a virtually hydrostatic pres-
librium boundary) , rhombic enstatite sure environment. The run conditions
disappears completely in less than 5 (20 kb, 600°C) were well within the
minutes when shearing is applied, clinoenstatite field of Boyd and Eng-
At 600 °C the reaction is complete in land. All data from prior piston-and-
less than '2 hours. Clinoenstatite cylinder experiments predicted con-
forms, even if rotational shearing is siderable reaction in 24 hours. In the
not directly applied, although the rate "hydrostatic cell," however, no reac-
of reaction is greatly decreased. This tion could be detected in 100 hours.
behavior may be explained by the The experiment reveals an effect of
fact that some shearing stress is al- shearing stress on the rate of the
ways present in opposed anvil de- rhombic enstatite -» clinoenstatite re-
vices, action, but it is inconclusive regard-
These results pose some interest- ing clinoenstatite stability. Thus, the
ing problems regarding the interpre- interpretation of the equilibrium
tation of the piston-and-cylinder boundary is still unclear.
equilibrium curve. There is undoubt- The influence of shearing stress on
edly a small component of shear in the stability of enstatite is an intrigu-
any solid-media apparatus. In the ing problem, which may be recom-
single-stage apparatus it is probably mended to the attention of crystallog-
no more than a few hundred bars. Is raphers. From a geologic viewpoint,
the enstatite transition, as deter- the apparent absence of clinoenstatite
mined, considerably influenced by this in terrestrial rocks, especially in
small shearing stress, or does the sheared ultrabasic rocks of orogenic
curve approximate a hydrostatic regions, is difficult to interpret in the
equilibrium boundary? In the first light of the present experimental
case, the determined curve would be data, and needs further study and dis-
valid for the piston-and-cylinder de- cussion.
vice only, and would be a reflection of
the anvilling effect of the fired pyro- High-Pressure Stability
phyllite plug and the thermocouple Relations of Spodumene
ceramic on the sample capsule. If the
second hypothesis were true, one
would expect to find clinoenstatite in Phase relations along the join
low-temperature hydrothermal ex- Li2OAl203-Si02 have long been of in-
periments ; however, this has not been terest to both mineralogists and cera-
reported. mists, but because of the unusual
To test for the effect of shearing in persistence of metastable phases, sub-
the piston-and-cylinder apparatus, a solidus studies of the mineral assem-
1 'hydrostatic" cell was prepared as blages at low pressure have been
follows: A graphite capsule contain- difficult to interpret. Spodumene, a
ing a rhombic enstatite-clinoenstatite pyroxene with composition LiAlSi206,
mixture was tightly packed in AgCl is an important phase on this join.
inside a larger Pt capsule, which was Although the natural environments in
then welded shut at both ends. The which spodumene is found (lithium-
larger capsule and furnace assembly rich granitic pegmatites) suggest
were accommodated in a piston-and- crystallization near the surface at
cylinder apparatus with a %-inch fairly low pressures, an investigation
bore. Based on post-run textural of the stability of spodumene at very
J. L. Munoz
GEOPHYSICAL LABORATORY 371
high pressures has been undertaken are: spodumene (pyroxene form, 16
in the hope of clarifying the low- kb, 1100°C), a = 9.480 ± 0.010 A,
pressure relationships. b = 8.391 ± 0.008 A, c = 5.208 ±
Two distinct types of "stuffed 0.005 A, p = 109°59' ± 5'; jff-spod-
silica" structures (Buerger, 1954) umene (keatite form, 1 atmosphere,
are encountered along the join Li20 1300°C), a = 7.539 ± 0.001 A, e =
Al203-Si02. The first example (p- 9.151 ± 0.001 A; /?-eucryptite,„ here-
spodumene, a tetragonal form ob- after referred to as /2-euc :s (LiAl
tained by heating spodumene above Si2Oe), (high-quartz structure, 10 kb,
800°C at 1 atmosphere) was shown 1100°C), a = 5.212 ± 0.001 A, e =
by Skinner and Evans (1960) to be 10.916 ± 0.003 A. Although /3-euc:iS
isomorphous with keatite, a silica (LiAlSi20G) is metastable at 1 atmos-
polymorph. The second variety has phere, it is readily synthesized at
the structure of /3-quartz, and is gen- pressures above 10 kb, both from
erally called /?-eucryptite (named for crystalline ^-spodumene and from
the high-temperature polymorph of spodumene. The stability of /?-eucS8
eucryptite, LiAlSi04). Both the /?- (LiAlSi206) relative to these two
spodumene and the ^-eucryptite phases under conditions of its for-
structures form rather extensive solid mation is thus demonstrated,
solutions along the join Li2OAl203- The effect of pressure on the
Si02, although a two-phase region (/?- relative stability of /?-eucss versus
eucryptitess + /?-spodumeness) sepa- ^-spodumene solid solutions can be
rates the two structures at 1-atmos- evaluated qualitatively by comparing
phere pressure. the molar volumes of the two phases
To synthesize appropriate starting for a constant bulk composition,
materials for this study, a homogene- Assuming ideality and neglecting
ous glass was prepared by J. F. the different effects of pressure
Schairer from Li2C03 (kindly sup- on the compressibilities of the
plied by D. B. Stewart), A1203, and two phases, increase in pressure
Si02. It was found that simultaneous should favor formation of solid solu-
fusion of all three components leads tions with smaller molar volumes. The
to measurable loss of lithia even with molar volume of /3-eucss (LiAlSi206)
gradual heating; hence a method es- (77.32 cc/mole) is less than that of
sentially analogous to the preparation ^-spodumene (78.31 cc/mole), and
of K20-Al203-Si02 glasses was em- thus pressure is predicted to stabilize
ployed (Schairer and Bowen, 1955, ^-eucryptite over ^-spodumene, as
p. 682). was found experimentally. This is
The single-stage piston-and-cylin- rare in silicates, since pressure gen-
der apparatus was used for all the erally decreases the range of a solid
experiments. Starting materials were solution if temperature is held con-
either previously crystallized single stant. Thus, in T-X section along the
phases or mixtures of two reactants. join Li2OAl203-Si02, the two-phase
Reaction routinely went to completion field /?-eucryptitess + /?-spodumeness
in 1-8 hours. must move toward Si02 and across
Three polymorphs having the com- LiAlSi206 composition with increase
position LiAlSi206 were found. Cell in pressure. Assuming ideal behavior,
parameters for these phases (ob- the pressure at which /?-eucss (LiAl
tained by least-squares refinement of Si206) becomes stable can be calcu-
powder data with the program of lated from the relationship x± =
Appleman, Handwerker, and Evans, Xe*Vip-Pl)/BT (e.g., Kullerud, 1953, p.
1963), and conditions of synthesis 105), where x = the composition of
372
CARNEGIE INSTITUTION
/>-eucryptitesg in equilibrium with
/?-spodumeness at pressure Px and
temperature T, x\ = the same equilib-
rium composition at pressure P, and
aY is the difference in molar volume
between /3-euc^ (LiAlSi20«) and
/?-spodumene. At a temperature of
1350°C the value of x was taken as
62 mole cc LiAlSi,06 in LiAlSi04
(from the interpolated solvus of Roy,
Roy, and Osborn, 1950). Solving the
equation, and taking maximum errors
in AV into account, xx = 100 when
P — Pi = 60 ± 3 kb, a result larger
by a factor of 7 than the observed
value (9.5 kb at 1350°C) . Because the
position of the 1-atmosphere solvus
would have to be shifted by more than
25 mole % toward LiAlSi206 to make
the calculated pressure agree with the
1700
600
500
1400
LIQUID
£-EUCss +
B LIQUID
o
o
o
Q>
Q.
E
,42 >+-l° *S 'H /
/5-SPOD
1300-
200
1100-
\zzz
900
SPODUMENE
/5-SP0D55+1
P -eucss
00D D
1 /3-SPOD-^ GLASS
\A /3-EUCss^/3-EUCss +GLASS
0/3-SPOD-^/3-EUCss
0 SP0D-^/3-EUCss
♦H GLASS ^-/3-SP0D
SyS-EUCss-^/3-SPOD
M J (B -EUCSS-^ /3-SP0Dss+ j3 EUCSS.
^[/3 -SP0Dss-^/3-SP0Dss+yS EUCSS
D/3-EUCSs-^ SPOD
BULK COMPOSITION^
Li20- AI203-4Si02
0
15 20 25
Pressure, kb
30
35
40
Fig. 17. Stability relations for bulk composition I^OA^OaMSiC^ as a function of pressure
and temperature. The size of the symbols represents experimental uncertainty.
GEOPHYSICAL LABORATORY
373
observed pressure, the difference is
most probably explained by non-
ideality of /?-eucryptite solid solu-
tions.
The results of the stability experi-
ments are shown in Fig. 17. Schairer
redetermined the liquidus of /?-spod-
umene at 1 atmosphere to be 1429°
± 1°C, six degrees higher than pre-
viously reported (Roy and Osborn,
1949). Melting at high pressure was
readily detected since liquid of
LiAlSi206 composition quenches to a
clear glass. Congruent melting per-
sists to about 8.5 ± 0.5 kb, 1460°C;
at higher pressures the liquidus
steepens sharply, and incongruent
melting to /?-eucryptitess + liquid
occurs. Below the solidus the two-
phase field that separates /?-eucss
(LiAlSi206) from /3-spodumene ter-
minates at a point on the solidus. The
existence of this two-phase region
was demonstrated by two experi-
ments for each P and T condition
with /?-spodumene and /?-eucHS (LiAl
Si206) as independent starting mate-
rials; /?-spodumene8S + /3-eucryptite
were observed in both the run prod-
ucts. The low-pressure terminations
of the solvi could not be detected be-
cause they are below the workable
pressure range of the piston-and-cyl-
inder apparatus. The strong curva-
ture of the phase boundary /?-eucss
(LiAlSi206) in equilibrium with p-
eucryptitess + /3-spodumeness at low
temperatures is unexplained but may
be related to inadequately known
pressure calibrations in this region.
The present data suggest that the
two-phase field may persist to rather
low temperatures and to pressures be-
low 5 kb.
The reaction spodumene <=± /?-
eucryptitess (LiAlSi206) is extremely
o
o
of
Zi
1400
1300
200
a>
Ql
£
a>
1— I 100
1000
900
£-EUCRYPTITEss
SPODUMENE
IN OUT
SP0D-^/3-EUCss o □
£-EUCss ^SPOD o □
DQ
/
0 20
Pressure, kb
30
Fig. 18. Reversed points for the reaction spodumene ^ /?-euCsS (L1AIS12O6) for both piston-in
and piston-out procedures. The difference in slope between the two curves is 6°/kb.
O I
CARNEGIE INSTITUTION
pressure sensitive and was studied by
both "piston-out" and "piston-in"
procedures. The results (Fig:. 18)
show a slope of 37°/kb for the piston-
out curve and a slope of 43°/kb for
the piston-in curve. These boundaries,
extrapolated as straight lines back
to 1 atmosphere, give temperatures
of 520 °C (for the piston-out proce-
dure) and 395°C (for the piston-in
procedure) for the maximum stabil-
ity of spodumene at 1 atmosphere.
These must be regarded as maximum
temperatures, since thermodynamics
demands that the intersection of this
reaction with the two-phase field must
cause some refraction of the spodu-
mene curve toward lower tempera-
tures. The differences observed
between piston-in and piston-out pro-
cedures probably represent the com-
bined effects of maximum friction
differences and reaction hysteresis
that are possible for the given tem-
perature, pressure, and furnace as-
sembly; hence the "true" slope for
the reaction is probably between the
two extreme values.
Neither ^-spodumene nor /?-eucryp-
tite has been reported in nature, thus
indicating rather low-temperature
limits for spodumene-bearing zones
in lithium pegmatites. The present
data described in this report suggest
that pressures as low as 5-10 kb and
temperatures not outside the range
of regional metamorphism will pro-
mote decomposition of spodumene,
forming a /3-eucryptite rather than a
ft-spodumene phase. The decomposi-
tion product would be difficult to dis-
tinguish from quartz without a chem-
ical analysis and hence might pass un-
noticed in the field.
X-ray Properties of Jadeite-Acmite
Pyroxenes
M. C. Gilbert
Pyroxenes in the system Na20-
AUCK-Fe^Oo-SiOa are important
phases in many rock types but par-
ticularly in blueschist facies rocks
and their associated eclogites. Study
of these pyroxenes was begun last
year with a report on the stability of
the end member acmite, NaFeSi206
(Year Book 65). X-ray powder dif-
fraction data are necessary in order
to measure the extent and nature of
the crystalline solubility of acmite in
jadeite.
A series of compositions ranging
along the NaAlSi206-NaFeSi206 join,
from JdiooAco to Jd47Ac53 (mole %),
have been synthesized into single-
phase pyroxenes at 40 kb and 1400 °C
by means of a piston-cylinder appa-
ratus. Reflections with a 20 (CuK«)
from 25° to 50° were indexed. The
computer program of Appleman,
Handwerker, and Evans (1963) was
used to calculate theoretical reflec-
tions, and comparisons were made
with the work on jadeite structure
(Prewitt and Burnham, 1966) and
with the acmite indexing of Nolan
and Edgar (1963). With NaF as an
internal standard, unit-cell param-
eters were determined and refined by
the least-squares program of Burn-
ham {Year Book 61) .
A systematic decrease in 2^ was
found in compositions ranging from
jadeite toward acmite, but no large
changes in separation between reflec-
tions were observed. Therefore, the
best measures of compositional
change are the reflections exhibiting
linear and large absolute variation.
In CuKa these are the 221 (Jd =
37.215°, Ac = 36.328°) and the 421
reflections (Jd = 44.277°, Ac =
43.148°). These data must be applied
with caution to natural jadeite-
acmites, for solubility of the most
common accessory cations, Ca + Mg
(diopside) , would lower the 29 values,
giving erroneously high acmite con-
tents.
Variations in the cell parameters
GEOPHYSICAL LABORATORY
375
9.700. r
9.600
9.500
9.400
1 r
j 1
NaAISi206
20
107.75
o
107.25
NaFeSi206
80 60 40
Mole per cent
Fig. 19. Variation in cell parameters of synthetic jadeite-acmite pyroxenes (space group C2/c)
and molar volumes are depicted in
Figs. 19 and 20. The straight lines
shown are drawn simply between the
values of the end-member parameters
and are not an attempt to fit all the
data. It is at once clear from inspec-
tion of Fig. 20, however, that there is
no significant deviation of the vol-
umes from a straight-line relation-
ship.
For the pyroxenes studied, com-
plete crystalline solubility holds. The
only chemical variation in the jadeite-
acmites is the substitution of Fe3+
(ionic radius, 0.64 A) for Al (ionic
radius, 0.51 A) in Mi, a 6-fold coor-
dinated site. The smooth relationships
encountered suggest random distri-
bution of Al and Fe3+ in this site.
376
CARNEGIE INSTITUTION
>
430
420
410
400
i i i
S
65-
%
NaAISi206
NaFeSi206
80 60 40
Mole per cent
Fig. 20. Variation in volume and molar volume of synthetic jadeite-acmite pyroxenes.
OTHER SILICATES
Yoderite : Synthesis, Stability,
and Interpretation of its
Natural Occurrence
W. Schreyer and H. S. Yoder, Jr.
Yoderite, a spectacular but rare
rock-forming silicate, was first dis-
covered by McKie (1959) in a talc-
yoderite - kyanite - quartz schist at
Mautia Hill, Tanzania. As a first ap-
proximation yoderite may be regarded
as a kyanite in which part of the
Al '- ions are replaced by Mg2+ with a
concomitant substitution of O2- by
OH1-. On the basis of three chemical
analyses of natural yoderites (McKie,
1959; McKie and Bradshaw, 1966), it
seems that their compositions may lie
close to an idealized end member
Mg2Al6Si4018(OH)2 (2:3:4:1). If one
accepts and includes the relatively
high Ca and Fe contents of the 1959
analysis with R2+, another theoretical
composition of yoderite might be
about Mg2.7Al5.4Si4Oi7.6(OH)2.4, which
could be simplified into Mg2Al4Si30i3
(OH)2 (2:2:3:1). Both the 2:3:4:1
and 2:2:3:1 compositions lie on the
supposed basic yoderite substitution
line Mg2+H1+ -> Al3+ (Fig. 21). Exten-
sion of this substitution line gives rise
to the interesting prospect of a hy-
drate of forsterite composition, e.g.,
Mg2Si04-H20 or Mg2Si03(OH)2.
On textural grounds McKie (1959)
concluded that the Mautia Hill yoder-
ite formed at the expense of a pre-
existing kyanite, relics of which are
still preserved as inclusions within
GEOPHYSICAL LABORATORY
377
Si02
Quartz
Pyrophyllite
Montmorillonite
Periclase,
Brucite
MgO
Talc
Anthophyllite
Enstatite
Kyanite
Sillimanite
M+A,Andalusite
Serpentine
Forsterite
lole per cent
Corundum
Diaspore
Boehmite
Gibbsite
Al203
Fig. 21. Projection of phases in the system MgO-Al2C>3-Si02-H20 onto the anhydrous plane.
Crosses indicate anhydrous portion of compositions studied. The line from A^SiOs to 2:2:3 rep-
resents the changes in composition resulting from the substitution MgH — > Al. Extent of solid
solutions is shown schematically.
yoderite. The proposed metamorphic
reaction can be given, for example,
as:
9ALSiO;
L2kJJ.W5
kyanite
+ 2Mg3Si4O10(OH)2 +
talc
H20 -> 3Mg2Al6Si4018(OH)2 + 5Si02
yoderite quartz
Kyanite and talc are not found in di-
rect contact at Mautia Hill. The bulk
composition of the schist projects into
the system MgO-Al203-Si02-(H20) in
the region denned by its final mineral
assemblage talc + yoderite + quartz.
Synthesis
Since natural yoderites contain
only about 8% of the oxides of (Fe +
Mn + Ca + Ti), it seemed likely that
the magnesium end member could be
synthesized in the system MgO-AL03-
Si02-H20. High yields of synthetic
Mg-yoderite were readily obtained in
runs only a few hours in duration in
solid-media apparatus. The pressures
of direct, unseeded synthesis ranged
from about 11 to 25 kb; the greatest
yield was obtained in the vicinity of
15 kb. The temperatures of synthesis,
varying with pressure, ranged from
600° to 900°C. Water was held in the
system through the use of sealed
silver capsules. Finely ground mix-
tures of crystalline synthetic Mg-
378 CARNEGIE INSTITUTION
cordierite and spinel with or without haps even a ternary range of solid
natural andalusite from Minas Gerais, solutions.
Brazil, were found to be the most The optical and X-ray properties
reactive starting- materials for voder- of the yoderites synthesized (Plate 2)
ite synthesis. On the other hand, re- are similar to those of the natural
action rates to form yoderite were minerals as given by McKie (1959).
negligible or slight in mixtures of In contrast to the natural purple and
natural kyanite with talc, chlorite, or green varieties, however, the syn-
forsterite. thetic Mg-yoderites are colorless.
The bulk compositions of the start- Their powder X-ray diffraction pat-
ing mixtures, disregarding H20, were terns exhibit the same non-Bragg re-
chosen according to potential compo- flections as indicated by McKie
sitions in the range of natural yoder- (1959). Slight displacements of sev-
ites (see above), as 2MgO2Al203' eral peaks toward higher 20 angles
3Si02 (2: 2 :3) , 2MgO3Al203-4Si02 are due to the somewhat smaller unit
(2:3:4), and an intermediate compo- cell expected for the Mg-yoderite end
sition, 4MgO5Al203,7Si02 (4:5:7), members,
illustrated in Fig. 21. Reaction prod-
ucts obtained invariably showed addi- Stability
tional crystalline phases amounting to The P-T field for the growth of
at least 3rr-5rr. The 2:2:3 mixture yoderite does not coincide with its
yielded small amounts of talc + chlo- range of stability. At least half of the
rite, indicating that the composition growth field is metastable owing to
of the yoderite synthesized is more the failure of kyanite to crystallize;
aluminous than 2:2:3. On the other yoderite is a metastable substitute of
hand, the 2:3:4 and 4:5:7 mixture the structurally related phase kyanite.
yielded talc, corundum, and some Synthetic yoderite was consumed un-
amorphous (probably Al-rich) ma- der the P-T conditions of its growth
terial resulting from the andalusite when seeded with kyanite-bearing
breakdown. The highest yield of syn- assemblages of corresponding bulk
thetic yoderite with only about 2% compositions. Yoderite frequently did
talc was obtained from a cordierite- not form from kyanite-bearing mix-
spinel mixture of 3:3:5 composition, tures.
Unless there is considerable loss of The presumably stable region of
Si02 to the vapor phase, this result yoderite was obtained with the use of
would indicate that the yoderites syn- a mix of 4: 5: 7 composition consisting
thesized are more siliceous than is of natural kyanite, natural chlorite
indicated by the MgH +± Al substitu- (clinochlore) , and synthetic talc
tion line. Additional substitutions in seeded with 10% of a 90% -yield
the yoderite structure leading to such yoderite run of identical bulk compo-
bulk compositions might be Mg2+Si4+ sition. Disappearance of yoderite
-» 2A13+ or perhaps Si4+ -> 4H+. The seeds was taken as evidence of its
main difficulty in determining the instability, whereas obvious growth
exact composition of the usual 5-mg was considered as an indication of its
samples produced in a single run in probable stability. The latter evidence
the solid-media apparatus was the un- was used with caution, however, be-
certainty in the composition of the cause of the possibility that yoderite
coexisting fluid phase, which occa- may grow metastably from kyanite
sionally was quenched as montmoril- + chlorite + talc if seeds of a still
lonite. It seems likely that yoderite more stable assemblage are not pro-
exhibits at least a binary, and per- vided. The field in which yoderite is
GEOPHYSICAL LABORATORY
379
25
20
a
o
a>
CO
<D
^_
0_
5 —
10
1
— i — ' — i — <— <r
-? t y— p
i 1 i 1 i
-
1
1
1
-
—
Various
3 • C
a
—
-
low- temperature
,
Various
—
-
breakdown |
assemblages i
Yoderite '
/
high-temperature
breakdown
assemblages
-
—
O
3 3 • /□
—
-
/ /
O 03 ® /
/ e /D
? / D
' ? ?
-
-
-
1
i . i i i i i i i
i 1 i 1 i
—
400
500
600
700
800
900
1000
100
1200
Temperature , ° C
Fig. 22. Preliminary pressure-temperature diagram for stability field of yoderite of possibly
variable composition based on starting materials seeded with synthetic yoderite. Solid circles in-
dicate growth of yoderite; open circles/ breakdown of yoderite to low-temperature assemblages;
open boxes, breakdown of yoderite to high-temperature assemblages; composite symbols indi-
cate negligible reaction of yoderite seeds. Breakdown assemblages are described in text.
stable can be ascertained only if all
the limiting univariant reaction
curves have been reversed by the use
of seeds of the related breakdown
assemblages demonstrated to be the
most stable.
Results of the preliminary efforts
undertaken are summarized in a P-T
plot in Fig. 22. They indicate that
yoderite is stable at temperatures be-
tween approximately 750° and 875 °C
over a large pressure interval. At
lower temperatures yoderite was
found to break down principally to
the assemblage kyanite + chlorite +
talc ; whereas at higher temperatures
the breakdown assemblages were
found to be kyanite + pure Mg-stau-
rolite (see a following section) +
enstatite at high pressures, and other
assemblages involving the additional
phases gedrite (see following sec-
380
CARNEGIE INSTITUTION
tion), corundum, cordierite, and liq-
uid at lower pressures. The most
astonishing result, however, was that
the yoderite field does extend down to
pressures even below 10 kb.
The Assemblage Y ode rite-Quartz
The experimental investigations re-
ported have limited bearing on the
occurrence of the mineral yoderite in
nature, inasmuch as they were re-
stricted to the bulk composition close
to yoderite itself. As noted above, the
Mautia Hill yoderite schist is an as-
semblage containing excess quartz. A
useful step in approaching the prob-
lem of the pedogenesis of this rock
seemed to be an experimental study
of the reaction kyanite + talc + H20
-> yoderite + quartz. This reaction
was considered by McKie (1959) to
be the fundamental process of yoder-
ite formation at Mautia Hill.
For this purpose a mixture was
made of 90 % kyanite + talc and 10%
seeds of synthetic yoderite + quartz,
having the composition 3MgO3Al203#
7Si02 and an excess of H20, i.e., near
the intersection of the kyanite + talc
and yoderite + quartz joins (Fig.
21). Over the largest part of the
yoderite stability field, i.e., between
about 11 and 25 kb, the seed assem-
blage yoderite + quartz broke down
readily to form more kyanite and talc.
Only at 10 kb, in the vicinity of
700 °C, was a small amount of growth
of yoderite + quartz seen.
These results indicate that the as-
semblage kyanite + talc, which was
present in the Mautia Hill schist be-
fore the formation of yoderite, is a
high-pressure assemblage formed
above approximately 10 kb. The sub-
sequent growth of yoderite and
quartz at their expense has probably
taken place at a slightly lower pres-
sure but still in the vicinity of 10 kb.
Because the yoderite growth involved
a hydration, it is likely that it was
also accompanied by a lowering of the
prevailing temperature after the peak
of metamorphism- with a hydrous
vapor phase still present in the sys-
tem. The small but not negligible
amount of Fe contained in the Mautia
Hill rock undoubtedly will favor the
yoderite + quartz assemblage versus
an assemblage of kyanite with talc,
and also assemblages involving cordi-
erite. Iron is not accommodated read-
ily in either talc or cordierite at these
high pressures, according to studies
on the analogous iron system.
The rarity of the mineral yoderite
in the natural environments encoun-
tered so far through geological ex-
ploration may thus be explained
mainly on the basis of two properties :
1. The extensive stability field of
the assemblage kyanite + talc re-
stricts the phase yoderite to silica-
poor (less than about 50%-56%
Si02) bulk compositions of the system
MgO-Al203-Si02-H20. Such bulk com-
positions are not common, especially
in the lower parts of the earth's crust
or the upper part of the mantle.
2. The stability field of the assem-
blage yoderite-quartz is very small.
Hence yoderite formation in more si-
liceous rocks such as the Mautia Hill
schist would be a rare event.
A Reconnaissance Study of the
System MgO-Al203-Si02-H20 at
Pressures between 10 and 25 kb
W. Schreyer
Experimental work on the syn-
thesis and stability of yoderite as
reported in the previous section indi-
cated that the phase relations in the
system Mg0-Al203-Si02-H20 at high
pressures are considerably different
from those known in the pressure
range below 10 kb ( Yoder, 1952 ; Roy
and Roy, 1955; Schreyer and Yoder,
1964; Fawcett and Yoder, 1966, and
Year Book 62, pp. 143-145). For this
reason the investigation was extended
to other bulk compositions, also
GEOPHYSICAL LABORATORY 381
shown in Fig. 21, that were expected drous composition of Mg-staurolite is
to yield critical evidence on the com- less siliceous than 4:9:8. Richardson
patibility relations of phases in the (Year Book 65, pp. 248-252) has
system. pointed out that the oxide ratio in
Phase relations in the system are Fe-staurolite is 4:9:7.5, probably be-
strongly influenced by the appearance cause of a 4H+ for Si4+ substitution in
of four crystalline ternary phases the lattice. Recent determination of
that are not stable at low pressures or the water content of synthetic Fe-
else have not been synthesized pre- staurolite supports this view (Rich-
viously at still lower pressures. In ardson, this report) .
addition to yoderite, discussed in the The growth field of Mg-staurolite is
preceding section, these phases in- considerably larger than its stability
elude pyrope (previously synthesized field (cf. yoderite synthesis field, pre-
at higher temperatures by Boyd and ceding section) , and synthesis of Mg-
England, Year Book 58, pp. 83-87), staurolite was possible at tempera-
Mg-gedrite, and Mg-staurolite. The tures as low as 650 °C, although it
last two phases, as well as yoderite, does not become a stable phase until
were synthesized for the first time about 100 °C higher,
during the present investigations. The tentative stability field of pure
Mg-staurolite is outlined in Fig. 23.
Mg-Staurolite: Synthesis and It is based on the results of runs with
Stability a 4:9:8 mixture of kyanite + spinel
A phase with optical and X-ray + MgO seeded with 10% of previ-
properties akin to natural and syn- ously synthesized almost pure Mg-
thetic Fe-bearing staurolites was syn- staurolite. Disappearance or obvious
thesized from a variety of composi- decrease in the number of seeds was
tions in the system MgO-Al203-Si02- taken as evidence for their instability,
H20. The phase assemblages in which whereas obvious increase was consid-
it occurred suggested that its bulk ered preliminary evidence for the
composition lies in the vicinity of the stability of Mg-staurolite. It is ap-
anhydrous composition 4MgO9Al203' parent that Mg-staurolite is stable
8Si02 (4:9:8, Fig. 21), which is the only at pressures above about 12 kb.
pure Mg analogue of the formula at- Its stability range overlaps and ex-
tributed by Naray-Szabo and Sasvari ceeds at its high-temperature end
(1958) to natural staurolites. The that of yoderite (cf. Fig. 22). Because
highest yields of Mg-staurolite were of this overlap and, especially, be-
obtained in runs at 25 kb, 850°- cause the lower temperature stability
900°C, on a 4:9:8 mixture consisting limit virtually coincides with that of
of natural kyanite with MgO and syn- yoderite and that of pyrope (see next
thetic Mg-spinel. Due to the similar- paragraph and Fig. 24), the reactions
ity of the crystal structures of stau- limiting the field of Mg-staurolite
rolite and kyanite, this mixture may stability are not yet clear,
provide seeds of structural units req- In the experiments Mg-staurolite
uisite for the growth of staurolite. broke down at low temperatures to
It is probably significant, however, kyanite -f chlorite + corundum. If
that the 4:9:8 composition never the yoderite field extends to only
yielded 100% Mg-staurolite but al- slightly lower temperatures than the
ways additional small amounts of Mg-staurolite field, however, the sta-
enstatite and relic kyanite, even in ble breakdown assemblage over this
the longest runs. This phase assem- narrow temperature interval could,
blage may indicate that the anhy- for example, be yoderite + kyanite
382
25
CARNEGIE INSTITUTION
:;
CO
o
_Q
_o
I/]
EL
Ql
10
-e — o
f-#-D-D-
¥
Various
low- temperature
breakdown
assemblages
O O |
D D
Mg-
Stau rolite
.?'
D
O
L,*
/
/a
/
A
/
Various
high-temperature
breakdown
assemblages
/
D
D a
400
500
600
700
800
900
1000
100
1200
Temperature ,° C
Fig. 23. Preliminary pressure-temperature diagram for the stability field of Mg-staurolite based
on runs seeded with synthetic Mg-staurolite. Solid dots indicate growth of Mg-staurolite; open cir-
cles, breakdown to low-temperature assemblages; open boxes, breakdown to high-temperature
assemblages described in text. Composite symbols indicate negligible reaction of Mg-staurolite
seeds.
+ corundum. Conversely, if the lower
stability limit of yoderite lies at
somewhat higher temperatures than
that of Mg-staurolite, the latter phase
might, over a certain P-T interval,
take part in the breakdown reaction
of yoderite. Possible intersections of
the lower stability limits of Mg-
staurolite, yoderite, and pyrope may
create several stable invariant points
involving these three phases plus
kyanite, chlorite, talc, and corundum.
At the high-temperature end of the
Mg-staurolite field, breakdown to the
assemblage pyrope + kyanite + co-
rundum was observed at 25 kb; at
lower pressures aluminous enstatite
substituted for pyrope — either meta-
stably or, outside the pyrope stability
range, stably. At the low-pressure
end of the Mg-staurolite field the as-
semblages kyanite + gedrite + co-
GEOPHYSICAL LABORATORY
383
rundum and kyanite + talc + corun-
dum were encountered. Sapphirine
was also observed among the break-
down products of Mg-staurolite, how-
ever, and the assemblage gedrite +
kyanite + sapphirine may well have
a stability field.
Staurolite approaching the compo-
sition of the pure Mg end member has
never been found in nature (Deer,
Howie, and Zussman, 1962). The
reason for this absence is probably
the lack of highly aluminous bulk
compositions in the upper parts of
the mantle where the P-T conditions
necessary for the formation of Mg-
25
20
to
o
O
13
CO
CO
<D
i_
CL-
IO
Various
low- temperature
breakdown
assemblages
Various low-pressure
breakdown assemblages
400
500
600
700
800
900
1000
100
1200
Temperature ,°C
Fig. 24. Preliminary pressure-temperature diagram for the stability of pyrope in the presence of
water based on runs seeded with synthetic pyrope. Solid circles indicate growth of pyrope; open
circles, breakdown of pyrope to low-temperature assemblages; open boxes, breakdown to low-
pressure assemblages; composite symbols indicate negligible reaction of pyrope seeds. The smaller
symbols represent experimental results reported by Boyd and England [Year Book 58, p. 84,
Fig. 1); solid line is their breakdown curve. The nature of the various breakdown assemblages is
discussed in the text.
384 CARNEGIE INSTITUTION
staurolite may prevail. Moreover, in which also breaks down to chlorite +
that part of the stability range in- talc + kyanite. The intersection of
vestigated, Mg-staurolite and quartz the pyrope and yoderite breakdown
are incompatible by virtue of the curves produces an invariant point
stable tie lines kyanite-talc, kyanite- involving the six phases yoderite, py-
gedrite, kyanite-enstatite, kyanite- rope, chlorite, talc, kyanite, and fluid,
yoderite, kyanite-chlorite, and ky- The breakdown of pyrope to the as-
anite-pyrope, described later (see semblage chlorite 4- talc + yoderite
Fig. 27A, B). would thus be limited to pressures
below this invariant point, whereas
Stability of Pyrope in the Presence *he breakdown to chlorite + talc +
of Water kyanite is limited to pressures above
this invariant point. Another inter-
The stability relations of the pure esting implication of this invariant
MgAl-garnet pyrope in the anhydrous point is that the low-temperature sta-
system MgO-AL03-Si02 under high bility limit of yoderite can be given
pressures and temperatures were in- only by the breakdown reaction to
vestigated by Boyd and England chlorite + talc + kyanite below this
(Year Book 58, pp. 83-87). By anal- pressure; above it, yoderite must
ogy with the behavior of other an- break down over a narrow tempera-
hydrous phases in the system it was ture interval to the assemblage chlo-
expected that, in the presence of ex- rite + kyanite + pyrope. It is not
cess H20, pyrope would break down yet clear whether or not the low-
into hydrous mineral phases. The temperature stability limit of pyrope
preliminary P-T diagram given in is further modified by reactions in-
Fig. 24 is based on the results ob- volving Mg-staurolite.
tained with a starting material con- Boyd and England reported the
sisting of 90% forsterite + cordierite breakdown of pyrope at high temper-
+ spinel seeded with 10% synthetic atures but low pressures to aluminous
pyrope, kindly provided by Dr. F. R. enstatite + sapphirine + sillimanite.
Boyd. Their experimental results are desig-
As described by Boyd and England, nated by the smaller symbols in Fig.
at approximately 23 kb and above, 24. Their low-pressure breakdown
pyrope forms very rapidly without limit was followed in the present in-
seeds at temperatures in excess of vestigation to still lower pressure-
about 950 CC. At lower pressures, and temperature conditions where it in-
at 25 kb even at temperatures below tersects the previously mentioned
950°C, pyrope grows only when seeds low-temperature breakdown curve,
are available. At temperatures be- thus delimiting the whole pyrope sta-
tween about 700° and 750°C in the £?£ field in the presence of water,
pressure range of about 14-25 kb the Wlthm the errors of the Mnfltt-
pyrope seeds plus the rest of the urements the slope of the breakdown
1T--X. i t j. i it curve determined by Boyd and Eng-
mixture were found to break down land .g maintained in the lower ex„
to the assemblage chlorite + talc tensi(m> although the breakdown
+ erite or chlorite + talc + assemblages encountered in the pres-
kyanite. The overall slope of this ence of water are different from those
lower temperature stability limit is reported by Boyd and England for
steep but clearly slightly negative, the dry system.
For this reason it must intersect the Thus, at the highest temperature
lower stability limit of yoderite, investigated (975°C) a liquid or fluid
GEOPHYSICAL LABORATORY 385
quenched mainly as glass was found Upper Pressure Stability Limit of the
in addition to the solid breakdown Assemblage Chlorite + Quartz
products enstatite and sapphirine. At
25 kb and temperatures of even The assemblage chlorite-quartz is
1000° and 1100° only relatively small very common in natural rocks of
amounts of glass were found to coat sedimentary, metamorphic, and mag-
euhedral garnet crystals making up matic origin. The assemblage is used,
at least 90% of the charge with no among others, to define the green-
other solid phases present. This result schist facies of regional metamor-
can be interpreted as a supercritical phism (Turner and Verhoogen, 1960,
phenomenon above a second critical p. 534). Fawcett and Yoder (1966)
end point. It would indicate that py- have determined that its upper tem-
rope in the presence of water no perature stability limit in the system
longer shows melting to a liquid at MgO-Al203-Si02-H20 is defined, at
much lower temperatures than in the low pressures, by the reaction chlo-
dry system, but that it displays only rite + quartz <p* cordierite 4- talc +
a more or less congruent solubility in vapor,
a supercritical hydrous fluid. To obtain an insight into the reac-
At about 900 °C the liquids encoun- tions limiting the chlorite + quartz
tered outside the pyrope field also dis- compatibility at higher pressures, ex-
appear, and the main breakdown as- periments were conducted with mix-
semblage was found to be enstatite tures of crystalline clinochlore (natu-
and corundum (which may be a meta- ral leuchtenbergite) + quartz, and
stable substitute for a sapphirine- natural kyanite and talc of a total
Al2Si05 assemblage). At still lower bulk composition approximating
temperatures hydrous phases make 15MgO*3Al203,23Si02, i.e., lying at
their appearance, one possibly stable the intersection of the two joins (Fig.
breakdown assemblage being gedrite 21). Either of the two mixtures was
+ enstatite + sapphirine. However, also used as seeding material for the
especially near the low-pressure ter- other one. In addition a 3:3:7 mix-
mination of the pyrope field, the rela- ture of the chlorite with quartz and
tions become extremely complicated, kyanite, with or without seeds of
As many as six crystalline phases — yoderite and talc, was used for com-
talc, yoderite, chlorite, gedrite, en- parative purposes. The results of
statite, and corundum — were found these experiments indicate that there
in the run products; this probably are two reactions limiting coexistence
indicates the proximity of at least of chlorite-quartz at higher pres-
two invariant points. sures: (1) chlorite + quartz <=*
The present results on the stability yoderite + talc, and (2) chlorite +
of pyrope support the view that this quartz ^ kyanite + talc. The pre-
mineral cannot be expected to occur liminary P-T plot of Fig. 25 shows
within the crust. About 15 kb is the that at least one of these reaction
minimum pressure for its formation curves has a pronounced negative
in the presence of water. Pyrope slope. Thus this assemblage (so com-
could not be stable in a mantle con- mon at low temperatures and pres-
taining sufficient water for its hydra- sures) is eliminated at higher pres-
tion below temperatures of approxi- sures because of the higher densities
mately 750 °C. Instead of pyrope, of the assemblages kyanite 4- talc,
assemblages with chlorite, talc, and and yoderite + talc. Because chlo-
kyanite d= yoderite might be expected rite and (aluminous) talc are mem-
in such environments. bers of solid solution series, the two
1S6
CARNEGIE INSTITUTION
25
20
15 -
o
.a
o
lOi —
-
i '
i
i
1 1 k
O
-
\
\
Kyani te
-
+ T
al c
-
\
?
-
\
\ °
O
t/>
\
0)
o
\
-
f ?
\
o
-
II
\
? o
•\
-
V
c
\°
—
• •
leper -1
<u 1
\ :
_
'
\ +
-
1
Chlorite
K
z
h
+
/
/
5 ?
Quartz
/
—
s
Talc
^V+Cordierite
<y + Chlorite
300
<?-<?
O
o
o o
Yoderite -I- Talc
+ Quartz
400
500
600
700
800
900
1000
1 100
Temperature, c C
Fig. 25. Preliminary pressure-temperature diagram for the stability of the assemblage chlorite-
quartz based on runs with starting materials containing chlorite + quartz as well as kyanite +
talc. Solid circles indicate growth of chlorite + quartz; open circles, growth of kyanite + talc;
composite symbols represent runs with negligible reaction between these two assemblages; open
box indicates probable growth of the assemblage yoderite + talc + quartz. The solid curve des-
ignated F — Y is taken from Fawcett and Yoder (1966); dashed curves as well as locations of
invariant points /3 and /$ are speculative.
reactions outlined are not univariant
but may take place over an extended
pressure range with the compositions
of chlorite and talc gradually chang-
ing. Only a unique composition of
chlorite coexisting with quartz at the
higher possible pressure can react
with quartz along a univariant curve
to form a unique aluminous talc +
kyanite.
Both the assemblages kyanite +
talc (+ quartz) and yoderite + talc
(+ quartz) were formed at least
once during the metamorphic history
of the unique schist at Mautia Hill,
Tanzania (see previous section on
yoderite). This provides additional
evidence for a high-pressure meta-
morphism of this rock, which, in an
earlier lower pressure metamorphic
event, may have been a simple chlo-
rite-quartz schist. The Mautia Hill
schist contains yoderite surrounded
by minute films of a chlorite mineral
GEOPHYSICAL LABORATORY 387
in direct contact with quartz. This cordicrite stability relations might
chlorite-bearing assemblage was most have to be modified. For this reason a
probably formed during the final up- number of theoretically possible
lifting of the rock. The experiments breakdown assemblages involving
conducted with a talc + kyanite mix- talc, chlorite, quartz, yoderite, and
ture seeded with 10% chlorite + kyanite were seeded with cordierite
quartz show that the growth of the and run under various pressures and
latter assemblage at the expense of temperatures.
the former is extremely sluggish. The results obtained thus far in-
Several reaction products obtained dicate that the cordierite previously
within the P-T range considered here, grown at 10 kb, 700 °C, is clearly met-
i.e., those at 15 kb, 500°-550°C, from astable with respect to the assem-
cordierite + spinel mixtures, showed blage yoderite + talc + quartz. On
additional peaks in the X-ray diffrac- the other hand, cordierite is the stable
tion patterns that might be attributed phase at 800 °C, 10 kb. At 800 °C and
to a phase with a chloritoid structure. 11 kb cordierite breaks down to talc
If stable, Mg-chloritoid-bearing as- + kyanite 4- quartz. At 900 °C cordi-
semblages may modify the stability erite was found to break down at a
limits of the chlorite-quartz assem- pressure between 11 and 12 kb to
blage as given here. either kyanite + gedrite + liquid or
kyanite + enstatite + liquid, this al-
Upper Pressure Stability Limit of ternative depending on the problem
Mg -Cordierite 0f gedrite stability (see later para-
The stability relations of cordierite, graph) . Although no runs were made
Mg2Al4Si50i8'£H20, have been stud- at still higher temperatures, it is ap-
ied by Schreyer and Yoder (1964) in parent that the upper pressure sta-
the pressure range up to 10 kb. At bility limit of cordierite lies — as al-
this pressure, cordierite was found ready suspected by Schreyer and
to be stable between approximately Yoder (1964) — only slightly above 10
700° and 950 °C, where incongruent kb and has, in general, a very flat
melting takes place. Subsolidus break- positive slope.
down of cordierite into the metastable These stability relations, as well as
assemblage chlorite + quartz + the modified breakdown reactions of
corundum occurred below 700 °C. cordierite at pressures and tempera-
Schreyer and Yoder speculated that tures below 10 kb, 700 °C, are sum-
the cordierite breakdown at pressures marized in a P-T plot (Fig. 26) that
in excess of 10 kb involved reactions is based partly on theoretical consid-
to the assemblages Al-talc + silli- erations, and needs further experi-
manite + quartz, gedrite -f silliman- mental confirmation through longer
ite + quartz, and Al-enstatite + silli- runs with regard to positions and
manite + quartz. slopes of curves. The most important
In the present investigation the re- implication of this plot is that in ad-
action relationships involved in the dition to the invariant point I5 sepa-
cordierite breakdown have been stud- rating the talc-kyanite and gedrite
ied more closely by means of both the (enstatite) - kyanite breakdown as-
gas-media apparatus and the single- semblages there must be three others
stage, solid-media apparatus. It was (74, 73, I2), which, in the order of
expected from the extension of the decreasing pressures and tempera-
previously unknown yoderite stability tures, separate the cordierite break-
field to pressures below 10 kb that at down assemblages yoderite + talc +
least part of the earlier determined quartz, yoderite + chlorite + quartz,
388
CARNEGIE INSTITUTION
o
_Q
O
<D
Z3
CO
CO
CD
:
Talc -f Kyanite
4- Quartz
Gedrite (or Enstatite
+ Kyanite + Quartz
?
:_
I. o a o +■
+£_ ^ h
/ M
I 1 ^-7" \_ Yoderite 4 Talc
I I /
-/H3 •
+T \ Yodente+Chlorite
+ Quartz
Various Crystals
+ Liquid
Cord i eri t e
^HH
600
700
800
900
1000
100
1200
Temperature ,° C
Fig. 26. Preliminary pressure-temperature diagram for the breakdown of cordierite at high pres-
sures in the presence of water, based on runs seeded with cordierite and various possible break-
down assemblages. Solid circles indicate growth of cordierite; crosses, open boxes, open circles,
and open triangles represent breakdown to the various assemblages as indicated in the figure.
The solid curves designated S + Y (1964) are taken from Schreyer and Yoder (1964); dashed
curves as well as locations of invariant points 1 2— U ar^ either not known accurately or are specu-
lative.
and finally kyanite + chlorite +
quartz. This last assemblage, prob-
ably to be modified with regard to
the stable Al2Si05 polymorph, would
then extend down to the invariant
point lt involving the additional
phases cordierite and pyrophyllite
which was shown by Schreyer and
Yoder (1964) to lie at approximately
5 kb, 525 °C. The invariant point I2
involving the phases chlorite, kya-
nite, quartz, yoderite, and cordierite is
also the origin of a univariant curve,
chlorite — kyanite + quartz z± yoder-
ite, which must represent the lower
stability limit of the phase yoderite
at pressures within the stability field
of the assemblage chlorite-quartz
(see above and Fig. 25).
Synthesis of a Pure Mg-Gedrite
In the present study a phase yield-
ing an X-ray powder pattern closely
related to that of natural gedrites as
well as to natural and synthetic an-
thophyllites has been synthesized
from various bulk compositions over
a relatively narrow range of tem-
peratures and pressures. The distinc-
tion between the phase obtained here
and Mg-anthophyllite, Mg7[Si8022]
(OH) 2 (Greenwood, 1963), is based
on appreciable shifts of X-ray peaks
to higher 20 angles, indicating a
GEOPHYSICAL LABORATORY 389
smaller cell size of the aluminous or- Book 59, pp. 90-91) and of pyrope
thoamphibole, which is in agreement (Boyd and England, Year Book 58, pp.
with the findings of Francis and Hey 83-87), and as an incongruent melt-
(1956). ing product of cordierite with excess
The highest yields of synthetic ged- water (Schreyer and Yoder, 1964).
rite (85%-90%) were obtained at 11 On the other hand, a subsolidus hy-
kb, 850 °C, with a mixture of ideal ged- drothermal synthesis of this phase or,
rite composition (Mg5Al2[Al2Si6022] conversely, a breakdown of sapphi-
(OH)2, i.e., 5:2:6 + H20, cf. Fig. rine in the presence of water into
21) , prepared from talc, kyanite, and hydrous assemblages has not been re-
MgO. Since only small amounts of ported. Yet the observation of hy-
sapphirine and relic kyanite were drous alteration products surround-
found to coexist, it is estimated that ing sapphirines in natural rocks
the gedrite synthesized contains suggests quite clearly that the min-
somewhat less alumina (about 10 eral must exhibit a lower temperature
mole % ) than the ideal composition, stability limit in the presence of H20.
Reactions have not been reversed, In the present study sapphirine co-
and it is therefore not clear whether existing with other phases was syn-
the gedrites are stable phases under thesized in the presence of water
the conditions of synthesis or whether from several bulk compositions at
they represent metastable intermedi- temperatures ranging from 850° to
ate products formed during the 975 °C and at pressures between 11
breakdown of talc in much the same and 17 kb. The formation of sapphi-
way as the first synthetic Mg-antho- rine seems to be relatively sluggish,
phyllites formed as intermediate the more so, the higher the pressure,
products of the reaction talc -> ensta- A starting mixture of 2: 2: 1 composi-
tite + Si02 + H20 (Bowen and tion consisting of cordierite and spi-
Tuttle, 1949). nel run at 15 kb, 900°C, first yielded
In the present study gedrites were the assemblage enstatite + corundum
obtained only at pressures between 11 + spinel, and only after 22 hours
and 20 kb and at temperatures in the under the same conditions had an
range 850°-900°C. Their failure to amount of some 40% sapphirine
form at higher pressures may be due formed at the expense of that meta-
to an upper pressure stability limit stable assemblage. At 25 kb, 900 °C,
similar to that proposed by Green- the same starting mixture did not
wood (1963) for anthophyllite, yield any sapphirine even after 23
which on the basis of thermochemical hours, but only corundum + spinel +
data is assumed to break down to pyrope. It is suspected that this be-
enstatite + talc in the vicinity of havior is an indication of the prox-
20 kb. imity of an upper pressure stability
limit for the phase sapphirine.
Stability of Sapphirine j± low-temperature breakdown of a
Sapphirine, the least siliceous an- natural sapphirine from Greenland
hydrous ternary phase of the system and of a synthetic sapphirine kindly
(Fig. 21), is stable at atmospheric provided by Dr. W. Foster was
pressure near liquidus temperatures achieved at 800 °C, 15 and 20 kb. In
(Keith and Schairer, 1952). It has also the presence of water the sapphirine
been encountered at higher pressures was found to break down to an as-
and temperatures among the break- semblage chlorite + corundum. Al-
down products of anhydrous cordi- though reactions have not been re-
erite (Schreyer and Yoder, Year versed, it appears that the lower tern-
390
CARNEGIE INSTITUTION
perature stability limit of sapphirine
lies in the vicinity of 825 °C at 15 kb.
At a pressure of 10,000 psi (-750
bars) chlorite and corundum are com-
patible only below approximately
500°C, whereas at all higher tempera-
tures the join cordierite-spinel is
valid (Roy and Roy, 1955). Also to
be clarified is whether or not Mg-
sapphirines can have the composition
2MgO2Al203"Si02 suggested by anal-
yses of some natural specimens. If so,
MgO
10 kb
600°C
Al203
14 kb
750°C
5 kb
700°C
GEOPHYSICAL LABORATORY
391
B
20 kb
775°C
20 kb
900° C
Al203
Fig. 27. (A,B). Compatibility triangles of the system MgO-A!2C>3-Si02-H20 for various repre-
sentative pressures and temperatures. The relations near the MgO corner of the diagrams have not
been investigated in this study. The possible instability of sapphirine at 20 kb even at tempera-
tures above 775°C (see text) has not been taken into account. Abbreviations as in Fig. 21.
the presence of such a sapphirine
would prohibit the stable coexistence
of cordierite and spinel.
Isothermal Isobaric Sections
In order to summarize the phase
relations in the system MgO-Al203-
Si02-H20 encountered in this study
at high pressures, a series of isother-
mal isobaric sections projected on the
anhydrous base are given in Fig.
27 (A, B). Although the sequence of
10 compatibility triangles chosen to
illustrate the formation and disap-
pearance of phases is largely self-
explanatory, some of the more salient
features of the varying compatibili-
ties will be pointed out:
1. At variance with many compati-
bility triangles presented by other
authors for lower pressures, no stable
tie line was found for the assemblage
spinel-Al-silicate (kyanite). A mix-
ture of 2:3:1 composition (Fig. 21)
made up of synthetic spinel and nat-
ural andalusite run at 20 kb, 800 °C,
exhibited complete breakdown of
these phases and the formation of
chlorite and corundum. The 4:9:8
mixture of kyanite + spinel + MgO
invariably yielded run products in-
volving corundum, unless large
392
CARNEGIE INSTITUTION
amounts of the Mg-staurolite formed.
2. Particular attention is drawn to
the compatibility of cordierite with
kyanite at 10 kb, 800°C, which had
previously been dismissed as an un-
stable pair (Schreyer and Yoder,
1964).
3. There is a drastic rearrange-
ment of tie lines in the temperature
range between 750° and 850 °C at
all pressures investigated, which is
mainly induced by the instability of
the chlorite-kyanite pair at higher
temperatures and its reaction to form
a talc-corundum assemblage.
4. The assemblages of kyanite with
talc, gedrite, and enstatite represent
an effective barrier over most of the
P-T range against the appearance of
the more Al- or Mg-rich phases to-
gether with free quartz. This barrier
may well be overcome at still higher
pressures when the denser forms of
SiOj, coesite and stishovite, become
stable.
The Aluminum Silicates
S. W. Richardson, P. M. Bell,
and M. C. Gilbert
Metamorphic petrologists consider
the system Al203'Si02 fundamental to
any petrogenetic grid. The occurrence
of the polymorphic forms kyanite,
andalusite, and sillimanite has been
regarded as a function simply of
total pressure and temperature and
not of the fugacities of the volatile
constituents in rocks. Thus it is im-
portant (1) to determine accurately
the stability fields of these poly-
morphs and (2) to ascertain what
effect a common volatile substance,
like H20, has on their stability re-
lations.
We here report the determination
of 17 reversed brackets of kyanite-
sillimanite equilibrium between 700°
and 1500 °C. Measurements of the
equilibrium pressure in solid-media
apparatus have been checked against
measurements in gas-pressure appa-
\c
1 !
1 1 1 1
i
to
o |0
-
■ inn
-
_o
2
—
■□□
—
<D
3
in 8
■
-
a.
6
□
1 1
i i i i
i
600 700 800 900
Temperature, °C
1000
Fig. 28. P-T coordinates of gas-apparatus ex-
periments in which growth of kyanite or silli-
manite could be detected. Filled squares, kya-
nite grew; open squares, sillimanite grew.
ratus; the data may be extrapolated
with some confidence to temperatures
of geological interest and used to as-
certain pressures of metamorphism.
Water does not affect the position of
the equilibrium.
Kyanite-Sillimanite Equilibrium
Figure 28 summarizes the data ob-
tained from 700°-850°C with gas-
pressure apparatus (Yoder, 1950a).
The experimental procedure and X-
ray method used to detect reaction
were reported last year (Year Book
65). In the light of further determina-
tions of the variation of the kyanite
(021) /sillimanite (120) peak height
ratio in the starting material, we are
now confident that sillimanite has
been produced in an experiment only
if the value of the K/S ratio shown
by the product falls below 0.30 or
that kyanite has been produced only
if the value of the ratio exceeds 0.60.
Runs were carried out at tempera-
tures up to 1500°C and at pressures
above 10 kb in the solid pressure-
media, piston - cylinder apparatus.
Both piston-in and piston-out brack-
ets of the equilibrium were obtained.
(For an explanation of piston-in and
piston-out conditions see Boyd, Bell,
England, and Gilbert, Year Book 65.)
GEOPHYSICAL LABORATORY
393
CO
o
O
CD
^_
r3
CO
CO
CD
^_
30
28
26
24
22
20
14
12
10
IN-STROKE
OUT-STROKE
—
-^
/\
DRY
V
TRACE
WATER
V
S7
▲
Z\
WET
V
V
^
X\
K
S
K
S
/\
-^-
VV
^
/\/\
V
z^
z\
^
800
900
1000
MOO 1200
Temperature, °C
1300
1400
1500
Fig. 29. P-T coordinates of piston-cylinder experiments in which growth of kyanite or sillimanite
could be detected.
At the same time it was possible to
compare a determination of kyanite-
sillimanite equilibrium made in the
presence of 20-50 wt % water, in
the presence of traces of water, and
dry. The results are summarized in
Fig*. 29.
Different amounts of water have
no detectable effect on the position of
the equilibrium boundary: Brackets
obtained by the same pressure-control
procedure at the same temperature
show overlap; and a straight line,
such as one passing through 12 kb,
800°C, and 29 kb, 1600°C, would sat-
isfy the data.
394
CARNEGIE INSTITUTION
24
:;
i i
en
^_
D
O
CD
^_
CO
<D
Q_
0
PISTON-CYLINDER
IN -STROKE OUT-STROKE
CRY
-y~
^
TRACE
WATER
T
▲
WET
y
•A
GAS-APPARATUS
(WET)
EXTREME SLOPES
FROM PISTON -CYLINDER
EXPERIMENTAL UNCERTAINTY IN KYANITE-
SILLIMANITE BOUNDARY APPLYING PISTON -
CYLINDER SLOPES TO GAS-APPARATUS BRACKETS
200
400
1000
1200
600 800
Temperature , ° C
Fig. 30. Brackets of kyanire-sillimanite equilibrium.
1400
1600
However, there is no straight line
that satisfies both the piston-cylinder
data and the data obtained in the gas
apparatus (see Fig. 30). We believe
that an inequality of pressure distri-
bution within the piston-cylinder
apparatus furnace cell (Boyd and
England, 1963, Fig. 1) explains this
lack of a straight line.
At 850°C and 12-14 kb nominal
pressure, the magnitude of the non-
frictional pressure discrepancy must
be between 1 and 3 kb. It is important
to note that this discrepancy is not
friction in the usual sense. Friction
between the piston and cylinder wall
in piston-cylinder apparatus can be
evaluated by making piston-in and
piston-out runs. The discrepancy is
apparently due to differences in the
shearing strength and compressibility
of the various components of the fur-
nace cell.
We can attempt to set limits of
error for kyanite-sillimanite equilib-
rium by applying the maximum and
minimum slopes that satisfy all our
piston-cylinder brackets to the limits
of the gas-pressure apparatus brack-
ets (Fig. 30) . These slopes fit the gas-
pressure apparatus data very well.
In cooperation with J. F. Hays at
Hoffman Laboratory, Harvard Uni-
versity, we have tried to determine
the magnitude of the nonfrictional
pressure discrepancy at high temper-
GEOPHYSICAL LABORATORY
395
ature and thus test the validity of the
slope assigned to the kyanite-silliman-
ite equilibrium. We have investigated
the equilibrium, dry, at 1300 °C in a
gas-pressure apparatus described by
Birch, Robertson, and Clark (1957)
and Clark, Robertson, and Birch
(1957). The pressure coefficient of
the manganin resistance pressure
gauge used at Hoffman Laboratory
was determined up to 10 kb in the
gas-pressure apparatus used here for
the lower pressure experiments.
Thus far, two successful runs have
been made: After 5 hours at 23.45 ±
0.30 kb, all sillimanite had been con-
verted to kyanite; after 5 hours at
22.23 ± 0.30 kb no reaction could be
detected. If the latter run is inter-
preted as being close to the equilib-
rium boundary (a likely interpreta-
tion in the light of our experience
with the piston-cylinder apparatus at
this temperature) it would appear
that the nonfrictional pressure dis-
crepancy in the piston-cylinder appa-
ratus decreases at 1300 °C to the ex-
tent that our runs under piston-out
conditions bracket the pressure of the
inversion. This is not surprising since
shearing strength decreases with tem-
perature, and if the pressure discrep-
ancy is related to variations in shear-
ing strength in the furnace cell, one
would expect it to be reduced at
higher temperatures. This result sug-
gests that the equilibrium boundary
has a slope close to the maximum
shown in Fig. 30.
Application of Kyanite-Sillimanite
Equilibrium
Assumption of equilibrium of the
kyanite-to-sillimanite isograd reac-
tion seen in metamorphic terrains,
coupled with an independent estimate
of the temperature of the isograd,
permits the estimation of load pres-
sure during metamorphism. For ex-
ample, the temperatures inferred
from oxygen isotope fractionation
between quartz-magnetite and quartz-
ilmenite pairs by Garlick and Epstein
(1967) may be used to estimate load
pressures during metamorphism of
the areas studied by them. Thus the
temperature of the isograd in north-
ern Idaho lies between that of speci-
mens AH-2096, 600°C, and AH-243,
685 °C. These temperatures corre-
spond to pressures between 4.7 and
7.7 kb (from Fig. 30). For the area
studied by Garlick and Epstein in
Duchess County, New York, the pres-
sure of the isograd would be slightly
below 5.0-6.3 kb; and for the central
Connecticut area, slightly above 6.5-
7.5 kb. The estimates given above do
not take into account possible errors
in the measurement of the Oi8/Oi6
ratio of the natural minerals and in
the experimentally determined refer-
ence system, quartz-magnetite-water
(O'Neil and Clayton, 1964).
Determination of kyanite-silliman-
ite equilibrium, a dehydration isograd
equilibrium, and the temperatures of
both isograds in the field, may be used
to calculate limits for the extent to
which /h2o (fugacity of water) at the
dehydration isograd departs from the
value of /h2o in a pure water phase
at the same Ptotai and T, An example
is given below.
The mineralogical changes called
the "second sillimanite isograd" in
kyanite-sillimanite type metamor-
phism are closely (but not exactly)
defined by the equilibrium
Muscovite + quartz <^ K-feldspar +
sillimanite + H20 (I)
(See, for example, Evans and Gui-
dotti, 1966.) Reaction I has been
studied by Evans (1965) at water
pressures up to 4 kb. In the area
of central Connecticut mentioned
above, Garlick and Epstein (1967)
placed the temperature of the second
sillimanite isograd at about 700 °C.
Since (1) sillimanite, not kyanite,
396
CARNEGIE INSTITUTION
was the stable ALSiOs polymorph
and (2) pressure at the kyanite-
sillimanite isograd in this area was
greater than 6.5 (see above), then,
if dP, dT of the geothermal gra-
dient was positive, the pressure at
the second sillimanite isograd lay be-
tween G.b and 8 kb. The values of /H„o
that would allow equilibrium I to
pass through 700°C, 6.5 kb, or 700°C,
8 kb. are therefore limiting.
These values of /H.,0 were derived
by constructing a logi0 /h2o against
air plot for three of Evans's points
on equilibrium I (the fourth point, at
1050 bars, does not permit a reason-
able interpretation of the other
three) . But Evans's data apply to the
condition Ph2o = Psoiws, so the log
/h9o of each point was next adjusted
to total pressures (Psoiids) of 6.5 and
8 kb (see Orville and Greenwood,
1965, for a lucid exposition of the
theory behind this technique) using
the relation
A log / =
'A' solids \l ■* experiment )
2.303RT
(II)
(Eugster and Wones, 1962, as quoted
by Orville and Greenwood) ; the value
taken for A7solids was —5 cc/mole.
The equilibrium I was thereby ex-
pressed in terms of log /h2o and 1/T
at two constant total pressures (6.5
and 8 kb), and the value of /h2o at
700 °C for each case was read from
the graph: At 700°C, 6.5 kb, /h2o «
4050 bars; at 700 °C, 8 kb, /h2o ~ 4450
9 -
o
O
CD
^_
t/J
X
o
5r-
Q_ -
3-
1
1 1 1
1 1 1 i 1
// '
Kyanite
Andalusite
/ -
- a
O This study
a / -
+
X Newton, 1966
^<^'^- ■•//' ^
-A uncertainty
— ^^
^^^ /^::';:;:yy ^Intersection uncertainty
—
/ /~~~~~^K-S uncertainty, this report
-
/\
! !/ 1 1 1 1 1 1 1
1 1
:oo
400
500 600
Temperature , °C
700
800
900
Fig. 31. P-T coordinates of the points where kyanite-andalusite equilibrium was reversed in the
presence of quartz and water, both in this study and in that of Newton (1966). Limits of uncer-
tainty of kyanite-sillimanite equilibrium from Fig. 30.
GEOPHYSICAL LABORATORY 397
bars. These values may be compared manite and kyanite-andalusite equi-
to the fugacities of water in a pure libria are so similar. To locate the
water phase under the same Ptota]-T position of the triple point with pre-
conditions: 5450 and 7950 bars, re- cision it will obviously be necessary to
spectively. (All values of /Ho0 were in- determine sillimanite-andalusite equi-
terpolated from the compilation of librium directly.
Anderson, 1964.) The results of the
calculation above may be compared The Composition of Synthetic
to the very low values of /h2o inferred Fe-Staurolite
for rocks from other areas by Wones s. W. Richardson
and Eugster (1965, p. 1267) and -. (y „ , Q_
Greenwood (1963, pp. 345-349). It 0J*™. ,year {YeaL?°l ?' PP* ?/~
• -a- j- j j-V 4. ±1 4.- n 251) it was reported that staurolite
is indicated that the assumption PKn0 , -. , ±.i • j • i i
~ p ~ p • r -f v •*. could be synthesized as a single phase
JL Jlui™~ d> ™?llc1lt or ex?„2„ from oxide mixes of composition
(Fyfe Turner, and Verhoogen, 1958, 4Fe0.9AL03-7i/2Si02 (molar propor-
?Ln16~18^7^e^r111d1 Ve^°gen, tions) with water. It was sug-
1960 pp. 506-507; Winkler, 1965, pp. ted f ollowing Schreyer and Chin-
11-14) m the facies concept as cur- ner (1966) that the discrepancy be.
rently applied to carbonate - free tween tMg composition and the
rocks, cannot be taken for granted; «ideal„ staurolite composition of
(19™^^ Naray-Szabo and Sasvari (1958)
\l™i> iybt>; clearly nave practical could be accounted for by a substitu-
significance at the second sillimanite tion of 4 h drogen atoms for x out of
isograd m central Connecticut. 16 gilicon atomg -n the structurej ^v.
rjr ., a j 7 .. r, .7., . ing 4 (OH) instead of 2 (OH) per
Kyamte-Andalusite Equilibrium 4g (Q?0H) formula ^ Analygis f Qr
Following Newton's (1966) experi- water provides a partial test for the
ments in solid pressure-media appa- suggestion.
ratus, we have made a preliminary A problem was encountered in at-
investigation of the metastable exten- tempting to synthesize, in a %-inch
sion of kyanite-andalusite equilibrium bore piston-cylinder apparatus, large
in gas-pressure apparatus, employing volumes of single-phase staurolite on
the experimental methods used in the the composition 4*9«71/2> because tem-
study of kyanite-sillimanite equilib- perature gradients over the sample
rium (above) . capsule cause enrichment of silica in
The results of runs containing ex- the hotter parts of the vapor phase;
cess water are shown in Fig. 31, on opening, the ends of the quenched
which also shows the coordinates of capsule were observed to contain
the limiting runs made by Newton, staurolite + corundum + opaque
There is excellent agreement between oxide, and the center, staurolite +
the two sets of data, which lends quartz. Therefore, a sample of stauro-
credence to the pressure-correction lite + quartz was prepared on 4-9 »8
procedure employed by Newton. But composition (see Year Book 65 for
even if limits of uncertainty are as- experimental details) .
signed on the basis of both sets of The sample of staurolite + quartz
data, there is still considerable uncer- was analyzed for water by Dr. I.
tainty as to the position of the triple Friedman, U. S. Geological Survey,
point where kyanite, sillimanite, and Denver, who reported 2.0 ± 0.2 wt %
andalusite may coexist at equilibrium, water. On the assumption that all the
because the slopes of kyanite-silli- iron in the original mix remains in
398
CARNEGIE INSTITUTION
TABLE 14. Synthetic Staurolite Composition
I II III*
Si02
27.94
Si02
26.66
Si
7.51
Al203
53.35
Al203
54.30
Al
18.03
FeO
16.71
FeO
17.01
Fe
4.01
H20
2.0
100.00
H20
2.04
100.01
(OH)
3.83
I, composition of material synthesized from
4-9-8 mix. II, composition of staurolite. Ill,
atomic proportions of staurolite.
* Per 48(0,OH) formula unit.
the ferrous state, the composition of
the sample is calculated as I of Table
14. The composition of the staurolite
in the sample is given as II of Table
14 ; II has been recalculated as atomic
proportions of a 48(0, OH) formula
unit in III.
The value 3.83(011) per formula
unit is well within experimental
uncertainty of the 4.0 (OH) in the
postulated staurolite formula. Crys-
tallographic methods will have to be
used to decide (1) whether the Naray-
Szabo and Sasvari formula does in-
deed reflect the fundamental struc-
ture of staurolite and (2) whether
synthetic and natural staurolites do
differ from the fundamental structure
by the substitution 4H <=± Si. It is still
possible that the observed silica de-
ficiency and observed water excess
are caused by two separate substitu-
tions.
The Stability of Fe-Staurolite
+ Quartz
S* W. Richardson
The compatibility of staurolite with
quartz is characteristic of a rather
narrow zone in some types of regional
metamorphic belts. Experimental
determination of the stability field of
staurolite + quartz should therefore
yield information on the pressure and
temperature prevailing during the
formation of those metamorphic
types.
As an approach to this problem the
relations in the system Fe-Al-Si-O-H
between Fe-staurolite, Fe-cordierite,
almandine, Fe-chloritoid, quartz, and
sillimanite have been investigated in
terms of temperature and fluid pres-
sure under conditions where the com-
position of the fluid was controlled by
the quartz-fayalite-magnetite oxygen
buffer (Eugster and Wones, 1962).
Because a full report of this work is
in preparation and will be published
shortly, only a summary of the re-
sults is given here. The reactions
studied were:
Fe-staurolite + quartz *± almandine
+ sillimanite + H20 (I)
Fe-staurolite + quartz ±± Fe-cordierite
+ sillimanite + H20 (II)
Fe-chloritoid + sillimanite ^
Fe-staurolite + quartz + H20 (III)
Fe-cordierite ±± almandine
+ sillimanite + quartz (IV)
The starting material for all ex-
periments had the bulk composition
4FeO-9Al203-15Si02 (with excess
H20), which lies at the intersection
of the joins staurolite-quartz-fluid
and cordierite-sillimanite-fluid. Start-
ing material for the investigation of
each reaction consisted of equal
amounts of previously synthesized
reactants and products (except that
natural sillimanite was used). At the
conclusion of the run, direction of re-
action was determined by both X-ray
and optical examination.
The pressure and temperature of
every run in which reaction could be
detected are shown in Fig. 32. The
data in Fig. 32 have been used to
evaluate the chemographic relations
of the system Fe-staurolite, Fe-cordi-
erite, almandine, Fe-chloritoid, silli-
manite, and quartz in equilibrium
with a fluid whose composition is de-
fined by the Q-F-M buffer (Fig. 33).
In the only other set of rotations that
GEOPHYSICAL LABORATORY
399
SiOc
QUARTZ
Fe-CORDIERITE#
SILLIMANITE
ALMANDINE.
Fe-CHLORITOID# Fe-STAUROLITE
FeO
AloO
2U3
o
o
ALM + SILL
4- Qz
CORD+SILL n
700
800
Q- r -
T3
3
5-
3-
1 1
i i
□
-
STAUR
+
Qz
□
□ y
/-
A'
•*•
jym m
m
CORD
—
■
+
SILL
400 500 600 500 600 700
Temperature, CC
600
700
800
Fig. 32. Fluid pressure-temperature projections of the experiments in which reaction could be
detected. Open squares, low-temperature assemblage grew; filled squares, high-temperature
assemblage grew. All experiments contained excess fluid buffered by Q-F-M.
could reasonably be allowed by the
data, the invariant point (cord) is at
very low (probably fictive) rather
than very high fluid pressure and is
metastable rather than stable; this
situation must come about if reac-
tion I has a lesser dP/dT than reac-
tion III.
Figure 34 is a Pnuia-T plot of the
stable parts of those reactions in Fig.
33 that involve quartz. They are not
necessarily the most stable reactions
in the system FeO-Al203-Si02-OH
fluid: For example, it has not been
demonstrated that the assemblage
diaspore-f errocarpholite-quartz is less
stable than any of the assemblages
illustrated in Fig. 34. More realisti-
cally, kyanite (this report, pp. 392-
396) and andalusite (Newton, 1966)
are more stable than sillimanite over
part of the P-T range illustrated; as
more information on the ALSi05 sys-
tem becomes available, the rotations
in Fig. 34 may be recalculated in
terms of the stable polymorphs. Nat-
urally occurring rocks indicate that
assemblages involving Fe-gedrite may
possibly have a stability field over
part of the P-T-X range considered
here.
The salient features of Fig. 34 are
(1) the low-pressure termination of
the Fe-staurolite + quartz field and
its rather narrow temperature inter-
val (^150°C) and (2) the upper-
pressure termination of the Fe-cordi-
erite field. Although no rocks are
known whose compositions lie in the
system Fe-Al-Si-O-H, the probable
400
CARNEGIE INSTITUTION
Fig. 33. Fluid pressure-temperature plot of the geometrical relations between reaction curves
and invariant points involving the six solid phases staurolite, cordierite, chloritoid, almandine, silli-
manite, and quartz.
GEOPHYSICAL LABORATORY
401
if)
o
O
LSE
if)
if)
CD
Ll_
500
600 700
Temperature, °C
800
Fig. 34. The reactions separating the stability fields of quartz-bearing assemblages. Solid
curves were experimentally determined (see Fig. 32). Dashed curves are inferred from Fig. 33.
No reliance should be placed on the precise location of the dashed curves or the invariant
point (SILL). The compositions of the phases in the stippled compatibility triangles may be seen in
Fig. 32.
effects of other compositional vari-
ables on the relations shown in Fig.
34 may be estimated qualitatively and
some preliminary inferences may be
drawn. It should be emphasized that
this treatment is no substitute for
continued work on staurolite stability
in more complex systems.
The occurrence in natural rocks of
cordierites with Mg/Fe ratios much
greater than those of coexisting stau-
rolite limits the stability of natural
staurolite + quartz to higher fluid
pressures than the li/2 kb shown in
Fig. 34 and also raises the pressure
at which cordierite becomes unstable
(see Schreyer and Yoder, 1964, and
this report, pp. 387-388, for a de-
termination of the pressure stability
of Mg-cordierite) . Reaction "with
402
CARNEGIE INSTITUTION
muscovite (Chinner, 1965), which is
present in many pelitic rocks, limits
the thermal stability of staurolite +
quartz to temperatures lower than
those of reaction I. The /0>, in natural
rocks may differ from that imposed
by the quartz-fayalite-magnetite buf-
fer: large changes in /0a wiU probably
render unstable the reactions investi-
gated (Fisher, Year Book 65), but
small changes may only affect their
position on the Pnuia-T diagram.
Tentative conclusions are: (1) The
compatibility of staurolite + quartz
in certain types of regional meta-
morphism indicates a fluid pressure
greater than l1/? kb. (2) In musco-
vite-bearing pelites the presence of
staurolite and quartz indicates tem-
peratures lower than 675°-700°C but
greater than 500°-550°C. (3) In
those types of regional metamor-
phism where cordierite is characteris-
tically absent at all grades, fluid
pressure was probably greater than
31/2 kb.
The Stability of Fayalite
D. R. Wones* and M. C. Gilbert
Knowledge of the stability of oli-
vine as a function of oxygen pressure
has been used in recent years to
establish oxidation-reduction trends
in igneous and metamorphic rocks.
The redox reaction of the iron end
member of the olivine system
3Fe2Si04 + 02 *± 2Fe304 + 3SiO,
has also been used as an important
reference state in experimental sys-
tems. Previously, this reaction was
determined only at high temperatures
or by thermochemical calculations.
We have studied fayalite oxidation at
600°, 700°, and 800°C at 800 bars
total pressure and at 700 °C at 2000
bars total pressure using a hydro-
thermal apparatus designed by Shaw
*U.S. Geological Survey.
(1963) . This apparatus allows hydro-
gen pressure to be varied at will and
measured independently of the total
pressure. Knowledge of the thermal
decomposition of water in the charge
then leads to a determination of oxy-
gen fugacity.
Various synthetic materials were
used as reactants. Fayalite was syn-
thesized from mixtures of FeCoCV
2H20 + Si02, Fe + Si02, Fe203 +
Si02, and Fe304 + Si02 (glass). The
d130 of the various synthetic f ayalites
ranged from 2.8263 to 2.8283 A. Mag-
netite was made by reducing Fe2Os.
All reactants gave uniform and con-
sistent results.
The reaction was reversed under
four sets of conditions, and the data
are shown plotted in Fig. 35 recalcu-
lated to 1 bar. The pressure correc-
tion procedure is discussed by Eugs-
ter and Wones (1962, pp. 90-92).
Size of symbols represents the actual
experimental uncertainties in T, Ph2,
and Ptotai for the bracketing runs.
For the calculation of /o2 from the
run data, the equilibrium constant for
the dissociation of H20 was taken
from the JANAF Tables (1960).
The agreement among our data is
excellent. Equally good is the agree-
ment with the earlier work of Schenk,
Franz, and Leymann (1932) and with
the recent work of Schwerdtf eger and
Muan (1966), plotted in the lower
left of Fig. 35.
Our data permit estimation of the
heat of reaction over the temperature
range 600°-800°C. The value ob-
tained, — 117,400 ± 2500 cal, appears
to agree with published heats of re-
action within the quoted uncertain-
ties.
One result of this study is to shift
the fo2-T curve for the oxidation of
fayalite to somewhat higher /o2 values
compared with those given in Eugster
and Wones (1962, Table 2). A report
giving full details of the experimental
GEOPHYSICAL LABORATORY
403
work and some of the implications is
in preparation.
SULFIDES AND RELATED
MINERALS
The pronounced correlations often
seen between certain magmatic ores
and the rocks in which they typically
occur are particularly well demon-
strated by coexisting noritic rocks
and copper-iron-nickel sulfides in the
Sudbury District, Ontario, Canada.
Massive field evidence indicates that
the sulfides were carried from a deep-
seated magma chamber either dis-
solved in or as emulsified droplets in
noritic liquid.
On cooling, sulfide liquid became
partly segregated from the crystalliz-
ing rock to produce ore deposits.
Some of the sulfide droplets remained
(/>
D
_Q
17
CM
O
M—
o
18
600
Temperature, °C
700
800
Standard State of Solid Phases 1 Bar
Reaction Reversed
$ at 800 bars PT
0.9
X 1000 ,°K
Fig. 35. f0 -T diagram showing experimental runs that limit the reaction 3Fe2Si04 + 02^
2Fe304 + Si02. Open symbols signify growth of FeaCXj + Si02; solid symbols signify growth of
Fe2Si04 (for the experiments reported here).
404 • CARNEGIE INSTITUTION
dispersed in the norite. Investigations relations in a portion of the Fe-S-0
of coexisting hypersthene and augite system. Experiments have also been
in the ore-bearing norite show that conducted in the Mg-Fe-Si-O-S sys-
on cooling from magmatic conditions tern to determine the compositions of
these minerals only partly responded coexisting pyrrhotite and pyroxene.
to the changes in temperature. Water is a ubiquitous participant
The Fe/Mg ratio of hypersthene in geological events. Its remarkable
versus that of augite indicates that influence on the melting relations of
reequilibration between these miner- numerous silicates is well known. We
als ceased when the cooling process have now studied the melting rela-
reached 1100°C. Because of the close tions of several sulfides in the pres-
association observed in the field be- ence of water and find no observable
tween these pyroxenes and the typical lowering of melting temperatures due
ore sulfides, it is likely that sulfides to its presence.
and silicates were exposed together to Sulfide mineral assemblages are
a temperature at least that high. the result of complex reactions oc-
Thus, knowledge of the high-tern- curring during the cooling period of
perature behavior of the appropriate ore masses. Applications of phase dia-
sulfide and silicate systems and of the grams to sulfide assemblages in ores
combined rock-sulfide systems is es- demonstrate that extensive equilibra-
sential to our understanding of these tion takes place among these minerals
magmatic deposits. Accordingly, we during the prolonged process of cool-
have studied the phase relations in ing from the temperature of ore dep-
the Cu-Fe-S, Cu-Ni-S, and Fe-Ni-S osition. For this reason it is neces-
systems over wide temperature sary to investigate the phase rela-
ranges. The results of our investiga- tions in pertinent systems also at low
tions of the Cu-Fe-S and Cu-Ni-S sys- temperatures. The stabilities of viola-
terns from 1200° to 700 °C are pre- r*te and of various members of the
sented in this report. Knowledge of violarite-polydymite solid solution se-
the ternary systems permitted us to ries have now been determined; the
explore the complex Cu-Fe-Ni-S sys- field boundaries of the Fe1^S-Ni1..S
tern, which includes more than 99% solid solution series have been brack-
of the sulfides occurring in Sudbury- eted down to 300 °C and an effort has
type ores. The phase relations in the been made at 600° and 400°C to esti-
Cu-Fe-S and Cu-Fe-Ni-S systems mate the variation in the partial pres-
demonstrate that a copper-rich sulfide sure of sulfur as a function of solid
liquid coexists with the pyrrhotite solution composition.
phase at temperatures as low as
850 °C. Segregation of such copper- High -Temperature Phase
rich liquid from pyrrhotite may well RELATIONS IN THE Cu-FE-S SYSTEM
be responsible for the high concen- G. Kullerud
(rations of copper, in the form of The phase relations at high tem-
chalcopyrite, observed in certain ores, peratures in the Cu-Fe-S system are
The sulfides in Sudbury-type ores of considerable potential importance
are ubiquitously associated with con- to better methods of extracting cop-
siderable amounts of magnetite. To per from copper-iron sulfide ores,
assess the role of this mineral during Knowledge of the behavior of phases
ore formation we have investigated in this system over a wide tempera-
melting relations between Fe304 and ture range is of interest to students of
the common ore sulfides and have de- copper-iron-sulfur and copper-iron-
termined the high-temperature phase nickel-sulfur ore deposits.
GEOPHYSICAL LABORATORY 405
Investigations of liquidus relations geneous liquid (Jensen, 1947). When
in the ternary Cu-Fe-S system were it first appears chalcocite contains no
undertaken by Schlegel and Schiiller iron and its sulfur content is
(1952), Greig, Jensen, and Merwin 20.2 wt %. Stoichiometric Cu2S in
(Year Book 5h) , and Kullerud (Year comparison contains 20.15 wt % S.
Book 63) . The results of these studies This cubic chalcocite-type phase on
together with new data obtained by further cooling forms solid solution
differential thermal analysis experi- that extends both toward Cu (to
ments now make it possible to draw Cu2S) and toward S. This phase
phase diagrams at various elevated forms extensive ternary solid solution
temperatures. with decreasing temperature.
A field of liquid immiscibility ex- The phase relations in the Cu-Fe-S
ists on the Cu-Cu2S join above system are shown at 1100°C in Fig.
1105°C. Schlegel and Schiiller found 36. This isotherm was drawn on the
that this field extends into the ternary basis of the results of the above-
system and that the presence of iron mentioned investigations and addi-
lowers the temperature of the ap- tional DTA experiments. It is noted
pearance of the field of liquid immis- that at 1100 °C, only 29 °C below the
cibility to 1077 °C. The two-liquid field temperature of congruent melting of
is extensive at elevated temperatures Cu2S, as much as 12 wt % Fe can go
and reaches within 5 % of the Fe-FeS into solid solution. Because of the
join at 1400 °C. considerable ternary solid solution
Kullerud (Year Book 63) showed displayed by this phase, we refer to it
that a second liquid immiscibility field as bornite solid solution and retain
spans the sulfur-rich portion of the the name chalcocite for compositions
system above 1083 °C. These liquid of or near Cu2S.
immiscibility fields are separated by a The divariant field containing bor-
field of homogeneous liquid. Thus, at nite solid solution and ternary liquid
very high temperatures, for instance is very narrow at 1100° C. Stoichio-
1400 °C, the phase relations in the metric bornite (Cu5FeS4) with 63.31
Cu-Fe-S system are dominated by wt % Cu, 11.13 wt % Fe, and 25.56
these three fields and the only solid wt % S contains 1.5 wt % more sulfur
phase in the system is metallic iron, than the maximum amount soluble
which melts at 1534°C. On cooling, in the bornite solid solution at 1100° C
the next solid phase to appear is pyr- and thus lies within the ternary liquid
rhotite (Fe^S), which crystallizes field of Fig 36, The hexagonal pyrrho-
from homogeneous liquid at 1192° tite phase contains a maximum of
(Jensen, 1942) . When it first appears about 5 wt % Cu in solid solution and
this phase contains no copper, it is is surrounded by a wide pyrrhotite +
strictly binary, and its sulfur content ternary liquid divariant field at
is 38.4 wt %. In comparison troilite 1100°C. On cooling below 1100°C
(stoichiometric FeS) contains 36.48 the liquid immiscibility field con-
wt % S. On further cooling the pyr- taining copper liquid and ternary
rhotite phase forms solid solutions ex- liquid narrows rapidly. At 1077 °C
tending both toward Fe (to FeS) and the two liquids can no longer coexist
S from the original composition and in the presence of vapor, and tie lines
at maximum can accommodate sev- are instead established between bor-
eral per cent copper in its hexagonal nite solid solution and iron solid solu-
NiAs-type structure. tion containing about 10 wt % Cu:
At 1129°C the chalcocite (Cu2S)- Cu liquid + ternary liquid -> bn ss +
type phase crystallizes from homo- Fe ss. Below this invariant point Cu
406
CARNEGIE INSTITUTION
Fig. 36. Phase relations in the Cu-Fe-S system at 1100°C. All phases and phase assemblages
coexist with vapor. The point S shows the average Sudbury composition when projected from
the Cu-Fe-Ni-S tetrahedron onto the Cu-Fe-S boundary plane.
liquid r bn ss + Fe ss coexist to
1067 °C. At this temperature the Cu
liquid crystallizes in a eutectic on the
Cu-S binary join, and metallic Cu
coexists with bornite solid solution
near Cu2S composition and with Fe
solid solution. Below 1067°C Cu con-
taining 2-3 wt % Fe in solid solution
coexists with bornite solid solution in
a divariant field that widens with de-
creasing temperature as the ternary
liquid diminishes in size.
The Fe ss-bn ss divariant field in-
creases in width with decreasing
temperature as the ternary liquid
field diminishes in size. At about
1040 °C the univariant assemblage bn
ss (containing about 21 wt % Fe) 4-
ternary liquid + Fe ss (containing
about 5 wt % Cu) + vapor is es-
tablished.
In the sulfur-rich portion of the
system liquid immiscibility can no
longer exist on the Fe-S join when on
cooling the temperature reaches
1083 °C. Below this temperature hex-
agonal Fei-aS is stable with liquid sul-
fur. Both the ternary liquid field and
the liquid immiscibility field retreat
from the Fe-S boundary and give rise
to a univariant field containing ter-
nary liquid, liquid sulfur, and pyrrho-
tite, in addition to vapor.
Solubility of Cu in the pyrrhotite
GEOPHYSICAL LABORATORY
407
T=IOOO°C
Fig. 37. Phase relations in the Cu-Fe-S system at 1000°C. All phases and phase assemblages
coexist with vapor.
phase increases with decreasing tem-
perature, and at 1000 °C the solubility
is about 7.0 wt % as noted in Fig. 37.
The ternary liquid field decreases
markedly in size between 1100° and
1000 °C as is seen by comparison of
Figs. 36 and 37. This is partly due to
the expansion of the homogeneous
bornite solid solution field and the
divariant bornite solid solution +
liquid fields. Figure 37 shows that the
maximum solubility of Fe in the bor-
nite solid solution is about 22 wt % at
1000 °C, an increase of 10 wt % be-
yond the limit of solubility at 1100°C
(Fig. 36) . At 1000 °C Cu5FeS4 compo-
sition lies barely within the bornite
solid solution field. On cooling below
1000 °C a series of important events
takes place to change the phase rela-
tions. At 988 °C a eutectic exists on
the Fe-FeS join at 31.3 wt % S, and
the ternary liquid field below this
temperature recedes from the binary
join into the ternary system.
At about 960 °C the chalcopyrite
phase crystallizes from the ternary
homogeneous liquid. When it first ap-
pears this phase contains about 35
wt % Fe, 32.5 wt % Cu, and 31.5
wt % S, corresponding to a metal-to-
sulfur ratio of about 2: 1.75. On cool-
ing, the sulfur content of the chal-
copyrite phase increases rapidly, and
the structure of the mineral can
tolerate larger variations in the
408
CARNEGIE INSTITUTION
Cu-to-Fe ratio. At about 940°C the
ternary liquid field is divided as tie
lines are established between the
bornite and pyrrhotite solid solutions.
At this temperature there is maxi-
mum solid solution of iron in bornite
(about 32 wt %) and of copper in
pyrrhotite (about 7.5 wt %). The
metal-rich liquid field gradually di-
minishes in size with decreasing tem-
perature and disappears in a eutectic
reaction L -> bn ss + po ss + Fe ss
at about 910 X. Schlegel and Schuller
(1952) gave the eutectic composition
as 30.5 wt % Cu, 39.5 wt % Fe, and
30.0 wt cc S; whereas Greig, Jensen,
and Merwin (Year Book 5J>) gave
21.5 wt % Cu, 49.0 wt °/o Fe, and
29.5 wt °/o S. The latter values are in
agreement with results obtained in
the present study.
Tie lines between the chalcopyrite
phase and pyrrhotite are established
at about 935 °C where maximum solid
solution of chalcopyrite in pyrrhotite
(about 7 wt % Cu) exists. The chal-
copyrite phase at this temperature
contains a maximum of about 37.5
wt c/o Fe. The bornite and chalcopy-
rite solid solutions become stable to-
gether at approximately 930 °C.
The phase relations at 900 °C are
shown in Fig. 38. It is noted that
CuFeS2 composition still lies within
the homogeneous ternary liquid field.
Below 900 °C the ternary liquid
T=900°C
Cu2S-^
Fig. 38. Phase relations in the Cu-Fe-S system at 900°C. All phases and phase assemblages
coexist with vapor.
GEOPHYSICAL LABORATORY 409
field decreases in area as its compo- iron sulfides (hexagonal + mono-
sition becomes increasingly copper clinic pyrrhotite 4- about 1% py rite)
rich. At 860 °C the liquid composition as Fe7S8, the estimated composition
has retreated sufficiently toward the projects into the Cu-Fe-S system in
Cu-S join so that tie lines between the point (6.1 wt % Cu, 55.1 wt °/o
ternary liquid and pyrrhotite can be Fe, and 38.8 wt % S) marked S on
replaced by chalcopyrite-liquid sulfur Fig. 36. At 1100°C bulk composition
tie lines. Similarly, tie lines are estab- S is accounted for by a mixture of
lished between the most iron-rich hexagonal pyrrhotite containing
member of the bornite solid solution about 5 wt % Cu in solid solution and
and liquid sulfur at about 850 °C. At a liquid (L) containing 10-11 wt %
813 °C the ternary liquid field disap- Cu. It is noted that a relatively cop-
pears in a binary monotectic reaction per-rich liquid may segregate from a
on the Cu-S join. Below this tempera- large range of bulk compositions
ture liquid sulfur coexists stably with within the pyrrhotite-liquid divariant
the entire bornite solid solution. field.
Pyrite appears on the Fe-S join at On cooling below 1100°C the pyr-
743 °C, and tie lines are established rhotite structure can accommodate
between this phase and the chalcopy- more Cu, and at 1000 °C the solubility
rite solid solution at 739 °C. The phase is about 7.0 wt %. The average Sud-
relations below 739 °C, except for bury composition when projected onto
minor variations in the homogeneous the Cu-Fe-S plane at 1000° and 900 °C
solid solution fields, remain un- is accounted for by a mixture of
changed to 700 °C. Yund and Kullerud fcopper-containing pyrrhotite and cop-
(1966) studied the phase relations in per-rich liquid. Segregation of Cu-
the entire Cu-Fe-S system from 700° rich liquid is therefore possible also
to 200 °C and to below 100 °C in se- at 1000 °C (see Fig. 37) and at 900 °C
lected portions of the system. Thus, (see Fig. 38) but is not possible below
at this time we possess considerable 860 °C. At this temperature tie lines
knowledge of the behavior of the are established between chalcopyrite
phases in this system over a very ex- and liquid sulfur, prohibiting coexist-
tensive temperature range. ence of pyrrhotite and ternary
Applications of phase relations in liquid. Thus at high temperature a
the Cu-Fe-S system to ore mineral mechanism exists in the Cu-Fe-S sys-
assemblages originally deposited at tern that may be responsible for cer-
temperatures below 700 °C were dis- tain copper-rich segregations ob-
cussed in considerable detail by Yund served in this type of ore.
and Kullerud (1966). In the follow-
ing, an example will be given of ap- High-Temperature Phase
plication of phase relations to ores Relations in the Cu-Ni-S System
of magmatic origin Hawley (1962) Q KuRerud and Q Moh ,
estimated that on the average the
magmatic Cu-Ni-Fe deposits of the Matousek and Samis (1963) stud-
Sudbury region contain chalcopyrite, ie^ the metal-rich portion of the Cu-
pentlandite, and iron sulfides (con- Ni"S system at 1200°C and presented
sisting of two kinds of pyrrhotite and a diagram showing the phase rela-
small amounts of pyrite) in the ratio tions at this temperature. The Ni3S2-
15:15:70 wt %. Cu2S join was investigated by Hay-
If we calculate chalcopyrite as ward (1915). The results of these
CuFeS2, pentlandite as (Fe,Ni)9S8, investigations together with new data
and the average composition of the * university of Heidelberg.
410
CARNEGIE INSTITUTION
obtained by differential thermal anal-
ysis and quenching experiments now
make it possible to draw phase dia-
grams at various elevated tempera-
tures.
At very high temperatures the
phase relations in this ternary sys-
tem are dominated by two extensive
fields of liquid immiscibility, which
are separated by a field containing
homogeneous liquid. One of the liquid
immiscibility fields is situated in the
metal-rich portion of the system, and
the other occupies a large area near
the sulfur corner of the system. The
metal-rich liquid immiscibility field
decreases rapidly in size with de-
creasing temperatures. The phase
relations at 1200 °C in the entire sys-
tem are shown in Fig. 39. The metal-
rich liquid immiscibility field at this
temperature extends from 1.5 to 19.8
wt °/o S along the Cu-S boundary and
about 25 wt % toward the Ni com-
position corner. The sulfur-rich liquid
immiscibility field is noted to span
the system at 1200 °C. This very large
field extends from 27 to > 98 wt % S
along the Cu-S boundary and from
54.5 to > 97.6 wt % S along the Ni-S
join. The phase relations in the cen-
tral portion of the ternary system at
1200 °C are dominated by a large
liquid field, which is bounded by
the liquid immiscibility fields, as
discussed above, and by the divariant
Fig. 39. Phase relations in the Cu-Ni-S system at 1200°C. All phases and phase assemblages
coexist with vapor.
GEOPHYSICAL LABORATORY
411
liquid + NiCu alloy field shown in
Fig. 39. The liquid field extends to
about 21 wt % Ni along the Cu-Ni
join and to about 88 wt % Ni along
the Ni-S boundary. The only solid
phase in the system at this tempera-
ture is an NiCu alloy, which extends
along the Ni-Cu join from the Ni
corner to about 62 wt °fo Cu. This
phase contains little or no sulfur in
solid solution. On cooling below
1200 °C the cubic chalcocite (Cu2S)
phase crystallizes on the Cu-S join
from homogeneous liquid at 1129°C
(Jensen, 1947).
The metal-rich liquid immiscibility
field diminishes rapidly in size. It is
not known whether it gradually re-
treats toward the Cu-S join to dis-
appear at 1105°C, where chalcocite
becomes stable with pure liquid Cu, or
whether it retreats from the Cu-S
join at 1105°C and persists to some
lower temperature in the ternary sys-
tem. Pure Cu crystallizes at 1083 °C,
and below this temperature cubic
chalcocite coexists with metallic Cu
and with CuNi alloy. Cu and Ni form
a complete solid solution series below
the melting point of Cu. The cubic
vaesite (NiS2) phase crystallizes on
the Ni-S join from homogeneous
liquid at 1007 °C. On further cooling
it becomes stable with liquid sulfur
at 991 °C, at which temperature the
sulfur-rich liquid immiscibility field
retreats from the Ni-S boundary into
the ternary system, creating a uni-
780°C
Cu2Sss+NiS2+L
NiS+Ni3.;rS2+L
Fig. 40. Phase relations in the Cu-Ni-S system at 780°C. All phases and phase assemblages
coexist with vapor.
412
CARNEGIE INSTITUTION
700 °C
Cu2Sss +
Ni3-*S2ss
+ L
NlhrS
Ni3±<rS2
Cu9S5N
Cu2 S
Fig. 41. Phase relations in the Cu-Ni-S system at 700°C. All phases and phase assemblages
coexist with vapor. The amounts of solid solution of Cu in the nickel sulfides and of Ni in the
chalcocite-digenite phase were determined at this temperature.
variant field that contains ternary
liquid — sulfur liquid -I- NiS2. The
hexagonal Ni^S phase crystallizes
from homogeneous liquid at 992 °C. Be-
low this temperature the phase rela-
tions remain essentially unchanged to
813 °C, where the chalcocite phase,
which forms solid solution extending
from Cu2S to Cu9S5, becomes stable
with liquid sulfur.
At 813 '"C the sulfur-rich liquid im-
miscibility field retreats from the
Cu-S boundary into the ternary sys-
tem. Thus, liquid immiscibility occurs
in the ternary system at lower tem-
peratures than on the binary bound-
aries. The high-temperature form of
the Ni3±a.S2 compound forms at 806 °C
through a reaction involving the
hexagonal aNi^S phase and ternary
liquid. At 788 °C tie lines are estab-
lished between the chalcocite and vae-
site phases, and the ternary liquid
field is thus divided into two liquid
fields, one sulfur rich and the other
metal rich. The phase relations at
780 °C are shown schematically in
Fig. 40. The sulfur-rich ternary liquid
field is noted to be very small at this
temperature, whereas the metal-rich
ternary liquid field is still rather ex-
tensive.
On cooling below 780 °C the sulfur-
rich liquid disappears at about 770 °C
GEOPHYSICAL LABORATORY
413
in a monotectic reaction resulting in
the formation of chalcocite, vaesite,
and liquid sulfur as a stable univari-
ant assemblage. The metal-rich liquid
field decreases rapidly in size below
780 °C and at 760 °C has already re-
treated sufficiently to permit estab-
lishment of tie lines between the chal-
cocite and the aNi^S phases. At
about 720 °C tie lines are established
between the chalcocite and Ni3±a.S2
phases. The phase relations at 700 °C
are shown in Fig. 41.
At this temperature the chalcocite
phase of or near Cu2S composition co-
exists with CuNi alloy ranging in
composition from pure Cu to a maxi-
mum of 79 wt % Ni. The chalcocite
(of near Cu2S composition), CuNi
alloy (containing 79 wt % Ni), and
ternary liquid (composed of 69 wt %
Ni, 12 wt % Cu, and 19 wt % S)
form a univariant assemblage. Chal-
cocite solid solution and ternary liquid
form a relatively wide divariant field.
The maximum copper content of the
ternary liquid is 19 wt % at 700 °C.
The Ni3-rS2 phase, which contains a
maximum of 3 wt % Cu, also coexists
with ternary liquid in a wide di-
variant field. Chalcocite solid solution
(containing about 21 wt % S and
about 0.5 wt % Ni), Ni3±a?S2 (contain-
ing about 2.5 wt % Cu and 27 wt %
S), and ternary liquid (composed of
61.5 wt % Ni, 24.5 wt % S, and 14
wt % Cu) form a univariant assem-
blage at this temperature. The max-
imum solubility of copper in the
aNii-a-S phase is about 1.2 wt % and
in the NiS2 phase, about 1.0 wt % at
700°C.
On cooling below 700 °C eutectic
conditions are encountered on the
Ni-S join at 635° C. The ternary liquid
field retreats from the binary join be-
low this temperature, and a univari-
ant assemblage containing Ni3±arS2 +
Niss + ternary liquid is created. The
ternary liquid disappears in a eutectic
reaction at 572 °C, where liquid of
composition 12 wt % Cu, 67 wt %
Ni, and 21 wt % S crystallizes to
produce a mixture of chalcocite,
Ni3±xS2, and CuNi alloy.
The Cu-Fe-Ni-S System
J. R, Craig and G. Kullerud
Continuing research in the geologi-
cally important Cu-Fe-Ni-S system
has yielded new data both above 650°
and below 550 °C, which now permit
discussion of phase relations in the
temperature range 1000° to 400 °C.
High-temperature phase relations are
shown schematically at 1000° and at
850 °C in Figs. 42 and 43. More de-
tailed discussion of phase relations in
the Cu-Fe-Ni-S system at high tem-
perature will be forthcoming (Craig
and Kullerud, 1967) ; thus only
limited description is presented here.
At 1000°C (Fig. 42) the Fe^S-
Ni^S monosulfide solid solution
(Mss) and the homogeneous quater-
nary sulfide liquid dominate phase
relations in the most geologically
significant portion of the system. The
Mss phase increases in size to span
the Fe-Ni-S face of the compositional
tetrahedron and the sulfide liquid
diminishes in size, retreating toward
the Cu-Ni-S face, as temperature de-
creases. At 850 °C (Fig. 43) the
quaternary sulfide liquid has with-
drawn sufficiently from the Mss to
permit coexistence of cp ss with Mss
compositions ranging from the Fe-S
join (point A) to a maximum of
about 30% Ni (point C) . Tie lines
fan from more Ni-rich Mss composi-
tions to the quaternary sulfide
liquid. Absence of the sulfide liquid
from the geologically applicable por-
tion of the system below about 850 °C
is significant in that there no longer
remains any mechanism by which a
Cu-enriched liquid may segregate
from a typical pyrrhotite ore.
Decrease in temperature below
850 °C results in further withdrawal
414
CARNEGIE INSTITUTION
Fig. 42. Schematic 1000°C isothermal diagram of the Cu-Fe-Ni-S system in the presence of
vapor. Tie lines in sulfur-rich two-liquid field and within the tetrahedron are omitted for clarity.
Regions of the Mss and the homogeneous sulfide liquid are shaded.
of the sulfide liquid toward the Cu-
Xi-S face and establishment of tie
lines between cp ss and «NiS at 830°
± 3°C, cp ss and vs ss at 825° ± 2°C,
bn ss and vs ss at 816° dz 4°C, cc ss
and vs ss at 790° ± 2°C, py ss and
cp ss at 739 °C (Roseboom and
Kullerud, Year Book 57), py ss and
vs ss at 729 °C (Clark and Kullerud,
1963) , cc ss and «NiS at 730° ± 2°C,
and cc ss and Ni3jS2 at 720 °C
(Sproule, Harcourt, and Renzoni,
1960).
Between 720° and 650 °C no signifi-
cant changes occur within the system
aside from gradual reduction in the
volume of the sulfide-rich liquid and
slight shrinkage of the solid solution
fields. Phase relations in the tempera-
ture interval 650° — 550° C were pre-
sented in Year Book 65, where rela-
tions were illustrated schematically
in Figs. 72 and 73.
Although precise phase relations
are not known in the entire Cu-Fe-
Ni-S system below 550 °C, it is possi-
GEOPHYSICAL LABORATORY
415
m>F*h±A
Fe
Fig. 43. Schematic 850°C isothermal diagram of the Cu-Fe-Ni-S system in the presence of
vapor. Many tie lines within the interior of the tetrahedron are omitted for clarity. Note that
the region of homogeneous liquid has diminished from 1Q00°C, that the tAss spans the entire Fe-
Ni-S face, and that cp and (Ni.Fe^^ are stable phases. Points A, B, and C are discussed in the
text.
ble to construct detailed relations in ture from 550 °C {Year Book 65, Fig.
the region of most geologic signifi- 73) to 400 °C — Fig. 44 — results in
cance while presenting general rela- only relatively minor changes in min-
tions in the remainder of the system eral assemblages in the po-py-cp-pn
down to 400 °C. Decrease in tempera- portion of the system. Bornite tie
416
CARNEGIE INSTITUTION
400 °C
FeNi:
Fig. 44. Schematic 400°C isothermal diagram of the Cu-Fe-Ni-S system in the presence of
vapor. The portion of the system most applicable to massive Ni-Cu ores is shown in the enlarged
insert.
lines to the Mss continue to withdraw
from the central portion of the Mss
toward each end such that at 400 °C
bornite coexists only with Mss com-
position containing less than about
10% Xi or less than about 10% Fe.
The Mss field is continually dimin-
ished in thickness but persists in
spanning the Fe-Ni-S face as a homo-
geneous phase. Although tie lines
from cp to po are prohibited above
334°C on the Cu-Fe-S face by the
cb-py assemblage (Yund and Kul-
lerud, 1966) , at 400 °C cp does coexist
with Mss compositions containing
more than 5% Ni.
In the 550°-400°C temperature in-
terval several phases appear for the
first time in the system outside the
po-py-cp-pn portion, whereas other
GEOPHYSICAL LABORATORY
417
phases, present at higher tempera-
ture, undergo structural transfor-
mations. Newly appearing phases in-
clude covellite (507° C), villimaninite
(503 °C), awaruite (503°C), idaite
(501°C), and violarite (461°C).
Stoichiometric Ni3S2 inverts at 556 °C
to the lower-temperature, heazle-
woodite form; because of solid solu-
tion of Cu and Fe in the high-tem-
perature form, however, inversion is
not complete until the temperature
has decreased to 468 °C. Chalcopyrite
of composition CuFeSi.92 inverts
rapidly at 547 °C from the high-tem-
perature cubic to the low-temperature
tetragonal form (Yund and Kullerud,
Year Book 60) ; chalcopyrite composi-
tions that are more S deficient or Cu
rich transform at lower temperatures
and at much slower rates. Chalcopy-
rite of composition CuFeSi.80 remains
cubic at 400°C. The Cu2S-Cu9S5 solid
solution breaks down to two separate
phases below 430° ± 10°C (Kullerud,
1964). At 400°C the separation of
these two distinct phases expresses
itself as two narrow solid solution
fields on the Cu-Fe-S face of the sys-
tem (Yund and Kullerud, 1966).
Between 550° and 400 °C phase as-
semblages within the quaternary sys-
tem are modified by the appearance of
the new phases and by changes in tie
lines on the bounding ternary faces.
The most significant tie-line changes
are reaction of bn + Fe to give po
+ Cu below 475 °C and reaction of id
+ S to give cv + py below 434 °C
(Yund and Kullerud, 1966). The ap-
pearance of violarite on the Fe-Ni-S
face through reaction of py + vs +
Mss at 461 °C, subsequent develop-
ment of solid solution of violarite to-
ward Ni3S4 composition, and the for-
mation of a pn + Ni7S6 pair through
reaction of Mss and Ni3S2 also alter
relations in the central portion of the
system.
The net result of the appearance of
new phases and tie-line changes is a
400 °C isothermal tetrahedron, as
shown in Fig. 44. Relations in the
most geologically significant portion
of the quaternary system are shown
in bold outline and enlarged, whereas
those in the remainder of the system,
which are less well known and of less
geologic interest, are shown by light-
weight lines. Pertinent stable mineral
pairs within this latter portion of the
system are cp 4- vs, cp + viol, cp +
aNiS, py + vill, id 4- vill, id + vs,
bn + vs, bn + «NiS, bn + Ni3S2, and
pn + bn. Although this list does not
include all stable mineral pairs within
the system, it is sufficient to outline
relations in the compositional regions
of the Cu-Fe-Ni-S system of most
probable geologic interest.
Minimum Melting of
nlckeliferous pyrrhotite ores
J. R. Craig and A. J. Naldrett
It has been proposed that many
iron-nickel-copper sulfide ores are
magmatic in origin (Hawley, 1962;
Souch, Podolsky, et al., 1967; Nal-
drett and Kullerud, 1967) . Skinner and
Peck (1967) have reported observa-
tion of a natural nickelif erous pyrrho-
tite-magnetite melt in the Alae lava
lake, Hawaii, at a temperature of
1065 °C. The minimum temperature
at which sulfide magmas of this type
may have been intruded is important
when this proposed origin is consid-
ered. Craig and Kullerud (1967) and
Kullerud, Yund, and Moh (1967)
have discussed melting relations in
the pure sulfide systems involving Fe,
Ni, and Cu. Craig {Year Book 65)
presented the melting of a single
Cu-Fe-Ni-S-0 composition. Naldrett
(elsewhere in this report) presents
data on melting relations in the
Fe-S-0 system and discusses its ap-
plication. The object of this study is
to extend observations on the Fe-S-0
system to compositions including Ni
and Cu as well.
418
CARNEGIE INSTITUTION
Since Cu is generally subordinate
to the Fe, Ni, S, and 0, and, further-
more, since the oxygen is present only
in the form of magnetite, the study
may be reduced to examination of the
appropriate portion of the Fe-Ni-S
system in the presence of magnetite.
The fact that the portion of the
Fe-Xi-S system applicable to massive
Xi-bearing pyrrhotite ores lies within
the compositional confines of the Mss
phase at temperatures above 600 °C
was helpful, since it allowed prepara-
tion of charges with only two reac-
tants, Mss of the appropriate composi-
tion and Fe304. The effects of small
amounts of Cu (up to 4% -5%) were
determined merely by incorporation
of this element in the Mss.
Experiments were carried out in
evacuated silica tubes. Reaction of
the charges with the tubes as de-
o
o
'050
1040
030
1020
ai
Jj 1010
o
L.
01
Ql
E
£ 1000
990
980
970
960
36
<y
mt +po+ liquid /rf
'/ / ''
20% Ni— *■// B J / mt + po
. (< H
/io%Ni /a
/ f* — Fe-S—O system
a
D
mt + po + liquid mt+po Ni,%
Cu,%
B
a
20
0
E
n
LJ
20
2
*r
4-
15
0
©
o
10
0
e
10
2
37 38
Weight per cent sulfur in pyrrhotite
39
Fig. 45. Melting experiments conducted with nickeliferous pyrrhotite ore compositions con-
taining 15—20 wt % magnetite. Minimum melting curves are shown for pyrrhotite-nickel contents
of 20 wt % (long-and-short dashes) and 10 wt % (long dashes). The effect of 2 wt % Cu
(dashed squares and circles) is shown by the short dashes. The minimum melting curve for the
Fe-S-O system is denoted by the solid line.
GEOPHYSICAL LABORATORY
419
scribed in the section on melting rela-
tions in the Fe-S-0 system (presented
elsewhere in this report) was slightly
retarded by the presence of Ni. All
experiments were held at the tem-
perature of reaction for one half hour
and then rapidly chilled by immediate
immersion in cold water.
Experiments have been conducted
using Mss compositions of 10, 15, and
20 wt % Ni. Experimental charges
were prepared by mixing 80 or 85 wt
% Mss with 20 or 15 wt % synthetic
magnetite. Charges in which no melt-
ing occurred contained only rounded
grains of unreacted pyrrhotite and
magnetite, whereas melting was
evidenced by the presence, in polished
section, of small amounts of very fine
(<2 fi) "fingerprint-like" or "hiero-
glyphic-like" textures, or both, of Mss
and Fe304, interstitial among larger
(25-50 fx) , rounded, unmelted grains
of the same phases.
The composition of the resultant
Mss in each experiment was deter-
mined by X-ray measurement of the
(102) reflection and application of
the spacing curves presented by Nal-
drett, Craig, and Kullerud (1967).
Naldrett (elsewhere in this report)
has discussed the applicability of such
spacing curves to composition deter-
mination. Although this discussion
concerns only pyrrhotite containing
Fe and S, there is no evidence that
the presence of Ni affects the applica-
tion of such curves.
Loss of Ni from the Mss and sub-
sequent incorporation in the magnetite
were checked by electron-probe anal-
ysis. Analysis of magnetite grains
from an experiment with 80 wt %
Mss (containing 20 wt % Ni), in
which considerable melting had
occurred, indicated that the loss was
equivalent to 0.07 wt % Ni from the
Mss.
The results of the present experi-
ments are shown in Fig. 45. For ref-
erence, the solidus curve from the
pure Fe-S-0 system (given by Nal-
drett elsewhere in this report) is
shown by the solid line. Although the
minimum melting curves shift to
lower S contents with 10% and 20%
Ni, it is apparent that 10% to 20%
Ni does not appreciably lower melt-
ing temperatures below those in the
Fe-S-0 system. It is also apparent
that the presence of 2% copper
lowers the melting temperature of
nickeliferous pyrrhotite-magnetite
mixtures 15°-20°C. The melting tem-
perature at 10% and 20% nickel
varies over a range of 50° to 60 °C
as sulfur content changes from about
36% to about 39% in the Mss.
The data derived in this study thus
indicate that the minimum tempera-
tures at which nickeliferous pyrrho-
tite-magnetite ores may be emplaced
as melts or crystal mushes varies be-
tween about 980° and 1040 °C, depend-
ing upon the bulk sulfur content.
Melting Relations over a Portion
of the Fe-S-0 System and Their
Bearing on the Temperature of
Crystallization of Natural
sulfide-oxide liquids
A. J. Naldrett
Nickel or nickel-copper sulfide de-
posits of the type found at Sudbury,
Canada, the Pechenga district,
U.S.S.R., and the Manitoba nickel
belt, Canada, are thought to have been
emplaced as ore magmas. In addition
to such economically important de-
posits, small amounts of iron-copper,
iron-nickel, or iron sulfides occur in
many of the classic, large, differen-
tiated intrusions. An origin proposed
for these sulfides is that they segre-
gated from associated bodies of mafic
or ultramafic silicate magma as im-
miscible sulfide liquids during intru-
sion or crystallization of the magmas.
In a discussion of the feasibility of
this proposed origin it is important to
consider the melting temperature of
the sulfide liquids.
420 CARNEGIE INSTITUTION
Iron sulfide (largely pyrrhotite) is and temperature of the ternary eu-
by far the dominant sulfide in all the tectic within the iron-rich portion of
occurrences mentioned above (e.g., the Fe-S-0 system. Neither of these
pyrrhotite accounts for more than studies is directly applicable to ore
1 of the total volume of sulfides in magmas, since the compositions cov-
most of the random samples of Sud- ered are all richer in iron than any
bury ore cited by Hawley, 19(5:2, p. natural magmas. In the present re-
117). The sulfides of many Sudbury port data on liquidus and solidus re-
deposits contain no more than 4*4% lations have been extended to compo-
nickel and 1%% copper. Study of sitions falling on the join between
the Cu-Fe-S and Fe-Ni-S systems magnetite and a pyrrhotite solid solu-
(Kullerud, Yund. and Moh, 1967) and tion containing 60 wt % iron, and
the Cu-Fe-Ni-S system (Craig and therefore cover the compositions of
Kullerud, 1967) has shown that many typical ore magmas.
copper and nickel in these proportions ^ . , 7 m , .
lower solidus temperatures reported Experimental Techniques
for the pure Fe-S system only Starting materials used in the
slightly. Oxygen has a much greater study consisted of iron, pyrrhotite
effect on melting temperatures in the solid solutions of different metal/
Fe-S system than nickel and copper, sulfur ratios, wustite solid solutions
Giani (reported in Oberhoffer, 1925, of different metal/oxygen ratios, and
pp. 98-99) first demonstrated this in magnetite (Fe304 composition). All
his work on melting relations between experiments were conducted in evac-
troilite and wustite. Magnetite is a uated sealed silica tubes. In the
characteristic mineral of iron-copper- experiments to determine the tem-
nickel ores commonly occurring in perature of the ternary eutectic, iron-
amounts up to 20 volume %. The con- wustite-troilite-liquid, the charges
sequent importance of the melting were loaded into small iron crucibles,
relations in that portion of the Fe- which were themselves sealed in silica
S-0 system covering wustite, magne- tubes. The precise composition of the
tite, and pyrrhotite made them the eutectic was determined by loading
subject of this study. the charges into small silver tubes and
High-temperature data on the Fe- sealing these in silica tubes. The use
0 system are summarized by Darken of silver containers is possible be-
and Gurry (1953) and on the Fe-S cause the values of /g2 and /o2 at the
system by Hansen and Anderko ternary eutectic are much lower than
(1958) and Kullerud {Year Book those required to stabilize Ag2S and
60). Vogel and Fulling (1948) and Ag20 at this temperature. The tem-
Hilty and Crafts (1952) have pub- perature of the eutectic as determined
lished liquidus diagrams of portions in the silver tubes lay between 904.5°
of the Fe-S-0 system. For reasons of ± 1°C and 905.5° ± 1°C, as corn-
experimental technique both of these pared with a bracket of 914° ± 1°C
studies were restricted to the iron- and 916.5° ± 1°C obtained by means
rich portion of the Fe-S-0 system, of the iron crucibles. The lowering
and few data were collected covering in temperature of about 10 °C induced
compositions poorer in iron than by the silver tubes is due to a very
those falling on the FeO-FeS join, small amount of silver (<1% was
Both groups of investigators outlined observed in polished section) dissolv-
the field of liquid immiscibility ing in the iron sulfide-oxide liquid,
(shown by a dashed line in Fig. Experiments on the Ag-Fe-S system
46 A) and determined the position (L. A. Taylor, personal communica-
GEOPHYSICAL LABORATORY
421
■o u
d> m-
-c o
to
G *-
^ g
>- P
-Q 5
> g
2 c
"^ LU
5 sr
o — '
"o .-
1_
£ g.
1 5
<J -C
~a -r
c £
D
e
>> §
X -Q
l_ ~
0 D
C C
o o
O C
■— •—
° o
to —
o a
c
•- to
o
TO to
<D D
_C -C
10 Q.
o
■D =
>.<
.£> .
0
OOll-
c -£
oo
£^
in
IO
ata sho
i this st
>
n
n
09
ations. D
are from
—
O
T. ">
jr
o aj
p"
-^
Q.
to
**
c
o
"5 o
O
O" to
-a
CD
BO
«-
— >.
>»
en
c
0)
U> -Q
J2
>
O
u
o.
.S c
o o
C
o
CO CO
to
'3
5
£ £
(1)
to T3
co "O
-a
^3
r^
O
CD
O o
—
\"5
to
to
c
• n
u_ -*-
o
l0-
\ >
1
i
O c
"to
o
a
£
1
g
Portion
nes are i
o
U
<
F
1
—
"*
#g>
1
a.
s^
=§1
o
oo x
CM
£
si
G
U- O
a
422
CARNEGIE INSTITUTION
tion) have shown that the ternary
monotectic silver - iron - troilite - liquid
has a temperature 24 °C below the
binary eutectic Fe-FeS but, within
the limits of measurement, has the
same Fe S ratio as the binary eu-
tectic. For this reason the presence
of a small amount of silver in the
Fe-S-0 eutectic liquid probably has
no measurable effect on the propor-
tions of iron, sulfur, and oxygen in
this liquid.
All other experiments were per-
formed with the charges in contact
with the walls of the silica tubes.
Because crystallization of the silica
glass during the experiment may
eventually expose the charge to the
atmosphere, experiments with charges
in contact with the silica tube walls
were of about 20 minutes' duration.
A second problem with silica tubes
is that fayalite is a stable reaction
product between the charge and silica
over much of the composition range
studied. Extremely small amounts of
a fine olive-green material, presum-
ably fayalite, were observed to form
at the contact between the charge and
tube in experiments in which the
pyrrhotite contained more than 62
wt % iron. The zones of recrystalliza-
tion in the glass tubes and the charges
themselves were repeatedly checked
for fayalite by X-ray powder diffrac-
tion methods, but none was detected.
The short duration of the runs is
probably the main reason that so
little fayalite forms.
Since it is likely that any sulfide-
oxide liquids that form in contact
with the walls of the silica tubes ap-
proach saturation with silica, the
temperature of the iron-wustite-
troilite eutectic was also determined
in silica tubes with no intervening
iron crucible. The temperature fell
between 915° ± 1°C and 916° ± 1°C;
this value, within the accuracy of the
determination, is no different from
the value determined with iron cruci-
bles. It is assumed, therefore, that
contact between charge and tube
walls has not measurably affected
temperature determinations in re-
gions of the system where such deter-
minations could not be checked in
other ways.
The composition of pyrrhotite at
the conclusion of many of the experi-
ments was checked by an X-ray spac-
ing method. The usual method of
2.670-
0<
2.620
63 62
Weight per cent Fe
Fig. 47. Plot showing variation of dioi values of pyrrhotite with composition.
GEOPHYSICAL LABORATORY
423
determining pyrrhotite composition
from its d102 value was not applicable
in those runs containing magnetite
and iron-rich pyrrhotite, since the
magnetite (004) reflection interferes
with the pyrrhotite (102) reflection.
Accordingly, a curve relating d101
values to pyrrhotite composition was
prepared (Fig. 47). Values of d101
were determined by standard diffrac-
tometer techniques using the (110)
peak of Lake Toxaway quartz as an
internal standard. The slope of the
curve in Fig. 47 increases with de-
creasing iron content of the pyrrho-
tite. In determining pyrrhotite com-
position in this way, it is assumed
that the pyrrhotite cell is not affected
by the presence of oxygen. Data re-
lating to this assumption are dis-
cussed below.
Experimental Results
The 900 °C isotherm of a portion of
the Fe-S-0 system is shown in Fig.
48. Pyrrhotite in equilibrium with
wustite and magnetite at this temper-
ature contains 62.8 ± 0.2 wt % iron.
The limits of the wustite solid solu-
tion at 900 °C are after Darken and
Gurry (1945). The rates of chilling
attained in the present experiments
were insufficient to quench wustite in
equilibrium with magnetite at 900 °C.
In polished section the wustite is
always seen to contain irregular,
50 %o
900°C
£> Hematite
C?
70
Weight per cent Iron
Pyrrhotite ss
50 %S
50%Fe
Fig. 48. Isothermal section through part of the Fe-S-O system at 900°C. All phases are i
equilibrium with vapor.
in
424
CARNEGIE INSTITUTION
lighter gray patches of exsolution
:netite: this observation agrees
with that of Foster and Welch
(1956).
The range in composition of mag-
netite is at present a matter of de-
Darken and Gurry (1946),
ii g their diagram partly on the
work of Greig, Posnjak, Merwin, and
S sman (1935), indicated that at
900°C it is less than 0.1 wt r;- oxygen.
On the other hand. Kullerud and
Donnay (Year Book 65) have shown
that in the presence of sulfur mag-
netite can range in composition from
Fe 04 to Fe^O . These cation-deficient
magnetites were produced by react-
ing Fe304 with elemental sulfur at
• with the resultant formation
of pyrite. The cation deficiency re-
sults in no superstructure reflections
and no variation in unit-cell size but
it can be detected by variations in the
relative intensity of certain reflec-
tions.
Experiments were performed to
determine whether the composition of
magnetite is changed when it equili-
brates with pyrrhotite, but no change
was detected. It was found, however,
that sulfur-rich pyrrhotites (<6.2
wt rr Fe) lose sulfur to the vapor in
the reaction tubes at temperatures
over 1000°C to alter their composition
as much as 0.5 wt ' .
The activity of sulfur in the experi-
ments of this study is much lower
than in those in which elemental sul-
fur is involved; this may be the rea-
son why cation-deficient magnetites
of the type studied by Kullerud and
Donnay were not produced in this
study. Because of the loss of sulfur
to the vapor in the reaction vessels,
compositions of experiments involv-
ing pyrrhotites containing less than
62 % iron are plotted on the basis of
the final composition of the pyrrho-
tite as determined from the (101)
spacing and not the initial composi-
tion of the charge.
Liquidus relations in the Fe-S-0
system (in the presence of vapor)
are shown in Fig. 46. The solid lines
in Fig. 46(A) are drawn on the
basis of data from this study, the
dashed lines on the basis of the data
of Iiilty and Crafts (1952), and the
dashed-dotted lines are interpretive.
Figure 46(B) is an enlargement of a
portion of Fig. 46(A).
In this as in previous studies, the
distinctive graphic intergrowths that
develop between the oxides and pyr-
rhotite were taken as the criterion
that partial or complete melting of
any sample had occurred. Where iron
has also dissolved in the liquid, on
very rapid chilling it appears in pol-
ished section as irregular stringers
that cut across the sulfide-oxide
graphic intergrowth or, where chill-
ing has been less rapid, as small
irregular masses within and around
areas of sulfide-oxide intergrowth.
Fig. 46 is characterized by fields of
iron, wiistite, pyrrhotite, and magne-
tite. The iron, wiistite, and pyrrhotite
fields meet at a ternary eutectic where
these three phases 4- liquid and vapor
are all stable. The wiistite, pyrrho-
tite, and magnetite fields meet at a
ternary reaction point at which mag-
netite + liquid react in the presence
of vapor to form pyrrhotite (contain-
ing 62.8 ± 0.2 wt % Fe) + wiistite.
Tie lines in Fig. 46(B) illustrate the
composition of pyrrhotite in equilib-
rium with iron oxide and liquid at
different temperatures along the
wiistite-pyrrhotite and magnetite-
pyrrhotite cotectic lines.
The relationship between the soli-
dus and the liquidus is illustrated in
Fig. 49. To the right of the vertical
dashed line the section is drawn along
the pyrrhotite-wiistite and pyrrho-
tite-magnetite cotectic lines. Conse-
quently, regions in which one solid
phase coexists with a liquid and
vapor do not appear on this section.
Most of the data points shown do not
GEOPHYSICAL LABORATORY
425
1 100
1050
P
O)
Q.
E
1=
1000
950
900
4- <X>
c .rr
5^
r
\
-\
\
\
\
I \
\
u \
M Liquid (+ ! solid phase) + vapor
is Liquid + 2 solid phcses + vapor
E3 3 solid phases + vapor
ED 2 solid phases + vapor
-6 .0
62 0 -
•62 5
WUSt;3+p0;
- +liq+
-Fe+wustss
+ po(troiiite)
W'JStjg
+ P°ss
mi + wust3S
+ poss (62JB%Fe)
62
69
68
67 66
Weight per cent
64
63
Fe
Fig. 49. Vertical section through part of the Fe-S-O system illustrating solidus and Iiquidus
relations. The portion shown by solid lines is drawn along the magnetite-pyrrhotite and wustite-
pyrrhotite cotectic lines. The portion shown by dashed lines is drawn along the extension of the
wustite-pyrrhotite cotectic line into the field of iron.
represent experiments with bulk com-
positions in the plane of the section
but have been projected onto this
plane.
The ternary eutectic, iron-wiistite-
troilite, occurs at 915 °C and has the
composition 68.2 wt % Fe, 24.3 wt
% S, 7.5 wt % 0; this composition
is slightly poorer in oxygen than that
determined by Hilty and Crafts
(8.9% 0) and slightly richer in oxy-
gen than that determined by Vogel
and Fulling (6.7% 0) but, like the
latter, falls on the join between
troilite and stoichiometric FeO. The
ternary eutectic lies within 0.5 wt
% iron of the join between iron-
saturated wtistite and troilite but not
on it.
The ternary reaction point occurs
at 934°C. As" shown in Fig. 49, the
solidus rises very steeply away from
the reaction point in the direction of
decreasing iron content (a mixture
of magnetite and pyrrhotite contain-
ing 62.5 wt % Fe would start to melt
at 1010 °C) and then flattens out as
the pyrrhotites in equilibrium with
426
CARNEGIE INSTITUTION
magnetite approach an iron content
of62wt c'c
Geological Applications
The results of this study set limits
on the solidus temperatures for many
iron-rich sulfide-oxide ore magmas
and demonstrate that these tempera-
tures rise with increasing sulfur-to-
iron ratio in the magma. Ores from
the Alexo Mine, Ontario (composition
of typical sulfides = 6.5 °/c Ni, 0.5%
Cu, 55rr Fe, 38% S), and the Strath-
cona Mine. Sudbury, Ontario (compo-
sition of typical sulfides = 4.5% Ni,
1.5% Cu, 55% Fe, 39% S), serve as
illustrative examples, provided that
the effects of nickel and copper are
ignored. Above 600 °C the nickel and
copper in both these deposits would
have been in solid solution in pyrrho-
tite. The metal/sulfur weight ratio of
the nickel- and copper-bearing Alexo
pyrrhotite would have been 1.63 and
that of the Strathcona nickel- and
copper-bearing pyrrhotite, 1.56. The
Alexo ore contains 4% magnetite,
whereas the Strathcona ore contains
147^ magnetite. These compositions
are plotted on Fig. 46(B), with cop-
per and nickel calculated as iron. The
estimate for the solidus temperature
of the Alexo ore obtained from this
figure is 1030 °C, whereas that for the
Strathcona ore is 1045°C.
It is shown elsewhere in this report
that the presence of small amounts of
nickel (<5%) in an iron sulfide-
oxide liquid has very little effect on
solidus temperatures and that copper
in amounts up to 2 wt % lowers the
solidus temperatures less than 20°C.
It is also shown elsewhere that the
presence of water has no measurable
effect on the solidus. Therefore the
temperatures outlined above for the
Alexo and Strathcona ore, subject to
a modification of up to 20°C on ac-
count of copper, are probably reliable
estimates of the temperature of the
beginning of melting of these and
many other sulfide ore magmas. This
conclusion depends on the reasonable
assumption that no other element was
originally present in the ore in large
enough amounts to have a significant
fluxing effect and was subsequently
removed.
Although the requisite liquidus
data have not been determined, it is
clear from Fig. 46(B) that because
of its higher magnetite content the
liquidus temperature of the Strath-
cona ore will be somewhat lower than
that for the Alexo ore.
The problem of the variable mag-
netite content of sulfide ore deposits
requires comment. Natural silicate
magmas tend to have compositions
close to eutectic points or cotectic
lines in the systems appropriate to
their compositions. This is primarily
because they are the products of proc-
esses such as fractional melting or
fractional crystallization. The same
observation is certainly not true of
magmatic ore deposits. As illustrated
by the Strathcona and Alexo ore
bodies, the proportion of magnetite
in such ores is highly variable and in
most cases is much lower than the
proportion of about 30 wt % required
for the composition to fall on the
magnetite-pyrrhotite cotectic line.
This is true because the composition
of an iron sulfide-oxide liquid at any
given temperature and pressure is
directly a function of its fSo and /o2.
The /s2 and f0o of small sulfide drop-
lets sparsely disseminated in a large
body of iron-bearing silicate magma
will be controlled by the magma. Any
tendency for the droplets to have a
higher /s2 or /o2 than the surrounding
magma will be countered by reaction
between the droplets and iron in solu-
tion in the magma and the addition
of more sulfide or oxide to the drop-
let. Nagamori and Kameda (1965)
presented a contour diagram illus-
trating the variation with composi-
tion of the fugacities of oxygen and
GEOPHYSICAL LABORATORY 427
sulfur at 1200 °C of liquids in the Fe- of the ore magma should occur, the
S-0 system that approximate the com- residual liquid would pass beyond the
positions of sulfide-oxide ore magmas, ternary reaction point to crystallize
Oxygen fugacities of basaltic and as a pyrrhotite-wiistite assemblage or
andesitic magmas (maximum values even reach the ternary eutectic to
of /o2 = 10"8'5 to 10~6-4 according to crystallize as a mixture of iron, wus-
the data of Fudali, 1965) plot well tite, and troilite. The wustite would
within the pyrrhotite field of the break down to a mixture of iron and
liquidus diagram if this is projected magnetite as the ore cooled below
onto Nagamori and Kameda's 1200 °C 560 °C (the lower stability limit of
isotherm. At 1200°C, oxygen fugaci- wustite) so that the final assemblage
ties of Fe-S-0 liquids close to the would consist of iron, magnetite, and
projected position of the magnetite- troilite. Native iron is never observed
pyrrhotite cotectic line are of the in large sulfide deposits of the type
order of 10~4-7, much higher than the under discussion, indicating that
fugacities of natural silicate magmas, strong isochemical fractionation of
Consequently, one would expect drop- an ore magma does not occur. One
lets of sulfide-oxide liquids segregat- obvious reason for this is that the
ing from silicate magmas of this type oxygen fugacities of liquids at or near
to contain appreciably less magnetite the ternary reaction point or eutectic
than liquids falling on the cotectic are so low that strong chemical gra-
line. dients would be established between
Once the sulfide-oxide ore magma the surrounding rocks and any
has segregated from the silicate sulfide-oxide liquid approaching these
magma, it may behave more like a compositions. These gradients are
closed than an open system. Assuming greater than can be tolerated in a
that the ore magma contains less than natural environment so that what
30% magnetite, pyrrhotite will start happens is that an interchange of
to crystallize once the temperature hydrogen and/or oxygen with the
falls to that of the liquidus and the surrounding rocks maintains the oxy-
residual liquid will become enriched gen fugacity of the sulfide magma at
in magnetite. The liquid will reach a higher value than that of liquids at
the pyrrhotite-magnetite cotectic line, the reaction point in the Fe-S-0 sys-
magnetite and pyrrhotite will crystal- tern. The ore magma therefore crys-
lize together, and the liquid will move tallizes as pyrrhotite and magnetite
along the line, becoming enriched in rather than as pyrrhotite and wiis-
iron. Assuming that no fractionation tite, or iron, wustite, and troilite.
occurs, the early-forming pyrrhotite _ ^
will itself be enriched in iron at this Faction between Pyrrhotite and
stage through reaction with the Enstatite-Ferrosilite Solid
liquid, and the ore magma will crys- solutions
tallize as a uniform mixture of pyr- A- J- Naldrett and G. M. Brown
rhotite and magnetite. It is possible Considerable debate has arisen
that during crystallization some of over whether certain ore deposits
the residual liquid will become sepa- were introduced as sulfide magma
rated from the early-forming pyrrho- suspended as droplets in liquid or
tite and crystallize as magnetite-rich partially crystalline silicate magma,
stringers or bands. Bands of this or whether they were introduced
type are common in the ore deposits after complete crystallization of the
of the Sudbury District. host igneous rocks. If limits are set
If strong isochemical fractionation experimentally on the compositions
42S
CARNEGIE INSTITUTION
of sulfides that can exist in equilib-
rium with pyroxenes of different
compositions, comparison between
those limits and the composition of
sulfides and iron-bearing silicates in
ore deposits may indicate whether
( 1 ) the natural sulfides ever achieved
widespread equilibrium with their
host rocks (i.e.. were introduced with
the mag-ma) or (2) they were intro-
duced after the host rocks crystal-
lized and thus achieved only local
equilibrium.
The slow reaction rates of silicates
means that potentially they are much
better "sliding-scale" geothermome-
ters than sulfides. Although sulfide-
silicate equilibria are unlikely to be
frozen at magmatic temperatures
(~1000°C), the reequilibration may
stop at a much higher temperature
than it does with sulfides alone and
so give useful minimum temperatures
for ore deposition.
The object of the experiments de-
scribed here was to study equilibria
between pyrrhotite and enstatite-
f errosilite solid solutions at high tem-
peratures in the presence of vapor,
excess quartz, and magnetite. Inter-
mediate members of the enstatite-
ferrosilite solid solution series were
kindly supplied by Dr. D. H. Lindsley,
who had synthesized them from oxide
mixes at 20 kb in iron capsules in a
solid-media piston-and-cylinder press.
In a few cases, where necessary,
these pyroxenes were inverted in a
solid-media press from the ortho-
rhombic to the monoclinic form. X-
ray diffractometer methods were not
very sensitive for diagnosis of com-
tional variation and were unsat-
isfactory where both pyroxene phases
may have been present in partially
inverted material. Polysynthetically
ined Cpx grains were used to
establish a -/-refractive-index work-
ing curve for the Cl-En-Cl-Fs series,
and only twinned grains were used
for product identification.
Bulk composition of our experi-
ments fell in the five-component sys-
tem Mg-Fe-Si-O-S. The phases pyrox-
ene, pyrrhotite, quartz, magnetite,
and vapor are present in all experi-
ments. By the phase rule, the system
therefore has two degrees of freedom
but at a given temperature will be
defined if the composition of one of
the phases of variable composition is
specified. Pyroxene and pyrrhotite
are the only phases showing appre-
ciable solid solution, and thus the tie-
line directions between these two are
of greatest interest.
Reversed tie lines were determined
in a single experiment by the use of
two silica tubes, one inside the other,
as shown in Fig. 50. Enstatite, to-
gether with excess quartz and mag-
netite and a small quantity of pyrrho-
tite, was placed in compartment A,
and an iron-rich member of the py-
roxene solid solution series, together
with a small quantity of pyrrhotite,
in compartment B. Sufficient pyrrho-
tite for an X-ray powder diffraction
mount was placed as shown at C. The
outer tube was sealed under vacuum,
and the experiment was maintained
at the required temperature for 6
weeks. After quenching, the new
compositions of the pyroxenes were
obtained from their -/ refractive in-
dices, and that of the pyrrhotite, from
the d value of the (102) peak.
Silica glass rod
Silica wool
Fig. 50. Design of silica glass tube used for
reactions between pyrrhotite and enstatite-
ferrosilite solid solutions. Position A, enstatite
+ quartz + magnetite + pyrrhotite. Position
B, iron-rich pyroxene + pyrrhotite. Position C,
pyrrhotite.
GEOPHYSICAL LABORATORY 429
One tie line has been established so basis of mineralogy that the sulfide
far at 980° ± 20 °C. This tempera- ore at Marbridge crystallized between
ture was chosen because it lies below 530° and 400 °C; the solidus tempera-
the solidus temperatures for the bulk ture for a bulk composition approxi-
compositions of our experiments (see mating the Marbridge ore is about
discussion of melting relations in the 1040°C. (2) The predominance of
Fe-S-0 system presented elsewhere primary pyrite in the Cyprus ores
in this report) . Pyrrhotite containing requires the bulk of this ore to have
48.3 atomic % iron was equilibrated crystallized below 743 °C (Kullerud
with enstatite, magnetite, quartz, and and Yoder, 1959), whereas the lowest
an intermediate pyroxene (Fs80En2o, temperature at which a sulfide liquid
mole °/o ) . The final composition of the containing any more than very minor
pyroxene in both compartments was amounts of iron can exist in the Fe-S
Fs45 ± 2 mole % FeSi03, and that of system is 1083 °C (Kullerud, Year
the pyrrhotite was 47.2 atomic % Book 60); the presence of copper in
iron. the proportions found in the Cyprus
deposits would not lower melting
Effect of Water on the Melting temperatures more than 20° to 30°C
of Pyrrhotite-Magnetite below 1083°C. Thus, if the magmatic
Assemblages origin proposed for either of the two
examples is correct, some component
A. J. Naldrett and S. W. Richardson ^ additional to those now occurring in
Recently a number of ore deposits the ores must have drastically low-
have been interpreted as the result ered the melting temperatures. Re-
of the injection and crystallization of connaissance experiments have shown
sulfide or sulfide-oxide ore magmas, that water at 2 kb total pressure will
These include the nickel-copper ores not depress melting temperatures
of the Sudbury District, Ontario sufficiently.
(Hawley, 1962; Souch, Podolsky, et The fluxing properties of water on
al., 1967; Naldrett and Kullerud, silicate melts are well established, and
1967), the copper-bearing pyritic water is therefore obviously a possi-
ores of Cyprus (Hutchinson, 1965), ble flux for sulfide melts. On the as-
the nickel deposit at Marbridge, Que- sumption of complete solubility in the
bee (Clark, 1965), and the copper- melt, Smith (1963, pp. 304-307) has
zinc ores and pyrite-pyrrhotite de- calculated that 10 wt % water would
posits of the Flin Flon region of the lower the melting point of pyrrhotite
Canadian shield (Koffman, Cairns, (Fe0.92S) -water mixtures by 300°C;
and Price, 1962). The mineralogy the object of this study was to test
and field relations of some of these the extent of water solubility in the
deposits are in accord with the high magnetic-pyrrhotite liquids by deter-
melting temperatures (>1000°C) in- mining the solidus temperature at 2
dicated for the ores by experimental kb fluid pressure,
work on appropriate systems. To make a reproducible determina-
In other deposits, however, the tion of the solidus temperature of
mineralogy and field relations seem pyrrhotite-magnetite mixtures in the
to require the ore magmas to have four-component system Fe-S-O-H,
been introduced at temperatures well three variables must be specified — P,
below the solidus temperatures ob- T, and the composition of the pyrrho-
tained from experimental study of tite. To control these variables, mix-
similar compositions. To take two ex- tures of pyrrhotite (of known com-
amples: (1) Clark estimated on the position), magnetite, and water were
430
CARNEGIE INSTITUTION
sealed in small gold tubes and brought
to pressure and temperature in an
internally heated gas-pressure appa-
ratus (Yoder, 1950a). However, two
problems were encountered:
1. Iron from the charge diffused
into the gold container, with which
it forms an extensive solid solution
at 1000°C. This problem was in part
overcome by making experiments of
short duration (15-20 minutes). Al-
though experiments of this duration
would be inadequate for the study of
silicate-water melting relations, the
fast reactions found in the dry sys-
tem Fe-S-0 (Naldrett, this report)
indicate that equilibrium is achieved
rapidly in the sulfide-oxide-water
system, well within the time limits
of the experiments. Before the charge
was brought to maximum tempera-
ture, it was held at 900 °C (below
the solidus temperature of the sys-
tem) for 5 minutes. Experiments in
which this holding period was varied
indicated that most of the iron loss
occurred during the first 5 minutes.
2. Gold is readily permeable to
hydrogen at the temperatures of the
experiments. Loss of hydrogen to the
pressure medium causes oxidation of
the charge. We therefore reduced the
hydrogen loss by enclosing the charge
capsule in a larger, thick- walled gold
capsule containing the assemblage
quartz-fayalite-magnetite- water.
We believe that these procedures
maintained a nearly constant charge
composition (within 0.2 wt %) dur-
ing the latter (high-temperature)
part of each experiment. After
quenching, the charges were exam-
ined in polished section for melting
textures, and the final composition of
the pyrrhotites was obtained by de-
termining their d1(ll values.
Our experimental results are shown
in Fig. 51, superimposed on the
solidus determined for magnetite-
pyrrhotite mixtures under dry con-
ditions. It is seen that, within experi-
IIOO
U
1050 -
fl> 1000
950
65 64
Weight per cent Iron
63
Fig. 51. Plot showing data points for the
melting of pyrrhotite-magnetite assemblages in
the presence of water at 2 kb total fluid pres-
sure superimposed on the pyrrhotite-magnetite
solidus obtained in dry, silica-tube experiments.
mental limits, solidus temperatures
determined in the presence of water
are no different from those determined
in the dry state. The experiments with
water, as compared with those with-
out water, are likely to have been
affected by two other factors besides
water — the 2-kb confining pressure
and the presence of gold. Although
the effects of both factors are unde-
termined, it is to be expected that
confining pressure would raise the
solidus temperatures 3° — 7°C/kb
(Kullerud, personal communication).
By comparison with the effect of
silver on the dry experiments, we
predict that gold would decrease soli-
dus temperatures perhaps 5°-20°C.
To determine these two effects inde-
pendent of the effect of water, tubes
containing the pyrrhotite-magnetite
mixture but no water were run con-
currently with two of the wet experi-
ments. The concurrent wet experi-
ments are labelled A and B in Fig. 51.
In the case of A the dry experiment
showed considerable melting, and in
the case of B, slight melting. The
final compositions of the pyrrhotites
GEOPHYSICAL LABORATORY
431
in the dry experiments of both A and
B were close to those of the wet ones.
The fact that dry experiment B just
melted, whereas the wet experiment
did not, is probably not significant in
view of the precision of our data.
What is clear from our experiments
is that the 2-kb confining pressure
and the presence of gold do not cam-
ouflage any appreciable effect by
water. We conclude, therefore, that
water has little influence on melting
temperatures of pyrrhotite-magnetite
mixtures and that it cannot be ap-
pealed to as a flux to explain the oc-
currence of natural, low-temperature
(<900°C) sulfide-oxide magmas. Al-
though our experiments have been
with iron sulfide-oxide mixtures, our
conclusion can almost certainly be
applied to oxide-free iron-bearing
sulfide magmas, and it casts a certain
amount of doubt on the existence of
low-temperature ore magmas. If such
magmas do exist, some other element
or elements, not occurring in the ores
as we examine them at present, must
be responsible for the fluxing.
Succession of Mineral
Assemblages in Pyrrhotite-Rich
Ni-Cu Ores
J. R. Craig, A. J. Naldrett, and G. Kullerud
The primary sulfide mineral as-
semblage most frequently encoun-
tered in magmatically derived Ni-Cu
ores of the Sudbury type is that con-
taining pyrrhotite of various kinds,
pentlandite, chalcopyrite, and pyrite.
Although numerous other assem-
blages have been reported from these
ores, the only ones of significance
are chalcopyrite-cubanite-pyrrhotite-
pentlandite and bornite-cubanite-
pyrrhotite-pentlandite. Magnetite in
amounts ranging up to 20% is always
present in these assemblages.
Investigation of pertinent portions
of the Cu-Fe-Ni-S system (Craig,
Year Book 65; Craig and Kullerud,
Year Book 65 and Year Book 66;
Naldrett, Craig, and Kullerud, 1967,
and previous sections of this report)
demonstrates that the mineral assem-
blages observed in ores are the net
result of a complex series of reactions
that occur on cooling of the ore mass.
The number and nature of reactions
involved in the paragenesis of a given
ore body or mineral concentration de-
pend upon the bulk composition
involved.
In the light of the earlier work and
additional new data it is now pos-
sible to reconstruct the succession of
mineralogical assemblages through
which typical ores have passed since
their injection or segregation as a
sulfide melt. This succession is shown
in Fig. 52. Development of two or
more succeeding assemblages from a
given assemblage is shown by branch-
ing of the pattern. Stages in which
one assemblage may result from two
preceding assemblages result in con-
vergence of the pattern. Relative con-
centrations of Cu, S, and Ni in the
mineral successions are indicated at
the top and bottom of the diagram.
Owing to the presence of several in-
dependent compositional variables, it
is not possible to define uniquely the
elemental concentrations in any por-
tion of the diagram.
The temperature scale at the left
side of the diagram, though presented
as specific values, should be viewed
only as a general guide to the limiting
thermal conditions under which spe-
cific mineral assemblages may exist.
The significance and limitations of
the individual calibration points for
this scale are discussed below. Al-
though these points have been de-
rived from low-pressure experiments
(equilibrium vapor pressure of the
appropriate assemblages), it is prob-
able that confining pressures of sev-
eral kilobars would modify the values
only slightly.
432
CARNEGIE INSTITUTION
CO
L±J
o
o
Q.
ID
<
1
00
00
UJ
CO
CO
<
<
o
u
w
o
o
o
o
o
CO
LO
o
o
to
<fr
00
<x>
r-
rO
O
m
lO
ro
K)
o
o
CO
v
q0 9jn|Djadiu9j_
o
Q.
E
o
c
o
o>
o
3
cr
O
en
■o
13
O
=3
cr
O ~ ~
a>
■o
o
Q.
O
I
CD
T3
O
CL
+
D
y .2
to
• J= z
<D — o
<D V
D *h -
<D
O
O
a
D
a
3
O O V>
7, E A
o
3
I "3= D
w x
•4= D
O E
.E 3
°> I
.E 8?
5 D
u
O •-
0)
<D •-
D) £
J -5
E IE
o >
— -o
D C
o
c ««
•i S
o .
IS
X 0s*
29
a co
° V
o ' — '
* o
D CI
E >.
*- <1>
3 >
•° CO
in
O _i
D o^
3 q
£ K
0) CO
s I
o CO
a *~
to
'*=■ O
- .5* </>
O -D
C
i/> o
c u
o __
'to O
to c
0) £
b ©
3 ■£
O
O
a
■
to
0) c
.E w
II
JC
a «>
D> _
.5 o
13 ■£
O j2
zn e
'o
a
<n c
>o o
■u
<D
c — ^
"U IT)
_>. 00
<D CO
££
i-5
o "?
c to
c
■4—
c
o
CO
TJ
c
D
V
o
o
a
2 Z
6) 2
h- SI
5A
GEOPHYSICAL LABORATORY 433
An approximate lower limit for the as a phase at temperatures as high as
existence of a sulfide-oxide melt from 610 °C. The bulk compositions of most
which ores of the Sudbury type may ore masses, however, are such that
be derived is 1000 °C. The exact tern- the pentlandite present is the product
peratures of final crystallization vary of exsolution at much lower tempera-
from about 1040 °C for Cu-free pyr- ture (Naldrett, Craig, and Kullerud,
rhotite-rich assemblages with S con- 1967). The effect of pressure on the
tents of approximately 38.5 wt % to temperature of appearance of ex-
about 985 °C for Cu- and Ni-bearing solved pentlandite is unknown,
pyrrhotite assemblages with about Yund and Kullerud (1966) have
36.5% S. The lowest temperature at shown that the chalcopyrite solid so-
which a Cu-enriched liquid may frac- lution that is stable at high tempera-
tionate from a coexisting nickelifer- ture divides at approximately 590 °C
ous pyrrhotite is 850 °C. Below this into two discrete solid solutions, chal-
temperature the liquid has crystal- copyrite and cubanite. Phase rela-
lized to Cu-bearing minerals such as tions are such that in copper-bearing
chalcopyrite. This is not to imply that assemblages containing less than 4%-
all chalcopyrite concentrations in Ni- 5% Ni, cubanite-pyrite tie lines pre-
Cu ores separated as melts from vent the existence of a pyrrhotite-
pyrrhotite-rich crystal "mushes" chalcopyrite assemblage between 590°
above 850 °C; rather, the implication and 334° C. On the other hand, where
is that if chalcopyrite concentrations 5% or more Ni is present, chalcopy-
did separate from pyrrhotite crystal rite and nickeliferous pyrrhotite can
mushes, the process must have taken coexist and no cubanite-pyrite assem-
place above 850 °C. blage is formed.
The maximum temperature at The maximum temperature at
which pyrite may exist is 743 °C. The which chalcopyrite or cubanite and
appearance of pyrite in nickeliferous pentlandite can coexist is approxi-
pyrrhotite ores, however, is limited to mately 575 °C (Craig and Kullerud,
temperatures below 675 °C, the upper Year Book 65). Below 334 °C any cu-
stability limit of pyrite and magne- banite-pyrite mineral pair formed in
tite. Although it is possible that py- Ni-poor assemblages reacts to form
rite may have formed in some ores as chalcopyrite and pyrrhotite (Yund
high as 675°C (e.g., the Strathcona and Kullerud, 1966) .
deposit, Naldrett and Kullerud, Monoclinic pyrrhotite first appears
1967) , it is probable that most of the as a phase in the pure Fe-S system at
pyrite in magmatic ores has exsolved 308°C (Clark, 1966). Assemblages in-
f rom the pyrrhotite at much lower volving this phase do not have appre-
temperatures (Naldrett, Craig, and ciable extension into the quaternary
Kullerud, 1967) . Cu-Fe-Ni-S system above about
Pentlandite, the principal Ni-bear- 300 °C. Accordingly, monoclinic pyr-
ing mineral, has a maximum stability rhotite, the most abundant mineral
of 610 °C (Kullerud, 1963) and thus phase in most massive Ni-Cu ores,
cannot form in an ore assemblage appears only below about 300° C. For-
above this temperature. Bell, Eng- mation of this phase results either
land, and Kullerud {Year Booh 63) through exsolution from hexagonal
have shown that increased confining pyrrhotite or through reaction of py-
pressure lowers the maximum ther- rite and hexagonal pyrrhotite.
mal stability of pentlandite 5°C per The final mineral assemblage that
1000 bars. In some S-poor, Ni-rich is frequently observed, but as yet un-
ores pentlandite may have appeared explained, is that of pyrite + pent-
434
CARNEGIE INSTITUTION
landite. Relations involving the for-
mation of this mineral pair are not
entirely clear; it appears, however,
that tlie pair is stable only below
200°C.
In conclusion, the mineral assem-
blages observed in Ni-Cu ores result
from a series of reactions occurring
over a wide temperature range and
represent at least partial reequilibra-
tion to temperatures below 200 °C. It
is further apparent that the concen-
tration of sulfur has a more pro-
nounced effect on the succession of
assemblages and the ultimate mineral
associations than have the concentra-
tions of nickel and copper.
The Fe-Ni-S System
Violarite Stability Relations
J. R. Craig
Violarite, the iron-nickel member
(reportedly FeNi2S4) of the linnaeite
(spinel) series of minerals, is fre-
quently present in small amounts in
iron-nickel sulfide ores. Although
most often formed as a secondary
mineral through deuteric alteration,
violarite also occurs as a primary and
principal ore mineral in several de-
posits (e.g., Falconbridge Nickel
Company, Marbridge No. 2 Mine,
Malartic, Quebec).
Synthesis of violarite in the pure Fe-
Ni-S system has heretofore been re-
ported only by Lundqvist (1947),
who reacted Fe, Ni, and S at 200°C.
The Ni end member of the linnaeite
series (polydymite, NioS4) was syn-
thesized by Kullerud and Yund
(1962) , who determined its maximum
stability as 356° ± 3°C and found
that increased confining pressure
raises its stability limit by 6°C per
1000 bars.
Violarite was first encountered in
the present study as an exsolution
product from the Fe^S-Ni^S Mss
phase at 300° and 400 °C. Subsequent
experimentation has demonstrated
that violarite is stable to 461° ± 3°C
in the presence of an equilibrium
vapor. Synthesis of violarite was at-
tempted by (1) direct reaction of
the elements at 300° and 400 °C and
(2) reaction of sulfur and the appro-
priate Mss composition at 300° and
400 °C. Reaction of the elements at
300° and 400 °C, even with repeated
grinding of the products, yielded only
10% -20% violarite after 5 months.
Other phases present were pyrite,
vaesite, and Mss. Reaction of homo-
geneous Mss (formed by previous re-
action of elements at 600 °C and S)
at 300 °C yielded pure violarite in
10 days. The exact composition of
the Mss used is not important so
long as the Mss is homogeneous and
contains Fe to Ni in the atomic ratio
of 1 to 2. Attempts to repeat this re-
action at 400 °C yielded 80% -90%
violarite and 10% -20% pyrite and
Mss. Since the violarite phase pro-
duced at 400 °C contains less iron than
is indicated by the formula FeNi2S4,
some of the pyrite formed is stable.
An additional portion of the pyrite
appears to have formed metastably
through rapid initial reaction of the
Mss and S. Such reaction leaves in-
sufficient sulfur to permit complete
reaction of the Mss.
Synthesis of a homogeneous solid
solution between violarite and polydy-
mite has been accomplished at
300 °C by sulfurization of various Mss
compositions. Attempts to synthesize
only homogeneous violarite - polydy-
mite solid solutions containing less
than about 5 wt % Fe (FeNi2S4 con-
tains 18.52 wt % Fe) , however, were
not successful. Although the principal
reaction product in these experiments
was a homogeneous violarite-polydy-
mite solid solution, initial reaction of
the Mss and S formed several per cent
vaesite. Formation of this phase,
which persists metastably, leaves in-
sufficient sulfur and nickel to com-
plete formation of the violarite-po-
435
Weight per cent Ni
10 20 30
40,
(Fe,N
500 °C
70
60
(Fe,Ni)S2
Co ^
50 40 30
Weight per cent Fe
Weight per cent Ni
10 20 30
20
10
40
60
450°C
(Ni,Fe)S2
50
60
(Fe,Ni)S2
50
40 30
Weight per cent Fe
Weight per cent Ni
10 20 30
20
40_
400°C
(Ni,Fe)S2
50
60
50
40 30
Weight per cent Fe
Weight per cent Ni
10 20
20
10
30
40,
300°C
(Ni,Fe)S2
60
50
40 30
Weight per cent Fe
20
10
Fig. 53. Appearance of violarite phase and subsequent development of violarite-polydymite
solid solution. Violarite becomes stable at 461° ± 3°C in the region bounded by (Fe,Ni)S2,
(Ni,Fe)S2/ and Mss; solid solution to the Ni-S join becomes complete at 356 °C.
436
CARNEGIE INSTITUTION
lydymite solid solution of the desired
composition; residual unreacted Mss
remains in the experimental charge.
Examination of polished sections of
violarite - polydymite solid solution
compositions prepared at 300 °C indi-
cates complete solid solution between
the two end members. The unit-cell
dimension of the solid solution varies
from 9.463 ± 0.003 A for FeNi2S4 to
9.480 ± 0.003 a for Ni3S4. In Fig. 53
the appearance of violarite in the Fe-
Ni-S system and subsequent develop-
ment of the complete solid solution
with NI3S4 are shown for four iso-
thermal sections at 500°, 450°, 400°,
and 300°C. At 500°C, Fig. 53(A),
FeXi2S4 composition lies within the
(Ni,Fe)S2 ss + Mss divariant field.
With decreasing temperature the Mss
corner of the (Fe,Ni)S2 + (Ni,Fe)S2
+ Mss univariant field shifts toward
more Ni-rich compositions, such that
at 461° ± 3°C, the temperature of
violarite appearance, FeNi2S4 compo-
sition lies within the univariant field.
At 461° ± 3°C the violarite phase
contains 17.0 ± 1.0 wt % Fe and still
lies within the region bounded by
(Fe,Ni)S2, (Ni,Fe)S2, and Mss. The
temperature of maximum stability
has been determined by breakdown
of homogeneous violarite above its
stability field and by its formation
from Mss 4- S in its field of stability.
Formation of violarite through reac-
tion of pyrite, vaesite, and Mss has
not been observed below 461 °C; for-
mation of violarite exsolution laths in
Mss has been observed, however, at
450 'C.
Phase relations of violarite in the
Fe-Ni-S system at 450 °C are shown
schematically in Fig. 53(B). With
decreasing temperature, solid solution
in the violarite phase increases to-
ward the Ni-S join, extending to a
composition of approximately 6.0%
Fe at 400°C, Fig. 53(C). Solid solu-
tion to NI3S4 composition becomes
complete with appearance of the lat-
ter phase on the Ni-S join at 356 °C.
Extension of the violarite phase to
include FeNi2S4 composition occurs
between 400° and 300°C. Stable phase
assemblages at 300 °C are shown in
Fig. 53(D).
Experimental investigation of
phase relations involving violarite is
not yet complete below 300 °C. With
the possible exception of a few "gray
violarites," described from the Mar-
bridge No. 2 Mine (Buchans and
Blows, 1967), which have uncertain
compositions, natural Fe-Ni-S spinel
phases closely approximate FeNi2S4
or Ni3S4. The absence of minerals
with intermediate compositions sug-
gests that the complete solid solution
observed between violarite and poly-
dymite at 300 °C may break down at
lower temperatures. Reported violar-
ite composition (FeNi2St) lies on the
pyrite-millerite join (FeS2-NiS). The
frequent occurrence of this latter as-
semblage represents either metasta-
bility or sufficient shift in violarite
composition at low temperature to
permit stable coexistence of pyrite
and millerite.
Partial Pressure of Sulfur in the
Vapor Coexisting with the
Fe^S-Nij-xS Solid Solution at 600°
and J>00° C
A. J. Naldrett and J. R. Craig
The partial pressure of sulfur is
one of the most important variables
to be considered in determining the
chemical conditions under which ore
deposits form. Naldrett, Craig, and
Kullerud (1967) have investigated in
detail the limits of the monosulfide
solid solution (Mss) spanning the
Fe-NiS system from Fe^S to Nii^S
between 600° and 300 °C and pointed
out that the bulk compositions of
many iron-nickel sulfide deposits fall
within the limits of this solid solution
at high temperature. A knowledge of
the partial pressure of sulfur in the
vapor in equilibrium with the Mss at
GEOPHYSICAL LABORATORY 437
different temperatures will permit the The electrum-tarnish method (Bar-
estimation of the partial pressure of ton and Toulmin, 1964) was used to
sulfur in ore deposits of this type at determine the partial pressure of sul-
the time of their formation and dur- fur over Mss in equilibrium with
ing their cooling" history. In an earlier pyrite and vaesite at 600°C. This
report Naldrett (Year Book 65) gave method is necessary because the sul-
some preliminary results obtained at fur pressure over this assemblage at
600 °C. The results presented here are 600 °C is higher than that over the
a continuation of this earlier study pyrite-pyrrhotite assemblage. Mix-
over the whole of the Mss field at tures of pyrite, vaesite, and Mss lost
600 °C and over much of it at 400 °C. sufficient sulfur when equilibrated
The experimental method used for with pyrrhotite to be reduced to Mss
most of our determinations was to alone. Most of the pyrrhotite in these
equilibrate an Mss sample with a sul- experiments was converted to pyrite.
fur-rich and a sulfur-poor pyrrhotite The results of this study are pre-
in a silica tube in a way that permits sented in Figs. 54 and 55 as two
sulfur but not nickel or iron to pass isotherms on which are drawn iso-
freely between all phases in the tube, pleths of log10 PS2 in bars. The accu-
The metal : sulfur ratios of the Mss racy of the isopleths has not been as-
and the pyrrhotites change until all sessed, although all the experimental
coexist with the same vapor ; the two points except two have values within
pyrrhotites have the same composi- 0.2 log unit of that indicated. The
tion at the conclusion of the experi- two exceptions lie within 0.3 log unit
ments. The partial pressure of sulfur of the appropriate isopleth.
in the vapor is then obtained from the The highest value of PS2 on the
composition of the resulting pyrrho- 600 °C isotherm, 1<H-8 bars, occurs at
tite with the use of the data of Toul- the composition at which pyrite and
min and Barton (1964). A more de- vaesite are in equilibrium with Mss.
tailed description of this method is The sulfur pressure decreases from
given in Naldrett's earlier report. this point to 10-20 for pyrrhotite in
In this study a problem was en- equilibrium with pyrite (data from
countered with sulfur-deficient, Toulmin and Barton, 1964) and to
nickel-rich Mss compositions, since 10-2.2 for Ni^S in equilibrium with
these are unquenchable and even on vaesite. On the sulfur-poor side of
the fastest cooling break down to a the Mss, sulfur pressure increases
less Ni-rich Mss and pentlandite, continuously with increasing nickel
Ni7S6 or Ni3S2. Consequently it was content from 10"12-5 for troilite (data
not possible to determine the composi- from Toulmin and Barton, 1964) to
tion of the Mss by the powder X-ray io-6-9 for high-temperature NiS.
diffraction method used for quench- The 400°C isotherm is only partially
able compositions. In the unquench- complete at present owing to the slow
able region, one determination was reaction rates encountered with sul-
obtained at 400 °C for the point at fur-rich Mss compositions. The pres-
which the Mss is in equilibrium with ence or absence of a maximum par-
pentlandite and Ni7S6 (see discussion tial pressure value for sulfur-satu-
of 400 °C isotherm of the Fe-Ni-S rated Mss with an intermediate
system presented elsewhere in this nickel : iron ratio has not been proved
report). Determinations were ob- although preliminary data indicate
tained at both 400° and 600 °C for a that the sulfur partial pressure
mixture of Mss and pentlandite with increases continuously with increas-
an iron : nickel ratio of 1 : 1. ing nickel content from Fe^S to
43S
CARNEGIE INSTITUTION
si-
te
&
u-> ic f-x ">
/0,-.
Li-
eu
Q.
.51
■%
lO
o
a
a
>
<D
D
-Q
<£
D)
O
^
3
«/>
H-
O
CO
k.
3
to
</»
a>
k.
a
CD
li.
"o
'*—
k_
C
CD
o
a
a
i_
c
cu
CL
c
o
SZ
CD
*^
D
*CU
^
D
>
O
JZ
-)—
O)
c
i
o
_c
a
o
oo
z
iu
-S
~- >o
i ,
o
*S a
rO
I/I
t: jc
o . -
A- £
O)
• c
V 'ZZ
V) .«*
. 'x
St
u- u
GEOPHYSICAL LABORATORY
439
v^
'X^ CO
S
Ate
^
10
-
<rcq
o
Q.
o
>
c
7i
o
«+-
VI
o
c
i/s
</>
0
Q.
O
a
*- a)
o
a
o
m
CM
oo
Z
0) g
■£: ^
.2 ^
t: -c
£■>
Ui
• c
IT) wi
'x
440
CARNEGIE INSTITUTION
NijJS. In contrast to the 600°C iso-
therm, our data indicate that at
400°C a maximum in the partial pres-
sure of sulfur occurs along* the sulfur-
deficient side of the Mss about mid-
way between troilite and high-tem-
perature NiS. This is to be expected
in view of the fact that with decreas-
ing temperature the pentlandite-Mss
field is becoming enlarged at the ex-
pense of the Mss and that at some
lower temperature (<250°C, Craig,
unpublished data) tie lines are estab-
lished between pyrite and pentlandite.
The geometry of the 400 °C isotherm
of the Fe-Ni-S system requires the
point at which the Mss-pentlandite-
Xi:S,, field abuts against the Mss to
be a local minimum in sulfur pres-
sure. Measurement of the sulfur pres-
sure over the univariant field Mss-
pentlandite-Ni7S6 showed this to have
a value of 10-10-8, in comparison with
10-101 and 10"10-7 at the nearest deter-
mined points on either side — a cor-
respondence between experiment and
theory that we attribute more to good
fortune than to good measurement.
400° C Isothermal Diagram
J. R. Craig, A. J. Naldrett, and G. Kullerud
Recent detailed investigations con-
ducted in the geologically significant
portion of the important Fe-Ni-S sys-
tem now permit construction of the
400 °C isothermal section of this sys-
tem shown in Fig. 56. This figure in-
corporates studies of the Fe^S-Ni^S
Mss (Naldrett, Craig, and Kullerud,
1967), the violarite-polydymite solid
solution (presented elsewhere in this
report) , and pentlandite-Ni7Se rela-
tions (discussed below) with the
basic features of the system as deter-
mined by Kullerud (Year Book 63).
In a detailed investigation of the
limits of homogeneity of the Mss
phase, Naldrett, Craig, and Kullerud
(1967) found that this phase spans
the Fe-Ni-S system to temperatures
even below 300°C. At 400°C the Mss
field is nearly straight on the sulfur-
rich side but is slightly concave to-
ward pentlandite on the sulfur-poor
side. The Mss has a width of about 3
wt °/o S on the Fe-S join, thinning to
about 2 wt % S on the Ni-S join.
The violarite-polydymite solid solu-
tion phase in the Fe-Ni-S system has
a maximum stability of 461° ± 3°C
(as detailed in another section of this
report). At 400°C the violarite-po-
lydymite solid solution has a composi-
tional range from 16.5 db 1.0 to 6.0
± 1.0 wt % Fe along the FeNi2S4-
Ni3S4 join. This phase coexists with
Mss, (Fe,Ni)S2, and (Ni,Fe)S2.
New experimental data have dem-
onstrated the stable coexistence of
Ni7S6 with pentlandite. This assem-
blage becomes stable at some temper-
ature above 400 °C through reaction
of the higher temperature assemblage
Ni3S2 + Mss.
Absence of Ni7S6 as a mineral spe-
cies has previously been ascribed to
the fact that this phase could not co-
exist with Fe-bearing phases such as
pentlandite. The nearly ubiquitous
occurrence of Fe-bearing minerals
thus was considered to preclude the
presence of Ni7S6 in ores. Demon-
strated stability of the Ni7S6 and
pentlandite pair over at least a cer-
tain temperature range now neces-
sitates reevaluation of the possible
existence of Ni7S6 as a mineral. Re-
ported natural assemblages of pent-
landite + millerite, pentlandite +
heazlewoodite, and heazlewoodite 4-
millerite suggest that either (1)
Ni7SG-pentlandite tie lines break down
at temperatures below 300 °C to yield
a heazlewoodite-millerite pair (with
one or both phases containing some
Fe in solid solution) or (2) Ni7S6
breaks down at some temperature be-
low 200°C to yield Ni3S2 and NiS.
Kullerud and Yund (1962) found no
GEOPHYSICAL LABORATORY
441
sign of breakdown of Ni7SG at 200 °C
in 240 days.
The homogeneity range of the pent-
landite phase is known only approxi-
mately and thus is represented only
schematically in Fig. 56. Phase rela-
tions at temperatures below 400 °C
are not completely known ; it appears,
however, that the only significant
changes at 300 °C are the presence of
millerite (the low-temperature form
of NiS), which coexists with pent-
landite and /?Ni7S6; extension of the
violarite-polydymite solid solution to
span the entire FeNi2S4-Ni3S4 com-
positional range; and the appearance
of monoclinic pyrrhotite.
Lundqvist (1947) has presented a
200 °C isothermal section of this sys-
tem. The marked similarity of Lund-
qvist's diagram with the present one
suggests that little additional change
occurs within the system as low as
200 °C. New data indicate that the
Mss, which Lundqvist shows extend-
ing as a homogeneous phase from the
Fe-S join to within 4% of the Ni-S
join, has separated into two distinct
Mss phases near the Fe-S join at
200°C.
400°C
(Fe,Ni)S2+L
(Ni,Fe)S2+L
(Fe,Ni)S
(Ni.Fe) S;
(Fe,Ni)9S6+Mss+Ni7S6
Mss+Ni7S6
6+Ni3S2
FeNi
y+(Fe,Ni)9SB+FeNi3
v y
FeNi3+ r +N;3S2
Fig. 56. Isothermal diagram of the Fe-Ni-S system at 400°C in the presence of vapor. Re-
sults of new experimental data incorporated in this diagram include the occurrence of the violarite
solid solution ([NijFe^S^, the coexistence of NiyS^ and (Fe,Ni)9Ss, and the presence of a homo-
geneous Mss phase spanning the system.
442
CARNEGIE INSTITUTION
o
o
- ; ?
1200
000
a> 800
k_
Z3
a
v_
0)
a BOC
E
200-
lilt
I i
i i i
! i i 1
1 1 1 1 1
en
-
CuS
— DTA
-
l2L52s-
■ —
""do""'
JH- ""
-
510 do
l_
□ Present Data _
-
• Previous Data _
III!
i I
i l l
1 1 1 1 1
! 1 1 1 1
8 10 12
Pressure, ki lobars
14
16
18
20
Fig. 57. Pressure-temperature curve for the inversion in CuS. Present data were obtained by
DTA (chromel versus constantan) with a single-stage piston and cylinder.
High-Pressure Differential
Thermal Analysis of CuS
G. KtiUerud, P. M. Bell, and J. L. England
The condensed Cu-S system is com-
plex at low temperatures because of
the existence of numerous as yet in-
completely defined compounds. The
phase relations become increasingly
complex when the system is exposed
to high confining pressures.
Two copper sulfides, Cu3S4 and
CuS2, which do not exist in the con-
densed system, have recently been
reported to be stable under high con-
fining pressures. In addition, Cu2S,
which occurs in three polymorphs in
the condensed system, has a fourth
high-pressure form. Covellite (CuS),
which melts incongruently to digenite
(Cu9S5) and sulfur liquid at 507°C
CKullerud, 1965), occurs in only one
known crystallographic form.
Kullerud (1965) found that the
univariant equilibrium curve for the
reaction covellite ^± digenite + liquid
Cor gas) passes through the points
510' C, 500 bars; 515°C, 1000 bars;
and 525' C, 2000 bars. These data are
plotted on the diagram of Fig. 57. We
have now investigated the melting
relations of covellite to 11 kb by the
differential thermal analysis tech-
nique described by Bell and England
(Year Book 63, pp. 176-178). In
these experiments covellite, synthe-
sized by the method of Kullerud
(1965), was contained in sealed gold
capsules. Reactions between gold and
covellite were not observed. The re-
sults of the high-pressure DTA ex-
periments are also plotted on the dia-
gram of Fig. 57. The P-T curve for
the covellite +± digenite + liquid
(gas) reaction is noted to have a
positive slope of about 10°C/kb.
Sulfide-Silicate Relations
G. Kullerud and H. S.s Yoder, Jr.
Experiments have been performed
to test the hypothesis that silicate
and sulfide liquids coexist at elevated
temperatures and to investigate the
mutual solubilities among these liq-
uids. The first test was to ascertain
the relationship at elevated tempera-
tures and pressures between a nat-
ural granite and mixtures of sulfides,
GEOPHYSICAL LABORATORY 443
commonly associated with granitic 650 °C and 5 kb. The effect of these
rocks, in the presence of excess H20 sulfides on the temperature of begin-
at 5 kb. ning of melting of G-l is therefore
The G-l reference granite was used less than 10 °C under this particular
with a mixture of galena and pyrite. pressure. Euhedral and subhedral
G-l granite in the presence of excess crystals of PbS and FeS2 and Fe^S
H20 begins to melt at about 655 °C were evenly distributed throughout
(Yoder, unpublished data) under a the recrystallized rock powder. PbS
confining pressure of 5 kb. The sul- was apparently transported at a
fide mixture contained 70 wt % PbS greater rate in the gas phase than
and 30 wt % FeS2, which is very either FeS2 or Fe^S. A considerable
nearly the minimum melting composi- amount of PbS was found to have
tion on the PbS-FeS2 join. Brett and been transported through the rock
Kullerud (1967) found that melting powder and deposited on the wall at
of such a mixture occurs at 724 °C the opposite end of the tube from
under the vapor pressure of the sys- where it was placed at the beginning
tern. The effect of confining pressure of the experiment. This PbS is asso-
on this melting temperature is not ciated with clear euhedral crystals of
known but is estimated to be about quartz; evidently the solubility of
5°C/kb. Si02 in the gas phase is also consider-
Experiments were conducted in able. Experiments lasting from 4 to
gold tubes and were, initially made at 21 hours with the same mixtures were
650 °C for 2 hours. One tube contained undertaken at 675°, 700°, and 725°C
G-l + H20, a second, the sulfide mix- and 5 kb. G-l in the G-l + H20 ex-
ture + H20, and a third, G-l -f sul- periments melted extensively. The
fide mixture + H20. The products of sulfides showed no sign of melting in
each run were studied in thin and the sulfide + H20 experiments. Loss
polished sections, and X-ray powder of sulfur to the gas phase and result-
diffraction charts were made on rep- ing formation of pyrrhotite were ob-
resentative samples. G-l showed no served to be more pronounced than at
sign of melting when heated with 650 °C. Increased transport of sul-
water at 650 °C and 5 kb. In the sul- fides through the gas phase was also
fide + H20 experiment melting was observed. In the experiment at 725 °C
not observed but considerable recrys- a small amount of gold occurred with
tallization, primarily of galena, was PbS. G-l also melted extensively in
evident. Some of the pyrite broke the G-l + sulfides + H20 experiment,
down to pyrrhotite in the presence of whereas the sulfides did not. The
H20. The sulfur released in this proc- quenched products showed that G-l
ess went into the gas phase during when molten develops a very pro-
the experiment but was at least in nounced meniscus convex toward the
part precipitated during the quench- sulfides. The sulfides appeared to be
ing period. There was considerable molded around the silica glass slug,
evidence that both PbS and FeS2 had The glassy phase was strongly ve-
been transported in the gas phase. sicular. The vesicles were commonly
In the G-l + sulfide + H20 experi- lined with sulfides, mostly pyrite and
ment melting of the granite or of the pyrrhotite, and with euhedral crystals
sulfides was not observed, but exten- with a morphology resembling high
sive recrystallization of the sulfides quartz. Polished and thin sections of
was obvious. Addition of the PbS + the glass reveal that complete melting
FeS2 phases does not produce an ob- did not occur. Sulfides, particularly
servable amount of silicate liquid at pyrrhotite but also some pyrite and
444
CARNEGIE INSTITUTION
very occasionally galena, were dis-
seminated throughout the glass and
were highly enriched in numerous
vesicles. A few grains of magnetite
were present. Biotite laths containing
grains of FeS2, Fe^-S, and Fe304
occurred in the glass. Clear, elongated
crystals of high sanidine occurred
with quartz and sulfides in many
vesicles. The combined amounts of
pyrite and pyrrhotite observed in the
polished sections was estimated at
lcc, most of which occurred in vesi-
cles. The solubility of pyrrhotite and
pyrite in the glass itself may not ex-
ceed 0.01%. The galena was again
highly enriched in the opposite end
of the gold tubes from where it was
initially placed. Galena reacted with
the gold tube.
Experiments lasting 8 hours at
740 °C and 1 hour at 750 °C with
the same components and under 5-kb
pressure revealed that PbS -f FeS2
mixtures with excess H20 melt at
both these temperatures. However,
extensive reactions occur between the
gold tubes and the galena. In polished
sections gold is observed to have ex-
solved from the galena phase during
the quench period. At the tempera-
tures of these experiments the ga-
lena containing some gold melted in
the presence of the iron sulfides.
The typical textures likened to finger-
prints by Brett and Kullerud (1967)
were observed in our polished sec-
tions. The solubility of gold in the
galena phase may lower the tempera-
ture of appearance of sulfide liquid
considerably. The conclusion to be
drawn from these experiments is,
therefore, that PbS + FeS2 mixtures
begin to melt somewhere between
725° and 740° C in the presence of ex-
cess H20 and gold and under a con-
fining pressure of 5 kb. Experiments
performed with G-l + sulfides +
H20 showed the coexistence of silicate
and sulfide liquids at 740° and 750°C.
The temperature of appearance of the
sulfide liquid was again influenced by
the presence of dissolved gold.
The gold-tube experiments show
that under a confining pressure of
5 kb and in the presence of excess
H20 an immiscibility gap exists be-
tween granitic and sulfidic liquids
above approximately 750 °C.
Thus, under natural conditions a
rock magma of granitic composition
would coexist with a sulfide (PbS +
FeS2) liquid at temperatures exceed-
ing 750 °C when the confining H20
pressure is 5 kb. If we compare the
beginning-of-melting curve of G-l
granite with the P-T curve for ap-
pearance of liquid on the PbS-FeS2
join (assuming a positive slope of
5°C/kb), we find that the liquid im-
miscibility gap may extend to as low
as 730 °C under a confining H20 pres-
sure of slightly less than 1 kb. These
experiments indicate that the solu-
bility of sulfides in the homogeneous
silicate liquid may be as low as 0.01%.
However, the volatility of sulfides,
particularly PbS, and of sulfur de-
rived from breakdown of FeS2, in the
gas phase is spectacularly high.
The second series of experiments
involved the ferrogabbro (EG 4330)
of the Skaergaard intrusion (see
melting behavior in Tilley, Yoder, and
Schairer, Year Book 63, pp. 92-97)
and pyrrhotite of FeS composition.
The FeS phase was synthesized at
600 °C from the elements in evacu-
ated, sealed silica tubes with mini-
mum vapor space. X-ray powder
diffraction charts made at room tem-
perature showed this FeS to have a
d102 corresponding to stoichiometric
FeS and to have the troilite super-
structure lattice. In the first experi-
ment FeS and ferrogabbro in ap-
proximately a 1:3 weight ratio were
mixed together and loaded into a
gold tube. This tube was firmly
pinched shut and inserted in a silica
tube, which in turn was evacuated
and sealed. The silica tube was heated
GEOPHYSICAL LABORATORY
445
for 4 hours at 1020 °C in a vertical
furnace. After it was quenched to
room temperature the gold tube was
observed to have expanded, the ex-
pansion evidently caused by gases
that had developed. The ferrogabbro
melted completely under these con-
ditions and when observed at room
temperature it was a glass exhibiting
strongly convexed menisci where it
was not in contact with the tube wall.
The glass contained numerous vesi-
cles, which commonly were filled or
lined with sulfide and considerable
amounts of silica.
Considerable amounts of silica and
some sulfide were observed in that
end of the gold tube that was on top
during the experiment. Most of the
sulfide was separated from the silicate
liquid during the experiment and in
the quenched products was observed
to lie molded around the glass. In
polished sections made on the silicate
glass, pyrrhotite, mostly in tiny
rounded blebs, is frequently observed.
Most of this pyrrhotite occurs in vesi-
cles, and it is not possible to estimate
with any degree of accuracy how
much of it occurs in homogeneous
glass. The solubility of the pyrrhotite
in ferrogabbro glass, however, could
not have exceeded 0.1%. Many of the
pyrrhotite blebs occurring in vesicles
in the glass, as well as near the bor-
der between the silicate glass and the
surrounding sulfide mass, appear in-
homogeneous even at low magnifica-
tion. On closer study, it is seen that
the FeS reacted with the silicate liq-
uid to create small amounts of a sec-
ond liquid. On quenching, the second
liquid crystallized to pyrrhotite and
a brownish phase that occurs in a
fingerprint-type pattern in the pyr-
rhotite blebs (see Plate 3). Electron-
probe investigation of this phase
shows that it is magnetite. Deter-
mination of the d102 value of the iron
sulfide indicated that the composition
of this phase changed during the ex-
periment. The quenched sulfide has
a composition of Fe48S52, as compared
with the Fe50Sr,o composition of the
starting material. Presumably some
of the iron removed from FeS was ab-
sorbed by the gold tube ; it is possible,
however, that some or most of this
iron reacted with the silicate liquid
to produce magnetite.
As described in another section,
melting occurs on the magnetite-pyr-
rhotite join at elevated temperatures.
Evidently the temperature of the pres-
ent experiments (1020°C) was suffi-
ciently high to produce a magnetite-
pyrrhotite liquid. Because of the large
amounts of pyrrhotite present, all
magnetite was incorporated in the
liquid and pyrrhotite coexisted with
this liquid. Under the conditions of
this experiment, ferrogabbro liquid
coexisted with small amounts of sul-
fide-oxide liquid and with large
amounts of pyrrhotite of or near
Fe48S52 composition. Thus, liquid im-
miscibility exists in the condensed
ferrogabbro-FeS system at 1020 °C.
The silicate liquid is essentially of the
composition postulated from field ob-
servations; but the composition of
the coexisting liquid lies on or near
the magnetite-pyrrhotite join, and it
should be emphasized that it does not
consist solely of sulfide.
Experiments with ferrogabbro and
FeS were also performed under a con-
fining pressure of 10 kb and at 1075°,
1150°, 1175°, and 1200°C. Because of
the melting of gold at 1083 °C, these
experiments were made in platinum
tubes. We had previously noted (Year
Book 64) that when ferrogabbro is
exposed to elemental sulfur in Pt
tubes at elevated temperatures, much
of the sulfur combines with Pt to
make PtS and PtS2. However, in the
present experiments FeS was used
instead of elemental sulfur with the
hope that such reactions would not
take place. This hope was justified in
part.
446
CARNEGIE INSTITUTION
The ferrogabbro melted in the ex-
periments conducted at 10 kb. Magne-
tite-pyrrhotite liquid was found to
have existed only in the experiments
conducted at 1175° and 1200°C and
10 kb.
Previous experiments (Year Book
6J+) on mixtures of ferrogabbro and
elemental sulfur yielded iron-deficient
pyrrhotite and magnetite. The pres-
ent experiments involving ferrogab-
bro and FeS also produced iron-de-
ficient pyrrhotite and magnetite.
Ferrogabbro in the first instance
gives up some iron to produce pyr-
rhotite by reaction with sulfur and
in the second instance accepts iron
from FeS to produce Fe^S. Magne-
tite is a ubiquitous phase in these re-
actions.
These experiments show that ferro-
gabbro liquid coexists with magne-
tite and pyrrhotite. The pyrrhotite
composition (of about Fe48S52) is
buffered in this assemblage and is
therefore uniquely controlled at any
given temperature and pressure. A
similar buffering mechanism is
exerted by noritic or peridotitic
magmas. Sulfides in mafic rocks of
these types are always associated with
magnetite, and the metal-to-sulfur
ratio of the combined sulfide minerals
usually lies within narrow boundaries
of 37.5 to 39.0 wt 7o S because of the
buffering process described. It is be-
lieved, therefore, that the underly-
ing principle which relates specific
ore deposits to specific rock types is
a direct result of the partition of
metals to sulfur and silicate.
INCLUSIONS IN DIAMONDS
Mineral Inclusions in Diamonds
Henry 0. A. Meyer
Mineral inclusions that occur in dia-
monds are interesting as possible
samples of the upper mantle. In cer-
tain cases they are syngenetic with
the diamond host and are chemically
unaltered, having been preserved
within the diamond.
Few studies have been made of
these mineral inclusions. Apart from
the visual identification of inclusions
by Bauer (Bauer and Spencer, 1904)
and Sutton (1921), and the observa-
tions of Williams (1932), no X-ray
identification was made until that of
Mitchell and Giardini (1953). The
last-named authors observed that sev-
eral of their olivine inclusions were
oriented relative to the diamond.
Futergendler and Frank-Kamenetsky
(1961) have also demonstrated crys-
tallographic control between inclusion
and diamond host, and this has been
substantiated in part by Harris,
Henriques, and Meyer (1966). This
report presents new data obtained
with an electron microprobe.
The minerals identified by X-ray
diffraction and electron-probe tech-
niques as primary inclusions in dia-
mond are olivine, orthopyroxene,
clinopyroxene, garnet, spinel, coesite,
and iron-nickel sulfides. All these
minerals occur as discrete crystals
within the diamond. The average size
of the inclusions now being studied is
of the order of 0.05 X 0.03 X 0.01
mm. Their minute size has necessi-
tated the development of special ana-
lytical techniques. The inclusions are
freed by oxidation of the host dia-
mond in air for several hours at ap-
proximately 800 °C. Cell dimensions
are calculated with the use of high 20
reflections (approximately 160°) ob-
tained from single-crystal Strau-
manis-type X-ray photographs. In the
case of the garnets, refractive index
measurements have also been made.
Selected crystals are mounted in
epoxy cement and polished in prepa-
ration for electron-probe analysis.
The initial analyses with the electron
probe were directed toward estima-
tion of the abundance of the constitu-
ent elements. Wide chemical variation
has not been observed. Quantitative
GEOPHYSICAL LABORATORY
447
"O
C
o
E
D
O
Q
0
Q.
O
E
E
D
CO
U">
CO.
<
O
E
to
*-
C
<D
E
_o
</>
LU
*«7>
ha
_>.
o
O
c
C
<
5
<L>
>
*«£!.
D
«/»
4—
•^
c
o
(1)
3
£
o
jd
LU
La
o
'o
^
CD
O
0.2
£ ."= <D
X O C
o a o
a E <*
a o
-Q.E
E £
= o
Z*
0) «/>
V 2
<1> ■*■
O c
o
2 §
a .2
■o
co o
!E <u
"o-S
O
o
co
o
U
II
E E
o o
c E
5-8
u
o
U
.<
7T. O
ii co
O)
*OMr-
NCNO
I I I
Or- 0>
O
U
CO
©
IX.
U
<
d)
<5 — '
o
o -Q o
«+-
c
o
o
co
a
'in
_2
>*
u
I-
c
c
a>
c
o
£
4)
a
o
>_
>.
Q.
CO
c
1—
o
o
o
E e
'Z o
«- U
© c
JZ —
<u >-
1 5
> -o
c
.£ 8
o a)
t: «"
© >-
u o
c
U
o
u
co
d)
O 00 o
CO 00 CN
00 CO IT)
7 i i
io co o
CMOOp-
T O IO *— 00 CO if)
O -Q O
to
c
©
©
X
o
o
c
-<
■<
d)
d)
u_
^
d)
CO
^
N
o
Z
_ D)
aT ©*"
UU
co *:
cd £
©
©
c
*a
co
co
<
O
CN
tot°
CNr-CS'O
io c> c> °"
6i oCQ.
© _
9- CN
©
a> c ■*-
•1° I
U
^ —
■o
©
c
E
k.
©
©
o
c
c
£
o
U
448
CARNEGIE INSTITUTION
analysis of a number of specimens is
under way. Preliminary results are
summarized in Table 15.
Olivines. The olivine inclusions are
forsterite. Forsterite has been shown
previously (Mitchell and Giardini,
1953) to be often crystallographically
oriented with respect to the host dia-
mond, the relation being that the b
axis of forsterite is perpendicular to
the {111} of diamond. A new method
has been employed (Meyer, Donnay,
and Donnay, this report) in which
X-ray precession goniometry is used
to examine inclusions in diamond in
which orientation is suspected. The
coexistence of two phases, olivine and
orthopyroxene, within a single host
diamond and also the coexistence of
olivine and garnet in another dia-
mond are significant and have not
previously been reported. The varia-
tion of minor elements (e.g., Al, Ni,
and Mn) between the coexisting
phases should yield valuable informa-
tion on the conditions during growth.
Garnets. All the fourteen garnets
so far examined are pyrope-alman-
'.80 —
x
cd
"O
CD
>
O
a
^_
CD
en
.78 —
.76 —
1.74
1.72 —
I
T
l Almandine
I
1 1
—
+
•
GARNET SOURCE
•
Peridotite, Pyroxenite Nodules
+
Nixon
I960
+
Eclogite, Granulite Nodules
—
A
o
X
Garnet, Garnet-Diopside Nodules
Kimberley area
—
* A
+
A
Inclusions in Diamond
—
A
A
1
o
o •
•
—
n°8
°o *• •• •
•• •
• X
X
•
x X X
•
•
■ "
-
• '
X
-
—
(Pyrope
, " Grossular
i i 1
,DU
1.55 11.60
Cell edge (A)
1.65
1.70
Fig. 58. Indices of refraction as a function of cell edge (A) for garnets in the field pyrope-
almandine-grossular. Data for garnets from various nodules are taken from Nixon (1960). Data
for garnets from Kimberley area and garnet inclusions are taken from the present work.
GEOPHYSICAL LABORATORY
449
dine. Nixon (1960; Nixon et al, 1963)
plotted the refractive index of various
garnets, subdivided on rock type,
from Basutoland kimberlites against
cell edge. On the same diagram (Fig.
58) are plotted nine new garnets
from the Kimberley area and also
seven inclusion garnets. The inclu-
sions all lie in the garnet nodule-to-
eclogite areas as plotted by Nixon.
Qualitative electron-probe results
have demonstrated that these garnets
contain appreciable quantities of
chromium, possibly more than suffi-
cient to balance the calcium present
in the uvarovite end member. In this
case the chromium-bearing end mem-
ber hanleite (Fermor, 1952) will
have to be considered (Nixon, 1960;
Nixon et ah, 1963.)
Spinel. Although spinel had been
previously identified in Russian dia-
monds (Futergendler and Frank-
Kamenetsky, 1961), this is the first
record of its occurrence in African
material. No quantitative analyses
have yet been made.
Coesite. Coesite is usually observed
in rocks metamorphosed by shock
(Chao, Shoemaker, and Madsen,
1960) and in tektite glass (Walter,
1965) whose origin is still a source of
controversy. Two natural diamonds
are known to contain coesite (Mill-
o
1600
1400
1200
E 1000
soo
600
400
Depth, km
100
200
300
cJC^
<p*
&
probable growth region
for natural diamonds
I
I
I
I
I
/
/
/
50 60 70
Pressure . kb
90
100
i20
Fig. 59. Phase equilibria data combined with geotherm for Precambrian shield. Geotherm from
Ringwood (Year Book 63, p. 153); reaction 4 enstatite + spinel &> pyrope + forsterite from
MacGregor (Year Book 63, p. 157); quartz *± coesite slope from Boyd and England (1960b)
modified to pass through 37.5 kb at 1600°C (Boyd, Bell, England, and Gilbert, Year Book 65, p.
412); graphite *± diamond from Bundy, Bovenkerk, Strong, and Wentorf (1961); and coesite-
stishovite curve from Stishov (1 963).
450
CARNEGIE INSTITUTION
edge, 1961; Harris. Henriques, and
Meyer, 1966), identified by X-ray
diffraction techniques, and this con-
stitutes the first occurrence of "pri-
mary" coesite. The presence of this
mineral in diamond is of great inter-
est, as it gives an upper limit to the
pressures involved in the formation
of the diamond, i.e., below the coesite-
stishovite equilibrium line (Fig. 59).
Undoubtedly the preservation of this
site owes much to the protective
effect of the diamond host.
Iron-nickel sulfides. Pyrrhotite and
also pentlandite have been recorded
as inclusions in South African dia-
monds (Sharp, 1966). The present
technique of oxidizing the diamond
host at 850°C destroys sulfides. Hem-
atite, which is the expected oxidation
product for an Fe-rich sulfide, has
been recovered from several dia-
monds.
Mica. Biotite has been recorded
(Williams, 1932, p. 421) as an inclu-
sion in diamond. The identification
was made by F. E. Wright of the
Geophysical Laboratory. During the
present study 2MX muscovite was
identified. Unfortunately the mica
was discovered after the diamond
host had been oxidized at 850 °C for
several hours; it is unlikely, however,
that the major cations would have
been lost during oxidation. Qualita-
tive analysis indicates that chemically
the specimen is a muscovite that con-
tains some paragonite.
Owing to the opacity of the host
diamond, it is not certain whether
this mica was a primary or secondary
inclusion. In spite of the uncertainty
the specimen is still significant, as
muscovite has not been recorded in
association with diamond.
In conclusion, it is evident from X-
ray and qualitative electron-probe
results that there are many similari-
ties between inclusions in diamond
and the primary minerals in kimber-
lite nodules. The principal differences
observed to date are the coesite and
muscovite identified as inclusions in
diamond. These minerals have not
been found as primary phases in the
nodules. It seems probable that dia-
monds form in the same geochemical
environment as the other primary
phases. The relative uniformity of
composition and the limited number
of major elements observed from
these qualitative results are striking.
The quantitative work now in prog-
ress will enable a more rigorous com-
parison to be made, and it will be
informative, in view of the lack of
major element variation, to determine
the possibility of significant variation
in minor element content, not only
between similar inclusions but also
between inclusions and the associated
primary minerals.
IGNEOUS PETROLOGY
Melting Relations of Igneous
Rock Series
C. E. Tilley, H. S. Yoder, Jr., and
J. F. S chair er
Experimental studies on the melt-
ing relations of both volcanic and in-
trusive members of igneous rock se-
ries are presented in this report in
continuance of earlier studies on vol-
canic rocks (Yrnr Book 65, pp. 260-
269 and references therein to earlier
work) .
Studies on the calcalkali volcanic
series of Paricutin, Mexico, reported
last year, have been extended to cover
a larger suite of analyzed lavas.
Strontium isotope ratios of basalt and
andesite in this succession are iden-
tical and characteristic of basalts.
These results appear to preclude the
possibility of the origin of the ande-
sites by assimilation of old sialic
crustal material. Melting work on the
unique tholeiitic volcanic series of
Thingmuli, Iceland — extending from
GEOPHYSICAL LABORATORY
451
picrite basalt to rhyolite as a frac-
tional crystallization trend — is com-
pared with the earlier reported ex-
perimental study on the historic Ki-
lauea tholeiite succession. Liquidus
determinations already reported on
the border facies or so-called "chilled
margins" of layered igneous com-
plexes have been extended to cover
the marginal facies of other intru-
sions. The experimental results em-
phasize the need to consider a re-
assessment of the validity of the con-
cept that these assemblages represent
parental magma of the individual
complexes.
Calcalkali Volcanic Series of
Paricutin, Mexico
The extension of the work pre-
sented in last year's report (Year
Book 65, pp. 266-268) now yields ex-
perimental data on lavas ranging
from 51% to 60% silica, involving
the high-alumina basalt ejected at the
beginning of the eruption in Febru-
ary 1943. The results are reported in
Table 16. The liquidus determinations
of these lavas range from 1227 °C
(high-alumina basalt) to 1197°C (hy-
persthene andesite). Despite the in-
creasing alkali and iron enrichment,
TABLE 16. Results of Melting Experiments on the Calcalkali and Tholeiitic Lavas
of Paricutin, Mexico, and Thingmuli, Iceland, and on the
"Border Facies" of Layered Intrusions (Table 17)
Rock Identification
Highest Temperature of
Crystallization of Major Phases
n of Glass
High-alumina basalt
Bomb ejected February 20, 1943 (Bl)
High-alumina basalt
Lava of March 1944 (HAB)
Hypersthene andesite
Lava of March 1943 (2)
Hypersthene andesite
Lava of September 1 947 (14)
Hypersthene andesite
Lava of September 1 950 (1 9)
Paricutin Lavas
Ol (1 227°); PI (1 208°); Opx, Cpx (1 1 45°)
PI (1220°), 01(1215°), Opx (11 65°), Cpx
(1155°)
PI (11 98°); Opx, Ol (1173°)
PI (1 203°), Opx (1 178°), Ol (1 170°)
PI (1197°), Opx (1145°)
Thingmuli, East Iceland, Lavas
Olivine tholeiite, east of Hottur (2) PI (1 210°)
Tholeiite, Eggjar Ridge (9) PI, Cpx (1 113°)
Tholeiitic andesite, Thingmuli Hill (1 3) PI (1 1 05°)
Tholeiitic andesite, south of Eggjar PI, Cpx (1 1 05°)
Ridge (12)
Icelandite (inninmorite), northwest of PI (1105°)
Hottur (17)
Norite, Bushveld (Bl
Border Facies of Layered Intrusions
PI (1 243°), Opx (1 1 93°), Ol (1 1 87°), Cpx
(1173°)
Hypersthene gabbro, Bushveld (B2) PI (11 95°); Opx, Cpx (1173°)
Chilled bronzite gabbro, Muskox, Ol (1 205°); Opx, Cpx (1 1 95°); PI (1 1 77°)
Canada (Mx)
Olivine basalt, Ubekendt Island, West Ol (1 267°); PI, Cpx (1 1 88°)
Greenland (LA)
1.570
1.553
1.553
1.543
1.540
1.605
1.593
1.575
1.555
1.525
1.584
1.583
1.595
1.605
452
CARNEGIE INSTITUTION
21
-
::
-
19
-
18
-
17
-
16
-
15
14
13
12
-
O
o> II
IE
-
10
9
8
7
6
5
4
3
2
1
rhyo
2
1 \^t
21
Fig. 60. Plot relating total iron oxides and magnesia of trends of the Paricutin (Mexico) cal-
calkali series and of the Thingmuli (Iceland) tholeiitic series. Squares: Pancutin series Bl , analysis
1, Table 2, p. 362, in Foshag and Gonzalez (1956); HAB, analysis 3 from Year Book 65, Table
5, p. 264; 2, 14, 19, analyses of Table 2, p. 294, in Wilcox (1954). Circles: rocks of the Thing-
muli series; numbered points, analyses of Table 9, p. 455, in Carmichael (1964). Liquidus tem-
perature of rocks experimentally treated are indicated.
the later andesitic lavas show a re-
markable similarity in their liquidus
temperatures and plagioclase is the
liquidus phase.
The trend of the Paricutin series is
illustrated in the iron oxide/magnesia
plot of Fig. 60. Whether the Paricutin
sequence is to be interpreted as a
straight fractional crystallization
series from the early high-alumina
basalts or sialic assimilation has
played a role has been debated. Xeno-
liths of granite and quartz-monzonite
are found in the ejecta and are now
High-alumina basalt (HAB)
Hypersthene andesite (19)
Rb
ppm
8.4
26.95
Sr
ppm
581
516
Sr87/sr86 Rb/Sr
0.7043
0.7040
0.014
0.052
GEOPHYSICAL LABORATORY 453
transformed by partial fusion and defined by the crystallization of
local reaction with their lava enve- olivine, an intermediate section (2-5,
lopes. Contamination by such sialic 9) dominated by the crystallization
country rock at depth has been sug- of plagioclase and pyroxene as well
gested by the studies of Wilcox as olivine, and a lower trend (9-21)
(1954). where the otherwise progressive iron
To examine this possibility, G. L. enrichment is rapidly decreased as
Davis and T. E. Krogh kindly car- magnetite becomes a phenocrystic
ried out strontium isotope studies on crystallizing phase,
two contrasted members of the suite Five analyzed rocks of this se-
— a high-alumina basalt (HAB, Si02 quence have been experimentally
= 54.08%) and a hypersthene ande- studied, ranging from olivine tholei-
site (analysis 19, Si02 = 59.93%). ite to icelandite, and the results of the
Their results are tabulated as follows, melting work are set down in Table
The close identity of the strontium 16. To avoid significant oxidation the
isotope ratios, values of which cor- fine powders were sealed in a plati-
respond to those characteristic of num tube and surrounded by iron fil-
basalts, would appear to preclude ings in separate platinum foil, and
contamination by old crustal sialic the whole enclosed in an evacuated
material for the Paricutin andesites. silica tube.
It is not precluded, however, that The liquidus temperatures with
assimilation of young sialic material plagioclase as the liquidus phase
— a differentiate of basic magma and range from 1210 °C (olivine tholei-
thus comagmatic — may have played ite) to 1105°C (icelandite), the more
a part in the form of internal hy- sialic members (13, 12, and 17) giv-
bridism. ing identical liquidus temperatures
It will be of interest to determine of 1105 °C.
the age and strontium ratios of the In the upper temperature region
visible sialic material thought to of the Thingmuli trend, the course
represent the underlying sialic coun- is similar to that of the historic Ki-
try rock, and such studies are in lauea trend as is revealed in the plot
progress. of Fig. 61. The Kilauea trend ex-
tends to 82 (lava of 1955 with liqui-
The Volcanic Series of <*us of 1135°). The Thingmuli trend
Thingmuli Iceland continues on the same course to 9
(liquidus 1113°) and then turns in
The tholeiitic series of Thingmuli, the direction of marked alkali en-
Iceland, investigated by Carmichael richment through the tholeiitic ande-
(1964) is a unique suite ranging sites to icelandite (17), which have
from picrite basalt through olivine much lower liquidus temperatures
tholeiite, tholeiite, tholeiitic andesite than the calcalkali andesites of the
to dacite (icelandite — inninmorite) Paricutin succession,
and rhyolite, and is believed to repre- In comparison with Kilauean
sent a fractional crystallization se- representatives of the same iron en-
quence. The trend of the series is richment, the olivine tholeiite has
shown in the iron oxide/magnesia plagioclase instead of olivine on the
plot of Fig. 60 and the FMA plot of liquidus, and this probably may be
Fig. 61. accounted for by its relatively high
In Fig. 60 the trend is shown di- alumina content (14.9%, normative
visible into three sections: an upper plagioclase 50%, An56). The Kilauean
section (25-2) where the control is representatives have only 13%-14%
454
CARNEGIE INSTITUTION
FeO + Fe205
MgO
Fig. 61. Trends of the historic Kilauea and Thingmuli (Iceland) tholeiitic series in the FMA sys-
tem— F (FeO + Fe203), M (MgO), A (Na20 + KzO). Circles: Kilauea series 1 840Pb, 1 840b,
K1921, 82 from Year Book 62, Fig. 13, p. 81. Triangles: Thingmuli series 2, 9, 13, 12, 17, anal-
yses of Table 9, p. 455, in Carmichael (1964). Also plotted is the residual liquid of the Skaergaard
[Skj] layered intrusion. Analysis in Year Book 63, Table 1, p. 93. Liquidus temperatures of rocks
experimentally treated are indicated.
alumina.* Plagioclase, however, is
the liquidus phase in Kilauea lavas of
higher iron enrichment, e.g., 1840b
* It is known also from previous studies
(Yoder and Tilley, 1962, p. 453) that the
presence of water delays the crystallization
of plagioclase as in the example of the War-
ner basalt there described (Fig. 28). It is
therefore of interest to record an experi-
mental run of the Thingmuli rock (2) at
1030" C, 2 kb, H20, for 2 hours in a sealed
platinum tube, which gave a product consist-
ing of euhedral olivine and clinopyroxene
with some acicular outgrowths of amphibole
in a light brown glass holding fine-grained
plates of biotite. Plagioclase had not yet
crystallized and either olivine or clinopyrox-
ene would clearly be the liquidus phase.
(1150°) and 82 (1135°), as recorded
in Fig. 61. The Kilauea trend ceases
at the 1955 lava (82) , but the marked
change in the Thingmuli trend, al-
ready referred to as associated with
the early phenocrystic separation of
magnetite at approximately this stage,
contrasts strongly with the extreme
iron-enrichment course followed by
some other rock suites, notably the
Skaergaard succession (Sk3 in Fig.
61).
Border Fades of Layered Intrusions
Experimental runs on border facies
or chilled margins of several layered
GEOPHYSICAL LABORATORY
455
intrusions have been previously re-
ported— notably those of the Skaer-
gaard and Stillwater layered com-
plexes (Year Book 62, p. 81, Fig. 13) .
A study has been conducted of the
melting relations of the reported
border facies of the Bushveld and
Muskox intrusions, and of an apha-
nitic facies of a zoned picrite basalt
intrusion from Ubekendt Island.
West Greenland; the melting results
are reported in Table 16 and plotted
in Fig. 62.
In the two Bushveld examples pla-
gioclase is on the liquidus but the
temperatures show a difference of
48°C. The Muskox and Ubekendt
examples have olivine on the liquidus
at temperatures of 1205° and 1267°C,
respectively. These ''border facies,,
assemblages have been variously in-
terpreted— first as a chilled repre-
sentative of the parental magma of
the particular layered intrusion, and
more recently, as in the case of the
Muskox and Ubekendt examples, as
a chill product of flowage differentia-
tion.
Field data on the Bushveld exam-
ples as marginal facies are in some
doubt, and it is unlikely that the
coarse-grained Daly example (Bl) is
correctly interpreted. The second ex-
ample (B2) is a much finer grained
rock, also with plagioclase on the
liquidus (1195°C), all the major
phases appearing over a narrow tem-
perature interval. The rock (Table
17, analysis 2) carries normative
olivine but none modally, an unex-
Fe0+Fe203
1840b (1150°)
^
\vn
o
**£
o
_y_
v
40
30
BIO 243°)
a
I? St (1245°)
a Mx(l205°)
□
Sk,
(1230°)
□ LA (1 267°)
(1435°)
1840 Pb
MgO
Fig. 62. Plot of analyses of border facies of layered intrusions in the FMA system, referred to
the historic Kilauea tholeiitic trend (1 840Pb, K1921, 1 840b, 82). Circles: Kilauea trend, lettered
points as in Fig. 61; squares: border facies of Skaergaard intrusion (Ski), Stillwater intrusion (St),
Muskox intrusion (Mx), Bushveld intrusion Bl and B2 (Table 17, analyses 1, 2); LA, lime-rich
aphanite of picrite basalt intrusion of Ubekendt Island, West Greenland (Table 17, analysis 4). Ski
and St from Year Book 62, Fig. 13, p. 81. Liquidus temperatures are indicated.
456
CARNEGIE INSTITUTION
TABLE 17. Chemical Analyses and Norms of Investigated Rocks
("Border Fades" of Layered Intrusions)
1
2
3
4
Analyses
Si02
51.45
50.55
51.33
47.01
AI2o3
18.67
15.23
13.69
13.58
Fe203
0.28
1.04
0.93
2.70
FeO
9.04
10.07
9.33
8.28
MnO
0.47
0.23
0.18
0.18
MgO
6.84
8.30
8.94
11.45
CaO
10.95
11.30
11.60
11.32
Na20
1.58
2.24
1.84
1.90
K20
0.14
0.19
0.50
0.18
H20-
0.34
0.24
0.37
0.69
H20-
0.05
0.98
Ti02
0.34
0.66
1.01
1.53
p2o5
0.09
0.12
0.10
0.15
Cr203
0.01
0.09
0.02 (C02)
NiO
. . .
. . .
0.02
0.02 (F)
0.00 (CI)
Totals
100.19
100.18
Norms
99.98
99.99
Less
O = F 0.01
99.98
Qz
2.94
0.36
Or
0.83
1.12
2.78
i.ii
Ab
13.37
18.95
15.72
16.24
An
43.44
30.96
27.52
27.80
Di
8.51
20.00
23.80
22.08
Hy
29.50
20.57
25.59
11.37
Ol
• • •
5.29
12.46
II
0.65
1.25
1.97
2.89
Mt
0.41
1.51
1.39
3.94
Ap
0.21
0.28
0.34
0.34
Remainder
0.34
0.25
0.44
1.71
Totals
100.20
100.18
99.91
99.94
1 . Norite, 1 mile south of Sjambok railway station, northwest of Pretoria, Transvaal (Daly,
1928, p. 727).
2. Fine-grained hypersthene gabbro, 4% miles east of Klip River, road between Lydenburg
and Roos, Transvaal (Atkins, 1965, p. 66).
3. Chilled bronzite gabbro, east margin of the Muskox intrusion, 25 feet from contact, from
drill hole 3500 feet north of the Copper Mine River, Northwest Territories, Canada. Analytical
Chemistry Section, Geological Survey of Canada.
4. Olivine basalt (lime-rich aphanite), nonporphyritic facies in the Igdlorssuit intrusion (picrite
basalt), Ubekendt Island, West Greenland (Drever and Johnston, 1 966; analyst G. O. Riddle).
pected feature in basic assemblages crystallize over a narrow temperature
of this gabbro type. If the Muskox range, the Ubekendt example shows
and Ubekendt examples are the prod- a wide interval between the liquidus
ucts of partial melting of a deep- phase temperature (olivine, 1267°C)
seated peridotite, the source of the and the onset of the succeeding major
latter has suffered the more drastic phases, plagioclase and clinopyroxene
partial melting, for whereas the at 1188°C. Smith and Kapp (1963)
major phases in the Muskox example consider that the Muskox intrusion
GEOPHYSICAL LABORATORY 457
has a higher olivine content than 3CaMgSi04 + CaMgSi2Oc ^
would normally be expected from a (Mo) (Di>
magma chemically similar to its chill 2Ca2MgSi207 + Mg2Si04
facies, and they conclude that the (Ak) <fo)
chill phase represents only the fluid „ , . . ,, „ _
portion of the solid olivine + fluid °f ™hich th\p-T ™™e **• Jem out-
mag-ma lined in part by Walter (1963) and
The melting data now assembled on ™re recently determined by Yoder
these border facies of layered com- <Jhls report). The alternative assem-
plexes reinforce the need to reassess ^ages .of *gn^us <™&* are illus-
the concept of these rocks as parental ^axted lx? %• 63 as <*> and c ' and
material, especially in the light of ft) and (d), respectively (c) and
the phases and their bulk composition <d> Portraying the assemblages with
as revealed in the early accumulate nephelme in the expanded basalt
assemblages of these complexes. tetrahedron (Schairer and Yoder,
The writers are greatly indebted to \ear . Bo°k 63> P- 65> Flg' }>• Tne
the donors of rocks under study in chemically more complex natural as-
this account, particularly F. B. At- semblages represented by the rocks
kins, I. S. E. Carmichael, H. I. of which analyses are set out m Table
Drever, C. H. Smith, and R. A. Wil- J?^ f}1™ contents ranging from
cox- and to F B Atkins H I Drever S1% to 40^' These and other rocks
and C. H. Smith for permission to Previously reported (Year Book U,
report the unpublished analyses of £able „3> and a nephelme basanite
Bushveld, Ubekendt Island, and Year Book 65, Table 5, analysis 2)
Muskox rocks, respectively, reported have been run experimentally at high
in Table 17 pressures and temperatures (1) after
heating the sample at 110 °C and (2)
with excess water added. The rocks
The Pyroxenite Facies Conversion contain sufficient H20+ to yield acces-
of Volcanic and Subvolcanic, sory hydrous phases (hornblende,
Melilite-Bearing and Other biotite) in some of the runs.
Alkali Ultramafic Assemblages Experiments have been carried out
C. E. Tilley and H. S. Yoder, Jr.
under pressures of 2, 5, and 10 kb
and at temperatures ranging from
The melting relations at 1 atmos- 900° to 1200°C. The most striking
phere of an assemblage characteristic result of the runs without excess
of alkali igneous ultramafics, namely, water is the transformation of the
melilite olivine nephelinites ("melilite rocks into assemblages in which the
basalts") was discussed in Year Book dominant phase is clinopyroxene with
6U (pp. 76-78). In this report atten- the minor accessory phases, biotite,
tion is devoted to these and other hornblende, olivine, and magnetite,
ultramafic rocks, namely, those types In runs with excess water, especially
designated as (1) melilite olivine at the lower pressures, there is a
nephelinite, (2) monticellite alnoite, prominent development of hornblende
(3) monticellite olivine nephelinite, along with clinopyroxene. Products
(4) olivine nephelinite, and (5) of the runs made without excess
nephelinite (Fig. 63). water are essentially pyroxenites
From experimental work on the holding one or more of the phases
system CaO-MgO-Si02 simplified as- olivine, hornblende, biotite, and mag-
semblages of this character are netite as accessories. The monticel-
known to be related by a reaction lite alnoite from Sutherland Com-
458
CARNEGIE INSTITUTION
1
o>
-s
*s
*- ^o
o>
-
4—
«/> Q.
ha
0)
o
*»j ~^
s
0> -*
M
u o
a
- 1
o
t5
° L
i
L- V.
o
Q. O
jjj
d)
c
!r. >-
D
a>
E
* .-•*
c
o a>
<D
~~ "O
J*
«• o
O
£ >-
o
2 -o
JZ
O C
■^
o 0
^c
a l.
«/)
<D .-
.*
*- O
u
i_ -c
O
o y
k_
J= <^>
D)
V)
3
^ 5
O
a>
c
o
-Q T3
— <l)
eO)
«/> o
#u
Oj £
*E
D
*u a>
D *-
_u
O
>
3
0) "5
in
a>
i; "D
"O
a. a>
i° c
Ll. o
*c
-c 0)
o
u
O
>
La
"5
0) C
2- *
E o
0
<u -*=
■4- «/>
'U'
!r, o
O u.
JE
■*-
£ o
o
o — .
a
a
«/>
2
D 15
* — ' c
0
•<; D
c
o
Q —
e
CO <L)
^ _
61
o
D) O
y
"o
^ a
o 8
DO,.
V)
2
E £
<D
i> W)
H-
"£ a>
>- D)
•4-
o
i/» 0
a> _d
l/>
£ E
o X
_D
-Q
c a
E
O <y
0 «/»
1 to
4) -c
■«- <?
</» ■*-
.
IS •/>
CO
o c .
<o
u O •—
O)
2 o o>
iE
E a UZ
GEOPHYSICAL LABORATORY
459
TABLE 18. Chemical Analyses and Norms of Investigated Rocks (Alkali Type)
1
2
3
4
Analyses
Si02
31.30
36.03
40.54
39.14
AI2O3
8.14
15.19
16.07
12.23
Fe203
5.43
5.94
4.85
5.28
FeO
7.80
9.55
6.14
8.02
MnO
0.27
0.17
0.28
0.20
MgO
19.23
8.60
4.33
12.86
CaO
18.16
15.52
12.77
13.11
Na20
2.27
4.23
5.47
3.95
K20
1.07
1.85
3.70
1.23
H20+
1.82)
0.51 S
0.58
(0.90
/ 0.33
0.48
H20~
0.16
Ti02
3.20
1.13
3.53
2.54
p2o5
0.49
1.38
1.07
0.58
co2
0.24
. . .
...
0.13
Zr02
. . .
0.21
100.38
. . .
Totals
99.93
99.98
99.91
Norms
Ne
10.37
19.31
24.99
18.18
Ks
3.63
6.32
• • •
4.11
Lc
• ■ •
...
17.00
. . .
An
8.77 (Ge)
16.96
8.62
11.95
Di
2.62
13.72
27.05
30.60
Ol
19.82
8.87
2.20 (Wo)
17.16
Mo
36.89
20.46
...
. . .
Cs
* • •
• • •
2.92
3.18
II
6.08
2.13
6.69
4.86
Mt
7.89
8.58
6.96
7.66
Ap
1.02
3.19
2.35
1.34
Ct
0.50
...
...
0.30
Remainder
2.33
0.79
1.23
0.64
Totals
99.92
100.33
100.01
99.98
1. Monticellite alnoite, Sutherland Commonage, Cape Province, South Africa. Analyst, J. H.
Scoon.
2. Monticellite olivine nephelinite, Shannon Tier, Tasmania (Tilley, 1928, p. 30).
3. Leucite nephelinite, Etinde, Cameroons (Tilley, 1953, p. 147).
4. Olivine nephelinite, dike in the Pali, Honolulu (Yoder and Tilley, 1962, p. 362).
monage, Cape Province, South Africa
(cf. Taljaard, 1937) requires sepa-
rate description. The rock (Table 18,
analysis 1) is built essentially of
olivine, monticellite (in part mantling
the olivine), melilite, and nepheline,
with magnetite and perovskite (Plate
5D). Clinopyroxene is absent, but
in the experimental runs this mineral
becomes a significant phase. In some
runs merwinite is recorded, but the
associated minerals, which include
olivine, monticellite, melilite, and
biotite, are also significant. These
assemblages have proved to be of
considerable complexity, and require
continued study in view of the possi-
bility of nonelimination of all resid-
uals. At 1 atmosphere the liquidus
phase at 1415°C is olivine.
The disappearance of melilite and
nepheline and the major part of the
olivine in the conversion of rocks
1, 3, 4, and 5 (Fig. 63) would indicate
that the original rock types are
strictly confined as volcanic and sub-
volcanic facies, represented at depth
by pyroxenites or their hornblende
and biotite-bearing variants.
Photomicrographs of several of the
460
CARNEGIE INSTITUTION
assemblages experimentally treated,
and of their conversion products
under the P-T conditions noted, are
set out in Plates 4 and 5. Pertinent
in this connection is the reported
presence of biotite-pyroxenites asso-
ciated with volcanic alkali ultrama-
fics, notably in the volcanic province
of Bufumbira, South West Uganda
(Holmes and Harwood, 1937), where
they appear as xenolithic ejecta.
Bowen (1922), in his study of the
pyroxene-bearing monticellite alno-
ites of lie Cadieux, Quebec, has
traced their derivation from olivine-
pyroxene assemblages by reactions at
a lower temperature with their own
alkaline liquid, producing melilite,
monticellite, and biotite.
An alternative explanation, in view
of the experimental studies now re-
ported, would involve the partial
melting of an olivine pyroxenite at
depth, yielding the alkalic liquid, and
the intrusion of the resulting crystal
mush, holding relict clinopyroxene
from the high-pressure environment,
to consolidate as the alnoite assem-
blages at higher, subvolcanic levels
in the crust.
Melting Relations of Some
Calcalkaline Volcanic Rocks
G. M. Brown and J. F. Schairer
Experimental studies on suites of
volcanic rocks from the circumoce-
anic island arcs of the West Indies
and the Solomons have been made to
help explain certain unique chemical,
mineralogical, and field relationships
in these regions. The island arcs are
the sites of recently active and highly
explosive volcanoes. Melting relations
under dry conditions were examined
as a first step. The results are related
to work previously reported by Tilley,
Yoder, and Schairer (Year Books
62-65) on tholeiitic and alkalic rock
suites, and on a calcalkaline suite
from Paricutin, a continental volcano
(Year Book 65). Some melting data
obtained under a PH.,o of 2 kb are
given. The influence of the H20 was
judged particularly significant in
considering the physical conditions
under which these calcalkaline mag-
mas originated.
Characteristic Features of the
Investigated Rock Suites
Calcalkaline rocks are often viewed
as the products, in part, of sialic as-
similation by basalt magmas, despite
the cogent argument advanced by
Osborn (1959) on the role played by
oxygen pressure. The plutonic com-
plexes formed under orogenic condi-
tions pose different problems from
the volcanic, while the latter ought
nowadays to be considered in two
different categories, based on envi-
ronment, petrology, and the chem-
istry of major and minor elements
and of isotopic ratios. These two
categories are in the circumoceanic
island arcs (e.g., West Indies, East
Indies, or Solomons) and the con-
tinental margins (e.g., North Ameri-
can Cascades or Paricutin). The first
group is particularly interesting,
since the field relations suggest the
eruption of magmas from a deep
source where the physical conditions
gave rise to a greater range in silica
contents than found in other oceanic
suites. Olivine and anorthite-rich
blocks are ejected, rhyolites are less
common than in the second group,
and the basalts, andesites, and dacites
are characteristically rich in norma-
tive anorthite and poor in potassium
and titanium (Table 19).
The differences between calcalka-
line and alkalic rocks are well
known (Holmes, 1921, pp. 304-305),
but distinctions drawn between tho-
leiitic and calcalkaline suites de-
pend upon the significance attached
to sialic contamination or the vary-
GEOPHYSICAL LABORATORY
461
>
CN
CO
CN
CN
O
CN
S3
CO
CN
to
to
to
CO
.— CNCOCO'OCOKOCOOOCOO
UOU^i— CO — CNCNO^t'O'—COr-
drNCOiodrNr-'cNddddd
V> r- —
OOCOCOO'OOOrj-iOCOOCNO
CSr-^r-O^;<>O«OOC0^;r-
tNo'cidi-'iori^cNodo
S3 r-
OOCOCNOO^NiOCNOOiOCN
N a «o q r- oo io fs in r- ^r 10 r-_
0>OO^OCNKCNi— 'i-^OOO
H3 «-
OO^tCOCON"— NOCN^-OCO
U3C0iOKr-c0iq^iq,O,O>q>-;
CD N CN IT) O lO
iO —
CN O •— O O O
O
</»
_>.
O
C
<
^f^qi-r-nOvCNar-COTfr-
lo^dtOr-'otdcoodd
S3 —
OOOCOCOCOOOSDCOSDiOO
K^SDC0'-;S3CNC>sSDC>CNiOr—
<> N CN xj O CN S3 CO i CD i—" CD O O
OOO^NOsCO — 'tO — C000
ocoioKr— iq^on^rcNcoq
KCS^XfO<)CNCNCDCDCDCDCD
^r- r-
(N^O^«NOTtOOt- C0C0O
'OCOTl-iOCNiOOvr-Ovr-'OKM
csNcoiod'ococod'-'ddd
IO r-
co"tooo>oNco — ooontoo
SD^fOv^CNCOf—OsiOCNTfTrCNCD
CO CN CO NQ O — — r-' r- r- d 6 d d
TfOMOOO'OK'OOOtCOOa)
00OOOCN — CNO^CO^CN^CNCD
oOf-'t'odri-'r-Vddddd
^" r— i— c—
CNO — OOOCNr-iO o CO »0 CN
f— r- co k q co r- k) co ^j- t^q
KOco<)CD'— cdr— "o r-' odd
Tf CN
o
o
o
o
CO
d
o
W<lLlL^^UZ^IIl-Q.U
—
o
*° "? *2
CN CN Is!
K — O
CO CO O-
CN CO d
CN
CO CN CO
CD CO CN
^f «o >d
CN
CN CO
On CO
00 >0
CO O
>o "O
CN CN
q no
wS CO
•o — —
*-> o —
Ntoti
CO «— «—
oo
«o
o O CM
<* — N
•oo^ON
• ^r ^t cn
d
o
^r a! io
CN CN
' oo d oo
CN
CN
O S3 <*
^ "". ^
• CN «— O
• CN — —
o
rs! c> co
t— CN
' On* O* O
CN r-
oo .— N
^ k "*
xr k cn
CO r- —
CO
CO
xr «o o
CN «— CN
• O Xf CN
• 00 CD K
o\
o
CO SD d
CN Xf
' 00 i- ■ On
CN
N O «0
• 00 O CN
N
r- CD *0
• q sd oo
d
iO ■^■'O
' 00 CN CO
O
— CO
CN
s
o
Z
CO
CN U-) IO
• OO O CN
Ov
r- cn q
• 00 00 IO
CN
1— ' I— ' r— '
' --' «o d
O
CN
^r-r-
O — CN
no q xr
CD cS ^n
co «— <—
CO
"1
NO
CN
CN
xr o"
o
CO
CNJ
xr
CN
odd
rfO^OO
O O CN
o o
CO
00
<>
CO
'
CO
d
o
K
iO
d
d
O Tf IO
oo k ^
d
CO
o
CN O
<o
"""
NO O
O CO K
COO^
— xr d
CN On — CN CN IO N
CO CO XT CO 00 >o «o
d >o i si cn d in Id
CN CN i—
CO — NO XT CN CN N
co xr co co oo co xr
o* is" xf vd d io d
r- CN ,—
OKCTOCMO^
•— Tf r- . vQ 00 OO CO
io cn is! cn d xf d
*- r- CO
— CO
K so
On r_
CO
>o
On
Nj--Q0Jc._>> — *; a
00<Z<Ql05^<
N
O .
£n
NO
—•"On
OJ *~
CQ __«.
Jo
c c
o o
to c
' — '2
w cO_
I -
E c
o o
o E
to _o
^ o
o to
O o
Z £
5 co
Cn °° -5
On -O ■ — -
r- C ««
o
CQ
"73
c
"a
c
D
c
o
O v*
c ^r
:= <o
^J CN
£ -
o *
to _Q
.2 £
"O o
c
^1^0
K
NO
O
£
-a
c
c
o
in
O
#0)
CO
On C
i— <U
u
c.E
£ .^
c
o _
C c • -
S o^
to "5 *:
— ^oo
c
o
£
_o
o
to
i5 -zr *-
j: co
I" "O
e
^ ^-^^^
0 .r
00
</>
O #U
3^
o
.5 ^
O .2
Z oo^:
° ^ >
N-Oi
^^ CN
to ^
w CO
_o —
>- —
a
c E
o 2
.9 -a
U (D
■*^ _**
GO .0
^ U
o — '
CO ^_.
c to
K O
CO X
O
< a
V.. — ,
CO CO
^ N
o o
Q -o
^ X
D
Q
T3 c:
C 3
— O
o -
.5 I
U -
D (1)
. CO
,^ o
-to
CN
CO
CO
»» — — ■*- <1>
Z't
u
G
.a
o o 5
- 0 o
«! J2 -Q
o O) G)
• ~ .— C — "w1 o
±: a a-r- -r ■ —
a; o
S ^
o o
0)
0L<
*- ' J
fl. « °
■2 c .£
<§o
oj cn
c
uo ._
k- u
OJ O
Q
"? .1=
II
O °
^ OJ
u c
•= o
>■■£
Q. *-
2 8.
> x
£ -^
o co
"D —
£ CN
o • — ■
— =
^1 o
^ o
5T§
NO u
CN ««
— ■ *J^
oj o
■*- V)
'u 0
O -Q
"° D.
OJ c
c "Z
*= O
D OJ
OJ .O
■9 »
« .E
lo
co to to to ^
CO •O N CN SD
CN CN
O —
CN CN
-J >
CO IN,
CN CN
462
CARNEGIE INSTITUTION
O
CD
22
21
20
19
18
17
16
15
14
13
12
I I
10
9
8
7
6
5
4
3
2
CALCALKAL1NE SERIES
I— 15 Solomons
16-19 St. Kitts
20-23 St. Lucia
24, SV St. Vincent
Pb-Pa Parfcutin
THOLEIITIC %,\
SERIES \ SK
I 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16 17 18 19 20 21
FeO-f- Fe203
Fig. 64. Plot relating total iron oxides and magnesia for calcalkaline volcanic rocks (unbroken
lines) compared with those for alkalic series of Hawaii, two calcalkaline rocks of Parfcutin, and
tholeiitic series of Hav/aii (and Skaergaard, SK) and Thingmuli, Iceland. Closed symbols represent
rocks studied for melting relations (Tables 19 and 20); circles (1—15), specimens from Solomon
Islands (Stanton and Bell, 1967); triangles (16-19), from St. Kitts, West Indies (Baker, 1963);
squares (20-23), from St. Lucia, West Indies (Tomblin, 1964); diamond (24), from St. Vincent,
West Indies (Robson and Tomblin, 1967). Crosses refer to analyses of St. Vincent plutonic-block
material (Lewis, 1964) including average of four rock analyses (SVR), interstitial basalt scoria
(SVS), average of four similar, separated pyroxene analyses (SVP), and average of four similar,
separated amphibole analyses (SVA). Hawaiian series from Year Book 64, Fig. 3; Parfcutin lavas
from Year Boot 65, Fig. 41, and Thingmuli trend (Tpb picrite basalt-basalts-basaltic andesites-
icelandites-T}. rhyolite) plotted from data by Carmichael (1 964).
GEOPHYSICAL LABORATORY
463
ing effects of oxygen pressures,
acting on basalts of the same initial
composition. It is now shown that
the calcalkaline suites of the West
Indies and the Solomons can be
viewed usefully in relation to the
tholeiitic suite of Skaergaard type ac-
cording to the parameters shown in
Fig. 64. The plots diverge markedly
at basalt compositions and illustrate
the probable effect of increasing
oxygen fugacity in the calcalkaline
basalts, resulting in iron-oxide for-
mation at the site of crystal frac-
tionation. For comparison, the Thing-
muli suite shows an initial tholeiitic
affinity but also a calcalkaline affinity
in the more siliceous rocks, which can
be correlated with the stage when
magnetite microphenocrysts appear
abundantly in the lavas (Carmichael,
1964, pp. 440-444).
High-alumina basaltic, andesitic,
and dacitic lavas and pyroclastics
comprise the bulk of the West Indies
islands (Baker, 1967; Robson, 1967;
Robson and Tomblin, 1967; Tomblin,
1967). Geophysical evidence suggests
intermediate velocity layers (7.1-
7.7 km/sec) between mantle and typ-
ical oceanic material (8.1 and 6.5 km/
sec), which are thicker and rise
closer to the surface beneath the vol-
canic islands (Officer, Ewing, et al.,
1959, p. 107). Plutonic blocks erupted
with the lavas and pyroclastics are
poor in silica (Wager, 1962; Lewis,
1964, 1967), the most siliceous min-
eral being augite (Si02 = 48.7%)
and the others being unusually calcic
plagioclase (An95), magnesian olivine
(Fo79), tschermakitic hornblende
(normative Ne = 9.9%), and Fe-Ti
oxides. Tomblin (1964) recorded
plagioclase cores of An84 in basalts,
An81 in andesites, and An90 in dacites
of Santa Lucia. Anorthites are char-
acteristic of calcalkaline rocks in
Japan (Kuno, 1950; Kawano and
Aoki, 1960) and the South Sandwich
Islands (Dr. J. F. Tomblin). The
23A
WEST INDIES
19 St. KittS
23 St. Lucia
24-27 St. Vincent
FeO + Fe2 03
Fig. 65. Expanded part of Fig. 64 showing
the distinctive trend toward low magnesia and
total iron oxides for the calcalkaline series of
the Solomons and West Indies, from olivine
basalt (24) through basalt (25), andesite (19)
and dacite (23), to rhyodacite (23A). Numbers
1 9 (St. Kitts), 23 (St. Lucia), and 24 (St. Vincent)
are reproduced from Fig. 64, the extra points
here relating to analyzed pairs of rocks (open
symbols) and separated groundmasses (closed
symbols). Numbers 1 9-1 9A from Baker (1963);
23-23A from Tomblin (1964); 24-24A to 27-
27A from unpublished analyses by Dr. J. F.
Tomblin (University of the West Indies, Trinidad,
1 967).
early accumulates of the Solomons
suite are rich in olivine and augite
rather than hornblende and anorthite
(Stanton and Bell, 1967), but the ba-
salts and basaltic andesites plot close
to the West Indies suite in Figs. 64
and 65 and are similarly rich in alu-
mina. In both regions, silica variation
in the lavas is extreme compared with
tholeiitic suites, and the accumulative
rocks are particularly low in silica.
Melting relations in picrite basalts,
basalts, andesites, and dacites were
studied. Specimens were provided by
P. E. Baker, J. D. Bell, J. F. Tomblin,
and R. L. Stanton.
Melting Studies
The analyses of the rocks used are
listed in Table 19, the experimental
results in Table 20, and the plotted
464
CARNEGIE INSTITUTION
TABLE 20. Results of Melting Experiments on the Calcalkaline Volcanic Rocks
of the West Indies and Solomon Islands
(Analyses in Table 1 9. See also Figs. 64 and 66)
Rock Starting Material
Highest Temperatures
of Crystallization
of Major Phases n of Glass
Picrite basalt (NG/367), New Georgia,
Solomons (3S)
Augite-rich picrite basalt (NG/443), New
Georgia, Solomons (5S)
Augite-rich picrite basalt (NG/450), New
Georgia, Solomons (7S)
Olivine-bearing basalt (NG/120/2), New
Georgia, Solomons (9S)
Olivine-bearing basalt cinders (KB/64), St.
Kitts, West Indies (16K)
Hypersthene-andesite pumice (KB/77A), St.
Kitts, West Indies (19K)
Dacitic, glassy matrix from hypersthene
andesite (KB/77B), (19K-A)
Vitrophyric, olivine-bearing basalt (L/83),
St. Lucia, West Indies (20L)
Hypersthene andesite from dome (L/290),
St. Lucia, West Indies (2U)
Biotite-bearing dacite from dome (L/267),
St. Lucia, West Indies (23L)
Olivine-bearing basaltic scoria (27713),
St. Vincent, West Indies (27V)
Ol (1363°), Cpx (1205°), PI (1155°) 1.609
Ol (1287°), Cpx (1195°), PI (1155°) 1.609
Ol (1257°), Cpx (1185°), PI (1155°) 1.590
PI (1213°), Ol (1 1 65°), Cpx (1 1 55°) 1 .571
PI (1280°) 1.595
PI (1 240°), Px (11 80°) 1.540
PI (11 80°), Crist (1150°) 1.498
PI (1245°) 1.585
PI (1255°), Px (11 80°) 1.538
PI (1275°), Px (11 80°) 1.519
PI(1215°),OI (1185°), Cpx (1175°) 1.584
points on Figs. 64 and 65 (filled sym-
bols) and Fig. 66.
The rocks poor in silica and having
a small iron ratio ( [FeO 4- Fe203] /
[FeO + Fe203 + MgO]) could be
run for 1 hour in sealed Pt tubes at
1 atmosphere. Initial tube-bursting,
due to contained water, was overcome
by loading small (20 mg) charges
and flattening the tubes before weld-
ing the tops. For the Saint Kitts and
Santa Lucia basalts, andesites, and
dacites, great difficulties were ex-
perienced in reaching equilibrium.
The usual short runs at temperatures
above 1200°C failed to eliminate re-
sidual grains, chiefly of plagioclase
but occasionally of pyroxene. (All
analyzed ^ 90-mesh powders were
reground mechanically in the same
way for 2 hours under acetone.)
Since these residuals persisted at un-
usually high temperatures, it was de-
cided to eliminate them at 1300°C
(achieved in all cases in 1/2 hour).
Then it was found that the glasses
would not nucleate as low as 900 °C
in 7-day runs. Each long run was
made under reducing conditions,
iron filings in Pt foil being placed,
with the sealed Pt tube containing
the charge, in an evacuated, sealed
silica-glass tube. Some loss of iron to
the Pt tube may have taken place in
the long runs, as is to be expected in
all experiments of this kind (Yoder
and Tilley, 1962, p. 374). For these
rocks, however, the iron content is
lower than in most (Fig. 64), and
for the purpose of achieving nuclea-
tion, even extreme iron loss would
probably not be important in highly
siliceous material such as the ande-
sitic and dacitic glasses. In all low-tem-
perature runs, the retention of glass
or quenching products would seem to
be influenced by the combination of
high Si02 and A1203 and low Na20
+ K20 in the rocks (Table 19) . Equi-
librium was attained eventually by
GEOPHYSICAL LABORATORY
465
making most runs for 6-24 hours (a
few were possible in 2 hours) and
using the sealed silica-tube technique
in order to prevent oxidation of the
charge. This was successful in all
cases for obtaining the liquidus tem-
peratures, residuals being eliminated
in these longer runs and the products
consisting of well-faceted, evenly
distributed crystals, in glass showing
little heterogeneity in refractive in-
dices. Runs at lower temperatures
generally resulted in the appearance
of a second crystalline phase (except
where omitted from Table 20) but
beyond that it was not possible to dis-
tinguish between residual and newly
formed crystals in most cases.
Interpretation of Results
As shown on Fig. 66, the melting
temperatures of Solomons lavas
(3S-9S) show a regular decrease as
the iron ratio increases. Olivine forms
on the liquidus in 3S, 5S, and 7S, but
the liquidus temperatures are sys-
tematically lower than for the less
augite-rich, picritic basalts of the
tholeiitic and alkalic series. Clino-
pyroxene appears at about 1200° C
in all three cases, and plagioclase at
about 1150°C (Table 20). Results ob-
tained on basaltic rocks from the
Solomons and from Saint Kitts, Santa
Lucia, and Saint Vincent in the West
Indies (9S, 16K, 20L, and 24V) are
plotted. Here, great differences are
apparent in the liquidus temperatures
(PI appearing in all cases). In com-
paring results on 9S with 16K or 20L
(Table 19) the reason for the differ-
ences seems to lie in the higher norma-
1450
1400
1350
O 1300
o
£ 1250
Z5
a
a3 1200
Q.
£
HII50
1100
1050
1000
i i i i i i i — i — i — i — i — i — r
\
3S X\
t— i — r
l — l — r
5S \ \ |6K
V ,A
_^\ \ 20L dIL
7SXV24V\D
9S
23L
□
I9K
A
I9K-A
#75
ic
&
J LJ 1 I L
J I I L
J L_J L_J L
035 0.40 045 0.50 0.55 0.60 0.65 0.70 0.75 0-80 0.85 0.90 0.95 1.00
FeO-f Fe203/Mg0 + Fe0+Fe203
Fig. 66. Plot relating liquidus temperatures and iron enrichment. Points for the calcalkaline rocks
have the same symbols and numbers as in Fig. 64 (see also Tables 19 and 20). The letters refer
to specimens from the Solomons (S), St. Kitts (K), St. Lucia (L), and St. Vincent (V). The unbroken
line is drawn tentatively and indicates a possible relationship in the absence of feldspar ac-
cumulation. The small-dash line connects augite-rich picrite basalts from the Solomons, the change
of slope representing an estimated stage at which plagioclase appears first on the liquidus. For
comparison, the tholeiitic and alkalic trends (large-dash lines) are reproduced from Year Book 65,
Fig. 40.
466 CARNEGIE INSTITUTION
five An and lower Ab and Or of the trast to the tholeiitic and alkalic
latter two rocks, but the same can- rocks, originated. The source might
not be said of 24V. For the andesites well lie within the intermediate veloc-
(21L, 19K) and dacites (23L, 19K-A) ity layers between the oceanic mantle
(Fig. 66) the liquidus temperatures and crust, the source material being
are high and variable and not directly possibly hydrous (as suggested also
related to iron ratio. The iron ratio by O'Hara, 1965, p. 37, and by Don-
is generally significant in comparing nelly and Rogers, 1967). Under mod-
volcanic rock suites but fails to have erate water pressures the basalts gen-
the same significance for the calcal- erated would tend to be aluminous
kaline rocks as for the tholeiitic and (Yoder and Tilley, 1962, Fig. 28),
alkalic rocks. and if the conditions favored precipi-
It is most likely, however, from the tation of a spinel and nepheline-nor-
data for 24V, 9S, and the ground- mative amphibole in addition to
mass dacitic material shown as olivine and a highly calcic plagioclase,
19K-A, that these calcalkaline rocks in higher-level magma chambers, a
have higher dry-melting temperatures silica-enriched suite of derivative liq-
than the other volcanics. uids could ensue. Although these ex-
Whatever argument is presented to treme conditions were not apparently
relate chemistry and melting tern- realized in the Solomons, they were in
peratures we must ultimately be faced the West Indies, and the ejected
with the implications in regard to plutonic blocks have the appropriate
the crystallization history of these mineralogy. Not only are the water
calcalkaline rocks. We would not pro- pressures important in the formation
pose, for example, that because of of the amphibole, but also the oxygen
some of the data presented in Fig. 66, fugacities are critical in formation of
the natural andesitic liquids were magnetite and generation of the
initially at a higher temperature than andesites (Fudali, 1965) and sili-
the basaltic, and the dacitic higher ceous dacites.
than the andesitic. The effects of un- The plutonic blocks in the West
resorbed xenocrysts and xenocrysts Indies are often layered and probably
or phenocrysts resorbed under cer- formed in a magma chamber filled
tain physical conditions, and variable with mantle-generated basalt. The
effects of water pressure on the differ- basalt interstitial to the large crystals
ent rock systems would all result in a is vesicular, and although Morey's
misleading correlation between chem- hypothesis on explosive volcanism is
istry and dry-melting temperatures, not generally applicable (Yoder, Year
in relation to the actual magmatic Book 6J>, p. 89), the large number of
conditions. It may be argued that for mineral phases present in these
these reasons alone we should have blocks could conceivably have led to
concentrated on the behavior of univariant conditions in this case.
aphyric rocks under pressure of It is significant, however, that the
water, but evidence from the dry ex- blocks are not erupted with basalt
periments is needed first if compari- magma but with andesitic and dacitic
sons are to be made with the other tuffs, suggesting that vesiculation in
rock suites. basalt magma was possible without
We consider that the results shown eruption and that the magma cham-
in Fig. 66 may have a significant im- bers were disrupted only at later
plication related to the physical con- fractionation stages. This suggests a
ditions under which the calcalkaline buildup of water pressure during
rocks of oceanic island arcs, in con- crystal fractionation and therefore a
GEOPHYSICAL LABORATORY
467
marked control by this component on
the crystal-liquid equilibria. It is un-
likely that the basalt magmas were
saturated with water at pressures
greater than about 2 kb; otherwise
plagioclase would have been sup-
pressed on the liquidus (Yoder and
Tilley, 1962, Fig. 28), although the
conditions must have been appropri-
ate for amphibole to form as well.
Water loss during eruptions must
have complicated the phase relations
in these volcanic regions also, as in-
dicated by the presence of hyper-
sthene phenocrysts in many andesites
and of amphibole and biotite in
dacites.
Clearly the whole calcalkaline prob-
lem now requires experimental study
under hydrothermal conditions. Pre-
liminary runs at 2 kb water pressure
have been made by H. S. Yoder,
Jr., in an internally heated, argon-
pressure apparatus, on 19K and 21L
(andesite) and 23L (dacite) . The
liquidus temperatures were reduced
by about 200 °C. Values observed
(19K, 1010°C; 21L, 1050°C; 23L,
1040° C) seem more realistic (e.g.,
23L now below 21L) . Anomalies were
still apparent, however; the siliceous
dacite (23L) melted at a slightly
higher temperature than the Saint
Kitts andesite (19K).
The Join Nepheline-Diopside-
Anorthite and Its Relation to
Alkali Basalt Fractionation
J. F. Schairer, C. E. Tilley, and
G. M. Brown
The simplified basalt tetrahedron
(Ne-Fo-Si02-Di) of Yoder and Tilley
(1962, Fig. 1) and the expanded
basalt tetrahedron of Schairer and
Yoder (Year Book 63, Fig. 1) showed
the relations of minerals in the syn-
thetic analogues of tholeiites, alkali
basalts, and a group of other alkali
rocks in which the feldspar is albite
or a highly sodic plagioclase. Since
most basaltic rocks carry a feldspar
with a moderate to large anorthite
content, it is clearly necessary to
ascertain phase-equilibrium relations
on synthetic analogues in which
anorthite plays the dominant role. In
order to do this the join nepheline-
diopside-anorthite was selected for
study, and the experimental results of
this investigation are presented here.
To show how the join Ne-Di-An fits
into the general scheme of anorthite-
hO
Fig. 67. The system Ne-Di-An, forming the
common face of the coupled tetrahedra, with
forsterite and akermanite, respectively, as the
apex phases of the two individual tetrahedra.
Phase relations in the system Ne-Ak-Di (Schairer
and Yoder, Year Book 63, Fig. 2), the back
face of the upper tetrahedron showing the
melilite field, and in the systems Di-Fo-An (Os-
born and Tait, 1 942) and Ne-Fo-An, the two
side faces of the lower tetrahedron showing the
spinel fields, are indicated. The fields of forster-
ite, a b c d on the back face and c e d in the
join Ne-Di-An, extend to form a smail volume
within the upper tetrahedron.
46$
CARNEGIE INSTITUTION
(CaO)
Fig. 68. The plane Ne-Di-An depicted in the
tetrahedron MgO-AI203-CaO-Na20 (silica omit-
ted). The positions of forsterite (Fo), spinel (Sp),
wollastonite (Wo), and the plane of the melilite
system akermanite (Ak)-soda melilite [StA]-
gehlenite (Ge) are indicated relative to the
Ne-Di-An plane.
bearing basaltic rock-forming min-
eral assemblages, two figures are em-
ployed. The location of this join as the
common plane of the coupled tetra-
hedra Fo-Ne-Di-An-Ak is depicted in
Fig. 67. Its position in the tetra-
hedron MgO-Al203-CaO-Na20 (silica
omitted) is shown in Fig. 68. In this
figure the positions of forsterite,
spinel, wollastonite, and the plane of
the melilite system akermanite-soda
melilite-gehlenite are indicated rela-
tive to the join Ne-Di-An.
The phase-equilibrium diagram for
the join Ne-Di-An is given in Fig. 69.
Thirty-four compositions, shown as
filled circles, were selected for study
by the method of quenching at 1-at-
mosphere pressure. Five piercing
points of univariant lines lie in the
join. If olivine (forsterite), spinel, or
/?A12CK appears during crystallization,
the composition of the liquid phase
must leave the join because the
composition of the solid phase sep-
arating does not lie in the join. All
compositions studied yield melilite
as a solid phase at some stage of the
crystallization. Attention is called
to the compositions of several of the
solid phases, which are solid solutions
with compositions that do not lie in
the join. The diopside is an aluminous
diopside, the anorthite is not pure
anorthite but must be a slightly al-
bitic plagioclase, the nepheline is a
solid solution with at least a small
content of anorthite and probably
albite, and the melilite is a solid solu-
tion of akermanite and soda melilite
(NaCaAlSi207) with a small gehlenite
content also. Because of solid solution,
compositions studied in the join lying
close to the nepheline-diopside side
line of the join failed to show anor-
thite as a solid phase. The piercing
points A, By and C lie close to each
other in temperature and composition
and must lie close to three quaternary
invariant points (F', C, and L) . The
flow sheet of Schairer and Yoder
(Year Book 63, p. 72) has been modi-
fied for rocks containing anorthite-
rich plagioclase, and is given here as
Fig. 70.
From an examination of Fig. 70 it
should be apparent that the two tem-
perature maxima, one in XF' and the
other in G'a, separate the quaternary
invariant points X, Y, and G' from
the other points that are nepheline
bearing. The temperature maximum
at c in YGf separates the points X
and Y (anorthite-bearing tholeiitic
basalts) from G', an extreme composi-
tion probably represented only in
metamorphic rocks.
Compositions that lie in the spinel
field (Fig. 69) and those compositions
studied that lie near the Ne-An side
line during crystallization proceed to
the quaternary invariant point H
(Fig. 70) at 1163° ± 2°C, where they
lose spinel by reaction, and at a tem-
perature only a few degrees lower
are joined by either forsterite or
melilite or both. They lose forsterite
GEOPHYSICAL LABORATORY
469
Diopside
r^CoO.Mg0.2Si02
I39J.5°
A = I 165*5*
B= 1160*2°
C= 1168*2°
D= 1263*3°
E= 1298*3°
I274±2°
1526±2°/ °
Nepheline 10
Na2O.AI203.2Si02
30
40 50 60
Weight per cent
70
80
I553±2=
90 Anorthite
CaO.A!2G3.2Si02
Fig. 69. Phase equilibrium diagram for the join nepheline-diopside-anorthite.
by reaction at F', however, and reach
the quaternary eutectic point L (Fig.
70). Note the very close proximity
in temperature and composition of H
(1163° ± 2°C), F' (1155° ± 3°C),
C (1152° ± 3°C), and L (1148° ±
3°C).
The assemblages developed in the
crystallization of this join have their
analogues in the more complex natu-
ral alkali basaltic rock suites. The
rock-type analogues are nepheline ba-
sanites represented at the quaternary
invariant point Ff (ne, di, an, fo),
and with loss of olivine, nepheline
tephrites (ne, di, an) from liquids
moving from F' to L; and melilite
olivine nephelinites represented at
the quaternary invariant point C
(ne, di, mel, fo), and with loss of
olivine, melilite nephelinites (ne, di,
mel) from liquids moving from C
to L. The olivine nephelinities (ne, di,
fo) between F' and C separate assem-
blages carrying anorthite and meli-
lite, respectively. The quaternary eu-
tectic L (ne, di, mel, an) is not rep-
resented among igneous rocks except
in the nonequilibrium assemblages
provided by limestone contamination
of olivine dolerite, such as those re-
ported from Scawt Hill, County
Antrim, Ireland (Tilley and Har-
wood, 1931, p. 464, and Plate XVII,
Figs. 4 and 5) . Assemblages recorded
there consist essentially of titanau-
gite, basic plagioclase, melilite
( Ak46Sm35Ge19) , and nepheline, the
470
CARNEGIE INSTITUTION
(Ne-Di-An)
cuts 1298*3'
(Ne-Fo-An) /S-AI2O3
»••«« NEss (
> SP /
NEss
si' AN
cuts
k
7£~7SP / I SP S
?temp?/3-AI203-(Ne-Di-An) +
NESS
AN
SP
(Ne-Fo-Di)
cuts
(Fo-An-Si02) / NEss
1260*5° .* SP
V
(Fo-Di-Si02)
1386*1°
FO
Diss
(Fo-Diop-An)
p0 1270*5° Y F0
rv a xDiss(NE-Di-An)
(En-D,An) \^±?
PROTO
1250*10°
V
(Fo-An-Si02) PR0T0
1222^5° AN
• —
TRIDY
■>
II50*I0C
Diss
TRIDY
AN
(Wo-An-Si02) TRIDY
1170*5° AN
I24I5*-5°F0/||63±2<
FO
PROTO
AN
/3-AI203
^'A,2°3 (Ne-Fo-An)
. "N , cuts
<-zrz >"| _
temp:
cuts I263±3°
NEss
AN
-SP
^|n^2°^An-Di)
f-LNEss .l3'7±5'
AN
SP
FO
NEss
AN
J
1215*5°
I2I3*5{
FO
(NE-DI-AN)Dlss
cuts 1160*5°
_.(Ne-Wo-Di)cuts
(Ne-AK-Di)cuts
Diss
MEL NEss (NE-AK-DI)
v
X 1262*5° AN
^•^FO h
Diss
PROTO
AN
PROTO
Diss
AN
PROTO
. Diss
FO
v r TRIDY (Fo-Di-SiOo)
AC
V
v 1 PROTO
Diss
TRIDY
.AN
(Fo-An-Si02)
1200*
MEL cuts 1178*5'
(Ne-Wo-Di)
cuts.
1193*3°
WO
1236*5°
A
WO' TG' Diss
^TRIDY
AN
WO
TRIDY
Diss
WO
1320*5°
'(Wo-Di-Si02)
Fig. 70. Flow sheet for basaltic liquids where anorthite [not highly albitic plagioclase) is the
feldspar mineral. Abbreviations for crystalline solid phases: forsterite (FO), protoenstatites
(PROTO), diopsides (Diss), anorthite (An), tridymite (TRIDY), wollastonite (WO), beta alumina
(/9-AI2O3), nephelines (NEss), spinel (SP), melilites (MEL), akermanite (Ak), enstatite (En).
melilite arising at the expense of ti-
tanaugite and plagioclase.
An olivine reaction clearly plays a
very significant role in these frac-
tionation processes. From liquids in
the join nepheline-diopside-anorthite,
olivine reacts with liquid to form
diopside, and for liquids in the diop-
side-nepheline side line the final
assemblages consolidating consist of
di, ne, and mel, with all olivine re-
acted out at temperatures below about
GEOPHYSICAL LABORATORY 471
1120°C (Schairer, Yagi, and Yoder, rocks. One of the major magma
Year Book 61, p. 97, Fig. 29). trends, originating with the olivine-
Evidence of olivine resorption; melilite nephelinites, depends on the
should therefore be expected to be stability of akermanite. The impor-
seen in the natural nepheline-bearing tant rock kimberlite is often described
suites, which are analogues of the as derived at high pressure from a
synthetic systems. Olivines mantled hydrous magma rich in melilite. It
by fine aggregates of clinopyroxene is desirable, therefore, to have full
observed in some of these rocks are understanding of the limits of stabil-
probably indicative of this olivine re- ity imposed on the rocks that contain
sorption. The reaction is the analogue a melilite rich in the akermanite com-
of the better known olivine resorption ponent.
characteristic of tholeiitic suites, The position of akermanite relative
though there the pyroxene product is to its related phases is shown in Fig.
a low-lime type (hypersthene, pigeo- 71. The stability field of akermanite
nite, or subcalcic augite). The meli- is given in Fig. 72, based on gas-
lite phase crystallizing from runs in media apparatus experiments, usually
the nepheline-diopside-anorthite join made in quadruplicate with synthetic
characteristically has a very low akermanite, its glass, and relevant
negative birefringence with ultrablue synthetic breakdown assemblages of
interference tints or is optically iso- akermanite composition as starting
tropic. In the less anorthite-rich com- materials in the presence of an ex-
positions at least, these isotropic cess of water. The "univariant"
members of the melilite phase have curves bounding akermanite repre-
an index of refraction very close to sent estimates of the midpoints of
1.630, indicating a composition near narrow regions of indifference to re-
Ak65Sm25Ge10 (cf. Schairer, Yoder, action or divariant regions resulting
and Tilley, Year Book 6h, pp. 99-100, from complex solid solution. In the
and unpublished work) . Whether mel- latter case, for example, akermanite
ilites with higher gehlenite content is believed to crystallize as a solid so-
crystallize in the more anorthite-rich lution close to but not on the end
compositions of the join has not yet member composition Ca2MgSi207.
been determined. Throughout the region of stability of
The character of the melilite-basic akermanite both diopside and wollas-
plagioclase relationship in synthetic tonite in variable proportions appear
systems remains a problem requiring with akermanite synthesized from
continued experimental study, and the starting materials of end-member
investigation of the systems albite- bulk composition. At the higher pres-
anorthite with akermanite and other sures diopside predominates, indicat-
members of the melilite group (Sm, ing that the akermanite solid solution
Ge) may point the way to an ultimate lies near the join akermanite-mer-
solution of the behavior of natural as- winite.
semblages. The beginning-of-melting curve,
which is independent of the water
Akermanite and Related Melilite- content, appears to be of a simple
Bearing Assemblages nature ; however, quenching products
H. S. Yoder, Jr. obscure the presence, if any, of small
A1 .. a. ,' „ amounts of possible incongruent melt-
Akermanite Stability Curve ing phases> eg> merwinite or diop.
Akermanite is the primary com- side. The quenching products are
ponent of the melilites of igneous either wollastonite + monticellite or
472
CARNEGIE INSTITUTION
CaO-
Weight per cent
Fig. 71 . Plot of the CaO-MgO-Si02 system showing the location of akermanite and its asso-
ciated phases.
diopside + merwinite and are limited
to their respective regions bounded by
the metastable extension of the reac-
tion curve relating those two assem-
blages. In the field labeled diopside +
liquid + gas, clear, faceted diopside
crystals are easily distinguished in
a radiating groundmass of quench
merwinite speckled with blebs of
diopside, irregular voids, and bubbles.
Kushiro (Year Book 63, p. 85) has
shown that merwinite, not diopside,
is the incongruent melting product at
high pressure in the absence of water.
The invariant point at A, 750°C and
6.3 kb, involving Wo, Mo, Mer, Di,
and Ak, is the origin of five univari-
ant curves. A schematic presentation
is given in Fig. 73, wherein some
possible solid solutions were neglected
and the most probable composition of
akermanite near, but not on, the com-
position Ca2MgSi207 was assumed.
The two reaction curves, indicated by
the absent phase in parentheses,
which were not realized because the
bulk compositions investigated were
exactly on the join Ca2MgSi207-H20,
are (Wo) and (Di). If the composi-
tion of akermanite lies on the join
Woss-MoR8 or the join DiKS-MerHfi, the
curves (Di) and (Mer) or (Wo) and
(Mo) become identical, respectively.
The invariant point at B, 1065 °C
and 10.2 kb, involving Di, Mer, L, G,
and Ak is also the origin of five uni-
GEOPHYSICAL LABORATORY
473
1 1 -
o —
9 —
8 -
o 7
o
1 1 j 1
1
1 1*1"
1 '
1 1
1 ' 1 ' 1
:
—
1 J— '
Diopside + Liquid + Gas
—
B
\0 1
1 1
O
- a
\ \
—
_ Diopside +
a
a/x
\ 1
\ 1
\ 1
Merwinite
1 1
\ ?
+ Gas
\i
—
\ i
\7
\ i
—
\
a
4o
>S p_.
—
2$\o
X
Liquid + Gas
—
— A A A A A A/ x
X
XX X
X X
po
—
_ Wollastonite + / *
A 1
Monticellite a \
—
+ Gas
Z^Ax
Akermanite + Gas
—
Ai x
x\p
—
. 1 . /I i
I
, I.I.
1 .
I.I. >^
I
5 —
I —
600 700 800 900 1000 1100
Temperature, °C
1200
1300
1400
Fig. 72. Pressure-temperature diagram for the composition Ca2MgSi207/ represented principally
by an akermanite on or near that composition, in the presence of an excess of water. The in-
variant points A and 8 are analyzed in Figs. 73 and 74, respectively. The seven smaller sym-
bols between 2 and 4 kb are the bracket runs of Harker and Tuttle (1956). The curve presented
here in that region is consistent with their results.
variant curves. A schematic presen- cause the akermanite solid solution
tation is given in Fig. 74, again neg- may melt congruently so that the
lecting solid solutions. The reaction curve is identical with (Di), and
curve (Mer) was not realized be- suitable bulk compositions were not
474
CARNEGIE INSTITUTION
investigated. The curve (G) was also
investigated under anhydrous con-
ditions and is now presented.
Akermanite -f Forsterite Reaction
The occurrence of an olivine with a
melilite places further restrictions on
the stability of the melilite-bearing
rocks. The reaction Ak + Fo -> Di
-:- Mo outlined by Walter (1963) was
extended to higher pressures, and its
termination was defined with the use
of a crystallized glass of the composi-
tion Di2oMoao (see Fig. 71). In Fig.
75 are presented the preliminary
data obtained in the presence of
excess water below the beginning of
melting of akermanite and under
anhydrous conditions above the be-
ginning of melting of akermanite
under hydrous conditions. The in-
variant point C, 1095°C and 10.6 kb,
is the origin of five univariant curves,
and these are illustrated schemati-
cally in Fig. 76. The univariant points
A (Fig. 73) and C (Fig. 76) are con-
nected through the reaction Ak -> Di
— Mer. In addition, the preliminary
data of Kushiro and Yoder (Year
Book 63, p. 82) and Kushiro {Year
(Wo)
(Mo)
(Diy(Merr7
Fig. 73. Sequence of univariant curves about
the invariant point A (Fig. 72). Heavy curves de-
termined; light curves theoretical assuming a
general composition for akermanite. Parentheses
indicate absent phase. Ak, akermanite; Di,
diopside; Mer, merwinite; Mo, monticellite; Wo,
wollastonite.
Fig. 74. Sequence of univariant curves about
the invariant point B (Fig. 72). Abbreviations
and symbols as in Fig. 73. G, gas; I, liquid.
Book 63, p. 85) have been supple-
mented by new data obtained from a
gas-media apparatus, and the result-
ing curve location corrections con-
sistent with most of the previous
data, obtained in solid-media appara-
tus, are included.
A particularly striking field illus-
tration of the low-temperature and
low-pressure breakdown of akerma-
nitic melilite, as well as its reaction
with olivine, has been described by
Willemse and Bensch (1964).
Graphic intergrowths of monticellite
and wollastonite, indicating the
breakdown of akermanite, and the
separation of olivine and akermanite
by intergrowths of clinopyroxene and
monticellite, indicating retrograde
reaction, are illustrated in inclusions
of carbonate rocks in the gabbro and
norite of the Bushveld complex. Also
of interest is the concomitant forma-
tion of an intergrowth involving
kalsilite and K-feldspar with another
unidentified phase or phases. The
identified phases may represent the
low-temperature breakdown of leucite
GEOPHYSICAL LABORATORY
475
o
-Q
O
a- 5
Di + Mer + Fo
A
A
A
A
Di + Mo
ss
600
700
800
1000 1 100 1200
Temperature, °C
500
1600
Fig. 75. Pressure-temperature diagram for the composition diopside 20, monticellite 80 by
weight, in the presence of an excess of water beiow the beginning of melting of the water-
saturated system and without water above. Fo, forsterite. Small symbols indicate runs of Kushiro
and Yoder (Year Book 63) in solid-media apparatus.
( Scarf e, Luth, and Tuttle, 1966, p.
728), which may itself have formed
as a dehydration product of phlogo-
pitic or eastonitic mica. Willemse and
Bensch already noted, on the basis of
the work of Walter (1963), that the
carbonate inclusions had been heated
to at least above 900 °C, a temperature
still below the solidus of the norite.
It is now possible to suggest that the
inclusions were cooled sufficiently
slowly for retrograde reaction at
depths less than about 20 km; thus
the base of the Bushveld igneous com-
476
CARNEGIE INSTITUTION
(Mer)
Fig. 76. Sequence of univariant curves about
the invariant point C (Fig. 75). Abbreviations
and symbols as in Fig. 73. Fo, forsterite.
plex must have been at a depth less
than 20 km.
In applying the observations sum-
marized in Figs. 72 and 75, the fol-
lowing generalizations may be ap-
plied to akermanitic melilite-bearing
rocks.
1. In the presence of an excess of
water akermanite is restricted to
depths equivalent to those within an
average continental crust.
2. Olivine melilite nephelinites and
monticellite alnoites are apparently
restricted to depths equivalent to
those in a continental crust, particu-
larly in the presence of an excess of
water.
3. Monticellite olivine nephelinite
probably bears a low-temperature
relationship to both olivine melilite
nephelinite (the so-called melilite
basalt) and monticellite alnoite (see
Tilley, Yoder, and Schairer, this re-
port) .
4. The wollastonite-monticellite as-
semblage of inclusions and contact
aureoles is confined to relatively low-
pressure environments.
5. Merwinite may be an important
mineral in the mantle, particularly in
calcium-rich and silica-poor prov-
inces.
6. The restriction of olivine meli-
lite ''basalt" to depths equivalent to
those within a continental crust in the
presence of an excess of water casts
some doubt on the concept that the
magmatic portion of kimberlite is
hydrothermally altered melilite basalt
magma.
7. Derivation of rocks containing
akermanite-rich melilites at the high
pressure existing in the region of
magma generation, where pure aker-
manite is not stable, will probably de-
pend on the presence of the soda meli-
lite component (see Kushiro, Year
Book 63, p. 92).
Incompatibility of Akermanite and
Albite
The incompatibility of akermanite
and albite, demonstrated at 1 atmos-
phere by Schairer and Yoder (Year
Book 63, pp. 65-74), has great im-
portance in nature, where melilite
and plagioclase form from two inde-
pendent magma series. In seeking
possible relationships between these
two major magma series, it was con-
sidered desirable to determine the in-
fluence of other conditions and other
melilite molecules. Preliminary ex-
periments at 10 kb demonstrate that
albite and akermanite remain incom-
patible, and it appears that this in-
compatibility persists within the
range of stability of pure akermanite.
On the other hand, in other ex-
ploratory experiments this year al-
bite and soda melilite were found to
be stable at the elevated pressures
where soda melilite is stable (Yoder,
Year Book 63, p. 88) up to 10 kb.
Furthermore, it has been known that
anorthite and akermanite are stable
at atmospheric pressure (de Wys and
Foster, 1956) . It is now necessary to
ascertain the limits of solid solution
of plagioclase that coexists with aker-
manite, i.e., the system akermanite-
albite-anorthite, as well as the limits
of solid solution of melilite that co-
exists with albite at high pressures,
i.e., the system akermanite-soda meli-
GEOPHYSICAL LABORATORY
477
Si02
Al7l3±5a
I8±3C
NaAISi308
5+2°
Weight per cent
CaAI2Si2Cs
Fig. 77. The system albite-anorthite-silica at 1 atmosphere replotted from the data of Schairer
(1957).
lite-albite. These studies may lead to
an understanding of the relationship
between melilite olivine nephelinites
and the basanites.
Albite-Anorthite-Quartz- Water
at 5 KB
H. S. Yoder, Jr.
Most granitic rocks contain the
anorthite component, and it is desir-
able to know the effects of that com-
ponent on the simplified granite sys-
tem Or-Ab-Qz-H20. The first step to-
ward that end was taken by Yoder,
Stewart, and Smith {Year Book 56,
pp. 206-214) in a study of Ab-An-Or-
H20 at 5 kb and by Stewart {Year
Book 56, pp. 214-216; 1958, p. 1648)
in a study of the An-Qz-H20 system
at 2 and 5 kb. Two compositions in
the Ab-Qz-H20 system have since
been studied on or near 5 kb by Luth,
Jahns, and Tuttle (1964, pp. 761,
763). A preliminary equilibrium dia-
gram for Ab-An-Qz at 1 atmosphere
has been determined by Schairer
(1957, p. 232) as part of the larger
system An-Ne-Qz. His data have been
replotted and are presented here as
Fig. 77.* Reconnaissance runs have
* The diagram may serve as a teaching
example of a type of ternary system having
a simple compound and a continuous series
of solid solutions. The system diopside-
albite-anorthite (Bowen, 1915), now com-
monly used as such an example, has been
demonstrated to be pseudoternary involving
complex solid solutions in diopside that have
not as yet been fully identified (see Osborn,
1942; Schairer and Yoder, I960).
478
CARNEGIE INSTITUTION
SiO
Ph2o=5 kb
748
NaAISi30(
LOW-QUARTZ
\
Weigh! per cent
1234°
CaAI2Si208
Fig. 78. Preliminary projection of the system albite-anorthite-quartz-water at 5 kb onto the
anhydrous plane.
been made on Ab-An-Qz-H20 at 5 kb
using Schairer's glasses and crystal-
lized glasses as starting materials, and
the results are presented in Fig. 78.
The significant observations are
that the presence of anorthite pro-
duces a relatively small increase in
temperature at which coprecipitation
of plagioclase and quartz occurs in
the simplified granite system and that
the quartz content of the liquids de-
creases with fractionation. Of special
interest is the narrow temperature in-
terval between the initial coprecipita-
tion of plagioclase and quartz and the
solidus; a maximum of 20° was ob-
served under equilibrium conditions
involving excess water. The interval
would increase if there were insuffi-
cient water to saturate the liquid.
The observations reported here are
in accord with the predictions of
Bateman et al. (1963). They noted
that the anorthite content of granitic
rocks of the Sierra Nevada batholith
appeared to have been controlled ini-
tially by a fractionation curve in the
plagioclase field, then primarily by
the plagioclase-quartz surface until
the liquid reached its common bound-
ary with alkali feldspar, and ulti-
mately by the effective "eutectic" of
the granite system.
GEOPHYSICAL LABORATORY
479
Effect of Pressure on the
Boundary Curve in the System
Diopside-Albite-Anorthite
D. H. Lindsley and R. F. Emslie
Yoder (Year Book 53, pp. 106-
107) suggested that high water pres-
sures in the diopside-albite-anorthite-
water system might move the water-
saturated diopside-plagioclase bound-
ary curve toward plagioclase. Accord-
ingly, simplified basaltic compositions,
which lie near the boundary curve
for low pressure (Bowen, 1915, p.
167), would have pyroxene as the
primary phase at high water pres-
sures; crystallization and removal
of this pyroxene might drive the
composition of the residual liquid
toward that of a simplified gabbroic
anorthosite. Yoder suggested that
this mechanism might explain the
genesis of anorthositic magmas.
Clark, Schairer, and de Neufville
(Year Book 61, p. 68) demonstrated
that high total (dry) pressures also
shift the diopside-anorthite "eutec-
tic" toward anorthite. Because many
anorthositic intrusions show no inde-
pendent evidence of high water pres-
sures, some workers have recently
suggested that high total pressures
might also cause a shift in the diop-
side-plagioclase boundary curve (e.g.,
Philpotts, 1966, pp. 54-55; Lindsley,
1967). Green (1966) has argued,
however, that removal of aluminous
pyroxene at high pressure would be
unlikely to enrich the residual liquid
greatly in components of intermediate
plagioclase. Because high-pressure
data for the join diopside-anorthite
are already available, and because the
diopside-albite "eutectic" lies close to
albite even at 1 atmosphere (Schairer
and Yoder, 1960, p. 279), high-pres-
sure experiments on one carefully
chosen composition in the diopside-
albite-anorthite system should serve
to test the boundary-shift hypothesis.
We chose the composition Ab32.6
An32.eDi34.8 (mole %; Ab34An36Di30,
wt %); see point A, Fig. 79. This
duplicates one of Bowen's composi-
tions and lies within the plagioclase
field with a liquidus temperature of
1298° C at 1 atmosphere (Bowen,
1915, p. 167). This composition is
intermediate between haplobasaltic
and simplified gabbroic anorthosite
compositions within the Di-Ab-An
system ; and for a shift in the bound-
ary curve to be an important petro-
logical mechanism, the boundary
curve should pass through point A at
relatively low pressures. In fact,
point A does not lie in the primary
field of pyroxene until pressures
greater than 15 ± 1 kb are achieved.
The probable form of the boundary
curve at 15 ± 1 kb is shown in Fig.
79 as a heavy dashed line. It appears
that considerably higher pressures
would be required to drive the bound-
CaMgSi206
NaAISi308
Weight per cent
15 kb
20 kb
Clark et al
CaAI2Si208
Fig. 79. Relations in the join CaMgSi2C>6-
NaAISi308-CaAI2Si208 (Di-Ab-An). One-atmos-
phere boundary curve after Bowen (1915). Ap-
proximate boundary curve at 15 kb based on
interpolation from the CaMgSi206-CaAl2Si20s
data of Clark, Schairer, and de Neufville [Year
Book 67, p. 67) and experiments on the compo-
sition A. At pressures below 15 kb, A lies in
the primary phase field of plagioclase; above
15 kb, A lies in the primary phase field of di-
opsidic pyroxene. Stippled area shows ap-
proximate range of simplified gabbroic anortho-
site.
480
CARNEGIE INSTITUTION
ary curve through compositions cor-
responding to gabbroic anorthosites
(stippled pattern in Fig. 79) — pres-
sures at which intermediate pla-
gioclase is unlikely to be stable in
complex natural systems (e.g., Green,
1966, p. 220). We therefore conclude
that although the shift of the bound-
ary curve with increasing pressure in
the Di-Ab-An system is in the right
direction, the magnitude of the shift
in the absence of water is too small to
make this a likely mechanism for the
generation of gabbroic anorthosites.
The Influence of Pressure on the
Composition of Eutectic Liquids
in the Binary Systems Sanidine-
SlLICA AND ALBITE-SlLICA
W. C. Luth
No data have been available on the
composition of the eutectic liquids in
either of the two binary feldspar-
silica systems as a function of pres-
sure. It has been implicit in many
applications of experimental studies
in the synthetic system albite-sani-
dine-silica-water that pressure either
has no effect on, or causes only minor
changes in, the composition relations
of the dry system albite-sanidine-
silica.
Equilibria in the dry systems pro-
vide important limiting conditions
regarding the crystallization of
water-undersaturated granitic mag-
mas in the earth's
quently, preliminary
have been performed
systems albite-silica
silica at pressures between 8 and 20
kb. The pressure-temperature coor-
dinates for the eutectic reactions
sanidine + quartz = liquid and albite
quartz = liquid, and the composi-
tion of the eutectic liquids have been
determined over this pressure range.
Single-stage, solid-media pressure
apparatus (Boyd and England, 1963)
was employed in this study, and the
crust. Conse-
experiments
in the binary
and sanidine-
techniques used are similar to those
described by Lindsley (1966) and
Bell and Roseboom (Year Book 6Jf) .
The starting materials are those used
by Luth, Jahns, and Tuttle (1964),
and the method of preparation has
been described by Luth and Ingamells
(1965). The starting materials were
placed in either carbon or platinum
capsules and were carefully dried at
temperatures of 900 °C (for crystal-
line material) or 1100 °C (for glass
or gel starting material). The start-
ing materials used for the determina-
tion of the P-T coordinates of the
solidus reactions (Fig. 80) were crys-
talline assemblages of the feldspar
20-
o
O
Q_
15 -
10-
5 -
1 1
i
//
//
r^v-^A^^J ~
-
/ /
/ /
/ t^3
/ /
(1) /
(2)
yk&
i i
i
IOOO
I IOO 1200
Temperature ,°C
1300
Fig. 80. Pressure-temperature curves denot-
ing solidus reactions in the systems KAISi30s-
Si02 (1) and NaAISi308-Si02 (2). The left and
right margins of each bracket represent the
nominal P-T conditions of each pair of experi-
ments. Thus each symbol represents two experi-
ments. The estimated probable error of each
experiment is ±10° and ±0.2 kb. Completely
crystalline starting materials were used.
GEOPHYSICAL LABORATORY 481
(high sanidine or high albite) and serburg, Heard, and Newton (1962)
quartz. For each of the systems stud- have determined the effect of pres-
ied the solidus was located within sure on the quartz-tridymite inver-
25 °C. On the high-temperature side sion. On the basis of the experimen-
of the bracket only quartz or feldspar, tally determined slope of this P-T
but not both, was observed with glass curve, 0.26°C/bar, invariant points
when either glass or crystalline as- in the systems albite-silica and sani-
semblages were used as starting ma- dine-silica may be postulated at pres-
terial. The low-temperature side of sures and temperatures of 1 ± 0.3
the bracket was obtained by using kb, 1060° ± 10 °C, where albite,
glass starting material and noting the quartz, tridymite, and liquid coexist,
growth of both feldspar and quartz, and at pressures and temperatures of
The starting material used for the 0.6 ± 0.3 kb, 990 °C, where sanidine,
determination of the isobaric T-X quartz, tridymite, and liquid coexist.
diagrams (Figs. 81 and 82) was glass These two invariant points are shown
or gel dried at 900 °C for 40 minutes, schematically in Fig. 83. The P-T-X
The presence or absence of a crystal- relations at these two postulated in-
line phase in the glass, determined variant points are only approximate,
by means of the petrographic micro- since the data are extrapolated from
scope, was the criterion used for the the vicinity of 10 kb.
location of these liquidus curves. In the system sanidine-silica at 10
In both aspects of this study and kb the temperature of the incongru-
in both systems equilibrium was not ent melting of sanidine was deter-
attained during the experiments (1 mined as 1285° ± 10 °C, in good
to 4 hours) ; however, the direction agreement with Lindsley's (1966)
of equilibrium is readily obtained. value of just greater than 1290 °C.
The experimental results bearing The composition of the univariant
on the composition of the eutectic liquid that coexists with sanidine and
liquids in these two systems as a leucite is Sa83±2Qi7±2, at 1285° ± 10 °C
function of pressure may be repre- and 10 ± 0.2 kb.
sented conveniently in terms of a The results of this study indicate
poly thermal P-X projection. A dia- a marked effect of pressure on the
gram of this type illustrates the composition of the eutectic liquids in
change in composition of coexisting the binary systems sanidine-silica and
phases along a P-T univariant curve, albite-silica over the 10-20 kb pres-
It is assumed that each of the feld- sure range. It is to be expected, as
spars and quartz are pure phases; shown schematically in Fig. 83, that
thus, only the change in composition a similar effect will be shown in the
of the univariant liquid is shown. In pressure range below 10 kb. We may
Fig. 83 the P-T and P-X projections also expect the position of the bound-
are given for the systems albite-silica ary curve separating the primary
and sanidine-silica. fields of quartz (or tridymite) and
Over the P-T range studied quartz homogeneous alkali feldspar crystal-
is the silica phase participating in the line solutions in the ternary system
equilibria. The studies of Schairer sanidine-albite-silica to shift as a
and Bowen (1955, 1956) in these two function of pressure,
binary systems indicate that tridy- If the position of the boundary
mite is the stable phase participating curve in the system albite-sanidine-
in the eutectic melting reaction at silica does shift with pressure, the
atmospheric pressure. Tuttle and £h2o in water-undersaturated melts
England (1955) and Kennedy, Was- cannot be related simply to values
482
CARNEGIE INSTITUTION
obtained on the saturation surface by
Turtle and Bowen (1958) and Luth,
Jahns. and Tuttle (1964).
Since the effects of pressure of
water and of water as a component in
the synthetic systems albite-silica-
water and sanidine-silica-water can-
not be treated separately with respect
to water-undersaturated liquids, it is
unlikely that these two effects can be
treated separately in applications of
the experimental studies to the crys-
tallization of water-undersaturated
granitic magmas.
1700
1600
1500
o
o
CD
o
CD
£
1400
1300
1200
100
1000
900
latm (Schairer and Bowen, 1955
20 kb
10 kb
10
20 30
40
50 60 70 80 90
KAISi30Q Si02
Weight per cent
Fig. 81. Isobaric (±0.2 kb) temperature-composition diagrams for the system KAISisOs-SiOa.
JL(Cr), l(Tr), L[Q), /.(Sa), and L(Lc) denote the liquidus curves at each pressure and the primary
phase present as, respectively, cristobalite, tridymite, quartz, sanidine, and leucite.
GEOPHYSICAL LABORATORY
483
O
o
CD
Z3
"5
^_
CD
CL
E
|2
I0001
NaAISi308 ,0
Si02
Weight per cenl
Fig. 82. Isobaric (±0.2 kb) temperature-composition diagrams for the system NaAISi308-Si02.
t(Cr), /.(Tr), L(Q), and £(Ab) denote the liquidus curves at each pressure, and the primary phases
present as, respectively, cristobalite, tridymite, quartz, and albite.
484
CARNEGIE INSTITUTION
■ < 1 ■ T T "I 1 T 1 '
■■ 1
i i i
OHr
^\"\ ^^^
\\ ^
\ 1
-
, . 1 1 1 1 . 1
■
1 ,L
o
o
o
<u
=j
o
i_
CD
Q.
E
|2
00
SJDqon^ * ajnssaJd
o .t: c
7 I D
S *£ •*
- — < o -C
--=• D
"O ,-i.
>» CN
O >o
CO Ov
E C
P c
o
O -o o
< . c
2 N < |
/> *• .«- "St
O
oo
o
to
<
c
D
«s
o
to
oo
o
<
D
z
E
<u
G
to
JC
o
Q.
c
O
t3
'o
a
i_
3
D
Q.
£
D
^3
Si 8"
° o .2>
o u_
2"g.S
"5 ° c
| .| .>
D
~j *> o
™ u> (to
!_•-«/>
D 5
3 S
c o 0
J -8 H
!'!©
w a _d
-t *c <
o ° -„
rr *E c
-; i o
'I - II
-°o
~a to
C C _|_
DO'
<U • - O
.eSw
"D Q. v>
D o £
to (J D
f *". ^
</> CN o
•55 00 -5
-c ^ "O
12 £
(to
c ~7
O ■*■
<~» C
D D
<D
U "D
— 1 >-
a
II »
_Q ..
< CO
. to
II
«/> ■»
<U >«
~ "O
D o
£ c
oo (to
(to ^
£ I
o <u
c
u
o
0) (J •-
_. (to a>
l- 'I .5
+ o 8.
-Q jo c
< *■ D
-~ "o *E
C o >
DC
CQ
^ >0 -j
II S +
+ JT^
- O +
> ° ^
L^ • - O
K !
a.
o 1 -Q
-C T"
1—
+
_Q
<
.- D)
3 :r
•*: cr
+
>- c
3 D
CO
00
cn
"".Eg
c c D
D <1>
~ > -=•
S ^r
c £ n
.t: 0
«/> Q. "D
O K C
a. <u o
u
aj
D)
to +
V)
-C _c *-
13*
to
d> C O
d^:-e
7 D) «
D "-1 ""
(D Oh
GEOPHYSICAL LABORATORY
CRYSTALLOGRAPHY
485
The Crystal Structure of
Ardennite
Gabrielle Donnay and Rudolf Allmann*
The crystal structure of ardennite
was solved by the symbolic addition
procedure of Karle and Karle (1963,
1966). It turns out to be a structure
closely related to that of epidote, as
had been suspected (Year Book 64,
pp. 209-210) , but it does show an un-
expected novel feature: linear Si3Oi0
groups of symmetry 2mm, with the
three tetrahedra on one side of the
line of coplanar edges. This is why
we were unable to guess the correct
trial structure. The epidote cell is re-
flected in "Ito mirrors" (001)0 and
(001) a (Fig. 84), so that the cell vol-
ume is doubled. The formulae for the
cell content of epidote and half the
cell content of ardennite can be writ-
* The Johns Hopkins University.
ten so as to show their structural re-
lation (Table 21). In ardennite there
are three independent octahedral
chains extending along b, whereas
there were only two in epidote (Fig.
85). The "mixed rings" of epidote
{Year Book 61+, p. 210) are also found
in ardennite, where they share the
central Si04 tetrahedron of the Si3O10
group, a tetrahedron that lies on the
mirror plane. The two five-membered
rings look like a large bow tie. The
Mn2+ coordination polyhedron is an
extremely distorted octahedron, with
one metal-oxygen distance so much
longer than the other five that only
five nearest neighbors can be con-
sidered. Jointly with Dr. F. Senftle,
of the U.S. Geological Survey, we are
investigating the expected abnormal
magnetic properties of ardennite.
The presumed substitution of As
and V for Si, which prompted this
Itom
• SiOAl ©AI,Fe ©Ca OOxygen Ooxygen and (OH)
Fig. 84. The epidote structure (ten cells in all, after Ito, Morimoto, and Sadanaga, 1954),
projected on (010). The Ito mirrors, through the cell origin, are a distance d (001) apart. The out-
line of the orthorhombic cell related to that of ardennite, also projected on (010), is shown with
dashed lines.
486
CARNEGIE INSTITUTION
O
C)
a
o
O
Q.
la
o
o
o
g
c
o
o
»_
•»—
u
3
a>
•*-
"E
c
o
-a
-a
c
o
c
t3
*o
o.
oo
D>
GEOPHYSICAL LABORATORY 487
TABLE 21. Structural Formulae of Epidote and Ardennite
Epidote Ca2 Ca2 (AI,Fe3+)2 (AIOH)2 Q (AIO)2 (Si207)2 (Si04)2
Ardennite Mn22+ \J2 (Mn2+,Ca)2 (AIOH)2 (AIOH)2 [(Mg,AI,Fe3+)OH]2 (As,V)O4Si3O10 (Si04)2
Q is a vacant site.
structure investigation, is not con- structural changes in the single mus-
firmed to any large extent, because a covite layers caused by isomorphic
separate (As,V)04 tetrahedron oc- substitutions.
curs in the structure (Fig. 85), The samples of muscovite and
astride the (001) mirror. The Si(l) phengite studied have been described
of the isolated Si04 tetrahedron (Fig. by Ernst (1963). The muscovite is
85) has a considerably lower temper- from a pegmatite from Georgia, U.S.
ature factor (B — 0.22 ± 7) than National Museum No. 105100, with
the other two silicons (B — 0.42 ± 8; the composition
B = 0.42 ±z 10) ; also the tetrahedron ,Q. A1 v , A1 T. ^ 3+
is slightly less regular. We therefore ^ls™Ai<>-»*> ^All-Mll^+e °^2
assume any substitution of As and V, * e °™mgo<>*> Ul°
when these atoms are present in (OH1.99Fo.oi)[Ko.86Nao.io(H±30)o.oi]
amounts of more than two atoms per Phengite is from glaucophane + ac-
cell, to take place in this tetrahedron, tinolite + chlorite + aragonite +
The recent analyses of Semet and quartz schist from Tiburon Penin-
Moreau, which were converted to suia> California, with the chemical
structural formulae (Year Book 65, formula
p. 294), require as much as 10 atomic
% As substituting for Si. This is the (Sis.sgAW) (Ak.43Tio.01
highest amount indicated by any ar- Fe3+o.o5Fe2+o.o9Mgo.5o)Oio
dennite analysis that we have been [(OHh.^Oo.os] (K0.87Nao.o7
able to find in the literature. Ba0.0iCa0.o2)
The X-ray diffraction symbol for
The Crystal Structures of 2Mx muscovite phengite, determined from
Phengite and 2Mx Muscovite precession photographs, is 2/mC-/c,
N. Gilven with the possible space groups Cc
Although numerous structure anal- and C2/c. Precise unit-cell dimen-
yses of micas have recently been sions (Table 22) were determined by
made, the structural factors govern- least-squares refinement (Burnham,
ing the stacking sequence remain un- Year Book 61 ) of data obtained by
known. In general terms, mica modi- means of a precision back-reflection
fications can be considered as differ- Weissenberg camera with CuKa radi-
ent stacking sequences of topologically ation (CuKai = 1.54051, CuKa2 =
similar but not identical units (single 1.54434) .
mica layers) . Differences in these Three-dimensional least-squares re-
units, which can be obtained from fmement was carried out with 583 ob-
precise analyses of the structures, served reflections (from 756 possible
may reveal the factors that control reflections) for muscovite and 741
the stacking sequence. In a recent observed reflections (from 1077 pos-
account of muscovite polymorphism, sible reflections) for phengite. In-
Giiven and Burnham (1967) showed tensity data collection procedures and
that the single layers of these poly- subsequent data processing were the
morphs are significantly different, same as described in Year Book 65
The present study considers the (pp. 283-285). One structure model
4SS
CARNEGIE INSTITUTION
TABLE 22. Unit-Cell Dimensions of Phengite and Georgia Muscovite
Georgia Muscovite
Phengite
a, A
fa, A
c, A
P ,
Unit-cell volume, A3
5.1906
9.0080
20.0470
95.757
932.614
0.0002
0.0003
0.0006
0.002
0.075
5.2112
9.0383
19.9473
95.769
934.77
0.0003
0.0004
0.0006
0.005
0.10
in space group 02/ c with the atomic
parameters of a %rsix Na-rich musco-
vite (Burnham and Radoslovich,
Year Book 63) and another structure
model derived from the first one in
space group Cc have been used for
the least-squares refinement. In the
space group C2/c, discrepancy in-
dexes converged after 6 cycles to un-
weighted R(2|F0-FC|/2F0) = 0.045
and weighted R [Xw (AF) 2/%wF021 *
= 0.023 for phengite, and to un-
weighted R = 0.036 and weighted R
= 0.024 for muscovite. Atomic pa-
rameters of phengite after the last
cycle are given in Table 23. The iso-
tropic temperature factors listed in
Table 24 were converted to aniso-
tropic Pa's and further refinement
was carried through three more
cycles for muscovite, resulting in un-
weighted R = 0.035 and weighted R
— 0.022. Atomic parameters of mus-
covite after the last cycle are given
in Table 24. Anisotropic refinement of
phengite was not possible, however,
owing to negative Pa's for some
atoms.
The refinement of the second model
in space group Cc with 19 atoms in
the asymmetric unit gave larger dis-
crepancy factors and negative tem-
perature factors for some atoms. The
least-squares correlation matrix
showed very high correlation coeffi-
cients between the atoms that are re-
lated to each other by the inversion
center in space group C2/c. Thus,
space group C2/c is regarded as cor-
rect for both mica structures.
Interatomic distances and angles
of muscovite and phengite were cal-
culated by an IBM 7094 program
ORFEE (Busing, Martin, and Levy,
1966), and the important bond
lengths are given in Table 25.
T-0 distances in the structure of
muscovite showed the absence of or-
dering of (Si,Al) over the two dis-
tinct tetrahedral sites, contrary to
the results of Radoslovich (1960)
but in agreement with those of Burn-
ham and Radoslovich (Year Book 63)
for Na-rich muscovite. In phengite
the two average tetrahedral metal-
oxygen bond lengths, with Ta-0 =
1.622 ± 0.003 and T2-0 = 1.633 db
0.003, indicate a slight ordering, al-
TABLE 23. Atomic Coordinates and Isotropic Temperature Factors in 2Mi Phengite5
Atom
Equipoint
X
y
z
6
K
4e
0.0
0.0964(5)
1/4
1.76(3)
Al
8f
0.2470(3)
0.0825(3)
0.0(0.0001)
0.74(3)
Oa
8f
0.9568(6)
0.4396(7)
0.0544(2)
1.03(8)
ob
8f
0.3933(6)
0.2496(9)
0.0537(2)
1.07(8)
OH
8f
0.9531(7)
0.0656(8)
0.0526(2)
1.21(8)
Tt
8/
0.9632(3)
0.4297(4)
0.1355(1)
0.60(3)
T2
8f
0.4525(3)
0.2581(3)
0.1354(1)
0.60(3)
oc
St
0.4426(5)
0.0931(8)
0.1678(1)
1.30(6)
od
8f
0.7372(7)
0.3257(5)
0.1601(2)
1.41(0.10)
Oe
8f
0.2326(7)
0.3574(6)
0.1682(2)
1.23(9)
Standard deviations are given in parentheses.
GEOPHYSICAL LABORATORY
489
TABLE 24. Atomic Coordinates and Isotropic Temperature Factors in 2Mi Muscovite
Atom
Equipoint
X
/
z
8
K
4e
0.0
0.0985(3)
1/4
1.60(4)
Al
Bf
0.2496(3)
0.0834(2)
-0.00005(8)
0.47(2)
Oa
Bf
0.9613(7)
0.4435(5)
0.0540(2)
0.84(7)
o„
Bf
0.3850(5)
0.2519(5)
0.0537(2)
0.77(6)
OH
Bf
0.9564(7)
0.0630(5)
0.0505(2)
0.75(7)
Tt
Bf
0.9648(2)
0.4295(2)
0.1355(1)
0.54(3)
T2
Bf
0.4510(3)
0.2584(2)
0.1355(1)
0.54(2)
Oc
Bf
0.4174(6)
0.0930(6)
0.1685(2)
1.07(6)
o,
Bf
0.7513(7)
0.3110(5)
0.1575(2)
1.34(8)
Oe
Bf
0.2522(7)
0.3705(5)
0.1689(2)
1.17(7)
Standard deviations are given in parentheses.
TABLE 25. Interatomic Distances (A) in 2Mj Muscovite and in 2Mi Phengite
Muscovite
Phengite
Tt tetrahedron
TrOa (apical)
TrOc
TrOd
TrOe
Mean T^-O
Oa-Oc
Oa-Od
Oa-Oe
Oc-Od
Oc-Oe
Od-Oe
Mean O-O
T2 tetrahedron
T2-Oj> (apical)
T2-Oc
T2-Od
T2-Oe
Mean T2-O
Ob-Oc
o6-od
Ob-Oe
oc-od
Oc-Oe
OdOe
Mean O-O
Al octahedron
AI-Oa
AI-Oa'
AI-Ob
Al-Ob'
AI-OH
AI-OH'
Mean Al-O
Unshared
Oa'-Ob
Oa-OH
Ob'-OH
Mean O-O
1.637
1.644
1.632
1.661
0.004
0.004
0.004
0.004
1.643
2.691
2.716
2.707
2.685
2.653
2.642
0.002
0.005
0.005
0.006
0.007
0.006
0.006
2.682 ± 0.002
1.642
1.647
1.647
1.635
0.004
0.005
0.004
0.004
1.643
2.702
2.728
2.698
2.643
2.642
2.678
0.002
0.006
0.005
0.005
0.006
0.007
0.006
2.682 ± 0.002
1.931
1.951
1.950
1.923
1.919
1.920
0.004
0.004
0.005
0.005
0.004
0.004
1.932 ± 0.002
2.956
2.860
2.856
0.005
0.004
0.005
1.618
1.620
1.621
1.629
0.004
0.007
0.005
0.004
1.622
2.662
2.702
2.671
2.642
2.615
2.592
0.003
0.006
0.006
0.006
0.008
0.008
0.006
2.647 ± 0.003
1.630
1.629
1.634
1.639
0.004
0.007
0.004
0.005
1.633
2.671
2.725
2.692
2.617
2.626
2.661
0.003
0.007
0.006
0.006
0.008
0.008
0.006
2.665 ± 0.003
1.985
1.950
1.956
1.961
1.938
1.946
0.004
0.005
0.007
0.007
0.006
0.004
2.891 ± 0.005
1.956 ± 0.002
2.918 ± 0.008
2.863 ± 0.004
2.874 ± 0.008
2.885 ± 0.004
490
CARNEGIE INSTITUTION
TABLE 25. Interatomic Distances (A) in 2Mj Muscovite and in 2Mi Phengite — [Continued)
Muscovite
Phengite
Shared
OH-OH'
0,0,'
Ob-Ob
Mean O-O
Interlayer cation
K-O,
K-Oj
K-O,
Mean K-Ojnner
K-Oc
K-Od
K-O,
Mean K-O0uter
K-OH
K-K
2.402
2.461
2.447
0.007
0.007
0.007
2.434 ± 0.004
2.842
2.879
2.845
0.003
0.005
0.005
2.855
3.290
3.511
3.284
0.002
0.003
0.005
0.005
3.362
3.997
10.179
0.003
0.003
0.001
2.489
2.509
2.484
0.009
0.008
0.006
2.497 ± 0.005
2.961
2.982
2.966
0.003
0.005
0.005
2.970
3.189
3.344
3.179
0.003
0.003
0.006
0.006
3.237
3.931
10.125
0.003
0.004
0.002
though isotropic temperature factors Atomic planes in micas are dis-
for both tetrahedral sites show no torted in plan and in elevation. In
difference. ideal atomic configurations both
Average octahedral metal-oxygen anions and cations form hexagonal ar-
distances are 1.932 A for Georgia rays with interatomic angles of 120°.
muscovite and 1.956 A for phengite, Deviations (2a) from this value give
as would be expected with the in- directly the distortions in the plane.*
creased number of larger cations in The average values of 2a are given in
the octahedra of the latter. There is Table 26 for 2Mi muscovite, phengite,
no ordering in octahedral substitu- and 3T muscovite. Distortions in the
tions in either mica, as shown by the plane of basal oxygens are twice as
refinement of the model in space great in 2Ma muscovite as in phen-
group Cc, which allows two distinct gite. In the plane of apical oxygens,
octahedral sites. however, they are similar in both
The interlayer cation is in six- structures, one direction being dis-
coordi nation with the basal oxygen tinctly greater than the others, as
atoms (three from bottom and three shown on Fig. 86(B) for 2Ma mus-
from top tetrahedral layers) with an covite. The tetrahedral and octa-
average distance of 2.855 A for mus-
covite and 2.970 A for phengite. The
first and second nearest K-O dis-
tances are given in Table 25.
* Approximately half the value of 2a is
usually referred to as the tetrahedral rota-
tion, since the rotations of two adjacent
tetrahedra contribute to 2a.
TABLE 26.
120° in
Average Deviati
2Mi Muscovite,
ons
2M
(2a)
, Phe
of Interatomic Angles from
ngite, and 3T Muscovite
2a, degrees
2Mi Muscovite
2Mi Phengite
3T Muscovite
Basal oxygens
Apical oxygens
Tetrahedral cations
Octahedral cations
22.73 ± 0.07
10.81 ± 0.10
1.86 ± 0.05
0.20 ± 0.12
12.05 ± 0.07
9.16 ± 0.12
1.59 ± 0.06
0.75 ± 0.14
23.68 ± 0.20
11.31 ± 0.30
2.50 ± 0.24
GEOPHYSICAL LABORATORY
491
O
c
c
'o
c
o
u
UJ
H
>
O
O
CO
z>
ro
O
■*a
UJ
>
o
o
CO
Z>
CO
OQ
>»
X
o
'a
o
c
<D
to
O
v.
a
c
O
a
u
—
0
CO
-a
-
CN
C
a;
x
o
a.E
**- "S
° o
c ■ — ■
£ <
E S
U) <+.
«= O
T3
SJ
81
00
_C
0
n
m
LU
^-
492
CARNEGIE INSTITUTION
hedral cations show very little dis-
placement from their ideal sites in
hexagonal arrays.
Distortions in elevation result in
the corrugation of the anionic planes,
as seen from differences in the z co-
ordinates of the atoms belonging to
the same plane. One of the basal
oxygens, Oj, is displaced by an amount
A.c
-4- 7
c • sin p = 0.22 ± 0.01 A
for muscovite and 0.16 ± 0.01 A in
the structure of phengite. This cor-
rugation of the basal oxygen layers
(usually referred to as tetrahedral
tilt) is characteristic for dioctahedral
2Mi micas.
Tetrahedra in phengite, however,
are 0.06 ± 0.01 A less tilted than in
muscovite. This difference in tetrahe-
dral tilt is the probable cause of c-
axis shortening in phengite. This
point can also be confirmed by calcu-
lating average thicknesses of tetra-
hedral layers in both mica structures
with the use of the relation
*o,
*o,
%on ~r Zqi
• c • sin p
This calculation gives 2.29 A for mus-
covite and 2.26 A for phengite, and
four times this difference is approxi-
mately equal to the shortening of the
c axis in phengite. In comparison, the
thickness of octahedral layers indi-
cated no difference for both mica
structures. The distance between the
lower and upper basal oxygen layers
above and below the interlayer cat-
ions is 3.27 A for muscovite and 3.25
A for phengite.
As shown by the experiments of
Crowley and Roy (1964) and Velde
(1965), significant solid solutions
occur along the join muscovite-
(AlMg) celadonite. Velde (1965) has
demonstrated that more Mg can be
accommodated in the muscovite struc-
ture with increasing pressure
(^Pn.,o) at lower temperatures. The
limit of Mg substitution is the phen-
gite composition. As shown above, Mg
substitution (which is necessarily ac-
companied by Al/Si substitution in
tetrahedra) causes shortening of the
c axis and decrease in distortions in
the muscovite structure. This may ex-
plain why phengite is stabler than
muscovite in a high-pressure, low-
temperature environment, as shown
by Ernst (1963). These specific P-T
conditions are not necessary, how-
ever, for the formation of phengites.
Provided that chemical activities are
favorable, the formation of a diocta-
hedral mica with fewer distortions
(phengite type) seems to require less
energy than the formation of a more
distorted muscovite-type structure, as
appears to be the case in nature.
A Mechanism of Stacking
Sequences in Dioctahedral Micas
N. Giiven
The structures of 2Mi and 3T mus-
covite have been compared in an ef-
fort to explain how the effects of
octahedral vacancies, distortions, and
substitutional order-disorder in single
mica layers give rise to multiple
modes in lattice translations along c,
which differ from the normal mode
of the 1M stacking sequence.
In Fig. 86 (D, E) the apical oxy-
gens of lower (light lines) and upper
(heavy lines) tetrahedral layers of
2M] and 3T muscovite are shown.
In ideal close-packing, apical oxygens
in each layer form hexagons around
OII~ ions, Fig. 86(A). In dioctahedral
micas, octahedral vacancies give rise
to severe distortions, Fig. 86 (B,
C), of these hexagons and result in
GEOPHYSICAL LABORATORY 493
shortening of four edges around oc- linked by octahedral cations that
cupied sites and stretching of two share apical oxygens of the tetrahe-
other edges of the hexagons around dra. From the arguments it is obvious
the vacant sites. The strong attraction that if the lower tetrahedral layers
between OH-octahedral cations dis- are undistorted as in Fig. 86 (A) ,
places OH from the centers of the there must be many possible positions
hexagons. Thus the Oa-0&-OH triangle for the upper layers. Otherwise upper
around an occupied site becomes dis- and lower tetrahedral layers must be
tinct. Suppose we take out the octa- linked in a way that represents a
hedral cations and keep the arrange- stable arrangement,
ment as it is in Fig. 86 (A, B, C). If, The above-mentioned displacements
now, octahedral cations were to come of apical oxygens cause immediate
back to the arrangement in Fig. rotations and tilting of the (Si,Al)
86(A), they could be accommodated tetrahedra around an axis parallel to
equally well in any one of six possible c* and passing through the silicons.
ways without any energy difference. Tetrahedral metals, therefore, do not
If, however, they came back to the show large displacements. As the av-
arrangement in Fig. 86 (B or C) they erage values of 2a (Table 26) show,
would most likely occupy the previous the basal and apical oxygens are,
positions. however, severely distorted into a
Referring now to Fig. 86(D), the pseudoditrigonal ring. Although ro-
hexagons of upper and lower apical tational distortions are similar in 2Mi
oxygens have to be so arranged and 3T muscovite structures, tetrahe-
around octahedral cations that a dral tilt is markedly different — Az
least-strained coordination around = 0.22A for 2MX muscovite and Az
the latter is possible. This arrange- = 0.14 A for 3T muscovite. The net
ment would be centrosymmetrical, result is a destruction of the 2-fold
with the center of symmetry in the axis in the single layer of 2MX mus-
vacant octahedral site, as in the 2M2 covite.
muscovite structure. In this structure The next step is the coordination
the single mica layer has Ci sym- of interlayer cation with basal oxy-
metry, and the tetrahedral cations gens. To simplify relationships be-
(Si,Al) are completely disordered, tween successive 10-A mica layers the
However, if there is (Si,Al) ordering, following designations will be used:
the above arrangement of upper and Ph Pu, lower and upper tetrahedral
lower apical oxygens around octahe- layers in the first 10-A mica layer;
dral cations will be less stable because IPh IPU, lower and upper tetrahedral
octahedral polyhedra will have two layers in the second 10-A mica layer;
highly charged oxygens along a and so on.
shared edge. The arrangement, found The pseudoditrigonal arrangement
in the 3T muscovite structure, with a of basal oxygens because of previous
2-fold axis passing through the va- distortions allows the interlayer cat-
cant octahedral site and the midpoint ion to form first-coordination bonds
of OH-OH', would be stabler (Fig. with three basal oxygens of the tetra-
86 E), because it avoids the occur- hedral layers. Interlayer cation-basal
rence of two highly charged oxygens oxygen distances are very similar in
along a shared octahedral edge. both 2MX and 3T muscovites. The two
For simplicity the above discussion tetrahedral layers, I*„ and IPZ, around
was expressed in terms of apical oxy- interlayer cations exert forces upon
gens. In reality we mean upper and them with opposite directions, for
lower tetrahedral layers, which are their vertices point in opposite direc-
494
CARNEGIE INSTITUTION
tions. The stable arrangement, then,
would be symmetrical. The position
of interlayer cations in muscovites
allows, however, only a 2-fold axis
between the tetrahedral layers. This
operation is equivalent to a 120°
rotation of the top tetrahedral layers
(II- j ) over I*i around an axis parallel
to c*, if the single layer has C2 sym-
metry. Now the hexagons of the
apical oxygens of IV j are distorted
in the same way as those of the
apical oxygens of Vu. The positions
of OH and Al are therefore fixed
as in Fig. 86 (B or C). The relative
positions of the II*i and IVU tetra-
hedral layers are decided by the
same factors as for Vi and Vu. With
the same arguments as above, the
next tetrahedral layer (III*i) will
again be related to II*« by a 2-fold
axis. If, howrever, a previous inver-
sion operation took place between
lYi and IF*, the IIP* would then be
identical with P?. Thus, the periodic-
ity will be established after two
layers, generating a 2MX stacking se-
quence. If IP; and IV u are related by
a 2-fold axis, III*, will be rotated 120°
relative to IPZ, generating a 3T stack-
ing sequence of (120°, 120°, 120°).
Thus the symmetry of the single
layer, determined by defects (vacan-
cies), isomorphic substitutions, order-
disorder, and distortions, generates
the periodicity of the stacking se-
quence. The relations of the top and
bottom tetrahedral layers around the
interlayer cation are the same for
2MX and 3T muscovite. The role of
interlayer cations, in the hypothesis
proposed here, is therefore secondary
in generating stacking sequences, in
contrast to its primary importance in
the hypothesis of Radoslovich (1960,
1963).
I would like to thank C. W. Burn-
ham for his critical review of this
report and for his constructive sug-
gestions.
Refinement of the Crystal
Structure of Pigeonite,
(Mgo.39Feo.52Cao.09) Si03
N. Morimoto and N. Gilven
It has been the purpose of this
study to refine the crystal structure
of pigeonite, with three-dimensional
data, and to elucidate the effect of Ca
atoms in clinopyroxene structures.
Our effort has been concentrated on
these problems: (1) the distribution
of divalent Ca, Fe, and Mg atoms
among the possible sites, Ml and M2;
(2) the nature of the coordination
around the Ml and M2 sites; and (3)
the reason for the diffuseness of the
h + k = 2n + 1 reflections.
Pigeonite from the Isle of Mull,
kindly provided by H. S. Yoder, Jr.,
has been used for the structure re-
finement in this study. The space
group is P21/c. The cell dimensions
were determined by the least-squares
refinement of measurements from
precision back-reflection Weissenberg
photographs, with the use of the pro-
gram of Burnham (Year Book 61,
pp. 132-135). The results, based on
38 independent measurements, are
as follows: a = 9.706 db 0.002, b =
8.950 dz 0.001, c = 5.246 ± 0.001 A,
and/? = 108.59 dz 0.01°.
The chemical composition of the
Mull pigeonite was determined to be
(Mgo.39Feo.5oCao.o9)Si03 by Hallimond
(1914). Electron-probe microanalysis
by F. R. Boyd gave a composition of
(Mgo.4iFe0.5iCa0.o8)SiOs, in good
agreement with that obtained by
Hallimond.
All the integrated intensities were
measured with the Supper-Pace au-
tomatic single-crystal diffractometer.
The intensity collection and data-
processing procedures are the same as
those described in Year Book 65 (pp.
283-285).
Since pigeonite has a pseudo C
lattice, the h 4- k = 2n + 1 reflec-
tions are generally weak and result
GEOPHYSICAL LABORATORY
495
in many reflections with integrated
intensity less than the minimum ob-
servable value. Thus, among the 1500
recorded, only 523 reflections could
be used for the refinement, of which
374 were h 4- k = 2n reflections and
149 were h + k = 2n + 1 reflections.
Full matrix, least-squares refine-
ment was carried out on the IBM
7094 computer with the use of a pro-
gram written by Prewitt (1962). Re-
finement was initiated with the
atomic coordinates of Ashio pigeon-
ite* by Morimoto, Appleman, and
Evans (1960), transformed to the
conventional unit cell. The distribu-
tion of metal atoms was assumed so
that Mg and Ca are ordered com-
pletely into Ml and M2 sites, respec-
tively. This assumption gave a start-
ing distribution for the Ml site of
0.78Mg and 0.22Fe and for the M2
site, 0.84Fe and 0.18Ca, on the basis
of the chemical composition by Halli-
mond.
After nine cycles of least-squares
refinement the R values reduced to
0.045. Three final cycles of refinement
with anisotropic temperature factors,
varying scale factors, and metal distri-
bution factors reduced the R values
to 0.037 (Table 27). The final results
* The Ml and M2 sites in this report
were named Mil and MI in the earlier paper
(Morimoto, Appleman, and Evans, 1960).
The reader must be cautious in referring to
the earlier paper.
of the least-squares refinement show
that the distribution of metal atoms
is 0.73Mg and 0.27Fe for the Ml site
and 0.75Fe and 0.25Ca for the M2
site.
The chemical composition demands
that, on the basis of the metal distri-
bution in the Ml sites, the M2 site
must have 0.05Mg, 0.77Fe, and
0.18Ca. This metal distribution gives
a total scattering factor almost the
same as that for the distribution of
0.75Fe and 0.25Ca, obtained by least-
squares refinement for the M2 site.
The structure factors were computed
with the new metal distribution. The
R value was 0.036. The final param-
eters and individual isotropic tem-
perature factors, calculated from the
anisotropic temperature factors, are
listed with their estimated standard
deviation in Table 27.
Discussion
Interatomic distances for the final
structure are given in Table 28. The
structure is essentially the same as
that obtained earlier (Morimoto, Ap-
pleman, and Evans, 1960).
Single silicate chain. There are two
crystallographically different kinds
of single silicate chain in this struc-
ture. The SiA chain is more extended
along the c axis (03A-03A'-03A"
= 167°) than the Si£ chain (035-
03£'-03£" = 142°), but is not so
fully extended as the chain in jadeite
TABLE 27. Final Atomic Coordinates and Isotropic Temperature Factors:
Atom
X
y
z
0.2328(8)
B
Ml
0.2507(4)
0.6548(2)
0.67
M2
0.2564(3)
0.0183(1)
0.2308(6)
1.26
SiA
0.0427(3)
0.3398(5)
0.2797(6)
0.62
Sifi
0.5504(3)
0.8367(5)
0.2372(6)
0.54
OlA
0.8659(8)
0.3404(13)
0.1715(15)
0.62
02A
0.1220(9)
0.4970(10)
0.3306(16)
0.93
03A
0.1037(5)
0.2633(6)
0.5779(13)
0.95
OIB
0.3743(8)
0.8342(13)
0.1344(15)
0.66
02B
0.6290(8)
0.9877(11)
0.3765(17)
1.23
03B
0.6053(5)
0.7087(6)
0.4773(12)
0.93
* Standard deviations, <r, given in parentheses.
496
CARNEGIE INSTITUTION
TABLE 28. Interatomic Distances
in Pigeonite
*
SiA tetrahedr
on
Si-OIA
1.627(8)
Si-02A
1.585(10)
Si-03A
1.636(7)
Si-03A'
1.650(7)
Mean SiA-O
1.625
01A-02A
2.742(9)
01A-03A
2.686(10)
01A-03A'
2.675(9)
02A-03A
2.497(11)
02A-03A'
2.658(10)
03A-03A'
2.634(1)
Mean O-O
2.649
SiA-SiA'
3.077(5)
Si8 tetrahed
ron
Si-OIB
1.620(8)
Si-02B
1.609(11)
Si-038
1.665[7)
Si-03B'
1 .660(7)
Mean SiB-O
1.639
01B-02B
2.759(9)
01B-03B
2.653(9)
01B-03B'
2.637(11)
02B-03B
2.688(10)
02B-03B'
2.578(12)
03B-03B'
2.725(3)
Mean O-O
2.673
SiB-SiB'
3.048
Ml octahed
ron
Ml-OIA
2.158(11)
Ml-OIA'
2.062(10)
Ml-OIB
2.162(9)
Ml-OIB'
2.068(10)
M1-02A
2.057(10)
M1-02B
2.086(10)
Mean Ml-O
2.099
M2 coordinc
ition
M2-01A
2.143(10)
M2-01B
2.156(10)
M2-02A
2.093(10)
M2-02B
2.006(10)
M2-03A
2.430(6)
M2-03B
2.614(7)
M2-03B'
2.968(6)
M2-03A'
3.470(6)
* Standard deviations, a, given in paren-
theses.
(03-03'-03" = 175°; Prewitt and
Burnham, 1966).
In all tetrahedra, Si-0 bonds to
chain-linking oxygens (03 A and
OSB) are significantly longer than
those to other oxygens. The mean dis-
tances of SiA-03A (1.643 A) and of
SiB-OSB (1.663 A) are to be com-
pared with those of SiA-0 1 A and
SiA-02A (1.606 A), on the one hand,
and with those of SiB-OlB and SiB-
02B (1.605 A), on the other. The
Si-Si distances are 3.076 A for the
SiA chain and 3.048 A for the SiB
chain. These values are in good agree-
ment with those for other monoclinic
pyroxenes.
The distortions of the O-Si-0 bond
angles from the tetrahedral values
are similar for both the SiA and SiB
tetrahedra. The 01A-SiA-02A and
01B-SiB-02B angles are both 117°,
whereas the angle 03A-SiA-03A' is
106.5° and 03E-SLB-03A' is 110°.
These distortions are also shown by
the 0-0 distances of the tetrahedra.
The 01A-02A and 01B-02B dis-
tances are 2.742 and 2.759 A, re-
spectively, compared with 03A-03A'
and 03£-03£' distances of 2.634 and
2.725 A, respectively. These distor-
tions are in agreement with those in
other metasilicates (McDonald and
Cruickshank, 1967).
Metal polyhedra. Both the Ml and
M2 sites are in general position in
pigeonite. Their deviations from the
C-lattice positions are, however,
rather small compared with those of
other atoms. The Ml site is coordi-
nated octahedrally by six oxygen
atoms, with Ml-0 distances that de-
part a maximum of 0.06 A from the
mean. These Ml octahedra form
"chains" parallel to the silicate chains
by sharing a common edge with the
length of 2.840 A (OIA-OIB). The
coordination of the M2 site is irregu-
lar as in other pyroxenes. The ob-
served M2 oxygen distances are given
in Table 28. *
GEOPHYSICAL LABORATORY
497
Thermal model. The diff useness ob-
served with the h 4- k = 2n 4- 1 re-
flections implies that the pigeonite
has possibly a domain structure. The
apparent anisotropic temperature fac-
tors obtained are considered to be the
statistical average of the effect of
substitutional replacement of metal
atoms in the Ml and M2 sites and of
the domain structure. The interpreta-
tion of the anisotropic temperature
factors obtained is under investiga-
tion on the basis of the domain struc-
ture of pigeonite.
Omission Solid Solution in
Magnetite
G. Kullerud, Gabrielle Donnay,
and J. D. H. Donnay*
Magnetite powder and sulfur,
heated in evacuated silica tubes with
minimum vapor volume, react to pro-
duce pyrite and iron-deficient magne-
tite (see Year Book 65, p. 356). In
the magnetite structure, vacancies
occupy centers of oxygen tetrahedra
and octahedra. (The iron atoms are
withdrawn from tetrahedral as well
as octahedral sites.) The distribution
of vacancies between the two posi-
tions depends on the temperature
of the reaction; it can also be altered
by heat treatment of the reaction
product after the reaction has gone
to completion. Properties such as
density, color, magnetic suscepti-
bility, and thermodynamic stability
depend, in turn, on the distribution
of vacancies. The cell edge a, for a
member of the Fe3_.04 solid solution,
is not uniquely defined by composi-
tion. Because of the occurrence of
tetrahedral as well as octahedral va-
cancies, the formula must instead be
written : [Fe2.x[Jx] oct [Fei-JOy] tet04,
with x + y = z. Although our experi-
mental data as yet do not permit an
accurate plot, they can be used to
* The Johns Hopkins University.
show the trend of the surface of a as
a function of z, the total number of
vacancies (0 < z < 1/3), and y, the
number of vacancies on tetrahedral
sites.
Figure 87 was constructed as fol-
lows. The literature data for the series
magnetite-maghemite, obtained by
oxidation of magnetite in air (Hagg,
1935), give three points in addition
to the end members. The resulting
curve is valid for y = 0, for Hagg
has shown that all the vacant sites are
octahedral sites in this case. Our first
two experiments yield the section of
the desired surface at 300z = 50.
Our third experiment gives a second
point on the curve of the section at
3002 = 100. One last bit of informa-
tion was used : The Alaska magnetite
(Year Book 65, p. 356), with its a
value larger than that of Hagg's pure
magnetite, seemed likely to have a
certain number of vacant sites. Its
diffraction-line intensities were stud-
ied, and good agreement between cal-
culated and observed values was ob-
tained for 300z = 11 ± 2 and SOOy
= 4.4 (that is, 40 ± 5% of 11) . This
i 10
300 v
H-2
30 / 50
H-3 H-4
t 70 | 90 100 u K
1 8.35
845
Fig. 87. Nomogram giving z, total number of
vacant sites per formula Fe3_204, and /, num-
ber of vacant sites that are tetrahedral sites, in
terms of density D and cell edge a. A, Alaska
magnetite. I, II, and III, our experimental
samples.
498
CARNEGIE INSTITUTION
natural sample thus provided one
more point of the surface. One family
of contour lines (Fig. 87) are the loci
of points of equal a values. Although
the nomogram claims no more than
the status of an explanatory sketch,
it does bring out the fact that the cell
edge of pure FesO, (a = 8.397,
Hagg's value) decreases on oxidation
if the vacancies occur on octahedral
sites but increases if they occupy
tetrahedral sites. In the one case it
may reach a minimum of about 8.34 ;
in the other it could attain a maxi-
mum of about 8.45, at Fe203 composi-
tion. The smallest values on record, in
addition to that of Hagg, a = 8.339
(A from kX), are a = 8.333 ± 4 (A
from kX, Haul and Schoon, 1939)
and a = 8.330 (Lihl, 1950). Other
values on record are a = 8.34 (A
from kX, Pouillard, 1950) ; a = 8.342
± 8 (A from kX, Mason, 1943) ; a =
8.35 (Sinha and Sinha, 1957) ; a =
8.350 ± 4 (A from kX, Haul and
Schoon, 1939). The largest value on
record in the literature for yFe203 is
a = 8.4 (Thewlis, 1931). It now
seems likely that the cell edges
greater than 8.42, reported for mag-
netites, actually refer to intermediate
Fe3_;04 compositions with y > 0.
Such are a = 8.429 ± 3 (A from kX,
Holgersson, 1927; from Nord-
marken) ; a = 8.430 (A from kX,
Pouillard, 1950 ; synthetic) ; a =
8.434 (A from kX, Holgersson, 1927;
synthetic) ; a = 8.44 (A from kX,
Krause and Thiel, 1934) .
The surface of density D as a func-
tion of y and z has also been calcu-
lated. Contour lines are drawn (Fig.
87) for five values of the density 5.1
to 4.7. The construction is carried out
as follows. For each D value, the cell
weight is calculated for all the rele-
vant a values (from 8.34 to 8.45).
From the curve of cell weight against
300z, we get the value of SOOz at
which the D value must be plotted on
the appropriate a contour line. The
family of D curves suffers from the
inaccuracy of the family of a curves,
but this imperfection does not vitiate
the conclusion that the density, for a
given composition, decreases as the
percentage of vacancies on tetrahe-
dral sites increases.
Note that the two families of
curves (a and D curves) occupy a
right-angle triangle. The two coordi-
nates are taken as 300?/ and 300z. It
follows that the maximum total num-
ber of vacancies reads 100 instead of
y3. The hypotenuse of the triangle is
the locus of the points for which all
the vacancies are on tetrahedral sites.
An improved figure, with more
rigorous experimental control, will
become a nomogram that will yield
z and y from a knowledge of a and D.
Relative Orientations of
Intergrown Crystals
H. O. A. Meyer, Gabrielle Donnay, and
J. D. H. Donnay
Silicate inclusions, such as garnet
and olivine, which occur in diamonds,
are of crystallographic and mineral-
ogical interest. Several authors
(Mitchell and Giardini, 1953; Futer-
gendler, 1958; Futergendler and
Frank-Kamenetsky, 1961 ; Harris,
Henriques, and Meyer, 1966) have
found that the diamond host affects
the habit of the olivine inclusions and
that there is an orientation relation
between host and inclusions. A
method that would permit a rapid and
accurate determination of such a re-
lation would be useful.
H. 0. A. Meyer has had occasion to
use the Laue technique for the pur-
pose and has found that repeated re-
mounting of the specimen is often
necessary, so that the procedure may
be time consuming. The precession
camera is well suited to the solution
of the problem : It has been used be-
fore as an X-ray goniometer, to
decipher twins (Donnay, Donnay,
GEOPHYSICAL LABORATORY 499
and Hurst, 1955; Hurst, Donnay and angular coordinates (</>, p) and the
Donnay, 1956). For determining cyclographic projection of the net
orientation relations between host plane is drawn with the aid of the
and inclusion, we proceed as follows. Vulf net; it is the great circle whose
The diamond specimen is mounted pole is U. Next two indexed reciprocal
randomly and centered in the X-ray lattice rows, [hkl] * and [h'k'l'] *, in
beam so that at least one inclusion, the net (uvw) * are selected for plot-
or part of it, is in the beam. Explora- ting on the stereogram. To determine
tory orientation photographs are the true angle ^ which a row [hkl] *
taken (^ = 15°, unfiltered Mo radia- makes with the trace t of the net on
tion) with various dial settings until the projection plane, we draw if
identifiable orientation circles for parallel to t on the film (V passes
both host and inclusion appear on one through the center of the film and is
film. Frequently this occurs on the perpendicular to the line on which
first trial exposure. We determine the r was measured) . Note that we write
missetting of a reciprocal lattice net t' at the left end of the V line (when
by locating the center of its orienta- looking in the direction center of film
tion circle and measuring its polar co- to center of orientation circle) . The
ordinates, $ and r, on the film. Note angie w between V and the reciprocal
that <f> is measured clockwise around lattice row [hkl] * is measured clock-
the center of the film, from the W-E wise from the labeled end of the line
line (the fiducial point W has <f> = 0). y on the film (tt is the projection of
The information obtained on the ^ ont0 the film) and, as can easily be
film is now transferred to a stereo- shown, the value of if, is obtained from
graphic projection, which is related the relation tan ^ = tan tt/cos p. The
to the X-ray geometry as follows. The value of ^ is laid off, with the aid of
sphere of projection is drawn around the Vulf net, on the cyclographic pro-
the origin 0 of the reciprocal lattice; jection of (uvw) *.
the plane of projection, through 0, is A recipr0cal lattice row [hkl] *, be-
perpendicular to the direct X-ray ing a face normal> pierCes the sphere
beam; the projection (like the film) of projecti0n in the pole of the face
is viewed by the observer m the direc- {my 0ur stereographic plot of
tion of the incident X-ray beam; the reciprocal lattice rows iSj therefore,
projection point lies behind, or m the usual projection of face poles, on
front of, the projection plane, ac- which it is possible to locate the three
cording as points in the front hemi- coordinate aXes in either reciprocal or
sphere or in the back hemisphere are direct provided the cell dimen-
to be projected. The symbols of the gions are known.* Each reciprocal
points are, respectively, full or empty lattice net f ound on the film thus en.
circles, squares, and triangles. Note ab]eg ug to find the orientation of the
that the empty symbols correspond to corresponding crystal, as in the f ol-
lmes that are directed into the film. iowmg. example
For each misset reciprocal lattice A diamond crystai (a = 3.567 A)
net (uvw)* under consideration, and two forsterite inclusions (a =
the eccentricity r of the center of the 475^ b - 10.195, c = 5.981 A) gave
orientation circle, measured on the an orientation film (Fig. 88) on
film in millimeters (see above), is which we could locate the nets (112) *
converted into an angular value p by
means of the appropriate chart. The * „T .,-, , . , 7 /+ . ,. .
, TJ j. ,, V i . j.i i 1 , i With low-symmetry crystals (triclmic,
pole U Of the net plane IS then plotted monoclinic) an additional precession picture
stereographically by means of its two may be necessary to remove the ambiguity.
500
CARNEGIE INSTITUTION
Fig. 88. Interpretation of orientation photograph (/*. = 15°;MoKa:, A = 0.71 07 A) of diamond
crystal with two forsterite inclusions. Midpoint of figure is C, center of film and origin of all three
reciprocal lattices. D, a reciprocal net of diamond; nodes shown as circlets; orientation circle
drawn in solid line, with center Cd- 1/ a reciprocal net of forsterite inclusion Ii; nodes shown as
squares; orientation circle drawn in dashed line with center Ci II, a reciprocal net of forsterite in-
clusion II; nodes shown as triangles; orientation circle drawn in dash-and-dot line, with center Cn.
of diamond, (730)* of forsterite in-
clusion I, and (123) * of forsterite in-
clusion II. The eccentricity of the
three nets is measured (Table 29)
and the stereographic net showing the
poles with the cyclographic projec-
tions of their nets is constructed (Fig.
89). The three desired sets of coordi-
nate axes are obtained by standard
procedures.
The test for 1-, 2-, or 3-dimensional
fit of translations of host and inclu-
sion consists in finding agreement in
directions and translation repeats for
direct-lattice rows of the two crystal-
line species. Using the known angular
relation between the two coordinate
systems, we first calculate, for each
row of the crystal with smallest cell
dimensions, which row, if any, is sub-
parallel to it in the other crystal.
When such a row exists, the ratio of
the two nearly parallel translations
is computed. If this ratio is close to
that of two small integers, a direc-
tional relationship has been found. If
two and only two such rows exist,
epitaxy occurs; if three noncoplanar
rows can be found to agree in direc-
tion and to repeat, the host and inclu-
sion show syntaxy.
The procedure described here may
also prove useful in the study of solid-
state reactions when an orientation
GEOPHYSICAL LABORATORY
501
Fig. 89. Cyclographic projections of reciprocal lattice nets and stereographic poles of reciprocal
lattice rows for diamond, forsterite I, and forsterite II.
relation between product and starting
material is suspected. With the short
exposure times of a few minutes or
even seconds made possible by the
Polaroid film attachment for the pre-
cession camera, such a reaction could
be followed step by step.
Another related use, which is being
tested with the cooperation of Dr.
W. H. Scott, lies in the field of petro-
f abrics, where the optical study of the
orientation relations between quartz
grains in a thin section of a rock can
be supplemented by an X-ray preces-
sion study of the same thin section.
We found that, when the thick glass
slide is removed and only the thin
cover-glass supports the specimen, it
can be safely mounted in wax on the
goniometer head. The cover glass is
oriented normal to the X-ray beam
and a fi = 15° orientation picture
with unfiltered Mo radiation gives an
interpretable film after a 1-hour ex-
posure.
A similar use of this technique is
its application to the orientation
study of isotropic minerals in meta-
morphic rocks, such as a porphyro-
blastic garnet in garnet-amphibolite.
502
CARNEGIE INSTITUTION
c
o
d)
o
U
~^-
II
£
CN
CO
<
-e-
o
o
o
CN
CN
CO
— o
o
O
O
o
CN
CO
O
CO
o
>o
CN
CO
CN
CN
CN
CN
O
co
CO
co o
K CN
o
*0
in ^r
CO K">
o
o
o o
0 o
r—
O
O O
CN "tf
K
S3
O W-)
«o »o
co
CO
io -t
CO CO
iO
o
■vj- io
CN O
o
o
0 o
o o
O
r—
O O
CN IO
N
o
NQ «0
«o lO
-p I-
O K ICN
O CO «—
iO
O
o
O
o
o
CO
-<*
CN
r—
CN
^-
CO
CO
-o
lO
CO
io
SD
CO
-*
o
CO
CO
[CN
K
u
-a
0)
o
c
o
In
0)
k.
E
</>
«/>
o
L_
b>
5
o
u_
o
u_
GEOPHYSICAL LABORATORY
503
Cross-Twinning as a Mechanism
of Phase Transition
G. Kullerud, J. D. H. Donnay,
and Gabrielle Donnay
There are experimental indications
that marcasite and pyrite are not
polymorphs: Marcasite appears to
contain less sulfur than pyrite
(FeS2) and may, in addition, contain
hydrogen. H-S bonds are postulated
(G. Kullerud) . It is by no means cer-
tain that the marcasite-pyrite reac-
tion is a straightforward polymorphic
transition. In the course of the study
of this transition, experiments were
performed (by G. Kullerud) on pre-
viously analyzed natural marcasite
from Joplin, Missouri, which consists
of over 99% marcasite and less than
1 % galena. The grains exhibit typical
marcasite anisotropism, and very in-
frequent twinning is observed (in less
than 5% of these grains). This mate-
rial was heated with H20 at various
temperatures for different time pe-
riods under confining pressures of 2
and 5 kb.
At the termination of each experi-
ment the cold-seal bombs were rapidly
cooled to room temperature under
pressure. Next the gold tubes were
opened, and polished sections and X-
ray powder diffraction charts were
obtained from representative samples
of the experimental products.
A series of experiments conducted
at 2 kb and 350 °C showed that twin-
ning in one direction is common in all
marcasite grains after 5 days. A few
grains show twinning in two inter-
secting directions. After 7 days twin-
ning is observed in two directions in
all grains. Where twin lamellae inter-
sect, a highly reflecting phase oc-
casionally appears in patches 2-5 /x
in diameter, too small for positive
identification but resembling pyrite.
The frequency of occurrence of cross-
twinning increases markedly with
time. After 2 weeks all grains are
highly cross-twinned, the highly re-
flecting phase is found in many points
of intersection of twin lamellae, its
individual grains have grown to ex-
ceed 50 ii in diameter, and it can now
be identified as pyrite in polished sec-
tions as well as by its powder pattern.
After 3 weeks the products consist al-
most entirely of pyrite.
Experiments at 5 kb with marcasite
+ H20 showed similar results. These
experiments, which all lasted 120
hours, were performed at 305°, 315°,
325°, and 335 °C. Twinning was ob-
served in polished sections of all prod-
ucts, but pyrite in tiny specks at in-
tersections of twin lamellae was ob-
servable only in the products of the
325° and 335°C runs.
X-ray powder diffraction patterns
made on the products of all experi-
ments show that the relative intensi-
ties of the marcasite reflections vary
as a function of the amount of twin-
ning. For instance, the marcasite 220
peak becomes increasingly weaker
with increasing twinning frequency.
In some patterns it is missing. The
pyrite observed in polished sections is
weakly but distinctly anisotropic. The
powder patterns obtained from this
inversion pyrite differ, in relative
peak intensities, from those commonly
observed on synthetic and natural
pyrites.
The interpretation of these phe-
nomena rests on a reexamination of
twinning. Traditionally twinning has
been considered a macroscopic phenom-
enon and twin operations have been
expressed as point-group operations.
For example, the iron-cross twinning
in pyrite, FeS2 (cubic, 2/m 3), has
been defined by means of any one or
all of the antisymmetry operations in
the black- white point group 4' [2] /m 3
2'/m', which symbolizes the symmetry
of the twin; the twin operation is
often taken as reflection in (110). In
marcasite (orthorhombic, 2/m 2/m
2/m, a:b:c = 0.7622:1:1.2205 in the
504
CARNEGIE INSTITUTION
old mineralogical setting), twinning
on (110) leads to the orthorhombic
twin symmetry 2' m' 2' W 2/m,
In recent years various attempts
have been made to refine the concept
of twinning in terms of crystal struc-
ture: The twin operation thus be-
comes a space-group operation. The
difficulty lies in finding an experi-
mental check on the postulated new
twin operations.
In pyrite (PaS; a = 5.418 A;
4FeS ) and marcasite (Pmnn; a =
3.381. b = 4.436, c = 5.414 A; 2FeS2)
the structural significance of twin-
ning was pointed out by Strunz and
Tennyson (1965) on theoretical
grounds. The twin plane (110) is a
twin glide plane in both species. In
pyrite the glide component* of the
glide reflection is 3/o[il0] » m marca-
site it is y% [001]. In pyrite the "com-
position plane" is a one-layer-thick
marcasite slab. In marcasite it is a
two-layer-thick slab that approxi-
mates the pyrite crystal structure,
the dumbbell-shaped S2 groups in one
layer being not quite perpendicular
to those in the other layer.
From the figures given by Strunz
and Tennyson it is clear that in mar-
casite twinned on (110) the dumb-
bells of the second layer (in crystal
II) would have to be rotated — in pro-
jection on (001) — about 15° counter-
clockwise to form a pyrite configura-
tion with the dumbbells of the first
layer (crystal I). For twinning on
(110), on the other hand, this 15°
*The twin glide plane (110) of pyrite,
with glide component M* [110], is called an
n glide plane by Strunz and Tennyson
(1965), in violation of convention. If the
(110) plane were an n glide plane, the
glide component would be one half the
diagonal of the mesh in the (110) net,
i.e., % [111], as in PmSn, P43n, and Prion.
Tl e (110) twin plane is a c glide plane in
marcasite, but in pyrite it must be desig-
nated by a new symbolic letter, perhaps s
for side (Greek sceles), as none is available
in the present notation.
correction would have to be applied
clockwise. If crystal I of marcasite
twins on both (110) and (110) to
give crystals II and III, respectively,
and the twinning is repeated, there
will be regions II-III where lamellae
of II and III intersect. The question
arises of which crystal orientation,
II or III, will be found in such regions
of cross-twinning. The experiments,
however, show neither orientation II
nor orientation III, but pyrite. This
observation is now easily explained.
The two orientations in (001) pro-
jection would turn the S2 dumbbells
in opposite senses. The two conflicting
forces neutralize each other and the
projected S2 dumbbells in the II-III
regions end up by being exactly per-
pendicular to those in the original
crystal I, thus resulting in the forma-
tion of a tiny crystal of pyrite. Once
begun, the crystallization of the new
phase can be expected to proceed and
spread away from the intersection of
the twinning lamellae, in perfect
agreement with observation.
In numerous crystallographic stud-
ies of phase transitions, the onset of
twinning has been observed as a
"premonitory" phenomenon (Ubbe-
lohde, 1963). In some very carefully
studied transitions, such as that of a
to p quartz (Young, 1963), twinning
could not be inhibited; as a result, it
has been interpreted as an essential
feature of the transition. Indeed,
where the individuals of the twin are
extremely small, the area occupied by
twin boundaries may become so large
as compared with the volume of the
specimen that it constitutes a new
phase. We may consider that the ap-
pearance of this highly disordered
phase is what triggers the actual
transition. The twinning of marcasite
that precedes the transition to pyrite
demonstrates, in a particularly clear
fashion, the role of cross-twinning as
the transition mechanism.
GEOPHYSICAL LABORATORY
505
STATISTICAL PETROGRAPHY
F. Chayes
Negative Variances among
Theoretical Open Variables Formed
from Harker Arrays
Closure correlation; a review of
general relations. A simple method of
estimating closure correlation, based
on the assumption that an observed
closed array is a (random) sample of
a theoretical closed array formed by
closure of an underlying open parent
whose variables are uncorrelated, was
described in last year's report. De-
noting the sample closed array by U
and the underlying open parent by X,
the expected closure correlation be-
tween variables Ui and Uj is, to a
first-order approximation,
In chemical petrography we are
usually faced by the additional com-
plication that the Xs cannot be ob-
served directly, and indeed may not
even exist in a physical sense.
Whether r%$ differs significantly from
pij remains an intuitively appealing
question, but if we are to use (1)
when the Xs cannot be directly ob-
served we are obliged to estimate the
variances or2, vf, and <jt2 indirectly,
from the sample statistics.
It has been shown (see Chayes and
Kruskal, 1966, p. 694) that, if U con-
tains m elements and u% =£ % for all
pij ^ (UiUjVt2 ~~ Uia-j
E
= i \ 1 — 2iii
Uj(Ti2)/SiSh (1)
crt ££
-, (2)
where uif Uj are the observed means
of closed variables i and j, sif Sj are
their observed standard deviations,
o-j2* and (Tj2 are the variances of the
open analogues of variables i and /,
and, finally, at2 is the sum of the ele-
ments of or, the vector of open vari-
ances.
In some practical situations the
open means and variances are di-
rectly measurable, the transforma-
tion from X to U is carried through
largely as a matter of convenience,
and the derivation and use of (1) are
fairly straightforward exercises in
the generation of covariance by ratio
formation. The interdependence of
one proportion with another is then
an example of correlation between
two fractions whose numerators are
common elements in the sum which is
their common denominator, and the
question at issue is whether an ob-
served correlation, rih differs signifi-
cantly from pij, the correlation which
would be expected between Ui and Uj
if all the Xs were uncorrelated.
i = l
Ui(l — Ui)
1 — 2ui
and
or
1 - 2ui
— (Si2 — Ui2<jt2),
i = 1, ra. (3)
These results do indeed permit us to
use (1) even when X cannot be ob-
served; they require only that X
"exist" in the numerical sense, i.e.,
that it be numerically possible. Petro-
graphic data often satisfy this re-
quirement, particularly when the
closed variances are not greatly dif-
ferent from each other. In a very
important class of geochemical data,
however, the closed means and vari-
ances are such that at least one of the
open variances calculated from (3) is
negative. Now variance, whether
open, closed, observed, or theoretical,
is by definition nonnegative, and the
emergence of negative elements in &
* Small sigma indicates an element of a
matrix or vector; large sigma indicates a
matrix or vector.
506
CARNEGIE INSTITUTION
brings the statistical analysis to an
abrupt halt. Negative elements in <j
may reflect unsatisfactory or insuffi-
cient sampling, they may indicate
that the underlying- statistical model
is unrealistic, or they may be the
consequence of an unwise choice of
variables. Each of these possibilities
requires careful study, but the pres-
ent report is concerned only with the
last. In developing the argument we
ignore the very troublesome question
of sampling technique, and presume
also that the statistical model is not
inherently unsound.
Algebraic controls on the signs of
elements in <j. Given that v2 > 0, the
relations between u, s, and the signs
of the elements in <j are easily found.
A general proof that ut2 > 0 is elu-
sive and perhaps impossible, but it is
easy to show that this must be so in
the situation of interest here, in
which o-i2 < 0 for some i whose mean
value is less than 0.5. For, rearrang-
ing (3) we have, to the limit of the
approximation,
1
a,2 = — — [sr— (l-2wi)c7i2]. (4)
aC-
lf Ui < 1/2 and o-i2 < 0, it is obvious
from inspection that the right side of
(4) is always positive.
But if id < V2 and <rt2 > 0 it is in
turn obvious from (3) that <n2 will be
positive or negative depending on
whether or not (Si/u\) > <jt. Negative
elements in ej are thus to be antici-
pated among variables whose (ob-
served) coefficients of variation are
relatively small. In particular, if
there is only one negative element in
ej it will be for the open equivalent of
the variable whose observed coeffi-
cient of variation is the smallest in
the array; if there are two negative
elements in cr they will be for the
open equivalents of the variables with
smallest and next-to-smallest coeffi-
cient of variation, and so forth.
Signs of the elements in cr calcu-
lated from Harker arrays. The Har-
ker diagram is usually employed for
graphical appraisal of associations
between Si02 and other oxides in
suites of subalkaline volcanic rocks.
In such suites the variance of silica
is ordinarily considerably greater
than the sum of other variances, and
it may be shown quite generally that
err > 0 if, for some i=£j, s{2 > 2(s;2).
The argument is rather involved and
will be published elsewhere; since the
negative elements so far encountered
in cr are exclusively for variables with
means less than 0.5, the proof that
o-r > 0 given in the preceding section
is sufficient. Of the numerous Harker
arrays of this type for which calcula-
tions have so far been performed,
only one yields a cr free of negative
elements. In all of the others cr con-
tains at least one negative element, in
nearly two thirds it contains two, and
in none does it contain more than two.
Where there is only one negative ele-
ment in cr, it is invariably for the
open equivalent of A1203; where there
are two, they are for the open equiva-
lents of A1203 and Na20. As might be
expected from the argument of the
preceding section, the coefficients of
variation of A1203 and Na20 are with-
out exception smaller than those of
any other essential oxides, and when
the only negative element in cr is for
the open equivalent of A1203, the co-
efficient of variation of A1203 is
smaller than that of Na20. Our ob-
jective is calculation of a null value
against which to test each of the
Harker correlations. Since the calcu-
lation involves o-*2, the sum of the
open variances, it is meaningless
if any of the elements of cr is
significantly negative. We have seen
that in the particular circumstances
of the Harker array, negative ele-
ments will occur only for variables
whose coefficients of variation are
(relatively) very small, and in the
GEOPHYSICAL LABORATORY
507
next two sections we examine the
possibility of expressing- the chemical
analyses by a different set of vari-
ables, a set so chosen as to greatly
reduce or eliminate the incidence of
negative elements in or.*
Modifying U by linear combination.
In forming V i= \'U we consider only
\s compatible with the restrictions
encountered in most petrographic
calculation, i.e., that U is initially
closed and that a materials balance —
of weights, molar proportions, or
numbers of atoms — must be main-
tained. Suppose, for instance, that Uj
is to be incremented by a multiple of
itself, say k, at the expense of Ui, a
procedure common in normative cal-
culations. The full transformation is
and
Vi = Ui — KjUj, {ha)
Vj = (l\+K,)Uh (56)
Vk = Ujc, k^ i and /. (5c)
The means and variances of the Vic
are then precisely those of the Uh-
For Vi and V3; however, it is evident
that the means are
Vi = Ui
KiU
JWJ)
and
Vj — (14- kj)uj.
(6a)
(6b)
Subtracting each line in (6) from the
relevant line in (5), squaring, sum-
ming over the sample, and dividing
by N — 1 where N is the number of
items it contains, we have
var (Vi) = s^ 4- kj2Sj2 — 2Kjsisjrij,
(7 a)
and
var (V^ = (14 kj)2Sj2,
(7b)
* This shuffling of variables may seem
rather arbitrary but is no more so than
procedures common in other branches of
geochemistry. Ordinarily the basis for
choice is some condition on the first mo-
ments; here the condition concerns the
second moments as well.
in which s^, S;2 are the variances of,
and Tij the coefficient of correlation
between, Ui and Uj.
Multiplying the differences (5a —
6a) and (5b — 6b) by each other in-
stead of squaring them, and then
summing and dividing as before, we
have, after some rearrangement of
terms,
cov (Vi,Vj) = (14- Kj) (SiSjTij -
Kr?y2). (8)
The covariances of Vi and Vj with
any Vlc may be obtained in analogous
fashion, so that the sample statistics
of V are completely implicit in those
of U. (In small samples it will usually
be quicker to transform U to V and
calculate the statistics of V directly,
but in large samples the indirect so-
lution may save much copying.)
Recalling that the transformation
is being used because some of the co-
efficients of variation in U are small
enough to generate negative elements
in (j, we note (1) that the coefficient
of variation of Vj is the same as that
of Uj, but (2) that v% is always
smaller than ui} and var (Vi) will be
larger than Si2 providing only that
KjSj/2si > Tij. Thus the coefficient of
variation of Vi will often be larger
than that of Ui, and if r^ < 0 it will
certainly be so. Accordingly, if <ji2 < 0
in the initial calculation, a new &, cal-
culated from V rather than U, may
be free of negative elements. If, how-
ever, gj2 < 0 the transformation will
be to no avail, since the coefficient of
variation of Vj is the same as that of
If both <ii2 and oy2 are negative,
this particular transformation may
be useful for its effect on variable i
but will have to be supplemented by
additional linear combinations which
either increase the coefficient of vari-
ation for variable j or decrease the
variance of some variable or variables
other than i and /. The much more
extensive calculations that may now
508 CARNEGIE INSTITUTION
be required involve no new principles, there is thus surely reason to suspect
However complicated the transfor- a strong tendency toward "molecu-
mation(s), it will still be true that lar" association of the same general
every mean, variance, and covariance sort in rock liquids of appropriate
of V can be found from the statistics composition. If this is so, the stand-
of U, and it will usually be possible ard choice of variables in the Harker
to determine, by inspection of the array seems rather unrealistic, and
latter, whether the transformation its modification along lines suggested
will shift the coefficients of variation in the preceding section may be in
in the desired direction. In practice, order.
however, at least in the present con- It will be noted that of the four
text, it turns out to be difficult or oxides involved in this association,
impossible to determine by inspection two — Na20 and A1203 — characteris-
whether the shift will be large enough tically yield negative elements in cr
to eliminate the negative element (s) and are responsible, singly or jointly,
in (j. for all such elements so far found.
Some proposed transformations for Allotting a molar equivalent of A1203
Harker data. In discussing the alge- to Na20 is an operation of exactly the
bra of the linear combination tech- type proposed in the preceding sec-
nique it is easy to lose sight of the tion. As shown there, it may shift the
fact that no transformation should be coefficient of variation of the remain-
used unless there is substantive justi- ing alumina in the appropriate direc-
fication for it. The mere fact that tion, but the coefficient of variation of
some transformation is needed and Na2OAl203 will be exactly the same as
this one works is not enough. that of Na20. Something more will be
Although our null model presumes required if the negative open vari-
no association (zero covariance) be- ance of Na20 is to be eliminated.
tween the elements of X, it is well Similarly, the complexing of K20
known that where the data are Har- with a molar equivalent of A1203 may
ker arrays there are strong functional eliminate the negative element in cr
associations between some of the Us, contributed by A1203, but will have no
or, more precisely, between some parts effect on that contributed by Na20.
of some of the Us, in the solid state. To dispose of the latter we are forced
In rocks of the sort under discussion, to rely on a transformation which,
for instance, pyroxenes and amphi- instead of increasing the coefficient
boles are subalkaline and f eld- of variation of Na20 or its equivalent
spathoids are lacking. Nearly all the linear combination, will drastically
Xa20 and most of the K20 occur in reduce <jt2, the sum of the open vari-
feldspar, in which the ternary molar ances. Both the transformation of
ratio R20/Al203/Si02 is 1/1/6. (This K20 to or and of Na20 to ab have this
situation is reflected in the CIPW effect.
norm, in which all alkalies are cast At the present writing, & has been
provisionally as R2OA1203 if (Na20 computed for six well-known Harker
+ K20) < AL03 and finally as (ab arrays after each of the following
4- or) if sufficient Si02 is present. In transformations:
Na20 to NaA102,
the rocks under discussion both con-
ditions are satisfied.) There is good
evidence that something very close to K20 to KA102,
the 1/1/6 molar ratio may be main- XT ~ -, T^ ^ j. tvt aia 1TrA1A
tained even in hydrothermal trans- Na2° and K2° to NaA1°2 and KA1°2>
port (Morey and Chen, 1955) and Na20 to ab,
GEOPHYSICAL LABORATORY
509
D
D
X
O
E
£
o
CO
o
>
e
o
>
C
o
a
O
O
D
<D
z
u
c
0)
c
o
CO
LLS
—J
QQ
<
o
■*-
a
V
-E
CM
£
b
to
<u
>■
c
a
o
i_
<
o
-*
i_
'</»"
a
a.
E
o
c
**™
<D
^^
t/>
£
D
— i
'to"
(1)
£
-*
D
o
0
u.
U
i.
*D
£
£
■#M
E
o
(1)
^
U_
<u
"e"
E
O
o
«/»
*u
-i£
E
o
^
<
T3
D
D
-^
a
c
D
O
o
■a
<
E
X
D
o
U
<J
?
a.
E
#o
o
£
o
«4-
«/>
E
O
i_
1-
>o ^o *o ^O CO «—
o
(I)
O
E
E
E
o
o
O
E
E
E
o ©
E E
o o
-O E E
O D -Q
E E
« ° °
-Q E E -Q
D -Q -O
E
o
O E
O D -Q
O
E
o
-Q -Q E
-Q U -Q
<M
o
2 ' z
2 1 £"2
O £0*0
OJ _5» CM CM
t
CM O
cO A
o
CM
Z o
E
D
CM « <M
^ Z ^
^ z ^
o
E
|e
'5
O
cm
< .
O °
<D T
o O
to . _
to OO
<u cs
jc co
E^
•go
a <
oO
._ *M
.•= ^
"O
■g'1
2 «j
**- i-
a»
.O.E
E* D "7^
O O CM
1 aO
£ E %
o u .
**- — . 4-
o 9 J?
■fc ^o
0
oQ
< D
_z
•|o
Q) CM
*^ D
oz
O o
.is
9 =
<u o
a a
o o
<D <D
,_ •- CM
is ii
7i O
G>
OO — o
m- ■£ E
o §>c
*- o —
E ^ ■-•
a> ^
^■* =
IT i-
<D O
U
c _
o; —
a o
o a,
w U W rt
H- «+- «-t- O
o o
*■ «+-
O O
vv
CM CM
b b
II II
o| §
E o
V o E
^ o
b P
a
*■ o
II o £
510
CARNEGIE INSTITUTION
K>0 to or,
Na_>0 and K.O to ab and or.
The occurrence of negative elements
in each of the resulting: o~s is shown
in Table 30. The negative element in
cr for the open equivalent of AL03 is
readily eliminated by the conversion
of alkalies to aluminates and in five
of the six examples the conversion
of Na.O alone is sufficient for this
purpose. But no transformation based
solely on alkalies and alumina suc-
ceeds in eliminating- the negative ele-
ment in (j for the open equivalent of
NasO. In five of the six examples the
or transformation alone eliminates
both original negative elements, but
in two of the five it introduces a new
one, for SiOj remaining after conver-
sion of KoO to or. No explanation is
at present available for this curious
effect, but considerable checking sug-
gests that it is not a computation
error. The most nearly successful
transformation is of Na20 to ab,
which eliminates negative elements
in a- for five of the six arrays. The
conversion of both alkalies to feld-
spar appears to introduce an over-
compensation, again tending to make
o-i2 for ab negative even though it is
not negative when only Na20 is con-
verted to feldspar.
BIOGEOCHEMISTRY
Production of Hydrocarbons from
the Organic Matter in a Recent
Sediment
R. M. Mitterer and T. C. Hoering
An instructive approach to the dia-
genesis of organic matter in sedi-
ments is to heat a Recent sediment in
the laboratory and analyze the prod-
ucts formed. This procedure gives
clues to the types of organic reac-
tions occurring under natural condi-
tions. The formation of low-molecular-
weight hydrocarbons from mild heat-
ing of Recent and ancient sediments
was reported in Year Book 62 and
Year Book 63. This year we have ex-
amined the higher-molecular-weight
normal hydrocarbons produced in
similar experiments. The formation
of isoprenoid hydrocarbons and por-
phyrins has also been noted. Experi-
ments in which octadecene and octa-
decanol have been heated with sedi-
mentary organic matter have yielded
octadecane.
About 1% to 4% of the total or-
ganic matter in a Recent sediment is
soluble in organic solvents and con-
sists of hydrocarbons and organic
compounds containing such elements
as oxygen, nitrogen, and perhaps sul-
fur. The remainder, kerogen, is in-
soluble in organic solvents and is iso-
lated by destroying the inorganic
minerals with hydrofluoric acid.
Recent marine sediment from the
San Nicolas Basin, described in the
following section of this report, was
extracted twice in an ultrasonic tank
with a mixture of benzene and meth-
anol (7:3 vol) to remove preexisting
soluble material. The insoluble kero-
gen remains in the sediment after
this treatment. The usual procedures
for isolation and identification of hy-
drocarbons— liquid - liquid extraction
(heptane: methanol), silica-gel col-
umn chromatography, urea adducti-
nation, and gas-liquid chromatog-
raphy— were followed.
About 30 ppm of soluble paraffinic
hydrocarbons on a dry sediment
weight basis was recovered from un-
heated San Nicolas sediments. Values
from nearby basins range from 10 to
130 ppm (Emery, 1960). The distri-
bution of normal hydrocarbons is
typical of that for Recent sediments,
molecules with 27, 29, and 31 carbon
atoms predominating.
To investigate the results of expo-
GEOPHYSICAL LABORATORY 511
sure of kerogen to elevated temper- rotation of the unheated sediment ex-
atures about 250-g portions of the tract was [«]D = +13.62, whereas
dried, extracted sediment, containing the extract of the sediment heated to
kerogen, were sealed into a stainless 225 °C had a value of [a]D — +0.71.
steel bomb, pressurized to about 200 The specific rotation is measured in
psig with nitrogen, and heated in a degrees of rotation/ (rim x g/ml).
constant- temperature oven to 162°, Certain organic compounds diag-
184°, 225°, and 250°C. The experi- nostic of life processes, such as iso-
ments lasted from 3 to 14 days, with prenoid hydrocarbons, are found in
the shorter runs at the higher tern- ancient sediments and petroleum.
peratures. The heated sediment was Pristane and phytane, two of the iso-
extracted, and the hydrocarbons were prenoid hydrocarbons, can be derived
separated as previously described, from the phytol side-chain of the
Table 31 lists the amounts of paraf- chlorophyll molecule. Pristane, but
finic hydrocarbons generated at the not phytane, has been reported in Re-
indicated temperatures and times. cent sediments. Because of their
The newly generated hydrocarbons branched nature, pristane and phy-
are different from those of the origi- tane do not form adducts with urea
nal sediment extract, for little or no and thus can be separated from the
odd-carbon preference is seen, and normal paraffins. The nonadducted
significant amounts of compounds fractions of the extracts from heated
with less than 27 carbon atoms are sediment contained substances with
found. The C-27, C-29, and C-31 n- similar gas chromatographic reten-
hydrocarbons, prevalent in Recent tion times of pristane and phytane.
sediment extract, are present in small The compounds were separated by
amounts in the heated samples, in- preparative gas-liquid chromatog-
dicating that the new hydrocarbons raphy, and the individual peaks were
could not be a result of incomplete analyzed on two different column sub-
extraction. Figures 90 and 91 show trates, together with pure pristane
the gas-liquid chromatograms of the and phytane. The retention times of
extractable hydrocarbons from un- the unknown compounds were identi-
heated Recent sediment and hydro- cal with those of pristane and phy-
carbons extracted from sediment tane on each column. The infrared
heated to 225 °C. spectra and the mass spectra of the
The newly formed hydrocarbons compounds agreed with those of pure
had little optical activity. The specific pristane and phytane. In a typical ex-
periment, e.g., 184 °C, pristane and
-.„._.«, f n «,.,,, phytane each comprised about 1% of
TABLE 31. Amount of Paraffinic Hydro- f, hvdrnoarhon^ forrnpd about the
carbons Generated by Mild Heating of tne UyarocarDOns lormea, SLDOUt tne
Recent Sediment same concentration as n-C-17 and
ti-C-18 hydrocarbons. Experiments at
other temperatures gave similar re-
sults.
Another group of compounds com-
monly found in petroleum is the por-
phyrins. They have not been found in
Recent sediments, though their pos-
sible precursors, pheophytins, have
been reported. Porphyrins were gen-
erated by heating Recent sediment at
225° and 250°C.
Original
Sediment
Pe
rcentage of
Total
Temperature
°C
Time,
days
Organic
Matter
0.04
162
14
0.04
184
8
0.04
225
8
0.16
225
3
0.05
250
3
0.24
512
CARNEGIE INSTITUTION
29
Hydrocarbon extract
San Nicolas Basin sediment
c
1/3
a
o
CO
o
CJ
CD
CO
Q
27
31
Column temperature 75-325 °C 47min
Fig. 90. Gas chromatogram of total hydrocarbons extracted with benzene-methanol from San
Nicolas Basin sediment. Column: 10-foot by Va-inch 7% Apiezon substrate on Chromosorb W.
Helium flow 40 ml/minute. Temperature programmed from 75° to 325°C at 4°/minute. The num-
bers at the peaks indicate the number of carbon atoms in the n-alkane.
The porphyrins were eluted from
a column of silica gel with benzene
and then purified by thin-layer chro-
matography. Visible absorption spec-
tra were obtained on a Beckman DB
spectrophotometer. Figure 92 shows
the spectrum obtained after demetal-
lation with methane sulfonic acid. The
large Soret peak at 390 m/j, is char-
acteristic of porphyrins. When ex-
cited with light of wavelength 492
mji the porphyrins fluoresced strongly
at 628 m/j..
The chemical steps leading to por-
phyrins from chlorophyll or pheophy-
tin are hydrogenation, hydrogenoly-
sis, and decarboxylation. Different
combinations of these reactions will
result in different porphyrins. The
presence of porphyrin indicates that
at least some of these reactions have
occurred.
The catalytic effect of the inorganic
minerals on organic compounds has
been suggested repeatedly and appar-
ently demonstrated in the instance of
a fatty acid and montmorillonite
(Jurg and Eisma, 1964). Hoering
and Abelson (Year Book 62), how-
ever, have previously shown that heat-
ing of Green River shale and kerogen
derived from it resulted in qualita-
GEOPHYSICAL LABORATORY
513
CD
CO
C
O
QJ
CO
CD
CD
Q
Hydrocarbons- 225 °C
San Nicolas Basin sediment
15
lJ
KjHU
23
25
27
29
1/11/
M
Column temperature 75-325°C 47min
Fig. 91. Gas chromatogram of normal hydrocarbons extracted with benzene-methanol from San
Nicolas Basin sediment heated to 225°C. Column: 10-foot by ^-inch 7% Apiezon substrate on
Chromosorb W. Helium flow 40 ml/minute. Temperature programmed from 75° to 325°C at 4° /
minute. Peak numbers indicate the number of carbon atoms in the n-alkane.
tively and quantitatively similar
amounts of low molecular weight hy-
drocarbons.
In the present work, Recent sedi-
ment and kerogen derived from it
by demineralization with HF were
heated to 250 °C in a similar experi-
ment to study the higher molecular
weight hydrocarbons formed. The
new results support the findings of
Hoering and Abelson; the yield and
distribution of hydrocarbons from
the sediment and kerogen were about
the same. The results of both experi-
ments show that mild heating is ca-
pable of releasing saturated hydro-
carbons from kerogen.
The probable precursors of the
normal hydrocarbons are normal
alkyl chains, chemically bonded
into the kerogen, and their re-
lease involves bond breakage and
hydrogenolysis. The formation of
normal hydrocarbons, isoprenoids,
and porphyrins in the heated sedi-
ment indicates that hydrogenolysis
has occurred and implies the presence
of a hydrogen donor in the sediment
that is capable of acting as a reduc-
ing agent. The following experiments
were performed to explore further
the reducing power of sediments.
A sample of sediment containing
organic matter was treated with deu-
terium oxide at 100 °C for several
weeks in a sealed flask to replace some
514
CARNEGIE INSTITUTION
500
.'• '. ; ength, m/x
400
Fig. 92. Absorption spectra in heptane of
metal-free porphyrin extracted from San
Nicolas Basin sediment heated to 225°C for
8 days.
of the easily exchangeable hydrogen
atoms with deuterium. After drying,
the sample was heated at 225 °C for
3 days in a sealed bomb. The infrared
spectrum of the newly formed hydro-
carbons had a new band at 4.3 /x, due
to C-D vibrations, indicating that
deuteration had occurred. Hoering
and Abelson had previously investi-
gated the deuteration of volatile hy-
drocarbons in a similar experiment.
In another series of experiments
various compounds capable of being
reduced were added to 100 g of dried,
unextracted sediment. The sediment
was suspended in pentane or hep-
tane, and about 1 g of 1-octadecanol,
1-octadecene, stearic acid, and lau-
rone were added to the mixture, which
was then evaporated at reduced pres-
sure. The sediment and the pure or-
ganic substance were heated at 225°C
for about 3 days.
Approximately 140 mg of octadec-
ane, the C-18 saturated hydrocarbon,
was formed from about 1 g of 1-octa-
decanol, the C-18 alcohol. Octadecene,
a C-18 hydrocarbon with one double
bond, was converted to the saturated
hydrocarbon in about 9% yield.
About 40% was recovered as unre-
acted octadecene and about 50% was
apparently incorporated into the ker-
ogen. The experiment was also per-
formed with separated kerogen con-
centrate as a possible hydrogen donor.
Octadecane was formed from octadec-
ene in about 17% yield. Again, about
50% of the octadecene was not re-
covered.
Less conclusive results were ob-
tained by heating stearic acid and
laurone with sediment. Only 2% to
3% of the stearic acid and laurone
was recovered as hydrocarbon. At
this level the indigenous hydrocar-
bons of the unextracted sediment and
those generated through heating be-
come prominent.
Identification of octadecane was
made by gas-liquid chromatographic
analysis, after paraffinic hydrocar-
bons were isolated by the standard
chemical procedures. No octadecane
impurity was detected in either the
starting octadecene or octadecanol al-
though the alcohol was found to con-
tain about 3% of other paraffinic
impurities.
These results provide further sup-
port for the view that kerogen in
Recent sediments is a major source
of hydrocarbons. Mild heating re-
leases much larger amounts of hydro-
carbons than are present in the nat-
ural sediment.
The present experiments have
given insight into the nature of some
of the reactions occurring in Recent
sediment. To form free hydrocarbons
and porphyrins, bond breakage and
hydrogenolysis or hydrogenation
must occur. The model experiments
with olefin and octadecanol and the
deuterium experiment illustrate the
reactivity and reducing power of Re-
cent sediments. Reduction is an im-
portant reaction of the sedimentary
organic matter accompanying the re-
lease of alkyl chains from kerogen.
GEOPHYSICAL LABORATORY
515
Organic Acids from the
Oxidation of Recent Sediments
T. C. Hoering
A small percentage of the organic
matter in a Recent sediment is sol-
uble in organic solvents, but the bulk
of it is a dark polymer of unknown
structure. Oxidative degradation per-
mits an estimation of the amounts
and chain lengths of normal alkyl
groups in the polymer. The necessary
procedures have been developed and
applied to a set of Recent marine
sediments, described in Table 32.
Experimental
The acids already present in the
sediments were separated first, then
the remaining organic matter was
treated with an oxidizing agent, and
new acids were generated. The mo-
lecular weight range of the acids is
too wide for a single analysis; there-
fore, the procedure was divided into
two parts. The first served to isolate
the volatile, low molecular weight,
aliphatic acids with 2 to 9 carbon
atoms. The second permitted the iso-
lation of monocarboxylic acids with
14 to 28 carbon atoms and dicarbox-
ylic acids with 4 to 28 carbon atoms.
Generalized flow sheets for the two
procedures are outlined in the follow-
ing:
1. Isolation of Steam-Distillable Acids
2.
from Sediments, Acetic through De-
canoic Acids
(A) Steam-distillation of sediment
in 6-Af sulfuric acid
Distillate in KHCOz
(B) Extraction of alkaline solution
with ether to remove neutral
substances
(C) Evaporation to give potassium
salts
(D) Regeneration of acids with 6-M
sulfuric acid, extract into di-
ethyl ether
(E) Concentration of acids in car-
bon disulfide
(F) Gas chromatography as free
acids
Residue
(G) Excess K2Cr207 added, reac-
tion, steam-distillation into
KHC03
(H) Purification, isolation, and anal-
ysis as in (B) through (F)
Analysis of Nonvolatile Acids, Sol-
vent Extraction
Soluble
(A) Solvent extraction
(B) Solvent fractionation
Free acids
(C) Silica-gel chromatographic sepa-
ration of free acids
(D) Esterification with methanol-
BF3
(E) Silica-gel chromatographic sepa-
ration into monocarboxylic acid
esters and dicarboxylic acid
esters
(F) Clathration of esters with pre-
formed urea-heptane adduct
(G) Isolation of methyl esters
TABLE
32. Surface Grab Samples Used
in Oxidation
Studies
Location
Latitude and Longitude
Station No.
Depth,
feet
Organic
Matter, %
San Pedro Basin
San Nicolas Basin
Tanner Basin
33°35'36"N, 1 1 8°27'00" W
33o0Tl5"N, 118059'15"W
32°58'13"N, 119°44'50"W
AHF 10894
AHF 10895
AHF 10896
2930
5750
5080
7
7
7
Surface sediments were taken from basins in the continental shelf off Southern California.
The samples are portions of a 1 -cubic yard Campbell grab sampler haul. An extensive discus-
sion of the sediments is given by Emery (1 960). The materia! was frozen from the time of col-
lection until it was used. Samples were dried under vacuum over sulfuric acid or were treated
with an excess of hydrochloric acid, filtered and washed, and dried under vacuum over solid
sodium hydroxide. The collections were made through the courtesy of the Allan Hancock Founda-
tion of the University of Southern California. The assistance of Dr. D. S. Gorsline and the crew
of the R. V. Valero IV is acknowledged.
516
CARNEGIE INSTITUTION
(H) Gas-liquid chromatography
Soluble polar molecules
(L) Oxidation, isolation, and anal-
ysis as in (/) through (K)
Insoluble
(I) Oxidation with KMn04 or
K,Cr,0:
(J) Extraction of acids from re-
action mixture
(K) Purification, isolation, and anal-
ysis as in (C) through (H)
Volatile acids. Aliphatic acids with
up to 9 carbon atoms can be steam
distilled from a sediment. A major
problem is the concentration, with-
out loss, of the acids in a large vol-
ume of water to a small volume of an
organic solvent for injection into a
gas chromatograph. The selection of
the proper substrate for gas chroma-
tography of free acids is critical.
These very polar molecules "trail"
badly in most columns. An effective
column packing for free C2 to C9
acids is a 7% FFAP coating (Wil-
kins Instrument Co., Walnut Creek,
California) on a Poropak Q support.
A typical chromatogram is shown in
Fig. 93.
Table 33 gives the concentrations
of free volatile acids and the acids
from oxidation. Identification is
based, so far, only on gas-chromato-
graphic retention times. The variabil-
ity in the yields of acetic acid is due,
in part, to losses by volatilization
during separation and concentration.
Acids with more than 9 carbon atoms
have vapor pressures too low for effi-
cient steam distillation.
Nonvolatile acids. The isolation and
analysis of higher molecular weight
acids is more difficult. The procedure
in outline 1 above is the result of a
great deal of experimentation. A few
crucial points will be mentioned.
Hydrochloric-acid-treated, dried
sediment samples were solvent -ex-
tracted to remove free organic acids.
It is difficult to judge whether ex-
traction is complete. Widely differing
solvents systems were used. They in-
cluded benzene-methanol, potassium
hydroxide in methanol, pyridine, and
glacial acetic acid. A black oil was
obtained. Microgram quantities of
C14-labeled palmitic acid were added
to monitor the extraction. At least
two or three successive extractions
were needed to get a good yield of
the tracer. Complete recovery was
never obtained. Subsequent analyses
(Table 34) showed that approxi-
mately the same amounts of fatty
acids were obtained by the different
solvents. The free fatty acids have
TABLE 33. Low Molecular Weight Acids Isolated from Recent Sediments3
San Nicolas
Basin
San Pedro Basin
Tanner
Basin
Cr03
Cr03
CrOs
Acid
Free Acids
O
xidation
Free Acids
Oxidation
Free Acids
Oxidation
Acetic
1810
384
1350
20,700
459
4680
Propanoic
97
539
179
1870
37
1310
Isobutanoic
13
521
36
1230
13
716
n-Butanoic
4
319
40
758
11
424
Isopentanoic
8
600
46
861
16
628
n-Pentanoic
• • •
512
24
500
1
557
Isohexanoic
« • •
79
• • •
120
n-Hexanoic
32
690
63
283
24
697
n-Heptanoic
41
574
113
124
13
628
n-Octanoic
10
530
25
104
12
522
n-Nonanoic
. . .
392
6
. . .
1
347
Total
2015
5140
1882
26,434
587
10,629
* Units are in parts per million of the organic matter in the sediment.
GEOPHYSICAL LABORATORY
517
CD
If)
o
Q.
CO
CD
O
CD
"CD
Q
Acids from Cr03 Oxidation
of Tanner Basin sediment
100 250
Column temperature 100 -250° C 47min
Fig. 93. Gas chromatogram of volatile acids from chromic acid oxidation of Tanner Basin sedi-
ment. Column: 10-foot by V&-incn 15% FFAP substrate on Poropak Q. Helium flow 40 ml/minute.
Temperature programmed from 100° to 250°C at 4°/minute. The symbols at the peaks indicate
the carbon structure of the individual acids.
apparently been completely extracted.
No free dicarboxylic acids could be
detected.
The bulk of the extracted organic
matter did not follow acid chemistry
but consisted of "polar molecules."
This material was oxidized with po-
tassium permanganate in 2-molar po-
tassium hydroxide for 1 hour. Mono-
carboxylic and dicarboxylic acids
could then be isolated from the oxida-
tion products.
The extracted sediment was oxi-
dized with potassium dichromate in
3-molar sulfuric acid or with potas-
sium permanganate in 2-molar potas-
sium hydroxide. Bonded alkyl groups
were cleaved and liberated as organic
acids, and extracted from the oxida-
tion mixture.
The three separate crude acid mix-
tures— the free acids, acids from the
oxidation of soluble polar material,
and acids from the oxidation of the
insoluble organic matter — were puri-
fied and separated.
51S
CARNEGIE INSTITUTION
TABLE 34. Free Fatty Acids Extracted from San Nicolas Basin Sediment by Various Solvents*
Solvent
Cn rhon
VU 1 WW) 1
No.
Benzene
-Methanol f
Pyridine
Methanolic KOH
Acetic Acid
12
5.2
4.0
1.5
6.3
13
2.6
2.0
• • •
3.1
14
19.3
11.4
47.4
59.5
69.3
15
5.2
1.4
7.9
11.4
12.5
16
74.2
82.8
126.0
212.0
170.0
17
5.2
3.3
7.9
10.7
12.6
18
26.1
37.9
35.6
66.5
51.3
19
2.6
11.4
4.0
6.1
1.6
20
14.1
21.4
20.2
45.0
23.5
21
3.6
2.8
6.3
6.1
2.4
22
30.8
37.0
24.5
57.3
33.0
23
8.0
5.6
10.3
8.4
5.7
24
72.6
60.0
43.2
96.3
51.3
25
14.5
7.8
9.6
11.7
3.4
26
113.4
93.6
56.0
115.0
53.7
27
19.2
9.0
8.9
10.7
28
138.5
93.7
49.5
97.5
35.5
Totals
555.1
479.1
407.4
624.9
382.2
Per cent
accounted for \
53
37
33.4
38.8
21.5
Tracer
recovery, % §
61
60
57.5
51
67
* Units are parts per million of organic carbon in sediment.
f Results of two separate experiments.
f Percentage of the total ester fraction that could be accounted for as peaks in the gas chro-
matogram.
§ The tabulated values have not been corrected for yields, based on recovery of added C14-
tagged palmitic acid.
Solvent fractionation of the acids,
shown in outline 2, step B, above, was
done by partitioning the mixture be-
tween aqueous methanol and heptane.
The acids went quantitatively into
the heptane phase. The dark polar
substances were in the methanol and
were subsequently oxidized. A chro-
matographic separation of the acids
was made on silica gel (step C) using
a mixture of 75 volume % benzene,
22% dioxane, and 3% acetic acid. A
considerable amount of nonacid ma-
terial remained on the column.
A clean separation of monocarbox-
ylic acid esters and dicarboxylic acid
esters (step E) was made by silica-
gel chromatography. The first group
was el u ted by 3 column volumes of
benzene, and the dicarboxylic esters
were eluted by 90 volume % benzene,
10% ethyl acetate. The quantitative
adductination of small amounts of
esters with straight carbon chains
(step F) is difficult. A good solution
to the problem was found by exchang-
ing esters in benzene with a pre-
formed urea-heptane adduct:
urea'heptane + ester +
heptane
urea-ester +
The equilibrium lies far to the right
and gives a high yield of pure normal
esters when used with submilligram
quantities of material. The molecular
weight range of the esters prepared
from the acids of sedimentary rocks
is wide. Only the most stable gas-
chromatographic substrates can be
used for the temperature program-
ming from 75° to over 300 °C that is
required. Apiezon L grease or type
GEOPHYSICAL LABORATORY
519
FFAP substrate, in packed columns
or in capillary columns, gave satis-
factory results.
Some typical chromatograms of
the esters of nonvolatile acids are
shown in Figs. 94 and 95. The rela-
tively simple spectrum of well-
resolved components indicates that
the separation and purification pro-
cedures are adequate. The recovery
was monitored with C14-labeled acids,
and yields of greater than 60% could
be obtained.
The structures of the individual
compounds were verified by several
techniques. The ester mixture was
chromatographed on a variety of sub-
strates, polar and nonpolar, and both
packed and capillary columns were
used. Each individual peak had the
expected retention time in all cases.
Individual esters were separated by
preparative gas chromatography in
a large column packed with Apiezon
L as a substrate. The individual,
separated esters were then co-injected
into analytical columns with an
equivalent amount of a pure ester
with the expected structure of the
unknown. A variety of substrates
was used. A single peak emerged at
the correct time. The infrared spectra
of individual, separated esters were
measured and found to be in good
agreement with those published in
the literature ; such an infrared spec-
trum is shown in Figs. 96 and 97. The
mass spectra of several separated
t
CD
(/)
o
Cl
CO
CD
CD
"CD
O
Esters of free acids
San Nicolas Basin
Column temperature IOO°-300°C 47m in
Fig. 94. Gas chromatogram of esters of free acids extracted with benzene-methanol from San
Nicolas Basin sediment. Column: 5-foot by %-inch 7% FFAP substrate. Helium flow 40 ml/minute.
Temperature programmed from 100° to 300°C at 4°/minute. The numbers at each peak indi-
cate the number of carbon atoms in the acid portion of the ester.
520
CARNEGIE INSTITUTION
c
in
c
o
Q_
en
CD
O
T3
CD
~cd
Q
Esters of dicarboxylic acids
n s
14
Cr03 on San Nicolas Basin sediment
12
10
8
k/U
6
18
20
J*
5C
22
Column temperature IOO°-350° 47min
Fig. 95. Gas chromatogram of esters of dicarboxylic acids. Chromic acid on San Nicolas
Basin sediment. Column: 200-foot by 0.018-inch capillary coated with Apiezon L. Helium flow
36 ml/minute. Temperature programmed from 100° to 350°C at 4°/minute. Symbols indicate the
carbon atoms in individual acids.
esters were measured and found to
agree with published values ; the mass
spectrum of methyl hexacosanoate
from the San Nicolas Basin sediment
is shown in Fig. 98. Derivatives of
the esters were made by catalytically
hydrogenating them to hydrocarbons
in a Beroza "carbon skeleton deter-
mination apparatus." The hydrocar-
bons were identified by gas chroma-
tography and corresponded to those
expected. The identification of the
peaks in Figs. 94 and 95 is fairly cer-
tain.
The concentrations of nonvolatile
acids from oxidized fractions of the
San Nicolas Basin sediment are given
in Tables 34 and 35. Several points
are noteworthy : Few monocarboxylic
acids with fewer than 14 carbon
atoms are isolated because of the ease
with which their esters can be lost
by volatilization during solvent evap-
oration. There is a discrepancy be-
tween the amount of steam-distillable
acids and the amount of nonvolatile
acids. The gap between 9 and 14 car-
bon acids is not bridged by these pro-
cedures. The amount of acid produced
by oxidative degradation is four to
five times as great as that which
exists free in the sediment. The acids
produced by oxidation have a higher
proportion of odd-carbon-numbered
GEOPHYSICAL LABORATORY
521
c
CD
F
(V
H —
T)
u
CD
o
CO
l_
D
r"
(/)
o
CO
()
U
D
U_l
—
■4 —
(O
o>
o
—
o
o
>,
2
r~
o>
o
^
CD
o
O
o
O
O
O
00
CD
'vT-
CM
CD
LO
<fr
- ro
c\j
O
CD
CO
cz
o
o
CD
r-
CD
in
- ^t
ro
O
C\J
6
L.
*E
o
c
o
JE
<c
c
o
c
3
o
E
_o
a
E
a
to
c
o
.E
a>
D
CQ
e
a
CO
oo _^ £
CD
> o
O ■£
UOISSIUUSUDJJ JU93 Jid^
E
3
O
k.
-u
U
(1)
E
a
to
o
0)
E
w
D
D
VI
i_
V)
D
^~
O
a
•
<+-
^O
o
o
<D
D>
~o
UL
a
522
CARNEGIE INSTITUTION
c
o
to
E
CO
c
Methyl Tetracosanoate from Extract
of San Nicolas Basin Sediment
2n + 2 = 24
Wavelength in microns
Fig. 97. An expanded portion of the infrared
spectrum of methyl tetracosanoate from the San
Nicolas Basin sediment. The section from 7.00
to 9.00 n shows the fine structure caused by
methylene wagging and twisting vibrations
(peaks numbered from 1 to 1 1 ). The number of
carbon atoms in the acid portion of the ester
is given by the relationship C = 2n + 2, where
n is the number of bands.
molecules and have a lower mean mo-
lecular weight. This is indicative of
nonspecific degradation of alkyl
chains by the oxidizing agent. Carbon
chains with more than 20 atoms are
common in all fractions. The analyti-
cal procedure does not isolate dicar-
boxylic acids with fewer than 5 car-
bon atoms.
A closer examination of the esters
obtained from the free fatty acids ex-
tracted from the sediments revealed
some interesting points. Only 20% to
40% of the weight of the esters frac-
tion could be accounted for by peaks
on the gas chromatogram. The re-
maining 60% to 80% was high mo-
lecular weight material that could not
be chromatographed. This was veri-
fied by molecular weight determina-
tions. The mean molecular weight of
the esters from the San Nicolas Basin
sediment was measured on a dilute
benzene solution in a Mecholab vapor
pressure osmometer and found to be
448. The mean molecular weight of
the assemblage of esters shown in
Fig. 94 was calculated to be 343.
Thus, the esters are composed of 32
wt % of the mixture chromato-
graphed and 68 wt % of material
with mean molecular weight 496 that
could not be chromatographed.
Although the gas chromatogram
gave no indication of unsaturated
esters, the infrared spectrum showed
a strong absorption at 10.35 jl, which
disappeared when the ester was
treated with bromine. This is a strong
indication of trans carbon-to-carbon
double bonds. It is possible that the
missing compounds are higher mo-
lecular weight polymerization prod-
ucts of unsaturated fatty acids. These
acids are major constituents of the
lipids of marine organisms but are
conspicuously absent in the sedi-
ments. Further investigation may dis-
close the contribution of such poly-
merization to the formation of the
insoluble organic matter that com-
prises the bulk of the organic carbon.
Some interesting insight into the
origin of the free fatty acids was
gained by the following experiment.
Recent sediment from the San Nico-
las Basin was exhaustively extracted
with benzene-methanol. The extracted
sediment was dried and then heated
to 225 °C for 2 days in a pressure
vessel under a nitrogen atmosphere.
The heated sediment was reextracted.
Appreciable new soluble material was
GEOPHYSICAL LABORATORY
523
i r
+£H0OO0fr2(3H0)£H0Olfr
S
o
a
o
c
^_
u
C
CO
0>
o
E
a
x
"O
CD
cu
X
en
3^
c
<-
• —
co
a>
n
^
CD
H—
(O
o
a
E
o
o
~>
i_
^
o
a>
c
a.
a
(7)
GO
CO
F
n
o
+O0i*J(3H0)£H0 6Z£ ~
+£H000023(2H0U9£
+£H000002(2H0)6£2 "
+£H0OO06,(zH0)Q2£
+£H00008l(3HO)ll£-
+£H00009I(2H0)I82
+£H0000 QI(SH0)Z92 ~
+
2H00000I(ZH0)66! 1
6-bl
+£H00009(£H0)£W-
^0 .
,gH0OO03(zH0)Z8 S — =
9Q^ ~ ~~=
.+209H£01?Z
SO
*0 —
SJ
oooooooo
oooooooo
oo f- co in *f ro cm —
|U3JjnQ U0|
E
>s
p
3
k.
k.
o
•»-
a»
U
c
o
a;
CD
a
o
CO
cd
o
_c
c
4a
H-
o
c
g
«/>
<L>
k.
o
,E
a>
u
o
X
a)
^5
CD
cd
E
</>
o
k_
jC
c
•*—
">»
*«7i
-C
o
Q.
LQ
•/>
"(U
V)
E
V)
D
**-
~o
E
o
y
D)
c
z
C
*«/»
o
c
3
^_
<J
c
a
o
O
CO
«4-
L.
o
J)
jC
U)
■+-
c
CD
E
*«/>
-C
o
•C
M-
«+-
O
"D
T
"U
0)
*0
c
U
D
0
D
t/i
ut
,
k.
■4—
X
C
c
o
E
o
•-
o
o
c
CD
E
k.
c
CD
E
c
o
O) ^2
o
*-
o
k.
</»
i_
o
*o
c
«+-
(J
c
•""
D
o
#u
c
t/»
_c
a> o
1-
O
_£Z
o x
t6
c
i
■CI CD
"E
D
«»
D -c
i_
c
o
^
o
E
M-
* —
i_
J^
o
o
O
c
k_
U
*o
o
•u
_«.
E
£
#o
CD
D
D
a
CD
a
D
O
O
•4-
c
o
J2
C
D
to
VI
k.
O
u
o
o
»
E
k.
k.
■D
a
to
o
o
>»
U
M-
^Z
O
(J
D
X
cd
'E
0)
1-
k.
o
'o
E
-C
o
m
to
•*-
CD
MM
o
JZ
>s
mmm>
>
•*-
>C
D
<A
CD
U
o
V
E
"E
N
a
u
</>
u
o
cd
1-
£
co
«/>
s
D
k.
o
o
c
o
U
U
U
1_
"o
0
a
«/>
t3
£
cd
o
c
CD
^
J2
1-
"D
o
T3
00
CD
k.
• mm
CD
Oo
3
CD
D
«/>
JZ
u
O
D)
CD
,Ja
IE
E
0
£
524
CARNEGIE INSTITUTION
TABLE 35. Monocarboxylic and Dicarboxylic Acids from the
Oxidation of San Nicolas Basin Sediment*
M<
Dnocarboxylic Acids
Dicarboxylic Acids
Soluble
Insoluble
Soluble
Insoluble
Carbon
Polar
Organic
Carbon
Polar
Organic
No.
Material
Matter
No.
Material
Matter
12
9.3
14.8
4
1.9
• • •
13
20.6
24.6
5
2.8
. . .
14
61.2
145.0
6
7.1
. . .
15
71.7
80.1
7
11.6
. . .
16
136.5
322.0
8
23.5
0.5
17
87.8
78.5
9
24.3
1.22
18
98.7
136.0
10
21.8
4.2
19
49.5
36.1
11
21.8
6.2
20
66.0
65.0
12
17.6
7.2
21
37.2
27.1
13
15.7
8.3
22
65.1
62.1
14
13.9
10.4
23
27.6
17.3
15
10.1
9.6
24
52.3
66.7
16
9.2
10.5
25
20.6
6.8
17
9.0
8.8
26
65.7
35.1
18
10.2
9.1
27
23.5
2.1
19
9.9
9.4
28
81.6
10.7
20
15.6
9.8
Totals
974.9
1130.0
21
20.4
8.4
22
29.0
6.6
23
29.4
3.2
24
33.7
2.9
25
26.8
3.2
26
23.1
2.9
27
15.3
0.7
28
10.0
0.5
Totals
413.7
123.7
* Units are in parts per million of the organic carbon in the sediment.
isolated. This was fractionated as de-
scribed previously, and the acids were
isolated and analyzed. New, free acids
were generated from the normal alkyl
chains of the insoluble organic mat-
ter. The concentrations are shown in
Table 36. It is interesting that new,
saturated, normal hydrocarbons were
formed at the same time. A surpris-
ing difference between the acids that
exist free in the original sediment
and those generated on heating is
that there is a much lower relative
abundance of acids with an odd num-
ber of carbon atoms in the second
case.
Discussion
The results confirm previous ob-
servations that oxidation converts
some of the organic carbon in a Re-
cent sediment to acids. The amounts
obtained are small — only 1% or less
is converted. The question arises as
to how well the acids reflect the struc-
tural units in the insoluble organic
matter.
The organic carbon in a Recent
sediment has been drastically altered
from the organic substances of living
organisms in the water column above
it. Its radiocarbon age is only a few
thousand years, but it has been trans-
GEOPHYSICAL LABORATORY
525
TABLE 36. Fatty Acids Produced by Heating
San Nicolas Basin Sediment at 225°C
for 2 Days*
Carbon No.
Concentration
12
13
• • •
14
13.6
15
2.2
16
41.2
17
2.2
18
13.7
19
• • •
20
4.7
21
• • •
22
77
23
• • •
24
8.2
25
26
3.8
27
• • •
28
1.9
Total
99.2
* Units are in parts per million of the total
organic carbon.
formed into a more stable assem-
blage. The oxidation methods needed
to cleave fragments are fairly drastic,
and the results indicate that consid-
erable secondary oxidation of pri-
mary fragments occurs. The high
yields of low molecular weight acids
and the nearly smooth distribution of
dicarboxylic acids indicate a random
attack of the oxidizing agent on car-
bon chains.
Oxidations with chromic acids and
potassium permanganate do not pro-
ceed smoothly to organic acids but
may stop at stable substances, such
as ketones, that resist further oxida-
tion. Long, straight-chained com-
pounds with other functional groups
are probably present.
The results suggest that the bulk
of the organic carbon does not exist
as alkyl groups. It is likely that it is
in the form of five- or six-membered
ring structures, with oxygen, nitro-
gen, or sulfur in some of the rings.
They would give oxalic acid or carbon
dioxide on oxidation.
The presence of acids with carbon
chain lengths greater than 20 is sig-
nificant. Oxidation will not synthe-
size them from smaller molecules. In-
asmuch as the free acids in sediments
contain many such structures, and
since additional ones are released by
mild thermal treatment, these long
carbon chains are structural units of
the sediments and are not artifacts
of the analytical procedure.
The fatty acids of marine phyto-
plankton contain normal alkyl chains
with mainly 16, 18, 14, and 20 car-
bon atoms — the first two predominat-
ing. Acids with more than 20 carbon
atoms or with an odd number of car-
bon atoms are not common. The
waxes of certain higher plants and
animals contain very long-chained
acids, but such waxes are not well
known in marine organisms.
The origin of the alkyl groups in
sediments with more than 20 carbon
atoms is puzzling. The predominance
of molecules with an even number of
carbon atoms shows that they are
probably biological products. The ob-
served distribution, however, would
have to be the result of a large, selec-
tive preservation of scarce molecules.
A careful search for high molecular
weight acids in marine organisms
might resolve this possibility.
The synthesis of long-chained acids,
starting with the common Ci6 and Ci8
acids of cells and other reactants, can
be accomplished in the laboratory,
but such synthesis would be unex-
pected in the environment of a Recent
sediment.
Unsaturated fatty acids with 16
and 18 carbon atoms can polymerize
to higher molecular weight material,
but such polymerization does not
yield linear structures. Branched and
cyclic structures are formed.
The new, improved analytical pro-
cedures developed this year have
pointed out clearly some interesting
features of simple organic structures
526 CARNEGIE INSTITUTION
in a Recent sediment. The procedures was placed in the amino acid ana-
can now be used to explore several lyzer.
possible modes of formation of long This instrument is based on con-
carbon chains. ventional design, but it has been
modified to attain high sensitivity for
Racemization of Amino Acids in sma11 samples. Output of the spec-
Fossil Shells trophotometer is monitored by a re-
corder and digital integrator. In the
P.E.HarcandP.H.Abelson present experiments the apparatus
The proteins of all living forms was operated at somewhat reduced
consist principally of L-amino acids, sensitivity. Accordingly, 10~8 moles of
The universal use of the one stereo- amino acid resulted in 50,000 counts
isomer is a phenomenon of special on the integrator. It is possible to
concern to persons interested in the measure a peak of 200 counts with a
origin and evolution of life. Why was probable error of 50 counts. Because
one isomer chosen, and what factors of this sensitivity and because of the
determined the choice? specificity and effectiveness of the L-
Present views of processes involved amino oxidase in destroying com-
in the origin of life emphasize a pe- pletely the L isomer of a number of
riod of abiologic synthesis during the amino acids, the appearance of
which amino acids would accumulate as little as a few per cent of D isomer
in a primitive ocean. It is therefore could be detected. In comparing the
of interest to study changes in amino amino acid content of solutions be-
acids with passage of time. The pres- fore and after enzyme treatment, gly-
ent study was devoted to determining cine constituted an ideal internal
the optical configuration of the amino standard, since it is inert to the var-
acids in Recent and fossil Mercenaria ious enzymes used in this study.
shells. Overall, the combination of amino
The shells were dissolved in HC1, acid analyzer and enzyme digestion
and proteins or peptides were con- provides a sensitivity for detection of
verted to free amino acids by hydrol- racemization that is at least three
ysis in 6 N HC1. Calcium was pre- orders of magnitude better than a
cipitated with HF, and the superna- typical electronic polarimeter.
tant solution was dried in vacuo. After The total amount of amino acids
adjustment of the pH, solutions of present in fossil shells decreases with
the amino acids were treated with age of the fossil. In Table 37 the
enzymes that specifically attack either amino acid composition of each sam-
the L or D forms, and destroy them, pie is indicated in residues per thou-
The enzymes employed included D- sand as well as the total micromoles
amino oxidase, L-amino oxidase, and per gram of shell. The distribution of
L-glutamic decarboxylase. Behavior amino acids changes with time, re-
of the enzymes was thoroughly tested fleeting different reaction rates for
with standard solutions of known the disappearance of each amino acid,
content and of known amino acid con- As pointed out in last year's report,
figuration. After digestion of the some amino acids not found in Re-
samples with enzymes, aliquots of the cent shells are present in fossil shells.
mixture were treated with enough In Table 38 data are summarized
25% trichloracetic acid to precipitate on the optical configuration of the
the enzymes and to give a final p*H. amino acids, which are present in sig-
of about 2. The mixture was centri- nificant quantities and which react
f uged, and the supernatant solution with the specific enzymes used in this
GEOPHYSICAL LABORATORY
527
TABLE 37. Amino Acid Composition in Residues per Thousand
for Inner Layer of Mercenaria Shells
Recent,
total
Upper Pleistocene
Total
Insoluble
Miocene,
total
Aspartic acid
122
133
Threonine
73
47
Serine
89
38
Glutamic acid
87
109
Proline
111
118
Glycine
93
88
Alanine
75
110
Cystine
25
4
Valine
46
65
Methionine
14
13
Alloisoleucine
* • •
8
Isoleucine
23
25
Leucine
32
43
Tyrosine
32
44
Phenylalanine
22
31
Ornithine
8
Lysine
71
68
Histidine
21
10
Arginine
38
37
Ammonia*
(62)
(108)
jU,M/g shell
16.1
6.5
142
35
38
<1
29
<1
105
139
111
120
67
104
56
166
25
62
127
15
10
1
42
33
34
47
78
56
38
41
52
5
9
77
46
14
79
m #
(60)
(780)
1.1
* Ammonia values not included in total residues.
study. The results show that complete
racemization has occurred in the Mi-
ocene specimen, and appreciable rac-
emization has occurred even in the
Upper Pleistocene from Wailes Bluff.
Data from the insoluble fraction from
Wailes Bluff indicate that some race-
mization occurred while the amino
acids were still bound in peptide
linkage in an insoluble state. As ex-
pected, all the amino acids in the
Recent shell are of the L configura-
tion.
From the experimental results
with specific enzymes it appears that
part of the L-isoleucine originally
present is converted to D-alloisoleu-
cine and that no appreciable amounts
of Z)-isoleucine or L-alloisoleucine are
formed.
Alloisoleucine is not a protein
amino acid. It is not present in Re-
TABLE
38. Per
centage
of
D
-amino
Acid
Isomer in Amino
Acids
from
Sr
.ell
s of
Mercenaria
Upper
Upper
*
Recent,
P!<
*istocene.
Pleistocene,
Miocene,
total
total
insoluble
total
Glutamic acid
<5
22
11
47
Proline
0
31
19
52
Alanine
0
40
10
51
Valine
0
28
5
52
Alloisoleucine
...
98
100
95
Isoleucine
0
0
0
<5
Leucine
0
26
5
48
Tyrosine
0
25
16
50
Phenylalanine
0
25
8
49
528 CARNEGIE INSTITUTION
cent shells but is present in fossils, metric properties is required to sep-
The relationship of L-isoleucine and arate them. In a prebiotic world it
D-alloisoleucine is intriguing: Isoleu- would be much more feasible to sep-
cine has two asymmetric carbon arate alloisoleucine and isoleucine
atoms, and as a result there are four than to separate racemic pairs of
possible isomers. D- and L-isoleucine amino acids such as alanine,
and D- and L-alloisoleucine. Racemi- The apparent characteristic time
zation of L-isoleucine involves the for racemization is of the order of
amino carbon and yields D-alloisoleu- 105 years. Enough racemization
cine. Racemization involving the sec- should occur in shorter periods so
ond asymmetric carbon atom proceeds that it should be possible to study the
very slowly, if at all. We have not process in specimens 30,000 years old
detected significant formation of D- that can be dated by C14. The rate of
isoleucine or L-alloisoleucine from L- racemization is affected by tempera-
isoleucine or D-alloisoleucine under ture. Laboratory tests conducted at
circumstances in which most amino 160°c for 1 week showed that amino
acids are completely racemized. ac^s m shens were racemized in that
Isoleucine and alloisoleucine behave -peri0(j
differently on cation-ion-exchange The;e regultg indicate that for
columns; alloisoleucine is eluted more mogt rf ^ amino .^ ^ mixture
reachlv than isoleucine. This indicates „ , - , , .
, , , . -, . . , -, u u. of stereo forms accumulating in a
that isoleucine is more hydrophobic. .... t -, , . TJ!
Thus, the attempted racemization of P™utive ocean would be racemic If
isoleucine results in products easily s<™e unknown process presented a
separable by nonbiologic processes. sll«*t advantage to one form or
In contrast, when amino acids such another, racemization would tend to
as alanine racemize, the physical- vitiate the advantage. On the other
chemical behavior of the products is hand> once a decision was made to
almost identical. They are equally hy- use the L form, supplies of it would
drophilic, and they cannot be sepa- be replenished by racemization from
rated by ion-exchange methods. An the D form as the L isomer was used
adsorption surface possessing asym- up.
GEOCHRONOLOGY
T. E. Krogh, G. L. Davis, L. T. Aldrich* and S. R. Hart,* with A. Stueber*
Geological History of the Grenville the rocks are in fact much older. In
Province our earlier work we have determined
Age measurements this year pro- a^e values between 1500 and 2000
vide the basis for a major revision of m-v- for rocks from this area usm£
the geological time sequences in the the whole-rock rubidium-strontium
Grenville province. The real ages of method. Now a major proportion of
the rocks in this area are uncertain our effort is directed toward evalu-
because age determinations on min- atm£ the effect of the 900-m.y. event
erals yield values of about 900 ± 100 on th^ whole-rock system. Certain m-
m.y. (Geological Survey of Canada, novations in sampling methods have
1960-1966), whereas certain other made it possible for us to draw defi-
geological arguments suggest that nite conclusions regarding the volume
of the rock that remained a closed
* Department of Terrestrial Magnetism. system with respect to migrations of
GEOPHYSICAL LABORATORY 529
the dating elements during the 900- shown on the diagram represent a
m.y. event. Most of the rocks in the special rock type and will be dis-
region are paragneisses and have cussed later. The relatively small
characteristics indicative of having standard deviation of 16 m.y. in the
undergone deep burial, intense heat- slope of the line through this series
ing, and plastic deformation. It has of plotted points has considerable sig-
been commonly assumed that these nificance both in the evaluation of the
characteristics were established dur- assumption of the whole-rock method
ing the 900-m.y. event. Through ap- and in a geological sense. An expla-
plication of our sampling techniques nation of this significance and of the
we have been able to show that these meaning of the mineral data also
characteristics were established dur- shown in Fig. 99 will require a di-
ing an earlier metamorphism about gression into certain fundamental
1500 to 1800 m.y. ago. concepts of Rb-Sr dating. In the
Our investigation has been twofold, whole-rock dating method represen-
In certain cases we have studied tative samples split from 5 to 50 kg
smaller and smaller details in order to of rock are analyzed for total Rb and
evaluate the dating technique. In Sr, as well as for the strontium iso-
other cases it has been necessary to topic composition. The age is deter-
measure widely spaced samples to mined by comparison of the relatively
learn about regional geological pat- small amount of radiogenic Sr87 gen-
terns. In detailed studies of two gran- erated in the rock with the amount
ite bodies a good fit of points to iso- of radioactive parental Rb87 present
chron plots for 1500 and 1700 m.y. in the rock. Because we have no way
was accomplished. This establishes of directly measuring the amount of
the occurrence of two discrete events Sr87 present when the rock was
and eliminates the possibility that formed, the assumption is made that
these ages were the result of a at that time the abundance of Sr87 rel-
younger, say 900-m.y., event acting on ative to one of the stable Sr isotopes
rocks with a much older true age. Our (Sr86 by convention) was the same
studies of widely spaced samples com- in all the samples. The assumption is
plement the more detailed ones, as we verified if the analytical data, ex-
have found evidence that two granites pressed on a plot of Sr87/Sr86 versus
in other parts of the region were em- Rb/Sr, lie on a straight line,
placed about 1500 m.y. ago and that The geological significance of a
another major granite body formed high degree of fit to a straight line
about 1700 m.y. ago. Similarly, stud- lies in the implication that all sarn-
ies of paragneiss samples collected pies did in fact contain isotopically
from a single outcrop yield age values identical Sr at one time. The isotopic
for the most part between 1500 and composition of the initial strontium
1800 m.y., as do samples collected is shown at the intercept of the line
over much of the northwest Grenville with the vertical axis. Any argument
area and over the eastern Grenville that is advanced to suggest that a
area along the St. Lawrence River, particular granite is older than the
The French River area as an ex- indicated isochron age must include
ample of the application of the some proposed mechanism by which
method. Measurements on whole-rock the isotopic composition of Sr can be
samples from a single granite body in made identical in all parts of the
the French River area of Ontario sampled body. This requirement is
(Fig. 99) yield a whole-rock isochron most severe when a vast volume of
age of 1725 m.y. The circled points granite acts as a buffer to external
530
CARNEGIE INSTITUTION
; -?:
: -;•; -
z 76:
£ 0.750
CD
C/)
0.740
0.730
0.720
0.710
0.700
/
ONT. 4-0.WR ■
APPARENT ISOCHRON 1725* 16 my. /y _
W 170-3
0.2
04 0.6
Rb/Sr
0.8
1.2
Fig. 99. Isochron plot for the French River, Ontario, granite. Solid line is the whole-rock iso-
chron. Broken line is the mineral isochron.
chemical and physical effects. Thus,
if a major granite mass can be shown
to have a high Rb/Sr ratio over an
extensive region, it is this entire re-
gion that must be purged of radio-
genic strontium in order to eliminate
traces of an earlier history. In con-
trast to major volumes of whole rock,
minerals need only to exchange their
strontium over a distance of milli-
meters in order to lose all specific
trace of a past history.
A broken line joins the whole-rock
data point of sample Ont-4-0 to the
data point for apatite in this rock.
The biotite data point plots far to the
right of this diagram. In this case
the data for the apatite, whole rock,
and biotite lie approximately on a
straight line. Data points for other
minerals in this rock, if analyzed,
should lie on or near the same line,
because apatite and biotite are prob-
ably the least and most reactive min-
GEOPHYSICAL LABORATORY 531
erals in the rock. This relationship mineral equilibration or during an
establishes that the minerals and earlier event.
whole rock all achieved an identical In the French River granite nar-
isotopic composition at a time indi- row (5-10 cm) amphibolite dikes
cated by the slope of the line through occur at one outcrop area. The vari-
the data points. The Sr87/Sr86 ratio ation of Rb and Sr in a suite of
of the strontium common to all min- samples cut from a section taken per-
erals was 0.738 (Fig. 99), and the pendicular to a dike shows that the
time of establishment of this isotopic Rb content drops to half and the Sr
composition was about 1000 m.y. ago. content doubles along the edge of the
The slight difference between the apa- dike (0-2 cm) over that present in
tite- whole-rock age value and the bio- the granite 7-15 cm from the dike.
tite- whole-rock age value probably Isotopic data for these samples (cir-
indicates that biotite continued to lose cled points, Fig. 99) demonstrate
radiogenic strontium after the apa- that complete isotopic mixing did not
tite had ceased to gain radiogenic occur between these samples at the
strontium. Any mineral bias intro- time of late mineral recrystallization
duced in sample splitting will move (1000 m.y.). These Rb and Sr vari-
the data points up or down this min- ations developed either during an
eral isochron and will thus introduce earlier metamorphic event or during
scatter into the whole-rock data. the intrusion of the dike.
The occurrence of mineral iso- French River paragneiss. Chemi-
chrons demonstrates intergranular cally layered rocks, which are thought
migration of strontium but it tells to be of sedimentary origin but which
us nothing regarding the distances now are intensely recrystallized and
involved during these migrations, deformed into paragneisses, are corn-
There is no requirement, for example, mon in the French River area. One
that the isotopic composition of Sr outcrop exposed approximately 2
was ever identical in all grains of a miles south of the granite was suit-
single mineral in one rock. ably layered for studying isotopic
Evaluation of the size of the closed migration. Specifically, we hoped to
system. To evaluate the distances in- determine the time at which the
volved in isotope migration we have gneiss formed from its sedimentary
obtained samples in which the min- precursor. Block samples represent-
erals were concentrated into discrete ing a continuous section of approxi-
layers instead of being in a three- mately 18 inches of rock perpendicu-
dimensional mosaic. Actually, mono- lar to the layering were obtained,
mineralic layers are not easily ob- Within each layer of granitic gneiss
tained; however, layers with finite the Rb concentration drops by about
differences in their Rb/Sr ratio are a factor of 3, whereas the Sr content
all that is required. We have found increases by about 50% as the edge
that in many cases chemical inter- of each biotite-amphibolite layer is
actions have occurred between rocks approached. The abrupt drop in Rb
with different chemical compositions and increase in Sr correspond to the
to form layers with intermediate or occurrence of a plagioclase-rich mi-
contrasting characteristics. Insofar crocline-free layer adjacent to each
as these can be shown to be reaction amphibolite layer. Further studies of
zones formed during intense heating these variations are in progress, but
of the region, we are able to deter- it is a reasonable conclusion at this
mine whether the chemical interac- point that we are examining the
tions occurred at the time of the last products of major chemical interac-
532
CARNEGIE INSTITUTION
tions that occurred at the time of
formation of this gneiss. With this
degree of chemical interaction it is
a reasonable assumption that the iso-
topic composition of Sr would be the
same in all parts of this section at
the time of these reactions. If the
reactions occurred during the 1000-
m.y. event, then data points for all
parts of the gneiss would lie on a line
with a 1000-m.y. slope. If, however,
the gneiss formed during an earlier
event, further modification of the iso-
topic relationships during the 1000-
m.y. event are to be expected. As
shown in Fig. 100, samples 1A and
2Z>, each approximately 1-cm wide,
are located at an abrupt Rb/Sr dis-
continuity. In Fig. 101 data points
for these two samples define a line
with a slope comparable to that for
minerals in the rock. The other sam-
1 , • ' i ' , i
WJW tc
?2 i/>s y//s/si mw/y/, x-
Fig. 100. French River, Ontario, paragneiss. Diagram showing changes in concentration of
rubidium and strontium along a section normal to the layering. The vertically dashed layers are
grantitic gneiss; the crosshatched layers are biotite-amphibolite.
GEOPHYSICAL LABORATORY
533
0.760
0.750
0.740
10
CO
c/5
CO
u5
0.730
APPARENT ISOCHRON
1622 my. ±. 70 m.y.
0.720
0.710
0.700
COMPOSITE ^2AB
31
I L
0.2
0.4 0.6
Rb/Sr
0.8
Fig. 101 . Isochron plot for the French River parctgneiss.
pies of gneiss, however, do not lie on
a single line with this slope but
rather lie approximately along the
isochron transferred from Fig. 99 for
the French River granite. An arith-
metic composite sample (1A, 2AB,
37) also appears to lie on this line,
suggesting that the enclosed volume
may have been closed to migrations
of the dating elements during the
1000-m.y. event. We conclude that the
gneiss formed during an earlier meta-
morphism and did not form during
the late heating event demonstrated
by the apatite in sample 66-88. If
we propose limited isotopic mixing
during a time when the major phases
remained stable, we should expect
low points on the Rb/Sr profile of
Fig. 100 to gain, and high points to
lose, radiogenic Sr87. Thus samples
2AB and 1C do lie below samples
with comparable Rb/Sr ratios on Fig.
101. Similarly, the samples occupying
the valleys in the profile of Fig. 100
do show an increase in radiogenic
534 CARNEGIE INSTITUTION
Sr over that which the higher sam- two dominant granite types present
pies would have had 1725 m.y. ago, are shown in Fig. 102. Samples from
One could also pass a mineral iso- three separate outcrops of the felsic
chron through samples SB and 3D as granite (Ont-1, 65-165, 65-162-0)
if they had locally obtained an iden- define the same line as the three sam-
tical isotopic composition approxi- pies from different parts of the bio-
mately 1000 m.y. ago. A least-squares tite-hornblende granite: 65-199, 201,
line through all the granitic gneiss 202. Again the minerals have under-
samples yields an age value of 1622 gone isotopic exchange long after the
± 70 m.y.: another isochron between time of formation of the rock. As
three samples of gneiss (not plotted noted on the diagram, sample 65-165
on Fig. 101) from other parts of the is a composite; it was made in the
outcrop defines a line for 1430 ± 26 field by taking samples approximately
m.y. The three high points on the iso- 5X5 X 2 cm at 1-meter centers over
chron yield age values between 1750 approximately 5 square meters of
and 1900 m.y., assuming an initial freshly blasted road cut. Results for
Sr87 SrS6 ratio of 0.704. The meta- a 5-kg sample from this outcrop, ana-
sediments clearly must be older than lyzed in duplicate last year, lie near
1725 m.y., the age of the granite that but not on the line defined by the
presumably intrudes these gneisses, other whole rocks. The implication of
The data demonstrate that a para- this test of sampling technique is
gneiss subjected to regional meta- that an average of many possible
morphism after the time of its for- open subsystems collected over a
mation may yield a variety of age large area of homogeneous material
values. Our best estimate of the time may approximate a closed system
of formation of this gneiss is ob- better than a single block sample.
tained by considering those samples Another composite, 162-0-9 (Fig.
farthest removed from variations in 102), was made in the laboratory by
the Rb/Sr profile shown in Fig. 101 combining hand-specimen-sized sam-
and by eliminating from considera- pies collected at 2-meter intervals
tion such samples as 3/ and 1A, across a freshly blasted road cut. Re-
which are susceptible to additions of suits for this composite and the 10-
radiogenic strontium. Sample 66-88 kg whole-rock sample from the same
represents about 30 kg of rock from outcrop are essentially identical. A
the thick layer 87-3; hence it and contact between the felsic granite
sample 3D probably define our best (65-162) and a biotite-hornblende
approximate age at about 1750 to granite occurs on this outcrop. A
1900 m.y. for an assumed initial ratio study of the Rb and Sr variation
of 0.704. The spread in ages possible across the contact indicates a gradual
for a single outcrop is instructive in drop in Rb from 170 to 95 ppm and
interpreting single whole-rock para- an increase in Sr from 140 to 220
gneiss age values reported later in ppm. Such a gradual variation could
this section. occur for a magmatic contact, or it
Lake Muskoka granite and other may have developed during the meta-
coeval granites. Additional work has morphism of the granites. The rocks
been completed on a granite body themselves show the effects of having
that occurs on the eastern shore of recrystallized under directed stress
Lake ^Iuskoka, approximately 150 with the development of lenticular
km southeast of the Grenville front pods, a few millimeters thick, of
(see Year Book 65). Whole-rock data quartz and feldspar or of mafic min-
for three locations from each of the erals. In this case the foliation is
GEOPHYSICAL LABORATORY
535
0.820
0.800
0.780
CD
00
an
>.
5 0.760
CO
0.740
0.720
0.700
I ' | » I ' I l I ' I '
-o-y
COMPOSITE 162
65-162-0 WR.
APPARENT ISOCHRON 1497 ± 10 my.
65-165 COMPOSITE
P
CONTACT (E)
65-199
*"" APATITE
(65-162-0
/
■ ONT. 1-2
1 ONT. I
i 65-202
CONTACT (A)
65-201
i ■ i ■ 1 . I
1.4
0.2 0.4 0.6 0.8 1.0 1.2
Rb/Sr
Fig. 1 02. Isochron plot for a granite at Lake Muskoka, Ontario.
1.6 1.8 2.0
parallel to the contact. Data for two
samples on the established Rb/Sr
gradient are shown in Fig. 102. These
were sawed from a block containing
the contact and were 10 cm apart.
Because their points do not define a
line whose slope is parallel to that of
the mineral isochron, the rock was
not isotopically mixed during the last
recrystallization (in this case 700-
800 m.y. ago). They do in fact lie,
within the error of our measure-
ments, on the line defined by widely
spaced samples from other parts of
the body. Therefore, the Rb/Sr gra-
dient of which these samples are a
536 CARNEGIE INSTITUTION
part must have been established ap- whole-rock analysis if we assume an
proximately 1500 m.y. ago. initial SrS7/SrS6 ratio. For example,
It is a reasonable assumption that for the French River paragneiss, age
the recrystallization exhibited by values between 1500 and 1900 m.y.
these rocks would be accompanied by can be calculated. Five out of six
isotopic mixing of strontium, on a paragneiss samples collected in a re-
scale of at least a few centimeters. If gion of approximately 300 X 150 km
this is true then the metamorphism lying east of Georgian Bay yield age
of these granites occurred approxi- values in this range. Three samples
mately 1500 m.y. ago. of paragneiss from the area between
In two other localities ages within the Saguenay River and Sept Isles,
lr'c or 2c'c of that obtained in this Quebec, on the southeast margin of
study have been found. Two whole the Grenville, yield age values be-
rocks from a single large outcrop tween 1550 and 1650 m.y. Three
near Round Lake, approximately 35 granitic rocks with age values be-
km southwest of Pembroke, Ontario, tween 1000 and 1300 m.y. occur in
yield two-point isochrons of 1510 this area, as well as three others with
m.y. A single sample from an outcrop age values between 1500 and 1600
south of Sandridge, Ontario, yields m.y. These values indicate that pre-
a whole-rock age of 1475 m.y., with Grenville rocks occur in the eastern
the use of an initial ratio of 0.703. as well as the northwestern parts of
Other paragneiss es and granitic the Grenville Province and that the
rocks in Ontario and Quebec. An age age patterns in the two areas may be
value can be calculated from a single similar.
STRUCTURAL GEOLOGY
Natural Slip Folds in which the the mechanism of folding is likely to
Fold Axes Nearly Parallel the have been quite different. The geom-
Slip Lines etry of many of these folds suggests
„ , , rr that the beds had little or no strength
Edward Hansen , . „ , ,. , , , ~r ,
during folding and have been offset
It has long been assumed by geol- into the folded form by slip parallel
ogists that fold axes in rocks are with the axial surfaces, somewhat as
oriented more or less perpendicular an edge of a deck of cards can be
to the direction of internal, relative folded by causing the individual cards
movement, the sliding between beds to slip past each other. Weiss (1959)
or slip across beds, that resulted in has shown from geometric consider-
the folds. This assumption is an out- ations that the axes of folds de-
growth of considerations of the ge- veloped by this mechanism of slip
ometry of sedimentary rocks that may be oriented at any angle from
appear to have been flexed into folds, perpendicular to nearly parallel with
the beds on the limbs having adjusted the slip lines. Nevertheless, though
to the folded configuration by sliding the theoretical basis has been given
past one another at right angles to by Weiss, it remains to be shown that
the fold axes. The obvious analogy is natural slip folds have indeed devel-
a flexed sheaf of papers. The validity oped with their fold axes significantly
of this assumption comes into ques- less than 90° to their slip lines,
tion, however, when it is applied to Mean slip-line orientations have
the folds in metamorphic rocks where been deduced by several independent
GEOPHYSICAL LABORATORY
537
geometric methods for three areas of
late slip folds, located in the medium-
grade metamorphic rocks of Troll-
heimen, Norway (Year Book 65).
These folds are assumed to be slip folds
for several reasons: (1) They exhibit
the geometric characteristics typical
of slip folds; they are strictly similar
in profile and nearly isoclinical, with
relatively long axial planes in profile
section and with a prominent schis-
tosity that parallels their axial sur-
faces. (2) Paths of earlier fold axes
rotated by the late folds are planar
and within 10° of parallelism with
the mean attitude of the late fold
axes as well as the axis of the next
lower-order fold on whose limb they
appear; theoretically, this geometry
would result from slip folding and
not from flexural slip (Weiss, 1959;
Ramsay, 1960). (3) The late folds
interfere with earlier folds to pro-
duce highly attenuated domes and
basins, the sides of which are within
20° of parallelism; this also indicates
that slip was the late folding mech-
anism because of the virtual impos-
sibility of tightly flexing beds across
a previous direction of folding with-
out the development of subordinate
folds and faults to accommodate the
beds. (4) Deduction of the slip-line
orientations recorded by the late
folds is also based upon the assump-
tion that they are slip folds; the con-
sistency of the results seems to sup-
port the assumption.
to -i
\
Cx
4
n=29
i — | — i — i — i — | — i — | — r
/O 20 30 40 50 60 70 80 90
fO -i
o
0
5 -
n
= 6
*
B
-i 1 1 — i — i — i — i — i — I — i 1 — i — r
O fO 20 30 40 50 60 70 80 90
tO -i
^
CH
n = 10
c
i.
i — i — i — i — i — i — r — i-
fO '20 30 <4Q 50 60 70 80 90
step Line/\ ~foCcL axe's — =*-
20 -,
fS -
\
^ fO
4
D
n = 4-5
1 — i — ' — i — ' — i — ' — i — >
fO 20 30 4-0 50 60 70 80 90
slip (cne/i -fold axis
Fig. 103. Histograms of the angle between mean slip-line orientations and fold axes of the
late slip folds in three areas, Trollheimen, Norway. (A), Area a; (B), area b; (C), area c; (D),
composite of areas a, b, and c.
533
CARNEGIE INSTITUTION
The histograms in Fig. 103 show
the angles between the mean slip-line
orientations and the fold axes of the
late slip folds. They show that the
angles between the fold axes and slip
lines spread from 0° to 40°, with the
mode in the 5° to 10° class interval.
The center of gravity of the linear
preferred orientations of fold axes in
each area has boon calculated as axis
3 of the Dimroth ellipsoid (Dimroth,
1962a, 1962b, 1963), with the use of
the program YEC-1 of Horace Win-
chell and William Scott. The angles
between the centers of gravity, so
calculated, and the mean slip-line ori-
entations are 6° for area a, 5° for
area b, and 5° for area c. In conclu-
sion, the slip lines and average fold
axes of the late slip folds in this part
of Trollheimen make an angle of ap-
proximately 5°.
Here, therefore, is a set of natural
folds for which a kinematic interpre-
tation based upon the assumption
that fold axes are formed perpendic-
ular to the direction of internal rela-
tive movement would be seriously in
error. This example indicates that
the orientations of fold axes and slip
lines of natural folds that may have
been produced by slip should be deter-
mined independently.
It is interesting to look at the kine-
matics of currently accepted models
of alpine-type mountain building in
the light of this relationship. Gener-
ally speaking, axes of the major folds
in most tectonic zones of the Alps
are approximately parallel with the
long axis of the mountain chain
(with the notable exception of the
late domal structures in the meta-
morphic core) . Geometry of folds in
the marginal, low-grade zones per-
mits the conclusion that flexural slip
was the prevailing mechanism of
folding and that the relative move-
ment within the sedimentary se-
quences during folding was directed
perpendicular to the fold axes and
the long axis of the mountain chain.
The geometry of folds in the central,
high-grade zone, however, is more in-
dicative of slip folding than flexural
slip. Nevertheless, current models of
mountain building assume movement
perpendicular to the fold axes and
to the long axis of the chain. Geolo-
gists reconstruct the stratigraphy
and tectonics of this part of the Alps
and of other, similar mountain belts,
wholly on the basis of this invalid
assumption. It would appear to be
time to collect some independent data
of slip-line orientations in such cen-
tral zones to establish a more rigor-
ous basis for the various models and
reconstructions.
Real versus Apparent
Displacement in Slip Folds
Edward Hansen and William H. Scott
Based upon the assumption that
fold axes form perpendicular to slip
lines, it is customary to consider the
amplitude of a fold in profile to be
a good approximation of the distance
that bedding has been displaced to
form the fold. As discussed in the
preceding section, however, this as-
sumption is valid for flexural-slip
folds but invalid for slip folds (cf.
Knopf and Ingerson, 1938, pp. 157-
160; Weiss, 1959, p. 92; Donath and
Parker, 1964, pp. 48-49); conse-
quently, a better measure of bedding
displacement during slip folding is
needed. Knowing the orientations of
the fold axis and the slip lines in a
slip fold and knowing its amplitude
measured perpendicular to the fold
axis (Fig. 104), we can define a dis-
placement ratio
P =
H
CSC a
where Y is the displacement parallel
with the slip lines, H is the apparent
GEOPHYSICAL LABORATORY
539
Fig. 104. The fold axis [fa], slip-line orienta-
tion [si), and amplitude (H) within the axial
surface [as) of a slip fold.
displacement or amplitude of the
fold, and a is the angle between the
slip lines and the fold axis. This
treatment assumes the fold axis to be
a fixed line of reference, but if the
fold axis were to rotate by differen-
tial slip within the slip surfaces dur-
ing folding {Year Book 65, Fig. 110) ,
the true displacement could differ
somewhat from that calculated.
Nevertheless, the simple displacement
ratio given here still gives a better
approximation than the amplitude of
a fold.
The late set of folds exposed in
Trollheimen, Norway (discussed in
the preceding section) , illustrates the
discrepancy between real and appar-
ent displacements in natural slip
folds. The average angle between the
mean slip-line orientations and the
mean attitudes of the late fold axes
was found to be 5°; therefore, the
displacement ratio P for those folds
is 11.5. Thus the displacement that
actually occurred during the develop-
ment of an average, late slip fold in
Trollheimen is an order of magnitude
greater than the displacement appar-
ent in its profile section.
Although this example is interest-
ing in itself, it introduces the possi-
bility that other groups of natural
slip folds may also have high dis-
placement ratios, a possibility of po-
tential importance to the study of
strain in metamorphic rocks.
EXPERIMENTAL TECHNIQUES
P-V-T Measurements on Hydrogen
up to 600°C and 1800 Atmospheres
D. C. Presnall
Previously, P-V-T measurements
on hydrogen were reported at 200°
and 300 °C (Presnall, Year Book 65,
pp. 415-418). This report presents
the results of a continuation of these
measurements at temperatures up to
600 °C together with a slight revision
of the previously published data at
200° and 300°C.
Figure 105 shows the new data and
the older revised data. The isotherms
at 200° and 300 °C have been changed
slightly based on a correction of the
thermocouple calibration and a re-
vised calibration of the pressure
gauge used at pressures less than 400
atmospheres. It will be noted that the
present results differ increasingly
from the empirically derived equa-
tion of Shaw and Wones (1964) as
the temperature increases. Within
the temperature-pressure range cov-
ered here, this difference causes a
maximum error of 2% in the fugac-
540
CARNEGIE INSTITUTION
WJO *( oi)(f-^)
<l> "D
^
;= c
c
D o
p
thers
haw
to
c
O
u
o to
«/»
I'S
o
D c
N
4- O
as:
a 0
**»_
</>
t 2.
o
o ST
CO o
k.
M- CD
a
0 _c
i/>
o
hose
rom
E
,]o,
•*- M-
k.
3
O 4-
«/>
O J5
O D
to
<D
k.
CN JJ
a
■♦- O
„
O O
Q.
d) to
k. <U
S^
° £
o
O =>
"~~ '
5: u
<L>
O o
3
.c
*X3 h-
O
c
1—
a
C o
E
atm
olbor
300
o
h-"
x H
»«.
c
- **r O
JD
0) O
O
3 <D°o
co O O
E
a> _c cn
E
1. *-
<j
a °
o a)
<D
3 B
£
3
200.
addy
o
>
1—
D
oO
"o
O "O
E
«- c
o o
I>
agaf
iebe
U
o
I!
O
O
>o
*t; o
o
T3
« *
c
O to
"£ o
O
-S -C
^
"O *-
o
<D -U
O
-c c
o
1— O
io
• •*
*
o
o
o <*>
o
o
J= CO
o
*t "O
o
O c
CO
«/> a
-
E -
o
k. O
o
O ,_
o
« o»
CN
8 Os
-t—
— CN
o
..
o
. o
k.
•o o
o
° 06
<o
On
<0
c
6> ^
o
>
GEOPHYSICAL LABORATORY 541
ity coefficients given by Shaw and 1.4% as the oxygen fugacity is re-
Wones. duced to that of the wiistite-iron
These data provide a basis for in- buffer,
terpreting hydrothermal experiments The problem of species in the gas
in which the oxygen fugacity is con- other than H2, II20, and 02 has been
trolled by a gas mixture. The appa- considered by Eugster and Wones
ratus described by Shaw (1963) per- (1962, p. 93). They found that rep-
mits the oxygen fugacity of a hydro- resentative solubilities of silicates
thermal experiment to be varied and oxides in water vapor would
continuously and independently of cause a negligible error in the calcu-
temperature, but the calculation of lated oxygen fugacity.
the oxygen fugacity imposed on a It is concluded that the present
sample during an experiment de- data on hydrogen, together with (1)
pends on (1) a knowledge of the fugacity values of Holser (1954) for
fugacities of H2 and H20 in mixtures water and (2) the assumption of
of these two gases and (2) the as- ideal mixing, provide a basis for cal-
sumption that there are no species culating oxygen fugacities in H2-H20
other than H2, H20, and 02 in the mixtures that is sufficiently accurate
gas phase in equilibrium with the for most purposes. For oxygen fu-
sample. At present there are no data gacities higher than those for the
on H2-H20 mixtures from which fu- magnetite-wiistite buffer, the appa-
gacity coefficients of these two gases ratus described by Shaw (1963) can
can be calculated. Alternatively, one be used with negligible error in the
can assume ideal mixing of H2 and calculated oxygen fugacity.
H20, that is, assume that the fugacity
of each species in the mixture is pro- High-Pressure, High-Temperature
portional to its mole fraction, and use X-Ray Diffraction
fugacity coefficients determined from
P-V-T measurements on the pure H. 0. A Meyer P. M. Bell, and
gases. Fugacity coefficients for H20 J' L' En^land
have been tabulated by Holser (1954), A new high-pressure, high-temper-
and those for H2 can be calculated up ature X-ray diffraction apparatus has
to 600 °C from the data presented been designed and built, with a boron
here. The error in the calculated oxy- carbide (B4C) pressure vessel. The
gen fugacity caused by deviations apparatus is particularly suited to
from ideal mixing would increase the study of either single-crystal or
with increasing pressure, decreasing polycrystalline nonquenchable phases
temperature, and increasing H2/H20 up to 1000 °C and approximately 25
ratio. At 700 °C and 800 bars the data kb. It can also be used to examine
of Shaw (1963) indicate the magni- the effect of pressure on the thermal
tude of the error at various H2/H20 expansion of various minerals. The
ratios. He found that the error in- device has been tested at ambient
creased from an immeasurable temperature by examination of the
amount at H2/H20 ratios less than compressibility of Solenhofen lime-
0.02, to 1.4% of the logarithm of the stone.
oxygen fugacity at an H2/H20 ratio This new apparatus is a substantial
of 1.8. Stated differently, the error in modification of a previous design
the calculated oxygen fugacity is un- published by Meyer in 1966. Although
measurable at oxygen fugacities the X-ray optics and pressure cell are
higher than that of the quartz-faya- similar, changes have involved inter-
lite-magnetite buffer and increases to f erence support of the cylinder in
542
CARNEGIE INSTITUTION
/^M y>^
~^vV Pillars
( ^
\ X. ' *' \
Top Section
Side Section
■* _ +/J
Rcys ^"""^ T~^
o. .
Boron Carbide/ /
Pressure Cylinder /
Steel Ring / /
Supporting /
Fiber /
Insulation
Fig. 1 06. High-pressure cell showing steel
supporting ring and slits to allow passage of
X rays.
keeping with the work of Boyd and
England (1960«). Also, the pressure
is now generated at the sample by
means of a hydraulic ram, which
allows advance and retraction on the
piston (Figs. 106 and 107). For pres-
sures greater than 25 kb it is probable
that end loading of the B4C cylinder
will be necessary.
The design of the pressure cell is
such that X-ray diffraction patterns
can be obtained by either powder or
oscillating single-crystal techniques
at 20 angles ranging from 0° to ap-
proximately 160°. The hollow cylin-
der permits utilization of linear sam-
ple geometry, which is favored for
powder X-ray diffraction. X-ray ab-
sorption by boron carbide is about
50% less than that of diamond, and
both CuK and MoK radiation may be
utilized. CuK radiation is especially
useful in the back-reflection region.
The small size of the apparatus
makes it possible to mount it on a
standard X-ray goniometer head. The
device can also be mounted in a
Weissenberg camera or on a diffrac-
tometer.
The specimen is heated by inser-
tion into a thin-walled nickel tube
that makes electrical continuity with
the upper and lower pistons. In the
case of CuK radiation the nickel
heater also acts to filter CuK/? radi-
ation.
Several tests have been made to
assess frictional effects. Comparison
of indentations made by the pistons
in steel plates with standard inden-
tations made in a test machine at
known pressures indicates that the
pressure on the specimen is approxi-
mately the same as that calculated.
Further steps to precisely determine
the pressure have involved direct
measurement of the friction. Two
identical presses were attached to-
gether so that the ram of one directly
opposed and pushed against the ram
of the other. Oil was pumped against
one ram and the pressure transmitted
through the other ram was read. In
this way the net friction of the two
rams could be measured. In a second
test the pressure cylinder, complete
with 1-mm pistons and sample, was
placed between the two opposed rams.
In this test the difference between
readings is a measure of total friction
in the pressure cylinder and in the
rams. As the friction in the rams is
known, the friction in the pressure
cylinder can be calculated. Experi-
ment has shown this to be approxi-
mately 21 % for this particular set of
pistons and cylinders. This type of
experiment, allowing direct measure-
ment of friction, is rare in high-
pressure research.
Fifteen-degree oscillation X-ray
diffraction photographs, with unfil-
tered MoK radiation, taken while
GEOPHYSICAL LABORATORY
543
Sample
Steel
Supporting Ri
Supporting
Sector(30e
I inch
To
Goniometer
Spindle
'0'
Rings
Tungsten Carbide
Piston
Boron Carbide
Pressure Cylinder
Tungsten Carbide
Base Plate
Fiber
Insulation
Fig. 1 07. Sectional view of hydraulic system and pressure cell in place.
Solenhofen limestone (calcite) was
being compressed, give results simi-
lar to those reported by Jamieson
(1957) and Davis (1964). The low
angle lines appear to diminish in in-
tensity and shift toward higher 20
angles with increasing pressure.
As these were preliminary photo-
graphs and not intended to give high
accuracy, a flat plate camera was
used in the front-reflection region.
However, if the measured friction is
taken into account, the data obtained
(Fig. 108) are in good agreement
with those of Bridgman (1925, 1938) .
Utilization of the back-reflection re-
gion in future experiments will pro-
vide more accurate cell parameters.
544
CARNEGIE INSTITUTION
co
o
O
a> 8 —
CO
CO
CD
•000
0.990
0.980
[■-«]
Fig. 103. Nominal pressure calculated from gauge readings versus the compressibility function.
Present uncertainties represent the extreme of measurements. Bridgman's uncertainties have been
estimated from his 1938 publication.
GEOPHYSICAL LABORATORY 545
Pressure Effect on the Platinum ety of pure metals and alloys. These
versus Platinum 10% Rhodium included, however, only a few of the
Thermocouple metals commonly used for thermo-
couples today. Moreover, his experi-
P. M. Bell, J. L. England, and ments extended to only about 100°C,
F. R. Boyd whereas data are now needed for
In order that a thermocouple may much higher temperatures,
be in close proximity with the sample In the present study, apparatus
in high-pressure, high-temperature (which will be described in detail in
geochemical experiments, the thermo- a future report) similar to that used
couple junction must itself be sub- in Bridgman's (1918) "one-wire" ex-
jected to pressure. Thermoelectric be- periment was employed. In this case
havior is affected by pressure, and a platinum versus platinum 10% rho-
questions have been raised as to the dium thermocouple was placed in a
accuracy of this type of temperature pressure vessel. The end of the vessel
measurement. In the present report containing the junction was heated
a technique originally devised by in a precisely controlled electrical-
Bridgman (1918) was employed to resistance furnace so that the t em-
determine absolute values for the perature could be independently
effect of pressure on thermal emf of measured in a small cavity located
a platinum-versus-platinum 10% rho- on the outside of the vessel. At the
dium thermocouple. cool end the leads were brought out
Pressure affects differently the through standard pressure seals lo-
thermocouple quality of each metal cated in an ice bath. An emf differ-
or alloy used in a thermocouple, and ence between the inside and outside
the total effect is the algebraic sum thermocouples was monitored with a
of these different effects. If it is de- potentiometer-electronic detector ac-
sired to measure the total effect on curate to 10~9 volts. Temperature was
the thermocouple, the experiment also monitored on both sides of the
must be designed to contain the fol- pressure seal to assure that no gra-
lowing features. There must be a dient existed in that region,
temperature gradient along the ther- Measurements were made at 20°,
mocouple from some high tempera- 185°, 305°, and 509 °C at various
ture T at the hot junction to 0° at pressures (Fig. 109). Figure 110
some distance from the hot junction, shows a plot of these data in terms of
The wires must be at pressure along temperature correction as a function
this thermal gradient, but the pres- of pressure and temperature. The ini-
sure P must be constant. The temper- tial inflection is reminiscent of that
ature must be kept constant at 0° demonstrated by Hanneman and
where the thermocouple leads are Strong (1964) in the plot of their
brought out of the pressure vessel, relative measurements at higher
from pressure P to atmospheric pres- pressures.
sure. It must also be possible to The apparatus employed in the
measure P and T accurately. These study is capable of achieving 13 kb
considerations make it necessary to at about 900 °C. Further measure-
carry out an absolute determination ments will be made in the future
of this type in an externally heated throughout this range with the plati-
gas apparatus. Bridgman (1918) num versus platinum 10% rhodium
used an apparatus of this type to and various other commonly used
determine the effect of pressure on thermocouples. The trends established
the thermoelectric quality of a vari- (Fig. 110) show that the pressure
546
CARNEGIE INSTITUTION
U
o
a
D
o
u
o
E
u
d)
-c
o
-»-
o
o
E
3
<\l
o
'-T3
d
O
-C
O
E
D
D
a
o
«»
o
ID
D
q
>
6
E
3
en
C
U—
"o
>
a
a
1
-*-
c
C
o
o
o
•4—
<D
<J
o
o
o
O
o
o
a>
—
<D
q
0>
i_
lb.
3
d
3
«/>
■B
0)
w.
i_
CD
a.
Ql.
E
(u
■<^
o
o
lO
o
o
o
sJDqo|i>j 'ajnssajj
D
o u
O c
.2>|
GEOPHYSICAL LABORATORY
547
200 300
Temperature, °C
Fig. 110. Pressure correction curves for the platinum versus platinum 10% rhodium thermo-
couple (based on the data given in Fig. 109).
effect is very significant and indicate measured at high pressures may be
that reevaluation of geochemical data necessary.
STAFF ACTIVITIES
Journal of Petrology Malcolm Brown, currently a fellow
Professor C. E. Tilley and Dr. and recently appointed to the chair
H. S. Yoder, Jr., of this laboratory at Durham University, England, is
continue to serve as editors of the the Senior Managing Editor. They
Journal of Petrology. Professor G. are pleased to announce that Profes-
54S CARNEGIE INSTITUTION
sor E. C. Vincent, Oxford University, Annual Meeting- of the American In-
England, and Professor J. F. G. Wil- stitute of Chemical Engineers at De-
kinson, University of New England, troit; as a lecturer ("Chemical Events
Australia, have agreed to serve as on the Primitive Earth") at the De-
coed iters, partment of Geology, University of
Volume 7 for 1966 consisted of 499 Toronto; as a participant ("Govern-
pages, with contributions from three ment Influence on the Conduct of
staff members, one research associate, Scientific Research" and "How Do
one fellow, and four alumni. The staff We Educate Tomorrow's Scien-
of the Geophysical Laboratory con- tists?") in a program on The Future
tinue to serve as critics with their of Science in the Liberal Arts College
international colleagues. at Wheaton College, Norton, Massa-
Emphasis remains on the more chusetts; as speaker ("Scientific Re-
thorough investigations of lasting search" ) at the University of Rhode
value that produce new data and Island's Honors Colloquium on Crea-
ideas. Definitive papers providing di- tivity in Modern Science ; and as lec-
rection to research in major petro- turer ("Science in America's Future")
logical problems are the hallmark of at the Sigma Xi Chapter of the Uni-
the Journal of Petrology. versity of Maine.
P. M. Bell presented an invited
Lectures paper at a Symposium on Properties
During the report year staff mem- of Crystalline Solutions at the annual
bers and fellows were invited to pre- meeting of the American Geophysical
sent lectures as follows: Union.
P. H. Abelson made a total of 18 F. R. Boyd was one of the invited
invited public appearances. He made lecturers at a short course on chain
speeches in the following capacities: silicates, sponsored by the American
as after-dinner speaker ("Science and Geological Institute, held at Palo
the Future of Society") at a Sympo- Alto, California, before the annual
sium on Electron Spin Resonance at meeting of the Geological Society of
Michigan State University, sponsored America.
by the Division of Physical Chemis- G. M. Brown lectured at the geol-
try of the American Chemical Society; ogy departments of Toronto Univer-
as a participant ("Who Shall Live?") sity, Princeton University, The Johns
in a panel discussion on "Science, Hopkins University, and Franklin
Society, and the Public Health — Eth- and Marshall College, and at the pre-
ical Issues" at The Johns Hopkins liminary course on chain silicates of
University School of Hygiene and the Geological Society of America,
Public Health 50th Anniversary Cele- sponsored by the American Geologi-
bration; as a banquet speaker ("Jus- cal Institute, at Palo Alto.
tifying Support of Research") at a F. Chayes was president of the
meeting of the Division of Inorganic Mineralogical Society of America
Chemistry of the American Chemical during 1967. During the spring he
Society at New York City; as a lee- delivered a series of eight lectures on
turer ("Status and Trends in Public ratio correlation to a graduate semi-
Policy") at the Delaware Section of nar in the Geology Department of
the American Chemical Society at the State University of New York
Wilmington; as a participant ("An at Binghamton.
Overall Look at Water Resources") J. R. Craig presented an invited
in a Symposium on Water Resources paper at a Symposium on Magmatic
and Government Policy at the 59th Ore Deposits, sponsored by the So-
GEOPHYSICAL LABORATORY 549
ciety of Economic Geologists, at Experimental Biology and Medicine
Stanford University. at the U.S. Naval Hospital, Bethesda,
Gabrielle Donnay was a visiting Maryland,
scientist at the Mineralogical Insti- H. G. Huckenholz lectured at the
tute of the University of Marburg, Lamont Geological Observatory, Co-
Germany, from June 15 to August lumbia University.
30, 1966. She gave invited lectures at T. E. Krogh presented a paper at
the Mineralogical Institutes of the the conference on Geochronology of
Universities of Marburg, Kiel, and Precambrian Stratified Rocks, spon-
Gottingen. The President of the Na- sored by the Geochronology Commis-
tional Academy of Sciences appointed sion of the International Union of
her to a three-year term on the Geological Sciences, the Geological
U.S.A. National Committee on Crys- Survey of Canada, and the Depart-
tallography. ment of Geology, University of Al-
M. C. Gilbert lectured at the de- berta, held at Edmonton, Canada,
partments of geology at the Univer- June 9-12, 1967.
sity of Missouri and the University G. Kullerud continued as Adjunct
of Pennsylvania and the departments Professor in Geochemistry at Lehigh
of geological sciences at Virginia University where he supervised the
Polytechnic Institute and the Univer- research of two Ph.D. candidates. He
sity of Colorado. He also presented an also served as a Visiting Professor at
invited paper at a Symposium on Heidelberg University and delivered
Properties of Crystalline Solutions at a series of lectures and conducted
the annual meeting of the American seminars in the Mineralogisch-Petro-
Geophysical Union. graphisches Institut of that univer-
N. Guven lectured at the Depart- sity. In addition, he gave two lectures
ment of Geology at the University of in the Mineralogisches Institut der
Illinois. Universitat Frankfurt. He also ad-
E. C. Hansen addressed the De- dressed faculties and students in the
partment of Geology at the Univer- Mineralogisches Institut, Erlangen-
sity of Maine. Nurnberg; the Mineralogisch-Petro-
P. E. Hare spent part of the second graphisches Institut der Universitat
semester as a Visiting Lecturer in Bonn; the Geologisches Institut der
the Geology Department at Yale Uni- Universitat zu Koln ; the Geologisches
versity. He also lectured at Woods Institut, Technische Hochschule
Hole Oceanographic Institution and Karlsruhe; Mineralogisches Institut
presented an invited paper before the der Universitat Saarbriicken; and the
Geological Society of Washington. Mineralogisch-Petrographisches In-
T. C. Hoering lectured at the de- stitut der Universitat Tubingen. He
partments of geology at the Univer- gave two lectures in the Mineralo-
sity of Tasmania, Australia, the Uni- gisch-Petrographisches Institut der
versity of Adelaide, South Australia, Universitat Bern, one lecture in the
and The Johns Hopkins University; Department of Mineralogy and Pe-
at the Department of Chemistry, trology, University of Cambridge, and
Australian National University; at one in the Department of Geology and
the Department of Geochemistry and Petrology, Oxford University. Kulle-
Mineralogy, Pennsylvania State Uni- rud also addressed a joint Geochemis-
versity; and before the Geological So- try, Mineralogy, and Geological
ciety of Western Australia. He also Sciences seminar at Pennsylvania
participated in a Symposium on Exo- State University and was the speaker
biology, sponsored by the Society for at a luncheon meeting of the U.S.
550
CARNEGIE INSTITUTION
Department of Agriculture Viking
Club. He also presented an invited
lecture at a Symposium on Magmatic
Ore Deposits at Stanford University
and served on the President's Ad-
visory Committee to the Department
of Earth Sciences at the University
of Toronto.
D. H. Lindsley lectured at the de-
partments of geology at the Univer-
sity of California at Santa Barbara,
California Institute of Technology,
and the University of Illinois and
conducted a three-week lecture series
on high pressures at the Department
of Geology at The Johns Hopkins
University. He also gave an invited
lecture at the Anorthosite Sympo-
sium at the State University of New
York College at Plattsburgh and pre-
sented an invited paper at a Sympo-
sium on Magmatic Ore Deposits at
Stanford University.
R. M. Mitterer lectured at the In-
stitute of Marine Sciences at the
University of Texas and at the Divi-
sion of Geosciences, Southwest Cen-
ter for Advanced Studies at Dallas.
A. J. Naldrett addressed the de-
partments of geology at Queen's Uni-
versity and the University of Toronto.
D. C. Presnall lectured at the Divi-
sion of Geosciences, Southwest Cen-
ter for Advanced Studies.
S. W. Richardson was an invited
lecturer at the departments of geol-
ogy at The Johns Hopkins University
and Princeton University.
J. F. Schairer addressed the joint
meeting of the Geological Society of
Pittsburgh and the Pittsburgh Sec-
tion of the American Chemical Society
at the Mellon Institute, and lectured
at the College of Mineral Industries,
Pennsylvania State University, and
at the Department of Geology, Uni-
versity of Pennsylvania. As a Visit-
ing Scholar in the Department of
Geological Sciences at Virginia Poly-
technic Institute, he gave two formal
lectures and held informal discussions
with faculty members and students.
Schairer also gave an invited lecture
at a Symposium on Ceramic Mate-
rials and their Properties, sponsored
by the American Ceramic Society, the
American Society for Testing and
Materials, and the National Bureau
of Standards at the NBS Gaithers-
burg facility. He conducted the
Southeast Tour of the Sigma Xi-
RESA National Lectureship Program
in the fall of 1966 with speaking
engagements at North Carolina State
University, Duke University, Wake
Forest College, Lithium Corporation
of America RES A Branch, Clemson
University, Georgia Institute of
Technology, Atlanta University Cen-
ter, Auburn University, University of
Georgia, Vanderbilt University, and
the Oak Ridge National Laboratory
RES A Branch. Schairer was sent by
the Foreign Secretary's Office of the
National Academy of Sciences to at-
tend the council meetings and sym-
posia at the first meeting of the In-
ternational Association of Geochemis-
try and Cosmochemistry at UNESCO
headquarters in Paris, France (May
8-11, 1967). Schairer is a member of
the Ad Hoc Committee of the Na-
tional Academy of Sciences on Geo-
chemistry and Its Place Internation-
ally, to which he made a favorable
report on his return.
W. H. Scott delivered a series of
four lectures to an undergraduate
class at the Department of Geology,
Yale University.
H. S. Yoder, Jr., gave two lectures
at the University of Toronto on pet-
rological problems and presented a
talk at the Inorganic Materials Divi-
sion of the National Bureau of Stand-
ards on the effects of high pressures
and temperatures on mineral stabil-
ity. He again participated in the
American Geophysical Union Visit-
ing Scientist Program by giving a
series of lectures at Queen's College
of City University of New York,
GEOPHYSICAL LABORATORY 551
Dartmouth College, Texas Technical University, Ottawa), March 14, 1967.
College, and Wittenberg University. "Interaction of hydrothermal solu-
In addition, an extensive series of tions with country rocks," by Ulrich
lectures was given in connection with Petersen (Harvard University), May
the Engineering for Executives Con- 2, 1967.
ference sponsored by the University
of Texas. Yoder was an invited lee- The section Summary of Published
turer at the Anorthosite Symposium Work briefly describes the papers
at the State University of New York published in scientific journals dur-
College at Pittsburgh and he gave ing the report year> In addition, the
two lectures at the School of Ocean- following papers have been prepared
ography, University of Rhode Island, for publication: P. H. Abelson, "Con-
on basalt problems related to the version of biochemicals to kerogen
ocean basins. An advanced course of and n_paraffins"; P. M. Bell, "Geo-
26 lectures on problems in experi- physical research at pressures above
mental petrology was given at the 30 kb"; P. M. Bell and J. L. England,
Summer Session of the University of -High-pressure experimental tech-
Colorado, niques"; F. R. Boyd, "Penological
p ± 7 - + > n h problems in high-pressure research";
reiroiogists Ltuo F chayes> «0n graphical appraisal
Seven meetings were held during of the strength of associations in pet-
the 56th year of the Penologists' rographic variation diagrams"; F.
Club. Metamorphic petrology and Chayes, "On locating field boundaries
geochemistry were the general sub- in simple phase diagrams by means
jects discussed. Specifically, the fol- of discriminant functions"; J. R.
lowing lectures were given during the Craig and G. Kullerud, "Phase rela-
meetings: tions and mineral assemblages in the
"A comparison of the Si02 and H20 copper-lead-sulfur system" ; J. R.
systems," by W. Barclay Kamb (Cal- Craig and G. Kullerud, "Phase rela-
ifornia Institute of Technology and tions in the Cu-Fe-Ni-S system and
Massachusetts Institute of Technol- their application to magmatic ore de-
ogy), November 3, 1966. posits"; G. Donnay, L. B. Coleman,
"The n-dimensional tie-line prob- N. G. Krieghoff, and D. 0. Cowan,
lem," by Hugh Greenwood (Prince- "Trimethylplatinum (IV) iodide and
ton University) , December 6, 1966. its misrepresentation as hexamethyl-
"The system NaAlSi308-KAlSi308- diplatinum"; N. Guven and C. W.
Si02-H20 in the presence or absence Burnham, "The crystal structure of
of an aqueous vapor phase," by Wil- 3T muscovite"; T. C. Hoering, "The
liam Luth (Massachusetts Institute of organic geochemistry of Precambrian
Technology) , December 20, 1966. rocks"; G. Kullerud, "Sulfide stud-
"Hydrothermal studies in the sys- ies"; G. Kullerud, "Geologic ther-
tem K20-MgO-Al203-Si02-H20," by mometry"; D. H. Lindsley, "Melting
Werner Schreyer (Mineralogisches relations of plagioclase at high pres-
Institut, Kiel), February 14, 1967. sures"; A. J. Naldrett, J. R. Craig,
"The Skaergaard intrusion : A sum- and G. Kullerud, "The central por-
mary of progress and problems," by tion of the Fe-Ni-S system and its
G. Malcolm Brown (University of bearing on pentlandite exsolution in
Oxford), February 28, 1967. iron-nickel sulfide ores"; A. J. Nal-
"An experimental study of the drett and G. Kullerud, "A study of
metamorphism of siliceous carbon- the Strathcona mine and its bearing
ates," by George Skippen (Carleton on the origin of the nickel-copper ores
552 CARNEGIE INSTITUTION
of the Sudbury District, Ontario"; alkali basalts"; D. B. Stewart, "Four-
Mi J. O'Hara and H. S. Yoder, Jr., phase curve in the system CaAl2Si208-
" Formation and fractionation of ba- Si02-H20 between 1 and 10 kilobars";
sic magmas at high pressures"; J. F. H. S. Yoder, Jr., "Experimental stud-
Schairer, "Phase equilibria at one at- ies bearing on the origin of anortho-
mosphere related to tholeiitic and site."
GEOPHYSICAL LABORATORY
553
SUMMARY OF PUBLISHED WORK
(1457) Isotopic composition of lead
from granitic rocks of North
America. G. R. Tilton. Sci.
Terre, 10, 247-259, 1965.
Intensive studies of the isotopic com-
position of lead in feldspar from granitic
rocks having ages of 450 m.y. (Appala-
chian Province), 1100 m.y. (Grenville
age), and 2700 m.y. (Superior Province)
are described. Considerable variation is
found within each group, but it is inter-
esting that the isotopic composition of
the least radiogenic lead in each case
agrees with that of an associated "con-
formable galena" as defined by Stanton
and Russell. Comparison of the model
lead age of the least radiogenic lead with
rock age suggests a systematic trend.
The model ages are 150-200 m.y. younger
than the age of the granitic rocks of Ap-
palachian and Grenville age, but 200-300
m.y. older than the rock age of the Su-
perior samples. The trend agrees rather
closely with one noted by Patterson and
Tatsumoto in a study of detrital feld-
spars from North America. Possible
mechanisms responsible for the model
age values are given. If lead from the
2700-m.y.-old rocks is used to calculate
an age for the earth in the usual way,
with the use of meteoritic lead as the
earth's initial lead, a value of 4730 m.y.
is obtained. Commonly quoted values are
4530-4580 m.y. Since there is some evi-
dence that the Superior Province leads
reported here may be anomalous, the new
age value must be considered tentative
until improved knowledge of the isotopic
composition of lead in very old rocks can
be obtained.
(1463) Mineral ages from the Finnish
Precambrian. O. Kouvo and G. R.
Tilton. J. Geol, 74, 421-442,
1966.
The number of isotopically dated zir-
cons from the basement complex of the
Karelian belt has been doubled to include
six samples. The ages are all discordant
and are compatible with loss of lead by
a continuous diffusion mechanism. Two
possible mechanisms are that diffusion
took place at a constant value of the
parameter, D/a2, or under conditions
such that D/a2 increased linearly in pro-
portion to the amount of radiation dam-
age. The lead age data are also compat-
ible with an episodic loss of lead about
500 m.y. ago, but there is no evidence
that episodic loss should have occurred at
that time. The regional metamorphism
which affected the rocks 1700-1900 m.y.
ago, as recorded by the mica age values,
did not seriously alter the uranium-lead
ratios in the zircons of the pre-Karelian
basement complex; however, zircon from
the gneiss domes to the west may have
experienced some lead loss as a result
of the regional metamorphism.
Four detrital zircon samples from
Svecofennian schists near Tampere have
discordant isotopic ages suggesting an
age of 2300 m.y. Since igneous rocks of
this age are as yet unknown in Finland,
the value is provisional. Zircon from a
granodiorite intruding the schists has a
nearly concordant age of 1900 m.y. Com-
parison of the zircon ages from the
schists and granodiorite shows that the
1900-m.y. age must represent the time of
crystallization of zircon in the granodi-
orite, not a time of resetting as a result
of intense metamorphism. These data
strengthen the conclusions of earlier
studies that plutonic rocks associated
with both the Karelian and Svecofennian
orogenic belts crystallized at about the
same time.
Data bearing on the retention of lead
in zircon are presented. Zircon in a lens
of schist 40 X 150 m enclosed in the
1900-m.y. granodiorite mass has not had
its age record erased in spite of the fact
that the zircons in the schist must have
been heated to the temperature of the
intrusive mass. Other data in the litera-
ture establish that loss of lead from zir-
con may take place under less severe
conditions. The uranium and thorium
content of the samples appears to be one
factor controlling lead loss, samples hav-
ing high radioactivity losing lead more
readily. This effect is probably caused by
radiation damage to the crystal structure
554
CARNEGIE INSTITUTION
or by inhomogeneous distribution of
radioactivity.
^1464") Polymorphism on the Cu5FeS.t-
Cu9S5 join. N. Morimoto and G.
Kullerud. Z. Krist., 123, 235-254,
1966.
At elevated temperatures complete
solid solution exists between Cu5FeS4
and Cu9S8 (Kullerud and Roseboom,
1958). Homogeneous solid solutions of
nine compositions on this join were syn-
thesized at temperatures above the
solvus. The erystallographic properties of
the solid solutions were studied by the
Weissenberg and precession methods at
various temperatures. At elevated tem-
peratures each composition exists as one
phase, commonly in well-developed single
crystals, and at subsolidus temperatures
two phases occur. All the synthetic mate-
rials, when studied over a temperature
range, display X-ray reflections anal-
ogous to those of the high-, transitional-,
and low-temperature forms of bornite
and digenite described by Morimoto and
Kullerud (1961, 1963). The structure of
the high-temperature form is described
on the basis of the cubic close-packing of
sulfur atoms with statistical distribution
of metal atoms. During the inversion
from the high- to the low-temperature
polymorph, the transitional metastable
forms temporarily appear. These are
superstructures of the high-temperature
form, and their structure types differ
owing to differences in composition. The
various superstructures are derived, by
ordered arrangements of metal-vacant
layers, from the structure of the high-
temperature form. The transitional
forms gradually change to the low-
temperature forms. Among the low-
temperature forms only two types cor-
responding to the structures of natural
bornite (Cu5FeS4) and digenite (Cu9-
rS5) are stable at room temperature.
Experiments were also carried out with
natural digenite and bornite, and the re-
sults are explained on the basis of the
structures of the transitional forms oc-
curring on the Cu-FeS4-Cu,,S.- join. The
mechanisms of the polymorphic trans-
formations are considered from a struc-
tural point of view.
(1465) An approximate statistical test
for correlations between propor-
tions. F. Chayes and W. Kruskal.
J. Geol, 7U, 692-702, 1966.
The observed means and variances of
data occurring as proportions or per-
centages may be used to estimate anal-
ogous parameters of a theoretical open
array, X, which, on closure, yields a
new array, Y, whose means and vari-
ances are exactly those of the observed
data, but in which the covariances have
been generated entirely by closure. The
correlations in Y, found directly from
the means and variances of X, are ap-
propriate null values against which the
observed correlations may be tested. A
testing procedure is outlined, and a
practical example is given.
(1467) The join Mg2Si206-CaMgSi2Ofi at
30 kb pressure and its applica-
tion to pyroxenes from kimber-
lites. B. T. C. Davis and F. R.
Boyd. /. Geophys. Res., 71, 3567-
3576, 1966.
At 30 kb pressure the join Mg2Si206-
CaMgSi206 is binary. The liquidus de-
scends in a smooth curve from the melt-
ing point of enstatite (1840°C) to that
of diopside (1715°C). The solidus is in-
tersected by a transition loop at 1765 °C
forming a peritectic where the composi-
tions of the solid and liquid phases in
equilibrium are L = en65di35, Diss =
en79di21, Enss = en92di8 (mole %). The
diopside-rich limb of the transition loop
(diopside solvus) is sharply inflected
toward diopside at about 1500 °C. This
inflection is attributed to a metastable
miscibility gap between clinoenstatite
and diopside solid solutions hidden
within the transition loop. At tempera-
tures above the inflection the range of
diopside solid solutions is much greater
than at atmospheric pressure. However,
at temperatures below 1400°C the diop-
side solvus at 30 kb is very close to the
boundary of the two-pyroxene field at
atmospheric pressure. Ca-rich and Mg*
rich pyroxene pairs from peridotite
nodules in kimberlites are relatively poor
in R2Oo, alkalies, and Fe. To a first ap-
proximation they can be treated as phases
GEOPHYSICAL LABORATORY
555
in the system Mg2Si206-CaMgSi206.
Seven modern analyses of Ca-rich pyrox-
enes from such nodules show a range of
solid solution corresponding to a range
of equilibration temperatures of about
950°-1300°C. At present we can only
speculate whether this range is primary
or whether some exsolution occurred
during eruption.
(1471) Upper stability of muscovite. B.
Velde. Am. Mineralogist, 51, 924-
929, 1966.
The maximum stability of muscovite,
K[Al2(Si3Al)O10(OH)2], was determined
experimentally under the hydrothermal
conditions PH 0 = ^totai- On the basis of
optical and X-ray observations, the upper
stability of muscovite at 1 bar = (550°-
570°) ± 6°C, 160 bars = (610°-630°)
± 6°C, 1000 bars = (650°-670°) ± 6°C,
2000 bars = (675°-700°) ± 6°C, and
8000 bars = (720°-740°) ± 6°C. The
reaction muscovite ^ sanidine + corun-
dum + H20 was observed to proceed in
both directions at the above intervals
with both synthetic and natural starting
materials.
(1472) Anorthite-forsterite and anor-
thite-enstatite reactions and their
bearing on the basalt-eclogite
transformation. I. Kushiro and
H. S. Yoder, Jr. /. Petrol, 7,
337-362, 1966.
The reactions between anorthite and
forsterite and between anorthite and en-
statite have been studied at high pres-
sures and high temperatures with solid-
media and gas-media apparatus. For the
1:1 (molecular ratio) composition of
anorthite and forsterite the following
three assemblages have been obtained
with increasing pressure at a constant
temperature: anorthite + forsterite,
clinopyroxene + orthopyroxene + spinel
+ anorthite, and garnet (pyrope-gros-
sularite series). Two univariant curves
between these three assemblages have
been determined. The amounts of pyrox-
enes relative to spinel and anorthite in
the second assemblage increase and the
contents of the Tschermak's components
in the pyroxenes increase with increasing
pressure.
The 1: 2 (molecular ratio) composition
of anorthite and forsterite was also
studied, and the univariant curve for the
reaction clinopyroxene + orthopyroxene
+ spinel «=* garnet (pyrope-grossularite
series) + forsterite has been determined.
For the 1:2 (molecular ratio) com-
position of anorthite and enstatite the
following assemblages have been ob-
tained with increasing pressure at a
constant temperature: anorthite + en-
statite, anorthite + orthopyroxene +
clinopyroxene + quartz, and garnet Cpy-
rope-grossularite series) + quartz. The
univariant curve between the last two
assemblages has been determined. Both
the pyroxenes in the second assemblage
contain Tschermak's components, which
increase in content with increasing pres-
sure at the expense of anorthite.
These experimental results suggest
that basalts, either saturated or under-
saturated with silica, transform into alu-
minous pyroxene-rich assemblages at
lower pressures or higher temperatures
than the stability field of eclogites, and
that pyroxene-spinel symplectites found
in some metamorphosed gabbro, trocto-
lite, and norite may have formed in the
stability field of the pyroxene-rich assem-
blages. The stability field of the pyrox-
ene-rich assemblages is probably included
within the pressure-temperature condi-
tions of the granulite facies. It is also
suggested that the garnet-free perido-
tites, such as the peridotite inclusions in
basaltic rocks and the intrusive perido-
tites in the orogenic belts, are stable at
lower pressures or higher temperatures
than garnet peridotites, such as the in-
clusions in kimberlite and intrusive gar-
net peridotite in the orogenic belts.
(1473) Molecular solid solution of tetra-
phenylporphin and silver tetra-
phenylporphin. G. Donnay and
C. B. Storm. Molecular Crystals,
2, 287-292, 1967.
The triclinic modifications of tetra-
phenylporphyrin (TPP) and silver tetra-
phenylporphyrin (AgTPP) are isostruc-
tural. Five intermediate compositions
were crystallized and studied by X rays.
556
CARNEGIE INSTITUTION
A complete series of solid solution was
found ; cell dimensions show very small
changes only. Between 8 and 53 mole %
AgTPP, a second tetragonal phase, co-
precipitates in small amounts. Its crystal
data are given.
(1474) The crystal structure of jadeite,
XaAlSi,0,. C. T. Prewitt and
C. W. Burnham. Am. Mineralo-
gist, 51, 956-975, 1966.
The crystal structure of naturally oc-
curring pure jadeite has been refined by
least-squares methods with the use of
single-crystal X-ray intensity data ob-
tained by counter-diffractometer tech-
niques. This jadeite is monoclinic, space
group C2/c, with cell dimensions a =
9.418 A. b = 8.562 A, c = 5.219 A, and
f3 = 107.58°. The structure is similar to
that of the pyroxene diopside and con-
tains parallel sheets of octahedrally co-
ordinated aluminum and 8-coordinated
sodium polyhedra connected by silicate
chains running parallel to the c axis. The
mean cation-oxygen distances are Si-0
= 1.623 A, Al-0 = 1.928 A, and Na-0
= 2.469 A.
(1475) Thermal stability of assemblages
in the Cu-Fe-S system. R. A.
Yund and G. Kullerud. /. Petrol.,
7, 454-488, 1966.
The phase relations in the Cu-Fe-S
system were determined from 700 °C to
approximately 200 °C in most portions of
the system and below 100 °C in restricted
areas. Approximate solid solution limits
for bornite, chalcopyrite, and pyrrhotite
were determined at elevated tempera-
tures. At low temperatures emphasis was
placed on establishing the stable assem-
blages and less on determining the com-
positions of coexisting phases. At 700°C
two extensive ternary solid solutions
dominate the phase relations in this sys-
tem. One of these solid solutions (born-
ite) includes the compositions Cu2S,
Cu18S, and Cur>FeS.t; and the other
(chalcopyrite) lies within the area
bounded by the compositions CuFeS2,
CuFe2S„ and Cu,Fe,SG. The two fields
are separated by approximately 10 wt %
copper at 700 °C. The chalcopyrite vol-
ume, as seen in a trigonal prism repre-
senting temperature and composition, is
intersected by a miscibility gap below
approximately 600 °C. Below this tem-
perature the two one-phase volumes are
referred to as chalcopyrite and cubanite.
Chalcopyrite is tetragonal at low temper-
ature but isometric above approximately
550 °C. The temperature of the transfor-
mation is a function of composition. Cu-
banite is isometric above 252 °C, tetrag-
onal from 252° to at least 213 °C, and
orthorhombic at lower temperature. The
temperature of the second transforma-
tion is unknown because the tetragonal-
to-orthorhombic transformation has not
been achieved in the laboratory. Bornite
and pyrite become stable together at
568°C and coexist down to 228°C. Covel-
lite appears with lowering temperature
at 507°C, and idaite at 501°C. Idaite-
pyrite and idaite-bornite are stable as-
semblages below 501 °C. The composition
of bornite coexisting with idaite changes
gradually toward digenite with decreas-
ing temperature, thus permitting the
change from the bornite-pyrite tie line
to the digenite-chalcopyrite tie line at
228 °C. Other major tie-line changes are
bornite-iron to pyrrhotite-copper below
475 °C and cubanite-pyrite to chalcopy-
rite-pyrrhotite below 334 °C. A new syn-
thetic phase, ^-bornite, which has a
composition close to bornite (Cu5FeS4)
but contains about 0.4 wt % more sulfur,
forms when sulfur-rich bornite synthe-
sized at high temperature is annealed
between 62° and 140 °C. Optically this
new phase is very similar to bornite, and
their X-ray powder diffraction patterns
are given for comparison.
The determined phase relations are ap-
plicable to numerous deposits. The tie-
line changes involving bornite + pyrite
reacting to produce digenite + chalco-
pyrite below 228 °C and cubanite (iso-
metric) + pyrite going to chalcopyrite
+ pyrrhotite below 334 °C are of con-
siderable geological interest. The rates
of these reactions are sufficiently slow to
allow the higher temperature assem-
blages to be observed in some ores. The
cubic-tetragonal inversion in chalcopy-
rite is often deduced in ores by inversion
twins. However, twinning is also com-
monly produced through deformation.
GEOPHYSICAL LABORATORY
557
Geological application of the inversion
therefore depends on correct interpreta-
tion of the twinning. Because of the
considerable solubility of copper in pyr-
rhotite the pyrrhotite-pyrite solvus of
the pure Fe-S system cannot be applied
indiscriminately to ores that also contain
chalcopyrite or cubanite, or both. The
new cc-bornite phase was identified with
the natural "anomalous bornites," which
when heated exsolve chalcopyrite and,
depending on their composition, also
digenite.
The experimental results indicate that
the mineral commonly identified as chal-
copyrrhotite is in reality tetragonal or
even isometric cubanite. Experimental
evidence could not be obtained for the
existence of a phase of Cu3Fe4S7 or
Cu2Fe4S7 composition, the older formulas
given for valleriite. The thermal break-
down of natural material supports the
idea that valleriite is a low-temperature
polymorph of chalcopyrite. The relatively
uncommon occurrence of idaite in com-
parison to covellite is attributed to the
greater difficulty in nucleating idaite.
The possibility of stable coexistence of
chalcocite and pyrite was investigated
but was found to be prohibited by tie
lines between bornite and digenite even
as low as 100 °C.
(1476) Ferrosilite III: A triclinic py-
roxenoid-type polymorph of fer-
rous metasilicate. C. W. Burn-
ham. Science, 154, 513-516, 1966.
The relationships between the triclinic
unit cell of ferrosilite III and those of
pyroxmangite, rhodonite, and wollasto-
nite lead to the hypothesis that this poly-
morph of ferrous metasilicate has a
pyroxenoid-type crystal structure with
single silicate chains that repeat after
every nine silicon tetrahedra. Vector re-
lations between the triclinic cell and an
apparent pseudomonoclinic cell support
this hypothesis. Although the proposed
silicate chain has a longer repeat length
than any now known, it represents a
logical extension of those found in other
pyroxenoids and suggests that even
longer repeat lengths may yet be found
among phases with pyroxene composi-
tions.
(1477) The join forsterite-diopside-iron
oxide and its bearing on the
crystallization of basaltic and
ultramafic magmas. D. C. Pres-
nall. Am. J. Sci., 264, 753-809,
1966.
The liquidus surface of the join for-
sterite-diopside-iron oxide has been stud-
ied by quenching techniques at a total
pressure of 1 atmosphere and at oxygen
fugacities from 10-0-68 to 10~8 atmos-
pheres. At oxygen fugacities from 10-0-68
to 10-6 atmospheres the minimum liqui-
dus temperature on the join forsterite-
diopside-iron oxide remains nearly con-
stant in temperature and composition at
about 1300 °C and 79% diopside, 21%
iron oxide (total iron as FeO) by weight.
Almost coincident with this minimum is
a piercing point at which olivine, diop-
side, and spinel are in equilibrium. As
oxygen fugacity decreases from 10-°-68 to
10-6 atmospheres, the quantity (Fe203)
(100) /(FeO + Fe203) in the liquid at
this piercing point decreases from 87.7
to 37.0 and the weight percentage of
fayalite in the olivine in equilibrium
with this liquid changes from 4 to 14 wt
%. At 10~6 atmospheres oxygen fugacity
an additional piercing point occurs at
which olivine, spinel, and magnesiowus-
tite are in equilibrium. At 10-8 atmos-
pheres oxygen fugacity, the primary
phase field of spinel is absent and no
piercing points occur.
Comparison of published data on the
chemistry of basaltic and ultramafic
rocks with data on the join forsterite-
diopside-iron oxide, inferred relation-
ships in parts of the system CaO-MgO-
FeO-Fe203-Si02 adjacent to this join,
and previous data on the system MgO-
FeO-Fe203-Si02 (Muan and Osborn,
1956) yields these conclusions: (1) The
Skaergaard intrusion represents the re-
sults of a close approach to fractional
crystallization of an essentially anhy-
drous magma at constant total composi-
tion. In this situation the oxygen fugac-
ity of the magma would be controlled by
the liquid and crystalline phases in equi-
librium in the magma. (2) If basaltic
magmas similar in composition to the
initial Skaergaard magma have frac-
tionally crystallized at relatively low
55S
CARNEGIE INSTITUTION
pressures within the earth's crust, and
if liquids derived from this fractional
crystallization have been produced at the
surface as lavas, the oxygen fugacities
of these basaltic magmas were controlled
by oxidizing buffers. (3) The equilib-
rium diagrams support the hypothesis of
Kuno (1950) that in volcanic rocks re-
action rims consisting of both hyper-
sthene and magnetite on olivine pheno-
crysts are the result of partial resorption
of olivine by liquid. (4) At low pressures
peridot ites can exist while partly liquid
at temperatures as low as about 1250 °C.
Complete melting would require temper-
atures in the vicinity of 1700°C. (5)
Fractional crystallization of ultramafic
magma under conditions of buffered oxy-
gen fugacity adequately explains the
origin of the major ultramafic rock types
in the Duke Island ultramafic complex in
Alaska.
(1478) Melting relations of KAlSi308:
Effect of pressures up to 40 kilo-
bars. D. H. Lindsley. Am. Miner-
alogist, 51, 1793-1799, 1966.
Under dry conditions the incongruent
melting of KAlSi3Os to leucite + liquid
persists to 19 ± 1 kb, whereas under
water pressure the incongruent melting
disappears at 2.6 kb. This illustrates the
profound effect of water activity in the
melt relative to total pressure on the
condensed phases. For systems in which
water pressure is less than total pres-
sure, the leucite field may be more im-
portant than has been presumed.
(1479) Optical determination of water
content in spherulitic vaterite.
J. D. H. Donnay and G. Donnay.
Acta Cryst., 22, 312-314, 1967.
From measured indices of refraction
for vaterite crystals (ne, 1.650 ; n0),
1.550; Johnston, Merwin, and William-
son, 1916) and for vaterite fibers (ne,
1.625; n0i 1.538; Meyer, 1965), Wiener's
formulas give two independent values
for the interstitial water content of the
fibers (5.9, 5.8 vol %). The lowering of
the density from 1.645 (crystal) to 1.54
(fiber) indicates 6.4 vol % water. The
decomposition product of CaC03#6H20
described by Johnston and co-workers
can now be identified as fibrous aggre-
gates of vaterite and water.
(1480) Annual report of the Director
for 1965-1966.
(1482) The microcline-orthoclase transi-
tion within a contact aureole.
R. H. Steiger and S. R. Hart.
Am. Mineralogist, 52, 87-116,
1967.
Microcline perthites from Precambrian
pegmatites in the Front Range, Colorado,
have been converted to orthoclase in a
zone around a Tertiary intrusive stock.
This stock, near the town of Eldor, is
2 miles across and shows fairly well-
exposed vertical contacts. The microcline-
orthoclase transition, the only major
manifestation of contact metamorphism,
was investigated optically and by X-ray-
powder diffraction along five traverses.
Outside the contact aureole the micro-
cline commonly consists of clear crystals
with typical cross-hatched twinning and
shows a 2V of 80° ± 5° and an obliquity
ranging from 0.84 to 0.93. The orthoclase
from near the contact is always turbid,
untwinned, with axial angles -L (010)
from 50° to 66°. The perthitization and
composition (Or77Ab23) of the feldspars
remain roughly constant across the con-
tact aureole, except within the last few
feet from the contact, where an increased
albite content or the disappearance of
the perthite lamellae may occasionally
be observed. Both the nature of the
microcline-orthoclase transition and its
distance from the contact depend strongly
on the configuration of the contact in
that particular area. In accordance with
the heat flow calculations, the transition
is relatively sharp and close (1200 feet
or less) to protruding corners of the in-
trusive, whereas, near reentrant sections
of the intrusive, the transition is grada-
tional and relatively distant (over 2400
feet) from the contact. Heat flow calcu-
lations further indicate that the micro-
cline-orthoclase transition may occur at
temperatures below 400 °C, i.e., more
than 50 °C lower than was hitherto known
from an experimental approach. The
order of the transition is not specified by
GEOPHYSICAL LABORATORY
559
this investigation, although the occur-
rence of some intermediate microcline in
the transition zone may be evidence for
a continuous type of transformation.
(1483) Phase relations and mineral as-
semblages in the Ag-Bi-Pb-S
system. J. R. Craig. Mineralium
Deposita, 1, 278-306, 1967.
Phase relations within the Ag-Bi-S,
Bi-Pb-S, and Ag-Pb-S systems have been
determined in evacuated silica tube ex-
periments. Integration of experimental
data from these systems has permitted
examination and extrapolation of phase
relations within the Ag-Bi-Pb-S quater-
nary system. In the Ag-Bi-S system
liquid immiscibility fields exist in the
metal-rich portion above 597° ± 3°C
and in the sulfur-rich portion above 563°
± 3°C. Ternary phases present corre-
spond to matildite (AgBiS2) andpavonite
(AgBi3S5). Throughout the temperature
range 802° ± 2°C to 343° ± 2°C the
assemblage argentite (Ag2S) + bismuth-
rich liquid is stable; below 343 °C this
assemblage is replaced by the assemblage
silver + matildite. Five ternary phases
are stable on the PbS-Bi2S3 join above
400°C— phase II (18 mole % Bi2S3),
phase III (27 mole % Bi2S3), "cosalite"
(33.3 mole % Bi2S3), phase IV (51 mole
% Bi2S3), and phase V (65 mole %
Bi2S3). Phase IV corresponds to the
mineral galenobismutite and is stable
below 750° ± 3°C. Phases II, III, and
V do not occur as minerals, but typical
lamellar and myrmekitic textures com-
monly observed among the Pb-Bi sul-
fosalts and galena evidence their pre-
vious existence in ores. Phases II and
III are stable from 829° ± 6°C and 816°
± 6°C, respectively, to below 200°C;
phase V, stable only between 730° ± 5°C
and 680° ± 5°C in the pure Bi-Pb-S
system, is stabilized to 625° ± 5°C by
the presence of 2% Ag2S. Experiments
conducted with natural cosalites suggest
that this phase is stable only below 425°
± 25 °C in the presence of vapor. In the
Ag-Pb-S system the silver-galena assem-
blage is stable below 784° ± 2° C, where-
as the argentite + galena mineral pair is
stable below 605° ± 5°C. Solid solution
between matildite and galena is complete
above 215° ± 15°C; below this tempera-
ture characteristic Widmanstatten struc-
ture-like textures are formed through
exsolution. Schematic phase relations
within the quaternary system are pre-
sented at 1050 °C, at 400 °C, and at low
temperature.
(1484) The Fe-Pb-S system. R. Brett
and G. Kullerud. Econ. Geol, 62,
354-369, 1967.
The phase relations in the ternary
system Fe-Pb-S were determined by
quenching and differential thermal anal-
ysis experiments in rigid silica-glass
tubes. The solubilities of FeS in galena
and in liquid Pb and the extent of solid
solution of PbS in pyrite and pyrrhotite
at 700 °C are too small to be determined
by the methods employed and are prob-
ably less than 0.1 wt %. Less than 1 wt
% S is soluble in liquid Pb at 700 °C.
Ternary compounds do not occur in
this system. An invariant point involving
galena, pyrite, sulfide liquid, liquid sul-
fur, and vapor occurs at 716 °C. The ter-
nary liquid at this temperature contains
about 60 wt % Pb, 13 wt % Fe, and 27
wt % S, and is thus situated slightly on
the sulfur side of the PbS-FeS2 join.
At 719 °C tie lines between galena and
pyrite are replaced by sulfide liquid-
pyrrhotite tie lines in an invariant re-
action involving vapor. The sulfide liquid
field expands rapidly and intersects the
PbS-FeS2 join at 723 °C. At 848 °C the
ternary liquid field intersects the FeS-
PbS join. Diagrams are presented to
illustrate changes in phase relations
from 700° to 1115°C.
The experimental results demonstrate
that mixtures of galena (about 70 wt % )
and pyrite (about 30 wt %) can crystal-
lize directly from a liquid at 716 °C.
Additional components may lower this
temperature so that it appears possible
that pyrite-galena ores may melt at tem-
peratures below 700 °C when exposed to
thermal metamorphism.
560
CARNEGIE INSTITUTION
BIBLIOGRAPHY
Boyd, F. R., sec Davis, B. T. C.
Brett. R., and G. Kullerud, The Fe-Pb-S
system, Econ. GeoL, 62, 354-369, 1967.
Burnham. C. W., Ferrosilite III: A triclinic
pyroxenoid-type polymorph of ferrous
metasilicate, Science, 154, 513-516, 1966.
Burnham, C. W., sec also Prewitt, C. T.
Chaves, F., and W. Kruskal, An approxi-
mate statistical test for correlations be-
tween proportions, J. GeoL, 74, 692-702,
1966.
Craig, J. R., Phase relations and mineral
assemblages in the Ag-Bi-Pb-S system,
Minemlium Deposita, 1, 278-306, 1967.
Davis, B. T. C, and F. R. Boyd, The join
Mg2Si206-CaMgSi206 at 30 kb pressure
and its application to pyroxenes from
kimberlites, J. Geophys. Res., 71, 3567-
3576, 1966.
Donnay, G., and C. B. Storm, Molecular solid
solution of tetraphenylporphin and silver
tetraphenylporphin, Molecular Crystals,
2, 287-292, 1967.
Donnay, G., see also Donnay, J. D. H.
Donnay, J. D. H., and G. Donnay, Optical
determination of water content in
spherulitic vaterite, Acta Cryst., 22, 312-
314, 1967.
Hart, S. R., see Steiger, R. H.
Kouvo, 0., and G. R. Tilton, Mineral ages
from the Finnish Precambrian, J. Geol.,
74, 421-442, 1966.
Kruskal, W., see Chayes, F.
Kullerud, G., see Brett, R.; Morimoto, N.;
Yund, R. A.
Kushiro, I., and H. S. Yoder, Jr., Anorthite-
forsterite and anorthite-enstatite reac-
tions and their bearing on the basalt-
eclogite transformation, J. Petrol., 7, 337-
362, 1966.
Lindsley, D. H., Melting relations of
KAlSi3Os: effect of pressures up to 40
kilobars, Am. Mineralogist, 51, 1793-1799,
1966.
Morimoto, N., and G. Kullerud, Polymor-
phism on the Cu5FeS4-Cu9S5 join, Z.
Krist., 123, 235-254, 1966.
Presnall, D. C, The join forsterite-diopside-
iron oxide and its bearing on the crystal-
lization of basaltic and ultramafic mag-
mas, Am. J. Sci., 264, 753-809, 1966.
Prewitt, C. T., and C. W. Burnham, The
crystal structure of jadeite, NaAlSi206,
Am. Mineralogist, 51, 956-975, 1966.
Steiger, R. H., and S. R. Hart, The micro-
cline-orthoclase transition within a con-
tact aureole, Am. Mineralogist, 52, 87-
116, 1967.
Storm, C. B., see Donnay, G.
Tilton, G. R., Isotopic composition of lead
from granitic rocks of North America,
Sci. Terre, 10, 247-259, 1965.
Tilton, G. R., see also Kouvo, O.
Velde, B., Upper stability of muscovite,
Am. Mineralogist, 51, 924-929, 1966.
Yoder, H. S., Jr., see Kushiro, I.
Yund, R. A., and G. Kullerud, Thermal
stability of assemblages in the Cu-Fe-S
system, J- Petrol, 7, 454-488, 1966.
REFERENCES CITED
Akimoto, S., E. Komada, and I. Kushiro,
Preliminary experiments on the stability
of natural pigeonite and enstatite, Proc.
Japan Acad., 42, 482-487, 1966.
Alfani, M., L'augite pneumatolitica raccolta
sulle lave del fondo del cratere vesuviano
nel 1929, Periodico Mineral. {Rome), 5,
77-96, 1934; Mineral. Abstr., 5, 440, 1934.
Amagat, E. H., Memoires sur Pelasticite et
la dilatabilite" des fluides jusqu'aux tres
hautes pressions, Ann. Chim. Phys., 29,
68-136, 1893.
Anderson, G. M., The calculated fugacity of
water to 1000°C and 10,000 bars, Geochim.
Cosmochim. Acta, 28, 713-715, 1964.
Appleman, D. E., D. S. Handwerker, and
H. T. Evans, The least-squares refinement
of crystal unit cell with powder diffraction
data by automatic computer indexing
method (abstract), Program, Annual
Meeting, American Crystallographic
Association, March 28, Cambridge, Mass.,
pp. 42-43, 1963.
Atkins, F. B., The pyroxenes of the Bush-
veld igneous complex, central Transvaal,
unpublished D. Phil, thesis, Oxford Uni-
versity, 1965.
Baker, P. E., The geology of the Mount
Misery volcano, Saint Kitts, unpublished
D. Phil, thesis, Oxford University, 1963.
Baker, P. E., The geological history of
Mount Misery volcano, Saint Kitts, West
Indies, Overseas Geol. Mineral Resources,
Gt. Brit., in press, 1967.
GEOPHYSICAL LABORATORY
561
Bancroft, G. M., R. G. Burns, and R. A.
Howie, Determination of the cation dis-
tribution in the orthopyroxene series by
the Mossbauer effect, Nature, 213, 1221-
1223, 1967.
Bartlett, E. P., H. L. Cupples, and T. H.
Tremearne, The compressibility isotherms
of hydrogen, nitrogen and a 3:1 mixture
of these gases at temperatures between
0 and 400° and at pressures to 1000
atmospheres, J. Am. Chem. Soc, 50, 1275-
1288, 1928.
Barton, P. B., and P. Toulmin, The elec-
trum-tarnish method for the determina-
tion of the fugacity of sulfur in labora-
tory sulfide systems, Geochim. Cosmochim.
Acta, 28, 619-640, 1964.
Bateman, P. C, L. D. Clark, N. K. Huber,
J. G. Moore, and C. O. Rinehart, The
Sierra Nevada batholith — a synthesis of
recent work across the central part. U.S.
Geol. Surv. Prof. Paper, bH-D, 46 pp.,
1963.
Bauer, M. (translated by L. J. Spencer),
Precious Stones, Griffin, London, 1904.
Beaman, D. R., Evaluation of correction
procedures used in electron probe micro-
analysis with emphasis on atomic number
interval 13-33, Anal. Chem., 39, 418-426,
1967.
Bell, P. M., and J. L. England, High pres-
sure experimental techniques, in Re-
searches in Geochemistry, Vol. 2, edited
by P. H. Abelson, John Wiley and Sons,
Inc., New York, 619-638, 1967.
Binns, R. A., J. V. P. Long, and S. J. B.
Reed, Some naturally occurring members
of the clinoenstatite-clinoferrosilite min-
eral series, Nature, 198, 117-778, 1963.
Birch, F., E. C. Robertson, and S. P. Clark,
Apparatus for pressures of 27,000 bars
and temperatures of 1400 °C, Ind. Eng.
Chem., U9, 1965-1966, 1957.
Bowen, N. L., The crystallization of haplo-
basaltic, haplodioritic, and related mag-
mas, Am. J. Sci., 40, 161-185, 1915.
Bowen, N. L., Genetic features of alnoitic
rocks at Isle Cadieux, Quebec, Am. J. Sci.,
3, 1-34, 1922.
Bowen, N. L., J. F. Schairer, and E.
Posnjak, The system CaO-FeO-Si02, Am.
J. Sci., 26, 193-284, 1933.
Bowen, N. L., and O. F. Tuttle, The system
MgO-Si02-H20, Geol. Soc. Am. Bull, 60,
439-460, 1949.
Bown, M. G., and P. Gay, Observations on
pigeonite, Acta Cryst., 10, 440-441, 1957.
Boyd, F. R., and J. L. England, Apparatus
for phase-equilibrium measurements at
pressures up to 50 kb and temperatures
up to 1750 °C, J. Geophys. Res., 65, 741-
748, 1960a.
Boyd, F. R., and J. L. England, The quartz-
coesite transition, J. Geophys. Res., 65,
749-756, 19606.
Boyd, F. R., and J. L. England, Effect of
pressure on the melting of diopside,
CaMgSi206, and albite, NaAlSu08, in the
range up to 50 kilobars, J. Geophys. Res.,
68, 311-323, 1963.
Boyd, F. R., J. L. England, and B. T. C.
Davis, Effects of pressure on the melting
and polymorphism of enstatite, MgSiOQ,
J. Geophys. Res., 69, 2101-2109, 1964.
Boyd, F. R., and J. F. Schairer, The sys-
tem MgSi03-CaMgSi206, J. Petrol., 5,
275-309, 1964.
Brett, R., and G. Kullerud, The Fe-Pb-S sys-
tem, Econ. Geol., 62, 354-369, 1967.
Bridgman, P. W., Thermo-electromotive
force, Peltier heat, and Thompson heat
under pressure, Proc. Am. Acad., 53, 269-
386, 1918.
Bridgman, P. W., Linear compressibility of
fourteen natural crystals, Am. J. Sci., 10,
483-498, 1925.
Bridgman, P. W., The high pressure be-
havior of miscellaneous minerals, Am. J.
Sci., 36, 81-94, 1938.
Brown, G. M., Pyroxenes from the early and
middle stages of fractionation of the
Skaergaard intrusion, east Greenland,
Mineral. Mag., 31, 511-543, 1957.
Brown, G. M., and E. A. Vincent, Pyroxenes
from the late stages of fractionation of
the Skaergaard intrusion, east Greenland,
J. Petrol, U, 175-197, 1963.
Brown, W. L., and J. V. Smith, High-tem-
perature X-ray studies on the polymor-
phism of MgSi03, Z. Krist., 118, 186-212,
1963.
Buchans, R., and J. H. Blows, Geology and
mineralogy of a millerite nickel ore de-
posit, Marbridge No. 2 Mine, Malartic,
Quebec, paper presented at the Canadian
Institute of Mining Annual Meeting,
Ottawa, March 1967.
Buddington, A. F., and D. H. Lindsley, Iron-
titanium oxide minerals and synthetic
equivalents, J. Petrol, 5, 310-357, 1964.
Buerger, M. J., The stuffed derivatives of
the silica structures, Am. Mineralogist,
39, 600-614, 1954.
Bundy, F. P., H. P. Bovenkerk, H. M.
Strong, and R. H. Wentorf, Diamond-
graphite equilibrium line from growth and
graphitization of diamond, J. Chem. Phys.,
35, 383-391, 1961.
562
CARNEGIE INSTITUTION
Busing, W. R., K. 0. Martin, and H. A.
Levy. ORFEE, a Fortran crystallographic
function and error prog-ram. Oak Ridge
XcJiojial Laboratory ORXL-TM-S06, 1966.
Carmichael, I. S. E., The petrology of
Thingmuli, a Tertiary volcano in eastern
Iceland. J. Petrol. 5." 435-460, 1964.
Chao, E. C. T.. E. M. Shoemaker, and B. A.
Madsen, First natural occurrence of
coesite, Science. 132, 220-222, 1960.
Chaves. F., and W. Kruskal. An approxi-
mate statistical test for correlations be-
tween proportions, J. Geol, 74, 692-702,
196o.
Chinner, G. A., The kyanite isograd in Glen
Clova, Angus, Scotland, Mineral. Mag.,
34, 132-143, 1965.
Clark. A. H., Stability field of monoclinic
pvrrhotite, Trans. Can. Inst. Mining Met.,
75, 232-235, 1966.
Clark, L. A., Geology and geothermometry
of the Marbridge nickel deposit, Malartic,
Quebec, Econ. Geo!., 6C, 792-811, 1965.
Clark, L. A., and G. Kullerud, The sulfur-
rich portion of the Fe-Ni-S system,
Econ. Geol., 58, 853-885, 1963.
Clark, S. P., E. C. Robertson, and F. Birch,
Experimental determination of kyanite-
sillimanite equilibrium relations at high
temperatures and pressures, Am. J. Sci.,
255, 628-640, 1957.
Coleman, L. C, Effect of ionic substitution
on the unit-cell dimensions of synthetic
diopside, in Petrologic Studies: A Volume
in Honor of A. F. Buddington, edited by
A. E. J. Engel, H. L. James, and B. F.
Leonard, pp. 429-446, Geological Society
of America, New York, 1962.
Craig, J. R., and G. Kullerud, Phase rela-
tions in the Cu-Fe-Ni-S system and their
application to magmatic ore deposits,
Sympo.-ium on Magmatic Ore Deposits,
Stanford Univ., Calif., Society of Eco-
nomic Geologists, in press, 1967.
Crowley, M. S., and R. Roy, Crystalline
solubility in the muscovite and phlogopite
groups, Am. Mineralogist, 48, 348-362,
1964.
Dallwitz, W. B., D. H. Green, and J. E.
Thompson, Clinoenstatite in a volcanic
rock from the Cape Vogel area, Papua,
.7. Petrol, 7, 375-403, 1966.
Daly, R. A., Bushveld igneous complex of
the Transvaal, Geol. Soc. Am. Bull, 39,
703-768, 1928.
Darken, L. S., and R. W. Gurry, The system
iron-oxygen, I, The wiistite field and re-
lated equilibria, J. Am, Chem. Soc, 07,
1398-1412, 1945.
Darken, I,. S., and R. W. Gurry, The system
iron-oxygen, II, Equilibrium and thermo-
dynamics of liquid oxide and other phases,
J. Am. Chem. Soc, 68, 798-816, 1946.
Darken, L. S., and R. W. Gurry, Physical
Chemistry of Metals, McGraw-Hill Book
Co., Inc., New York, 1953.
Davis, B. L., X-ray diffraction data on two
high-pressure phases of calcium carbon-
ate, Science, 145, 489-491, 1964.
Deer, W. A., R. A. Howie, and J. Zussman,
Rock-Forming Minerals, Vol. 1, John
Wiley and Sons, Inc., New York, 1962.
de Wys, E. C, and W. R. Foster, The binary
system anorthite (CaAl2Si208)-akerma-
nite (CaoMgSio07), J. Am. Ceram. Soc,
39, 372-376, 19~56.
Dimroth, Erich, Untersuchungen zum
Mechanismus von Blastesis und Syntexis
in Phylliten und Hornfelsen des siidwest-
lichen Fichtelgebirges, Tschermaks Min-
eral. Petrog. Mitt., 3rd Ser., 8, 248-274,
1962a.
Dimroth, Erich, Eine Theorie der Kornge-
fiigestatistik, Neues Jahrb. Mineral,
Monatsh., 218-229, 19626.
Dimroth, Erich, Fortschritte der Gefiigesta-
tistik, Neues Jahrb. Mineral, Monatsh.,
186-192, 1963.
Donath, F. A., and R. B. Parker, Folds and
folding, Geol. Soc Am. Bull, 75, 45-62,
1964.
Donnay, G., J. D. H. Donnay, and J. V.
Hurst, Precession goniometry to identify
neighboring twins, Acta Cryst., 8, 507-
509, 1955.
Donnelly, T. W., and J. J. W. Rogers, Crust
versus mantle derivation of eastern Antil-
lean igneous rocks (abstract), Trans. Am.
Geophys. Union, 48, 253, 1967.
Drever, H. I., and R. Johnston, A natural
high-lime silicate liquid more basic than
basalt, J. Petrol, 7, 414-420, 1966.
Duncumb, P., and P. K. Shields, The present
state of quantitative X-ray microanalysis,
Part I, Physical basis, Brit. J. Appl
Phys., 14, 617-625, 1963.
Eckermann, H. von, The alkaline district of
Alno Island, Sveriges Geol. Undersokn.,
Arsbok, Ser. Ca, Avhandl Uppsat., no.
36, 1948.
Eckermann, H. von, The alkaline and car-
bonatite dikes of the Alno formation on
the mainland north-west of Alno Island,
Kungl Svenska Vetenk. Handl, Fj. S.,
7, no. 2, 1958.
Emery, K. O., The Sea Off Southern Cali-
fornia, John Wiley and Sons, Inc., New
York, 1960.
GEOPHYSICAL LABORATORY
563
Ernst, W. G., Significance of phengitic
micas from low-grade schists, Am. Miner-
alogist, 48, 1357-1373, 1963.
Ernst, W. G., Synthesis and stability rela-
tions of ferrotremolite, Am. J. Sci., 264,
37-65, 1966.
Eugster, H. P., and D. R. Wones, Stability
relations of the ferruginous biotite, an-
nite, J. Petrol, 3, 82-125, 1962.
Evans, B. J., S. Ghose, and S. Hafner,
Hyperfine splitting of Fe57 and MgFe or-
der-disorder in orthopyroxenes (MgSiOs-
FeSi03 solid solution), J. Geol, 75, 306-
322, 1967.
Evans, B. W., Application of a reaction rate
method to the breakdown equilibria of
muscovite and muscovite plus quartz,
Am. J. Sci., 263, 647-667, 1965.
Evans, B. W., and C. V. Guidotti, The silli-
manite-potash feldspar isograd in west-
ern Maine, U.S.A., Contrib. Mineral.
Petrol, 12, 25-62, 1966.
Fawcett, J. J., and H. S. Yoder, Jr., Phase
relations of chlorites in the system MgO-
Al203-Si02-H20, Am. Mineralogist, 51,
353-380, 1966.
Fermor, L. L., On a new chrome-garnet,
Geol. Mag., 89, 145-147, 1952.
Foshag, W. F., and J. Gonzalez R., Birth
and development of Paricutin volcano,
Mexico, U.S. Geol Surv. Bull, 965-D,
355-489, 1956.
Foster, P. K., and A. J. E. Welch, Metal-
oxide solid solutions, Part 1, Lattice-con-
stant and phase relationships in ferrous
oxide (wustite) and in solid solutions of
ferrous oxide and manganous oxide,
Trans. Faraday Soc., 52, 1626-1635, 1956.
Francis, G. H., and M. H. Hey, The unit-
cell constants of anthophyllite, Mineral.
Mag., 31, 173-186, 1956.
Fudali, R. F., Oxygen fugacities of basaltic
and andesitic magmas, Geochim. Cos-
mochim. Acta, 29, 1063-1075, 1965.
Futergendler, S. I., X-ray study of solid
mineral inclusions in diamond, Kristallo-
grafiya, 3, 494-496, 1958.
Futergendler, S. I., and Va. A. Frank-
Kamenetsky, Orientated growth of oli-
vine, garnet, and chrome-spinel in dia-
mond, Zap. Vses. Mineral. Obshchestva,
90, 230-236, 1961.
Fyfe, W. S., F. J. Turner, and J. Verhoogen,
Metamorphic reactions and metamorphic
facies, Geol. Soc. Am. Mem., 73, 1958.
Garlick, G. D., and S. Epstein, Oxygen
isotope ratios in coexisting minerals of
regionally metamorphosed rocks, Geochim.
Cosmochim. Acta, 31, 181-214, 1967.
Geological Survey of Canada, Age deter-
minations by the Geological Survey of
Canada, Geol Surv. Canada Papers 60-
17-66-17, 1960-1966.
Green, T. H., High pressure experiments on
the genesis of anorthosites, in Petrology
of the Upper Mantle, Publication no. 444,
Department of Geophysics and Geochem-
istry, Australian National University,
206-233, 1966.
Greenwood, H. J., The synthesis and sta-
bility of anthophyllite, J. Petrol, 4, 317-
351, 1963.
Greig, J. W., E. Posnjak, H. E. Merwin, and
R. B. Sosman, Equilibrium relationships
of Fe304, Fe20.3, and oxygen, Am. J. Sci.,
30, 239-316, 1935.
Guven, N., and C. W. Burnham, The crystal
structure of 3T muscovite, Z. Krist., in
press, 1967.
Hagg, Gunnar, Die Kristallstruktur des
Magnetischen Ferrioxyds, y-Fe20o, Z.
Physik. Chem., 29, B, 95-102, 1935. '
Hallimond, A. F., Optically uniaxial augite
from Mull, Mineral Mag., 17, 97-99,
1914.
Hanneman, R. E., and H. M. Strong, Pres-
sure dependence of electromotive force of
thermocouples to 1300 °C and 50 kb, Gen-
eral Electric Research Lab. Rept. 64-RL-
3725Y, Schenectady, N. Y., 11 pp., 1964.
Hansen, M., and K. Anderko, Constitution of
Binary Alloys, McGraw-Hill Book Co.,
Inc., New York, 1958.
Harker, A., Metamorphism, Methuen and
Co., Ltd., London, 1950.
Harker, R. I., and O. F. Tuttle, The lower
limit of stability of akermanite, Ca2MgSi2
Or, Am. J. Sci., 254, 468-478, 1956.
Harris, J. W., R. C. Henriques, and H. O. A.
Meyer, Orientation of silicate mineral in-
clusions in natural diamond (abstract),
Acta Cry st., 21, 259-260, 1966.
Haul, R., and Th. Schoon, Zur Struktur des
Ferromagnetischen Eisen (IH)-Oxyds
y-Fe203, Z. Physik. Chem., 44, B, 216-226,
1939.
Hawley, J. E., The Sudbury ores: their
mineralogy and origin, Can. Mineralogist,
7, 1-207, 1962.
Hayward, C. R., The equilibrium diagram of
the system Cu9S-Ni3S„, Trans. AIME, 48,
141-152, 1915.
Henry, N. F. M., Lamellar structure in
orthopyroxenes, Mineral Mag., 26, 179-
189, 1942.
564
CARNEGIE INSTITUTION
Hess, H. H., Pyroxenes of common mafic
magmas, Am. Mineralogist, 26, 515-535,
573-594, 1041.
Hess. H. H., Chemical composition and
optical properties of common clinopyrox-
enes. Part I, Am. Mineralogist, S4, 621-
666, 1940.
Hess. H. H., Stillwater igneous complex,
Montana: a quantitative mineralogies!
study, Geol. Soc. Am. Mem., SO, 1-230,
1960.
Hess, H. H.. and E. P. Henderson, The
Moore County meteorite: a further study
with comment on its primordial environ-
ment. Am. Mineralogist, 34, 494-507,
1949.
Hilty, D. C, and W. Crafts, Liquidus sur-
face of the Fe-S-0 system, J. Metals, 4,
1307-1312, 1952.
Holborn, L., and J. Otto, liber die Isother-
men einiger Gase bis 400° und ihre
Bedeutung fur das Gasthermometer, Z.
Physik, 23, 77-94, 1924.
Holgersson, S., Rontgenographische Unter-
suchungen der Mineralien der Spinell-
gruppe und von synthetisch dargestellten
Substanzen, von Spinelltypen, Lunds
Univ. Arsskr. Avd. 2, 23, no. 9, 5-107,
1927.
Holmes, A., Petrographic Methods and Cal-
culations, Murby and Co., London, 1921.
Holmes, A., and H. F. Harwood, The petrol-
ogy of the volcanic area of Bufumbira,
Geol. Surv. Uganda, Mem. HI, part II,
1-300, 1937.
Holser, W. T., Fugacity of water at high
temperatures and pressures, J. Phys.
Chem., 58, 316-317, 1954.
Hurst, J. V., J. D. H. Donnay, and G. Don-
nay, Staurolite twinning, Mineral. Mag.,
31, 145-163, 1956.
Hutchinson, R. W., Genesis of Canadian
Massive sulfide deposits reconsidered by
comparison to Cyprus deposits, Bull. Can.
Inst. Mining Met., 58, 972-986, 1965.
Ito, T., N. Morimoto, and R. Sadanaga, On
the structure of epidote, Acta Cryst., 7,
53-59, 1954.
Jamieson, J. C, Introductory studies of
high-pressure polymorphism to 24,000
bars by X-ray diffraction with some com-
ments on calcite II, J. Geol., 65, 334-343,
1957.
JANAF (Joint Army, Navy, Air Force)
Tables of Thermo chemical Data, compiled
by the Dow Chemical Co. Thermal Labo-
ratory, Midland, Michigan, 1960.
Jensen, E., Pyrrhotite: melting relations
and composition, Am. J. Sci., 240, 695-
709, 1942.
Jensen, E., Melting relations of chalcocite,
Avhandl. Norske Videnskaps-Akad. Oslo,
I. Mat.-Naturv. Kl, 6, 14 pp., 1947.
Johnston, J., H. E. Merwin, and E. D.
Williamson, The several forms of calcium
carbonate, Am. J. Sci., 41, 473-512, 1916.
Jurg, J. W., and E. Eisma, Petroleum
hydrocarbons: generation from fatty
acids, Science, 144, 1451, 1964.
Karle, Isabella L., and J. Karle, An applica-
tion of a new phase determination pro-
cedure to the structure of cycle (hexa-
glycyl) hemihydrate, Acta Cryst., 16,
969-975, 1963.
Karle, J., and I. L. Karle, The symbolic
addition procedure for phase determina-
tion for centrosymmetric and noncentro-
symmetric crystals, Acta Cryst., 21, 849-
859, 1966.
Kashkai, M.-A., The augites from Talysh
(Azerbaidzhan), Dokl. Akad. Nauk SSSR,
43, 351-353, 1944.
Kawano, Y., and K. Aoki, Some anorthite-
bearing volcanic rocks in Japan, Sci. Rept.
Tohoku Univ., Ser. 3, 6, 431-437, 1960.
Keith, M. L., and J. F. Schairer, The stabil-
ity field of sapphirine in the system MgO-
Al203-Si02-H20, J. Geol, 60, 181-186,
1952.
Kennedy, G. C, G. J. Wasserburg, H. C.
Heard, and R. C. Newton, The upper
three-phase region in the system Si02-
H20, Am. J. Sci., 260, 501-521, 1962.
Knopf, E. B., and Earl Ingerson, Structural
petrology, Geol. Soc. Am. Mem., 6, 270 pp.,
1938.
Koffman, A. R., R. B. Cairns, and R. L.
Price, Recent evidence concerning the
occurrence and deposition of sulfides in
the pre-Cambrian shield (abstract), Bull.
Can. Inst. Mining Met., 55, 194, 1962.
Krause, O., and W. Thiel, Uber Keramische
Farbkorper, I, Spinellfarbkorper und
Spinelle, Ber. Deut. Keram. Ges., 15, 101-
110, 1934.
Kullerud, G., The FeS-ZnS system, a geo-
logical thermometer, Norsk Geol. Tidsskr.,
32, 62-147, 1953.
Kullerud, G., Thermal stability of pent-
landite, Can. Mineralogist, 7, 353-366,
1963.
Kullerud, G., Review and evaluation of re-
cent research on geologically significant
sulfide-type systems, Fortschr. Mineral.,
41, 221-270, 1964.
GEOPHYSICAL LABORATORY
565
Kullerud, G., Covellite stability relations in
the Cu-S system, Freiberger For-
schungsh., C, 186, 145-160, 1965.
Kullerud, G., and E. H. Roseboom, Jr.,
Cu9S5-Cu5FeS4 system (abstract), Geol.
Soc. Am. Bull, 69, 1602, 1958.
Kullerud, G., and H. S. Yoder, Jr., Pyrite
stability relations in the Fe-S system,
Econ. Geol., 54, 533-572, 1959.
Kullerud, G., and R. A. Yund, The Ni-S
system and related minerals, J. Petrol., 3,
126-175, 1962.
Kullerud, G., R. A. Yund, and G. Moh,
Phase relations in the Fe-Ni-S, Cu-Fe-S,
and Cu-Ni-S systems, Symposium on
Magmatic Ore Deposits, Stanford Univ.,
Calif., Society of Economic Geologists,
in press, 1967.
Kuno, H., Petrology of Hakone volcano and
the adjacent areas, Japan, Geol. Soc. Am.
Bull, 61, 957-1020, 1950.
Kuno, H., and K. Nagashima, Chemical com-
positions of hypersthene and pigeonite
in equilibrium with magma, Am. Min-
eralogist, 47, 1000-1006, 1952.
Larsen, E. S., Alkalic rocks of Iron Hill,
Gunnison County, Colorado, U.S. Geol.
Surv. Prof. Paper, 197 -A, 1942.
Lewis, J. F., Mineralogical and petrological
studies of plutonic blocks from the
Soufriere volcano, St. Vincent, B.W.I., un-
published D. Phil, thesis, Oxford Univer-
sity, 1964.
Lewis, J. F., The origin of the ejected
plutonic blocks of the Soufriere, St. Vin-
cent, Trans. Caribbean Geol Conf., 4th,
Trinidad, in press, 1967.
Lewis, J. F., and E. W. White, Pyroxene
relations in basalts and basaltic andesites
from the Soufriere volcano, St. Vincent,
West Indies (abstract), Trans. Am.
Geophys. Union, 48, 227, 1967.
Lihl, F., tiber ein pyrophores Eisenoxyd,
Oesterr. Akad. Wiss., Math.-Naturw.
Kl, Abt. lib, 159, 632-646, 1950.
Lindsley, D. H., Melting relations of
KAlSi308: effect of pressures up to 40
kb, Am. Mineralogist, 51, 1793-1799, 1966.
Lindsley, D. H., Melting relations of plagi-
oclase at high pressures, in Origin of
Anorthosite, Proceedings, edited by Y.
Isachsen, New York State Geological
Survey, in press, 1967.
Lundqvist, D., X-ray studies on the ternary
system Fe-Ni-S, Arkiv Kemi, Mineral,
Geol, 24A, no. 22, 12 pp., 1947.
Lupanova, N. P., Basaltic hornblende and
augite from monchikite of Hibina Moun-
tains, Trudy Petroar. Inst. Akn/1. Navk
SSSR, no. 6, 53-64, 1934; Mineral Abstr.,
6, 420, 1937.
Luth, W. C, and C. O. Ingamells, Gel prep-
aration of starting materials for hy-
drothermal experimentation, Am. Min-
eralogist, 50, 255-258, 1965.
Luth, W. C, R. H. Jahns, and O. F. Tuttle,
The granite system at pressures of 4 to
10 kb, J. Geophys. Res., 69, 759-773, 1964.
Macres, V. G., Geometrical considerations
involving electron beam, target, and
analyzed X-ray beam in the electron
microprobe, unpublished manuscript,
1963.
Marzolf, J., J. Dehn, and J. Salmon, Moss-
bauer studies of tektites, pyroxenes, and
olivines, in The Mossbauer Effect and
its Applications in Chemistry, edited by
Robert Gould, Advances in Chemistry
Series, American Chemical Society, in
press, 1967.
Mason, B., Mineralogical aspects of the
system FeO-Fe203, MnO-Mn203, Geol.
Foren. Stockholm Forh., 65, 97-180, 1943.
Mason, B. H., Cummingtonite from the
Mikonvi River, Westland, New Zealand,
Am. Mineralogist, 38, 862-865, 1953.
Mason, B. H., and T. Berggren, A phos-
phate-bearing spessartite garnet from
Wodgina, western Australia, Geol. Foren.
Stockholm Forh., 63, 413-418, 1941.
Matousek, J. W., and C. S. Samis, The
thermodynamic properties of Cu-Ni-S
mattes, Trans. AIME, 227, 980-985, 1963.
McDonald, W. S., and D. W. J. Cruickshank,
A refinement of the structure of sodium
metasilicate, Acta Cryst., 22, 37-43, 1967.
McKie, D., Yoderite, a new hydrous magne-
sium iron aluminosilicate from Mautia
Hill, Tanganyika, Mineral. Mag., 32, 282-
307, 1959.
McKie, D., and N. Bradshaw, A green
variety of yoderite, Nature, 210, 1148,
1966.
Meyer, H. J., Bildung und Morphologie des
Vaterits, Z. Krist., 121, 220-242, 1965.
Meyer, H. O. A., High-pressure-high-tem-
perature X-ray diffraction apparatus (ab-
stract), Acta Cryst, 21, A 220, 1966.
Milledge, H. J., Coesite as an inclusion in
G.E.C. synthetic diamonds, Nature, 190,
1181, 1961.
Mitchell, R. S., and A. A. Giardini,
Oriented olivine inclusions in diamond,
Am. Mineralogist, 38, 136-138, 1953.
566
CARNEGIE INSTITUTION
Morey. G. W., and W. T. Chen, The action
of water on some feldspars, Am. Min-
eralogist. 10, 996-1000, 1955.
Morimoto, X., D. E. Appleman, and H. T.
Evans, Jr., The crystal structures of
clinoenstatite and pigeonite, Z. Krist., 114,
120-147, 1960.
Morimoto. N., and G. Kullerud, Polymor-
phism in bornite, Am. Mineralogist, 40,
1270-1282, 1961.
Morimoto, N., and G. Kullerud, Polymor-
phism in digenite, Am. Mineralogist, 48,
110-123, 1963.
Muan, A., Silver-palladium alloys as cruci-
ble material in studies of low-melting
iron silicates, Am. Ceram. Soc. Bull., 42,
344-347, 1963.
Muan A., and E. F. Osborn, Phase equilibria
at liquidus temperatures in the system
M^O-FeO-Fe.Og-SiO.,, J. Am. Ceram.
Soc, 39, 121-140, 1956.
Muir, I. D., The clinopyroxenes of the
Skaergaard intrusion, eastern Greenland,
Mineral Mag., 29, 690-714, 1951.
Miiller, K., Augit vom Vesuv, Neues Jahrb.
Mineral, Centralbl, A, 116-122, 1936.
Xagamori, M., and M. Kameda, Equilibria
between Fe-S-0 system melts and CO-
CO.-S00 gas mixtures at 1200°C, Trans.
Japan Inst. Metals, 6, 21-30, 1965.
Xaldrett, A. J., J. R. Craig, and G.
Kullerud, The central portion of the Fe-
Ni-S system and its bearing on pent-
landite exsolution in iron-nickel sulfide
ores, Econ. Geol, 62, 826-847, 1967.
Xaldrett, A. J., and G. Kullerud, A study of
the Strathcona mine and its bearing on
the origin of the nickel-copper ores of
the Sudbury District, Ontario, J. Petrol.,
8, 453-531, 1967.
Xaray-Szabo, I., and Sasvari, On the struc-
ture of staurolite, HFe2Al9Si4024, Acta
Cry st., 11, 862-865, 1958.
Newton, R. C, Kyanite-andalusite equilib-
rium from 700° to 800°C, Science, 153,
170-172, 1966.
Xixon, P. H., A mineralogical and geochem-
ical study of kimberlites and their
associated xenoliths, Ph.D. thesis, Univer-
sity of Leeds, 1960.
Xixon, P. H., O. von Knorring, and J. M.
Rooke, Kimberlites and associated in-
clusions of Basutoland: a mineralogical
and geochemical study, Am. Mineralogist,
48, 1000-1132, 1963.
Xolan, J., and A. D. Edgar, An X-ray in-
vestigation of synthetic pyroxenes in the
system acmite-diopside-water at 1000
kb/cm2 water-vapour pressure, Mineral.
Mag., S3, 625-634, 1963.
Oberhoffer, P., Das Technische Eisen, Julius
Springer, Berlin, 1925.
Officer, C. B., J. I. Ewing, et al., Geophysical
investigation in the eastern Caribbean:
summary of 1955 and 1956 cruises, Phys.
Chem. Earth, 3, 17-109, 1959.
O'Hara, M. J., Primary magmas and the
origin of basalts, Scot. J. Geol., 1, 19-40,
1965.
O'Neil, J. R., and R. N. Clayton, Oxygen
isotope geothermometry, in Isotopic and
Cosmic Chemistry, edited by H. Craig,
S. L. Miller, and G. J. Wasserburg, pp.
157-168, North Holland Publishing Co.,
Amsterdam, 1964.
Orville, P. M., and H. J. Greenwood, Deter-
mination of AH' of reaction from experi-
mental pressure-temperature curves, Am.
J. Sci., 263, 678-683, 1965.
Osborn, E. F., The system CaSiOs-diopside-
anorthite, Am. J. Sci., 240, 751-788, 1942.
Osborn, E. F., Role of oxygen pressure in
the crystallization and differentiation of
basaltic magma, Am. J. Sci., 257, 609-
647, 1959.
Osborn, E. F., and D. B. Tait, The system
diopside-forsterite-anorthite, Am. J. Sci.,
Bowen Vol., 413-433, 1942.
Peacor, D. R., Refinement of the crystal
structure of a pyroxene of formula
MiMn (Sii.5Alo.5)06, Am. Mineralogist,
52, 31-41, 1967.
Perrotta, A. J., and D. A. Stephenson,
Clinoenstatite: high-low inversion, Sci-
ence, 148, 1090-1091, 1965.
Philibert, J. A., A method for calculating
the absorption correction in electron-probe
microanalysis, in Third International
Symposium on X-ray Optics and Micro-
analysis, edited by H. H. Pattee, Jr., V. E.
Cosslett, and A. Engstrom, pp. 379-392,
Academic Press, New York, 1963.
Phillips, B., and A. Muan, Stability rela-
tions of iron oxides: phase equilibria in
the system Fes04-Fe203 at oxygen pres-
sures up to 45 atmospheres, J. Phys.
Chem., 64, 1451-1453, 1960.
Philpotts, A. R., Origin of the anorthosite-
mangerite rocks in southern Quebec, J.
Petrol, 7, 1-64, 1966.
Pouillard, E., Sur le comportement de
l'alumine et de l'oxyde de titane vis-a-vis
de oxydes de fer, Ann. Chim. (Paris), 5,
164-214, 1950.
GEOPHYSICAL LABORATORY
567
Presnall, D. C, The join forsterite-diopside-
iron oxide and its bearing on the crystal-
lization of basaltic and ultramafic mag-
mas, Am. J. Scl, 264, 753-809, 1966.
Preston, J., An unusual hourglass structure
in augite, Am. Mineralogist, 51, 1227-
1233, 1966.
Prewitt, C. T., Unpublished least-squares
computer program, 1962.
Prewitt, C. T., and Charles W. Burnham,
The crystal structure of jadeite,
NaAlSi206, Am. Mineralogist, 51, 956-
975, 1966.
Pulfrey, W., Ijolitic rocks near Homa Bay,
western Kenya, Quart. J. Geol. Soc.
(London), 105, 425-459, 1950.
Putman, H.-M., Analyse chimique d'une
augite du Cap Tourmente, Naturaliste
Canadien, 69, 261-263, 1942.
Quareni, S., and R. de Pierri, La struttura
dell'andradite, Mem. Accad. Patavina,
Sri., Lettre, Arte, CI. Sci. Mat. Nat, 78
(1956-1966), 1966.
Radoslovich, E. W., The structure of mus-
covite, KAl2(Si3Al)O10(OH)2, Acta
Cryst., 13, 919-932, 1960.
Radoslovich, E. W., The cell dimensions and
symmetry of layer lattice silicates, IV,
Interatomic forces, Am. Mineralogist, 48,
76-99, 1963.
Ramsay, J. G., The deformation of early
linear structures in areas of repeated
folding, J. Geol, 68, 75-93, 1960.
Reed, S. J. B., Characteristic fluorescence
corrections in electron-probe microanaly-
sis, Brit. J. Appl. Phys., 16, 913-926, 1965.
Riecker, R. E., and T. P. Rooney, Shear
strength, polymorphism, and mechanical
behavior of olivine, enstatite, diopside,
labradorite, and pyrope garnet: tests to
920° and 60 kb, U.S. Air Force Office of
Aerospace Res., Environmental Res.
Paper, 216, 1966.
Robson, G. R., The pyroclastic deposits of
the 1902-3 eruptions of the Soufriere
volcano, St. Vincent, Trans. Caribbean
Geol. Conf., 4-th, Trinidad, in press, 1967.
Robson, G. R., and J. F. Tomblin, Catalogue
of the Active Volcanoes of the World,
Part 20, West Indies, International Asso-
ciation of Volcanology, Rome, 1967.
Roy, D. M., and R. Roy, Synthesis and sta-
bility of minerals in the system MgO-
Al20.s-Si02-H20, Am. Mineralogist, 4.0,
147-178, 1955.
Roy, R., and E. F. Osborn, The system
lithium metasilicate-spodumene-silica, J.
Am. Chem. Soc, 49, 2086-2095, 1949.
Roy, R., D. Roy, and E. F. Osborn, Composi-
tional and stability relationships among
the lithium alumino-silicates, eucryptite,
spodumene, and petalite, J. Am,. Ceram.
Soc, 33, 152-159, 1950.
Scarfe, C. M., W. C. Luth, and O. F. Tuttle,
An experimental study bearing on the
absence of leucite in plutonic rocks, Am.
Mineralogist, 51, 726-735, 1966.
Schairer, J. F., Melting relations of the
common rock-forming oxides, J. Am.
Ceram. Soc, 40, 215-235, 1957.
Schairer, J. F., and N. L. Bowen, The
binary system CaSiOs-diopside and the
relations between CaSi03 and akermanite,
Am. J. Sci., 240, 725-742, 1942.
Schairer, J. F., and N. L. Bowen, The sys-
tem K20-Al203-Si02, Am J. Sci., 253,
681-746, 1955.
Schairer, J. F., and N. L. Bowen, The sys-
tem Na20-Al203-Si02, Am. J. Sci., 254,
129-195, 1956.
Schairer, J. F., and H. S. Yoder, Jr., The
nature of residual liquids from crystal-
lization, with data on the system nephe-
line-diopside-silica, Am. J. Sci., Bradley
Vol., 258-A, 273-283, 1960.
Schenk, R., H. Franz, and A. Leymann,
Gleichgewichtsuntersuchung iiber die
Reduktions-, Oxidations- und Kohlungs-
vorgange beim Eisen, XI, Z. Anorg. All-
gem. Chem., 206, 129-151, 1932.
Schlegel, H., and A. Schiiller, Die Schmelz-
und Kristallisations Gleichgewichte im
System Kupfer-Eisen-Schwefel und ihre
Bedeutung fur die Kupfergewinnung,
Freiberger Forschungsh., B, 2, 31 pp.,
1952.
Schreyer, W., and G. A. Chinner, Staurolite-
quartzite bands in kyanite quartzite at
Big Rock, Rio Arriba County, New
Mexico, Contrib. Mineral. Petrol, 12, 223-
244, 1966.
Schreyer, W., and H. S. Yoder, Jr., The
system Mg-cordierite-IbO and related
rocks, Neues Jahrb. Mineral., Abhandl,
101, 271-342, 1964.
Schwerdtfeger, K., and A. Muan, Activities
in olivine and pyroxenoid solid solutions
of the system Fe-Mn-Si-0 at 1150°C,
Trans. AIME, 236, 201-211, 1966.
Sclar, C. B., L. C. Carrison, and C. M.
Schwartz, High-pressure stability field of
clinoenstatite and the orthoenstatite-
clinoenstatite transition (abstract) ,
Trans. Am. Geophys. Union, 45, 121, 1964.
Segnit, E. R., Some data on synthetic alu-
minous and other pyroxenes, Mineral
Mag., 30, 218-226, 1951.
56S
CARNEGIE INSTITUTION
Sharp, W. E., Pyrrhotite: a common in-
clusion in South African diamonds,
Nature, :il, 402-403, 1966.
Shaw. H. R., Hydrogen-water vapor mix-
tures: control of hydrothermal atmos-
pheres by hydrogen osmosis, Science,
UP. 1220-1222. 1903.
Shaw. H. R., and D. R. Wones, Fugacity
coefficients for hydrogen gas between 0°
and 1000 °C, for pressures to 3000 atmos-
pheres, Am. J. Scu, 262, 918-929, 1964.
Sinha. K. P., and A. P. B. Sinha, Vacancy
distribution and bonding in some oxides
of spinel structure, J. PJiys. Chcm., 61,
758-761, 1957.
Skinner, B. J., Physical properties of end-
members of the garnet group, Am. Min-
eralogist, 41, 428-436, 1956.
Skinner. B. J., and H. T. Evans, Jr., Crystal
chemistry of /?-spodumene solid solutions
on the join L^OAbCh-SiCb, Am. J. Sci.,
Bradley Vol., 25S-A, 312-324, 1960.
Skinner, B. J., and D. L. Peck, The solubil-
ity of sulfur in basic magmas, Symposium
on Magmatic Ore Deposits, Stanford
Univ., Calif., Society of Economic Geolo-
gists, in press, 1967.
Smith, C. H., and H. E. Kapp, The Muskox
intrusion, a recently discovered layered
intrusion in the Copper Mine River area,
Xorthwest Territories, Canada, Mineral.
Soc. Am. Spec. Paper, 1, 30-35, 1963.
Smith, F. G., Physical Geochemistry, Addi-
son-Wesley Publishing Co., Inc., Reading,
Mass., 1963.
Souch, B. E., T. Podolsky, and geological
staff of the International Nickel Com-
pany of Canada, Ltd., The sulfide ores of
Sudbury: their particular relation to a
distinctive inclusion-bearing facies of the
Xickel Irruptive, Symposium on Mag-
matic Ore Deposits, Stanford Univ.,
Calif., Society of Economic Geologists, in
press, 1967.
Spijkerman, J. J., F. C. Ruegg, and L. May,
Mossbauer spectroscopy in coordination
chemistry, in Mossbauer Effect Methodol-
r>f)V, Vol. 2, edited by I. J. Gruverman, pp.
85-93, Plenum Press, New York, 1965.
Sproule, K., G. A. Harcourt, and L. S.
Renzoni, Treatment of nickel-copper mat,
J. Metals, 12, 214-219, 1960.
Stanton, R. L., and J. D. Bell, Volcanic and
associated rocks of the New Georgia
group, British Solomon Islands, Over-
seas Geol. Mineral Resources, Gt. Brit.,
in press, 1967.
Stephenson, D. A., C. B. Sclar, and J. V.
Smith, Unit cell volumes of synthetic
orthoenstatite and low clinoenstatite, Min-
eral. Mag., S5, 838-846, 1966.
Stewart, D. B., The system CaAl2Si208-
SiO-2-HjO (abstract), Geol. Soc. Am.
Bull, 69, 1648, 1958.
Stishov, S. M., Equilibrium between coesite
and the rutile-like modification of silica,
Dokl. Akad. Nauk SSSR, 148, 1186-
1188, 1963.
Strunz, H., and Ch. Tennyson, Strukturelle
Deutung der Pyrit- und Markasitzwil-
linge, Neues Jahrb. Mineral., Montash.,
247-248, 1965.
Sutton, J. R., Inclusions in diamond from
South Africa, Mining Mag. (London), 19,
208-210, 1921.
Taljaard, M. S., South African melilite
basalts and their relations, Trans. Geol.
Soc. South Africa, 39, 28-316, 1937.
Thewlis, J., The structure of ferromagnetic
ferric oxide, Phil. Mag., 12, 1089-1106,
1931.
Tilley, C. E., A monticellite nepheline basalt
from Tasmania: a correction to mineral
data, Geol. Mag., 65, 29-30, 1928.
Tilley, C. E., The nephelinite of Etinde,
Cameroons, Geol. Mag., 90, 145-151, 1953.
Tilley, C. E., and H. F. Harwood, The
dolerite-chalk contact of Scawt Hill, Co.
Antrim, Mineral. Mag., 22, 439-468, 1931.
Tomblin, J. F., The volcanic history and
petrology of the Soufriere region, St.
Lucia, unpublished D. Phil, thesis, Oxford
University, 1964.
Tomblin, J. F., The geology of the Soufriere
volcanic centre, St. Lucia, Trans. Carib-
bean Geol. Conf., 4th, Trinidad, in press,
1967.
Toulmin, P., and P. B. Barton, A thermo-
dynamic study of pyrite and pyrrhotite,
Geochim. Cosmochim. Acta, 28, 641-671,
1964.
Turner, F. J., H. C. Heard, and D. T.
Griggs, Experimental deformation of
enstatite and accompanying inversion to
clinoenstatite, Intern. Geol. Congr., 21st,
Copenhagen, 1960, Rept. Session, Norden,
Part 28, 399-408, 1960.
Turner, F. J., and J. Verhoogen, Igneous
and Metamorphic Petrology, McGraw-Hill
Book Co., Inc., New York, 2nd ed., 1960.
Tuttle, O. F., and J. L. England, Prelimi-
nary report on the system Si02-H20,
Geol. Soc. Am. Bull, 66, 149-152, 1955.
Tuttle, O. F., and N. L. Bowen, Origin of
granite in the light of experimental
studies in the system NaAlSi308-
KAlSi308-Si02-H20, Geol. Soc. Am. Mem.,
74, 153 pp., 1958.
GEOPHYSICAL LABORATORY
569
Tyler, R. C, and B. C. King, The pyroxenes
of the alkaline igneous complexes of east-
ern Uganda, Mineral. Mag., 36, 5-21,
1967.
Ubbelohde, A. R., Premonitory phenomena
in phase transformation (abstract), Acta
Cry st., 16, A183, 1963.
Velde, B., Phengite micas: synthesis, sta-
bility, and natural occurrence, Am. J. Sci.,
263, 886-913, 1965.
Verhoogen, J., The chemical potential of a
stressed solid, Trans. Am. Geophys.
Union, 32, 251-258, 1951.
Vogel, R., and W. Fulling, Das system eisen-
eisensulfid (FeS)-Wiistit (FeO), Festkr.
Tillagnad J. Arvid Hedvall, 597-610, 1948.
Wager, L. R., The major element variation
of the layered series of the Skaergaard
intrusion and a re-estimation of the aver-
age composition of the Hidden Layered
Series and of the successive residual mag-
mas, J. Petrol, 1, 364-398, 1960.
Wager, L. R., Igneous cumulates from the
1902 eruption of Soufriere, St. Vincent,
Bull, volcanol, ser. 2, 24, 93-99, 1962.
Wager, L. R., and G. M. Brown, Layered
Igneous Rocks, Oliver and Boyd, Edin-
burgh, 1967.
Wager, L. R., and W. A. Deer, Geological
investigations in east Greenland, III,
Petrology of the Skaergaard intrusion,
Kangerdlugsung, east Greenland, Medd.
Groenland, 105, 355 pp., 1939.
Walker, L. R., G. K. Wertheim, and V.
Jaccarino, Interpretation of the Fe57
isomer shift, Phys. Rev. Letters, 6, 98-
101, 1961.
Walter, L. S., Experimental studies on
Bowen's decarbonation series, I, P-T uni-
variant equilibria of the "monticellite"
and "akermanite" reactions, Am. J. Sci.,
261, 488-500, 1963.
Walter, L. S., Coesite discovered in tektites,
Science, 147, 1029-1032, 1965.
Weiss, L. E., Geometry of superposed fold-
ing. Geol. Soc. Am. Bull, 70, 91-106,
1959.
Wertheim, G. K., Mossbauer Effect: Prin-
ciples and Applications, Academic Press,
New York, 1964a.
Wertheim, G. K., Mossbauer effect in chem-
istry and solid state physics, Science, 144,
253-259, 19646.
Wiebe, R., and V. L. Gaddy, The compressi-
bilities of hydrogen and of four mixtures
of hydrogen and nitrogen at 0, 25, 50, 100,
200 and 300° and to 1000 atmospheres, J.
Am. Chem. Soc, 60, 2300-2803, 1 038.
Wilcox, R. A., Petrology of Paricutin
volcano, Mexico, U.S. Geol. Surv. Bull.,
965-C, 281-353, 1954.
Willemse, J., and J. J. Bensch, Inclusions
of original carbonate rocks in gabbro and
norite of the eastern part of the Bushveld
Complex, Trans. Geol. Soc. South Africa,
67, 1-87, 1964.
Williams, A. F., The Genesis of the Dia-
mond, Vol. 2, E. Benn, Ltd., London, 1932.
Wimmenauer, W., Beitrage zur Petro-
graphie des Kaiserstuhles, Teil IV, Die
Gesteine der phonolithischen Familie;
Teil V, Die subvulkanischen Breccien,
Neues Jahrb. Mineral., Abhandl., 98, 367-
415, 1962.
Winkler, H. G. F., Die Genese der Metamor-
phen Gesteine, Springer- Verlag, Berlin,
1965.
Wittry, D. B., Methods of quantitative
electron probe analysis, Am. Soc. Testing
Mat. Spec. Tech. Publ., No. 349, 128-150,
1963.
Wones, D. R., and H. P. Eugster, Stability
of biotite: experiment, theory and appli-
cation, Am. Mineralogist, 50, 1228-1272,
1965.
Yagi, K., Petrochemical studies of the
alkalic rocks of the Morotu District,
Sakhalin, Geol. Soc. Am. Bull, 64, 769-
810, 1953.
Yoder, H. S., Jr., High-low quartz inversion
up to 10,000 bars, Trans. Am. Geophys.
Union, 31, 827-835, 1950a.
Yoder, H. S., Jr., The jadeite problem, Am.
J. Sci., 248, 225-248, 312-334, 19506.
Yoder, H. S., Jr., The MgO-Al203-Si02-H20
system and the related metamorphic fa-
des, Am. J. Sci., Bowen Vol., 569-627,
1952.
Yoder, H. S., Jr., Role of water in metamor-
phism, Geol. Soc. Am. Spec. Paper, 62,
505-524, 1955.
Yoder, H. S., Jr., and C. E. Tilley, Origin
of basalt magmas : an experimental study
of natural and synthetic rock systems,
J. Petrol, 3, 342-532, 1962.
Young, R. A., Mechanism of the phase
transition in quartz (abstract), Acta
Cryst., 16, A180-181, 1963.
Yund, R. A., and G. Kullerud, Thermal sta-
bility of assemblages in the Cu-Fe-S sys-
tem, J. Petrol, 7, 454-488, 1966.
570
CARNEGIE INSTITUTION
PERSONNEL
Scientific Staff
Director: P. H. Abelson
Emeritus Research Associate: E. G. Zies,
Chemist
Physical Chemists: F. R. Boyd, T. C.
Hoering, J. F. Schairer
Petrologists: C. W. Burnham,1 F.
Chaves, D. H. Lindsley, H. S. Yoder,
Jr.
Geochemists: G. L. Davis, T. E. Krogh,2
G. Kullerud
Organic Gcochemist: P. E. Hare
Geophi/sicist: P. M. Bell
Physicist: J. L. England
Crustallographer: G. Donnay
Fellows: G. M. Brown, Oxford Univer-
sity:3 J. R. Craig, Lehigh University;
G. \V. Fisher, Johns Hopkins Univer-
sity ;4 M. C. Gilbert, University of Cali-
fornia at Los Angeles; N. Giiven, Co-
lumbia University; E. C. Hansen, Yale
University; H. 0. A. Meyer, Univer-
sity College, London;3 R. M. Mitterer,
Florida State University; N. Mori-
moto, Osaka University;5 J. L. Munoz,
Johns Hopkins University;6 A. J. Nal-
drett, Queen's University, Canada; D.
C. Presnall, Pennsylvania State Uni-
versity;7 S. W. Richardson, Oxford
University, England; W. F. Schreyer,
University of Bochum, Germany;8 W.
H. Scott, Yale University;6 K. L. Wil-
liams, Australian National University9
Guest Investigators: J. Calder, Univer-
sity of Texas ; J. D. H. Donnay, Johns
Hopkins University; A. El Goresy,
Max Planck Institut fur Kernphysik,
Germany; R. F. Emslie, Geological
Survey of Canada; P. Geiser, Johns
Hopkins University; A. A. Godovikov,
Novsibirsk State University, U.S.S.R. ;
R. L. Harris, Jr., Texas Technological
College; H. G. Huckenholz, University
of Cologne, Germany; A. Long, Smith-
sonian Institution ; W. C. Luth, Massa-
chusetts Institute of Technology; I. D.
MacGregor, Southwest Center for Ad-
vanced Studies ; Z. Matthes, University
of Wiirzburg, Germany; V. R. Meenak-
shi, Duke University; S. A. Morse,
Franklin and Marshall College; D. C.
Presnall, Southwest Center for Ad-
vanced Studies; A. Saleuddin, Duke
University; L. A. Taylor, Lehigh Uni-
versity; C. E. Tilley, Cambridge Uni-
versity, England; J. O. Waller, Vir-
ginia Polytechnic Institute
Operating and Maintenance Staff
Executive Officer: A. D. Singer
Accountant: C. B. Petry
Editor and Librarian: Dolores M.
Thomas
Stenographers: Patricia S. Garrett, Mar-
jorie E. Imlay
Clerk: H. J. Lutz
Electronic Technician: C. G. Hadidiacos
Research Assistant: J. F. Kocmaneck10
Chief Mechanician: F. A. Rowe
Instrument Makers: C. A. Batten, L. C.
Garver, W. H. Lyons, 0. R. McClunin,
G. E. Speicher11
Mechanic and Carpenter: E. J. Shipley
Electrician: E. C. Huffaker
Machinist: J. R. Thomas
Building Engineer: R. L. Butler
Mechanic's Helper: M. Ferguson
Janitor: A. T. Lewis
1 Resigned August 31, 1966, to accept ap-
pointment as Associate Professor of Min-
eralogy at Harvard University.
2 Appointment from August 1, 1966.
3 Appointment from September 1, 1966.
4 Appointment terminated December 31,
1966, to accept position as Assistant Pro-
fessor at the Department of Geology, The
Johns Hopkins University.
5 Appointment from September 15, 1966,
through January 15, 1967.
6 Appointment from July 1, 1966.
7 Appointment terminated March 31,
1967, to accept position as Assistant Pro-
fessor, Geosciences Division, Southwest
Center for Advanced Studies, Dallas, Texas.
8 Appointment from October 1, 1966,
through May 31, 1967.
9 Appointment from April 1 through
June 30, 1967.
10 On leave of absence without pay from
September 1, 1966, through January 15,
1967.
11 On leave of absence without pay from
September 16 through October 15, 1966.
PLATES
Plate 1 . (A) Orthorhombic bronzite crystal with thin lamellae of Ca-rich pyroxene exsolved
11(100). Stillwater intrusion. Crossed nicols. X 42.
(B) "Inverted pigeonite" crystal, now orthorhombic hypersthene with thick augite lamellae
in herringbone pattern. Texture suggests exsolution of augite | j (00 1 ) of twinned monoclinic
pigeonite, prior to inversion of the latter to rhombic hypersthene. Skaergaard intrusion. Crossed
nicols. X 42. (See Brown, 1957.)
(C) "Inverted pigeonite" crystal showing a second, thin set of lamellae exsolved 1 1 (1 00) in the
hypersthene, presumably after inversion from pigeonite. Bushveld intrusion. Crossed nicols.
X 130. (See Atkins, 1965.)
(D) Complex pattern of exsolution in the hypersthene host of an "inverted pigeonite"
crystal (base). This is additional to the coarse, pre-inversion lamellae and the parallel thin
lamellae (top right). The angle between these two sets of lamellae does not suggest exsolution
of coarse lamellae along the (001) plane of pigeonite [cf. C). Bushveld intrusion. Crossed nicols.
X 130. (See Atkins, 1965.)
Plate 1
Geophysical Laboratory
r\
c
Plate S
Geophysical Laboratory
fev \rr&i&*>
i JF
J^-"S
0.1 mm
,JV\
Plate 2. Photomicrograph of synthetic yoderite produced at 800°C and 16 kb for 17 hours
in a solid-media apparatus. Refractive index of mounting media is 1.550. Starting materials
were synthetic cordierite, synthetic spinel (1:1), and natural andalusite from Minas Gerais,
Brazil, to make a bulk composition of 4:5:7 with excess water. Small traces of talc, chlorite,
Mg-staurolite, and corundum present, but not visible in photograph.
Plate 3
Geophysica I ha b o i a t o t y
Plate 3. Photomicrograph showing textures produced when a pyrrhotite-magnetite liquid
crystallizes. Magnetite (dark gray) occurs in pyrrhotite (light gray). The highly reflecting phase
is platinum from reaction vessel. Polished section; X 375.
Plate
Geophysical Laboratory
0.4 mm
Plate 4. P-T transformations of melilite olivine nephelinite, Uvalde, Texas.
(A) Melilite olivine nephelinite, Uvalde, Texas (analyzed rock), showing augite, melilite
(rectangular grain, left center), olivine (left edge), nepheline, apatite, and iron ore.
(B) Conversion of the Uvalde rock to a pyroxenite-type assemblage at 900°C, 2 kb, for
1 1 2 hours. The photomicrograph shows dominant clinopyroxene with accessory hornblende,
biotite, olivine, and iron ore. A little nepheline remains.
(C) Uvalde rock heated at 1200 C and 10 kb for 1 hour. The dominant colored clino-
pyroxene is associated with accessory biotite, olivine, and iron ore. Lath sections of biotite
are seen intergrown with clinopyroxene on the left.
(D) Conversion of the Uvalde rock to an assemblage built of brown hornblende with biotite.
Clear areas, voids in section. A run at 1 1 25 C, 5 kb, H2O, 1 hour. The biotite is seen as cores
to the quench amphibole (right edge).
Plate 5
G e op hysica I L a b or a t ory
Plate 5. P-T transformations of monticellite olivine nephelinite and nepheline basanite.
(A) Monticellite olivine nephelinite. Shannon Tier, Tasmania, showing olivine (center) and
clinopyroxene (upper edge, center), in a groundmass of irregular shaped monticellite enclosing
nepheline and iron ores.
(B) The same rock treated at 1175°C, 10 kb, for 2 hours. The assemblage is built of
euhedral clinopyroxene associated with iron ore and accessory biotite.
(C) Nepheline basanite, Inverell, New South Wales (analyzed rock), heated at 950°C, 5 kb,
for 4 hours. A relatively coarse-grained hornblende-clinopyroxene assemblage with olivine
(center) mantled by iron ore.
(D) Monticellite alnoite, Sutherland Commonage, Cape Province, South Africa. The section
shows phenocryst olivine mantled by monticellite in a groundmass of monticellite, melilite,
nepheline, iron ore, and perovskite.
Department of Embryology
Baltimore, Maryland
James D. Ebert
Director
Contents
Introduction 575
Ribonucleic Acid and Protein Synthesis during Amphibian Development . . . 580
Ribosoma] RNA and Its Genes during Oogenesis and Development .... 580
Hybridization method 580
Linkage of ribosomal genes 581
Evolutionary conservation of ribosomal genes 584
rDXA content of different tissues 585
Summary 588
Synthesis of Ribosomal Proteins in Xenopus laevis 589
Synthesis of Collagen during Amphibian Embryogenesis 590
Studies on Frog Oocyte Mitochondrial DNA 592
Cell Differentiation and Viral Susceptibility 598
Is Cellular DNA Synthesis Stimulated by Rous Sarcoma Virus? 598
Susceptibility to Rous Sarcoma Virus of Cell Clones Derived from
Sternal * Cartilage 600
Growth of chondrocytes after infection by RSV 600
Virus production by chondrocytes 601
Virally induced morphological changes in chondrocytes 601
Some Uses of Clonal Cell Cultures of Differentiated Cells 602
Studies on the Developing Heart of the Chick Embryo 606
Properties of Embryonic Heart Cells in Culture 606
Cell morphology and spontaneity 609
Spontaneity and growth 611
Summary 611
Electrophysiological Recording from Heart Cells in Culture 612
Growth media and oxygen tension 612
External potassium ion concentration 613
Cell density 614
Morphogenesis of the Early Heart 615
Neural Regulation of the Embryonic Chick Heart 618
Dissociation and Culture of Brain Tissue 618
Biochemistry and Physiology of the Gonads 619
Germinal Vesicle Breakdown in Response to Steroids and Gonadotropins in
Rana pipiens 619
Testicular Sorbitol Dehydrogenase 621
The Mammalian Embryo in Relation to Its Environment 626
Early Cleavage of the Mouse Egg in vitro 626
Implantation in the Rabbit 626
Anatomy and Physiology of the Placenta 632
Radioangiography of placental circulation 632
Gross anatomy of the placenta of rhesus monkeys 634
Uterine activity studies 636
The Collection of Human Embryos 638
The Early Development of the Nasal Pit in Staged Human Embryos . . . 639
Staff Activities 639
Bibliography 641
Personnel 643
Carnegie Institution Year Book 66, 1966-1967
INTRODUCTION
It was just about ten years ago peptides and proteins. Moreover, it is
that Jane Oppenheimer wrote, in her clear that gene action may be regu-
pro vocative essay "Embryological lated on at least three levels: (1)
Concepts in the Twentieth Century" The information content of the DNA
(happily reprinted this year as one may be regulated; that is, the number
of her collected Essays in the History of copies of a given gene may differ
of Biology and Embryology) , "The from cell to cell or from time to time
fact remains, however, that the con- in the life history of a given cell. (2)
ceptual chasm between the gene and The frequency with which a given
the finally differentiated cell con- gene is transcribed into RNA copies
tinues wide and deep. . . . When bio- may vary. Evidence from many
chemical steps intervening between organisms is accumulating that this
the gene and its effect can be fol- is a common way of genetic regula-
lowed . . . , the study of development tion, although we know very little
becomes itself molecular biology, about the mechanisms which bring
although unfortunately it must be about this regulation. And, finally,
remembered that the ultimate and (3) transcribed gene copies may or
really only significant step leading may not be translated into proteins —
directly back to the gene itself that is, controls may operate at the
remains as yet to be followed for level of translation,
metazoan material." Ursprung has reminded us that the
That step has now been taken. The investigator concerned with unravel-
chasm has been bridged, at least at a ing the complex pathways leading
few significant points. To be sure, not from gene to biochemical and mor-
all the underpinnings are securely phological traits often chooses a
in place, but almost every issue of developmental mutant as a starting
Journal of Molecular Biology (and to point for his analysis, since such a
a lesser extent, perhaps, the several mutant indicates that normal gene ac-
embryological journals) bring us a tion — direct or indirect — is required
little closer to a goal that, a scant 10 for the particular developmental
years ago, appeared, to all but a few, event. Next, a precise biochemical
insurmountable. and morphological catalog of events
The molecular events that consti- at stages prior to the first obvious
tute gene action have been resolved manifestation of the mutant trait is
in recent years with a remarkable established. A detailed knowledge of
degree of precision. In the first step, the biochemistry and morphology of
the information contained in the normal development of an organism
DNA molecule is transcribed into an is essential if we are to understand
RNA copy. This transcription gen- the role that genes play in its devel-
erally occurs in the cell nucleus. The opment. Genes have one primary
transcribed copy of the gene then effect, the production of RNAs, this
usually leaves the nucleus, and in the direct 1/1 relationship being followed
cytoplasm of the cell becomes en- by another 1/1 relationship, in that
gaged in the process of translation, the primary RNA products are used
during which the linear information for making proteins,
of nucleotides is translated into a Ten years ago, in studying the
linear sequence of amino acids which patterns of synthesis of specific pro-
become linked together to form poly- teins, using the techniques of enzy-
575
5 i 6
CARNEGIE INSTITUTION
mology and immunochemistry, we
made our first major efforts to center
our attention on products that were,
in Ursprung's words, "close to the
genes.'1 But it is now technically
possible to analyze the functioning of
many individual genes during devel-
opment by assessing- the most direct
products, the RNAs, and through
them to study the genes themselves.
For the past several years, the
program of Donald Brown and his
colleagues has been based on the
argument that the ideal expression of
the activity of a gene is the rate of
synthesis of its RNA. They have
taken the position that in studying
direct gene products, the following
criteria should be applied: (1) When-
ever possible, one should study two
or more linked genes which define
different but functionally related
RNA products. (2) Mutations in one
or more of these genes should be
available for study, thus facilitating
their mapping and ultimately their
quantitation and isolation. (3) The
direct RNA products should be easy
to purify and assay, and stable
enough in vivo for pertinent data to
be obtained regarding their rates of
synthesis. And, finally, (4) the syn-
thesis of the gene products should
be regulated in vivo; that is, there
should be clear evidence that the
activity of these genes changes at
different times during the develop-
ment of the organism. Brown's
application of these criteria to the
ribosomal RNAs (rRNAs) and the
genes which define their structure in
the toad, Xenopus, has led in recent
years to a series of notable dis-
coveries. The current year has been
no exception.
It is clear that oocytes and em-
bryos of Xenopu8 laevis synthesize
rRNA at vastly different rates which
are related to the stage of the e^
or embryo. Thus, the action of riboso-
mal genes must be controlled in a
sensitive way during oogenesis and
embryogenesis.
Brown's continuing studies of
rRNA and its genes, this year prin-
cipally in collaboration with Carl
Weber and John Sinclair, have yielded
the following information. The genes
for 28S and 18S rRNA while
highly redundant (about 800 of each
in X. laevis) are clustered together
on a single autosome. They are much
intermingled within the cluster and
may be strictly alternating. The
molecular hybridization technique
which demonstrates this linkage also
shows that the functionally related
genes for 5S rRNA are not linked to
the 28S and 18S genes, since they
are on separate DNA fragments and
are not deleted in the anucleolate
mutant. This is of particular interest
because 5S RNA accumulates coordi-
nate^ with 28S and 18S RNAs and is
present only as a constituent of the
ribosomal particle. The specific hy-
bridization technique developed for
these experiments has been used to
measure the content of 28S, 18S, and
5S RNA. Somatic tissues with widely
different rates of ribosome synthesis
contain similar numbers of genes for
28S, 18S, and 5S rRNA. However,
oocytes have an enormous increase in
genes for 28S and 18S RNA. Thus,
somatic tissues synthesize ribosomes
at different rates from an invariable
genome, while in the oocyte the actual
number of genes for 28S and 18S
rRNA is increased several hundred-
fold.
What is the fate of these multiple
copies of genes for 28S and 18S
rRNA? Preliminary evidence sug-
gests that during cleavage, when
rapid nucleolar division is taking
place, these extra copies of rDNA are
not replicated along with the chromo-
somal DNA. By gastrulation, hybrid-
ization experiments show that the
percentage of DNA which anneals
with rRNA is no greater than for
DEPARTMENT OF EMBRYOLOGY 577
erythrocyte DNA. Furthermore, The low molecular weight deoxyri-
since rRNA synthesis does not take bose-containing substances of the egg
place during cleavage, the excess are likely to constitute a precursor
copies (which are presumably in the pool for the intense DNA synthesis
cytoplasm) are nonfunctional. The during early development. Until some
extra genes (rDNA) therefore re- evidence to the contrary is presented,
main inert in the cytoplasm during it may be assumed that DNA syn-
cleavage and probably are ultimately thesis in the embryo proceeds in the
degraded. usual way using deoxynucleoside-
This extra rDNA is not to be con- triphosphatases,
fused with the cytoplasmic DNA The major unresolved problem or
associated with mitochondria. Dawid set of problems concerning differen-
has firmly established that the cyto- tiation is that of its control or "regu-
plasm of amphibian eggs contains lation." Whether positive or negative
substantial amounts of DNA, the control mechanisms operate in differ-
bulk of which is associated with entiation or, indeed, whether the
mitochondria. There is ample evi- problem can be stated in these terms
dence that the DNA is an integral borrowed from bacterial genetics, is
component of these particles and that still largely unknown. Frequent ref-
DNA is a general constituent of the erence is made to the interaction of
mitochondria of all tissues of all nucleus and cytoplasm, to negative
organisms so far studied. In a series controls, i.e., to the possible role of
of elegant studies, Dawid and his repressors of gene action.
colleague David Wolstenholme have One of the most striking examples
established that the mitochondrial of the cytoplasmic control of nuclear
DNA of frog oocytes is circular, activity was found in earlier experi-
having the structure of a "twisted ments by Gurdon and Brown. It will
circular" molecule. be recalled that the synthesis of
There is a substantial pool of rRNA that predominates during
low molecular weight deoxyribose- oogenesis in Xenopus stops at matu-
containing substances in eggs, some- ration, not to be resumed until gas-
times erroneously spoken of as "cyto- trulation, and not to be resumed in
plasmic DNA." The bulk of these significant amounts until substan-
substances may be monodeoxynucleo- tially later.
tides. There is no compelling need to Gurdon and Brown transplanted
postulate, as some have, the existence nuclei from late embryos or young
of oligodeoxynucleotides in eggs, and tadpoles into eggs. Following trans-
Dawid has emphasized that a con- plantation, the nucleoli that were
sideration of their function should prominent in the embryonic nuclei
be deferred until their presence is disappeared within 40 minutes, and
demonstrated. rRNA synthesis was not detected
The fate of mitochondrial DNA again until gastrulation. The tech-
during development is not known, but nique for studying the synthesis of
it seems highly probable that it re- rRNA was not sensitive enough to
mains an integral part of the mito- measure a small amount of rRNA
chondria and is distributed along which might have been synthesized
with these particles into the cells of immediately after transplantation.
the embryo. Mitochondrial DNA is However, dRNA was synthesized by
expected to be replicated whenever middle cleavage (about 6000 cells)
the multiplication of mitochondria is just as in a normal cleaving embryo,
initiated during development. However, the point to be emphasized
578
CARNEGIE INSTITUTION
is that, as in normal embryos, the
synthesis of rRNA in transplanted
embryos began at gastrulation.
Therefore, genes that were active in
rRNA synthesis in cells of late em-
bryos were "turned off" when placed
into egg cytoplasm. However, they
were not permanently repressed, for
they resumed their function "on
schedule" at gastrulation.
Brown's more recent studies point
up an even more attractive un-
resolved problem of regulation.
As already observed, in the anu-
cleolate mutant of X. lac vis the major
ribosomal RNAs, 28S and 18S, are
absent. The third ribosomal rRNA,
5S rRNA, also appears to be absent.
However, the absence of 28S and
18 S rRNAs on the one hand, and of
5S rRNA on the other, must be
explained by different (although
somehow related) mechanisms. As
we have seen, in the anucleolate
mutant, the genes for 28S and 18S
rRNA are deleted. However, the
genie material, the rDNA, for 5S
rRNA is not deleted.
It is clear, therefore, that in the
anucleolate mutant, the structural
genes for 28S and 18S rRNA are
lacking. The absence of 5S rRNA,
however, must be explained by a
repression of the activity of the genes
for 5S RNA, which are present. In
thinking of the "construction" of a
ribosome, we must visualize linked
structural genes for 28S and 18S
rRNA and an independently located
rDNA for 5S RNA. Although they
are independent, they must somehow
be coordinated, or regulated. Several
possibilities come to mind, the most
attractive to the writer being that in
the mutant the failure of production
of 28S and 1 8S rRNA somehow feeds
back and "turns off" the genes for
5S rRNA. The search for the nature
of this mechanism ought to be very
interesting.
These two programs, of Brown and
Dawid and their respective col-
leagues, have been singled out for
special mention for several reasons.
It is not just that they are helping
shape their respective fields, although
that reason alone would be more than
enough. These programs illustrate
three other features of life in the
Department and the Institution as a
whole: (1) the continuing interac-
tion with colleagues of kindred in-
terest and spirit in the Department
of Terrestrial Magnetism; (2) the
continuing, in fact increasing, in-
teraction with other laboratories
throughout the world; and, (3), the
important contributions of post-
doctoral fellows and visiting investi-
gators.
The importance of the continuing
exchange of advice and criticism be-
tween the Baltimore and Washington
laboratories hardly needs to be ampli-
fied. However, special mention should
be made of two examples of inter-
actions with other groups. For the
two years ending in the early spring
of 1967, Dawid was a guest in the
Max-Planck-Institut fiir Biologie in
Tubingen, Germany, through the
courtesy of Professor W. Beermann.
We hope that this will not be the last
exchange between these two labora-
tories. The range of opportunities
open to today's fellows and graduate
students is also exemplified by Merry
Schwartz's "three-way" program. As
a graduate student, her "formal"
studies, apart from research, are
carried out in the Department of
Biology at Johns Hopkins. Her re-
search on collagen biosynthesis,
which is conducted in this Depart-
ment, where she is sponsored by
Brown, is also carried out in coopera-
tion with Professor Howard Green
of New York University School of
Medicine, one of the leaders in that
field.
The contributions of postdoctoral
fellows to the Department's pro-
DEPARTMENT OF EMBRYOLOGY
579
grams cannot be emphasized too
strongly. Mention has been made of
John H. Sinclair, formerly of the
University of Chicago, and Carl S.
Weber, of the University of Illinois,
Fellows of the U. S. Public Health
Service, in their first and second
years in the Department, respec-
tively. A third postdoctoral fellow in
Brown's laboratory, Harold E. Kasin-
sky of the University of California,
also supported by the U. S. Public
Health Service, took up residence in
the Department late in the year
covered by this report.
Public Health Service Fellows con-
tributed to two other major pro-
grams. Harold H. Lee, who is
completing his second year of work
in cooperation with Ebert and
Kaighn, will assume a new position
as Assistant Professor of Biology at
the University of Toledo in the fall
of 1967. They have studied the stimu-
lation, by Rous sarcoma virus, of the
incorporation of the DNA building
block, thymidine, into the nuclei of
mature muscle in vitro. Their work
suggests that the nuclei in this differ-
entiated tissue, normally incapable
of DNA synthesis have been some-
how "derepressed" and enabled to
resume the synthesis of cellular
DNA. Their studies open up several
new avenues in the search for under-
standing of the way in which cells
are transformed by tumor viruses.
Helge Stalsberg, who has been on
leave for one year from the Uni-
versity of Oslo, has contributed
importantly to the program on car-
diogenesis, headed by Robert De-
Haan.
In the summer of 1966, Hayden G.
Coon arrived to take up his responsi-
bilities as a Carnegie Institution
Fellow. Already recognized for his
studies on the stable expression of
the differentiated phenotype in clonal
cultures, Coon has inaugurated a new
program related to, and based in part
upon, the line he had previously
pursued. He has begun to apply the
techniques of cell hybridization to
clonal populations of differentiating
and differentiated cells. Thus far, his
work has centered on the technical
problems of producing hybrid strains
from clonal cultures of chondrocytes,
pigment cells, fibroblasts, and myo-
blasts. His results, which are de-
scribed beginning on page 30, are
encouraging. The somatic hybridiza-
tion techniques have been "promis-
ing" for several years, but it is only
recently that they have begun to
provide definite evidence of an inter-
action between genomes in hybrid
cells.
In recent experiments by David-
son, Ephrussi, and Yamamoto, of
Western Reserve University, hybrids
between cells of a pigmented Syrian
hamster melanoma line and each of
three unpigmented mouse lines were
isolated and maintained in active
proliferation in vitro for up to 100
cell generations. These hybrids have
thus far remained unpigmented un-
der conditions in which the mela-
noma cells become heavily pigmented,
and they exhibit no dopa oxidase
activity, suggesting that some step
in the process of pigment formation
in these cells is under negative con-
trol.
Professor James F. Case, on leave
from the University of California at
Santa Barbara, spent about three
months in the Department as a Car-
negie Institution Fellow in the spring
of 1967, getting reacquainted with
the chick embryo, and carrying on
exploratory experiments on the ori-
gin of the neural regulation of the
chick heart.
Two new Fellows of Carnegie In-
stitution, Masako Fukada and Shuhei
Yuyama, arrived on June 1, 1967, to
work in cooperation with Ebert.
The records of the Department,
undoubtedly incomplete, show that at
580
CARNEGIE INSTITUTION
least 54 visiting investigators spent
from one week to the entire year in
cooperative research, in learning
techniques, or in using the collection
of human embryos. Among those
whose contributions are discussed in
these pages are Allen W. Schuetz and
David Whittingham, colleagues of
Professor John Diggers at the Johns
Hopkins University School of Hy-
giene and Public Health; C. B.
Martin, Jr., of the Medical College of
Georgia; Martin Donner of Johns
Hopkins University School of Medi-
cine; and Professor John Bonica and
his associate Marlene Eng, of the
University of Washington School of
Medicine. The last four contributed
importantly to the program of studies
of the placenta conducted under the
leadership of Elizabeth M. Ramsey.
During frequent visits through-
out the year Professor M. Wharton
Young of Howard University College
of Medicine pursued his studies on
the developing human ear. A pre-
liminary report of some of his find-
ings appeared during the year in the
Anatomical Record, vol. 157, p. 341.
Two changes in the departmental
roster should be mentioned. David W.
Bishop, a member of the staff since
1952, retired on June 30, 1967.
M. Edward Kaighn has completed his
term of appointment and has taken
up new responsibilities at the New
York Blood Center. It is hoped, how-
ever, that the Department will be
able to continue to draw upon
Kaighn's insight and skills in study-
ing the interactions of oncogenic
viruses and differentiating cells, a
program in which both laboratories
are keenly interested.
RIBONUCLEIC ACID AND PROTEIN SYNTHESIS
AMPHIBIAN DEVELOPMENT
DURING
ribosomal rna and its genes
During Oogenesis and
Development
D. D. Brown, C. S. Weber, and
J. H. Sinclair
We have taken ribosomal RNA
(rRXA) and its genes as a model sys-
tem to study the control of gene ac-
tion. Earlier studies demonstrated
that oocytes and embryos of Xenopus
laevis synthesize rRNA at vastly dif-
ferent rates which are related to the
stage of the egg or embryo (see Year
Book 6U, pp. 446-452 and Year Book
65, pp. 512-515) . Thus, it is clear that
the action of ribosomal genes must be
controlled in a sensitive way during
oogenesis and embryogenesis. In the
past year we have devised a tech-
nique for directly studying the genes
for ribosomal RNA by molecular
hybridization. The method has high
specificity and a unique versatility
not found in other techniques. The
substance of our report can be con-
veniently divided into four parts:
(1) the hybridization method, (2)
linkage of ribosomal genes, (3) evo-
lutionary conservation of ribosomal
genes, and (4) quantity of ribosomal
genes in different tissues.
Hybridization Method
28S, 18S, 5S and 4S RNAs of X
laevis have been purified from adult
kidney cells which had been exposed
to (H3) -uridine in tissue culture for
several generations. The RNA prep-
arations ranged from 100 to 450
CPM/m/xg. The hybridization tech-
nique combines a preliminary cesium
chloride (CsCl) centrif ugation of the
DNA with the nitrocellulose filter
method of Gillespie and Spiegelman.
The DNA (single- or double-stranded)
is centrifuged to equilibrium in
CsCl and fractionated by drop collec-
tion. The DNA in each fraction is
DEPARTMENT OF EMBRYOLOGY 581
denatured with alkali, neutralized, mal RNA. We conclude that this
and then entrapped on a Millipore small amount of hybridization is not
filter. The individual filters are then due to nucleotide sequences homolo-
baked in an oven at 70°C overnight gous with 28S and 18S rRNA. The
to fix the DNA on the filter, and the anucleolate DNA, therefore, has no
DNA filters are hybridized with the detectable sequences homologous with
appropriate radioactive RNA solu- 28S and 18S RNA.
tion. As many as 50 filters containing
DNA have been hybridized at the
same time in the same vial. The Linkage of Ribosomal Genes
specificity of the technique resides
in several features. The CsCl cen- The genes for 28S and 18S rRNA
trifugation step separates DNA must be clustered, since a single de-
with respect to buoyant density, letion removes all of them. Saturation
which is determined by its guanylic- experiments originally performed by
cytidylic acid (G+C) content. The Wallace and Birnstiel indicate that
genes for 28S and 18S RNA have a these genes comprise about 0.1% of
high G+C content (60%) compared the total DNA of X. laevis (or 0.2%
with the total DNA (40%), and these of the nucleotide pairs). From this
genes separate from the bulk of the value, the molecular weight of indi-
DNA as a satellite with a higher vidual 28S and 18S genes (1.6 and
buoyant density. The CsCl step also 0.6 X 106, respectively), and the
removes impurities in the DNA which amount of DNA in each cell (3 ppg
interfere with standard hybridization for haploid DNA content) , we calcu-
techniques. Finally, a competition ex- late that there are about 800 28S and
periment can be performed with un- 800 18S genes clustered on one chro-
labeled RNA isolated from another mosome of each haploid set. We wish
tissue of X. laevis. We purify cold to know how these redundant genes
28S, 18S, and 5S RNA from isolated are linked with respect to each other
egg ribosomes and cold 4S RNA from in the "rDNA" cluster. There are
the postribosomal supernate of adult two extreme linkage models:
liver. Because the radioactive and un- 1. Contiguous model. There are 800
labeled RNAs are isolated from dif- 28S genes followed by 800 18S genes.
ferent tissues, the unlabeled prepara- 2. Alternating model. 28S and 18S
tion probably will not have the same genes alternate 800 times. In addi-
contaminants as the labeled RNA. tion, any number of intermediate
Therefore, only hybridizable counts models can be postulated,
which are competed out of the cold Since 800 clustered 28S and 18S
RNA reflect true homology of DNA genes must be included on a very long
with the RNA in question. An ex- piece of DNA (at least 3 X 109 dal-
ample of a competition experiment is tons), ordinary DNA isolation tech-
shown in Fig. 1. DNA isolated from niques will break the cluster into
anucleolate (0-nu) X laevis embryos many pieces. According to the con-
is compared with wild-type DNA for tiguous model, almost all fragments
ability to hybridize with 28S and 18S of rDNA would contain either 18S or
rRNA. The competition with unla- 28S genes. In contrast, the alternat-
beled RNA reduces only radioactivity ing model predicts that each frag-
hybridized by control DNA in the ment of rDNA larger than 3.2 X 106
satellite region. The small amount of daltons (the presumed size of double-
hybridization with 0-nu DNA is not stranded DNA coding for one 28S
competed out by excess "cold" riboso- RNA molecule) will contain sequences
582
CARNEGIE INSTITUTION
E
O
CO
o
d
0.8
: 4
Control (2-nu)
Tube no.
Anucleolate (0 - nu)
2000
I 000 ?
J 400 .E
£
Q--r-n--9-fl'
-8-8'
<r*
200
0
Tube no.
Fig. 1 . CsC! fractionation of 0-nu and control DNA followed by hybridization with (H3)-28S
and (H3)-l 8S rRNA. Purified DNA from wild-type X. laevis erythrocytes and O-nu embryos was
centrifuged to equilibrium in CsCI in a fixed angle 65 rotor of the Spinco-L2. The contents of
each tube were fractionated, denatured with alkali, neutralized, and passed through millipore
filters. The DNA filters were baked overnight at 70°; then each filter was split in half. One
set of half filters was hybridized with (H3)-28S and (H3)-l 8S rRNA purified from cells labeled in
culture (dots). The other set was hybridized in the same solution with the addition of a 20-
fold excess of nonradioactive 28S and 1 8S rRNA purified from X. laevis egg ribosomes (circles).
of both 18S and 28S genes. As the
molecular weight of DNA drops
below 3.2 X 10° daltons, an increasing
proportion of the 28S DNA will be
unlinked, i.e., 28S sequences will be
present on DNA fragments which
do not contain 18S DNA sequences.
These predictions and models have
been tested. Birnstiel and Wallace
have shown that when X. laevis is
hybridized with saturating amounts
of rRNA and then centrifuged to
equilibrium in CsCI, the resulting
rRNA/rDNA hybrid has a buoyant
density of 1.77 compared with a
density of 1.72 for the rDNA alone.
The high buoyant density of RNA
(1.90) and the high RNA/DNA ratio
of the hybrid causes this dramatic
change in the equilibrium position of
rDNA. We can make use of this fact
by hybridizing X. laevis DNA with
either 18S or 28S rRNA before cen-
trifugation in CsCI. If both genes
are on the same DNA fragment
(linked) then both 18S and 28S
rDNA will be made denser by hybrid-
ization with either 18S or 28S rRNA.
Any fragment of DNA containing
sequences homologous to only one
of the genes will be altered in
density only by hybridization with its
homologous RNA; prehybridization
with the other RNA should not affect
its buoyant density. One such experi-
ment is shown in Fig. 2, in which
two preparations of DNA with differ-
ent molecular weights (6 X 10G and
0.5 XlO5 daltons) have been prehy-
bridized with 18S RNA, centrifuged,
DEPARTMENT OF EMBRYOLOGY
583
fractionated, and the fraction hy-
bridized again with radioactive 28S
or 18S RNA. The results demonstrate
that fragments of DNA of 6 X 10°
molecular weight (one 28S and 18S
gene should be 3.2 and 1.2 X 10° dal-
tons, respectively) are still com-
pletely linked, since prehybridization
with 18S rRNA has affected the den-
sity of both 28S and 18S DNA to the
same extent. As expected, the low
molecular weight fragments are
almost completely unlinked, since
prehybridization with 18S has no
effect on the buoyant density of 28S
genes. Therefore, 28S and 18S genes
must be highly intermingled in the
rDNA cluster to the extent that there
cannot be more than three 28S genes
in a row before an 18S gene is inter-
spersed. With a preparation of X.
laevis DNA which is about 2-3 / 10°
daltons, it will be possible to deter-
mine if the 28S and 18S genes are
precisely alternating.
We have tested the linkage of 4S
and 5S genes to the 28S and 18S
genes. In Fig. 3 it is clear that 5S
genes are on separate fragments of
DNA from 28S and 18S genes. The
same is true for 4S genes. By the
appropriate competition experiment
it can be shown that there cannot be
more than one 4S or 5S gene for
0.8
0.4
=1.
E
o
CO
C\J
ci
d
Control
r, r, ^ g\ ^ ^ *■>
6.7 x 10 Daltons
Prehybridized
n r o o.
400
200 f
A
2.0
0
Control
0.53x10 Daltons
Prehybridized
J\ f\ Lmmii
-J&&±.
CO
200
00
Tube no.
Fig. 2. Buoyant density of 28S and 1 8S genes after hybridization of DNA with cold 1 8S RNA
and CsCI centrifugation. DNA of 6 and 0.5 X 1 06 daltons was denatured by alkali and hybrid-
ized with or without purified cold 1 8S X. laevis RNA at 70°C for 65 minutes. The preparations
were treated with RNase and centrifuged to equilibrium in CsCI. Fractionation of the gradients
and preparation of the DNA filters were performed as described in the legend of Fig. 1. The
filters were split and half were hybridized with (H3)-28S RNA (dots) and the other half with
(H3)-l 8S RNA (circles).
584
CARNEGIE INSTITUTION
0.8
0.4
6
0.8
0.4
Native DNA
-r RNA
-600
5S RNA
Denatured DNA
Tube no.
400
-200
0
ii
e
"i
in
O
200
0
Fig. 3. Hybridization of DNA with (H3)-28S, 1 8S, and (H3)-5S RNA. Folowing centrifugation
of native and denatured DNA, fractionation and preparation of the DNA filters, the filters were
split; half were hybridized with (H3)-28S and 1 8S RNA (dots) and the other half hybridized
with (H3)-5S RNA (circles).
every ten 28S and 18S genes in the
28S and 18S rDNA cluster.
Evolutionary Conservation of
Ribosomal Genes
The specific hybridization tech-
nique described in the preceding sec-
tion has been used to test the extent
of homology of X. laevis rRNA with
DXA from a variety of biological
sources. Figure 4 demonstrates that
toad rRNA contains some sequences
that are homologous with DNA from
widely different organisms. Further-
more, the rDNA of species like X.
laevis behaves to varying degrees as
a satellite on the heavy side of the
DNA, presumably due to its high
G + C content. To date the DNA
from more than 20 eukaryotes has
been found to hybridize with X.
laevis rRNA, but the DNA from 8
nonnucleated organisms has no de-
tectable nucleotide sequences in com-
DEPARTMENT OF EMBRYOLOGY
585
0.4
0.2
3.
E
O
CD
00
Q
d
0.4
0.2
Aerobacter aerogenes
-100
Barley
0
-4000-
-2000
Starfish
Cow
r
0
Tube no.
jt=jszsL
rfl--o
200
800
400
800
400
£
CO
u
0
Fig. 4. Hybridization of X. /aeWs rRNA with various DNAs. Hybridization was performed in the
manner described in the legend of Fig. 1. Half filters were hybridized with (H3)-28S and 1 8S
RNA (dots) and in the presence of competing nonradioactive and 28S and 1 8S RNA (circles).
mon with toad rRNA. A summary of
these heterologous reactions is pre-
sented in Table 1. These results
merely represent a qualitative demon-
stration that the DNA of those spe-
cies does have some homology with
X. laevis rRNA. The extent of this
homology is under study.
rDNA Content of Different Tissues
Differentiation is believed to result
from "differential gene action" in
cells containing identical genomes.
We have directly tested this idea by
measuring the amount of ribosomal
DNA and 5S DNA in DNA from
different tissues in X. laevis. The
experiment involves hybridizing the
fractionated DNA filters with a mix-
ture of (H3)-rRNA and 5S RNA and
comparing the extent of hybridiza-
tion of these two unlinked markers.
The results (Fig. 5) show that DNA
from somatic tissues that are mark-
edly different in the rates of rRNA
synthesis they support, nevertheless
contain the same relative fraction of
their genome homologous with 28S
and 18S rRNA and 5S RNA. This
finding agrees with a study per-
formed by Ritossa and Spiegelman
in which DNA from different tissues
of the chicken were found by stand-
ard saturation measurements to have
5S6 CARNEGIE INSTITUTION
TABLE 1 . Homology of X. laevis rRNA with Nucleotide Sequences in Various DNAs
Organisms with nucleotide sequences homologous with X. laevis rRNA
Mammalia rat, guinea pig, cow
Aves chicken
Amphibia frogs, salamanders, toads, mud puppy, and congo eel
Pisces salmon, lung fish, shark, toadfish
Echinodermata sea urchin, starfish
Echiuroida Urechis
Insecta Drosophila
Crustacea crab
Arachnoidea horseshoe crab
Mollusca clam
Coelenterata sea anemone
Protozoa Tetrahymena
Spermatophyta barley, wheat, spinach
Eumycophyta Neurospora
Myxomycophyta cellular slime molds
DNAs not homologous with X. laevis rRNA
Cyanophyta Plectonema (blue-green alga)
Bacteria C. perfringens, 8. subiilis, A. aerogenes, E. coli,
M. lysodekticus, P. aeruginosa
Phage I?., T7
Mitochondrial DNA from X. laevis
identical fractions of their genome that each nucleolar chromosome con-
homologous with 28S and 18S RNA. tains extra copies of ribosomal genes.
However, the one cell type which We have proved this supposition by
has an increased number of genes for hybridizing egg DNA and the DNA
28S and 18S rRNA, is the oocyte, from isolated germinal vesicles of
Several groups, including our own, oocytes with ribosomal RNA by the
have predicted that the growing techniques described above. We de-
oocyte has extrachromosomal copies signed the experiments to demon-
of rDNA, the evidence for which strate a specific enrichment of ribo-
comes from several sources. To begin somal genes over the rest of the
with, the studies of Callan, Gall, and chromosomal DNA. As shown in Fig.
others demonstrated the remarkable 6, we have compared the relative
fact that the tetraploid nucleus of an abundance of 28S and 18S rDNA
amphibian oocyte has literally hun- with 5S DNA in an experiment simi-
dreds of nucleoli instead of the 4 lar to the one shown in Fig. 5. The
nucleoli expected from the number experiment compares hybridization
of chromosome sets. Recently Miller of erythrocyte, eggf and 0-nu DNA
established that each nucleolus con- with a mixture of H3-rRNA and 5S
tains its own chromosome and Callan RNA (in a ratio of about 1 to 10).
has shown that these chromosomes Mutant DNA does not hybridize with
are formed from the nucleolar orga- 28S and 18S rRNA but has a normal
nizer sites of the lampbrush chromo- complement of 5S DNA. In contrast,
somes. Coupled with the established egg DNA hybridizes with rRNA but
relationship between the nucleolar the level of 5S hybridization is very
organizer as the site of rDNA and the low. This small 5S RNA hybridiza-
nucleolus as the site of rRNA syn- tion is expected from the small
thesis, these observations indicated amount of nuclear DNA present in
DEPARTMENT OF EMBRYOLOGY
587
0.4
0.2
£
O
CD
C\J
Q
6
0.2
Erythrocyte
Early gastrula
rRNA
Liver
5SRNA
Heterozygote ( l-nu)
2000
1000
£
CO
O
000
0
10
0
Tube no.
0
Fig. 5. Hybridization of DNA from different tissues of X. laevis with a mixture of (H3)-28S/ 1 8S,
and 5S RNA. The same fractionation technique was used as previously. The hybridization mix-
ture contained about equal amounts (in fig) of (H3)-5S/ 28S, and 1 8S RNA.
Erythrocyte DNA
Egg DNA
Mutant DNA
-• — •— « — e-
1200
800
400 -
Tube no.
Fig. 6. Hybridization of control, Onu and egg DNA with (H3)-28S, 1 8S, and 5$ RNA. The
experiment is the same as shown in Fig. 5 except that the relative amount of rRNA to 5$ RNA in
the hybridization mixture was about 1 to 1 0.
588
CARNEGIE INSTITUTION
the egg DNA preparation. The opti-
cal density in the egg DNA is due to
mitochondrial DXA, which can read-
ily be shown to have no detectable
sequence homology with 28S and 18S
rRNA (Fig. 7). Eggs are enriched
for DXA homologous with 28S and
18S rRNA by at least several hun-
dredfold over the amount predicted
from the tetraploid chromosome com-
plement. This enrichment corre-
sponds to the number of extra
nucleoli.
From our experiments, and those
of Miller, Callan, and others, we can
reconstruct the sequence of events
surrounding the formation, function,
and fate of these extra 28S and 18S
genes. Specific replication of the
"nucleolar organizer sites" of the
chromosomes of an oocyte takes place
during a short period of oogenesis at
early lampbrush chromosome stage.
About 100-200 replicas must be made
from each "master" copy. After they
detach from the site of their forma-
tion each one becomes a nidus for a
single nucleolus independent of the
lampbrush chromosomes. Miller has
reported that there is a single chro-
mosome in each nucleolus. Since there
are 400-1000 nucleoli and about 400-
1000 times as much rDNA in an
oocyte nucleus, we assume that each
chromosome contains a replica of the
entire redundant rDNA cluster at the
"nucleolar organizer,, locus. After
supporting massive ribosome syn-
thesis during oogenesis the multiple
nucleoli of a mature oocyte become
dormant. At germinal vesicle break-
down (meiotic reduction) the nucle-
oli disappear and presumably enter
the cytoplasm. During development
the extra rDNA never functions,
since there is no synthesis of rRNA
until gastrulation. The normal two
nucleoli reappear in gastrula cells
whose DNA once again contains the
same ratio of 5S DNA to 28S DNA
and 18S DNA (see Fig. 5), proving
that only chromosomal DNA has been
replicated during early embryogene-
sis, whereas the extra rDNA copies
have been either diluted out or
destroyed.
Summary
Our continuing studies of rRNA
and its genes have yielded the follow-
ing information. The 28S and 18S
genes although highly redundant
(about 800 of each in X. laevis) are
clustered together on a single auto-
some. They are extensively inter-
mingled within the cluster and may
be strictly alternating. The technique
that demonstrates this linkage also
Erythrocyte DNA
Egg DNA
Mitochondrial DNA
» • •
80
ol
400 5
10 O 10 0 10
Fig. 7. Hybridization of erythrocyte, egg, and mitochondrial DNA with (H3)-28S and 1 8S RNA.
DEPARTMENT OF EMBRYOLOGY 589
shows that the functionally related taminated with RNA. This RNA is
genes for 5S RNA are not linked to hydrolyzed by treatment of the pro-
the 28S and 18S genes, since they are tein fraction with 10% TCA at 37°C
on separate DNA fragments and are for 20 hours. The precipitated pro-
not deleted in the anucleolate mutant, teins are further extracted with
This is of particular interest because ether, ethanol-ether (1/1), again
5S RNA accumulates coordinately with ether, and dried. This pro-
with 28S and 18S RNAs and is pres- cedure is repeated; then the dried
ent only as a constituent of the ribo- proteins are dissolved in freshly
somal particle. The specific hybridiza- deionized 9 M urea, and the pJl
tion technique developed for these adjusted to 8.5, after which the pro-
experiments has been used to meas- teins are treated with dithiothreitol
ure the content of 28S, 18S, and 5S (DTT), a disulfide reducing agent,
genes in different tissues. Somatic This treatment completely solubilizes
tissues with very different rates of the proteins, a step that is essential
ribosome synthesis contain similar for the subsequent fractionation pro-
numbers of genes for 28S, 18S, and cedure. After treatment with DTT
5S RNA. However, oocytes have an the pH is lowered to about 5.0 and
enormous increase in genes for 28S the protein solution is dialyzed
and 18S RNA. Thus somatic tissues against an acetate-buffered 8 M urea
synthesize ribosomes at different solution. The ribosomal proteins
rates from an invariable genome, recovered at this stage are 95% to
while in the oocyte the actual num- 98 % of those in the original ribosome
ber of ribosomal genes is increased preparation,
several hundredfold. The ribosomal proteins are then
fractionated on carboxymethyl-cellu-
Synthesis of Ribosomal Proteins lose (CMC) at pH 5>55> Because of
IN Xenopus Laevis their basic nature the great majority
Richard L. Hallberg 0f ribosomal proteins adsorb to
Work on the characterization of CMC. Elution is effected with a nar-
eggs of Xenopus laevis has been con- row, linear salt gradient. To ensure
tinued. The procedures used are reproducibility and greater than 95%
ribosomal proteins from unfertilized recovery of protein, freshly deionized
designed to allow us to detect ribo- 8 M urea is used routinely, and DTT
somal proteins in a mixture of total is included in all solutions,
cell protein so that any ribosomal The final analysis of the ribosomal
proteins not yet associated with proteins, either total or fractionated,
mature ribosomes would also be is done by polyacrylamide gel elec-
analyzed. We plan to examine the trophoresis. When radioactive pro-
synthesis of ribosomal proteins dur- teins are used, individual bands are
ing oogenesis and embryogenesis of cut out, solubilized, and counted in
normal and anucleolate embryos. a liquid scintillation spectrometer.
The methods used for purification, Proteins extracted from whole
fractionation, and characterization of oocytes, unfertilized eggs, and em-
ribosomal proteins are as follows, bryos have been subjected to this
Ribosomes are isolated by standard procedure. Independent of the stage
techniques. The RNA-protein com- of embryos used, 80% to 85% of the
plex is dissociated in 8 M urea-2 M total protein can be solubilized by
LiCl and the RNA is precipitated, this method. These proteins were
The proteins, which remain soluble treated with DTT in 8 M urea and
under these conditions, are still con- chromatographed on CMC using con-
590
CARNEGIE INSTITUTION
ditions identical for ribosomal pro-
tein fractionation. Protein yields
from column fractionation vary
between 75 c'c and 90 cb. The majority
of these proteins are either unad-
sorbed or eluted at very low salt con-
centrations. Only occ to brb of the
total protein applied to the column is
eluted in that portion of the gradient
where the majority of the ribosomal
proteins are eluted. It can be shown
that the presence of excess cellular
protein does not affect the behavior
of the ribosomal proteins during their
isolation or fractionation on CMC.
Therefore the column fractionation
substantially purifies the ribosomal
proteins. To show this, protein sam-
ples from identical positions from
CMC fractionation of total embryonic
protein and purified ribosomal pro-
teins were subjected to electro-
phoresis side by side on a single
polyacrylamide gel. Many of the pro-
teins from the total protein sample
migrated with the purified ribosomal
proteins.
Currently we are analyzing the
CMC elution profiles of proteins
from radioactive embryos of various
stages. We are interested both in the
gross change in count patterns as
development proceeds and, more
important, in the change in radio-
activity associated with ribosomal
proteins. When we have sufficiently
characterized the pattern of synthesis
of ribosomal proteins during the
development of normal embryos, we
will then determine whether anucleo-
late mutant embryos can or can not
synthesize ribosomal proteins.
Synthesis of Collagen During
Amphibian Embryogenesis
Merry C. Schwartz
Studies initiated during the pre-
vious year in collaboration with Pro-
fessor Howard Green of New York
University School of Medicine have
further delineated the pattern of col-
lagen synthesis during early amphib-
ian development. Collagen had been
chosen for study for three reasons:
first, it is a differentiated protein
which is sjmthesized by a specific cell
type, the fibroblastic cell; second, it
is unique in having hydroxyproline
as a constituent amino acid; and
third, hydroxyproline arises by the
selective hydroxylation of proline
residues in the polypeptide precursor
of collagen. Therefore, radioactive
proline serves as a precursor of all
proline and hydroxyproline residues
in proteins. The ratio of counts in
these two amino acids is thus a meas-
ure of the differential rate of collagen
synthesis relative to total protein
synthesis.
As described in Year Book 65 (pp.
517-518) Xenopus laevis embryos at
specified stages of development were
pulsed for 3 hours by injection with
L-proline 3,4-H3. Upon hydrolysis of
the protein fraction and purification
of the proline and hydroxyproline by
ion exchange chromatography, total
radioactivity in both amino acids was
determined.
As is shown in Table 2, the dif-
ferential rate of collagen synthesis,
expressed as the ratio of hydroxy-
proline radioactivity to proline radio-
activity, undergoes more than a
700-fold increase between fertiliza-
tion and the feeding tadpole stage.
Due to trailing of a small percentage
of the proline counts across the
hydroxyproline region of the elution
profile, values obtained for hydroxy-
proline radioactivity in the cleaving
embryo are an upper limit. The first
unequivocal appearance of H3-hy-
droxyproline is detected at gastrula-
tion. Only a slight further increase is
observed during neurulation.
Finally, after hatching there is a
rapid increase in the ratio of radio-
active hydroxyproline to proline in
the total protein synthesized. The
pattern of this curve, as illustrated in
DEPARTMENT OF EMBRYOLOGY
591
TABLE 2. Differential Rate of Collagen Synthesis in Embryos of X. laevis
(expressed as ratio of hydroxyproline radioactivity to proline radioactivity)
Number of
Developmental Stage
Embryos
Counts/minute
Counts/min
ute
Corrected
(Nieuwkoop-Faber)
Analyzed
Proline
Hydroxypro
line
(Hypro/pro) %
2—3 Early cleavage
46
26,71 1
<4
< 0.030
119
125,161
<6
<0.01
6—7 Late cleavage
100
141,000
<5
< 0.007
10-11 Gastrula
35
18,346
7
0.078
15—16 Neurula
76
134,300
69
0.103
20 Tailbud
19
9,752
13
0.266
30-31
28
1 6,745
74
0.884
39-40
20
11,090
184
3.32
46—47 Feeding tadpoles
3
4,210
109
5.18
Fig. 8, resembles the pattern pre- A persistent question raised by
viously observed both for the in- these studies is, what is the nature of
creases in many enzymatic activities the proteins that are synthesized in
and for the accumulation of new substantial amounts during cleavage ?
ribosomes. Are they transient proteins, made
6. Or
5.0 -
4.0 -
o
rr
o.
>
x
3.0 -
o
or
Q.
2.0 -
20
40
60 80
100 12
1 i
TIME (HRS.) AT 22-24°C
i i i i i . i
i 1
10 20 32 39 46
Nieuwkoop-Faber Developmental Stage
Fig. 8. Change in the ratio of radioactive hydroxyproline to proline during development of X.
laevis.
502
CARNEGIE INSTITUTION
specifically to function during; that tural proteins of the later embryo
time; or are some of them destined and adult?
to be definitive enzymatic or struc-
STUDIES ON FROG OOCYTE MITOCHONDRIAL DNA
Igor B. Dawid
In continuation of work reported
in Year Book 65 (pp. 518-522) which
showed that the major DNA com-
ponent of amphibian eggs is mito-
chondrial DNA, a study of some of the
physical properties of this DNA was
carried out during a stay at the Max
Planck-Institut fur Biologie in Tubin-
gen in collaboration with David R.
Wolstenholme.
Mitochondrial DNA from frog
oocytes was found to be circular and
to have the structure of a "twisted
circular" molecule as defined by Vino-
grad and his collaborators in their
study of polyoma virus DNA. The
assignment of this structure is based
on the following evidence. Freshly
prepared samples of mitochondrial
DNA were examined in the electron
microscope and were found to contain
98 % to 99% circular molecules. The
mean contour length of the circles
is 5.4 /j., which corresponds to a molec-
ular weight of 10.6 X 106 daltons.
Two classes of molecules could be
recognized: "twisted circles,,, charac-
terized by a tight coiling of the fila-
ment into one or several axes (Plate
1A, B) ; and "open circles/' show-
ing a much more extended structure
(Plate 1C). Up to 64% of the
material could be obtained in the
twisted form, which is considered to
be the native form of mitochondrial
DNA. In this circle, both strands of
the double helix are covalently closed;
free rotation is not possible in a
double helix, so that the twists of the
circle are locked in (see Fig. 9). The
open circles arise from the twisted
ones by the introduction of a single-
strand break, which creates a swivel
opposite the break and allows the
molecule to "untwist." These two
forms of mitochondrial DNA could
also be distinguished in band cen-
trif ligation experiments, by their
different sedimentation rates (Fig.
9). Treatment with small amounts of
DNase I, an enzyme known to break
only one strand of the DNA at a time,
converted the twisted circles into
open ones.
Two components were also found
in sedimentation in alkaline solutions.
The slow component is considered to
be a mixture of circular and linear
single-stranded DNA, which are the
expected denaturation products of the
open circles. The fast component is
regarded as the denaturation product
of the twisted circles in which both
strands of the original helix are
coiled very tightly, resulting in a
molecule with a very high sedimenta-
tion coefficient. A sedimentation
velocity titration was carried out; it
indicated, in analogy to the results
with polyoma DNA, that the dena-
turation of mitochondrial DNA pro-
ceeded through a partly untwisted
intermediate, component I', whose
structure is also depicted in Fig. 9.
The results on the structure and
size of frog oocyte mitochondrial
DNA outlined above and summarized
in Fig. 9 completely agree with the
results of Borst and his colleagues on
the mitochondrial DNA from liver
tissue of several other vertebrate
species. Results on insect and sea-
urchin mitochondrial DNA also indi-
cate a twisted circular structure for
these molecules, but it is not yet clear
whether these circles are the same
size or slightly smaller than those
from vertebrate mitochondria.
DEPARTMENT OF EMBRYOLOGY
593
Fig. 9. Schematic presentation of the different forms of mitochondrial DNA. The diagram is
adapted from the paper by Vinograd and his collaborators on polyoma DNA. Component I
(twisted circular form) is converted to component II (open circular form) by breaking one strand,
and to component III (linear form) by breaking both strands at the same position. Denaturation
of component I proceeds through form V, which resembles component II, and leads to the
tightly coiled component lalk, which is formed from both strands of the original helix. Denatur-
ation of component II leads to a mixture of open and circular single-stranded molecules, which
were not separated during band centrifugation (component l!aik). All S values quoted in the dia-
gram are S2o,w values of the Na-salts except in the case of component lalk/ where Sobserved in
CsCI solutions of the density 1 .35 is given.
In agreement with the reports
from several other laboratories, we
observed that mitochondrial DNA
renatures rapidly and virtually com-
pletely, as tested by density in CsCI
and hyperchromicity in a second
melting cycle. The product of the
renaturation reaction was studied by
CsCI density equilibrium centrifuga-
tion and by electron microscopy.
Figure 10 gives a schematic repre-
sentation of a reaction mechanism
that is consistent with most observa-
tions. The starting material in our
experiments was a preparation of
open circles of mitochondrial DNA,
i.e., molecules consisting of one cir-
cular and one linear strand. During
denaturation and reannealing, addi-
tional breaks are expected to occur at
random places. The result of this
process is a collection of strands with
circularly permuted sequences. In the
reaction of two homologous strands
594
CARNEGIE INSTITUTION
Fig. 10. Model for the proposed reaction mechanism of renaturation of mitochondrial DNA. Lines
represent single strands of DNA; letters symbolize sequences. For description see text.
with different starting points an
intermediate molecule arises which
consists of both double-stranded and
single-stranded regions. Such mole-
cules were present in partially
renatured mitochondrial DNA as
indicated by CsCl density centrifuga-
tion. During further annealing
several reactions can take place.
Intermediate molecules can fold back
on themselves forming circles, as indi-
cated in Fig. 10 d and e. Circles with
no apparent discontinuity of the fila-
ment and a contour length in the
range of 5.4 /x ("perfect circles")
were found in samples of rena-
tured mitochondrial DNA. If a circle
is formed during reannealing by
strands that are incomplete because
of breakage, the circles that arise will
contain a single-stranded region.
Such circles, symbolized in Fig. 10, a,
b, and c, were also found in partially
renatured samples. The single-
stranded regions of intermediate
molecules can also react with homol-
DEPARTMENT OF EMBRYOLOGY
595
ogous regions in other molecules,
leading to larger complexes. One pos-
sible reaction leading to such a com-
plex is shown in Fig. 10, /. Plate
1(D) shows this type of molecule,
which was frequently observed in
partially renatured mitochondrial
DNA. The distinct linear and looped
filaments seen in this complex are
considered to be double-stranded
DNA, whereas the indstinct, "puddle-
like" material is regarded as single
stranded. During prolonged rean-
nealing the complementary single-
stranded regions of such intermedi-
ates react with each other, eventually
leading to an entirely double-helical
product. Very large masses of
apparently interconnecting double-
stranded filaments were the predomi-
nant component in extensively rean-
nealed mitochondrial DNA. The high
molecular weight of these molecules
was also indicated by the rapid for-
mation and low width of the bands
arising during CsCl equilibrium cen-
trifugation.
The formation of large networks
by DNA during reannealing has been
termed concatenation. Britten and
Waring (Year Book 6U, pp. 318-322)
and Britten and Kohne (Year Book
65, pp. 81-87) have described this
phenomenon in their work with
nuclear DNA. These authors further
observed that two jrelated types of
DNA, when reannealed in mixture,
concatenated to form common com-
plexes, whereas unrelated DNAs
formed separate complexes under
these conditions. The presence of
separate or common complexes can
be observed in CsCl gradients if the
two DNA samples studied have dif-
ferent densities. A sequence homol-
ogy test can be based on these obser-
vations. The test is only qualitative
but offers the advantages of being
sensitive, using only small amounts
of unlabeled DNA, and not being
invalidated by cross contamination
of the starting materials.
The nuclear and mitochondrial
DNAs of X. laevis were subjected to
this test. Figure 11, a-d shows the
bands formed by the reactants in
CsCl gradients; the band to the right
is the reference DNA. Figure 11, e
shows the result of joint reannealing
of nuclear and mitochondrial DNA:
Two bands were formed, one at the
density of renatured mitochondrial
DNA (Fig. 11, 6), the other at the
density of reannealed nuclear DNA
(Fig. 11, d). This result indicates
that no general homologies exist
between mitochondrial and nuclear
DNA. The presence of one or several
copies ("master copies,,) of mito-
chondrial sequences in nuclear DNA
would not be detected by this experi-
ment and remains a possibility.
Sequence homologies were found in
the mitochondrial DNAs of X. laevis
and the chicken (Fig. 12). The top
two frames show the bands formed
by chicken DNA in the native, the
denatured, and the renatured states;
Fig. 12, c shows the band formed by
the product of joint renaturation.
The appearance of a single band of
intermediate density in this last
experiment shows that common com-
plexes were formed by the two
DNAs, thereby indicating the pres-
ence of sequence homologies between
them. In an analogous experiment
with the mitochondrial DNAs of
Xenopus and yeast there was no indi-
cation of the presence of homologies.
These results point to a correlation
with evolutionary distance and can
therefore be interpreted as suggest-
ing the preservation of mitochondrial
DNA during evolution. This conclu-
sion supports the hypothesis, based
on several lines of evidence, that
mitochondrial DNA has an important
role in the replication and function of
the mitochondria.
596
CARNEGIE INSTITUTION
Fig. 1 1 . Absence of detectable sequence
homologies between nuclear and M— DNA of
X. laevis. In this and the next figure the band
to the right is denatured DNA of Pseudomonas
aeruginosa, added as a reference. Bands in
CsCI are shown of: (a) native and denatured
mitochondrial DNA, (b) renatured mitochondrial
DNA, (c) native and denatured nuclear DNA,
(d) re-annealed nuclear DNA, (e) mitochondrial
DNA and nuclear DNA reannealed in mixture
(note the formation of two sharp bands at the
positions of renatured mitochondrial DNA,
1.703, and reannealed nuclear DNA, 1.717).
1.703 1.717 1.739
DEPARTMENT OF EMBRYOLOGY
597
1.709
1.702
1.711
1.739
Fig. 12. Hybridization of mitochondrial DNA
of X. laevis and the chick. Bands in CsCI are
shown of: (a) native and denatured chick mito-
chondrial DNA, (b) renatured chick mitochon-
drial DNA, (c) separately renatured mitochon-
drial DNA of Xenopus and the chick, (d) mito-
chondrial DNA of Xenopus and the chick re-
natured together. For the shape and position
of native, denatured and individually renatured
Xenopus mitochondrial DNA see Fig. 1 1 .
598 CARNEGIE INSTITUTION
CELL DIFFERENTIATION AND VIRAL SUSCEPTIBILITY
M. E. KaigJin, H. H. Lee, and J. D. Ebert
{assisted by D. Somervxlle and B. Smith)
Is CELLULAR DNA Synthesis mononucleated myoblasts and multi-
Stimulated by nucleated myotubes were infected
Rous Sarcoma Virus? with a high dose of semipurified RSV
(Bryan's high-titer strain). The
Our previous studies have demon- virus was allowed to absorb for 4
strated that muscle clones derived hours. The cultures were gently
from embryonic chick skeletal muscle shaken every 15 minutes during the
are susceptible to Rous sarcoma virus first 2 hours. At intervals of 7, 24,
(RSV) infection and are able to sup- 48, and 72 hours postinfection, cul-
port the synthesis of this RNA virus tures were pulse labeled with H3-
{Year Book 65, pp. 524-526). We thymidine (1 jxc/ ml of labeled thymi-
have also reported that, beginning at dine having a specific activity of 1.9
16 hours after infection, virus- c/mmole). A thymidineless F12
specific antigen can be detected on the medium was employed. Following this
membranes of multinucleated myo- pulse, cultures were allowed to grow
tubes as well as mononucleated myo- another hour in regular growth
blasts. In order to assess the ability of medium. The cultures were then
myotubes to produce infectious virus, fixed with Carnoy fixative, washed
attempts were made to isolate myo- thoroughly with water, and air dried.
tubes free of myoblasts. Thus far, Liquid emulsion NTB-2 was used for
these attempts have been unsuccess- autoradiography. Dishes with emul-
ful. sion were exposed for 4-7 days at
Some oncogenic DNA viruses have room temperature and then developed
been reported to stimulate the syn- with D-ll, fixed, and washed. After
thesis of cellular DNA in fibroblastic being stained lightly with hematoxy-
cells. In these studies "contact- lin, the petri dishes were covered
inhibited" populations of cells at high with 45-mm circular coverglasses by
density were usually employed. In means of Gurr's mounting medium.
such populations the rate of DNA Observations were made with phase
synthesis is reduced to a low level but optics.
residual synthesis still occurs. The Because of the inconsistency from
transformation of embryo chick fibro- embryo to embryo of viral infection
blasts by RSV is thought to require and transformation in cell culture,
cellular DNA synthesis, but crucial each experiment was carried out with
evidence is lacking. We thought it muscle cells derived from an individ-
would be of interest, therefore, to see ual embryo. Thus the results of indi-
whether RSV infection could induce vidual experiments vary. However,
DNA synthesis in the nuclei of the differences between the control
mature multinucleated myotubes in groups and the infected groups are
which DNA synthesis had ceased of the same order of magnitude in
entirely upon the attainment of mul- all experiments,
tinuclearity. At 7 hours postinfection, there is
Techniques of muscle cloning have no difference between the control and
been reported previously ( Year Book infected cultures. Labeled nuclei can
6U, pp. 483-489 and Year Boole 65, be found in myotubes at 24 hours
pp. 524-526). Seven to 10 days after postinfection (Plate 2C) but at very
seeding, the cultures containing both low frequency. The labeled myotubes
DEPARTMENT OF EMBRYOLOGY
599
have no more than one or two nuclei
with grains. At 48 hours after infec-
tion when transformed myoblasts
begin to appear, more myotubes
incorporate H3-thymidine and more
nuclei per myotube are labeled (Plate
2D). However, not more than 10%
of the myotubes contain radioactivity.
Results of pulsing at 72 hours after
infection do not differ markedly from
those obtained after 48 hours. Trans-
formation at this time, 72 hours post-
infection, is very pronounced. The
fraction of mononucleated cells which
incorporates H3-thymidine after in-
fection is also greater than in sister
control cultures. Figure 13 shows the
results of one representative experi-
ment in which nearly 50% of the
mononucleated cells are labeled in the
infected culture as compared to 20%
in the controls. The distribution of
grains in infected and control cul-
tures is summarized in Table 3.
Incorporation is stimulated in myo-
blasts in infected cultures. Incorpora-
tion clearly occurs in infected myo-
tubes, while in control cultures the
only labeled nuclei associated with
myotubes are located terminally or
peripherally — a very infrequent phe-
nomenon. Moreover, in infected cul-
tures, the number of grains per
nucleus in myotubes is less than one
third of that in morphologically nor-
mal mononucleated cells. In no case
were grains observed elsewhere than
over nuclei.
Accepting the assumption that
incorporation of H3-thymidine in an
50 -
40 J
^ 30 -
o
o
Q.
20-
10 -
Effect of RSV on the Uptake of H3-Thymidine
by Mononucleated Myoblasts
Infected
Control
i
7
24
48
72
Hours after Infection
Fig. 13. Effect of RSV on the incorporation of H3-thymidine by myoblasts. Each bar represents
cells counted in a 2- or 3-unit area in a clone. The total number of cells in each bar is about 500.
600
CARNEGIE INSTITUTION
TABLE 3. Incorporation of H3-thymidine into
the Nuclei of Mononucleated and Multinucleated
Components in Muscle Clones
(expressed as grains counts per nucleus
in autoradiograms)
Grain Counts
Cultures
per
Nucleus
Infected cultures
Multinucleated myotubes
21
Mononucleated myoblasts
60
Transformed cells
17
Control cultures
Multinucleated myotubes
none
Mononucleated myoblasts
60
acid-insoluble form is an indication
of DNA synthesis, the experiments
outlined indicate that RSV infection
does stimulate cellular DNA syn-
thesis. DNA synthesis is thus reac-
tivated in infected myotubes in
which, in the normal process of
differentiation, incorporation of H3-
thymidine and DNA synthesis has
stopped. This conclusion is based on
the following: (1) The time re-
quired for labeled mononucleated
myoblasts to form myotubes is longer
than the 2-hour pulse-chase period
employed in the present experiments.
(2) Myotubes containing 20 or 30
nuclei take days rather than hours to
form. (3) Assuming newly fused
myoblasts are still able to make
DXA, they would not be located in
the center of a myotube as shown
(Plate 2D,E). (4) Labeled nuclei in
mature myotubes with 20 or more
nuclei were never detected in control
cultures.
If the number of grains per nu-
cleus is an indication of the quantity
of DXA made, our results suggest
that only part of the genome is stimu-
lated by RSV. But before taking up
the question of amplification of spe-
cific regions of the genome, it will be
necessary to prove that the incor-
poration observed does represent
DNA synthesis rather than repair,
and that the DNA made is, in fact,
cellular DNA.
Susceptibility to Rous Sarcoma
Virus of Cell Clones Derived
from Sternal Cartilage
The results of our preliminary
studies on the susceptibility of em-
bryonic chondrocytes to RSV carried
out in collaboration with T. S. Okada
of Kyoto University were reported in
Year Book 6U (p. 486) . Evidence was
obtained which showed that clones
derived from embryonic femoral
cartilage could be productively in-
fected by RSV. However, as we
pointed out at the time, we could not
be certain of the identity of the cell
type involved because most of the
colonies that grew in muscle-condi-
tioned medium were of fibroblastic
morphology.
During the past year we have over-
come these technical shortcomings by
employing the methods developed by
H. G. Coon which make it possible to
obtain homogeneous populations of
clonally derived chondrocytes. Coon
and Cahn have shown that the carti-
lage phenotype, i.e., synthesis of an
extracellular matrix containing chon-
droitin sulfate, is suppressed by
concentrations of embryo extract of
the same order as that present in our
conditioned medium. From this, it is
clear that many of the fibroblastic
colonies that had grown earlier in our
conditioned medium were actually
covert chondrocytes.
Groivth of Chondrocytes after
Infection by RSV
Clonal cultures were prepared
from cell suspensions of sternal
cartilage from chick embryos of 12
to 13 days of incubation. For the
experiments reported here a single
cartilage-making colony (CMC) was
picked and allowed to grow in a large
DEPARTMENT OF EMBRYOLOGY
601
homogeneous population. No fibro-
blastic colonies were observed at this
or at subsequent passages. The effect
of infection on growth of such cells
in mass culture is shown in Fig. 14.
In this experiment cells were infected
at 3 different multiplicities in sus-
pension, then plated. The variation
shown in this growth curve during
the first 3 days is not due to virus
infection; rather, it is due to loss of
some cells which tend to come loose
during the early phases of growth
with some batches of fetal calf serum.
In other experiments the infected cul-
tures appear to grow faster than
controls. In 5 days all plates had
become confluent and contained about
4 x 106 cells per plate. Thus, under
these conditions, RSV exerted no sig-
nificant effect on chondrocyte growth.
Virus Production by Chondrocytes
Since we had homogeneous popula-
tions of chondrocytes available, it
was a simple matter to determine
directly their ability to produce infec-
tious virus. A population of chondro-
cytes derived from a single CMC was
passed and plated at 2 X 105 cells per
plate. Virus was added at the time
of plating and left overnight to give
maximum opportunity for infection.
Beginning on day 2, samples were
taken for assay followed by a com-
plete medium change. It is apparent
that the rate of virus production does
not differ greatly with varying virus
input (Fig. 15). However, this result
suggests that with increasing virus
concentration a greater proportion of
cells becomes productively infected.
Further, in all cases the final rate of
virus production approaches a com-
mon maximum, suggesting that ulti-
mately all cells are infected and yield
virus.
Virally Induced Morphological
Changes in Chondrocytes
Within 48 hours cells with a char-
acteristic spindle shape begin to
appear in cultures exposed to the
highest concentrations of virus
(Plate 2B). These cells increase in
number with time. At lower virus
concentrations, their appearance is
delayed but is correlated with virus
LiJ
<
_J
\
C/)
UJ
o
0
IO!
0
2 3 4 5 6
DAYS AFTER INFECTION
Fig. 14. Growth of chondrocytes after infection by Rous sarcoma virus. Chondrocytes were in-
fected in suspension at multiplicities ranging from 0.08 to 80 (focus-forming units) per cell. The
cells were then diluted and plated at 2 X 1 05 per plate in 4 ml of medium. At indicated times
replicate plates were treated with collagenase at 37°C for 30 minutes, and the cells were counted
in a hemocytometer.
602
CARNEGIE INSTITUTION
UJ
»-
VIRUS INPUT
<
_l
Q.
7
-
80ffu/cel
H-
2
-
8.0
3
6
.
O
z
5
cr
-
0.8
o
5
-
u.
Cn
=>
O
-
0.08
o
Li.
4
o
O
o
_l
■*
i
0 1 2 3 4 5
DAYS AFTER INFECTION
Fig. 15. Production of Rous sarcoma virus by
chondrocytes. One cartilage-making clone was
passed and the resulting cells used after 1 1
days. Plates (50-mm Falcon plastic) were seeded
with 2 X 105 cells in 4-ml medium (10%
fetal calf serum, 0.5% bovine serum albumin,
Ham's Ft2 with 2 X amino acids). Virus was
added as indicated at the time of plating.
One-ml samples were taken daily for virus
assay and the medium was then changed. Dur-
ing the course of this experiment the total num-
ber of cells per plate increased about tenfold.
production. Although cells with
spindle shape predominate in infected
culture, many distorted, granulated,
bizarre-shaped cells as well as mor-
phologically normal elements are
present (Plate 2B). A crucial diag-
nostic property of functional chon-
drocytes is their ability to secrete an
extracellular matrix demonstrable by
its metachromatic staining with to-
luidine blue. This material is never
seen in, or associated with, spindle
cells. Injection of as few as 200 of
these cells into the wing web of sus-
ceptible chickens produced tumors
within 13 days, whereas freezing and
thawing of the cells delayed tumor
induction. However, since the tumors
are histologically indistinguishable
from virally induced tumors, it is not
clear whether these preliminary
studies indicate in vivo growth of
transformed cartilage cells or only
their ability to release infectious
virus.
SOME USES OF CLONAL CELL CULTURE
OF DIFFERENTIATED CELLS
H. G. Coon
(with the technical assistance of Isabelle Williams)
Clonal culture of differentiated
cells has become almost commonplace
since the demonstration by Konigs-
berg (Year Book 6U, pp. 483-489)
that it could be done. The primary
questions of the stability and propa-
gability of the differentiated pheno-
type in cell culture have been
answered. Skeletal and cardiac mus-
cle, pigmented retina, and cartilage
have each been studied, and basically
similar culture conditions found to
support exceedingly stable expression
of the differentiated phenotype in
clonal cultures. These findings have
given new hope to efforts to use cell
culture techniques in a fashion anal-
ogous to those that have proved so
fruitful in microbial genetics. The
differentiated state of cells from
chick embryos is sufficiently stable to
permit its study in culture.
Our previous work has shown that
substances that could inhibit expres-
sion of a differentiated cellular phen-
otype in culture (but that did not
inhibit cell growth) were present in
high molecular weight fractions of
chick embryo extract obtained by gel
filtration. It has been shown that
this inhibitor acts reversibly to inter-
fere with the expression of the differ-
entiated function of cartilage cell
clones (similar results have been
obtained for pigmented retina cells
by R. D. and M. B. Cahn) . We have
shown that a direct relationship
between dose and percentage of in-
DEPARTMENT OF EMBRYOLOGY
603
hibited (nonfunctioning) cartilage
colonies exists in the presence of
increasing concentration of the inhib-
itor fraction (Fig. 16) . It has not yet
been possible to obtain quantitative
data from the pigment cell system.
However, recent experiments, under-
taken in collaboration with M. E.
Kaighn, have allowed us to show that
the same or a similar inhibition of
myoblast fusion (and therefore the
expression of muscle cell differentia-
tion in culture) is also proportional
to the concentration of the high mo-
lecular weight fraction of embryo
extract. The data show (Fig. 17),
that unlike cartilage and pigmented
retina, which can form differentiated
colonies in the absence of an embryo
extract, colonies derived from skele-
tal muscle require some component of
embryo extract for growth. At a con-
centration of 2%, this embryo extract
fraction seems to strike an optimal
balance between supplying the re-
quired nutrient and minimizing the
inhibition of strap formation. Knowl-
edge of the existence of these sub-
stances in sera and embryo extract
which inhibit the expression of differ-
entiation in culture may help to ex-
plain the previously widely accepted
conclusion of phenotypic instability
in cell cultures.
Work begun nearly 2 years ago in
collaboration with Giovanni Marzullo
of the University of Pennsylvania has
been continued and is now nearing
completion. In the synthetic pathway
for chondroitin sulfate the stage at
which the reversible inhibition in-
duced by embryo extract has its effect
has been located. Marzullo has devel-
oped procedures permitting the sepa-
00 r
o
o
1.0
0.5
H0.f H7.0
dose response
o = % CMC
* = % PE
15 day sternal
cartilage cells
3 4
% H
Fig. 16. Dose response curve relating to the concentration of high molecular weight fraction of
embryo extract (H) to the percentage of cartilage forming colonies (% CMC). The plating effi-
ciency (% PE) is plotted on the same scale to show that the loss of the expression of chondrocyte
colony function in higher concentrations of the inhibitor is not simply the result of selective killing.
604
CARNEGIE INSTITUTION
IOO-1
• A
• MUSCLE GI50
A MUSCLE GIOO
x
<
0)
O
O
UJ
-I
o
en
10 -
% H
Fig. 17. Dose-response curve relating the con-
centration of high molecular weight fraction (H)
to the percentage of colonies which exhibited
fusion to form muscle straps. The fraction of
colonies which showed strap formation is
plotted in turn as a percentage of the maxi-
mum response — 61% for the G150 horse
serum experiment, and 51% for the G100
horse serum experiment. Both curves showed
maxima at 2% H; both are repetitions of the
same experiment in media containing different
horse serum fractions. Plating efficiency (circles)
is plotted on the same scale to show that selec-
tive cell killing alone cannot account for the
lowered frequency of differentiation.
ration by thin-layer chromatography
of the phosphorylated and uridine
diphosphate (UDP) derivatives of
hexosamines and acetylhexosamines,
which are intermediates in the bio-
synthesis of chondroitin sulfate by
cartilage cells. After incubating cells
in the presence of glucosamine-C14,
the products of its metabolism may
be separated, visualized, and quanti-
tated. We have been able to show
that UDP-N-acetylglucosamine and
UDP-N-acetylgalactosamine accumu-
late in cells that are actively syn-
thesizing chondroitin sulfate (as
determined by Na2S3504 incorporation
into chromatographically isolated
chondroitin sulfate) , but that in cells
inhibited by the embryo extract frac-
tion both of these compounds are
missing, as are the products farther
along in the pathway. Intermediates
between glucosamine and UDP-N-
acetylglucosamine are regularly dem-
onstrated in homogenates from
functioning and nonfunctioning chon-
drocyte clones. Surprisingly, inhibited
cells show activity of both the UTP-
condensing enzyme and the subse-
quently acting epimerase, comparable
to that found in normally functioning
cartilage cells. Work is continuing in
Marzullo's laboratory on this inter-
esting separation of a differentiated
cell's heritable production of the
enzymatic machinery from the ac-
cumulation of the specialized product.
In our laboratory we are trying to
probe the mechanism or mechanisms
of cell heredity indicated by the
phenotypic stability to iterative clo-
nal culture and to extend our culture
work to those embryonic stages in
which cells are first forming commit-
ments to their future differentiated
state. We are attempting to apply
the techniques of cell hybridization of
Barski and Ephrussi and the tech-
nique of virus-mediated cell fusion of
Okada and Harris to homogeneous
populations of differentiated and
differentiating cells.
In studies of the interaction of
different nuclei in the same cytoplasm
(Year Book 65, pp. 21-22) Harris
has applied Okada's technique of cell
fusion induced by Sendai virus. His
work led to a dramatic series of
discoveries. He found that the nuclei
of fused cells sometimes respond to
their new environment in a hetero-
karyon by initiating the synthesis of
nucleic acids and the synthesis of
protein after these functions have
become quiescent during the course
of cellular differentiation. This posi-
tive result indicates that the inter-
action of at least some "control
functions" within differentiated cell
DEPARTMENT OF EMBRYOLOGY 605
combinations might be studied by technical problems of producing hy-
generating heterokaryons and "hy- brid strains (propagating mononu-
brid strains" from the artificial cleate cells result from artificially
fusion of differently differentiated fused cells which contain marker
cells or their embryonic progenitors, chromosomes from both parent cell
Initial experiments of this kind have types) from cloned cultures of differ-
already been performed by Davidson, entiated cells. We have used chondro-
Ephrussi, and Yamamoto with the cytes, myoblasts, pigment cells, and
result that a pigmented cell line fibroblasts from the turkey and the
(melanoma) hybridized with a non- goose as parent cells in these initial
pigmented, so-called fibroblast cell experiments because the karyotypes
line, did not show pigment produc- as well as the interphase nuclei may
tion, nor could characteristic mel- be distinguished in these two species.
anin-forming enzyme activity be (Marker chromosomes are not yet
demonstrated. These results have available in chickens). Our experi-
been interpreted to mean that differ- ence confirms that of other workers
entiated functions of cells may be in this field: It is exceedingly difficult
expressed — or selected from the en- to derive propagating hybrid strains
tire genetic repertoire — by a kind of from cultures of virus-induced het-
negative control, i.e., cells in which erokaryons. One approach to over-
a function is not expressed (fibro- coming this potential difficulty in the
blasts) may possess mechanisms for technique of virus-mediated cell fu-
ensuring that such irrelevant func- sion is to attempt fusions between
tions (e.g., pigment formation) are synchronized populations of cells,
suppressed — somehow turned off. Our approach has been to fuse cells
Presumably the fibroblast's portion of arrested in the metaphase of mitosis
the hybrid celPs genome was able to by the drug colcemid. Initial re-
suppress pigment-forming activity on suits are very promising. We have
the part of the pigmented cell's por- confirmed the observations of Stub-
tion of the hybrid genome. Marzullo's blefield and Klevecz that colcemid-
procedures should make possible induced metaphase arrest could be
analysis of metabolic interactions reversed without change in the
between differentiated cartilage cells karyotype or division rate of the
and other kinds of differentiated or surviving cells, provided that the
not yet functional embryonic cells. concentration and length of exposure
We have done preliminary control to the drug were carefully controlled,
experiments with cells from a natu- We have found, furthermore, that
rally occurring heterokaryon, skeletal cells from chicken embryos were able
muscle, and have shown that cells to resume cartilage or pigment for-
treated with the UV-inactivated Sen- mation following relief of 8 to 12
dai virus may fuse and form nor- hours of colcemid-induced metaphase
mally contractile muscle straps arrest. More important, karyotype
during subsequent culture. Myoblasts analyses of mass cultures grown
fused artificially by treatment with from cells which had been fused at
the killed virus are able to participate metaphase show that the incidence
in normal myotube formation, al- of "hybrid" division figures may be
though it is not yet known whether as high as 10% of the total number,
initially unincorporated myoblasts Since it may prove exceedingly diffi-
also contributed to these virus- cult to use selective techniques based
induced myotubes. upon metabolic inhibitors, a large
Our work has centered on the increase in the frequency of hybrid
606
CARNEGIE INSTITUTION
cell formation may be crucial to the
exploitation of this technique with
natural populations of diploid cells.
While we do not know the mecha-
nism by which the incidence of hy-
brids is increased following fusion of
metaphase arrested cells and the sub-
sequent release of the block, it is
tempting to suppose that the chromo-
somal masses are able to mix and
assort on a common spindle structure
when the colcemid block is released.
Perhaps, by varying the length of
time cells are permitted to remain
blocked after fusion, it will be possi-
ble to control the degree of mixing of
the parental chromosome comple-
ments occurring in the absence of the
spindle structure in colcemid arrested
cells. Special interest might attach to
strains resulting from partial mixing
of the chromosomal complements of
the parent cells.
STUDIES ON THE DEVELOPING HEART OF THE
CHICK EMBRYO
R. L. DeHaan, H. Stalsberg, E. E. Legum, I. Polinger,
S. Gottlieb, J. Graham, and C. Nuttall
Properties of Embryonic
Heart Cells in Culture
Some of the properties of 7-day
embryonic heart cells in tissue culture
were described in Year Book 65
(pp. 526-536), and subsequently in
two other publications. Dissociated
cells inoculated into a nongrowth
medium attach to the surface of the
culture dish; after 24 hours of culti-
vation, two morphological configura-
tions, termed M cells and F cells, are
observed in the population. The
fraction of these cells exhibiting
spontaneous activity is found to be
dependent upon treatment during dis-
sociation and culture, as well as upon
the specific components of the culture
medium. However, the vast majority
of spontaneously active cells falls in
the M-cell category, whereas very
few F-cells (less than 1%) are pace-
makers.
The extracellular concentration of
potassium ions has been shown to
influence pacemaker capacity mark-
edly. Evidence also indicates that
this ion can have more permanent
effects on heart cells. Cultures main-
tained overnight in low-potassium
medium (629A1, 1 mEq/1) are
found to exhibit a higher percentage
of latent pacemakers (i.e., those
stimulable by low potassium) than
similar cultures kept in media con-
taining potassium at normal serum
levels or higher (4-12 mEq/1). This
difference in number of latent pace-
makers (about 14%) appears to be
irreversible. Consequently, it was
suggested that in at least a portion
of heart cells from a 7-day embryo,
pacemaker capacity results from a
differentiate phenomenon which
can be influenced by the potassium
ion (K>) level. Results obtained dur-
ing the past year have caused us to
question whether this tentative con-
clusion is entirely justified.
It was observed that cultures incu-
bated overnight in maintenance
medium (M 629A1) exhibit more
spontaneously active cells (higher %
BC) if counted the next day, early in
the morning rather than later in the
day, and that the % BC decreases
with time in culture. This decrement
is illustrated in Table 4, which shows
incomplete data from a study still in
progress. Cells plated in medium
629 Al yielded a count of 47.3% BC
when counted 22 hours later. When
these plates were counted 27 hours
after the beginning of cultivation
(i.e., 5 hours after the first count)
DEPARTMENT OF EMBRYOLOGY
607
TABLE 4. Decrease of Spontaneous Activity in Chick Embryo Heart
Cells with Time in Culture
Time in culture, hours
22
27
48
72
120
629A, Low-K+ maintenance medium containing 4% serum but no CEE
%
BC
SE
N
%
BC
SE
N
47.3
42.7
28.5
1.5
2.0
1.9
19
17
3
729 A, Low-K+ growth medium, containing 4% serum plus 6% CEE
48.2 47.2 28.8 18.8
2.1 1.7 2.2 1.9
4 18 6 4
15.0
2
% BC had diminished to 42.7%.
After 48 hours of cultivation only
28.5% were still beating; by 5 days
that fraction had dropped to 15%.
The difference between 22 and 27
hours is quite consistent from one
culture to the next. Fifteen of the
19 plates counted at 22 hours were
counted again at 27 hours, for direct
comparison. Of those, 13 had de-
creased in percentage of beating
cells, one remained constant, and one
had increased. The resultant mean
difference was 4.7%; the standard
error of the difference between these
two was 1.2, indicating that the
difference is highly significant
(P = 0.01) . This progressive decrease
of spontaneity with time must be
taken into account when experiments
are designed to test the effects of any
manipulation of heart cell cultures,
especially those requiring treatment
of the cells for several hours. It is
also of interest in another respect:
In medium 629A few cells are seen
to divide; if maximum spontaneity
is taken as indicating the normal and
healthy state of these stable heart
cell cultures, then this decline in %
BC must result from the death or
reduced function of pacemaker cells
because of culture damage. On the
other hand, it has been suggested
that the trypsinization and cultiva-
tion procedures cause at least some
heart cells to become active.
Seven-day hearts may be minced
into small fragments, and these tis-
sue fragments seeded into plates of
standard medium (629A1) without
having undergone trypsinization.
After a few hours, the heart frag-
ments attach to the surface of the
culture dish, and by 24 hours are sur-
rounded by a halo of cells which have
emigrated from the tissue. Many of
these may be distinguished as M- or
F-cells. Although the central frag-
ments usually continue to beat rhyth-
mically for many days, at no time
do more than 1% of the cells that
become isolated at the periphery beat
spontaneously. We must bear in
mind, however, that those cells
capable of emigrating from a frag-
ment may be selected by their migra-
tory capacity as nonspontaneous cells.
However, the experiment does at
least suggest that the high levels of
pacemaker activity (occasionally up
to 50% BC) seen in cultures of
trypsin-dissociated cells may result
from the disaggregation procedure.
If this idea is correct, it follows
that the spontaneous decline in pace-
maker activity indicated in Table 4
could result from the gradual repair
of some part of the cell rather than
from its deterioration. But what
(VK
CARNEGIE INSTITUTION
part? Attention is centered on the
cell membrane, for the following rea-
sons: If pacemaker capacity of a cell
is associated with its having a
"leaky" membrane, that is, if the
diastolic depolarization typical of
pacemakers represents the unchecked
How of medium or potassium or both
down their respective concentration
gradients across the membrane, and
if treatment with proteolytic en-
zymes can cause quiescent cells to
become spontaneously active — as
shown by Holtzmann and Agin in
skeletal muscle fibers — the decline in
pacemaker activity might be depend-
ent upon the synthesis of new mem-
brane by damaged cells. It would not
be unexpected that such synthesis
might be fostered by normal levels
of K+, or inhibited by reduced levels.
Viewed against the background of
these considerations, the concept of
the "inducibility" of pacemaker activ-
ity by low K+ takes on a totally differ-
ent significance from that expressed
above.
To help distinguish between these
alternatives, experiments were de-
signed to test the relations between
two intervals: (1) the interval from
trypsinization to treatment with
high- or low-K+ medium, and (2) the
interval from treatment of the cells
to the time of counting them for
spontaneous activity. Results of six
experiments are shown in Table 5,
each experiment representing a mini-
mum of 800 cells counted in 4 repli-
cate plates. Experiment I illustrates
the depressive effects of high external
potassium-ion concentration [K+] e,
without allowing recovery. Cells were
trypsinized, washed, and cultivated in
medium 629A for 22 hours with the
use of techniques described earlier.
Each plate was washed in its respec-
tive medium, reincubated for 2 hours,
and then counted. In low-K+ medium
41.9% of the cells were active. At 4
and 12 mEq/1 K+, respectively, 23.1
and 5.2 of the cells beat spontane-
ously.
If, after 22 hours in 4 or 12
TABLE 5. Effect of Sequence and Duration of Treatment with High-K+
Low-K+ Medium on % BC
and
Experi
ment
No.
Initial [K+]e
mEq/l
Duration, Wash [K+]e Duration, Rewash [K+]e Duration,
hr mEq/l hr mEq/l hr % BC
f
1
22
1
2
41.9
1
4
22
4
2
... .
23.1
1
12
22
12
2
...
5.2
1
22
1
2
40.5
II
4
22
1
2
. * « «
26.9
I
12
22
1
2
...
26.9
I
I
1
6
1
16
1
2
44.3
III
4
6
1
16
1
2
44.2
12
6
1
16
1
2
43.9
1
20
1
6
1
2
39.9
IV
1
20
4
6
1
2
32.0
I
1
20
12
6
1
2
33.1
I
1
1
21
1
2
1
3
42.6
V
1
21
4
2
1
3
41.6
I
1
21
12
2
1
3
39.6
J
)
1
20
1
5
47.9
VI
4
20
1
5
• • • •
. .
43.3
I
12
20
1
5
...
40.8
DEPARTMENT OF EMBRYOLOGY
009
mEq/1 K+ the cells were washed in
low-K+ medium (experiment II) and
counted 2 hours later, they did not
recover completely to control levels
of % BC (40.5%). In fact, only
26.9% regained the capacity for
spontaneous activity. That is, one
third of the cells which could act as
pacemakers in low-K+ medium had
gained (or retained) that capacity as
a direct consequence of their resi-
dence in a medium containing only 1
mEq/1 K+.
Is this difference in recovery re-
lated to an "inductive" effect of low-
K+ medium, or to a reparative phe-
nomenon? If the first, the magnitude
of the difference should depend upon
the length of time the cells reside in
low-K+ medium after treatment with
high-K+. If the second, the length of
time in high-K+ medium would be
most important. Experiments III and
IV were designed to distinguish be-
tween these alternatives. In experi-
ment III, cells were trypsinized,
washed, and plated as described
above in medium 629 A containing 1,
4, or 12 mEq/1 K+. After 6 hours, all
plates were washed with low-K+
medium, and cultivated another 16
hours. The plates were then washed,
again with 629 Al (1 mEq/1 K+),
reincubated for 2 hours, and counted.
In all cases the level of spontaneity
was the same as if the cells had been
kept the full 24 hours in low-K+ me-
dium. There was thus no evidence of
"repair" (i.e., irreversible depression)
of the pacemaker membranes in high-
K+ media in the initial 6 hours.
In experiment IV the sequence was
reversed. The cells were plated ini-
tially in low-K+ medium, followed by
a 6-hour test period in either 1, 4, or
12 mEq/1 K+, and a 2-hour rewash
in low-K+ medium. With this pro-
cedure, recovery from the depressive
effects of high-K+ is incomplete. If
the recovery period is extended to 3
hours and the high-K+ wash reduced
to only 2 hours, almost complete
recovery occurs, as seen in experi-
ment V.
The critical experiment in this
series is not complete at the time of
writing. If cells maintained 22 hours
in high-K+ medium do not regain
maximum spontaneity after 2 hours
in low-K+ medium (experiment II)
or even after 5 hours (experiment
VI), will they do so in 8-10 hours?
If they do it may be concluded that
the K+ level during the initial 20-22
hours after trypsin treatment of the
cells is not relevant (experiments II
and IV). It could then be postulated
that what does determine the level
of spontaneous activity is the period
of time the cells have been in low-K+
medium before counting. On the
other hand, if after an extended
period in high-K+ medium and 8-10
hours in a low-K+ environment, the
cells still do not regain maximum
spontaneity, perhaps membrane re-
pair (i.e., depression of % BC) oc-
curs during the initial period of high
levels of K+, but not at 1 mEq/1.
Cell Morphology and Spontaneity
A second factor complicating the
analysis of the differentiation of
spontaneity is that heart cells are
represented by two morphological
types, one of which (F-cells) includes
very few latent pacemakers. M-cells
and F-cells may represent two com-
pletely separate populations. If F-
cells never under any circumstances
included more than 1% BC, any
"inductive" effect of low-K+ media on
pacemaker activity would be relevant
only for the M-cell group. However,
another possibility is that the M and
F configurations may represent two
different states of the same cells;
that is, if a single cell, during the
course of its division cycle or in
response to changes in its environ-
ment, could convert from a state in
610
CARNEGIE INSTITUTION
which it exhibits a thick, refractile
morphology and has one nucleolus, to
one in which it spreads thin on the
substratum and shows multiple nu-
cleoli, that cell would be counted as
an M cell in its first phase and an F
cell later.
That such a conversion might take
place in some cells was suggested by
data presented a year ago, although
at that time its significance was not
recognized. We had found (Year
Book 65, p. 529) that about 3% more
F-cells are found in medium contain-
ing 4 mEq/1 K+ than in low-K+
medium. This difference has since
been confirmed in a more extensive
series of tests, which also includes
counts on plates cultivated in 12
mEq/1 K+ medium. The results shown
in the upper left quadrant of Table 6
indicate that cultures maintained
for 20-22 hours in medium 629A
having 1, 4, or 12 mEq/1 K+ contain,
respectively, 33.4%, 36.8%, and
39.7% F-cells. More important, how-
ever, when the medium in these
plates is replaced by a low-K+ solu-
tion (upper right quadrant, Table 6),
the F-cells equalize at the level ini-
tially found in low-K+ medium; that
is, the % F in the 12 mEq plates
drops from 39.7 to 33.2 within 5
hours after low-K+ medium is added.
Presumably this represents a conver-
sion of F-cells to M-cells.
An analysis of this problem, em-
ploying time-lapse cinematography to
trace the constancy of morphological
configuration of individual cells, is
being completed by Charles Nuttall.
The findings thus far indicate that
the majority of F-cells remain recog-
nizable as such throughout their life
in culture, and when they divide (in
growth medium) , they tend to "breed
true." However, both M- and F-cells
spend some time in a rounded, highly
refractile state, and some time as
well-spread cells thinned out on the
surface of the dish. It is thus difficult
to be certain that an F-cell that is
round and refractile will not be
counted in the M category, and a
flattened, quiescent M-cell, among the
F group. In any event, we have yet
to determine what fraction of the
recovery of pacemaker capacity of
cells transferred from high-K+ to
low-K+ media is explainable by this
conversion of F-cells to M-cells.
Another approach to the deter-
mination of the properties of the M-
and F-cell populations can be made.
It was noted some time ago that F-
cells are more adhesive than M-cells
to the surface of the plastic tissue
culture dishes (which is the reason
F-cells spread so well on the bottom
of the dish). Iris Polinger has ex-
ploited this fact to devise a sep-
aration technique that results in
TABLE 6. F Cells as a Function of [K+]e per cent
Initial [K+]„
5-hour Wash
[K+l
e
mEq/1
%F
N
SE
mEq/1
%F
N
SE
Plating
medium 629A
1
33.4
28
1.5
1
31.4
28
1.1
4
36.8
28
1.5
1
30.2
24
1.1
12
39.7
9
2.9
Plating
1
medium 729A
33.2
25
1.1
1
34.0
6
2.9
1
34.9
10
2.4
4
35.5
6
2.2
1
48.7
6
1.7
12
49.7
6
3.3
1
48.0
6
3.2
DEPARTMENT OF EMBRYOLOGY
611
90%-95% pure M-cell or F-cell popu-
lations. She is also beginning to ex-
tend the techniques of clonal culture
and electron microscopy to these
cells, in preparation for a study on
the heritability and ultrastructure of
their properties.
Spontaneity and Growth
Yet another relationship which has
confused the issue of determination
of latent pacemakers in culture is
that of spontaneous activity to cell
division. Cells plated in medium 629A
remain viable and beat spontaneously
for many days. But they do not
divide. As indicated a year ago, how-
ever, chick embryo extract added to
629A promotes growth. Moreover, if
the embryo extract is first dialyzed
to remove potassium ions, the per-
centage of spontaneously contractile
cells appears to be unaffected. In
medium 729A (6% dialyzed CEE),
for example, when plates are counted
for spontaneous activity at 24 hours,
% BC is essentially identical with
that in control plates in nongrowth
medium. Moreover, the percentage of
pacemakers declines with time along
a curve similar to that produced by
cells in medium 629A (compare
upper and lower halves of Table 4).
However, in this medium the number
of cells per plate increases by approx-
imately tenfold after 4 days in
culture.
Although the number of spontane-
ously active cells is similar in growth
medium and in 629A, the morpho-
logical and physiological properties
of the cells are not identical. In
medium 729A, for example, the dis-
tinction between M-cells and F-cells
is more difficult to make. After 24
hours of culture the cells are usually
more flattened on the substratum,
and intranuclear inclusions tend to be
fragmented and granular. Although
the % F-cells in low-K+ growth and
nongrowth media is approximately
the same (compare upper and lower
left quadrants, Table 6), as many as
10% of the cells judged to be F-
cells by morphological criteria beat
spontaneously.
This large percentage of beating
F-cells was quite unexpected; in fact,
a change in the opposite direction
had been predicted on the basis
of the time-lapse cinematographic
studies mentioned earlier. It was
noted that, in growth media, F-cells
divide much more frequently than
M-cells. Moreover, just prior to and
during cytokinesis, F-cells stopped
migrating and took on a rounded
configuration similar to typical M-
cells. The average generation time of
F-cells that were observed to divide
on the films analyzed was 12 to 15
hours, while the rounded-up phase
generally lasted approximately 1
hour. Therefore it could be predicted
that, in growth medium, up to 7%-
8% of the F-cells at any given time
would be in a morphological state
where they could easily be mistaken
for quiescent M-cells. In nongrowth
medium this source of error would
be absent but, despite this fact, no
difference in the % F-cells in 629A1
and 729A1 was found. However, the
tendency for high-K+ media to aug-
ment the % F-cells was exaggerated
in the presence of embryo extract.
In medium 629A12, F-cells comprised
39.7% of the total population, where-
as in 729 A12, 49.7% were classed as
F-cells (Table 6). Moreover, after 5
hours in low-K+ medium, the % F
does not return to a lower number,
as happened in nongrowth media.
Summary
The distribution of cell types in the
7-day embryonic chick heart ap-
peared, at this time last year, to be
relatively simple. Handling cells by
standardized techniques in non-
growth, short-term cultures gave
evidence that, at least for a first
612
CARNEGIE INSTITUTION
approximation, the heart was com-
posed of only two major cell types.
F-eells were thought to include fibro-
blasts of the cardiac connective tissue
and endothelium. In the M-cell cate-
gory were grouped all myocardial
cells: pacemakers, latent pacemakers,
and nonspontaneous contractile cells.
It was thought that F- and M-cells
were completely separate and dis-
crete populations, with indeterminate
or intermediate cell types in very
small — essentially negligible — num-
bers. Recently, Mark and Strasser,
working with neonatal rat heart ven-
tricle cells have taken a similar posi-
tion, and have concluded that all
heart muscle cells are capable of
independent spontaneous contraction;
that is, that any cell which is flat-
tened and nonspontaneous is not a
heart muscle cell.
The results of the past year,
described above, indicate that this
is a highly oversimplified view. It
certainly seems reasonable to believe
that there are true fibroblasts dis-
sociated from embryonic heart tissue,
and that these would be classed as F-
cells. However, an error as high as
7% -8% could be made in categoriz-
ing this group alone, depending on
what fraction of these cells at any
given moment were rounded up in
mitosis. It is also clear that 65%-
70% of the heart cell population
obtained in culture by the techniques
described can be classed as M-cells,
and that as high as 70%-80% of
these (i.e., 45%-50% of the total
population) may be spontaneously
contractile in appropriate media at
a given time. However, changing the
culture conditions alters not only the
percentage of beating cells but also
the proportions of M- and F-cells;
that is, in only a few hours, some
cells appear to convert from one
morphological category to the other.
Moreover, the physiological switch
(spontaneously active to quiescent)
and the morphological switch (M to
F) are under the control of environ-
mental agents some of which are
identical for the two, and some differ-
ent. For example, in nongrowth
medium, increasing [K+]e from 1 to
12 mEq/1 depresses the % BC from
45% to 5%, while it raises the per-
centage of F-cells by about 7%. On
the other hand, with embryo extract
in the medium, the same increase in
K+ has less of a depressive effect on
% BC (about 11% at 12 mEq/1)
but increases the % F-cells by 16%
(up to 50% of the total population).
No clear picture of the relation
between cell morphology, mitosis, and
spontaneity has emerged yet from
these studies. There are still many un-
explained observations and unknown
variables. However, by attempting to
answer the many questions raised by
these results we hope eventually to
be able to give an accurate and
meaningful description of the differ-
entiation of function in embryonic
heart cells.
Electrophysiological Recording
from Heart Cells in Culture
In collaboration with Sheldon
Gottlieb, we have made much prog-
ress in improving techniques for elec-
trical recording from cultured heart
cells. It has been clearly demon-
strated that the resistance of the cells
to destruction by impalement ("im-
palability") and the electrical prop-
erties recorded with intracellular
microelectrodes, are both greatly
influenced by the conditions of cul-
ture, especially ambient oxygen ten-
sion, the presence of embryo extract,
and [K+]e.
Growth Media and Oxygen Tension
Insertion of a microelectrode into
a cell in a dense monolayer culture in
medium 629A in an atmosphere con-
taining 40% oxygen frequently re-
DEPARTMENT OF EMBRYOLOGY
613
suits in the immediate death of the
cell. Under these conditions, even
successful impalements rarely last
more than 2-3 minutes before the
cell membrane ruptures. Under simi-
lar conditions in medium 729 A (6%
CEE), impalements are more fre-
quently successful and often last 5
minutes or more.
In view of the observations of
Gerda Mark and her collaborators
(described at the Gordon Conference
of August 1966) indicating that a
high-oxygen atmosphere can be dele-
terious to rat heart cells in culture,
tests were run this past fall and
winter to see if chick heart cell mem-
branes were more fragile in an
atmosphere of high-oxygen tension
(40%) than in normal (20% 02) or
low-oxygen (10%) environments.
The results were striking. Cells
maintained for 48 hours in medium
629 A in 10% 02 can be impaled and
recordings made continuously for up
to 15 minutes, although the average
recording time under these conditions
is less than 7 minutes. In medium
729 A in 10% 02 atmosphere, im-
palability of cells in dense cultures
improves remarkably. Under these
conditions (now standard) the great
majority of impalements are success-
ful and recording can be continued
routinely for 15-30 minutes. Impale-
ments are usually terminated as a
result of withdrawal of the electrode
from the cell rather than cell mem-
brane rupture. On occasion, in fact, a
cell has been reimpaled and record-
ings continued.
Aside from their value in practical
application, these observations give
rise to numerous questions: Is the
difference in impalability of cells in
629A versus 729A a direct effect of
chick embryo extract on the structure
of the membrane, or is it indirect,
resulting from synthetic capacities of
mitotically active cells? Does the
fragility of the membranes of cells
cultured in 629 A in 40% 02 result
from their treatment with trypsin,
or from the subsequent culture con-
ditions? Is the greater impalability
of cells in 729 A in 10% 02 consequent
upon synthesis of new membrane
material? If so, does it follow that
membrane synthesis can be controlled
in heart cells by oxygen tension and
the presence or absence of embryo
extract? Answers to all of these
questions await further studies.
External Potassium Ion
Concentration
Means and standard deviations for
measurements of the properties of
action potentials recorded from 84
cells in culture are summarized in
Table 7. These data represent record-
TABLE 7. Electrical Properties of 7-Day Embryo Heart Cells in Culture, mV
Plating Medium
Property
729A1
729A4
729A12
Number of cells impaled
Resting potential
Action potential*
Overshoot
Positive afterpotential
Threshold
41
55 ± 6
75 ± 8
13 ± 5
8
9
31 f
63 ± 4
85 ± 7
21 ± 4
3
8
12}
48 ± 4
70 ± 7
25 ± 3
7
* Maximum negative-to-positive difference,
f In 1 1 of these 31 cells, no PAP occurred.
\ Five of these 1 2 cells were quiescent and therefore had no AP or overshoot.
614 CARNEGIE INSTITUTION
ings from cells in contact on all sides sampling error, we are testing the
with neighbors, in dense monolayer response of the RP to K+ of individ-
cultures. The medium was 729A con- ual cells, exploiting our ability to
fanning 1. 4, or 12 mEq/1 K+. Am- maintain a given impalement for 30
bient atmosphere was 10 cc 0L>, 5 % minutes or more. In this study a cell
C02j and 85 r'c N in the incubator in low-K+ medium is impaled and its
during the 48 hours of cultivation RP and AP measured. Without dis-
prior to impalement, and 10 % C02 turbing the preparation a small in-
to air during the recording session, crement of KC1 is injected into the
In all cells in low-K+ medium and culture dish, 5-10 minutes are al-
most of those in 729A4 a period of lowed for equilibration, and the new
hyperpolarization followed each ac- RP and AP parameters are measured
tion potential. This positive after- on the same cell. Although this study
potential (PAP) made exceedingly is not yet complete, it has already
difficult the definitive determination confirmed that as [K+]e is increased
of the level of the resting potential in small steps from 1 mEq/1 to about
of these cells. It was therefore de- 5 mEq/1, RP rises (i.e., polarizes),
cided to arbitrarily define the resting As [K+]e is elevated beyond 5mEq/l
potential (RP) as the point where the cell gradually depolarizes, most
the convex curve of the afterpotential cells becoming quiescent at 10-12
joined the concave curve of the mEq/1 K+. These results agree well
diastolic depolarization. For cells In with the potassium dose-response
high-K+ media, which did not exhibit curve given in this report last year
afterpotentials, RP was taken as the in terms of % BC. Extension of this
level of the stable diastolic voltage, study should permit an analysis of
The action potential (AP) was the electrophysiological mechanisms
defined as the total peak-to-peak underlying the capacity of K+ to
magnitude of each spike from the top switch pacemaker cells on and off.
of the overshoot to the trough of the
PAP. The threshold potential (THR) (jen Density
was taken as the difference between
the RP and the point of juncture Last year we had to report failure
between the concave foot of the of our attempts to record from cells
upstroke and the beginning of the completely isolated from one another
maximal rising velocity phase of the in low-density cultures. We made the
action potential. point that cells in complete isolation
The AP and RP are greatest at a from all neighbors had membrane
[K>] e of 4 mEq/1. In either higher properties different from similar cells
or lower levels of K+, both param- in contact with neighbors in groups
eters decrease. By contrast, the over- or monolayers. Although the second
shoot increases directly with the observation has since been amply con-
potassium level. firmed — very real differences seem
The peak-to-peak amplitude of the to exist, depending on cell density —
AP of 85 mV compares well with we can now report modest success
those published by other workers on in impaling such isolated cells. At
intact heart tissue, as does the rest- this writing, recordings have been
ing potential of 63 mV. The depres- made from 12 isolated cells. The RP
sion of the RP by high-K+ media was of these cells ranged between 10 and
expected, but the depolarization in 1 17 mV. From two cells low-level
mEq/1 K+ was not. To confirm that rhythmic waves were recorded in
this result was not due to a statistical synchrony with the visible beat.
DEPARTMENT OF EMBRYOLOGY
615
These had amplitides of 3-5 mV, and
durations of approximately 200
msec. It should be emphasized that
all of these recordings were made
from cultures in which cells in more
densely populated areas of the same
dish could be impaled with ease using
the same microelectrode, and yield
recordings of AP and RP of the order
reported in Table 7. We are continu-
ing our attempts to devise special
techniques to record from isolated
cells without producing evidence of
damage.
Morphogenesis of the
Early Heart
Since September 1966 Helge Stals-
berg, on leave from the Oslo Uni-
versity Medical School, has been
investigating the mode of formation
of the early chick heart tube and the
developmental mechanisms underly-
ing its early asymmetry.
The technique for grafting small
pieces of endoderm and mesoderm
from embryos labeled with tritiated
thymidine to homologous sites in
unlabeled recipient embryos of the
same stage, as well as the later trac-
ing of the fate of the implants by
autoradiography on serial sections,
was employed previously in this
laboratory by Rosenquist (Year Book
6U, pp. 477-481). An analysis of the
implants giving rise to heart tissue
was published by Rosenquist and
DeHaan in volume 38 of Contribu-
tions to Embryology. The observation
that an implant of precardiac meso-
derm retains its integrity as a
coherent fragment during the forma-
tion of the epimyocardial tube seemed
to support the idea that the topo-
graphical fate of a preepimyocardial
cell is fully determined by its loca-
tion in the head-process embryo. It
seemed possible, therefore, by reduc-
ing the size of the implants and by
some refinements of the technique of
analysis, to map accurately the pre-
conal, preventricular, and preatrial
subdivisions of the precardiac area.
This analysis is still in progress, but
some points have already been clari-
fied. It was reported previously by
Rosenquist and DeHaan that the
right and left heart-forming regions
were connected across the anterior
midline. To test this idea further, 26
implants 0.10-0.15 mm wide have
been made within 0.3 mm of the mid-
line on both sides, in the anterior
"bridge" region of stage 5 embryos.
None of these implants gave rise to
labeled cells in the epimyocardium.
Thus a gap of at least 600 microns
must exist between the preepimyo-
cardial areas anteriorly. This gap
was missed in the earlier studies
because of the larger size of the
implants. The endocardium takes its
origin from an area nearly identical
to that of the epimyocardium, though
it appears to extend more medially
than the latter. Out of 11 fragments
placed across the anterior midline,
only 1 gave rise to 2 labeled endo-
cardial cells, while the remaining 10
implants, as well as some paramedian
grafts, contributed to endocardial
cells. Thus there also appears to be a
gap, measuring about 400 microns,
between left and right preendocardial
areas at stage 5. In general, the
labeled cells in the endocardium are
found in the same region of the
heart as are the labeled epimyo-
cardial cells from the same implant.
The preendocardial cells must there-
fore also be regionally organized at
stage 5, though the regional bound-
aries appear to be less definite than
those of preepimyocardial tissue.
To aid in the understanding of the
morphogenetic movements that link
the fate map of the stage 5 embryo
to the heart tube seen 24 hours later,
microdissection of intervening stages
has proved to be very useful. After
treatment of the explanted embryo
with 1% trypsin solution for 1-3
616
CARNEGIE INSTITUTION
minutes, the endoderm can usually
be separated from the splanchnic
mesoderm by dissection with fine
tungsten needles. The addition of 1
or 2 drops of embalming* fluid after
trypsinization hardens the tissues
slightly, which facilitates the dissec-
tion procedure and helps to preserve
the form of the folded layers. Expos-
ing the entire layer of mesoderm in
successive developmental stages re-
veals the lateral mesoderm remaining
as a coherent sheet from stage 5 on,
continuous with paraxial mesoderm
medially and with head mesoderm
anteromedially. The sheet of meso-
derm becomes folded on each side
between the ectoderm of the lateral
body fold and the endoderm of the
wing of the anterior intestinal portal.
The anterior end of the heart differ-
entiates close to the crest of these two
original folds after they have met
under the floor of the foregut. Obser-
vations made in these microdissection
studies confirm the existence of a gap
of considerable width between right
and left heart-forming material in
the stage 5 embryo.
The cause of the bending of the
heart tube to the right is generally
believed to be a difference in regional
growth rates between the left and
the right side of the heart, but infor-
mation on when, where, and how
such differences arise, and indeed
whether they occur at all, is very
scanty. There is some evidence in
the literature that in the amphib-
ian heart, the bulboventricular loop
forms as a result of a substantial
overgrowth of ventricular material
originating from the left side. In
Year Book 65, pp. 536-537, van
Praagh described preliminary obser-
vations on the relative sizes of the
heart subdivisions formed on each
side after a "cardia bifida,, operation
had been performed, indicating that
a similar growth difference may be
present in the chick.
To test this idea further, the por-
tion of epimyocardium contributed
by each side of the cardiogenic cres-
cent has been identified by labeling
one side of chick embryos with tri-
tiated thymidine at stages 5-6 and
subsequently analyzing the serial
cross sections with the techniques of
autoradiography, after the embryos
had developed to stages 10~ to 12+.
The label was introduced in a 1 X 2-
mm piece of 25-micron-thick milli-
pore filter that had been soaked in
H3-thymidine solution (80 /xC/ml,
sp. activity 15.9 /xC/mM). On em-
bryos implanted endoderm side up
by the technique of New, one such
piece of filter was placed on the
endodermal surface on one side of
the embryo, and a similar piece,
soaked in the same concentration of
cold thymidine, was placed on the
contralateral side. The radioactive
piece was placed on alternate sides in
different embryos. Following a label-
ing period of 20 minutes at 37.5 °C,
the filter pieces were removed by
flooding the embryos with saline.
The embryos were then transferred
to fresh medium containing excess
cold thymidine (1 /xg/ml in 50% egg
albumen, 50% Howard-Ringer's salt
solution) . Although some radioactive
label always spread to the contra-
lateral side, a distinction between the
heavily and the weakly labeled re-
gions could always be made and was
usually very clear. The drawings
shown in Fig. 18 are based on graphic
reconstructions of 25 embryos so
treated. They illustrate typical find-
ings at 10, II-, and 12. At stage
10 the heart has already started
to bend to the right side. This is
reflected in the curved course of the
line of fusion between epimyocar-
dium originating from the two sides.
As development proceeds, the ventral
midline swings to the right, and the
left-sided material soon forms the
entire ventral surface of the tube in
DEPARTMENT OF EMBRYOLOGY
617
//
//
: .'* "li
/ ' .•'•'.•'
•"■•;■
'• * ' ■ f
/ {■).■■'■■
1 •"••*.'■•.'■.'
IV:-.^-:
Fig. 18. The location of epimyocardial tissue originating from right and left precardiac material.
Drawings of graphic reconstructions of embryos unilaterally labeled with tritiated thymidine are
shown in ventral view. Epimyocardium originating from the left side is stippled. Cross sections
are oriented with dorsal side up, and represent the levels indicated by transverse lines. The
developmental stages represented are 10, 11~,and 12.
the mid-ventricular region. However,
it can be seen from the accompany-
ing cross-sectional drawings that this
is not caused by overgrowth of left-
sided ventricular material, but mainly
by a 90° rotation of the heart tube in
this region, made possible by the
formation, elongation, and partial
disappearance of the dorsal meso-
cardium in the mid-ventricular re-
gion. The distinction between heavily
and weakly labeled regions can also
be made in the endocardium. As in
the grafting experiments, labeling of
the endocardium corresponds in gen-
eral to that of the epimyocardium
adjacent to it. Thus, in the flattened
endocardial tube of the bulboven-
tricular region, the transition from
right-sided to left-sided material is
found at each of the acute bends of
the circumferences.
As a further approach to the ques-
tion of differential growth in the
forming heart, the mitotic activity in
precardiac and cardiac mesoderm is
being investigated. A technique has
been developed whereby regional
fragments of the epimyocardial tube,
as well as fragments of precardiac
mesoderm, are cut out and squashed
on microscopic slides. This provides
preparations well suited for cell
counts. In chick embryos cultured for
2V2 hours on medium containing
Colcemid (0.3 /ig/ml), the relative
number of cells arrested at meta-
phase can be recorded in regional
squash preparations. The most con-
spicuous finding so far is that as the
precardiac mesoderm is incorporated
into the epimyocardial tube, the
number of mitotic figures drops from
the level of 15%-30% at stages 5-
7 to 0%-l% in the bulboventricular
region of stages 11-12, and sub-
sequently increases to 3%-8% at
stages 13-14, apparently following a
specific regional pattern.
61S
CARNEGIE INSTITUTION
NEURAL REGULATION OF THE
EMBRYONIC CHICK HEART
James F. Case
An investigation of neural regula-
tion of the embryonic chick heart
was begun, predicated upon the early
establishment of vagal influence,
which has already been shown by
others probably to develop by six
days of incubation. Other elements
of the cardioregulatory system have
not been demonstrated and these
specifically were sought.
Methods were devised by which
electrical activity of hearts of unre-
strained embryos with nearly intact
circulation could be recorded with
flexible suction electrodes applied
directly to the heart. With such
preparations it was possible to con-
firm that injection of acetylcholine
into the IVth ventricle of the brain
caused reversible slowing of the 6-
day heart. At that time, however, no
other elements of the heart control
system appeared to be functional.
Thus partial extravasation, hypoxia,
or elevated carbon dioxide concentra-
tions caused only slowing and abrupt
cessation of heartbeat in response to
extreme stimuli. Sixteen-day embryos
appeared to have a more complete
regulatory mechanism, since all of
these stimuli invariably produced at
least transient acceleration, and since
mechanical pressure of the cervical
carotid artery induced slowing of the
heart.
The precise timing of the establish-
ment of these reflexes remains to be
determined, as well as their signifi-
cance in circulatory homeostasis dur-
ing later embryonic development.
DISSOCIATION AND CULTURE OF BRAIN TISSUE
R. L. DeHaan
Despite the importance of tech-
niques for cultivating nervous tissues,
attempts by other workers to obtain
growth and differentiation of nerve
cells after dissociation from brain or
spinal cord have met with only
limited success. Intact ganglia or
fragments of nervous tissue often
survive well, but normally do not
undergo mitosis.
Both heart cells and brain cells are
electrically active. Much of our atten-
tion in the past 3 years has been
devoted to devising gentle techniques
of dissociation and culture for heart
cells, techniques specifically designed
to maintain or augment the cells,
electrogenic properties. It therefore
seemed reasonable at least to con-
sider whether the precautions and
special techniques that have proved
successful for cardiac tissue might
not have similar salutary effects on
nervous tissue.
This idea has been tested, with
what appear tentatively to be positive
results. Seven-day embryonic chick
brain, dissected carefully from its
meningeal membranes, was disso-
ciated with the trypsinization tech-
nique described last year (Year Book
65, p. 527) . The resultant suspension
of isolated brain cells was plated into
medium 729A1 and incubated in a
low-oxygen atmosphere.
As these cells settle out of the
medium onto the bottom of the dish,
they manifest behavioral properties
strikingly different from those of
heart cells treated with the same
techniques. Heart cells remain sep-
arate from each other and attach
DEPARTMENT OF EMBRYOLOGY
619
rapidly (within 3-4 hours) to the
plastic substratum. Brain cells, on the
other hand, come to rest on the bot-
tom of the dish, but do not begin to
adhere to it for 20-30 hours. How-
ever, during this period they are very
adhesive to one another, sticking
tightly whenever they collide. In this
manner, the initial suspension which
contains 90% or more of single cells
is soon converted to aggregates of 2-
20 cells. Within the first 24-48 hours,
most of these cell clumps attach to
the bottom, and gradually begin to
spread on the surface. For this rea-
son, care must be exercised in re-
freshing the medium the first time, to
avoid washing all the cells off the
dish.
Cultures thus established have been
maintained up to 15 days. The cul-
tures are characterized by two major
cell types. Within the first few days
most of the cells which remained
isolated or formed groups of only two
or three take on a fibroblastic appear-
ance and spread to form a ground
mat on the dish. After a week of
culture, cells appear on top of this
ground mat which have the appear-
ance of typical neural elements: bi-
polar and multipolar neurons, and
glial cells. Mitotic activity has been
evidenced in both the fibroblastic and
"neural" cells by obvious increases
in density of cells in the dish over a
2-week period, and with the aid of
autoradiography.
Cultures were incubated with 0.1
/xc of tritiated thymidine for 2 hours
on the fourth, eighth or fifteenth day
of cultivation. At the end of the thy-
midine incubation, the plates were
washed gently with three changes of
fresh medium and three changes of
Howard-Ringer solution, and were
then dried quickly in a stream of
warm air. They were postfixed with
95% ethanol, redried, and coated in
the dark with Kodak NTB2 nuclear
emulsion. After storage over silica
gel in the dark for 7-10 days, these
plates were developed in D-19
developer and mounted in Gurr's
water-mounting medium under cover
slips.
Examination of these preparations
is not yet complete, but thymidine
incorporation has already been con-
firmed in cells of both the ground-mat
type and those of neural morphology.
It must be emphasized, however, that
at present no evidence is available,
either of a histological or physiolog-
ical nature, that the cells in question
are really differentiated nerve cells.
They derive from brain tissue; their
morphology is suggestive. Further
studies are required to determine
their true properties.
BIOCHEMISTRY AND PHYSIOLOGY OF THE GONADS
Germinal Vesicle Breakdown in
Response to Steroids and
Gonadotropins in Rana pipiens
Allen W. Schuetz
Normal initiation of ovulation and
meiotic maturation in frog oocytes
results from the action of the pitui-
tary gland on the ovary. Meiotic
maturation in oocytes is preceded by
the breakdown of the germinal vesi-
cle (GVBD). Previous studies have
demonstrated that steroid hormones
are capable of inducing breakdown
of the vesicle in isolated individual
oocytes in vitro. Steroid hormones
with progestational, androgenic or
mineralocorticoid, but not estrogenic,
activity are capable of inducing
GVBD.
Experiments were conducted utiliz-
ing this test system to further eluci-
date the following: (1) the cellular
and biochemical mechanisms in-
620
CARNEGIE INSTITUTION
volved in progesterone- and pituitary-
induced vesicle breakdown; (2)
whether steroid hormones are impor-
tant in mediating pituitary-induced
GYBD; and (3) the function of the
follicular cells in the process.
Addition of 1-10 pg of progester-
one to an oocyte in vitro results in
essentially 100 cc GYBD in large pig-
mented oocytes within a 24-hour
period. Progesterone is ineffective in
producing GYBD, even at much
higher concentrations, in small, un-
pigmented oocytes which have not
undergone vitellogenesis. The first
signs of breakdown are observed
between 10 and 18 hours after the
administration of progesterone, and
within a 5-hour period GVBD is
completed. The time at which GVBD
is initiated varies considerably from
frog to frog; however, the time dur-
ing which GYBD is completed is rela-
tively constant in all animals. The
presence of the steroid is required in
the incubation medium for a short
period of 5-15 minutes. Oocytes
washed free of progesterone and
placed in medium alone after this
period undergo GVBD normally. The
breakdown induced by progesterone
or pituitary glands is inhibited when
oocytes are incubated with or without
follicular cells, at 4°C.
Protein synthesis appears to be
involved in the process of GVBD.
Introduction of puromycin (an inhib-
itor of protein synthesis) to the incu-
bation medium up to 5 hours after
the addition of progesterone inhibits
GYBD; however, it does not inhibit
breakdown if added 5-6 hours prior
to normal initiation. Although these
data suggest that progesterone-
induced protein synthesis occurs over
a relatively short period, 4-5 hours,
puromycin also has a direct effect
upon oocytes. Oocytes incubated with
puromycin for varying periods of
time are responsive to progesterone
if they are washed free of the
inhibitor within a 5-hour period.
Steroid hormones, which are nor-
mally inhibitors of protein synthesis
(glucocorticoids), or antagonistic to
progesterone (estrogens) are ineffec-
tive in preventing GVBD. Similarly,
Su-4885-Metyrapone, an inhibitor of
enzymes concerned with steroid
metabolism, is ineffective in prevent-
ing GVBD. If protein synthesis is
initiated by progesterone it does not
appear to require the synthesis of
new messenger RNA. The addition of
actinomycin D, an inhibitor of mes-
senger RNA synthesis, does not pre-
vent GVBD. Studies on the action of
these metabolic inhibitors on pitui-
tary-induced GVBD reveal markedly
different results from those obtained
when progesterone is used. Addition
of puromycin or actinomycin D to
oocytes exposed to pituitary glands
is equally effective in preventing
GVBD, whereas progesterone is not
inhibited by actinomycin D. Addition
of actinomycin D at varying times
after administration of pituitary
glands reveals that after 1-4 hours
(depending upon the frog tested)
actinomycin is ineffective in prevent-
ing GVBD. Experiments comparing
the time of GVBD after the addition
of either progesterone or pituitary
glands to oocytes show that oocytes
responded to progesterone with
GVBD 1-5 hours before those ex-
posed to pituitary glands.
In an attempt to determine the site
of action of the pituitary, pituitary
glands were added to oocytes in
vitro with and without follicular
cells. Three types of responses were
observed: (1) Oocytes from some
frogs responded to pituitary treat-
ment with or without the follicle cells
being present. Actinomycin D pre-
vented GVBD in all of them. (2) In
other animals removal of follicular
cells from oocytes prevented GVBD
initiated by pituitary gland treat-
ment. (3) Oocytes from one animal
DEPARTMENT OF EMBRYOLOGY
621
were stored for one day in the refrig-
erator. Upon resumption of incuba-
tion, in the absence of any hormone,
those oocytes with follicular cells
underwent GVBD, whereas the nu-
cleus in those eggs whose follicle cells
were removed, remained intact.
The findings indicate that pituitary
glands and progesterone are equally
effective in inducing GVBD. The
differential susceptibility of pitui-
tary-induced GVBD to actinomycin
D inhibition strongly suggests the
presence of ovarian intermediary
compound (s) responsible for the
initiation of GVBD. Whether this
substance is messenger RNA or is
steroidal is not clear. The effect of
removing the follicular cells from
the oocytes may suggest that the
pituitary gland acts directly on the
oocyte as well as on the follicular
cells.
Testicular Sorbitol
Dehydrogenase
D. W. Bishop, E. C. Muecke,
A. Mussehnan, and W. Schrank
Emphasis during the past year has
been placed upon the nature of the
testicular enzyme, sorbitol dehydro-
genase (SDH), and its relation to
germ cell development.
We have investigated the distribu-
tion, localization, and extraction
characteristics of SDH from guinea-
pig testes, and its potential utility as
an assay system for germinal epithe-
lial content in developing and mature
testes. The rationale behind this
undertaking was twofold: (1) The
complex and well-regulated process
of spermatogenesis may be elucidated
to some degree if one or more readily
identifiable and cell-specific biochemi-
cal markers are available for study.
(2) A germ cell-specific biochemical
label could be used as a reliable, quan-
titative assay for germinal epithelium
in normal and experimentally modi-
fied testicular tissue, as well as in
analyses of germ cell development in
tissue and organ cultures and in
implantation studies. These analyses
have too long been retarded by the
lack of sensitive and reproducible
methods of acquiring meaningful
data that might be subjected to sta-
tistical treatment. An unanticipated
aspect of these investigations, fur-
thermore, concerns the effects that
pure preparations of SDH and anti-
SDH serum may have on normal
processes of spermatogenesis in the
guinea pig. The precise roles of SDH
elaboration and activity vis-a-vis
germ cell differentiation and sperm
metabolism are under investigation.
A brief report of some of these
data was presented at the Federation
meetings (1967), and a more com-
plete discussion is in preparation for
the William C. Young memorial
volume, Reproduction and Sexual
Behavior.
Sorbitol dehydrogenase may be
readily extracted from homogenized
mammalian testes and tested spec-
trophotometrically for activity by
determination of A O.D./min at 340
mp (Year Book 64, pp. 496-504) . The
optical change is a function of DPNH
produced per minute per milligram
of protein. The substrate, sorbitol, a
6-carbon hexahydric alcohol, is oxi-
dized to fructose in this test system
(Fig. 19).
The extraction procedure developed
for the removal and purification of
SDH from mammalian testes is a
modification of the technique em-
ployed by Libby and Williams-Ash-
man for liver enzyme. Of considerable
significance is the treatment of the
first supernatant fraction (Fig. 20)
with ethanol-chloroform, which not
only removes heme and other com-
ponents, but also inactivates lactic
dehydrogenase, or removes it, or
both. Lactic dehydrogenase is other-
wise a constant and major contami-
622
CARNEGIE INSTITUTION
CH90H
I *
H-C-OH
CH^OH
C = 0
+ ' +
OH-C-H + DPN ^=^ OH-C-H 4- DPNH + H
H-C-OH
I
H-C-OH
CH2OH
H-C-OH
I
H-C-OH
CH20H
Fig. 19. Enzymatic oxidation of sorbitol to fructose catalyzed by sorbitol dehydrogenase.
nant. Column chromatography on
DEAE-cellulose, followed by CM-
cellulose, yields a sharp peak (Fig.
21) when serial fractions are assayed
for SDH activity. The most active
fraction in such an extraction has
given a specific activity of 13 pinoles
DPXH formed/min/mg protein at
21 CC, approximately a 500-fold in-
crease over activity of the initial
supernatant fraction. The final prod-
uct has no alcohol, lactic, or glucose
dehydrogenase activity and does not
catalyze the reduction of DPN by
glycerol; it does, however, contain a
trace of malic dehydrogenase. Sub-
strate specificity of the purified
enzyme preparation, like that of
crude testicular extract, is selectively
TESTIS H0M0GENATE
I2,000g
SUPERNATE
EtOH CHCI3
!2,000g
AQ. LAYER
DIALYSIS
40-70% A.S.
I2,000g
CENTRIFUGATE
DIALYSIS
DEAE
-» SDH
CM
Fig. 20. Flow chart of extraction procedure for
sorbitol dehydrogenase of guinea pig testis.
but not specifically absolute: sorbi-
tol > ribitol > d-xylitol > mannitol
> 1-arabitol. Similarly, both the
crude and the pure preparations show
an increase in activity when the
cofactor analogue, 3-acetylpyridine
DPN, is substituted for DPN. A pB.
optimum for purified SDH appears at
9.8 in Tris buffer. The final prepara-
tion loses activity upon treatment
with ammonium sulfate; activity is
retained, however, if the purified
enzyme is stored in 50% glycerine at
— 70°C.
The active, guinea-pig SDH frac-
tions, collected from the CMC column,
are highly antigenic. Antisera, pro-
duced in rabbits in response to SDH,
combined with complete adjuvant,
give high titers and show one, or at
most two, bands when reacted in the
Ouchterlony immunodiffusion system
(Plate 3A). In agar, SDH migrates
electrophoretically toward the cath-
ode and can be subsequently reacted
with specific, absorbed antiserum
(Plate 3B). These serological results
suggest a high degree of purification
of the extracted enzyme preparation,
confirming the specific activity data
noted above. An unexpected ancillary
finding indicated by preliminary tests
is the probable deleterious effect that
pure guinea-pig SDH, employed as a
sensitizing agent, may have in guinea
pigs; in adult males, testicular lesions
DEPARTMENT OF EMBRYOLOGY
623
0.3
o
d
0.2
0.
DEAE
CM
80 90 50
TUBE NUMBER
Fig. 21. Sorbitol dehydrogenase activities in successive (5-ml) fractions eluted from DEAE-cellu-
lose and CM-cellulose columns.
have occasionally appeared after
intramuscular injection of SDH
combined with adjuvant.
Preliminary studies of mammalian
testes have indicated that SDH is a
constant component of the mature
gonad. Evidence acquired more re-
cently demonstrates that the enzyme
is present in testes of a wide variety
of animals from invertebrates to man
(Table 8). This suggests an impor-
tant, if as yet unknown, role for the
enzyme in germ cell differentiation or
sperm metabolism.
Previous findings that enzyme
activity increases with age of the
testes in guinea pigs led to attempts
to localize SDH with respect to stage
of cell differentiation. Cytochemical
procedures employing the tetrazolium
reduction system were developed by
modification of the LDH technique.
Within the limitations of the pro-
cedure, the results clearly show in-
tense enzyme activity in the more
mature germ cells of the seminiferous
tubules of adult testes and virtually
no activity in neonatal tubules (Plate
4A-D).
Cellular localization of SDH was
further clarified by utilization of
fluorescent-antibody labeling by
means of a "sandwich-type" proce-
dure. The presence of the enzyme was
demonstrated throughout the tails of
mature sperm (Plate 5A). Of par-
ticular significance, moreover, was
the localization of SDH in developing
624
CARNEGIE INSTITUTION
TABLE 8. Sorbitol Dehydrogenase Activity of
Testes from Several Groups of Animals
Animal
A O.D. /Minute/Sample*
Guinea pig
0.031
Mouse
0.024
Macaque
0.019
Rat
0.015
Man
0.004
Rabbit
0.004
Rooster
0.014
Squid
0.010
Grasshopper
0.052
* Equivalent samples on a wet weight basis,
tested in presence of DPN and sorbitol in Tris
buffer, pH 8.2
germ cells (Plate 5C) which have
been tentatively identified as growth
stages of the first spermatocyte cell
generation (Plate 6A). The spectac-
ular appearance of SDH in these cells
confirms the marked increase in
enzyme activity in testes of 4-6 weeks
of age in these Hartley-strain guinea
pigs, and supports the general con-
cept of these cells' intense synthetic
capacity.
As an assay system for functional
germinal epithelium, the determina-
tion of SDH activity in testicular
homogenates has proved particularly
rewarding. This useful determination
is rapid, reliable, sensitive, readily
quantified. Moreover, the results with
mammalian testes which have been
impaired by either immunologically
induced aspermatogenesis or by ex-
perimental cryptorchidism, demon-
strate unequivocally that the germinal
epithelium and not the interstitial
tissue suffers damage; under both
these conditions, Leydig-cell activity
remains high, as judged by cyto-
chemical tests for androgen-secretory
function.
Enzymatic assay of an immuno-
logically impaired aspermatogenic
testis may serve only to confirm
obvious histological damage when the
lesion is severe. The advantages of
the spectrophotometry analysis, how-
ever, become clear when quantita-
tion is desirable, many samples are
to be averaged, and statistical treat-
ment is required. Moreover, when
only mild or subtle alterations in the
germinal epithelium are produced,
the SDH assay becomes highly
significant — in such investigations,
for example, as those that involve
minimum dosage levels, time studies,
or reversibility effects. The minimum
effective sensitization dose of asper-
matogenic factor as a function of
testicular SDH is shown in Fig. 22;
from these data one or a few milli-
grams of CPM ("chloroform-purified
material/' extracted from the guinea-
pig testes) appear sufficient to ini-
tiate the testicular response.
The SDH assay system has also
proved useful in studies of experi-
mental cryptorchidism in the guinea
pig in which, it has been found,
cryptorchidism can be followed by
orchiopexy with near-complete recov-
ery of the testis, if the initial opera-
tion is performed in young animals.
In adult guinea pigs, cryptorchidism
results in rapid loss of germinal
epithelium and an accompanying
decrease in SDH titer (Table 9). In
animals cryptorchidized at 10 days of
age, however, limited testicular devel-
opment occurs and the germinal epi-
thelium remains responsive to full
growth conditions when the testis is
again exteriorized by orchiopexy
TABLE 9. Changes in Sorbitol Dehydrogenase
Activity with Time in Adult Cryptorchid
Guinea Pigs
Days in
Control
Cryptorchid Cor
idition
Activity, %
0
100
3
93
6
74
9
50
12
33
15
25
DEPARTMENT OF EMBRYOLOGY
625
0.03
0.02
"i
Q
6
<
I
S o.oi
en
<:
>
I 10' 10* \06
CONCENTRATION CPM - pgm/ml
0
Fig. 22. Sorbitol dehydrogenase activity in testes of guinea pigs sensitized with small amounts
of aspermatogenic factor (CPM) combined in adjuvant. Each animal received 1 ml; each point is
the average of 1 0 tissue assays.
(Plate 6B-D). SDH assays of the tes of similar ages (Table 10).
impaired and recovered gonads give Within any one group of experimen-
an accurate and clear picture of the tal animals, SDH-activity determina-
degree of testicular inhibition and tions offer an excellent basis for
recovery as compared to normal tes- statistical analysis (Table 11). In
TABLE 1 0. Inhibition and Recovery in Testes of Cryptorchid and
Orchiopexed Guinea Pigs,* control SDH activity, per cent
Weeks
Cryptorchid
Weeks
Orchiopexed
Inhibition, %
Recovery, %
13
8
6
14
14
14
90
85
50
53
79
82
* First operated on at 1 0 days of age.
TABLE 11. Testicular Sorbitol Dehydrogenase Activity 90 Days after Cryptorchidism and 100
Days after Orchiopexy of 1 0-day-old Guinea Pigs
SDH Activity, A O.D. /Minute
Condition
No. of
of Testes
Animals
Mean
S.D.
S.E.
Normal
neonate
7
0.0031
0.0016
±0.0006
Cryptorchid,
90 days
18
0.0047
0.0013
±0.0003
Orchiopexed,
100 days
14
0.0129
0.0051
±0.0014
Normal
adult
14
0.0278
0.0028
±0.0007
626
CARNEGIE INSTITUTION
summary, the SDH assay serves as an
accurate biochemical index of germi-
nal epithelial content and, recipro-
cally, the systems here surveyed
indicate that the enzyme is princi-
pally, if not entirely, associated with
the germ plasm, and possibly with
key stages of development during
germ cell differentiation.
THE MAMMALIAN EMBRYO IN RELATION TO ITS
ENVIRONMENT
Early Cleavage of the
Mouse Egg, in vitro
D. G. Whittingkam
The development of techniques for
the cultivation of mouse embryos is
providing the foundation for what
should be an effective experimental
approach to the study of the pre-
implantational stages of pregnancy
in the mammal.
The mouse zygote may be cultured
to the early 2-cell stage in a simple,
chemically defined medium and from
the late 2-cell to the blastocyst stage
under similar conditions. However,
the zygote will not develop straight
through to the blastocyst in vitro
unless it is exposed to the environ-
ment of the fallopian tube for a short
period of time between the first and
second cleavage divisions. The energy
requirements of the early mammalian
embryo appear to be demanding. The
first cleavage division in vitro is
dependent upon pyruvate or oxaloace-
tate being present in the basic
medium. It will not occur with lac-
tate, phosphoenol-pyruvate, or glu-
cose, or in the absence of an energy
source. Studies with isotopically
labeled lactate and pyruvate have
shown that the ability of pyruvate
but not lactate to maintain the first
cleavage division is probably attrib-
utable to its greater utilization.
If follicular (cumulus) cells are
placed in the medium with 1-cell
embryos, the first cleavage division
occurs when lactate, phosphoenol-
pyruvate, or glucose is present as the
only energy source, Such evidence
indicates that the follicular cells are
able to metabolize these substrates
and provide the zygote with the
necessary energy substrate for cleav-
age. In addition, evidence that utiliz-
able energy sources, such as pyruvate,
are secreted by the mouse fallopian
tube has been obtained from studies
with ovarian bursal fluid and organ
cultures of the ampullary region of
the fallopian tube. A high percentage
(69.5%) of zygotes devoid of cumu-
lus cells will cleave to the 2-cell stage
in centrifuged ovarian bursal fluid.
Fluorometric methods have shown
pyruvate to be present in this fluid
at concentrations near optimum for
development in vitro (10-4 M to 10-3
M) . Zygotes without cumulus cells
will also develop within the ampul-
lary region of the fallopian tube
grown in organ culture. These results
suggest that pyruvate, an essential
energy requirement for first cleav-
age, is made available to the early
mouse embryo, in vivo, from two
sources: the cumulus cells and the
fallopian tube epithelium. Thus, the
maternal environment plays a vital
role in providing the necessary
energy requirements for the early
stages of embryonic development in
the mouse.
Implantation in the Rabbit
Bent G. Boving
Rabbit egg transport and attach-
ment to the uterus continue as two
principal lines of investigation, but
there have been significant changes
in technique and approach. The hy-
DEPARTMENT OF EMBRYOLOGY
627
pothesis that equidistant spacing of
blastocysts in the uterus depends on
uterine distention by blastocysts was
thought to deserve experimental
checking in spite of a secure founda-
tion of statistically analyzed normal
material. It was expected that the
previous quantitative determination
of the normal location for each indi-
vidual blastocyst could be used for a
statistical judgment of whether blas-
tocysts were displaced by an intra-
uterine device about the diameter of a
blastocyst. That approach has been
tried and abandoned.
A second approach, observation of
waves of uterine contraction by time-
lapse photography, began as a means
of recording uterine activity at im-
plantation time just before the uteri
for the preceding quantitative study
were excised for measurement of the
resulting blastocyst locations. A sim-
ple but significant technical improve-
ment, a ring of lights encircling the
camera lens, was discovered to reflect
from the wet cylindrical uterus as a
pair of bright lines parallel to the
uterine axis except where the surface
of the uterus is not perpendicular to
the line of sight. Most important, the
lines indent, merge, or disappear
wherever there is a local uterine con-
traction. Where the uterus bulges
between contractions, there is a
bright spot or small elongated ring.
Where the uterus is distended, as by
a blastocyst within ("dome," Plate
7), there is a large circle of lights.
The progress of waves of contraction
can be followed easily during projec-
tion at about 20 times the natural
speed and can even be measured with
some precision. Waves of contrac-
tion clearly originate from the uterus
where it is distended by a blastocyst
and from the tubal and cervical ends
as well. Moreover, it is equally
obvious that contraction waves travel
along the uterus in both directions
with equal facility (Plate 7), con-
tradicting the former attribution of
rabbit blastocyst spacing to contrac-
tion waves "milking" uterine con-
tents from the tube to the cervix and
then relaxing to permit reflux back
up through the uterus.
A third and very recent approach
superimposes experimentation on the
preceding observation procedure and
shifts from implantation time, when
equidistant spacing has already been
achieved, to a day earlier (6 days
after mating) when the spacing
mechanism is in full swing. Glass
beads were placed in the uterus —
through the cervix after vaginal
incision, to avoid injuring the uterus.
Beads were 4.5, 5.5, and 6.7 mm in
diameter and were inserted in that
sequence. Waves of contraction origi-
nated at each bead and from both
ends of the uterus, and moved the
two smaller beads back and forth.
By the end of the experiment, the
large bead had moved less than its
own diameter, whereas the two
smaller beads had been repelled to
the tubal end of the uterus. A most
revealing observation was that con-
traction waves from the large bead
and the small bead arrived at the
middle bead simultaneously; each
cancelled the propulsive effect of the
other.
This method of recording waves of
uterine contraction did confirm the
hypothesis that equidistant spacing
of blastocysts depends on the disten-
tion of the uterus by the blastocysts.
In addition, three points concern-
ing the experiment itself deserve
comment. First, the experiment has
the advantage that a greater or lesser
effect, according to diameter, is
shown by a single uterine horn. It is
therefore a much "cleaner" experi-
ment than the seemingly more ele-
gant statistical approach, which,
while using intrauterine devices of
constant size, unavoidably placed
them in uteri of different diameter
628
CARNEGIE INSTITUTION
and so actually proceeded from an
inconstant stimulus (distention). The
acute experiment also avoids extra-
neous effects on spacing from uterine
incision, sutures, healing', adhesions,
etc. Second, in spite of complete
agreement with expectations, this
experiment and all bead experiments
tell what the uterus does with beads,
not blastocysts. The biological valid-
ity derives from agreement with
observations of uterine reactions to
normal blastocysts, the second ap-
proach. Third, there may be a rela-
tion to a recent statistic from over
twenty thousand women in whom
small intrauterine devices proved less
effective than large ones in prevent-
ing pregnancy. Since uterine con-
traction and discomfort may attend
undue distention, one must forego the
obvious suggestion of using larger
devices exclusively, but choosing
IUDs for size and perhaps shape
appropriate to the individual uterine
lumen does seem reasonable — fitting
them as we now fit diaphragms. In
addition, the extraordinary ability of
the uterus to hypertrophy in response
to distention (pregnancy) suggests
possible value in periodic checks of
the fit of the IUD.
More intrauterine "pingpong ex-
periments" are planned to try to
determine (1) how fine a size dis-
crimination the uterus can make, (2)
what size bead is required to block
a contraction wave, (3) the largest
bead that can be moved by a uterine
contraction wave, and (4) the small-
est bead that can stimulate a contrac-
tion wave.
The second line of study, egg
attachment to uterus, continues to
explore the chemical and mechanical
aspects of trophoblast adhesion to
and penetration of uterine epithe-
lium. The experiments on trophoblast
knob extrusion mentioned in the past
two reports have been summarized
in a motion picture. The projected
micromanipulator studies on the
interior of the living uterus have
been deferred. Priority has been
given to examining the intrauterine
environment, partly for guidance in
providing physiological conditions
for the tissues exposed in the micro-
manipulator experiments, but mainly
for its own implications. It may be
recalled that trophoblast adhesion
and penetration were attributed to an
epivascular pB. rise. The rise is con-
sidered to result from residues
of alkaline carbonate left when
bicarbonate passes from the blas-
tocyst and yields carbonic acid which
loses carbon dioxide to the maternal
circulation. This reaction is promoted
by endometrial carbonic anhydrase
whose activity, in turn, has been aug-
mented by progesterone. The evidence
was consistent but indirect. Direct
confirmation has now been obtained.
Continuous records were made from
an expanded-scale pH meter whose
electrodes had been placed in uteri
near the time of implantation. When
blastocysts were ruptured so their
fluid contents reached the electrode,
there was an instant of instability, a
rapid fall of pH, and then a gradual
rise to a pll higher than that of the
previously empty uterus. The rate
and pH level of the rise were the
same as when the same segment of
uterus was injected with 0.2 ml
sodium carbonate solution (9.3
gm/L) that had been charged with
carbon dioxide (Fig. 23). In the
second instance, the pH level was
lower because its charge of carbon
dioxide was heavier than that of the
blastocyst. In nonpregnant control
tests, the pll drop was similar, but
the recovery was slower and to a
lower level. In general, our data are
in agreement with published average
values of 7.4-7.5 for normal intra-
uterine pH; however, minor varia-
tions have been found from one place
to another and from one time to
DEPARTMENT OF EMBRYOLOGY
629
x
7.7
7.61
7.5
7.4
731
7.2
71
7.0
® = 0.2ml Na2C03 + C02 solution
® = Blastocyst fluid
10 15
minutes
20
25
Fig. 23. The change of intrauterine pH with time is shown: (A) after injecting sodium carbonate
solution charged with carbon dioxide, and (B) after an adjacent blastocyst was ruptured.
another in the same uterus, and from
animal to animal, in addition to the
variations that occur when an elec-
trode stretches the uterus. Not only
inadvertent artifacts but a com-
parable effect by a blastocyst's
stretching of the uterus are in ques-
tion. Studies have been extended to
the several days before and after
implantation. The techniques are
being refined and standardized, and
oxygen tension measurements are
about to be added. Just as carbon
dioxide discharge to vessels promotes
attachment there, so the attached
trophoblast may be expected to find
at the same locus a particularly con-
centrated source of oxygen available
for passage in the other direction.
Other reasons for concentrating
interest on circulation and gas ex-
change are more theoretical and will
be explained in the following section.
Rabbit egg attachment to the
uterus has been studied not merely
to understand a circumscribed field
of embryological and possibly clinical
interest but to examine and experi-
ment with scientific method. If
science encompasses not only dis-
covering the facts of nature but also
constructing from them some scien-
tific understanding, then it is per-
tinent to consider not only how facts
are established but also whether
they are chosen and related so that
they extend understanding of na-
ture. Description and experimenta-
tion have been well tested by
scientists for discovering facts. Yet
in constructing systematic under-
standing, description and experimen-
tation have been left largely to
supposedly intuitive philosophers and
mathematicians, and many scientists
may have rested unduly content with
intuition slipped in behind a tradi-
tion that unbiased understanding is
best reached by letting facts accumu-
late until they speak for themselves.
But the listening intuitions will vary
among persons and decades, and the
facts accumulated at time of listen-
ing may add up to an incomplete and
630
CARNEGIE INSTITUTION
consequently biased story. The listen-
ing approach, congenial to early,
descriptive stages of exploration,
proves insufficient for more advanced,
primarily experimental study where
the conversation with nature most
profitably may be directed through
aggressive questioning and where
failures of such direction let ever
more rapidly accumulating facts
yield noise rather than comprehen-
sion. Practical causes and general
consequences have been especially
well stated by Forscher (Science
142:339, 1963)'. Theoretical problems
in the egg attachment study and
related fields offer a case history with
interesting implications.
Why worry whether the right
questions are being asked, so long as
answers keep coming? It has been
more than a century since compara-
tive anatomy and embryology derived
the generalization of evolution, in
small part, at least, from observa-
tions that various species do different
things with homologous structures.
The fact that various mammals do
different things with similar and
presumably homologous hormones
appears to offer a comparable oppor-
tunity, but the variations do not fol-
low phylogeny consistently, and no
other broad explanation has been
achieved. Let us conceive that the
straightforward comparative, one
agent-one function approach may be
inappropriate for physiology. For
example, a comparative understand-
ing of respiratory physiology would
hardly be obtained by comparing gills
and homologous structures, such as
ears and parathyroids. On the other
hand, recognizing the process of
respiratory exchange being served by
primarily vascular mechanisms in
the lung, skin, or placenta of various
species not only permits interesting
and even quantitative comparisons
but points out that physiology makes
better sense when dealing not with
homologues but with analogues and
units such as mechanisms and proc-
esses. Accordingly, there is some
importance in defining the objective
as understanding the process of
implantation and in seeking it first
in a single species rather than
through comparisons of homologous
aspects in several species.
How should one explore a process
as a whole? If a process is a chain
of mechanisms, then the main prob-
lem is to avoid missing any. To that
end, prerequisites and consequences
of each mechanism were sought until
a connected sequence of causes and
effects could be traced. The practice
of such sequential tracing of cause
and effect led to the realization that
it may waste effort to establish a
fact or statistic with greater security
than necessary to draw and test an
inference. Testing by implication
rather than just replication not only
achieves economy of effort but also
checks reasoning and technique.
Those advantages are especially
appropriate for the solitary investiga-
tor who uses a variety of techniques
and invents new ones, and whose
reasoning is neither continually
buffeted by collegial criticism nor
confined within safely conventional
lines.
A second means for avoiding major
omissions was systematic considera-
tion of all major categories presum-
ably involved in the natural situa-
tion: mechanical and chemical factors
of both fetal and maternal systems
and the interplay among all of them.
A diagram was drawn to visualize
how mechanisms of rabbit implanta-
tion relate to the major categories
and to time. The picture became very
complex, but an abstraction was
derived that proved to have some
heuristic interest and usefulness
(Fig. 24): (1) It points out that
interdependence of structure and
function is not simple, and that, there-
DEPARTMENT OF EMBRYOLOGY
631
<_v\<_\ovft\ca\ fcrc\\M^o\o<^j
> ^ACHNHTM MJIATOMY
cYvfettUCoA fettO_T>f o\o<^
[FLACilNITAL I2C€IHIAIN1__
_____
Fig. 24. This diagram suggests: (1) that the process of implantation and placenta formation in-
volves interaction of chemical and mechanical factors of both fetal and maternal systems and the
interplay among them; (2) that the fetal development or change with time is supported by mater-
nal structures and functions that also change with time; and (3) that the joint chemical function,
namely placental exchange, takes place against a background of joint structure, namely placen-
tal anatomy, which includes mechanical function.
fore, it is prudent to begin investiga-
tions by descriptive studies that avoid
altering the natural situation, rather
than by experimentations that may
influence unappreciated factors in
unrecognized ways. It is also clear
that several disciplines must be
brought to bear for a reasonably com-
plete explanation of implantation and
placentation. (2) It generates appre-
ciation for the completeness and
balance of the classical Flynn and
Mossman definitions of a placenta as
an apposition or fusion of fetal mem-
branes and maternal tissues for
physiological exchange. (3) It illus-
trates that the half-century-old egg-
versus-uterus controversy stems from
complementary omissions by experi-
ments with inadequate reference to
the natural situation. Brachet cul-
tured eggs without the uterus and
concluded that the uterus was
unnecessary, whereas Loeb induced
decidual reactions in uteri without
ova and concluded that the egg had
no specific role in forming the ma-
ternal part of the placenta. Beyond
that, however, we can see the less
obvious and still current error of
Loeb's basic assumption that deciduo-
mas are equivalent to the maternal
part of the placenta, when in fact
they lack the principal functioning
part: the maternal blood circulating
through the intervillous space. (4)
If, as the diagram indicates, chemical
exchange is a principal function of
the placenta, then rather special
attention seems due the circulation
that so especially serves it, as well as
the substances exchanged. The verj
close association between trophoblast
and maternal vessels in the earliest
implantations available for study
from man and macaques is consistent
with the rabbit trophoblast's "aim-
ing' ' for maternal vessels when it
632
CARNEGIE INSTITUTION
adheres to and penetrates uterine
epithelium. (5) Taking- the imme-
diately preceding- considerations back
to just before implantation, we may
make a provocative interspecies com-
parison. The rabbit endometrium
acquires increased carbonic anhy-
drase activity (which facilitates car-
bon dioxide discharge from egg to
maternal capillaries, leading to at-
tachment through steps reported
before). The human endometrium,
on the other hand, acquires capillar-
ies in the previously avascular,
subepithelial region (Bartelmez's
zone I). Both changes result from
progesterone stimulation. Both
changes may be presumed to promote
gas exchange across uterine epithe-
lium. Both changes precede and very
probably promote implantation. But
in the one case we recognize the
action of progesterone by observing
an anatomical change and in the
other case, a biochemical change,
realizing, of course, that the two may
or may not involve the same, as
yet unknown, underlying molecular
events.
What manner of biological relation-
ship can it be? There is no question
of homology. It has some resemblance
to analogy, but with one effector
mechanical and the other chemical,
it is beyond the usual biological mean-
ing of the term. Yet the identities of
stimuli and consequences are there to
be pondered. Thoughts go to coinci-
dence, convergent evolution, and
mimicry — even to the brink of tele-
ology when one recalls that, in the
few species studied, eggs graduate
from dependence on glycolysis and
suddenly acquire a greatly increased
capacity for aerobic metabolism at
about the time that the maternal sys-
tem improves the uterine gas ex-
change, and implantation proceeds.
With further study of developing
systems as processes, their curious
and flexible ability to achieve and
even anticipate what is "necessary"
seems likely to become more under-
standable in chemical and mechani-
cal terms.
Anatomy and Physiology
of the Placenta
Radioangiography of
Placental Circulation
E. M. Ramsey, M. W. Donner, and
C. B. Martin, Jr.
The opening, in January, 1967, of
the splendidly equipped and staffed
Radiologic Research Laboratory at
the Johns Hopkins Hospital has
added a fine new physical dimension
to the fruitful collaboration between
the Department of Embryology and
the Johns Hopkins Department of
Radiology. In this eighth year of
association Ramsey, Donner, and
Martin have continued their investi-
gation of circulation in the fetal
placenta of the rhesus monkey and
commenced study of two related
aspects of the general subject of
placental circulation.
Fetal placental circulation. The
two avenues of approach to the fetal
placental vascular bed described in
last year's annual report have been
employed in this year's studies with
particular attention to refining the
surgical techniques involved so as to
minimize shock to the fetus. This has
been carried to the point where such
operations can now be regarded as
tools, rather than as ends in them-
selves, and it is proposed to employ
them in the future for study of fetal
circulation under experimental con-
ditions induced by anesthesia and
other drugs.
The observations made while ex-
perience with intrauterine surgery
was being accumulated have provided
information on the configuration of
the fetal cotyledon which is particu-
larly significant in the light of
Martin's anatomical studies of this
DEPARTMENT OF EMBRYOLOGY 633
structure. The subject is reviewed as inhalation anesthesia (N02 + 02)
a whole elsewhere in this report and skeletal muscle relaxants and
(pp. 62-66). then with conventional pentobarbital
Anesthesia. The possibility that the sodium. The animal's blood pressure
anesthesia employed during studies was continuously monitored, and de-
of uterine motility and placental cir- terminations of blood gases and pH
culation in the monkey might be a were made at intervals. Injections of
source of artifact has long been a radiopaque material were made in
matter of concern. Although most of both uterine relaxation and contrac-
tile applicable procedures can be tion with each anesthetic agent. The
carried out in human patients with- number and character of the streams
out the use of anesthesia, it is not of maternal blood entering the inter-
practical to use unanesthetized mon- villous space ("spurts") were taken
keys because of handling problems, as criteria upon which the different
Throughout our studies pentobarbital types of anesthesia were evaluated,
sodium has been the anesthetic agent Study of the serial X rays made
employed, administered intravenously during the test periods showed that
for induction and intramuscularly the variation in number of spurts
for maintenance of as light a degree from one anesthetic agent to the
of anesthesia as compatible with quiet other, under comparable conditions
relaxation. Until now we have felt of myometrial activity, was minimal,
that there were so many variables always well within the physiological
and imponderables associated with range previously described. The uni-
the experiments themselves that it formity in the results can be taken
was unwise to introduce another one as reassuring evidence that previous
by varying the anesthesia also. This use of pentobarbital sodium did not
year, with base lines fixed in the affect the experimental results. In-
major experimental areas, the time deed, it seems both safe and reason-
seemed ripe to tackle the anesthesia able to plan to use pentobarbital
problem and lay at rest, if possible, sodium in future work, since it is
the concern about its artificial effect, simpler to administer than an inhala-
We have had the great good for- tion anesthesia,
tune to have the collaboration in this Hypotensive and hypertensive
study of John Bonica, Professor of states. A question of theoretical in-
Anesthesiology at the University of terest, and potential clinical impor-
Washington School of Medicine in tance, arises from the observation in
Seattle. Professor Bonica spent two both monkeys and human patients
weeks in Baltimore in the spring of that strong uterine contractions cur-
1967 bringing with him his associate tail or even prevent inflow of mater-
Marlene Eng and enlisting the enthu- nal blood to the placenta. One won-
siastic cooperation of Professor ders whether raising maternal blood
Donald W. Benson, who generously pressure might overcome the obstruc-
placed the staff and facilities of the tion caused by myometrial constric-
Johns Hopkins Department of Anes- tion of uterine arteries, and con-
thesiology at our disposal. versely whether dropping the blood
The now adequately standardized pressure below some critical level
technique of arteriography of mater- might decrease inflow in the absence
nal placental circulation was carried of uterine contraction. Bonica has
out in the test animals, each of which been particularly anxious to investi-
served as its own control. All of the gate these matters and, combining
animals were studied first with basal his interest and our own with the op-
634
CARNEGIE INSTITUTION
portunities provided by his visit to
Baltimore and the facilities and col-
laboration available here, a start was
made toward answering the key ques-
tions.
Under nitrous oxide-oxygen anes-
thesia, with normotensive baselines
established, animals were given drugs
which raised and lowered systemic
blood pressure as desired, with or
without effect upon myometrial activ-
ity. To date the effects of hypotension
induced by Arfonad (trimethaphan
camphorsulphonate) and of hyper-
tension with Vasoxyl (methoxamine
hydrochloride) have been studied.
Syntocinon, used to induce uterine
contractions, has a transient hypoten-
sive effect. Ether relaxes the uterus
when desired, without significant
effect upon systemic blood pressure.
Halothane depresses both myometrial
activity and blood pressure. As in the
anesthesia studies, counts and quali-
tative observations of arterial entries
to the intervillous space were used as
criteria of results.
Preliminary results indicate that
markedly lowered blood pressure
reduces placental perfusion and that
this accentuates the effect of myo-
metrial contractility. During uterine
relaxation arterial entries are not
significantly decreased in number but
the "spurts" are smaller and slower
in appearing. With the hypertensive
drug employed (Vasoxyl) even a
very much elevated systemic blood
pressure did not overcome the stop-
page of arterial inflow caused by a
strong contraction. This may be the
result of the vasoconstrictive action
of this particular drug. Experiments
in other species (sheep and dogs)
have shown that sympathomimetic
drugs produce vasoconstriction of the
uteroplacental arteries. The matter
will be explored further in our
studies by the use of other hyperten-
sive agents. It is also planned to
follow up the significant clue pro-
vided by the Vasoxyl experiment and
investigate the action of drugs upon
uterine spiral arteries. The "inter-
mittent functioning" of the arteries
has been established, though the
causation of the phenomenon is still
unexplained.
Gross Anatomy of the Placenta of
Rhesus Monkeys
C. B. Martin, Jr., E. M. Ramsey, and
M. W. Donner
The use of the rhesus monkey as an
experimental model in primate repro-
duction was pioneered in this country
at the Carnegie Department of Em-
bryology. In recent years, this species
has become very popular for studies
of placental and fetal physiology.
Despite some obvious differences
between the placentas of Macaca
mulatta and Homo (number of pla-
cental discs, depth of implantation,
etc.), many investigators seem to
have assumed that these placentas
are structurally identical. The early
stages of placental development and
the fine structure and histochemistry
of the macaque placenta have been
described in detail. There has been
no systematic description of the
architecture of the mature rhesus
placenta, however, and we have
therefore examined it by several
methods.
Some placentas were studied by the
classical anatomical techniques of
injection and serial sectioning. Others
were dissected after partial trypsin
digestion. In still other cases, we
made corrosion casts of both ma-
ternal and fetal placental vasculature.
Placental anatomy in vivo was stud-
ied by means of radioangiograms of
the maternal and fetal placental
circulations.
The rhesus monkey placenta is
usually bidiscoid, but single-disc
placentas have been encountered in
about 15% of pregnancies in the
DEPARTMENT OF EMBRYOLOGY 635
Carnegie colony. The two discs are corrosion cast shown in Plate 8(A)
located on the anterior and posterior and is illustrated in vivo by the fetal
walls of the uterus (when it is in the placental angiogram, Plate 8(B).
anatomical position). The umbilical Cotyledons are more apparent on
cord inserts near the center of one the maternal surface of the macaque
of the discs, usually the larger of the placenta after delivery than in the
two, which is conventionally called human placenta. Each placental disc
the primary placenta. The two pla- commonly contains 8-12 cotyledons
centas are connected by one or more but the variation here is considerable,
sets of interplacental vessels which Single-disc placentas have no more
run in the chorion. cotyledons than one disc of a double
The arrangement of arteries and placenta, but the individual cotyle-
veins on the chorionic plate resembles dons tend to be larger in the single
that in human placentas. The inter- placenta.
placental vessels are fairly direct The stubs of spiral arteries may be
continuations of major branches of identified easily on the maternal sur-
the umbilical vessels, although they face of the placenta after delivery;
give off side branches in their course Plate 8(C). A single spiral artery
over the primary placenta. enters each cotyledon near its center.
The rhesus placenta, like the Occasionally two or three spiral
human placenta, is of the villous arteries appear to enter a single
hemochorial type. It contains propor- cotyledon. In these instances, further
tionally more main-stem villi than dissection usually reveals two or
does the human placenta, but the three subcotyledons, each with a
monkey stems are smaller and less single spiral artery. This relationship
complexly branched. The villi of the between the cotyledons and the en-
rhesus placenta are grouped together tries of the maternal arteries has
into fairly prominent cotyledons, been confirmed in vivo by simultane-
Most of the stem villi of a single ous radioangiography of the maternal
cotyledon are supplied by branches and fetal placental circulations; Plate
of the same artery on the fetal sur- 8(B). The relationship is also appar-
face of the placenta. Both in its ent in the photograph of the doubly
pattern of branching and its vascu- injected corrosion specimen,
larization, the main-stem villus of the Large veins may also be identified
rhesus placenta appears to corre- on the endometrial surface of the
spond to the second-order branch of placenta. It appears that these
the human main-stem villus. usually arise at the edges of the
Each cotyledon is composed of a cotyledon,
few large or medium-sized stem villi These studies indicate that,
and many smaller ones. The larger although there are many similarities
stems are located toward the center in structure between rhesus and
of the cotyledon, and the smaller human placentas, there are also some
stems lie toward the periphery. In differences. These interspecific differ-
areas between cotyledons the stems ences must be taken into account
are quite small and simply branched, when data obtained from rhesus mon-
Peripheral branches of the stem villi keys are transferred to human preg-
interdigitate extensively within coty- nancy. They do not, however, appear
ledons but to a much lesser extent to invalidate the use of the rhesus
between adjacent cotyledons. Each monkey as an experimental model in
cotyledon is thus a fairly discrete the study of most aspects of human
unit. This is demonstrated in the uteroplacental physiology.
636
CARNEGIE INSTITUTION
Uterine Activity Studies
C. B. Martin, Jr.
The studios of uterine contraction
patterns in nonpregnant monkeys
described in previous Year Books
have been continued and expanded to
include long-term continuous obser-
vations of patterns in unanesthetized,
restrained animals.
The observations of uterine activ-
ity reported previously, which were
based on acute experiments, provided
a sharply limited sampling of uterine
activity. Following anesthetization
and insertion of the recording cathe-
ter, uterine activity was monitored
for periods of hours only. In addition,
the anesthetic agent (usually pento-
barbital sodium) and the uterine
stimulation attendant on placing the
recording catheter were recognized
as possible sources of artifact in
these experiments. The chronic re-
cording technique was adopted in
order to avoid these pitfalls.
To obtain the chronic records, the
recording catheters were inserted and
fixed in place while the animal was
anesthetized. In some animals, the
catheter was inserted per vaginam
and sutured to the uterine cervix. In
other animals, the catheter was in-
serted through the fundamental myo-
metrium at laparotomy. These latter
catheters were passed downward
through the cervix and out the
vagina, and were anchored by suture
to a plastic button on the peritoneal
surface of the uterus. The animals
were then allowed to recover from
the anesthesia and were placed in
restraining chairs. Either of these
techniques permits continuous record-
ing of uterine contraction patterns
through entire menstrual cycles. The
presence or absence of ovulation in
these animals was established by
means of serial vaginal smears, and
in some instances was confirmed by
laparotomy.
Contraction patterns in ovulatory
cycles. The continuous recordings
have clearly shown the changing pat-
terns of uterine contractions during
the ovulatory cycle. The contractions
increase in intensity and particu-
larly in frequency during the follicu-
lar phase. By the time of ovulation,
contractions of 100-150 mm Hg
intensity are occurring at a fre-
quency of 2-4 per minute. This pat-
tern persists with little change until
about the third postovulatory day,
when there is a notable decrease in
the frequency and intensity of con-
tractions. Quiet periods, which may
be recorded at any time of the cycle,
become much more prominent about
this time and may last for several
hours, interrupted only by infrequent
isolated contractions or occasional
bursts of contractions. Very slow
contractions of 30-40 mm intensity
lasting 5 minutes or more, often with
more rapid peaks superimposed upon
them, also occur during this time.
The contractions increase in intensity
and frequency during the premen-
strual week until, with menstruation,
contractions of up to 300 mm Hg
intensity occur at 1- and 2-minute
intervals. These uterine contraction
patterns of rhesus monkeys resemble
closely those recorded in normal
women, except for the greater con-
traction pressure observed in the
monkeys.
Anovulatory cycles. The progres-
sive changes in uterine contraction
patterns described above do not
occur in anovulatory cycles. In them,
the preovulatory pattern persists un-
til the onset of bleeding. Quiet peri-
ods may occur in anovulatory cycles,
but they are not as prolonged as those
following ovulation. When quiet
periods do occur, they alternate with
periods of fairly intense and rhyth-
mic uterine contractions. No correla-
tion is yet apparent between the
quiet periods and fluctuations in
DEPARTMENT OF EMBRYOLOGY
637
estrogenic activity as reflected in the
vaginal cornification count.
Short-term hour-to-hour variations
in uterine contraction patterns occur
in both ovulatory and anovulatory
cycles in these unanesthetized ani-
mals. At any stage of the cycle, the
uterus may suddenly stop contract-
ing, and, equally abruptly, resume
contractions. Changes in frequency
or intensity of contractions, or both,
have been observed. Contractions
may suddenly become coupled in
groups of 2 or 3, where before there
had been isolated contractions. The
shift from one contraction pattern to
another may be abrupt or gradual.
These variations are occasionally so
pronounced as to make it quite diffi-
cult to identify the predominating
contraction pattern. The changes
sometimes occur in association with
identifiable stimuli, as for example
with alarm or arousal when the
observer enters the room. More often,
however, no stimulus can be recog-
nized. The short-term nature of these
changes suggests that autonomic
nervous system activity, rather than
changes in the level of ovarian
steroids, is the mechanism by which
the changes are produced.
Artificial cycles. To study the con-
tribution of the ovarian steroids to
the cyclic changes in uterine activity,
artificial menstrual cycles have been
produced in 2 castrate rhesus fe-
males. Cycles have been produced
with estrogen alone, with estrogen
followed by estrogen plus progester-
one in varying combinations, and by
estrogen plus progesterone through-
out the cycle.
Estrogen alone produces the pre-
ovulatory contraction pattern. When
estrogen administration is continued
for more than 2 weeks without an
increase in dose, the contractions
begin to decrease in intensity. The
preovulatory contraction pattern per-
sists until the onset of bleeding after
an estrogen-only cycle, just as it does
in spontaneous anovulatory cycles.
When progesterone is added to the
estrogen at mid-cycle, there is a
decrease in frequency and intensity
of contractions qualitatively similar
to that which occurs after ovulation.
One must use some 25 mg of proges-
terone per day (in an animal primed
with 1-2 mg of conjugated equine
estrogens per day) to obtain uterine
quieting equivalent to that seen on
the third and fourth days after ovu-
lation. Uterine contractions cannot be
completely suppressed, however, even
with 50 mg of progesterone per day.
Progesterone withdrawal is followed
by an increase in contractions similar
to that preceding menstruation. The
uterine contraction pattern at the
onset of progesterone-withdrawal
bleeding is much more rhythmic and
intense than that at the onset of
estrogen-withdrawal bleeding.
Studies with IUDs. The effect of
intrauterine contraceptive devices
( IUDs ) has been studied in five rhesus
monkeys. Four animals were studied
by the acute recording technique, and
two of these animals had ovulatory
cycles with the IUD in place. One
animal has been studied, by means of
the continuous recording technique,
through four ovulatory cycles with
the IUD in place. Significant altera-
tions in the uterine contraction pat-
terns were observed in all animals.
In the animals studied by the acute
technique, the normal contraction
patterns were replaced at all phases
of the cycle by rhythmic, laborlike
contractions occurring at 1- to 3-
minute intervals. Brief periods of
rapid, less intense contraction were
observed at mid-cycle in the two
ovulatory animals, and the frequency
of contraction decreased slightly dur-
ing the luteal phase in these animals.
Quiet periods were observed during
the luteal phase in one of the ovula-
tory animals, but these quiet periods
638
CARNEGIE INSTITUTION
wore not as prolonged as during the
control cycle.
Contraction patterns during the
cycles following IUD insertion were
much more variable in the animals
studied by the continuous technique
than in the four animals reported
above. The laborlike contraction pat-
tern has been prominent during the
preovulatory and premenstrual por-
tions of the cycle, but even at these
times the contractions have not been
as uniform as in the animals studied
with the acute method. On the third
and fourth postovulatory days, when
implantation would ordinarily occur,
uterine contractions diminished con-
siderably in intensity in this animal,
and quiet periods appeared, though
not to the degree observed during the
control cycle. The quiet periods were
shorter, and during the active periods
the contractions were more rhythmic
and intense. Uterine activity nor-
mally begins to increase about a week
after ovulation. In the IUD cycles,
the build-up of contractions began
on the fourth or fifth postovulatory
day.
Although the number of ovulatory
IUD cycles studied so far is small, it
is apparent that the IUD produces
a more rhythmic and laborlike (and
probably more expulsive) uterine
contraction pattern than is normally
present at most times of the men-
strual cycle. The increased uterine
activity observed at the time of
implantation is especially note-
worthy, for it is just at this time that
the device must exert its contracep-
tive effect.
Monkey colony. The colony has
been maintained at the usual level of
3 males and 31 females. The gratify-
ing pregnancy rate of 63.3% is testi-
mony to the health and well-being of
the animals and to the skilled and
devoted care given them by James
Abbott and William Cleary and the
staff. There was no illness in the
colony throughout the entire year. It
remains tuberculin negative.
Besides its use in the placental
circulation studies, the colony was
made available to two extramural
colleagues. Ralph Wynn of the New
York Downstate Medical Center ob-
tained specimens of monkey placenta
for electron microscopic studies of
trophoblast. Roberto Narbaitz of the
University of Maryland School of
Medicine was given all the monkey
infants for histochemical studies of
tooth bud development. In addition,
several of the pregnant mothers were
subjected to X-irradiation two weeks
prior to cesarean section performed
in connection with other experiments
so that Narbaitz could study the
effect of this treatment upon fetal
gonads.
THE COLLECTION OF HUMAN EMBRYOS
B. G. Boving and E. M. Ramsey
In the year covered by this report,
E. M. Ramsey examined 20 specimens
sent by 13 physicians from four
states. Of these specimens 11 were
discarded as of no research or mu-
seum value at the end of 3 months
after reporting to the donor and in
the absence of instructions to the
contrary. Nine specimens had suffi-
cient value to justify preservation.
The percentage of specimens merit-
ing preservation is the highest since
1958 and the second highest since
1951.
The embryo collection was reviewed
in the past year to assemble all
specimens of embryos from mothers
with a history of rubella or exposure
to rubella during pregnancy. All of
the 24 specimens found had been
DEPARTMENT OF EMBRYOLOGY
639
carefully studied grossly; seven of
them had been sectioned and the sec-
tions reviewed for evidence of devel-
opmental anomaly. Since there is no
longer anyone in the Department
equipped to make an expert eval-
uation of such material, all 24
specimens, with several hundred
microscopic preparations and tran-
scripts of pertinent documents, were
sent to the laboratory of Lorenz
Zimmerman, Chief of the Division of
Ophthalmic Pathology at the Armed
Forces Institute of Pathology where
a study of postrubella ophthalmic
lesions of the fetus is in progress.
The Early Development of the
Nasal Pit in Staged Human
Embryos
R. O'Rahilly
The nasal fields appear as bilateral
ectodermal thickenings near the ros-
tral neuropore during stage 12 (26
postovulatory days). The nasal disc
is better denned by stage 13, is
flattened or even concave at stage 14,
and presents a pit by stage 15. The
nasal pit is lined by terminal bars,
underlying which is a superficial
mitotic zone, and the epithelium rests
on a coextensively invaginating base-
ment membrane. The arrangement is
thus comparable to that of the otic
disc and pit (stages 9 and 10), the
optic primordium and sulcus (stage
10), the retinal disc and optic cup
(stages 13 and 14) , and the lens disc
and pit (stages 13 and 14). The
nasal fin and the vomeronasal groove
are found during stage 16, and the
inferior conchal swelling at one stage
later. From stage 17 onward, cellu-
lar strands and nerve fibers are seen
leaving the wall of the nasal sac and
the vomeronasal groove, and proceed-
ing to the brain. The cells, which are
comparable to those migrating at
earlier stages from the optic and otic
vesicles, may be termed the "termi-
nal-vomeronasal neural crest." By
stage 18 (44 postovulatory days),
septal, preseptal, paranasal, and
nasal capsular condensations are
present in the surrounding mesen-
chyme.
STAFF ACTIVITIES
Faithful readers of these reports
know that over the past decade the De-
partment of Embryology has played
a major role in the training of in-
vestigators. Members of the staff are
often asked how they arrive at the
proper mixture of postdoctoral fel-
lows, graduate students, and under-
graduates, especially when, in most
years, opportunities can be provided
for only one out of eight or nine
applicants. We try to give considera-
tion to those who, judging by their
past performance, appear likely to
contribute seriously to developmental
biology. It should be noted that the
field of past performance is not
specified, whether biology, chemistry,
medicine, or some other discipline.
Moreover, the status of the applicant,
whether postgraduate, graduate, or
undergraduate is of secondary im-
portance; but other things being
equal, preference is given to the
better prepared applicant. Thus, the
number of undergraduate, students
has remained — and likely will remain
— small. Exception is made during
the summers, when slightly larger
numbers of undergraduates may be
offered opportunities as assistants.
There is a tendency for many of
these students to be drawn from
Johns Hopkins, where several mem-
bers of the staff take an active part
in teaching; but the roster of the past
two summers has included students
from Barnard and Pembroke Col-
640 CARNEGIE INSTITUTION
leges. Drew, Duke, and Howard Uni- Among the symposia and confer-
versities, and the Universities of ences in which various members of
Colorado, Maryland, and Pennsyl- the staff participated during the past
vania. year were the following: Inter-
Our staff's formal teaching activi- national Symposium for Cellular
ties are largely confined to the Johns Chemistry, Otsu, Japan; Internation-
Hopkins Department of Biology, but al Cancer Congress, Tokyo; Confer-
during the year lectures were offered ences on Comparative Aspects of
in other departments of the Univer- Reproductive Failure (Dartmouth
sity as well, among them Biomedical College) , Embryology and Pathology
Engineering, Biophysics, Obstetrics of Congenital Heart Disease (Uni-
and Gynecology, Pathobiology, and versity of Florida Medical Center),
Physiological Chemistry. Etiology and Morphogenesis of Con-
Other activities directed largely genital Heart Disease (National
toward teaching included the partici- Heart Institute) , Fetal Homeostasis
pation of members of the Department (New York Academy of Medicine),
in the Embryology Training Pro- Gene Expression (Neurosciences Re-
gram, Marine Biological Laboratory, search Program), and Obstetrical
Woods Hole, Massachusetts; in the and Placental Physiology (Yale Uni-
Intensive Study Program in the versity).
Neurosciences, held at Boulder, Special lectures included the
Colorado; and service on the Com- Lewis A. Conner Memorial Lecture
mission on Undergraduate Education "Development of Form and Function
in the Biological Sciences. One Staff in the Embryonic Heart: An Experi-
Member offered a series of holiday mental Approach" (American Heart
science lectures for high school Association), and addresses before
students, sponsored jointly by the the Northeastern New York Ob-
A.A.A.S. and the University of Miami, stetrical and Gynecological Society,
Other addresses were presented in Rocky Mountain Section of the So-
the Baltimore City Public Schools ciety for Experimental Biology and
High School Lecture Series on the Medicine, Sinai Hospital, Washington
Biological Sciences, and at the Hospital Center, and the Worcester
Science Talent Institute. Foundation.
A series of lectures was offered for Members of the group took part in
high school teachers at the National meetings of a number of learned
Association of Biology Teachers Re- societies, including, in addition to
gional Seminar, held at Bowdoin those already mentioned, the Ameri-
College, Maine. can Association of Anatomists,
During the year Staff Members, American College of Obstetricians
Fellows, and Visiting Investigators and Gynecologists, American Insti-
presented lectures at a number of tute of Biological Sciences, American
campuses, including City University Society of Biological Chemists, Amer-
of New York (Brooklyn College and ican Society for Cell Biology, Ameri-
Queens College), Cornell University, can Society of Zoologists, Federation
Goucher College, Harvard Medical of American Societies for Experi-
School, Ithaca College, Marquette mental Biology, Society for Develop-
University, and the Universities of mental Biology, Society of General
Delaware, Illinois, Kyoto, Nagoya, Physiologists, Society for Gyneco-
Osaka, Pennsylvania, Texas, Ver- logic Investigation, and the Tissue
mont, and Wisconsin, and Yale Uni- Culture Association,
versity. Advisory and consultative services
DEPARTMENT OF EMBRYOLOGY
641
included membership on the editorial
boards of Developmental Biology,
International Journal of Cancer,
Journal of Embryology and Experi-
mental Morphology, Excerpta Medica
(section on Human Developmental
Biology), and Current Topics in
Developmental Biology.
Members of the staff continued to
serve on the Science Development
Advisory Panel, National Science
Foundation; and the Visiting Com-
mittees of the Departments of Bi-
ology, Massachusetts Institute of
Technology, Western Eeserve Uni-
versity, and the University of Toledo.
Members of the staff also acted in
these capacities: Member of the
Board of Scientific Overseers, Jack-
son Laboratory; Trustee, Marine Bio-
logical Laboratory; Member of the
Board of Scientific Counselors,
National Cancer Institute; and Chair-
man-elect, Developmental Biology
Division, American Society of Zoolo-
gists. A sampling of other posts with-
in professional societies includes the
following: in the American Associa-
tion for the Advancement of Science
Advisory Committee for American
Men of Science, Committeeman-at-
Large for Medical Sciences, Com-
mittee on Science in the Promotion of
Human Welfare, Newcomb Cleveland
Prize Committee; in the American
Association of Anatomists, Repre-
sentative to Division of Medical
Sciences of the National Research
Council, Representative to the Na-
tional Society for Medical Research;
in the American Institute of Biologi-
cal Sciences, Chairman of the Com-
mittee on Laboratory Animal Care,
Member-at-Large of the Governing
Board, Council of Past Presidents.
Seminars. The roster of speakers at
the seminars organized by the De-
partment to serve all those working
in developmental biology in the re-
gion included J. M. Ashworth, Leices-
ter University; Jean-Pierre Chan-
geux, Pasteur Institute; Katsuma
Dan, Tokyo Metropolitan University;
E. J. Furshpan, Harvard Medical
School; Shinya Inoue, University of
Pennsylvania; R. E. Marshall, Na-
tional Heart Institute; B. S. McEwen,
Rockefeller University; Uzi Nur, Uni-
versity of Rochester; Richard Pann-
backer, Washington University; John
Paul, Royal Beatson Memorial Hos-
pital, Glasgow; Robert Perry, Insti-
tute for Cancer Research; and R. L.
Trelstad and Barbara Wright, both
of Massachusetts General Hospital.
BIBLIOGRAPHY
Austin, C. R., see Whittingham, D. G.
Beck, A. J., and F. Beck, The origin of
intra-arterial cells in the pregnant uterus
of the macaque (Macaca mulatto,). Anat.
Record, 158, 111-114, 1967.
Beck, F., see Beck, A. J.
Biggers, J. D., see Schuetz, A. W.; Whit-
tingham, D. B.
Bishop, D. W., W. W. Schrank, A. D.
Musselman, and E. C. Muecke, Testis
sorbitol dehydrogenase (SDH) and activ-
ity changes during induced aspermato-
genesis and cryptorchidism. Federation
Proc, 26, 646, 1967.
Boving, B. G., Some mechanical aspects of
trophoblast penetration of the uterine
epithelium in the rabbit, in Egg Implan-
tation, G. E. W. Wolstenholme, ed., Ciba
Foundation, London, pp. 72-82 and dis-
cussion extending to p. 93, 1966.
Boving, B. G., Tabulated outline of prenatal
human development. Reprinted after cor-
rections from Obstetrics, 13th edition,
J. P. Greenhill, ed., Saunders, Philadel-
phia, 1966, as an appendix in Human
Development, F. Falkner, ed., Saunders,
Philadelphia, pp. 603-619, 1966.
Boving, B. G., Antiexperimentalism. Science,
15k, 1603-1604, 1966.
Boving, B. G., Chemo-mechanics of implan-
tation, in Comparative Aspects of Repro-
ductive Failure, K. Benirschke, ed.,
Springer- Verlag, New York, pp. 142-153,
1967.
642
CARNEGIE INSTITUTION
Boving, B. G., Rabbit trophoblast knobs
extrude svncytiotrophoblast at high pH.
Anat Record. 157, 406, 1967.
Boving, B. G., Proposed standards for ani-
mal care. A letter to the Director, Animal
Health Division. U. S. Department of
Agriculture. BioScicncc, 17, 177-178, 1967.
Brown. D. D., The nucleolus and synthesis
of ribosomal RNA during oogenesis and
embrvogenesis of Xcnopus laevis. Nat.
Cancer Inst Manogr. 23, 297-309, 1966.
Brown, D. D., and J. B. Gurdon, Size dis-
tribution and stability of DNA-like RNA
synthesized during development of anu-
cleolate embryos of Xenopus laevis. J.
Mol Biol.. 19," 399-422, 1966.
Brown, D. D., and E. Littna, Synthesis and
accumulation of DNA-like RNA during
embrvogenesis of Xenopus laevis. J. Mol.
Biol^ 20, 81-94, 1966.
Brown, D. D., and E. Littna, Synthesis and
accumulation of low molecular weight
RNA during embryogenesis of Xenopus
laevis. J. Mol. Biol, 20, 95-112, 1966.
Calm, R. D., see Coon, H. G.
Coon, H. G., Clonal stability and phenotypic
expression of chick cartilage cells in vitro.
Proc. Natl. Acad. Sci. U.S., 55, 66-73,
1966.
Coon, H. G., and R. D. Cahn, Differentiation
in vitro: effects of Sephadex fractions of
chick embryo extract. Science, 153, 1116-
1119, 1966."
Dawid, I. B., see Jungwirth, C; Wolsten-
holme, D. R.
DeHaan, R. L., Development of form in the
embryonic heart: an experimental ap-
proach. Circulation, 35, 821-833, 1967.
Denis, H., Gene expression in amphibian
development. I. Validity of the method
used: interspecific and intrasepcific hy-
bridization between nucleic acids. Proper-
ties of messenger RNA synthesized by
developing embryos. J. Mol. Biol., 22,
269-283, 1966.
Denis, H., Gene expression in amphibian
development. II. Release of the genetic
information in growing embryos. J. Mol.
Biol, 22, 283-304, 1966.
Donner, M. W., and E. M. Ramsey,
Radioangiographische Studien iiber die
Dynamik der Blutzirkulation in der
mutterlichen Plazenta (Experimentelle
Untersuchungen am Rhesusaffen) .
Fortschr. Gebiete Rontgenstrahlen, Band
10U, Heft 6, 796-808, 1966.
Donner, M. W., see also Martin, C. B., Jr.;
Ramsey, E. M.
Ebert, J. D., Developmental interactions at
cellular and molecular levels, in Ontogeny
of Immunity, R. T. Smith, R. A. Good and
P. A. Miescher, eds., U. Florida Press,
Gainesville, Fla., pp. 3-4, 1967.
Ebert, J. D., and M. E. Kaighn, The keys
to change: factors regulating differentia-
tion, in Major Problems in Developmen-
tal Biology, M. Locke, ed., Academic
Press, New York, pp. 29-84, 1966.
Gurdon, J. B., Control of gene activity dur-
ing early development of Xenopus laevis,
in Heritage from Mendel, R. A. Brink,
ed., U. Wisconsin Press, Madison, Wise,
pp. 203-244, 1967.
Jungwirth, C, and I. B. Dawid, Vaccinia
DNA: separation from host cell DNA.
Arch. ges. Virusforsch., 20, 464-468, 1967.
Kaighn, M. E., see Ebert, J. D.
Kaiser, I. H., see Ramsey, E. M.
Kimmel, C. B., The response of lysosomes
in the chick embryo spleen in the graft-
versus-host reaction, in Ontogeny of
Immunity, R. T. Smith, R. A. Good, and
P. A. Miescher, eds., U. Florida Press,
Gainesville, Fla., pp. 103-111, 1967.
Konigsberg, I. R., The application of clonal
techniques to problems of cytodifferen-
tiation, in Ontogeny of Immunity, R. T.
Smith, R. A. Good, and P. A. Miescher,
eds., U. Florida Press, Gainesville, Fla.,
pp. 25-29, 1967.
Martin, C. B., Jr., E. M. Ramsey, and M. W.
Donner, The fetal placental circulation in
rhesus monkeys demonstrated by radio-
angiography. Am. J. Obstet. Gynecol, 95,
943-947, 1966.
Martin, C. B., Jr., see also Ramsey, E. M.
Meyer, R. K., see Sager, D.
Muecke, E. C, see Bishop, D. W.
Musselman, A. D., see Bishop, D. W.
O'Rahilly, R., The early development of the
nasal pit in staged human embryos.
Anat. Record, 157, 380, 1967.
Ramsey, E. M., Circulation in the uterus
and the intervillous space in the primate
placenta. Transcript of the Third Roches-
ter Trophoblast Conference, pp. 6-27,
1965.
Ramsey, E. M., Embryology and develop-
mental defects, in Textbook of Obstetrics
and Gynecology. D. N. Danforth, ed.,
Hoeber Medical Division, Harper & Row,
Publishers, Inc., New York, pp. 112-129,
1966.
Ramsey, E. M., Vascular anatomy of the
uterus, in Cellular Biology of the Uterus,
R. Wynn, ed., Appleton-Century-Crofts,
Division of Meredith Publishing Com-
pany, New York, pp. 33-52, 1967.
Ramsey, E. M., C. B. Martin, Jr., H. S.
McGaughey, Jr., I. H. Kaiser, and M. W.
Donner, Venous drainage of the placenta
in rhesus monkeys: radiographic studies,
Am. J. Obstet. Gynecol, 95, 948-955, 1966.
Ramsey, E. M., C. B. Martin, Jr., and M. W.
DEPARTMENT OF EMBRYOLOGY
643
Donner, Circulatory anatomy of the
macaque placenta. Anat. Record, 157, 408,
1967.
Ramsey, E. M., C. B. Martin, Jr., M. W.
Donner, Fetal and maternal circulations:
simultaneous visualization in monkeys by
radiography. Am. J. Obstet. Gynecol., 98,
419-426, 1967.
Ramsey, E. M., see also Donner, M. W.;
Martin, C. B., Jr.
Rudnick, D., Localization of glutamotrans-
ferase activity in the chick embryo during
the first six days of incubation. Arch.
Zool. Ital., 51, 137-147, 1966.
Sager, D., A. W. Schuetz, and R. K. Meyer,
Effect of estrone and progestational ste-
roids on human chorionic gonadotrophin
(HCG) -induced ovarian augmentation in
parabiotic rats. Endocrinology, 78, 445-
452, 1966.
Schrank, W. W., see Bishop, D. W.
Schuetz, A. W., Effect of steroids on the
germinal vesicle of oocytes of the frog
(Rana pipiens) in vitro. Proc. Soc. Exp.
Biol. Med., 124, 1307-1310, 1967.
Schuetz, A. W., and J. D. Biggers, Evidence
for an ovarian factor capable of inducing
meiotic maturation in starfish oocytes.
J. Cell Biol, 81, 159A, 1966.
Schuetz, A. W., see also Sager, D.
Whittingham, D. G., A critical phase in the
cultivation of mouse ova in vitro. J. Cell
Biol., 31, 123A, 1966.
Whittingham, D. G., Light induction of
shedding of gametes in Ciona intestinalis
and Molgula manhattensis. Biol. Bull.,
132, 292-298, 1967.
Whittingham, D. G., and C. R. Austin,
Cytological studies on the inhibition of
early cleavage by estradiol 17B in Ar-
bacia punctulata. Biol. Bull., 181 , 412,
1966.
Whittingham, D. G., and J. D. Biggers,
Fallopian tube and early cleavage in the
mouse. Nature, 213, 942-943, 1967.
Wolstenholme, D. R., and I. B. Dawid, Cir-
cular mitrochondrial DNA from Xenopus
laevis and Rana pipiens. Chromosoma
Berlin) 20, 445-449, 1967.
Young, M. Wharton, The embryology,
physiology and pathology of the "fissula
ante fenestram" in hearing and in fa-
milial deafness. Anat. Record, 157, 346,
1967.
PERSONNEL
Year Ended June 30, 1967
(including those whose services began or ended during the year)
Research Staff
David W. Bishop, General Physiology
Bent G. Boving, Physiology
Donald D. Brown, Biochemistry
Igor B. Dawid, Biochemistry
Robert L. DeHaan, Experimental Em-
bryology
James D. Ebert, Director
Elizabeth M. Ramsey, Placentology
and Pathology
Assistant Investigator
M. Edward Kaighn
Research Associates (extramural)
Louis B. Flexner, Philadelphia
Arthur T. Hertig, Boston
Irwin R. Konigsberg, Charlottesville
Samuel R. M. Reynolds, Chicago
Fellows
James F. Case, Fellow of Carnegie
Institution
Hayden G. Coon, Fellow of Carnegie
Institution
Masako Fukada, Fellow of Carnegie
Institution
Harold Kasinsky, Fellow of U. S.
Public Health Service
Harold H. Lee, Fellow of U. S. Public
Health Service
John Sinclair, Fellow of U. S. Public
Health Service
Helge Stalsberg, Fellow of U. S. Public
Health Service
Carl Weber, Fellow of U. S. Public
Health Service
Shuhei Yuyama, Fellow of Carnegie
Institution
Students
Sheldon H. Gottlieb, Undergraduate,
Johns Hopkins University
John Graham, Undergraduate, Johns
Hopkins University
R. L. Hallberg, Predoctoral Fellow,
U. S. Public Health Service
C. Nuttall, Undergraduate, Johns Hop-
kins University
Iris S. Polinger, Graduate, Biology,
Johns Hopkins University
644
CARNEGIE INSTITUTION
Merry C. Schwartz, Predoctoral Fellow,
National Science Foundation
Visiting In vest i gators
R. J. Alperin, Philadelphia, Pa.
John Bonica, Seattle, Wash.
Louis E. DeLanney, Ithaca, N. Y.
Martin W. Donner, Baltimore, Md.
Cary M. Dougherty, Baton Rouge, La.
Marlene Eng, Seattle, Wash.
M. J. T. Fitzgerald, Seattle, Wash.
R. M. Fry, Gainesville, Fla.
Peter Gruenwald, Baltimore, Md.
M. H. Hast, Iowa City, Iowa
W. P. Luckett, Madison, Wise.
C. B. Martin, Jr., Augusta, Ga.
R. D. Martin, Atlanta, Ga.
Harry S. McGaughey, Jr., Charlottes-
ville, Va.
E. C. Muecke, New York, N.Y.
Roberto Narbaitz, Baltimore, Md.
Ronan O'Rahilly, St. Louis, Mo.
Dorcas H. Padget, Baltimore, Md.
A. A. Pearson, Portland, Ore.
Glenn C. Rosenquist, Baltimore, Md.
Adolph H. Sellman, New Orleans, La.
C. B. Severn, Ann Arbor, Mich.
E. S. Tank, Boston, Mass.
Helen A. Toews, Baltimore, Md.
Richard van Praagh, Boston, Mass.
Ralph Wynn, Brooklyn, N. Y.
M. Wharton Young, Washington, D. C.
Clerical and Technical Staff
James E. Abbott, Recorder
Grace M. Andrews, Secretary-Recep-
tionist
Mary N. Barton, Librarian (part time)
Linda Berlin, Laboratory Helper
(part time)
James Blackwell, Custodian
Paul Blackwell, Custodian
William I. Cleary, Recorder
William H. Duncan, Senior Technician
Henry Fuller, Jr., Custodian
Wilbur F. Garde, Assistant Recorder
Richard D. Grill, Photographer
Elizabeth L. Hallberg, Technician
Ernest Harper, Chief Custodian
Virginia Hicks, Laboratory Helper
Eddie Jordan, Laboratory Helper-
Technician
Elizabeth Legum, Technician
Edna G. Lichtenstein, Secretary
Thomas F. Malooly, Business Manager
Juanita Mandy, Laboratory Helper
Ray O. Means, Assistant Recorder
Arlyne Musselman, Technician
John Pazdernik, Building Engineer
Margaret J. Proctor, Secretary
Martha Rebbert, Technician
Arthur G. Rever, Fiscal Officer
Bessie Smith, Laboratory Helper
Delores Somerville, Technician
Clinton Watterson, Custodian
Isabelle P. Williams, Technician
Leroy Williams, Custodian
David Wilmoth, Assistant Recorder
John L. Wiser, Machinist
Student Assistants
Lynn Billingsley, University of Mary-
land
John Chase, Johns Hopkins University
Gerald K. Goodenough, University of
Colorado School of Medicine
John Graham, Jr., Johns Hopkins
University
Fayette Marsh, University of Penn-
sylvania
Suzanne Riggs, Pembroke College
David Rosenfeld, Johns Hopkins Uni-
versity
Charlotte Rundles, Duke University
Winslow W. Schrank, University of
Maryland Medical School
Carol Sheppard, Barnard College
Jeff Sollins, Drew University
Arthur Winfree, Johns Hopkins Uni-
versity
PLATES
Plate. 1 . Electron micrographs of DNA prepared by David B. Wolstenholme by a modified
Kleinschmidt method. All micrographs X 85,000. (A— C) Molecules of native DNA from oocyte
mitochondria of X. laevis. (A, B) Twisted circles. (C) An open circle with a contour length of
5.6 fi. (D) A complex from a sample of partially renatured mitochondrial DNA. The distinct
filaments of double-stranded DNA are continuous with "puddle"-like material considered to be
single-stranded DNA.
Plate 1
Department of Embryology
Plate 2. (A) Normal chicken embryo chondrocytes in vitro fixed and stained with toluidine blue.
Note matrix between cells. X 400. (B) A sister culture 4 days after infection by RSV at high
multiplicity. Spindle cells have no discernible metachromasia. X400. (C) Autoradiogram of
chick embryo myotube in vitro, 24 hours after infection by RSV. Two myotube nuclei have
incorporated H3-thymidine. Note also a myoblast showing heavy incorporation (arrow). X 600.
D) A similar preparation, 48 hours postinfection. Almost every nucleus has incorporated H3-
thymidine. X 600. (E) A myotube showing some incorporation, 72 hours postinfection. X 600.
Plate 2
Department of Embryology
-"''"' • •• . ^
A
Pla :
Department of Embryology
*o
mmam
z=
jBtowi.tn.inl
Plate 3. (A) Immunodiffusion reactions in agar; extracted and purified SDH run against
homologous anti-SDH sera produced in rabbits. Left: SDH antigen from guinea-pig testis in
center well; outer wells contain antiserum (9780), unabsorbed at top, and previously absorbed
v/ith increasing amounts of SDH, running clockwise. Right: liver SDH antigen in center well;
outer wells contain antiserum (9777), unabsorbed at top, and absorbed with heart, liver, testis,
male and female sera in wells 2, 3, 4, 5, and 6, respectively. (B) Electrophoretic migration
toward cathode of extracted SDH (top) and crude testis supernate (just below) in agar.
Immunoelectrophoresis of testis supernate and extracted SDH reacted against anti-SDH serum
9780) in lower half of picture. At top, cathode at left; at bottom, cathode at right.
Plate U
Department of Embryology
/*?*.' .,
■ ♦# • --
*5r
& :•'• .« « <?
..•j -'j ./, %*
:^#-m
PELb*SS&S$
fl^PSK^-f-i, ^* r-'.r *^ ->'..'•• r-.V* ;•'-';,%•;.:
'.to ■ ".
t
..">
V. \'a-. -^
D
Plate 4. Cytocehmical reactions for SDH. All tests on frozen sections quick frozen in 2-
methylbutane at — 80°C. (A) Mature seminiferous tubule showing positive reaction in mature
germ cells (spermatids and spermatozoa) toward lumen, with some reactivity near basement
membrane and possibly in interstitium. (B) Control for (A), a similar preparation but with sorbitol
omitted from reaction mixture. (C) Neonatal testis with complete incubation mixture. (D) Control
for (C).
Plate 5
Department of Embryology
33
La
o
Q
ST
•if
•• ...
• • -4,
t
•
•
•
V
•
3
J2
3
*■
to
3
O
L.
O
M-
"c
E
<v
10
L>
c
0)
>
D
- —
O
c
N
O
1/)
D
E
"0
k_
01
-O
a
_CJ
</>
<D
M-
-Q
O
_D
I
c
Q
0
00
U
to
CD
5"
O
3
<
33
-0
01
0
m
■*—
_*
_D
1-
D
Q.
O
<D
<U
1/)
<
to
<L>
13
3
u
D"
"c
E
_c
<u
01
Q)
a)
-
>
(J
u
D
i
<D
■0
*~
c
O)
0
c
to
z
i
0
0
to
>^
_0
_Q
3
-D
CT
3
c
*■
—
>-t-
<L>
O
_Q
O
0)
■ —
u
_,_
c
c
<u
0)
t/>
(J
t/l
<D
<L>
O
O
3
3 Q=
cc
0
1/1
u
£ Q_
E .
— U
10 -
O
_QJ .
£
O
C
O
a.
(J
Plate G
hepart went of h'mhryolofjy
Plate 6. (A) Frozen section of testis stained with hematoxylin and eosin showing large first
spermatocytes in growth phase, tentatively regarded as the fluorescently labeled cells in Plate
5(C). (B— D) The reversible response to cryptorchidism demonstrated by neonatal guinea pigs.
(B) Normal 1 0-day-old testis at time of experimental cryptorchidism. (C) Cryptorchid testis after
90 days. (D) Recovery shown by contralateral testis 100 days after orchiopexy that followed
90 days of cryptorchidism.
Plate 5
Depart»i cut of Embryology
33
D
Q
ST
3
_Q
3
■*-
</>
3
o
<u
M-
c:
'e
<d
i/>
k-
c
<L>
>
D
- —
O
C
N
o
«/)
D
E
>_
<D
T5
a
_a>
to
<u
M-
_Q
o
_D
X
c
Q
<D
CO
u
en
Q~
O
3
<]
T3
>o
<U
<D
0£
■*—
_*
D
D
Q_
Q
a;
<u
to
<
«/}
0)
~a>
3
u
cr
*c
E
_c
a>
0)
D)
■*"
<1>
_
>
D
£
T3
1—
c
D)
a
c
V)
::
£
<U
O
_c
</>
>«
J)
-Q
3
_D
D)
3
c
-•-
- —
M-
a;
o
J2
a
<u
—
(J
^_
c
c
<L>
(J
to
0)
<D
o
o
3
3
<+:
3
£
Q-
_E
u
LO
"o
0)
1_
o
C
o
a.
u
Plate 6
Department of Embryology
Plate 6. (A) Frozen section of testis stained with hematoxylin and eosin showing large first
spermatocytes in growth phase, tentatively regarded as the fluorescently labeled cells in Plate
5(C). (B— D) The reversible response to cryptorchidism demonstrated by neonatal guinea pigs.
(B) Normal 1 0-day-old testis at time of experimental cryptorchidism. (C) Cryptorchid testis after
90 days. (D) Recovery shown by contralateral testis 100 days after orchiopexy that followed
90 days of cryptorchidism.
Plate 7. Rabbit uterus 7+ days after mating, illuminated by ring of lights, reflects parallel
lines except where a local uterine contraction brings lines together or eliminates them. A more
or less elongated small ring or spot of light is seen between contractions. A large ring is
reflected wherever the uterus is distended by an egg, forming a "dome." The white pointers
to highlights at the edge of a contraction mark the progress of two contractions moving simul-
taneously in opposite directions. The numbers indicate seconds "IUD" and black pointer shows
5 X 1 0-mm poly-ethylene tube inserted 2 days previously. L-586.
Plate 7
Department of Embryology
DOME* i 1L DOME
Wtm. Wat&fflm. ;:;^'
Piatt \
Department of Embryology
>*
>»
to
to
o
C
CD
0
1_
cv
"D
01
>+-
CO
<D
D
D
CO
CD
c
«4-
_E
<D
O
*.
'
0
TS
D
£
CD
10
0)
D
CO
0
-C
LU
i
^
to
CO
0)
!X
~D
c
^_
0
D
0)
c
:>
J3
D
a>
D
U
to
"5
D
*>r
D
^~
0
>
'q.
u
'
«/>
c
D
0
c
<
to
D
"D
>^
O
c
£
>.
en
O
"D
0
u
»j
10
Q.
~o
-»—
10
to
0)
D
C
>^
M-
a
.2?
O
.21
D
CM
D
O
IE
c
lO
1—
£
'-
jD
_c
"0
c
O"
c
u
i
<L>
CO
(D
"D
CD
D
£
10
0
_c
CO
u
.51
"D
>-
'c
c
O
c
O
c
CD
~U
:>
."t:
%
to
^D
«+-
^j
D
D
-a
M-
_CD
U
a>
to
wo
c
D
O
>
1—
15
<d
£
a
>>
3
L.
a
0
CD
u
a>
L.
u_
a
O
O
a
>.
"a
-*-
CD
c
c
L-
c
IS
1—
a>
u
a
3
0
a
0
"D
E
>^
a
D
c
0
0
E
0
«4-
O
JD
3
c
0
£
E
D
1—
0
D>
<u
O
c
to
_c
'-0
0
a>
D
■73
_c
M-
O
a;
_a>
.
>^
D)
c
co
-•—
c
0
0
u
£
O
D
c
■MM
O
a
C
D
"a;
M-
to
a
Q.
u
c
U_
u
c
to
0
O
>v
•^
10
D
a>
0
"q.
O
~D
0
>.
i_
O
Q.
(D
0
»o
to
_^
U
(/>
c
3
0
<T
ccf
_o
E
K
>
"° 1=
>^
<D C
00
<U
0
>- O
V
<U c
a»
c
c
.> O)
0
0
— <u
1) ll
-d a
Genetics Research Unit
Cold Spring Harbor, New York
Alfred D. Hershey
Director
Contents
Introduction 647
Structure and Function of Phage DNA's 650
Nucleotide distribution in A DNA 650
Campbell's model for prophage insertion 651
Deletions in A DNA 657
Base-sequence similarities between A and coli DNA's 659
Genetic transcription in bacteria infected with phage A 660
DXA replication in bacteria infected with phage T4 661
Sedimentation rates of polynucleotides 663
The States of a Gene Locus in Maize 664
The states of a^1 664
The states of a^2 665
Bibliography 672
Personnel 672
Carnegie Institution Year Book 66, 1966-1967
INTRODUCTION
Because nearly all living things
exhibit sexuality, one is led to believe
that sex plays nearly indispensable
roles in biological economy. Among
diploid organisms, whose physiology
depends in part on the continuous
interaction of two unlike genomes,
the roles of sexual processes are
multiple and diverse. Among orga-
nisms that reproduce in the haploid
phase, the single intent of sex may
be to ensure opportunities for genetic
recombination. Having reached that
conclusion, a perceptive biologist
might have anticipated what has been
found, namely, that among haploid
organisms, opportunities for genetic
recombination are often provided by
asexual means. The diversity of
means itself proves to be a remark-
able discovery, which I shall recount
here in an appreciative, as opposed to
authoritative, manner.
Until 1946, bacteria were denned
as asexual creatures. In that year
Lederberg and Tatum described for
Escherichia coli what is still one of
the few known examples of sexual
recombination in bacteria. Not the
least remarkable feature of their
experiment was that it turned up,
against considerable odds, just what
they were looking for, in violation
of the rule of serendipity that pre-
vailed for some years afterwards in
microbial genetics. Even so, more
than a decade of intensive work was
necessary before the general features
of sexual recombination in E. coli
could be appreciated.
Sexual differentiation in E. coli
depends on the status of the cells with
respect to a fertility factor called F,
first found in lines of E. coli called
F+ and absent in other lines called F~.
The crosses of Lederberg and Tatum
happened to be F+ X F~ and yielded
few recombinants because, as it
turned out, F+ cultures owe their
fertility to a third category of cell,
called Hfr, present in small numbers.
The name Hfr, which stands for high-
frequency recombination, is of his-
torical significance only. Hfr cells are
males (more strictly, donors), and F~
cells are females (recipients), in a
process of genetic transfer.
F+ cultures may be called vector
lines in that they carry F as an
extrachromosomal factor, demon-
strable in extracts as a small piece of
DNA, and transfer it with high fre-
quency at cell contact to F- cells, con-
verting them to F+. The F factor can
be lost, whereupon F+ cells revert to
F-.
F+ cells also give rise to Hfr
(male) lines with low frequency.
Male lines no longer carry F in infec-
tious form, the F factor having been
inserted into the bacterial chromo-
some. Male cells can mate with female
(F~) cells by forming cell pairs con-
nected by a cytoplasmic bridge
through which DNA passes from
male to female, not, apparently, by
passive transfer but by a process
associated with DNA replication in
the male. The female cell thus be-
comes a zygote, and shortly reverts to
the haploid condition, preserving the
female genotype except as modified
by genetic recombination that may
have occurred in the zygote.
Clues to male function came rap-
idly with detailed genetic study of
Hfr X F" crosses. The transfer of
genes proved to be a linear process
characterized by a temporal order
that could be determined experimen-
tally by interrupting mating at vari-
ous times. Several methods were used
for this, the most effective being to
separate conjugal pairs by stirring
647
648
CARNEGIE INSTITUTION
the cell suspension. Distance between
genes came to be measured as differ-
ence between times required for
transfer of those genes. Complete
chromosome transfer proved to take
about 100 minutes after formation
of conjugal pairs.
The next step called for isolation
and comparison of several male lines.
Then it was found that different
males transfer chromosomal markers
in different orders that are cj^clic
permutations of one another and,
moreover, can be either forward
(ABC) or reverse (CBA) in direc-
tion. The male determinant itself is
exceptional, and is always the last
to be transferred. Therefore males
differ with respect to the point of
insertion of F into the chromosome,
the genetic map is circular, and the
chromosome appears to be a ring
opened by F.
It follows that F functions in two
ways: first, to modify the cell
exterior so as to facilitate conjuga-
tion; second, to direct the transfer of
chromosomal material. These two
functions are clearly distinct, since
either F+ or Hfr cells can conjugate
with P", but only Hfr cells transfer
chromosomal genes during conjuga-
tion.
Why such a complicated process?
Given the decision of E. coli to mate
by DNA transfer rather than by cell
fusion, one answer is clear. The mat-
ing pair is a precarious structure,
which seldom persists until chromo-
somal transfer is complete. Thus, if
all genes are to be transferred with
equal frequency, a necessary condi-
tion if purposes of genetic recom-
bination are to be achieved, a regi-
ment of specialized males is needed.
The process described is sexual in
two senses: it involves two mating
types and, potentially, it allows
interaction between two intact chro-
mosomes. In fact, though, the trans-
fer of genetic material is usually
fragmentary and not very different
in effect from other processes of
genetic transfer that are clearly
asexual.
Asexual recombination was named
"transduction" by Zinder and Leder-
berg, who discovered one of the well-
known examples and quite properly
wished to emphasize the differences
between sexual and asexual processes.
Somewhat ironically, it turns out that
the fertility factor itself engages in
transduction in E. coli, Hfr lines can
revert to F+, and the revertants
usually carry wild-type F factor.
Occasionally, however, an F factor
recovered from an Hfr chromosome
proves to contain one or another
small chromosomal fragment of bac-
terial origin. Elements of this sort,
called F' factors, function like F in
most respects but in addition carry
their extra load of genetic determi-
nants with them, determinants that
accordingly behave as contagious ele-
ments or as authentic chromosomal
duplications, depending on their
status, extrachromosomal or chromo-
somal, in the carrier culture.
The conjugation system in E. coli
is not strictly species specific. F' fac-
tors can be transferred from E. coli
to Proteus and Serratia species, for
instance. In the foreign species they
are carried as extrachromosomal ele-
ments whose persistence is recogniz-
able by the functioning of the partic-
ular genes they contain. Turning this
process around, one can imagine that
certain "transfer factors" found in
E. coli are F' factors that originated
in another species. At any rate, some
colicine determinants and drug-
resistance factors behave in E. coli as
do F' factors from E. coli in Proteus.
Their relation to F is indicated by
their capacity for transfer by cell
contact, and by their maintenance as
extrachromosomal elements in carrier
cell lines. Their failure to interact
directly with the E. coli chromosome
GENETICS RESEARCH UNIT 649
suggests a foreign origin. Whatever transduces in a similar way genes
their origin, they function in E. coli concerned with tryptophan synthesis.
rather like viruses, carrying genetic By extension, this highly specialized
determinants not recognizable as transduction mechanism may also
chromosomal genes, and pursuing serve all the purposes by which sex
evolutionary ends of their own. They can be imagined to benefit E. coli,
can be called neither sex factors nor Note that the specialized transducing
transducing elements. phages combine features of F' and
Transduction, as a plausible nat- PI transduction,
ural means of genetic exchange The classical example of pure
among bacteria, was first discovered transducing agent is of course DNA
in Salmonella, where it is brought itself, though DNA transduction
about by bacteriophage P22. Only a doesn't seem to work for E. coli and,
few transducing phages have been to be sure, hardly seems necessary to
studied, but those few clearly belong that species. In the pneumococcus
to two quite dissimilar classes. and related bacteria, according to
One class of transducing phage, work of Hotchkiss and others, DNA
exemplified by the coliphage PI, transduction plays an authentic role
works by picking up more or less at in nature and promises to display its
random a phage-sized piece of bac- own adaptive specialties,
terial DNA and incorporating it, in In short, at least seven mecha-
place of a PI chromosome, into what nisms provide opportunities for ex-
is otherwise a phage particle. As a change of hereditary determinants in
result, that particle can transduce its bacteria. None is typically sexual,
content of bacterial genes by virtue though one closely resembles sexual
of phage-specific mechanisms of at- processes and several involve mating
tachment to bacteria. This process, between differentiated cell types,
called generalized transduction by Genes passed from cell to cell by the
phage, mimics sex at the population various mechanisms may differ with
level, since all bacterial genes are respect to origin, manner of transfer,
transduced with similar frequencies, and mode of replication in the
Here no mechanism is provided for carrier-cell line. In the face of such
maintenance of the transduced genes modalities, the meaning of many
in the recipient cell. That requires words useful in other contexts dis-
genetic exchange between the trans- solves — sex, fertility, conjugation,
duced fragment and the bacterial transduction, meromixis, episome,
chromosome. Phage P22 also is a virus — a forceful reminder that liv-
generalized transducing phage, but ing things embrace opportunities
hasn't been studied as thoroughly as without much regard for ideological
PI. categories.
A very different kind of transduc- Bacteria and phages, more literally
tion is carried out by phage A, as but not more truly than other
described in detail in later pages of organisms, are not seen through the
this report. Phage A specifically eye of the beholder, to which only the
transduces genes responsible for ga- raw materials of creation are directly
lactose and biotin metabolism in E. visible. Escherichia coli, for instance,
coli, and provides both a mechanism beset by the dual transience of mor-
for cell-to-cell transfer and a mecha- tality and shifting opportunities,
nism for insertion of the transduced exists as we know it against over-
genes into the chromosome of the whelming odds. Its story is therefore
recipient cell. The related phage <£80 trivial, unless it invites reflection.
650
CARNEGIE INSTITUTION
What we see on reflection is an inti-
mation of history, another witness
interrogated as to what living things
are about. What we see is also a
human construction, given us by the
dozen or so people who looked with
more than their eyes.
STRUCTURE AND FUNCTION OF PHAGE DNA'S
A. D. Hershey
The following report derives from
the work of Phyllis Bear, Elizabeth
Burgi, Laura Ingraham, Shraga
Makover, Anna Marie Skalka, Rudolf
Werner, and myself. Bear, Makover,
and Skalka are Carnegie Institution
Fellows. Our work as a whole is
partly supported by a research grant,
HD0i228, from the National Insti-
tute of Child Health and Human
Development, U.S. Public Health
Service.
Nucleotide Distribution in X DNA
Burgi, Skalka, and Hershey
Last year we described methods
for analyzing the distribution of
guanine and cytosine along the length
of the \ DNA molecule. These meth-
ods depend chiefly on breaking the
DXA into fragments of known aver-
age length and then sorting the frag-
ments with respect to nucleotide com-
position. The resolution that can be
achieved in this way depends on the
length of the fragments. Pieces of
fractional length 0.12, for instance,
fall into three discrete classes (Year
Bool: 65, pp. 559-562) . Pieces of frac-
tional length 0.06 fall into four
classes, as illustrated in Fig. 1. These
and other results show that the X
DXA molecule contains three large
segments that differ in composition.
From left to right these measure 0.44,
0.10, and 0.46 in fractional length,
and 57, 37, and 46 mole per cent
guanine plus cytosine (GC) in compo-
sition. The two central components
visible in Fig. 1 come from the 46%-
GC segment, which is therefore made
up of two subclasses measuring 43%
and 48.5% in GC content. The com-
position of right-terminal fragments
of various lengths shows that the
43%-GC DNA is more abundant to-
ward the molecular center than to-
ward the molecular end. However, a
short stretch poor in guanine and
cytosine at the right end of the mole-
cule (see below) complicates analysis
of terminal fragments.
The molecular ends provide a spe-
cial opportunity in that terminal
fragments can be isolated individu-
ally, owing to the specific left-to-right
joining of terminal cohesive sites
(Year Book 68, pp. 581-585). Burgi
has found that left and right molecu-
lar ends of fractional length 0.14 do
not differ appreciably in composition
from the larger terminal segments
from which they come. When reduced
to fractional length 0.012, however,
left ends contain only 48% GC, and
Density
Fig. 1 . Density distribution of mercury com-
plexes of A, DNA fragments. Fractional length
of fragments, 0.06. Hg/nucleotide ratio, 0.22.
CS2SO4 concentration, 42.8%. Centrifugation,
40 hours at 44,770 rpm in the Spinco analyti-
cal centrifuge.
GENETICS RESEARCH UNIT 651
right ends only 42%. These results feet and lyse bacterial cells, produc-
support the anticipated conclusion ing numerous phage progeny in the
that the shorter the fragment, the process. During the lytic cycle, ge-
less it is obliged to resemble its neigh- netic recombination occurs between
bors in composition. More interest- suitably marked phage chromosomes,
ingly, the two molecular ends show a Wollman and Jacob (1954) and Kai-
common tendency toward diminishing ser (1957) constructed the genetic
GC content. Since the two terminal map summarizing the results of such
genes in the genetic map function late crosses. I shall call their map, based
during the phage growth cycle, and on ordinary phage crosses, the ordi-
late functions are generally asso- nary genetic map of A. About 20
ciated with DNA of high GC content genes have been ordered so far, leav-
(Year Book 6U, pp. 526-529), it may ing a considerable gap in the center
be desirable to locate the terminal of the map where functions not essen-
genes with respect to the changing tial to the lytic cycle remain to be
base composition near the molecular identified. The length of DNA in A,
ends. about 44,000 nucleotide pairs, could
accommodate 30 or 40 genes.
Campbell's Model for Prophage 1 Second, like some but not all other
Insertion phages, A can take up residence in
the bacterial chromosome, giving rise
Escherichia coli and its phages T4 to viable cell lines of modified inher-
and X are in many ways the best- itance. Such cell lines are said to
known biological species in existence, carry prophage and to be lysogenic.
each having presented first-rate bio- (To keep matters of classification
logical problems and favorable oppor- straight, I note that among phage
tunities for their investigation, species able to produce typical lyso-
Lambda is important mainly because gens, some do and some do not occupy
it can recombine genetically with its identifiable sites in the bacterial
host, mingling cellular and viral in- chromosome, and among those that
heritance in ways that are fascinat- do, some, like A, occupy a unique site
ing to contemplate and, very likely, and some may occupy any of several
of practical importance to humans, sites.) Lwoff (1953) defined the main
In short, A is one of a small class of attributes of lysogenic cells. They are
biological elements to which Jacob immune to the lytic development of
and Wollman gave the name episome, the prophage they carry and to super-
in expectation, I presume, that many infection by phage particles of the
and varied examples remained to be same species. On the other hand,
discovered. lysogenic cells can spontaneously re-
Phage A was found by Esther Led- generate phage particles when their
erberg (1951) as a prophage residing immunity breaks down. Initiation of
in the K12 strain of E. coli, with the lytic cycle in this way is called
which J. Lederberg and Tatum had induction and can in many instances
first demonstrated sexual recombina- be brought about by irradiation with
tion in bacteria 5 years earlier. To ultraviolet light,
understand A, one must keep in mind The characteristic structure and
that it is a creature with three poten- properties of the chromosome of E.
tialities, each of which can be studied coli lysogenic for A is revealed by the
separately in other, less versatile, ele- following facts,
ments. 1. Crosses between lysogenic and
First, like other phages, A can in- nonlysogenic bacteria bring to light
652
CARNEGIE INSTITUTION
a determinant of lysogeny that is
linked to a cluster of galactose-
fermentation {gal) genes (Lederberg
and Lederberg. 1953; Wollman, 1953) .
2. When the bacterial chromosome
is transferred from a lysogenic donor
to a nonimmune recipient during bac-
terial mating, the recipient cell lyses
(Jacob and Wollman, 1956). Also, in
crosses between lysogens carrying
marked prophages, the prophage
markers show linkage to gal (Woll-
man and Jacob, 1954). The determi-
nant of lysogeny identified by bac-
terial crosses is therefore the pro-
phage itself, which is inserted into
the bacterial chromosome between gal
and a biotin gene called bio (Roth-
man, 1965).
3. Genetic analysis of lysogens also
shows that the order of genes in the
prophage is a cyclic permutation of
the order in the ordinary map of A
( Calef and Licciardello, 1960; Roth-
man, 1965) .
4. When a lysogenic bacterium re-
verts to nonlysogeny, a phenomenon
conveniently observed if the prophage
is \dg, the "cured" cell originates a
line that can, as a rule, be lysogenized
again by reinfection with phage.
Thus lysogenization is reversible with
respect to both phage and bacterial
chromosomes.
In its third manifestation, A and a
few related phages, unlike many
other phages, can form transducing
phage lines of a special kind. The
best-known examples are called "gal
transducing" because they carry
genes concerned with galactose fer-
mentation from the cell in which they
originate to cells they infect. This
propensity of A is called specialized
transduction, to distinguish it from
the rather different phenomenon of
generalized transduction discovered
in phage P22 by Zinder and Leder-
berg (1952).
When a culture of E. coli lysogenic
for \ is induced to produce phage, the
lysate contains, with rare exceptions,
A of the genetic type with which the
culture was originally infected. The
exceptions prove instructive.
Morse (1954) found that lysates
obtained by induction of X lysogens,
used as a source of phage to infect
galactose-negative bacteria, yielded
about one bacterial clone able to fer-
ment galactose per million infecting
phage particles. He had discovered
specialized transduction, character-
ized as follows by the work of Morse,
Lederberg, and Lederberg (1956) .
1. The genetic modification of the
recipient bacteria is brought about by
phage particles, and is allele specific
for gal markers present in the lyso-
genic donor.
2. Only markers adjacent to the
prophage site in the bacterial chromo-
some, namely, gal and (as found
later) bio, are transduced by X.
3. Transducing phage particles
originate only by induction of lyso-
gens, not by infection with ordinary X.
4. Cells genetically modified by
transduction are typically immune to
A, either because they are lysogenic
or because they carry a defective pro-
phage.
5. The genetically modified cells re-
vert to galactose-negative with rather
high frequency. The revertants typi-
cally show the gal genotype of the
recipient; that is to say, reversion
occurs by loss of genes brought in
by the phage.
6. When a culture made galactose-
positive by lysogenization is induced
in turn, and the lysate is used to in-
fect galactose-negative bacteria,
about one galactose-fermenting clone
per 10 phage particles is obtained.
These secondary lysates are called
"high frequency transducing," as op-
posed to the "low frequency trans-
ducing" lysates obtained by induction
of an ordinary A lysogen.
Item 1 above is the criterion of
transduction of all kinds. All the
GENETICS RESEARCH UNIT 653
other features of transduction by X particle per million wild- type parti-
are peculiar to specialized transduc- cles.
tion. Note particularly item 5, which 2. Transducing particles adsorb to
shows that gal transduction is usually bacteria and lyse them but produce no
brought about by addition of genes to offspring. Transducing particles are
the bacterium, not by substitution of found only in mixed yields coming
gene for gene. Note also that com- from bacteria doubly infected with
parison of the properties of nontrans- transducing and wild-type particles.
ducing, low frequency transducing, Situations of this kind are well
and high frequency transducing ly- known, and are understandable, since
sates shows that transducing phage gene function defective in one chro-
lines originate by some rare event mosome can often be supplied by the
associated with multiplication or in- corresponding gene in another chro-
duction of prophage. However, the mosome present in the same cell.
nature of this event could not be in- Arber named the transducing par-
vestigated until an important source tides Xdg, signifying "defective" and
of confusion had been cleared up. "gal transducing."
Clarification came from the simul- 3. "Transductants," that is, galac-
taneous work of Arber, Kellenberger, tose-fermenting clones of bacteria
and Weigle (1957) and Campbell originating by transduction, are of
(1957), work that depended for its two sorts: defective lysogenic, pro-
success on the simple expedient of duced rarely by infection with a sin-
making sure that bacterial cells in gle particle of \dg, and actively lyso-
the experimental cultures were in- genie, produced more frequently after
fected with single phage particles, simultaneous infection with both Xdg
Experiments performed with that and wild-type particles. Actively lyso-
precaution quickly showed that trans- genie transductants also arise with
ducing lysates contain a mixture of low frequency when single particles
normal phage particles and transduc- of Xdg infect bacteria already lyso-
ing phage particles, and that trans- genie for A.
duction clones of the sort previously 4. Defective lysogenic transduc-
studied usually arise by double infec- tants yield, on induction, sterile ly-
tion with particles of both kinds. It is sates. Actively lysogenic transduc-
worth noting that Morse and the tants yield, on induction, the high
Lederbergs (1956) and Weigle frequency transducing lysates al-
(1957) had performed experiments ready defined.
that narrowly missed bringing out 5. Defective lysogenic transduc-
these facts. tants revert spontaneously to galac-
The conclusions reached by Arber, tose-negative clones that are nonlyso-
Kellenberger, and Weigle (1957) and genie. Actively lysogenic transduc-
Campbell (1957) may be summarized tants revert to galactose-negative
as follows. clones that are typical stable lysogens
1. High frequency transducing ly- or, sometimes, nonlysogenic (Camp-
sates contain about equal numbers of bell, 1963).
transducing particles and wild-type These facts permit a number of
particles, which can be separately important inferences. First, the cor-
enumerated because both kinds of relation between defectiveness and
particles kill bacteria but only wild- transducing power in Xdg shows that
type particles form plaques. By infer- the inclusion of bacterial genes in a
ence, low frequency transducing ly- phage particle entails the loss of
sates contain about one transducing phage genes. Second, the instability
654
CARNEGIE INSTITUTION
of lysogens carrying xdg shows that
Xdg is defective as a prophage, as
well as defective in functions ex-
pressed after induction, though the
nature of the prophage defect is not
yet clear. Finally, phage genes and
bacterial genes must be combined in
a single structure, the Xdg chromo-
some, for two reasons. Both cate-
gories of gene can multiply during
the lytic cycle of phage growth to
produce high frequency transducing
lysates (Weigle, 1957). Both cate-
gories of gene are lost simultaneously
when a defective lysogenic transduc-
tant reverts to galactose-negative
(Campbell, 1957). Thus prophage in-
sertion and the origin of Xdg are
different manifestations of genetic
recombination between phage and
bacterial chromosomes.
Arber (1958) made crosses be-
tween xdg and X that revealed two
remarkable features of the structure
of the Xdg chromosome. First, the
missing phage genes, notably a host-
range marker, correspond to a sub-
terminal section of the ordinary ge-
netic map. Second, in different lines
of Xdg, the right ends of the deleted
segments fall at the same point in the
map, just to the right of the host-
range marker.
In the meantime, Campbell (Year
Book 57, pp. 386-389) had discovered
in X what are now known as sup-
pressor-sensitive (amber) mutations.
These mutations are important not
only because of their role in the elu-
cidation of gene function but also
because they can be found in any
gene that has an essential function.
Campbell (op. cit. and subsequent
papers) exploited them, by what is
known as the overlapping deletion
method, to analyze the structure of
Xdg in great detail. He found, first of
all, that Xdg' 8 of independent origin
tend to differ from one another,
whereas the Xdg particles in a high
frequency transducing lysate derived
from the same transduction clone are
identical. This conclusion meant that
xdg's, like wild-type phage, are stable
during replication in both the pro-
phage condition and the lytic cycle
(verified by Campbell, 1960). It also
meant that the events by which Xdg's
originate, doubly exceptional because
they give rise to Xdg's of various
kinds, occur at the moment of induc-
tion, not during replication of X pro-
phage (verified by Campbell, 1963).
The common feature of all Xdg's,
besides the inclusion of gal genes for
which they are selected, is the ab-
sence of a single block of phage genes.
The deleted region is variable in
length, but always includes genes I,
J, K, and part of L on the right. The
left end of the deletion may fall at
any of 25 or more places in any gene
A through L. (The genes identified
by complementation tests with sup-
pressor-sensitive mutants, and rele-
vant here, were named as follows,
starting from the left end of the
ordinary genetic map: A through H
in alphabetical order, then M, L, K,
I, J. Thus the right-hand ends of all
the deletions, placed to the right of
the host-range marker by Arber, lie
to the right of J, which may be the
same gene.)
Weigle, Meselson, and Paigen
(1959) analyzed Xdg's by physical
methods. They found that lines of Xdg
of independent origin differ in den-
sity, each line maintaining its charac-
teristic density during growth. They
ascribed the differences in density
to differences in DNA content per
particle, correctly as it turned out.
Thus crossovers between phage and
bacterial DNA's giving rise to Xdg's
are nonequational in the sense that
the lengths of phage DNA deleted
and of bacterial DNA inserted are
unequal.
The facts recapitulated above im-
pose severe restrictions on possible
mechanisms of recombination be-
GENETICS RESEARCH UNIT 655
tween A and E. coli. Normal prophage tion in the prophage map explains in
insertion and excision, conserving turn the fixed right end and variable
both chromosomes intact, already im- length of different deletions. In an
pose restrictions. Campbell (1962) analogous manner, bio transducing
proposed what seems to be the only particles can arise concomitantly
acceptable model. It is presented in with deletions at the left end of the
Fig. 2, which is intended to be self- prophage. The bio transducing parti-
explanatory, cles, unlike gal transducing particles,
Note what the model accomplishes, need not be defective (Wollman,
The ring configuration makes pro- 1963), presumably because the dis-
phage insertion and excision credible tance in the prophage map between
as consequences of single, reciprocal N and bio is shorter than that be-
crossovers. The ring structure also tween gal and J (see Fig. 2).
permits, though it does not demand, Note too that the model predicts
map permutation. Given the map per- permutations that are cyclic, denned
mutation, the model explains the symbolically as ABC ^ BCA (Web-
origin of Xdg as a terminal prophage ster's New International Dictionary,
deletion associated with substitution second edition, 1934). Cyclic permu-
of bacterial DNA (Kayajanian and tation should be clearly distin-
Campbell, 1966) . Their terminal posi- guished from circular permutation,
pp
N
£ A phage particles
1 1
Configuration
gal n bio at crossover
I 1
P'
J bio
I
I
I
Prophage
Adg phage
N R particles
Fig. 2. Structural relations of A, A prophage, and Adg according to Campbell's model. Genetic
maps are indicated by the phage genes A, J, N, R and the bacterial genes gal and bio. Phage
and bacterial components ore indicated by shading, terminal cohesive sites by arbitrary sym-
bols. The symbol pp stands for the locus of permutation points and ii for the locus of prophage
insertion sites, loci that become crossover regions ip and pi in the prophage map.
656
CARNEGIE INSTITUTION
defined by Streisinger, Edgar, and
Denhardt (1964) as a process giving
rise to sequences among which all
cyclic permutations are equally fre-
quent.
Kayajanian and Campbell (1966)
clarified certain features of the re-
combinations giving rise to Xdg's.
They selected lines of Xdg containing
some but not all of the gal genes, thus
fixing within narrow limits one ter-
minus of the substituted bacterial
DXA. If xdg's arise as terminal pro-
phage deletions, the selected lines
should contain a fixed amount of bac-
terial DXA and their density should
be inversely related to the size of the
deletion. Kayajanian and Campbell
found a strict correlation. Equally
important, they found that the
lengths of the deletions varied over
the same range as deletions in un-
selected Xdg's. Thus restriction of the
crossover point to a specified region
in the bacterial DNA does not mark-
edly restrict its location in the phage
DXA. If the crossovers depend on
local similarity of base sequences,
matching sequences are not ordered
in the same way in the two DNA's.
This is a very clear demonstration of
illegitimate crossing over.
The selected Xdg's varied in DNA
content from —14.2% (deletion end-
ing in A) to +3.4% (deletion ending
in M), expressed in relation to the
DXA content of wild-type phage par-
ticles. The range of variation signi-
fies that the genetic distance between
genes A and M corresponds to about
18% of the length of the X DNA
molecule. If no genes in this region
remain to be found, the measurement
requires 1100 nucleotide pairs per
gene, which is a reasonable number.
The results of Weigle, Meselson,
and Paigen (1959) as well as those of
Kayajanian and Campbell show that
Xdg, like wild-type A, enters and
emerges from the prophage state
without gain or loss of DNA. There-
fore both types of phage, though
differing in the frequency with which
they enter and leave the bacterial
chromosome, must do so by crossing
over at points separated by a fixed
distance, either a unique pair of
points or, as originally suggested by
Campbell, points determined by legit-
imate crossing over in homologous
regions of appreciable length.
A proper appreciation of Camp-
bell's model, which, together with the
facts that go with it, is one of the
important contributions of micro-
biology to genetic thought, calls for
an assessment of its historical
origins. (Its actual origin, in the in-
ventive mind of its creator, is another
matter.) Note first of all that the
structures invoked, with the excep-
tion of a properly situated region of
homology between phage and bac-
terial DNA's, have by now been dem-
onstrated. Only the origins of the
structures are subject to hypothesis.
Note, too, that prophage excision and
insertion are strictly analogous to
deletion and insertion in general. The
formal models for genesis of all chro-
mosome rearrangements are basically
the same as Campbell's. (See, for ex-
ample, Fig. 46 in Sturtevant and
Beadle's Introduction to Genetics,
1939, where the general scheme is
attributed to Serebrovsky.) Camp-
bell's perceptiveness lay, perhaps, in
recognizing early that the prophage
has to be inserted, in spite of some
experimental results that seemed to
point in another direction.
Sturtevant and Beadle pointed out
in their book that crossing over is a
process concerning which only the
results are known, and defined ille-
gitimate crossing over as crossing
over between nonhomologous chromo-
somes or between different parts of
the same chromosome. Therefore no
hypothesis is implied in the use of
these terms except, perhaps, that nor-
mal prophage excision and the origin
GENETICS RESEARCH UNIT 657
of Xdg share common mechanisms. It Questions that can be asked about
is doubtful too whether use of the the origin of Xdg are severely limited
word "homology" implies a hypoth- because rare events are involved. All
esis, and whether it should be used we know is that, in effect, a single
at all in the present connection. In piece of DNA is cut out of the lyso-
Campbell's model, "region of ho- genie bacterial chromosome, the ends
mology" could be translated as "locus of the piece are joined, and additional
of crossover points." The meaning of cuts are made to create ends re-
homology is clear when it refers to sembling the ends of normal X DXA
chromosomes identical except for molecules. And, of course, that bac-
experimentally introduced genetic teria can sometimes survive deletions,
markers. Otherwise, in molecular con- These facts are plausibly accounted
texts at least, it is better to speak of for if deletions in general result from
lengths and distributions of common single, reciprocal crossovers, as in
base sequences, on the one hand, and the models of Serebrovsky and Camp-
crossover frequencies on the other, bell. Experiments designed to detect
the relation between the two being bacterial deletions in the vicinity of
unknown. The solid fact coming from the prophage, arising concomitantly
the analysis of Xdg's is that illegiti- with prophage induction, might serve
mate crossing over, if it depends on to indicate whether or not the illegiti-
common base sequences in this in- mate crossovers that generate trans-
stance, can occur between DNA seg- ducing phage are reciprocal events.
ments in which the common se- Such experiments should be feasible
quences are short compared with with phage <£80 if not with A (Frank-
lengths of genes, and are, moreover, lin, Dove, and Yanofsky, 1965).
erratically distributed. Campbell's model beautifully or-
The basic question here concerns ganizes facts and clarifies questions.
the mechanism of crossing over in The same cannot be said of other
general, but one can ask more modest models that might be experimentally
questions, too. These have to do with equivalent to it.
similarities or differences between
the illegitimate crossing over that n . . . n\7J
gives rise to Xdg's and the crossing Deletions m X DNA
over that is responsible for normal Skalka and Burgi
prophage excision. Do the two proc-
esses depend on the same or different A mutant of phage X known as Xb2
enzymes? Current work may answer contains a DNA molecule 15% to
this question (Signer and Beckwith, 20% shorter than that of wild-type X
1967; Zissler, 1967) . Are both exam- (Kellenberger, Zichichi, and Weigle,
pies of crossing over reciprocal? Pro- 1961; Burgi, Year Book 62, p. 482).
phage insertion is reciprocal in the It has suffered a deletion near the
sense that, in the overall process, four chromosomal center to the right of
strands are cut and rejoined in new the host-range marker (Jordan,
combinations. Prophage excision is 1964). The mutant grows normally
probably reciprocal in the sense that except that it does not produce stable
a single lysogenic bacterium can give lysogens. Indirect evidence suggests
rise to daughter cells that lyse and that it has lost the crossover locus
daughter cells that are nonlysogenic within which genetic recombination
(Weisberg and Gallant, 1967) . These with the bacterium occurs during
processes are also reciprocal in the lysogenization (Campbell, 1965).
sense that they are reversible. Burgi and Skalka have found that
658
CARNEGIE INSTITUTION
b: DNA lacks all or most of the 37%-
GC segment present in the DNA of
wild-type A. Therefore b2+ function
resides in or adjacent to that seg-
ment, and the segment does not con-
tain genes whose functions are essen-
tial during the lytic cycle of phage
growth.
Skalka has also analyzed DNA
from a defective strain of \ known
as xdg (A-J). which is a typical de-
fective, ga /-transducing phage whose
deletion spans genes A through J in
the left third of the genetic map.
Her results reveal two features of
the DXA of xdg(A-J) that are best
seen by analysis under somewThat
different conditions (see legends,
Figs. 3 and 4) . As shown in Fig. 3,
a large fraction of the DNA present
in wild-type phage, including all the
left-terminal 57%-GC section, has
been replaced in Xdg (A-J) by DNA
having the GC content characteristic
of the DNA of E. coll. As shown in
Fig. 4, the 37%-GC section also has
been deleted. The deleted DNA there-
fore represents a continuous stretch
measuring at least 54% of the wild-
type molecular length taken from the
left arm of the DNA molecule (but
not including the extreme tip, because
cohesive function remains). This
physical structure, like the genetic
results, is best interpreted as the con-
sequence of a terminal prophage dele-
tion according to the model shown in
Fig. 2. Given that interpretation,
four conclusions follow.
1. The 37%-GC segment must lie
near the right end of the prophage
map.
2. The locus of permutation points
in the X DNA molecule lies near or
to the right of the right end of the
37%-GC section, at a position distant
at least 54% of the molecular length
from the left molecular end.
3. The 37%-GC segment cannot lie
in a region of exact homology be-
tween A, and E. coll. If it did, loss of
the A homologue would be compen-
10
20 30
Fraction number
Fig. 3. Substitution of bacterial for viral DNA in Xdg. Solid line, H3-labeled DNA from Xdg[A— J).
Broken line, P32-labeled DNA from wild-type A. Size of fragments, 4 X 1 06 daltons. Hg/nu-
cleotide ratio, 0.3. CS2SO4 concentration, 45.5%.
GENETICS RESEARCH UNIT
659
20 30
Fraction number
40
Fig. 4. Absence of 37% -GC section in DNA of Xdg. Solid line, H3-labe!ed DNA from Ac/g(A-J).
Broken line, P32-labeled DNA from wild-type A. Size of fragments, 1 .9 X 1 06 daltons. Effective
Hg/nucleotide ratio, 0.20. CS2SO4 concentration, 42%. The two DNA's were mixed before
shearing and centrifuged after the addition of unlabeled T2 DNA fragments.
sated by gain of the bacterial homo-
logue in the origin of Xdg,
4. All Xdg's should lack the 37%-
GC section plus varying fractions of
the 57%-GC section depending on the
lengths of their deletions, that of Xdg
(A-J) being one of the longest.
Independently of models, compari-
son of the DNA's of Xb2 and Xdg
(A-J) shows that the genes A
through J, present in b2 but not in
Xdg, must lie in the 57%-GC segment
of the molecule.
The physical data show that Xb2
has suffered a 10% deletion of DNA
near the right prophage terminus and
an additional 5%-10% deletion of
DNA of unknown location. We sug-
gest as a plausible hypothesis that
Xb2 arose by an illegitimate crossover
deleting some of the DNA corre-
sponding to both prophage ends.
Base-Sequence Similarities betiveen
X and E. coli DNA's
Ingraham and Hershey
From previous work, we concluded
that base-sequence similarities be-
tween the DNA's of phage X and E.
coli are generally though not uni-
formly distributed throughout the
length of A, DNA {Year Book 65, pp.
562-565). As far as it goes, this con-
clusion is consistent with the hypoth-
esis that the illegitimate crossovers
giving rise to diverse lines of trans-
ducing phage depend on irregularly
distributed base-sequence similarities.
The preferred crossover locus, pp
in Fig. 2, responsible for normal pro-
phage insertion and excision, must be
something different. To account for
the constancy in DNA content of
phage A through its lysogenization
660
CARNEGIE INSTITUTION
cycle, that locus must contain either a
unique crossover point or a region of
matching* base sequences that is co-
linear in the DNA's of A and coli.
Skalka concluded that the Xdg she
analyzed had arisen from a prophage
inserted by crossing over near the
right end of the 37%-GC section of
the DXA molecule. To examine base-
sequence similarities in this region,
we broke A DNA into fragments of
fractional length 0.32, removed re-
joined ends (Year Book 63, pp. 581-
582) . and then selected mercury com-
plexes of the remaining central sec-
tions that were rich in adenine and
thymine. Such fragments should con-
tain most of the 37%-GC section and
an adjacent part of the 46%-GC sec-
tion, and should be shorter than un-
selected fragments. When the selected
fragments were reduced in size to
about 0.1 of the original molecular
length and were again analyzed by
Hg-Cs2S04 fractionation, two distinct
components, with densities corre-
sponding to 37% and 46% GC, in the
approximate ratio 1/1.5, were recov-
ered. The effectiveness of the selec-
tion was also indicated by complete
absence of a 57%-GC component.
Both components recovered from the
molecular centers proved to bind
poorly to the DNA of E. coli, as
shown in Table 1.
TABLE 1 . Hybridization Tests with A DNA
Fractions
E. coli
DNA Filters,
A DNA Filters,
A DNA Fraction
100 jxg
10 jug
37%-GC centers
6,6
83
46%-GC centers
7,8
86
Right-end thirds
17, 19
87
Left-end thirds
11, 13
83
Unfractionated
13, 13
82
The numbers express percentages of soni-
cated, P32-labeled DNA fractions bound to
unlabeled DNA attached to membrane filters.
Tv/o numbers signify duplicate measurements.
Procedure according to Denhardt (1966).
We also examined short right and
left molecular ends (fractional length
0.14), donated by Elizabeth Burgi.
They were indistinguishable from
the corresponding terminal thirds in
tests like those shown in Table 1.
We conclude that the measurable
base-sequence similarities between A
and coli DNA's are strongest (in A
DNA) near the right molecular end,
weakest in the 37%-GC and 46%-GC
sections near the molecular center,
and intermediate near the left molec-
ular end (see also Year Book 65, pp.
562-565).
These results do not of course ex-
clude the possibility of a critical re-
gion of exactly matching base se-
quences near gal in E. coli and near
the molecular center in A DNA, be-
cause genetic considerations suggest
that the crossing over responsible for
prophage insertion ought to be rather
precisely denned, and that a short
matching sequence in a region of poor
matching would be all to the good.
The concentration of matching se-
quences in other parts of the mole-
cule, however, argues for a recogni-
tion device that does not depend on
homology alone. Signer and Beckwith
(1967) and Zissler (1967) propose,
in fact, that phage A employs a spe-
cial enzyme that somehow directs the
normal insertion and excision of the
prophage. In principle, such an en-
zyme could act by recognition of one
matching base sequence among many,
or even different, specified sequences
in phage and bacterial DNA's.
Genetic Transcription in Bacteria
Infected with Phage A
Skalka
Skalka and Harrison Echols (Uni-
versity of Wisconsin) have studied
the effects of mutational defects in A
on production of messenger RNA
during phage growth. Their results,
which are being published in detail
GENETICS RESEARCH UNIT 661
elsewhere, identify two genes whose fected cell could control its rate of
primary function may be control of DNA synthesis in either of two
transcription. ways: by varying the number of
Mutations in gene N block produc- growing points per unit length of
tion of all messenger excepting a DNA, or by varying the rate of syn-
small amount similar to that formed thesis at individual growing points.
when protein synthesis is inhibited Werner has examined this question,
by chloramphenicol. The simple in- and has found that the variable fac-
ference, also suggested by the work tor is the number of growing points.
of R. Thomas, is that a product of The rate of DNA synthesis at indi-
gene N directly initiates messenger vidual growing points remains con-
synthesis characteristic of the lytic stant.
cycle of phage growth. Werner performed experiments of
Mutations in gene Q, which do not three types, described here in terms
block DNA synthesis or production of specific examples,
of early-phase messenger, selectively Experiment 1. Infect thymine-
depress transcription of genes re- requiring bacteria (E. coli B3) with
sponsible for late functions. The sim- thymine-requiring phage (T4M8)
pie inference is that a product of and allow growth to proceed in
gene Q specifically facilitates tran- medium supplemented with 2 /^g/ml
scription of those genes. Needless to of thymidine. At minute 40, add 20
say, the facts are more reliable than ^g/ml of IP-labeled 5-bromouracil. At
the inferences at this time. minute 42, add a large excess of un-
labeled thymidine and permit grow-
DNA Replication in Bacteria ing points to move away from the
Infected with Phage TU labeled sections in the DNA. Extract
DNA from the cells at minute 45,
break samples into fragments of
Last year Werner reported that a various sizes, and measure both the
T4-infected cell contains a number of 5-bromouracil content of the DNA
sites of DNA replication, about one and the size to which the DNA must
for each molecular equivalent of be broken to liberate pieces, identifi-
phage DNA. His measurements were able by their density, that contain
made at 45 minutes after infection of one heavy and one light strand. The
cultures growing at 25 °C. Such cul- critical size turns out to be 0.10 of the
tures are entering a steady state of length of a T4 DNA molecule. The
phage growth in which a constant 5-bromouracil content of the DNA
rate of DNA synthesis is matched by corresponds to 6.0 DNA molecules
an equal rate of phage particle for- per cell. Therefore the individual cells
ma tion to maintain an intrabacterial contain an average of 6.0/0.1 or 60
pool of replicating DNA of constant growing points at 40 minutes after
size. infection. In this experiment, the
When bacteria are infected with total amount of DNA per cell is not
phage T4, DNA synthesis starts measured.
about 10 minutes later and attains a Experiment 2. Start the infection
rapid rate very quickly, a rate that is in medium supplemented with C14-
clearly not proportional to the thymidine, then switch to HMabeled
amount of phage-precursor DNA 5-bromouracil by centrifugation and
present in the cells. The approach to washing. The procedure is the same
the steady state therefore calls for as in experiment 1 except for the
regulation of DNA synthesis. The in- additional measurement of the
Werner
662
CARNEGIE INSTITUTION
amount of DXA synthesized after
infection. Because some cells lyse dur-
ing- centrif ligation, the only reliable
mates are L, the average length of
DXA segments of hybrid density, and
F, the fraction of the recovered DNA
containing: 5-bromouraeil in place of
thymine. The ratio L/F gives the
amount of DXA per growing point.
This ratio measures 1.5 molecular
equivalents at 45 minutes after infec-
tion.
Experiment 3. Start growth in
thymidine-containing medium and
add an excess of IP-labeled 5-bromo-
uracil at time t. After an additional
interval. At, extract DXA, reduce it
to small fragments by sonication, and
measure the ratio between H3 counts
in heavy DXA (both strands labeled)
and in hybrid DXA (one strand
labeled). This ratio increases in pro-
portion to At, and results can be in-
terpolated to find At corresponding to
the ratio heavy /hybrid equal to 0.5.
The interpolated At is roughly the
interval during which two growing
points move over an average segment
of replicating DXA. Since the rate of
movement of growing points is
known, the length of DXA between
growing points can be calculated. The
distance between growing points, D,
depends both on the time t at which
5-bromouracil is added and on the
thymidine concentration of the me-
dium. At 2 iJ.g/m\ thymidine, D —
0.17 of a T4 length when t = 20, and
0.29 of a T4 length when t = 40. At
higher concentrations, D is greater at
t : : 40, about 0.73 of a T4 length.
Results of these three types of ex-
periment permit the following con-
clusions.
1. Growing points first appear at
10 minutes after infection and in-
crease in number at the rate of 3 per
bacterium until they number 60 at
30 minutes, after which the number
remains constant (experiments of
type 1 ) . The number of growing
points found during a short pulse
with 5-bromouracil does not depend
on the amount of DXA in the cultures
as influenced by the prevailing thy-
midine concentration.
2. Individual growing points move
at the rate of 5% of a T4 length per
minute in the presence of 5-bromo-
uracil. This rate does not depend on
the time after infection at which the
measurement is made (experiments
of types 1 and 2) .
3. The distance between growing
points in replicating structures varies
from about 0.2 of a T4 length at early
times to nearly 1 at late times (ex-
periments of type 3). Replicating
DXA can therefore take the form of
a multiply branched structure.
4. During the steady state of phage
growth, cultures maintained in the
presence of excess thymidine produce
4.5 phage particles per bacterium per
minute and synthesize DXA at the
equivalent rate. If there are 60 grow-
ing points per cell, they are moving
at the rate of 0.075 of a molecular
length per minute. The local rate of
DXA synthesis measured in the
presence of 5-bromouracil is 0.05.
Brief labeling with H3-thymidine or
with C14-5-bromouracil shows that
growing points move 1.5 to 2.0 times
faster in the presence of thymidine
than in the presence of 5-bromoura-
cil. Thus each result checks fairly
well with two independent measure-
ments. The same is true at earlier
times when the rate of DXA syn-
thesis per cell is increasing.
Werner concludes that the rate of
DXA synthesis in the presence of
thymidine is controlled by the num-
ber of growing points, not by their
rate of movement, and that there is
no severe limitation to the number of
growing points per length of DXA or
per cell. A similar conclusion was
suggested by Sueoka and his col-
leagues concerning replication of
bacterial DXA in Bacillus subtilis.
GENETICS RESEARCH UNIT
663
According to Werner's results,
growing points accumulate rapidly
but move rather slowly and tend to
remain clustered in the templates on
which they originate. This conclusion
suggests an unanticipated role for
genetic recombination: to distribute
growing points over the newly syn-
thesized DNA. Eckhart found that
genetic markers introduced into a T4-
infected cell by a superinfecting
phage replicate mainly after recombi-
nation with markers contributed by
the primary infection. His finding
can perhaps be explained, wholly or
in part, by the clustering effect men-
tioned above.
Sedimentation Rates of
Polynucleotides
Ingraham
Molecular weights of the polynucle-
otide chains released by denaturation
of DNA can be measured from their
rates of sedimentation in alkaline
solutions according to an equation of
the type T>2/'D1 = (M2/Mi)tt, where
the D's and M's refer to distances
sedimented and molecular weights of
two DNA's spun in the same tube.
The exponent a was estimated at 0.40
by Studier and at 0.38 by Abelson
and Thomas, the difference possibly
reflecting the use of different refer-
ence DNA's and different solvents.
In any case, it remains uncertain
whether or not the exponent a is
really constant over a wide range of
molecular weights. With the assist-
ance of Dr. Gobind Khorana, Ingra-
ham has made a partial check.
Khorana supplied two samples of
OMabeled thymine deoxyoligonucleo-
tides, one containing the heptanucleo-
tide, the other containing mixed nu-
cleotides of somewhat greater length.
Since the mixed sample was larger,
it was calibrated against the hepta-
nucleotide and then used as a refer-
ence in other measurements.
In principle the check is simple.
The molecular weight of single
strands of A DNA is 15.5 million, that
of the heptanucleotide 2320. Measure-
ment of the relative distances sedi-
mented permits an estimate of «.
Owing to the very great difference in
sedimentation rates, the comparison
has to be made in several steps.
Ingraham used a sample of DNA
broken by stirring and a sample of
enzymically hydrolyzed DNA as in-
termediate references of unknown
molecular weight. Other than the
thymine oligonucleotides, the mate-
rials were prepared from P32-labeled
A DNA.
Adjacent pairs in the molecular
weight series were spun in concen-
tration gradients containing 5%-
20% sucrose, 10-3 M ethylenediamine-
tetraacetate, 0.3 M NaOH, and 0.7 M
NaCl. Time and speed of centrifuga-
tion were chosen to bring the faster-
sedimenting member of the pair well
down the tube.
Results for two or three trials of
each kind are given in Table 2. They
show that the molecular weight ratio
15.5 X 106/2320 corresponds to a dis-
tance ratio falling between 28.0 and
31.4. Substitution of these limits in
the equation gives estimates of a
ranging from 0.378 to 0.392. Ingra-
ham also verified that the exponent
0.38 serves in the range of molecular
weights between 15.5 million and
320,000 under the conditions speci-
fied. The provisional molecular
TABLE 2. Relative Sedimentation Rates of
Polynucleotides
DNA Pair
Distance Ratio
Intact A— sheared A
Sheared A— enzyme digest
Enzyme digest—
oligonucleotide pool
Oligonucleotide pool—
heptanucleotide
Intact A— heptanucleotide,
calculated
2.78, 2.81, 2.82
3.15, 3.17, 3.23
2.73, 2.81, 2.88
1.17, 1.20
28.0-31.4
664
CARNEGIE INSTITUTION
weights she cited last year are there-
fore about right (Year Book 65, pp.
564-565) .
The cheek is incomplete in at least
one respect. Nobody knows how
strongly the sedimentation rates of
oligonucleotides may depend on their
composition.
THE STATES OF A GENE LOCUS IN MAIZE
Barbara McClintock
The locus of a gene whose action
is governed by a known system of
controlling elements may undergo a
change that alters the pattern of ex-
pression of the gene in mature tis-
sues. The change arises as a single
event and thus resembles a mutation.
Because the effect is on the gene-
control mechanism, the alteration is
not termed a mutation but rather a
change in "state" of the locus. The
states reflect the presence of a mech-
anism that is able to modify gene ex-
pression during development. The
nature of the states and their signifi-
cance will be discussed in this report.
The distinctiveness of the various
states of a gene locus may be illus-
trated by selected examples. For this
purpose, some of the states of a^'1
and axm'2 will be considered. These
symbols were assigned to designate
the effects on the A1 (Anthocyanin)
locus in maize of two independent in-
ceptions of control of the action of its
gene by the Spm (Suppressor-muta-
tor) system. This action is required
for production of anthocyanin pig-
ment in both plant and kernel. The
states of cii"1"1 differ from those of
a,/"-2 in their modes of regulating gene
action. Some aspects of the difference
were mentioned in previous Year
Books, but its breadth was not em-
phasized because knowledge of some
of the states of axm-2 was insufficient
for characterization. Recent investi-
gations have made it possible to com-
pare a number of states of a^'2 with
those of a,!™'1 previously studied. The
comparisons demonstrate the manner
in which a single control system pro-
vides diversity of regulation of gene
expression. Although distinctions
among states are made apparent in
the phenotypes of both plant and ker-
nel, only those manifested in the ker-
nel will be considered here, as they
can be readily illustrated.
The States of a1m~1
The states of a^'1 were derived
through modifications of the original
state that arose when the first in-
stance of inception of control of At
gene action by the Spm system oc-
curred in my stocks. Each state was
recognized initially by the appearance
of an altered type or distribution of
anthocyanin pigment in the progeny
of a plant carrying the original state,
and most often in a single individual
of the progeny. Investigation of the
cause of the changed phenotypes was
begun with such individuals and con-
tinued with their progeny. The
studies indicated that each change
was initiated at the Ax locus by the
element of the control system residing
there, and that it represented one
type of this element's response to the
Spm element, located elsewhere in the
chromosome complement. In the
presence of a fully active Spm ele-
ment, the states of a1m'1 are distin-
guished from one another by differ-
ences in the time of "turning on" of
Ai gene action during development of
a tissue, by the frequency of occur-
rence of such action in the cells of a
tissue, and by the pigment intensity
in those cells in which it is produced.
They are also distinguished by the
intensity of pigment produced in the
GENETICS RESEARCH UNIT 665
absence of an active Spm element. The States of axm-2
One state produces no pigment
when Spm is inactive. Each of the Some of the states of a*1""2 are so
other states does produce pigment, multifaceted in expression that an
which is uniformly distributed over understanding of them requires
the aleurone layer of the kernel. The forms of analysis not demanded by
pigment is intense with one state but any of the states of a^'1. The infor-
pale to faint with the others. There is mation obtained from the analyses
no direct relation between the expres- has provided additional evidence of
sion given in the presence of a fully the way in which gene action may be
active Spm element and that given in regulated differentially by a single
its absence. It should also be men- system of controlling elements. The
tioned that when two states are com- states of a/*-2 may be divided into
bined as alleles in a plant or kernel, two major classes: those having an
gene action is regulated independ- Spm element at or close to the Ar
ently by each, and when an active gene locus, and those that show no
Spm element is present, the pattern evidence of the presence of Spm at
of anthocyanin distribution and in- the locus but respond to that element
tensity produced by one allele over- in a distinctive manner when it is
laps that produced by the other. Illus- located elsewhere in the chromosome
trations of the phenotypes of kernels complement. The original state, from
produced by the different states of which most of the other states were
cia™'1, both in the presence and in the derived, belongs in the first class. An
absence of a fully active Spm ele- Spm element is present, either within
ment, and the overlapping patterns or just distal to the locus of the struc-
produced by combinations of states tural gene(s).
as alleles, appear in the Brookhaven The original state of axm-2. The
Symposia in Biology, Number 18. phenotype produced by the original
Analysis of the states of ar_1 was state of a±m~2 depends altogether on
not complicated. Simple rules could the phase of activity of each of the
be formulated that allowed prediction components of Spm, If component- 1
of the phenotype each state would is inactive, no pigment is produced
produce in response to changes in ac- in the plant or in the aleurone layer
tivity of the components of Spm. of the kernel. When this component is
These components — component- 1, the active, the gene is activated and pig-
suppressor, and component-2, the ment is produced. The type and dis-
mutator — were considered in Year tribution of the pigment depend on
Book 6J+. The distinctions among the activity of component-2. If this
states relate to the gene-associated component is inactive, the aleurone
element of the system, which resides layer is lightly pigmented. If com-
at the A1 locus. How this element ponent-2 acts only late in the develop-
operates at the level of the gene and ment of a kernel, then small, deeply
within the nucleus to accomplish such pigmented spots appear in a lightly
diverse modes of regulation of gene pigmented background. If it is active
action is not yet known. It is im- at all stages of development, both
portant to recognize, nevertheless, large and small deeply pigmented
that the gene-associated element of a spots appear in the lighter back-
two-element control system does pro- ground. The kernels shown in Plate
vide a means for directing a broad 1(A) illustrate responses of the initial
range of types and patterns of gene state of a±m-2 to these different phases
expression during development. of activity of the components of Spm.
666 CARNEGIE INSTITUTION
The phenotype produced by this state with others. Three kernels with inter-
does not depend solely on the action mediate but differing levels of back-
of the Spm element that is situated ground pigmentation are shown in
at or close to the locus. If either com- Plate 1(B), (C), and (D). Each of
ponent of that element is inactive, a the stable states is inherited in the
fully active Spm element located else- same manner as any stable mutant
where in the chromosome complement allele of the gene. Intralocus cross-
will induce the same types of re- over studies have shown that the
spouse. phenotype each produces relates to a
The responses of the original state component residing at a particular
of aS1''2 to phases of activity of com- site within the locus.
ponent- 1 of Spin, the suppressor or New states, either responding or
inhibitor component, are the reverse not responding to Spm, arise only
of those given by most of the states from those states that react to com-
of axm-\ Responses to component-2, ponent-2 of Spm. The event respon-
the mutator component, are alike, sible for an altered state must occur
The term "mutator" is applied to this in a cell of the germ line and the
component because it induces re- modified state must be included in a
spouses that modify the organization gamete. A zygote produced by the
of the locus and thus the expression functioning of this gamete will give
of the gene. One such modification is rise to a plant having the new state,
responsible for the origin of the The behavior of the state may then
different states that are under dis- be analyzed in this plant and its
cussion in this report. Some states progeny. With some states of both
are produced that respond to com- axm'x and a/1"2, responses to com-
ponent-1 of Spm but not to com- ponent-2 of Spm occur only very late
ponent-2. Others show no evidence of in development of a tissue. None may
a response to component-1 but do occur in the germ-line cells. Thus
respond to component-2. Still others, these states remain unaltered through
which respond to neither component, successive plant generations even in
are termed "stable states," and each the presence of an Spm element with
of them gives rise to a distinctive an early-acting component-2. All
phenotype in both plant and kernel. states are inherited unaltered if corn-
Only a few of the stable states de- ponent-2 is inactive or if its action is
rived from the original state of axm'2 effective only very late in develop-
resemble that of the Ar locus before ment.
control of its gene action was taken The derived states of axm'2. The
over by the Spm system. They pro- phenotypes produced by four modi-
duce deep anthocyanin pigmentation tied states of axm~2 appear in Plate
in both plant and kernel, and in the 1(E). A fully active Spm element,
kernel the pigment is uniformly dis- placed at or close to the Ax gene locus,
tributed over the aleurone layer. Most is present in each kernel. No visible
of the stable states give rise to a dis- evidence of a response to com-
tinctive pattern of pigment distribu- ponent- 1 of Spm is shown by any of
tion and intensity in plant and kernel, these states. The dark background
In the kernels, clusters of cells with pigmentation in kernels like the one
more intense pigment than the sur- on the left, and the colorless back-
rounding cells are distributed over grounds of kernels carrying the other
the aleurone layer. The intensity of three states, are the same regardless
the background pigmentation ranges of the phase of activity of compo-
from faint with some states to deep nent-1 in the kernel. In the kernel on
GENETICS RESEARCH UNIT 667
the left the small deeply pigmented colorless background. Some of the
spots represent one response of the stable derivatives of this state give
state present in the kernel to com- rise to kernels whose aleurone layer is
ponent-2. The size of such spots is al- uniformly pale-pigmented, but most
ways small. Another response to com- of them produce colorless kernels.
ponent-2 occurs early enough for the Five additional states, each derived
modified locus to be included in a from the initial state, have been ex-
number of gametes produced by a amined. None of them gives evidence
plant carrying this state. The modi- of the presence of an Spm element at
fications give rise to stable states the locus of the At gene. Each re-
that no longer respond to compo- sponds in a distinctive manner, how-
nent-2. In the kernel, the aleurone ever, to an active Spm element
layer is uniformly pigmented, and located elsewhere. The phenotypes
the intensity of color is the same as produced by two of these states in
the background pigmentation pro- the presence of a fully active Spm
duced by the parent state (the dark are shown in Plate 1(F). The two
kernel in the photograph) . left-hand kernels have one state.
The second kernel from the left Their deeply pigmented spots are
illustrates one of the responses of the distinguished from those in most of
state present in that kernel to com- the other kernels illustrated by a near
ponent-2. The response produces pig- absence of diffusion rims. Usually,
mented areas, within which the pig- when deeply pigmented spots or areas
ment distribution is similar to that in are formed, a product that diffuses
the kernels shown in (B), (C), and from the pigment-producing cells into
(D). A second type of response is the surrounding cells, and often
not visibly registered in the kernel through several rows of cells, allows
but is made apparent by the stable pigment to be formed there even
derivatives this state produces. Some though the Ax gene is not functioning
of them give rise to phenotypes re- in these cells.
sembling those in (B), (C),and (D). The common expression given by
Others, however, produce no pigment the other state shown in Plate 1(F)
in the aleurone layer of the kernel. is seen in the second kernel from the
The state of axm~2 present in the right. Small deeply pigmented spots
second kernel from the right is dis- appear in a colorless background,
tinguished by the range of phenotypic This state gives rise frequently to a
expressions brought about by its re- new state, characterized by a marked
sponses to component-2. Areas in the increase in the number of pigmented
aleurone layer exhibit phenotypes spots, as in the adjacent kernel on
similar to those produced by several the right.
other states. One such area is visible The remaining three states have
in this kernel; it has many deeply been examined in greater detail than
pigmented spots in a pale-pigmented most of the others, as they provide
background. Other kernels carrying information about the operation of
this state may have sharply defined the Spm system that could not be
areas containing a number of deeply deduced readily from the behavior of
pigmented spots in a colorless rather other states. Two of them, although
than a pale background. All the spots independently isolated, are so much
may be small, or some may be large alike that they may be considered
and some small. jointly. These states (7977B and
The kernel on the right in Plate 7995) furnished the initial evidence
1(E) has only very pale spots in a of the "presetting" and "erasure"
668 CARNEGIE INSTITUTION
mechanism that was outlined and were: (1) no active Spm element;
illustrated in Year Book 63 (pp. 592- (2) one Spm element with both
602) and further commented on in components fully active; (3) one Spm
Year Book 6^ (pp. 527-536). This element with an active component-1
aspect of their behavior will not be and a late-acting component-2; (4)
restated here. Another aspect, not two Spm elements, both as in (2);
discussed earlier, will be considered (5) two Spm elements, one as in (2)
in conjunction with the behavior of and one as in (3). In addition, some
the remaining state, 8004. These plants had an Spm element whose
three states differ greatly from the component-1 was inactive and re-
original state of Oi**"2 in their re- mained inactive in most plants and in
sponses to the changes in activity of their progeny, returning to the active
the Spm element that can occur in phase only very rarely; the returns
individual cells during development could be observed in small regions
of the endosperm of the kernel. within individual kernels. Numerous
It is recognized that the compo- tests were conducted with plants hav-
nents of Spm undergo changes in ing these different constitutions. For
phase of activity. Control of the time the purposes of this discussion, the
and frequency of their occurrence evidence obtained from only a few
resides in the Spm element itself, kinds of test need be mentioned,
each change regulating in a distinc- When state 8004 is propagated in
tive manner the period when the next the absence of an active Spm element,
will take place. The changes in phase the aleurone layer of the kernel is
of activity of component-1, from ac- colorless. If, however, an active Spm
tive to inactive and back to active, are element is present, pigment is pro-
unambiguously registered by each of duced. Its type and distribution then
the states of of**. The same unam- depend on the nature of the Spm and
biguity applies to the states of some the number of elements present. If
of the other gene loci that have come component-2 of the Spm element (s)
under the control of the Spm system, is inactive or becomes active only late
and it applies to the original state of in development, the aleurone layer is
axm-2. The phenotype that will appear uniformly light-pigmented. If this
after each phase change can be pre- component is active initially, different
dieted and the predictions validated phenotypic expressions of the gene
by tests that accurately determine the appear. When two or more such ele-
phase. Illustrations are given in the ments are present, pigment in the
Brookhaven Symposium paper men- aleurone layer may range from nearly
tioned earlier. The states of axm~2 colorless in some kernels to dark pale
designated 7997B, 7995, and 8004, in others, with, in most kernels, one
however, respond in a distinctly dif- or several very small deeply pig-
ferent manner to the alternating mented spots. If, however, a kernel
phases of activity of Spm. The dif- starts development with a single ac-
ference is especially well registered tive Spm element that undergoes
when the changes in phase occur dur- change in phase of activity in some
ing development of the kernel. State cells, early in development, then pig-
8004 will be discussed first. ment intensities in the aleurone layer
State S00U of axm'2. Examination of of the mature kernel are strikingly
this state was conducted with plants modified. Examples are seen in the
having different constitutions with kernels of Plate 2 (A) -(D). These
respect to the Spm element. These kernels have both large and small
GENETICS RESEARCH UNIT 669
areas outlined by rims of deep pig- mottled, as illustrated in Plate 2(C),
ment. The background pigmentation (D). Some of the mottling is due to
in all but one kernel is faint: such very small areas in which a diffus-
kernels were purposely selected for ible complementing substance is pro-
clear illustration of the deeply pig- duced. If a number of rimmed areas
mented borders. The kernel with a are present in a kernel, the intensity
darker background (B) is included of the background pigmentation
to suggest the range in intensity of within the areas is often the same.
background pigmentation among In some kernels, however, it may
these kernels. differ in one or more of the areas;
The rimmed areas consist of de- see Plate 2(B) and legend. It is sus-
scendants of individual cells in which pected that the differences among
a change in phase of Spm activity has kernels in the intensity of pigment
occurred. The deep pigment outlining within rimmed areas, exclusive of the
an area is produced by the outermost rims, reflect initial differences in or-
cells of the area. They receive a ganization of the a±m'2 locus in the
diffusible substance from the cells kernels, or differences that may arise
surrounding the area, in which Spm during kernel development,
is fully active. This substance allows That such differences do arise was
the border cells to make a pigment learned in tests conducted to deter-
that is more intense than that in the mine the cause of interruptions of a
cells either inside or outside the area, rim along a segment of an otherwise
Thus the deep pigment of the rims is rimmed area. Such interruptions, ex-
the result of a complementation reac- pressed as absence of deep pigment,
tion. The rims indicate that the prod- are often noted. Some of them are
uct of action of the Ar gene (or not continuous but are arranged in
genes) differs in the cells within and sequence along a continuous segment,
without the areas. The larger It was found that the interruptions
rimmed areas often contain small are due in some instances to loss of
areas that also are rimmed with deep ability of the adjacent outer cells to
pigment. Again, the rims are the produce a complementing diffusible
product of a complementation reac- substance that will allow the border
tion. Here, however, the diffusible cells of the rimmed area to form deep
substance comes from cells within pigment. The tests were made by
the small area and enters the cells introducing wxm'8 along with state
surrounding it, where the deep pig- 8004 of atm-2 into the primary endo-
ment is produced. It could be demon- sperm nucleus.
strated that the small rimmed areas The gene at the Wx (Waxy) locus
within the larger ones are composed functions to convert amylopectin into
of descendants of cells in which Spm amylose in the starch granules of
has returned to an active phase. Not cells in the endosperm. Changes in
all returns are made visible in this action of this gene during develop-
way. Some are accompanied by a ment may be detected in the mature
change at the a-^-2 locus that alters kernel visually and also quite pre-
its capacity to produce a substance cisely by staining the starch with
that can complement. an iodine-potassium iodide solution.
In some kernels the pigment with- Gene action at the wx"1-8 locus is
in a rimmed area is much more in- under the control of the Spm system,
tense than that outside the area. It and its responses to Spm result in
is not uniform in intensity, but the production of amylose in the
670 CARNEGIE INSTITUTION
starch granules of the cells. Because number of Spm elements present in
all the cells of the endosperm below the kernel. When one is present,
the aleurone layer — the outermost many areas exhibit this change, and
layer — have starch granules, each many of them are large. With two
change in action of the gene during Spm elements, the areas are all small.
endosperm development is registered With three elements, no areas or only
in the mature endosperm. The de- some very small ones are formed. The
scendants of a cell in which a change same relationship governs the produc-
has occurred form a well-defined sec- tion of the rimmed areas with state
tor in which the altered action of the 8004 of axm-2. Few or no rimmed areas
gene is expressed in every cell. Most are formed if two or more active Spm
of the large sectors terminate in the elements are present in the kernel.
aleurone layer. Small sectors, pro- Both large and small areas are pro-
duced by changes occurring late in duced if only one Spm element is
kernel development, also may termi- present.
nate in the aleurone layer. Thus, An additional aspect of the be-
when state 8004 of ajrr2 and wxm-8 havior of state 8004 should be men-
are both present in a kernel, the cell tioned. If a plant carrying this state
lineage of the aleurone layer over- and also one active Spm element is
lying such a sector, either large or utilized as pollen parent in a cross to
small, is also sharply defined. Parts a plant that lacks an active Spm ele-
of some of these sectors are adjacent ment, the kernels receiving the Spm
to parts of a rimmed area, and some- element from the pollen parent may
times these adjacent rim-area cells show large rimmed areas, and in
do not form deep pigment. The inter- some kernels the pigment within the
ruption of the rim is precisely defined areas is dark. Kernels that do not
by the common region of contact of receive the Spm element from the
cells of the sector with those of the pollen parent are colorless. Removal
rimmed area. It is evident that the of Spm from the nucleus by meiotic
sector is formed from descendants of segregation does not induce a setting
a cell in which changes have occurred of the locus that allows pigment of
coincidentally at the loci of axm'2 and various intensities to be produced
?/\-/'m-s. The change at axm-2 alters its subsequently among the kernels, as
ability to make a diffusible substance happens with states 7997B and 7995.
that can be utilized by the adjacent States 7997B and 7995 of atm-2. An
cells in the rimmed area to form in- earlier report (Year Book 63, pp.
tense pigment. 592-602) showed that kernels having
In Year Book 57 the first evidence one of these states and also one or
of phenotypic change produced by more fully active Spm elements de-
alternating cycles of activity of Spm velop a number of small, deeply pig-
was outlined. The studies were con- mented spots in a more lightly pig-
ducted with a selected state of a2m~1. mented background. Should only one
(A2 is another locus in maize whose Spm element be present and should
gene is involved in the biosynthetic this element undergo change in phase
pathway leading to anthocyanin for- of activity during kernel develop-
mation.) It was learned that the num- ment, the response of either state to
ber and size of pigmented areas in a the change results in a darkly pig-
kernel, produced as the consequence mented area in the aleurone layer,
of a change of Spm from an active to Many of these dark areas contain
an inactive phase, depend on the small colorless areas. Examples are
GENETICS RESEARCH UNIT 671
seen in the two kernels in Plate 2(E), gree of expression of the end product
each of which carries state 7995. of action of a series of genes is
Kernels with a faintly pigmented mediated through control of the ac-
background were chosen for the illus- tion of one of these genes. Only the
tration, so that the dark areas might fact that anthocyanin pigment — the
be noted readily. In the parts of the end product of such a series — is not
kernels where the Spm element was vital to the plant makes it possible
active, the characteristic pattern of to learn about the many kinds of reg-
small, deeply pigmented spots ap- ulation such a system can provide. It
pears. Such spots are absent in the is well known that the various races
darkly pigmented areas. Except for and strains of maize are distin-
the small colorless spots, the pigment guished from one another by a re-
within the dark areas is uniform in markable diversity with respect to
intensity. The edges of the areas are distribution of anthocyanin to parts
not defined by deeply pigmented rims, of the plant and its intensity in any
as are comparable areas in kernels one part. The patterns are so varied
carrying state 8004. Rims will appear that they defy a meaningful classifi-
if the two states 8004 and 7995 are cation. The same is true of the differ-
present as alleles in a kernel. An ex- ent distributions of pigment produced
ample is shown in Plate 2(F). Here by the states of a1m'1 and, especially,
the uniformly dark areas, produced the states of a1m-2. Studies conducted
by state 7995, are bordered by deeply by maize geneticists have shown that
pigmented rims, produced by state in some instances different alleles of
8004. The small colorless spot within a gene locus involved in anthocyanin
the area on the left is also bordered production are responsible for the ap-
by a pigmented rim resulting from pearance of different patterns of pig-
the action of state 8004. ment distribution. The kind of regu-
The pigment produced by state lation exercised by some of these
7995 when a change in phase of Spm alleles resembles that afforded by
activity occurs during kernel develop- some of the states described here. In
ment differs markedly both in inten- the studies of the alleles, with two
sity and in pattern of distribution exceptions, it has not been possible
from that produced by this state to determine the presence at the gene
when Spm is removed by a somati- locus of a control-mechanism compo-
cally occurring transposition of Spm nent that could be responsible for the
or by means of meiotic segregation, differences in action of the alleles.
In a kernel that develops from the That is understandable, for a means
functioning of a gamete that has lost of detecting such a component usually
the Spm element by such means, the is not available. It can be suspected,
aleurone layer has a distinctively dif- however, that many of these alleles
ferent pattern of pigment distribu- represent different "states" of the
tion, as illustrated in Year Book 63 loci, in the sense of the term defined
(pp. 592-602 and Plate 2) . in this report. Without a means of
This review of the states of a gene distinguishing between a mutant of
locus is intended to illustrate the the structural gene itself and a
extraordinary diversity in capacity of mutant that is produced by a regula-
a single system of controlling ele- tory component at the locus, the term
ments to regulate the action of a gene "allele" must be retained even though
during development. It also demon- its significance in any one instance
strates that the distribution and de- will remain ambiguous.
672
CARNEGIE INSTITUTION
Since 1962. the Brookhaven Na- tion of this courtesy, and of the gen-
tional Laboratory has provided gar- erous attitude and cooperative atten-
den space and cultivation facilities tion of those persons at Brookhaven
for growing my maize plants. I who are responsible for the care and
should like to express my apprecia- maintenance of plant materials.
BIBLIOGRAPHY
Butler. B.. see Skalka, A.
Echols, H., see Skalka, A.
Goldberg, E. B., The amount of DNA be-
tween genetic markers in phage T4. Proc.
Xatl. Acad. Sci. U.S., 56, 1457-1463, 1966.
Mosig. G., Distances separating genetic
markers in T4 DNA. Proc. Natl. Acad.
Sci. U.S., 56, 1177-1183, 1966.
Skalka. A., Multiple units of transcription
in phage lambda. Cold Spring Harbor
Symp. Quant. Biol, 31, 377-379, 1966.
Skalka, A., B. Butler, and H. Echols, Genetic
control of transcription during develop-
ment of phage \. Proc. Natl. Acad. Sci.
U.S., 58, 576-583, 1967.
Smith, M. G., A replicating form of X phage
DNA, in IX International Congress for
Microbiology, Symposia, Moscow, pp. 483-
492, 1966.
Smith, M. G., Isolation of high-molecular-
weight DNA from normal and phage-
infected E. coli, in Methods in Enzymol-
ogy, vol. 12, part A, L. Grossman and
K. Moldave, eds., Academic Press, pp.
545-550, 1967.
PERSONNEL
Year Ended June 30, 1967
Phyllis D. Bear, Carnegie Institution
Fellow
Elizabeth M. Bocskay, Chief Clerk
Jennie S. Buchanan, Curator of Drosoph-
ila Stocks
Elizabeth Burgi, Associate in Micro-
biology
Ruth Ehring, Carnegie Institution Fel-
low
Agnes C. Fisher, Secretary to Director;
Editor
Alfred D. Hershey, Director
Laura J. Ingraham, Research Assistant
Barbara McClintock, Cytogeneticist
Shraga Makover, Carnegie Institution
Fellow
Anna Marie Skalka, Carnegie Institu-
tion Fellow
Carole E. Thomason, Technical Assist-
ant
Rudolf Werner, Associate in Research
Temporary and Part-Time
John B. Earl, Technical Assistant
PLATES
Plate 1 . The kernels in (A), (E), and (F) are viewed from above. Those in (B), (C), and (D)
are viewed from the abgerminal side.
^A) The original state of a?'"2 is present in each kernel. The differences in phenotypic
expression reflect differences in the Spm elements they carry. From left to right: fully active
Spm; inactive component-1 ; active component-1 and late-acting component-2; active com-
ponent-! and inactive component-2.
[B— D) Mottled pigment distributions produced by three stable states of ci\m~2, each derived
from the original state. The pericarp layer has been removed in order to reveal more clearly
the pigment distribution in the aleurone layer.
[E] Phenotypic expressions engendered by four states of aim'2, each derived from the
original state. A fully active Spm element is present in each kernel.
(F) Phenotypes produced by three additional states of aim~2 in the presence of a fully
active Spm element. The two kernels on the left have one of these states. The adjacent
kernel to the right has another. This state, in turn, often gives rise to the one whose expression
is shown in the right-hand kernel.
Plate 1
(',<>.ric I u-:- f!( ■(■<), fh. TJrdt
A
i
B
D
I
Plate 2. The kernels in (A), (E), and (F) are viewed from the side; those in (B), (C), and (D)
from above. One Spm element that underwent change in phase of activity during kernel
development is present in each. Kernels in (A) through (D) have state 8004 of a\m~2; (E) has
state 7995. In (F) states 7995 and 8004 are both present as alleles.
(A) Two kernels that received the same Spm element from the pollen parent. Note the large
and small areas with deeply pigmented rims. Within the larger rimmed area in each kernel
are small rimmed areas. The background pigmentation in both kernels is faint.
(B) Rimmed areas in a kernel with dark background pigmentation. Note the lighter
pigment intensity within all but one of the rimmed areas. The pigment intensity in the small
rimmed area at the lower position is the same as that of the background.
(C, D) Rimmed areas in two kernels with nearly colorless backgrounds. Note pigment dis-
tributions and intensities within the rimmed areas. In (D) the rim of the large area is interrupted
along part of the lower edge.
(E) Two kernels from the same ear, both having state 7995 of ai'"'2. In each, the large
pigmented areas contain small colorless areas, which are not bounded by deep-pigmented
rims. Note the small, intensely pigmented spots in the areas of nearly colorless background,
and their absence in the large pigmented areas.
(F) Overlapping of phenotypes produced by states 8004 and 7995 of a?'""2 when both are
present in a kernel as alleles. The uniformly pigmented areas represent the action of state
7995, and their rims represent the action of state 8004, as does also the rim of the small
internal spot within the large area on the left.
Plate 2
Genetics Research Unit
B
C
D
;/CJs;:g|
Bibliography
July 1, 1966-June 30, 1967
PUBLICATIONS OF THE INSTITUTION
Carnegie Institution of Washington Year
Book 65. Octavo, xi + 71 + 630 pages,
20 plates, 249 figures, Washington, D. C,
December 1966.
The Search for Understanding, edited by
Caryl P. Haskins. Octavo, xxiv + 330
pages, 13 figures, Washington, D. C,
September 1967.
Drosophila Guide: Introduction to the
Genetics and Cytology of Drosophila
Melanogaster. Eighth edition. M. Demerec
and B. P. Kaufmann, in collaboration
with Jennie S. Buchanan (curator),
Agnes C. Fisher (editor), and Henry S.
Jones (photographer). Octavo, iii + 45
pages, 14 figures, July 1967.
PUBLICATIONS BY THE PRESIDENT
Caryl P. Haskins
Report of the President. Reprinted from
Carnegie Institution of Washington
Year Book 65, 71 pages, 4 plates,
9 figures, December 1966.
The Report of the President on the Six-
tieth Anniversary of the Carnegie In-
stitution for 1961-1962. Excerpts re-
printed from "Report of the President,"
Carnegie Institution of Washington
Year Book 61. Science and Society,
Norman Kaplan, ed., 1965, pp. 156-174.
Discussion, "Behaviour of Social Insects,"
by Edward O. Wilson. Insect Behaviour.
Symposia of the Royal Entomological
Society of London, P. T. Haskell, ed.,
1966, No. 3, pp. 93-94.
Two faces of science. Ventures, Vol. 6,
No. 2, Fall 1966, pp. 21-28.
The Scientific Revolution and World
Politics (see Year Book 63, p. 615),
translated into Arabic by Al Karnak
Publishing House, Cairo, Egypt, Feb-
ruary 1967.
Message from the incoming President of
the Society of the Sigma Xi. American
Scientist, Vol. 55, No. 1, March 1967,
pp. 1-2.
Books and papers and communication in
science. Mills College Magazine, pp.
2-11, Spring 1967.
Mr. Jefferson's sacred gardens. The Vir-
ginia Quarterly Review, Vol. 43, No. 4,
Autumn 1967, pp. 530-544.
Some challenges of the future for Sigma
Xi. American Scientist, Vol. 55, No. 4,
December 1967, pp. 361-374.
PUBLICATIONS BY THE
EXECUTIVE OFFICER
Edward A. Ackerman
Population, natural resources, and tech-
nology. The Annals of the American
Academy of Political and Social Science,
Vol. 369, January 1967.
Population and natural resources. World
Population Conference, 1965, Volume 1:
Summary Report, United Nations, 1966.
Review of L 'Organisation de Vespace.
Elements de geographic volontaire, by
Jean Labasse. Hermann, Paris, 1966. In
Science, Vol. 157, No. 3792, September
1, 1967, pp. 1031-1032.
673
Report of the Executive Committee
To the Trustees of the Carnegie Institution of Washington
Gentlemen :
In accordance with the Provisions of the By-Laws, the Executive Committee
submits this report to the Annual Meeting of the Board of Trustees.
During the fiscal year ending June 30, 1967, the Executive Committee held
four meetings. Printed accounts of these meetings have been or will be mailed
to each Trustee.
The estimate of expenditures for the fiscal year beginning July 1, 1967, has
been reviewed by the Executive Committee.
The terms of office of the Vice-Chairman and Secretary of the Board of
Trustees, and the Chairmen of all Committees of the Board expire at the annual
meeting of the Board of Trustees on May 5, 1967. The terms of office of the
following members of Committees also expire on May 5, 1967 :
Executive Committee Retirement Committee
Amory H. Bradford Garrison Norton
Finance Committee Nominating Committee
Alfred L. Loomis Garrison Norton
Henry S. Morgan
Robert A. Lovett
Auditing Committee
Keith S. McHugh
Alfred L. Loomis
Juan T. Trippe
May 5, 1967 Henry S. Morgan, Chairman
675
Report of Auditors
Lybrand, Ross Bros. &, Montgomery
To the Auditing Committee of Carnegie Institution of Washington:
We have examined the statement of assets, liabilities and funds balances of Carnegie
Institution of Washington as of June 30, 1967, and the related summary statement of changes in funds
for the year then ended and the supporting exhibits and schedules, which have been prepared on the
general basis of cash receipts and disbursements and accordingly do not reflect accrued income,
accounts payable nor provision for depreciation. Our examination was made in accordance with
generally accepted auditing standards, and accordingly included confirmation from the custodian of
securities owned at June 30, 1967, and such tests of the accounting records and such other auditing
procedures as we considered necessary in the circumstances. We previously examined and reported
upon the financial statements of the Institution for the year ended June 30, 1966.
In our opinion, the accompanying financial statements and supporting exhibits and schedules
present fairly the assets, liabilities and funds balances of Carnegie Institution of Washington at
June 30, 1967 and 1966, and the changes in funds for the year ended June 30, 1967, on the basis
described above consistently applied.
/Upt (/^^^AluJ^^
Washington, D. C.
August 10, 1967
677
STATEMENT A ASSETS, LIABILITIES, AND FUNDS BALANCES
JUNE 30, 1967 and 1966
ASSETS
967 1 966
Cash $ 467,505.87 $ 720,674.22
Advances 40,694.08 33,062.31
Investments (cost) - Schedule 2:*
Mortgage 18,682.65 20,845.95
Governmental obligations 2,907,695.31 4,709,608.47
Nongovernmental bonds 41,152,380.99 42,724,498.21-
Corporate stocks 33,890,108.41 29,891,824.79
Land (cost) 362,147.71 362,147.71
Buildings and equipment (cost) 6,378,397.45 6,180,464.58
Prepaid insurance 25,926.97
Total assets $85,217,612.47 $84,669,053.21
LIABILITIES AND FUNDS
Liabilities:
Income taxes, etc., withheld $ 354.54 $ .6,743.84
Funds:
Operating Fund - Exhibit 1 898,044.58 1,110,057.77
Restricted Grants - Exhibit 2 106,318.98 147,473.81
Endowment and Special Funds - Exhibit 3 77,472,349.21 76,862,165.50
Land, Building, and Equipment Fund - Exhibit 4 6,740, 545. 16 6, 542,612.29
Total funds 85,217,257.93 84,662,309.37
Total liabilities and funds $85,217,612.47 $84,669,053.21
Approximate market value on June 30, 1967: $109,187,766.
678
r*.
ro
On
o
ro
ro
o
rs
-
nO
no
■«r
00
v>
LO
LO
LO
LT)
NO
r-C
o
ro
o
o
in
ON
•r
o
*■
00
ro • •
LO
ro
SO
00
ro
O
f 1
r t
00
NO
00
00
ro
o
On
LO
T
o~-
NO
o
r--
nO
ro
-
«
CS —
o t
co —
— rs
i_n ro
NO LO
•*T ro
■vT —
LO —
nO
LO
ro
ro
00
00
LO
rs
CO
LO
nO
LO
LO
rs
T
LO
CO
*J
.»!<?
ON
-O — Q. <->
rS
C T3 ~ ~
RJ — 2.^
NO
cqU-1 X
rs
LLI
T
T3 ^
LO
c
CO
NO
</>
00 ro
•vT LO
ro —
O nO
On LO
TT no"
h>. ro
rs
. TT
•«r
CO
NO
rs
• ro
LO
rs
ro
LO
rs
ro
O
LO
T
LO
ro
ro
o
NO
Q
Z
->
LL
1^
NO
Z
ON
to
o*
LLI
ro
z
<
X
u
LU
Z
— >
( i
LL
III
O
Q
h-
Z
z
LU
LLI
fY!
i.
<
LLI
III
H
V
<
r-
LLI
to
J_
>-
H
c£
ct.
<
o
£;
LL
z
Z3
CO
o
w
LO
4J "(_> /-N
LO
NO
c cm
<0 3 ^
E LL •-
£-£_£
rs
^•^
NO
CO
iS »B
no*
CO
&>
"° rs
o:
^_^
t->
M —
r^
c
•
rati
und
ibit
LO
o
<D LL .c
«
cl"" x
o
O LLI
—
w
a,
w
LO
LO
O
ON
LO
o
^r
LO
LO
NO
ro
NO
00
ro
rs
CO
NO
CO
CO
o
ro
■nT
ON
On"
LO
NO
CD
o o
o o
r»' d
no rs
O ro
oo" w
LO
rs —
O O
o o
r- o
no rs
O ro
co"
LO
NO
"""
r^
ro
ro
ro
00
O
^r
O
00
rs
NO
. . "T
ON
On
. . rs
. ^- ..... .
ON
ON
rs
ro
00
o
NO
rs
ON
ro
rs
T
r^
»■•%.
"V
CO
rs
co"
LO
T
ro
O
r»"
NO
ro
no"
00
o
^-^
r>»
ro
00
r^
CO
ON
T
-^r
00
00
L0
—
rs
—
ro
00
_
NO
LO
T
o
■nT
^^
*&
v>
CD
LLI
LU
H
co
NO
NO
ON
o
c
«J
CO
ao
E
rci
o
*->
u
r
u.
CO
u
c
TJ
<L»
<U
b
, #
N
4— '
irt
(/}
C
o
>
c
*->
VL
-o
-o
<
0)
■a
-E Q -Z
-a
N
TO TO
<n
l/l
E
o
CD
-1
u
■»-»
c
u
1—
0)
CU
a;
or (J O LLI
tj t: 2
U 0.
X
T3
- ^ LL
4_, <U LL.
TO 00 (u M
Q. => TO «
o OQ X oo
a.
o
i_
a.
Q.
TO
<L>
*->
TO
U
_o
"to
c
3
Q.
Q.
<
TO
C
CD
E
Q.
*3
cr
CD
3 O
(/)
00
O ci ^
•Z3 X -^
o LLI Q
V
a
TO
CD
>%
BO
c
<D
c
TO
o
NO
ON
o
ro
a>
c
cu
u
c
_ro
co
QQ
679
EXHIBIT I CHANGES IN OPERATING FUND
FOR THE YEAR ENDED JUNE 30, 1967
Balance July I, 1966 $1,110,057.77
Appropriations — Statement B:
Budget, July I, 1966
to June 30, 1 967 - Exhibit 3 $4,158,067.00
Hale Fund 320.00
Transfers - Statement B:
Unallocated appropriations (284, 033. 16) 3,874, 353.84
Total available for expenditures 4, 984, 41 1.61
Expenditures:
Salaries 1,985,143.42
Laboratory 430,516.43
Employee benefits 395,379.91
Equipment 232, 123. 16
Fellowship program 189,420.26
Operating 154,666.82
Building 128,674.09
Publications 124,180.61
Awards 98,349.24
Travel 85,302.62
Consulting fees and insurance 76,460.76
Financial advisory services 71,551.65
Taxes 58,665.36
Shop 24,023.11
Rent 20,525.21
Dormitory 5,943.37
Entertainment 5, 441 .01
Total expenditures 4,086,367.03
Balance June 30, 1967 $ 898,044.58
680
EXHIBIT 2
CHANGES IN RESTRICTED GRANTS
FOR THE YEAR ENDED JUNE 30, 1967
Expenditures
American Cancer Society.
Anonymous
Carnegie Corporation of
New York
General Services
Administration ....
Holt, Rinehart and
Winston, Inc
Jet Propulsion Laboratory
National Aeronautics and
Space Administration .
National Science
Foundation
Office of Naval Research.
Public Health Service . .
University of Chile . . .
Balance
July I, 1966
$ 103.02
14,500.00
Grants
Salaries
Other
$ 103.02
14,500.00
121,075.34 $ 80,000.00
6,900.00
1,214.65
479.47
73.65
45.31
Balance
June 30, 1967
87,702.97 $113,372.37
5,444.35 1,455.65
1 , 1 4 1 . 00
434. 16
18.076.00 $9,206.97 23,182.12 (14,313.09)
1,383.48
58.8!
6,362.40
2,296.64
216,000.00
28,341.64
54, 154.00
13,658.30
4,300.42
13,220.79
25,338.87
2,700.00
207,514.37
24,870.90
26,826.09
13,254.94
5,568.69
(9,691.24)
8,351.44
Total $147,473.81 $417,129.94 $54,767.05 $403,517.72 $106,318.98*
* Does not include grants to be received as follows:
National Aeronautics and Space Administration . . . $281,924.00
National Science Foundation 668,331.94
Office of Naval Research 35,066.36
$985,322.30
681
o
o
ro
T sO
O ON
fN •
sO
4 J"n
o
o
T
O T
o rs
o
o
00
CO vO
O ON
'
u —
o
ON
fN O
O fN
r-^ •
c
o
CM
LO —
O CO
so
~ o
_*
-
*
»
»
~^ ft
o
00
CO rs
O CO
ra
o
o
ON
ON O
O no
00 • . . .
0Q -
o
o
00
fN
in in
f-
c
3
fN
o
o
in
fN
N"
— '
rs
m
m
fN
CO o
fN CO
fN
O
00
so r^
O fN
r^ in
r~» fN
— o\ (>
in SO — SO
NT
fN
o
m
ON
in
n-
00
T
fN
IjN
o
ON
in
m
no
fN
SO
On
00
ON
no
o
o
fN
CN
CN
fN
ON
N"
m
LD
IT)
nt
00
00
V3
On
ON
ON
no
fN
NU
Lit
NT
Ql
o
a.
a.
<
o
o
so
so
in
fN
w
. . o
o
o
o
O
O
o
o
o
o
O
o
vO
CO
in
LT)
O
ON
LP
ON
o
ON
r--
00
ro
Os
o
rN
CO
LD
on
Q
i/)
Z
!t
73
(0
LL
sO
_l
ON
<
u
o
m
Ql
LU
00
Z
Q
Z>
/
<
Q
UJ
H
D
z
Z
UJ
LU
>
%
c£
n
<
n
LU
7"
>-
■o
LU
LU
7"
__
'JZ
\-
(Z
00
LU
O
u.
<
sJ
■5 £
4->
c
u
<D
E
E
*->
o
i/i
( )
u
c
>
c
S c
so
so
o
ON
u
—
c
a
— _
a
Nv
DU
3
—i
sO
O
fN
hs
in
fN
fN
fN
m
v>
so
N"
ON
fN
CO
o
o
o
o
sO
m
fN
00
fN
SO
o
o
o
o
o
o
T
o
m
fN
in
ro
o
o
o
o
o
o
o
o
m
in
CO
o
in
fN
fN
fN
o"
no
LO
so r-.
o fN
r^ m
h> rs
fN
O
CO
in so — so
ro
m
ON
o
"r
o
o
o
o
o
O CO
O no
m in
fN
fN
N"
CO
fN
ro
m
fN
CO
CO*
N"
so
00
N"
sO
o
ON
ro
CO
sO
fN
N"
LO
NO
ON
no
LO
in
so
fN
ro
ON
LO
no
(NO
in
NO
NO
OO
no
n-
N-
ON
ON
NO
fN
CO
N"
LO
rN
o"
fN
ON
o
N-"
s-
rN
O
rN
O
LTl
o
CO
NO
N"
ro
o-
NO
m
in
nC'
fN
rN
rN
ro
rN
no
OD
in
in
m
so
o
m
fN
LO
LP
ON
rN
LO
00
00
o
o
LO
so
• ON
NT
o
no
N"
o
rN
in
no
in
SO
fN
ON
CO
hC
ON
on"
o
n-
LD
fN
SO
00
SO
LL.
4->
4^
<u
<i>
c
c
TD
r:
c
Z3.
c
o
ra
LL
rs
+j
«>o
00
o
ra
—
~n
Ml
o
-^
ra
ro
o
gg
o
U
o
>-
Q.
ra
o
ex
ro
U
a.
ro
u
U
<u
■o
T3
TJ
a>
2
N
U
N
— (U
U
-j:
X
111
Jf c ra
<u
= S < U
or
$ J^-
-g O
c
LU
<
«f O cr: —
« a>
LL
ra
U
O w
p ra
a) LU 0. ^
oi
V)
a>
■»-> ••«... •
"ra
>-.
o
o «5
'Zi • • o
u ^ P ^
— 3 O C >n
■S II c ' fc ' > » '
3 o^-3 ra — ^,
u- mbw — —u-OJO
— i/iuororaajo
^<CQUXX|-5:
CL
00
682
u
ID
CD
sO
sO
no
LD
~"
CN
00
ro
m
no
CO
sO
O
no
no
ON
sO
sO
rS
S"
CN
rS
rs
o
LO
r-
0s
r-
r-
o
no
CO
00
rs
SO
rs
rs
00
ON
no
r-
rs
-T
SO
r-
rs
LO
rs
no
rs
r-
r-
CN
rs
O
CO
SO
r-
o
o
LO
CO
no
LO
sO
no
-sT
LO
r-
r-
■sr
rs
On
rs
rs
no
sO
Os
CO
</,
r-
rs
Is-
CO
CO
CN
LD
o
r-
r-
o
CN
CO
r-
rs
CN
-sT
LO
rs
CN
sO
CO
LD
•sT
rs
rs
r-
rs
Is-
» no
LD
o
rs
rs
S3
no
•z:
r-
3
ll
<D
sO
CN
IS*
u
c
C3
o
m
LU
vO
(0
i.
ON
CQ
c
Q.
Z3
3
rv
o
m
'
LU
LU
Q
<*>
LU
-7
U
z
Li
in
1
cC
ID
<
OJ
LU
a.
>-
1 \
7"
LU
<
X
_J
H
■7"
oct
o
to
LL.
LU
KD
u
OJ
Q-
X
LU
SO
so
<u
rN
u
__
c
ro
ro
CQ
rs
rs
O
ON
rs
o
—
SO
"sT
ro
no
CN
o
sO
LD
—
rs
CN
rs
sO
o
ON
T
in
CN
—
ON
O
ro
o
CN
T
—
rs
ro
sO
CO
sO
sO
Ln
h-
"sT
IT)
LO
Is-
CO
o
o
sO
"T
rs
ON
sO
no
-sT
rs
CO
rs
rs
ro
rs
O
r-
__
_
sO
</»
</»
LO
rs
sO <
Ln
o
ON
LO
CO
sO
"sT
•sT
Ln
o
r-
rs .
rs
SO
o
CN
CO
r-
CN
Ln
rs
no
Ln
in
ro
rs
r-.
ZZZ
so
ro
&
trt
sO
CO
ro
rS
Is-
so
CN
—
ro
CO
T
so
rs
no
o
o
O
r-
CO
Ln
Is-
no
rs
m
LO
00
ro
so
ro
m
ro
sO
00
r-
in
so
CN
r-
OS
T
Is-
o
ON
-sT
00
Is-
—
rs
LO
—
CO
CO
ro
no
v»
W
ro
o
o
Is-
CO
r-
rs
Ln
Is-
CN
CN
o
CN
rs
SO
CN
no
CN
m
SO
no
LO
CO
ro
O
rs
ON
CO
so
s-
Is-
o
sO
ro
rs
so
■sT
rs
Ln
Ln
CO
no
no
r-
OS
o
no
o
Is-
rs
Is-"
O
o
rs
T
m
_
—
sO
<&
w
so rs
— no
no
O
rs
CO
—
Is-
rs
rs
no
■<r
rs
tf>
ro
LO
sO
in
so
no
in
so
CQ
X
X
LU
o
CQ
4->
c
_ro
0.
c
E
j->
>_
ro
a.
a
Q
o
ro
_J
"ro
u
Q.
O
<u
(J
O
c
DO
ro
0)
Q
>-
O
O
>s
E
LU
1)
Q
u
0)
E
<
o
E
O
E
Q.
-o
c
0)
Q.
X
U
LL
c
ro
U
ro
U
X
UJ
ro
u
N-
Q_
o
»~m
2 g
a 3
a. — i
<
ON
on
sO
CN
rs
in
o
in
sO
so
in
CD
rs
ON
in
in
NT
in
o
sO
in
ro
m
ON
ro
CM
CN
rs no
co m
rs rs
no rs
o r^
sO O
ON — O
— — o
^ o\ 6
on on in
on in r^-
t in —
r^
NO
sO
ON
—
NT
h-
vO
ro
Sfi
h«.
no
NO
sfi
( )
00
LO
O
r^
NO
NT
QQ
—
o
NO
ro
no
o
h
—
o
o
r--
00
T
ON
o
ro
V
00
NO
in
un
NO
CN
no
M
CN
00
NO
ON
m
CO
in
in
no
no
O
no
f
00
CN
ON
NO
NO
_
m
in
00
o
00
—
o
—
NO
—
h*
sO
lo
—
no
NO
m
• in
rN
r^
00
m
N"
—
NO
ON
sO
o
ON
^r
ON
m
o
o
ON
co
o
• in
so
SO
CTN
in
O
00
in
NO
—
in
00
O
r--
—
^r
_
o
in
—
r-
. r^
o>
h»
—
no
NT
r^
00
V
<*-
</)
Q
CO
2
NO
h-
L_
ON
o
.
z
o
no
h-
UJ
<
Z
tiC
Z)
LU
Q_
O
Q
LU
LL
Q
O
Z
>-
LU
oC
Od
<
<
y
UJ
i.
>-
en
LU
X
h-
H
lU
\J
on
"O
Q
o
<u
~>
U-
-a
CO
c
<u
Q.
X
0)
CO
4->
c
<v
r
E
m
*->
o
<
CO
c
c
c
l_
Q.
O
i_
CL
Q.
<
CO
c
sO
o
sO
—
ON
CO
—
fc_
•
2^
—
o
>*
n
Q.
_)
<
1
00
rs
in
in
rs
in
no
lo
m
CN
ro
ON
On
ON
N"
so
o
00
CO
ON
SO
no
so
CO
ON
m
no
ro
in
o
o
v_x
-.^
v^^
-
lO
1 '
ro
m
ro
ro
SO
ON
LO
lo
CN
o
m
ON
CO
ro
•nT
CO
ro
NO
in
CN
no
PO
rs
ho
m
n0
O
NO
rs
CO
LT)
CN
ON
v>
CN
CN|
O
LO
r^
CN
so
o
rs
^r
ON
o
NO
oo
LT)
ro
ro
•sr
On
no
CN
r*.
LO
CN
o
vO
o
—
LP)
ON
m
CN
ON
r-
ON
uo
N"
CN
CN
sO
sO
o
V*
N-
s-*^
•ST-
CN
00
ON
CO
00
cxi
o
^r
U~)
o
^r
U")
—
sO
NO
o
m
o
m
no
o
m
CO
LP,
ro
sO
ON
NO
—
o
CN
no
^r
r^
h«.
ro
O
en
no
ON|
NT
LO
ON
o
CO
o
in
m
LO
_r
so"
—
—
in m o
— -sr in
on o n-
CN so" CN
*s CN N"
CN
SO O so
CN 00 no
O
r^
.
rsj — O
in n n
of) tCO
00
ON
no so CN
CN
no
ON
o o
o o
o o
o o
o o o
o o o
o
o
o
o
o
o
o o
o o
r^ in
o o
o o
o in
o o o
o in rs
in no no
o
00
o
o
o
00
no
V CN
cn m
215,
77,
T CO
ocT
ON
00
CD
00*
in
N"
CO o
m no
r^ rN
ON so
CN co rN ON
no
T On so
CN —
no
no
on
CN
CO
o-i
NT
no
NT
in
r-
CN
r-*
o
O
r^
no
<«■
—
—
_
CO
CO
CN
NT
CO
ON
CN
no
</>
LU
-I
=>
Q
LU
X
u
>s o
l_ 4->
o a
t! >
ra
x _
o
<U 4-.
oo a)
-Cl cz
O °o
5 X
o» M
: .2 2
uo >■ *- —
o
= CQ P
f3
Q.
Q
S o
a c «;
0. ^ Z r- LU
o
H
a.
o
— <u
i_ £ <u
u
CO 0)
co
10
2 <J or
o
o
1-
1)
Cl
X
(U
ra
k.
OJ
c
OJ
M
C C 0)
ro o a>
— *♦_
O CO W)
ro to ^
w C 3
W •— (A
■- E c
c -a o
E < nJ
■o
<
a; .2
Q. >
o -o
0)
i >
**■ a*
txo W
- ro o Z ■- ■-
CO
__
0J
ro
a)
"u
—
<L>
<\)
CX
OJ
C
o o
CL Cl.
E E
O — <L>
U LL U LU LU
ro
4->
o
I-
Q.
O
i—
o.
Q.
CO
•o
<L>
■•->
ro
u
_o
"ro
c
=)
o
h
SCHEDULE 2
INVESTMENTS, JUNE 30, 1967
Principal
Amount Description
Mortgage
$ 18,682.65 Alfred D. Hershey and Harriet D.
Hershey, 5V2 s
Federal Agencies and
United States Government Bonds
$1,025,000 Federal National Mortgage Association,
4V8s
400,000 Federal National Mortgage Association,
4s/8s
500,000 Federal National Mortgage Association,
5V8s
1,000,000 Federal National Mortgage Association,
Part. Certificates, 5V2 s
$2,925,000 Total
Approximate
Market
Maturity Book Value Value
1974 $ 18,632.65 $ 18,683
1970
1970
1972
973
$1,015,070.31 $ 978,875
394,500.00 388,000
498,125.00 490,000
1,000,000.00 990,000
$2,907,695.31 $2,846,875
Foreign and International
Bank Bonds
$ 700,000 Alberta Government Telephone,
Commission, Deb., 434 s
750,000 Alcan Aluminum Corporation,
Prom. Note, 4% s
489,000 Aluminum Co. of Canada, Ltd.,
S. F. Deb.,4^s
146,000 Australia (Commonwealth of), AV2 s . .
114,000 Australia (Commonwealth of), 5 s . . .
466,000 Australia (Commonwealth of), 5V2 s . .
750,000 Bell Telephone Co. of Canada,
1st Mtg. Series "X," 478 s ....
250,000 British Columbia Power Commission,
S. F.Deb. Series "L," 43/8s . . .
750,000 Industrial Acceptance Corp. Ltd.,
Sec. Note Series "Z," 5*4 s . . .
125,000 Intl. Bank for Reconstruction &
Development, 3 s
125,000 Intl. Bank for Reconstruction &
Development, 3% s
250,000 Intl. Bank for Reconstruction &
Development, 20 Yr. Bonds, AV2 s .
150,000 Noranda Mines Ltd., S. F. Deb., 434 s .
782,000 Quebec Hydro-Electric Commission,
S. F. Deb., 5s
200, 000 Shawinigan Water & Power Co., I st Mtg.
& Collat. Tr. S. F. Series "M," 3 s
1989
$ 700,000.00 <
574,000
1984
750,000.00
648,750
1980
495,016. 12
429,098
1971
143,810.00
139,795
1972
1 14,000.00
109.440
1982
468, 163.76
438, 040
1988
747,300.00
630,000
1987
245,000.00
202,500
1982
750,000.00
637,500
1976
125,000.00
104,375
1975
123, 125.00
107,500
1977
250,000.00
227,500
1968
150,000.00
137,438
1988
768,315.00
660,790
1971
201,440.00
178,500
685
INVESTMENTS - Continued
Approximate
Principal Market
Amount Description Maturity Book Value Value
Foreign and International
Bank Bonds (Continued)
SI. 000, 000 Shell Funding Corp., Collat. Tr.
Series "B," 434s 1985 $1,000,000.00 $ 860,000
5CC. CCC Toronto (Municipality of Metropolitan),
S. F. Deb., 5s 1979 498,637.50 457,500
S 125,
000
250,
000
237,
000
300,
000
4,
000
200,
000
250,
000
400,
000
100,
000
200,
000
250,
000
250,
000
236,
000
200,
000
400,
000
500,
000
250,
000
200,
000
500,
000
216,
000
300,
000
$5,368
,000
S~. 547. 000 Total $7,529,807.38 $6,542,726
Public Utility Bonds
Columbia Gas System, Inc., Series "B,"
3 s 1975 $ 126,047.25 $ 99,375
Columbia Gas System, Inc., Series "F,"
3-ss 1981 245,937.50 200,000
Columbus & Southern Ohio Electric Co.,
IstMtg., 3^s 1970 239,188.65 215,670
Consolidated Edison Co. of N. Y.,
1st & Ref. Mtg. Series "N," 5s . . . 1987 301,567.28 270,000
Consumers Power Co., 1st Mtg., 434 s . . 1987 4,017.16 3,530
Minnesota Power & Light Co., I st Mtg.,
3'eS . . 1975 201,276.09 163,000
Niagara Mohawk Power Corp., Gen. Mtg.,
359s 1986 252,089.82 193,125
Niagara Mohawk Power Corp., Gen. Mtg.,
4',s 1987 402,338.67 364,000
Ohio Power Co., 1st Mtg., 3*/4s 1968 101,500.00 97,125
Pacific Gas & Electric Co., 1st & Ref.
Mtg. Series "X," 3l/8s 1984 200,991.58 143,000
Pacific Gas & Electric Co., 1st & Ref.
Mtg. Series "BB," 5s 1989 251,336.54 230,000
Pacific Power & Light Co., 1st Mtg.,
436s 1986 252,031.99 199,375
Potomac Electric Power Co., Deb., 4%s . 1982 239,625.96 212,695
Public Service Co. of Indiana, 1st
Mtg. Series "F," 3V8s 1975 201,397.91 165,000
Public Service Co. of Indiana, 1st
Mtg. Series "L," 474S 1987 400,000.00 342,000
Public Service Electric & Gas Co.,
1st & Ref. Mtg., 47/8s 1987 503,201.09 455,000
Southern California Edison Co., I st &
Ref. Mtg. Series "H," 4 V4s 1982 251,093.80 200,625
Southern California Edison Co., I st &
Ref. Mtg. Series "J," 47/8 s 1982 201,311.01 175,000
Tenneco Inc., Deb., 5s 1982 503,750.00 427,500
Tenneco Inc., 1st Mtg. Pipe Line, 5V4s. . 1977 216,000.00 201,960
Washington Water Power Co., 1st Mtg.,
476s 1987 300,000.00 262,500
Total $5,394,702.30 $4,620,480
686
INVESTMENTS - Continued
Principal
Amount Description
Communication Bonds
$ 400,000 Illinois Bell Telephone Co., 1st
Mtg. Series "E," 4V4s
200,000 Mountain States Telephone &
Telegraph Co., Deb., 3V8s
100,000 New York Telephone Co., Ref. Mtg.
Series "E," 3V8s
200, 000 Pacific Telephone & Telegraph Co.,
Deb., 3y4s
250,000 Southern Bell Telephone & Telegraph
Co., Deb., 4s
300,000 Southwestern Bell Telephone Co.,
Deb., 3V8s
$1,450,000 Total
Railroad Bonds
$ 100,000 Chesapeake & Ohio Railway Co., Gen.
Mtg., 4V2s
267,000 Fort Worth & Denver Railway Co., 1st
Mtg., 438s Guar
.$ 367,000 Total
Industrial and Miscellaneous Bonds
$ 242,000 Aluminum Co. of America, S. F. Deb.,
4V4s
1 , 000, 000 Aluminum Co., of America, Conv. Sub.
S. F. Deb., 5V4s
262,500 Bethlehem Steel Corporation, Sub. Deb.,
4V2s
1,000,000 Boeing Co., Notes, 63/8s
550,000 C. I. T. Financial Corp., Deb., 43/4s . .
750,000 Colonial Pipeline Co., Sec. Note
Series "A," 434 s
1,000,000 Columbia Broadcasting System, Inc.,
Prom. Note, 5%s
400,000 Commercial Credit Co., Note, 3%s . . .
700,000 Commercial Credit Co., Note, 434s. . .
30,000 Corn Products Co., Sub. Deb., 458s . .
375,000 Crown Zellerbach Corp., Prom. Note,
43/8s
547,000 Erie Mining Company, I st Mtg.
Series "B," 4%s
500,000 FMC Corp., S. F. Deb., 3.8s
500,000 First National City Bank, Capital
Conv. Notes, 4s
Maturity Book Value
Approximate
Market
Value
1988 $ 403,406.96 $ 333,000
1978 200,420.00 160,000
1978 100,493.07 73,250
1978 201,324.34 154,000
1983 250,727.82 206,875
1983 302,750.00 222,000
1982.
1991
1990
1986
1970
1990
1991
1976
1982
1983
1981
1983
1981
1990
$1,459,122.19 $1,154,125
1992 $ 99,500.00 $ 81,000
1982 268,292.95 210,930
$ 367,792,95 $ 291,930
$ 242,000.00 $ 208, 120
1,000,000.00 I, 167,500
183,637.50
,000,000.00
536,937.50
750,000.00
,000,000.00
403,865.56
700,000.00
30,973.38
375,000.00
531,071.36
500,000.00
228,375
970, 000
528,000
635,625
932,500
324,000
586,250
26,400
328, 125
475,890
415,000
512,215.00 463,750
687
Principal
Amount
INVESTMENTS - Continued
Description.
Industrial and Miscellaneous Bonds
(Continued)
Maturity
Approximate
Market
Book Value Value
222.000 Four Corners Pipe Line Co., Sec.
Note. 5 s
500,000 General Electric Credit Corp. (N.Y.),
Prom. Note, 5 s
500, 000 General Electric Credit Corp. (N.Y.),
Sub. Note, 434 s
200,000 General Motors Acceptance Corp.,
Deb., 3%s. ..'... :
480,000 General Motors Acceptance Corp.,
Deb., 4 s
1 , 000, 000 General Motors Acceptance Corp.,
Deb., 47,sS
200,000 General Motors Acceptance Corp.,
Deb., 5s
200,000 General Motors Acceptance Corp.,
Deb., 5s
150,000 General Portland Cement Co.,
Conv. Sub. Deb., 5s
750,000 Household Finance Corp., Deb., 4% s .
1,000,000 Hystron Fibers, Inc., Notes> 53/4s . . .
423,038.68 Instlcorp, Inc., Collat. Tr. Note, A— 16 .
353,422.45 Instlcorp, Inc., Collat. Tr. Note, A-19.
205,888.09 Instlcorp, Inc., Collat. Tr. Note, A-21 .
258,802.21 Instlcorp, Inc., Collat. Tr. Note, A-23 .
808,299.31 Instlcorp, Inc., Collat. Tr. Note, A-36 .
400, 000 Intl. Harvester Credit Corp., Deb.,4%s .
251 , 000 Kaiser Aluminum & Chemical Corp.,
1st Mtg., 5^s
700,000 Kresge (S..S.). Company, Prom. Note,
47aS
200,000 Montgomery Ward Credit Corp., Deb.,
478s
95,000 National Dairy Products Corp., Deb.,
23/< s
700,000 Owens-Illinois, Inc., Notes, 5s •. . . .
150,000 Scovill Mfg. Co., Deb., 43/4s
525,000 Sears Roebuck Acceptance Corp., Sub.
Deb., 4%s
1,000,000 Shell Oil Company, Deb., 5s
250,000 Spiegel, Inc., Deb., 5s
475,000 Statewide Stations Inc., Sec. Note,45/gs
215, 000 Talcott (James) Inc., Senior Note, 5V2s .
700,000 Texas Gulf Sulphur Co., Prom. Note,
4.7s
582,381.01 Trailer Train Co., 47;s
340, 000 Trans World Airlines, Inc., Conv. Sub.
Deb., 4s
348,000 Tremarco Corporation, 1st Mtg. Series
"E," 5s . . . .
1982 J
& 222,000.00
$ 202, 575
1975
500,000.00
457, 500
1987
500,000.00
396,250
1972
200,409.14
177,750
1979
435,037.50
393,600
1987
990,000.00
865, 000
1977
195,000.00
185,500
1981
199,000.00
183,500
1977
154,500.00
126,000
1993
746,250.00
626,250
1986
1,000,000.00
957,500
1991
408,907.72
393,425
1991
341,728.29
326,915
1991
198,682.03
190,446
1991
254,609.73
239,392
1992
775,288.01
719,386
1979
398,000.00
332,000
1987
251,000.00
222,135
1983
700,000.00
609,000
1980
199,000.00
168,250
1970
94,802.62
85,975
1991
700,000.00
593,250
1982
147,686.32
133,500
1977
511,505.00
459,375
1991
1,000,000.00
895, 000
1987
250,000.00
220, 000
1994
475,000.00
401,375
1980
212,850.00
189,200
1989
700,000.00
593,250
1976
582,381.01
556, 174
1982
1992
371,445.00
313,650
1983
348,000.00
320,160
INVESTMENTS - Continued
Approximate
Principal Market
Amount Description Maturity Book Value Value
Industrial and Miscellaneous Bonds
(Continued)
$ 1,000,000 Union Carbide Corporation, S. F.
Deb., 5.3 s
700,000 United Air Lines, Inc., Notes, 5s .
680,000 United Shoe Machinery Corp., S. F.
Deb., 53/4s
542,500 U. S. Steel, S. F. Sub. Deb., 4s/8s .
I 000,000 Westinghouse Electric Corp., Deb..
53/8s
250,000 Whirlpool Corporation, S. F. Deb.,
3V2s
500,000 Woolworth (F. W.) Company, Prom.
Note, 5 s
$26,711,831.75 Total
1997
$ 1,000,000.00
$ 972, 500
1984
700,000.00
614,250
1992
678,300.00
666,400
1996
443,873.50
457,734
1992
1,000,000.00
950,000
1980
250,000.00
195,000
1982
500,000.00
427, 500
$26,400,956. 17
$24,106,202
$44,060,076.30
$39,562,338
Approximate
Market
Book Value
Value
$44,368,831.75 Bonds, funds invested . .
Number
of
Shares Description
Common Stocks
28,200 Alcan Aluminum Limited $ 828,977.7! $ 771,976
30 Allied Chemical Corp 1,235.29 1,133
25,600 American Cyanamid Co 850,430.03 771,200
23,110 American Electric Power Co., Inc 162,703.74 855,070
24,700 American Smelting & Refining Company . . . 1,360,570.39 1,732,088
42,352 American Telephone & Telegraph Company . 1,161,275.59 2,414,064
16,000 Armstrong Cork Company 131,908.39 826,000
20 Anaconda Company 926.88 953
10,000 Avon Products, Inc 826,947.35 1,057,500
31,200 Burlington Industries, Inc 494,681.48 1,029,600
24,000 Caterpillar Tractor Co 97,534.09 1,068,000
11,000 Celanese Corp 568,604.02 613,250
24,000 Chesebrough-Pond's Inc 746,765.43 870,000
16,500 Chicago Pneumatic Tool Co 601,964.31 707,438
32,000 Clark Equipment Co 772,393.17 1,032,000
19,600 Coca-Cola Company (The) 628,984.09 2,415,700
57 Columbia Gas System, Inc 1,503.38 1,532
30 Consolidated Natural Gas 900.00 859
18,000 Continental Oil Company (Del.) 146,960.65 1,199,250
2,500 Corning Glass Works 59,631.83 801,250
3,180 E. I. du Pont de Nemours & Co 477,281.33 477,795
689
INVESTMENTS- Continued
\..~re- Approximate
of Market
Shares Description Book Value Value
Common Stocks (Continued)
20.586 Eastman Kodak Company $ 233,651.32 $ 2,748,231
12.000 Federated Department Stores, Inc 582,805.81 799,500
15.104 First National City Bank 348,278.77 817,504
19.400 Ford Motor Company 577,047.36 991,825
30,337 General Electric Company 740,512.49 2,635,527
35.419 General Motors Corporation 1,143,847.99 2,722,836
36.200 Gillette Company 1,239,112.08 1,981,950
21,712 Goodyear Tire & Rubber Company .... 488,401.64 947,186
20,006 Gulf Oil Corporation 154,333.51 1,292,888
20 Hugoton Production Co 840.00 865
15.834 International Business Machines Corp. . . 851,095.87 7,877,41.5
24.470 International Nickel Co. of Canada Ltd. . 1,296,074.87 2,266,534
12.900 Johnson & Johnson 750,762.93 999,750
17.000 Kellogg Company 332,482.68 639,625
45.630 Kennecott Copper Corporation 1,208,688.57 2,024,831
8 Liggett & Myers Tobacco Company . . . . 598.00 573
15,000 May Department Stores Company 859,378.02 476,250
15.000 Merck& Co., Inc 107,286.55 1,230,000
30,700 Mobil Oil Corporation 1,099,916.18 1,220,325
40.000 Niagara Mohawk Power Corp 863,803.67 860,000
34,000 Ohio Edison Co 587,855.31 905,250
16,400 Panhandle Eastern Pipe Line Co 431,553.54 551,450
9,000 Pennsylvania Railroad Company 614,232.68 616,500
15,000 Philip Morris Incorporated 493,240.88 705,000
4,600 Sears, Roebuck and Co . 207,078.03 253,000
22.361 Standard Oil Co. (New Jersey) 578,043.17 1,364,021
33,300 Stevens (J. P.) & Company 988,349.72 1,436,063
24,190 Texaco Inc 249,172.89 1,693,300
14,500 Texas Gulf Sulphur Co 1,257,455.80 1,774,438
7,600 Texas Utilities Co 104,621.78 406,600
19,400 Travelers Corp. (The) . . 372,113.68 654,750
14,000 United Aircraft Corporation 1,223,902.37 1,508,500
26,400 United Air Lines, Inc 1,077,231.61 1,917,300
28,400 U. S. Plywood-Champion Papers Inc.. . . 697,928.16 1,476,800
16,000 Virginia Electric & Power Co 220,973.07 674,000
20,000 Whirlpool Corporation 943,953.26 850,000
6.000 Xerox Corp 1,041,305.00 1,639,500
090,556 Common stocks, funds invested . . . $33,890, 108.41 $ 69,606,745
Aggregate investments $77,968,867.36 $109, 187,766
690
SUMMARY OF INVESTMENT TRANSACTIONS
FOR THE YEAR ENDED JUNE 30, 1967
Investments July I, 1966
$77,346,777.42
Sales and Redemptions
Mortgage ....
Bonds
Preferred stocks
Common stocks .
Capital
Capital
Gain
Loss
Book Value
$ 2,163.30
$ 5,846.21
$1
,227,794.89
109,125.33
15,349,805.74
410,805.00
1,024,474.50
1,056.04
3,491,426.91
19,254,200.95
1,030,320.71
1
,337,976.26
Realized capital loss,
net - Statement B . . 307,655.55
$1,337,976.26 $1,337,976.26
Amortization
Acquisitions
Bonds 11,982,936.94
Common stocks 7,900,515.53
Investments June 30, 1967 ,
7,161.58
58,085,414.89
19,883,452.47
$77,968,867.36
691
Abstract of Minutes
of the Sixty-Ninth Meeting of the Board of Trustees
The annual meeting of the Board of Trustees was held in the Board Room
of the Administration Building on Friday, May 5, 1967. Chairman James N.
White called the meeting to order.
The following Trustees were present: Vannevar Bush, Carl J. Gilbert, Caryl P.
Haskins, Alfred L. Loomis, Keith S. McHugh, Henry S. Morgan, William I.
Myers, Garrison Norton, Richard S. Perkins, Elihu Root, Jr., William W. Rubey,
Frank Stanton, Charles P. Taft, Charles H. Townes, Juan T. Trippe, and James
N. White.
The minutes of the Sixty-Eighth Meeting were approved.
The Chairman notified the Trustees of the death of Walter S. Gifford and of
Barklie McKee Henry. Mr. Root and Mr. Morgan spoke of the Trustees' high
esteem for Mr. Gifford and for Mr. Henry, and of their services and contributions
to the Institution. Mr. Root proposed the following resolution, which was unani-
mously adopted by the Trustees :
Be It Resolved, That the Trustees of the Carnegie Institution of Washington
desire to record their sorrow and deep sense of loss at the death of their friend
and fellow member, Walter S. Gifford.
And Be It Further Resolved, That this resolution be entered on the minutes
of the Institution and that a copy be sent to Mrs. Gifford.
Mr. Morgan proposed the following resolution, which was also unanimously
adopted by the Trustees :
Be It Therefore Resolved, That we, the Trustees of the Carnegie Institution
of Washington, record our grief and deep sense of loss at the death of our
friend and companion, Barklie McKee Henry.
And Be It Further Resolved, That this resolution be entered on the minutes
of the Institution and that a copy be sent to Mrs. Henry.
Sir Eric Ashby and William McChesney Martin, Jr., were elected members of
the Board of Trustees.
By unanimous action, Seeley G. Mudd was reelected a member of the Board of
Trustees (Article I, Section 2, of the By-Laws).
Mr. Morgan was reelected Vice-Chairman of the Board, and Mr. Norton was
reelected Secretary of the Board, both for terms ending in 1970.
The following were elected for one-year terms : Henry S. Morgan as Chairman
of the Executive Committee, Richard S. Perkins as Chairman of the Finance
Committee, Carl J. Gilbert as Chairman of the Nominating Committee, Keith S.
McHugh as Chairman of the Auditing Committee, and Omar N. Bradley as
Chairman of the Retirement Committee.
Vacancies in standing committees, with terms ending in 1970, were filled as
follows: Amory H. Bradford was elected a member of the Executive Committee,
693
694 CARNEGIE INSTITUTION
Alfred L. Loomis, Henry S. Morgan and Keith S. McHugh were elected members
of the Finance Committee, Crawford H. Greenewalt and Keith S. McHugh were
elected members of the Nominating Committee, Keith S. McHugh, Alfred L.
Loomis and Juan T. Trippe were elected members of the Auditing Committee,
and Garrison Norton was elected a member of the Retirement Committee.
The reports of the Executive Committee, the Finance Committee, the Auditing
Committee, and the Retirement Committee were accepted.
The resignations of two Trustees, Mrs. Margaret Carnegie Miller and Mr.
Elihu Root, Jr., were reported and accepted with regret.
To provide for the operation of the Institution for the fiscal year beginning
July 1, 1967, and upon recommendation of the Executive Committee, the sum of
S4, 490, 137 was appropriated, the appropriation to be made specifically in the
amount of $4,455,690 from the Working Capital Fund, $1,160 from the Bush
Fund, $10,046 from the Colburn Fund, $390 from the Hale Fund, $430 from the
Harkavy Fund, $981 from the Teeple Fund, and $21,440 from the Harry Oscar
Wood Fund.
The annual report of the President was accepted.
Articles of Incorporation
JfiftjHiglrtjr Congress of % SMieo Slates of America;
^t tlxje j&ejtond Session,
Begun and held at the City of Washington on Monday, the seventh day of December, one
thousand nine hundred and three.
.A.3ST ACT
To incorporate the Carnegie Institution of Washington.
Be it enacted by the Senate and House of Representatives of the United
States of America in Congress assembled. That the persons following, being persons
who are now trustees of the Carnegie Institution, namely, Alexander Agassiz,
John S. Billings, John L. Cadwalader, Cleveland H. Dodge, William N. Frew,
Lyman J. Gage, Daniel C. Gilman, John Hay, Henry L. Higginson, William
Wirt Howe, Charles L. Hutchinson, Samuel P. Langley, William Lindsay, Seth
Low, Wayne MacVeagh, Darius 0. Mills, S. Weir Mitchell, William W. Morrow,
Ethan A. Hitchcock, Elihu Root, John C. Spooner, Andrew D. White, Charles
D. Walcott, Carroll D. Wright, their associates and successors, duly chosen, are
hereby incorporated and declared to be a body corporate by the name of the
Carnegie Institution of Washington and by that name shall be known and have
perpetual succession, with the powers, limitations, and restrictions herein contained.
Sec. 2. That the objects of the corporation shall be to encourage, in the
broadest and most liberal manner, investigation, research, and discovery, and
the application of knowledge to the improvement of mankind; and in particular —
(a) To conduct, endow, and assist investigation in any department of
science, literature, or art, and to this end to cooperate with governments,
universities, colleges, technical schools, learned societies, and individuals.
(b) To appoint committees of experts to direct special lines of research.
(c) To publish and distribute documents.
(d) To conduct lectures, hold meetings, and acquire and maintain a library.
(e) To purchase such property, real or personal, and construct such building
or buildings as may be necessary to carry on the work of the corporation.
695
696 CARNEGIE INSTITUTION
(f) In general, to do and perform all things necessary to promote the
objects of the institution, with full power, however, to the trustees hereinafter
appointed and their successors from time to time to modify the conditions and
regulations under which the work shall be carried on, so as to secure the
application of the funds in the manner best adapted to the conditions of the time,
provided that the objects of the corporation shall at all times be among the
foregoing or kindred thereto.
Sec. 3. That the direction and management of the affairs of the corporation
and the control and disposal of its property and funds shall be vested in a board
of trustees, twenty-two in number, to be composed of the following individuals:
Alexander Agassiz, John 8. Billings, John L. Cadwalader, Cleveland H. Dodge,
William X. Frew, Lyman J. Gage, Daniel 0. Gilman, John Hay, Henry
L. Higginson. William Wirt Howe, Charles L. Hutchinson, Samuel P.
Langley, William Lindsay, Seth Lowr, Wayne MacVeagh, Darius 0. Mills,
S. Weir Mitchell, William W. Morrow, Ethan A. Hitchcock, Elihu Hoot,
John C. Spooner, Andrew D. White, Charles D. Walcott, Carroll D. Wright,
who shall constitute the first board of trustees. The board of trustees shall
have power from time to time to increase its membership to not more than
twenty-seven members. Vacancies occasioned by death, resignation, or otherwise
shall be filled by the remaining trustees in such manner as the by-laws shall
prescribe ; and the persons so elected shall thereupon become trustees and also
members of the said corporation. The principal place of business of the said
corporation shall be the city of Washington, in the District of Columbia.
Sec. 4. That such board of trustees shall be entitled to take, hold and
administer the securities, funds, and property so transferred by said Andrew
Carnegie to the trustees of the Carnegie Institution and such other funds or
property as may at any time be given, devised, or bequeathed to them, or to such
corporation, for the purposes of the trust ; and with full power from time to time to
adopt a common seal, to appoint such officers, members of the board of trustees or
otherwise, and such employees as may be deemed necessary in carrying on the
business of the corporation, at such salaries or with such remuneration as they may
deem proper; and with full power to adopt by-laws from time to time and such rules
or regulations as may be necessary to secure the safe and convenient transaction
of the business of the corporation; and wTith full power and discretion to deal
with and expend the income of the corporation in such manner as in their
judgment will best promote the objects herein set forth and in general to have
and use all powers and authority necessary to promote such objects and carry out
the purposes of the donor. The said trustees shall have further power from time
ARTICLES OF INCORPORATION 697
to time to hold as investments the securities hereinabove referred to so transferred
by Andrew Carnegie, and any property which has been or may be transferred
to them or such corporation by Andrew Carnegie or by any other person,
persons, or corporation, and to invest any sums or amounts from time to time
in such securities and in such form and manner as are permitted to trustees
or to charitable or literary corporations for investment, according to the laws
of the States of New York, Pennsylvania, or Massachusetts, or in such securities
as are authorized for investment by the said deed of trust so executed by Andrew
Carnegie, or by any deed of gift or last will and testament to be hereafter made
or executed.
Sec. 5. That the said corporation may take and hold any additional
donations, grants, devises, or bequests which may be made in further support of
the purposes of the said corporation, and may include in the expenses thereof
the personal expenses which the trustees may incur in attending meetings or
otherwise in carrying out the business of the trust, but the services of the
trustees as such shall be gratuitous.
Sec. 6. That as soon as may be possible after the passage of this Act a
meeting of the trustees hereinbefore named shall be called by Daniel C. Gilman,
John S. Billings, Charles D. Walcott, S. "Weir Mitchell, John Hay, Elihu Eoot,
and Carroll D. Wright, or any four of them, at the city of Washington, in
the District of Columbia, by notice served in person or by mail addressed to
each trustee at his place of residence; and the said trustees, or a majority
thereof, being assembled, shall organize and proceed to adopt by-laws, to elect
officers and appoint committees, and generally to organize the said corporation;
and said trustees herein named, on behalf of the corporation hereby incorporated,
shall thereupon receive, take over, and enter into possession, custody, and
management of all property, real or personal, of the corporation heretofore known
as the Carnegie Institution, incorporated, as hereinbefore set forth under "An Act
to establish a Code of Law for the District of Columbia, January fourth, nineteen
hundred and two," and to all its rights, contracts, claims, and property of any
kind or nature; and the several officers of such corporation, or any other person
having charge of any of the securities, funds, real or personal, books or property
thereof, shall, on demand, deliver the same to the said trustees appointed by this
Act or to the persons appointed by them to receive the same; and the trustees
of the existing corporation and the trustees herein named shall and may take
such other steps as shall be necessary to carry out the purposes of this Act.
Sec. 7. That the rights of the creditors of the said existing corporation
known as the Carnegie Institution shall not in any manner be impaired by the
69S
CARNEGIE INSTITUTION
passage o'L this Act, or the transfer of the property hereinbefore mentioned, nor
shall any liability or obligation for the payment of any sums due or to become
due. or any claim or demand, in any manner or for any cause existing against
the said existing corporation, be released or impaired; but such corporation hereby
incorporated is declared to succeed to the obligations and liabilities and to be held
liable to pay and discharge all of the debts, liabilities, and contracts of the said
corporation so existing to the same effect as if such new corporation had itself
incurred the obligation or liability to pay such debt or damages, and no such action
or proceeding before any court or tribunal shall be deemed to have abated or been
discontinued by reason of the passage of this Act.
Sec. 8. That Congress may from time to time alter, repeal, or modify this
Act of incorporation, but no contract or individual right made or acquired shall
thereby be divested or impaired.
Sec. 9. That this Act shall take effect immediately.
President of the Senate pro tempore.
By-Laws of the Institution
Adopted December 13, 1904. Amended December IS, 1910, December 13, 1912, December
10, 1937, December 15, 1939, December 13, 1940, December 18, 1942, December 12, 1947,
December 10, 1954, October 24, 1957, May 8, 1959, May 13, I960, May 10, 1963,
May 15, 1964, and March 6, 1967.
ARTICLE I
The Trustees
1. The Board of Trustees shall consist of twenty-four members with power to increase its
membership to not more than twenty-seven members. The Trustees shall hold office contin-
uously and not for a stated term.
2. In case any Trustee shall fail to attend three successive annual meetings of the Board
he shall thereupon cease to be a Trustee.
3. No Trustee shall receive any compensation for his services as such.
4. All vacancies in the Board of Trustees shall be filled by the Trustees by ballot at an
annual meeting, but no person shall be declared elected unless he receives the votes of two-
thirds of the Trustees present.
5. If, at any time during an emergency period, there be no surviving Trustee capable of
acting, the President, the Director of each existing Department, and the Executive Officer,
or such of them as shall then be surviving and capable of acting, shall constitute a Board of
Trustees pro tern, with full powers under the provisions of the Articles of Incorporation and
these By-Laws. Should neither the President, nor any such Director, nor the Executive
Officer be capable of acting, the senior surviving Staff Member of each existing Department
shall be a Trustee pro tern with full powers of a Trustee under the Articles of Incorporation
and these By-Laws. It shall be incumbent on the Trustees pro tern to reconstitute the Board
with permanent members within a reasonable time after the emergency has passed, at which
time the Trustees pro tern shall cease to hold office. A list of Staff Member seniority, as
designated annually by the President, shall be kept in the Institution's records.
ARTICLE II
Officers of the Board
1. The officers of the Board shall be a Chairman of the Board, a Vice-Chairman, and a
Secretary, who shall be elected by the Trustees, from the members of the Board, by ballot
to serve for a term of three years. All vacancies shall be filled by the Board for the unexpired
term; provided, however, that the Executive Committee shall have power to fill a vacancy
in the office of Secretary to serve until the next meeting of the Board of Trustees.
2. The Chairman shall preside at all meetings and shall have the usual powers of a pre-
siding officer.
3. The Vice-Chairman, in the absence or disability of the Chairman, shall perform the
duties of the Chairman.
4. The Secretary shall issue notices of meetings of the Board, record its transactions, and
conduct that part of the correspondence relating to the Board and to his duties.
699
700 CARNEGIE INSTITUTION
ARTICLE III
Executive A dm in titration
The President
1. There shall be B President who shall be elected b}r ballot by, and hold office during the
pleasure of. the Board, who shall be the chief executive officer of the Institution. The Presi-
dent, subject to the control of the Board and the Executive Committee, shall have general
charge of all matters of administration and supervision of all arrangements for research and
other work undertaken by the Institution or with its funds. He shall prepare and submit to
the Board of Trustees and to the Executive Committee plans and suggestions for the work
of the Institution, shall conduct its general correspondence and the correspondence with
applicants for grants and with the special advisers of the Committee, and shall present his
recommendations in each case to the Executive Committee for decision. All proposals and
requests for grants shall be referred to the President for consideration and report. He shall
have power to remove, appoint, and, within the scope of funds made available by the Trustees,
provide for compensation of subordinate employees and to fix the compensation of such
employees within the limits of a maximum rate of compensation to be established from time
to time by the Executive Committee. He shall be ex officio a member of the Executive
Committee.
2. He shall be the legal custodian of the seal and of all property of the Institution whose
custody is not otherwise provided for. He shall sign and execute on behalf of the corporation
all contracts and instruments necessary in authorized administrative and research matters
and affix the corporate seal thereto when necessary, and may delegate the performance of
such acts and other administrative duties in his absence to the Executive Officer. He may
execute all other contracts, deeds, and instruments on behalf of the corporation and affix
the seal thereto when expressly authorized by the Board of Trustees or Executive Committee.
He may, within the limits of his own authorization, delegate to the Executive Officer author-
ity to act as custodian of and affix the corporate seal. He shall be responsible for the expendi-
ture and disbursement of all funds of the Institution in accordance with the directions of the
Board and of the Executive Committee, and shall keep accurate accounts of all receipts and
disbursements. Following approval by the Executive Committee he shall transmit to the
Board of Trustees before its annual meeting a written report of the operations and business
of the Institution for the preceding fiscal year with his recommendations for work and
appropriations for the succeeding fiscal year.
3. He shall attend all meetings of the Board of Trustees.
4. There shall be an officer designated Executive Officer who shall be appointed by and
hold office at the pleasure of the President, subject to the approval of the Executive Com-
mittee. His duties shall be to assist and act for the President as the latter may duly authorize
and direct.
5. The President shall retire from office at the end of the fiscal year in which he becomes
sixty-five years of age.
ARTICLE IV
Meetings and Voting
1. The annual meeting of the Board of Trustees shall be held in the City of Washington,
in the District of Columbia, in May of each year on a date fixed by the Executive Committee,
or at such other time or such other place as may be designated by the Executive Committee,
or if not so designated prior to May 1 of such year, by the Chairman of the Board of Trustees,
or if he is absent or is unable or refuses to act, by any Trustee with the written consent of the
majority of the Trustees then holding office.
2. Special meetings of the Board of Trustees may be called, and the time and place of
meeting designated, by the Chairman, or by the Executive Committee, or by any Trustee
BY-LAWS 701
with the written consent of the majority of the Trustees then holding office. Upon the written
request of seven members of the Board, the Chairman shall call a special meeting.
3. Notices of meetings shall be given ten days prior to the date thereof. Notice may be
given to any Trustee personally, or by mail or by telegram sent to the usual address of such
Trustee. Notices of adjourned meetings need not be given except when the adjournment u
for ten days or more.
4. The presence of a majority of the Trustees holding office shall constitute a quorum for
the transaction of business at any meeting. An act of the majority of the Trustees present
at a meeting at which a quorum is present shall be the act of the Board except as otherwise
provided in these By-Laws. If, at a duly called meeting, less than a quorum is present, a
majority of those present may adjourn the meeting from time to time until a quorum is
present. Trustees present at a duly called or held meeting at which a quorum is present may
continue to do business until adjournment notwithstanding the withdrawal of enough
Trustees to leave less than a quorum.
5. The transactions of any meeting, however called and noticed, shall be as valid as though
carried out at a meeting duly held after regular call and notice, if a quorum is present and if,
either before or after the meeting, each of the Trustees not present in person signs a written
waiver of notice, or consent to the holding of such meeting, or approval of the minutes
thereof. All such waivers, consents, or approvals shall be filed with the corporate records or
made a part of the minutes of the meeting.
6. Any action which, under law or these By-Laws, is authorized to be taken at a meeting
of the Board of Trustees may be taken without a meeting if authorized in a document or
documents in writing signed by all the Trustees then holding office and filed with the Secretary.
7. During an emergency period the term "Trustees holding office" shall, for purposes of
this Article, mean the surviving members of the Board who have not been rendered incapable
of acting for any reason including difficulty of transportation to a place of meeting or of com-
munication with other surviving members of the Board.
article v
Committees
1. There shall be the following Standing Committees, viz. an Executive Committee, a
Finance Committee, an Auditing Committee, a Nominating Committee, and a Retirement
Committee.
2. All vacancies occurring in the Executive Committee, the Finance Committee, the
Auditing Committee, the Nominating Committee, and the Retirement Committee shall be
filled by the Trustees at the next regular meeting. In case of vacancy in the Finance Com-
mittee, the Auditing Committee, the Nominating Committee, or the Retirement Committee,
upon request of the remaining members of such Committee, the Executive Committee may
fill such vacancy by appointment until the next meeting of the Board of Trustees.
3. The terms of all officers and of all members of Committees, as provided for herein, shall
continue until their successors are elected or appointed.
Executive Committee
4. The Executive Committee shall consist of the Chairman, Vice-Chairman, and Secretary
of the Board of Trustees, the President of the Institution ex officio, and, in addition, not less
than five or more than eight Trustees to be elected by the Board by ballot for a term of
three years, who shall be eligible for re-election. Any member elected to fill a vacancy shall
serve for the remainder of his predecessor's term.
5. The Executive Committee shall, when the Board is not in session and has not given
specific directions, have general control of the administration of the affairs of the corporation
and general supervision of all arrangements for administration, research, and other matters
undertaken or promoted by the Institution. It shall also submit to the Board of Trustees a
702 CARNEGIE INSTITUTION
printed or typewritten report of each of its meetings, and at the annual meeting shall submit
to the Board a report for publication.
0. The Executive Committee shall have power to authorize the purchase, sale, exchange,
or transfer of real estate.
Finance Committee
7. The Finance Committee shall consist of not less than five and not more than six mem-
bers to be elected by the Hoard of Trustees by ballot for a term of three years, who shall be
eligible for re-election.
S. The Finance Committee shall have custody of the securities of the corporation and
general charge of its investments and invested funds, including its investments and invested
funds as trustee of any retirement plan for the Institution's staff members and employees,
and shall care for and dispose of the same subject to the directions of the Board of Trustees.
It shall have power to authorize the purchase, sale, exchange, or transfer of securities and
to delegate this power. It shall consider and recommend to the Board from time to time
such measures as in its opinion will promote the financial interests of the Institution and of
the trust fund under any retirement plan for the Institution's staff members and employees,
and shall make a report at each meeting of the Board.
Auditing Committee
9. The Auditing Committee shall consist of three members to be elected by the Board of
Trustees by ballot for a term of three years.
10. Before each annual meeting of the Board of Trustees, the Auditing Committee shall
cause the accounts of the Institution for the preceding fiscal year to be audited by public
accountants. The accountants shall report to the Committee, and the Committee shall
present said report at the ensuing annual meeting of the Board with such recommendations
as the Committee may deem appropriate.
Nominating Committee
11. The Nominating Committee shall consist of the Chairman of the Board of Trustees
ex officio and, in addition, three Trustees to be elected by the Board by ballot for a term of
three years, who shall not be eligible for re-election until after the lapse of one year. Any
member elected to fill a vacanc}- shall serve for the remainder of his predecessor's term, pro-
vided that of the Nominating Committee first elected after adoption of this By-Law one
member shall serve for one year, one member shall serve for two years, and one member
shall serve for three years, the Committee to determine the respective terms by lot.
12. Sixty days prior to an annual meeting of the Board the Nominating Committee shall
notify the Trustees by mail of the vacancies to be filled in membership of the Board. Each
Trustee may submit nominations for such vacancies. Nominations so submitted shall be
considered by the Nominating Committee, and ten days prior to the annual meeting the
Nominating Committee shall submit to members of the Board by mail a list of the persons
so nominated, with its recommendations for filling existing vacancies on the Board and its
Standing Committees. No other nominations shall be received by the Board at the annual
meeting except with the unanimous consent of the Trustees present.
Retirement Committee
13. The Retirement Committee shall consist of three members to be elected by the Board
of Trustees by ballot for a term of three years, who shall be eligible for re-election, and the
Chairman of the Finance Committee ex officio. Any member elected to fill a vacancy shall
serve for the remainder of his predecessor's term.
14. The Retirement Committee shall, subject to the directions of the Board of Trustees,
be responsible for the maintenance of a retirement plan for staff members and employees
BY-LAWS 703
of the Institution and act for the Institution in its capacity as trustee under any such plan,
except that any matter relating to investments under any such plan shall be the responsibility
of the Finance Committee subject to the directions of the Board of Trustees. The Committee
shall submit a report to the Board at the annual meeting of the Board.
ARTICLE VI
Financial Administration
1. No expenditure shall be authorized or made except in pursuance of a previous appro-
priation by the Board of Trustees, or as provided in Article V, paragraph 8, hereof.
2. The fiscal year of the Institution shall commence on the first day of July in each year.
3. The Executive Committee shall submit to the annual meeting of the Board a full
statement of the finances and work of the Institution for the preceding fiscal year and a
detailed estimate of the expenditures of the succeeding fiscal year.
4. The Board of Trustees, at the annual meeting in each year, shall make general appro-
priations for the ensuing fiscal year; but nothing contained herein shall prevent the Board
of Trustees from making special appropriations at any meeting.
5. The Executive Committee shall have general charge and control of all appropriations
made by the Board. Following the annual meeting, the Executive Committee may allocate
these appropriations for the succeeding fiscal year. The Committee shall have full authority
to reallocate available funds, as needed, and to transfer balances.
6. The securities of the Institution and evidences of property, and funds invested and
to be invested, shall be deposited in such safe depository or in the custody of such trust
company and under such safeguards as the Finance Committee shall designate, subject to
directions of the Board of Trustees. Income of the Institution available for expenditure
shall be deposited in such banks or depositories as may from time to time be designated
by the Executive Committee.
7. Any trust company entrusted with the custody of securities by the Finance Committee
may, by resolution of the Board of Trustees, be made Fiscal Agent of the Institution, upon
an agreed compensation, for the transaction of the business coming within the authority of
the Finance Committee.
8. The property of the Institution is irrevocably dedicated to charitable purposes, and
in the event of dissolution its property shall be used for and distributed to those charit-
able purposes as are specified by the Congress of the United States in the Articles of In-
corporation, Public Law No. 260, approved April 28, 1904, as the same may be amended
from time to time.
ARTICLE VII
Amendment of By-Laws
1. These By-Laws may be amended at any annual or special meeting of the Board of
Trustees by a two-thirds vote of the members present, provided written notice of the pro-
posed amendment shall have been served personally upon, or mailed to the usual address
of, each member of the Board twenty days prior to the meeting.
Index
Numbers in italic type refer to the Report of the President.
Abbott, James, 638, 644
Abelson, Philip H., vi, 38, 43, 54, 55, 70, 320,
326, 548, 512, 551, 570, 663
studies, 526-528.
report of the Director, Geophysical Lab-
oratory, 317-570
Abt, Helmut A., 265
Ackerman, Edward A., viii
publications, 673
Adams, J. L., 64, 130
publication, 125, 126
Adams, L. H.
publication, 124
Adelman, Saul J., 267, 314
Agassiz, Alexander, xi, 695, 696
Akimoto, S., 348, 560
Aldrich, L. Thomas, vi, 46, 3, 7, 28, 42, 129,
320
publications, 125, 126
studies, 12, 528-536
Alfani, M., 346, 560
Allen, M. B., 196
publication, 197
Allen, Terence, 258
Aller, Lawrence H., 269, 295
Allmann, Rudolf, 320, 325
studies, 485-487
Alperin, R. J., 644
Amagat, E. H., 540, 560
Amesz, Jan, 67,68, 144, 145, 245
publications, 165, 171, 243, 244
studies, 149-155, 155-160, 160-165, 165-
171
Anderko, K., 563
Anderson, Carl D., 249
Anderson, Christine M., 245
Anderson, Christopher M., 278, 314
Anderson, D. L., 31
publication, 125
Anderson, Fred, 315
Anderson, G. M., 560
Anderson, Jan M., 164, 165, 243, 244
Anderson, Kurt S., 282, 314
Anderson, Lawrence, 258
Andrews, Grace M., 644
Anfinsen, Carol, 130
Aoki, K., 564
Aparicio, Pablo, 130
Appleman, D. E., 371, 374, 495, 560, 565
Archambeau, C. B.
publication, 125
Armstrong, Jean M., 240
publication, 242
Arnold, W., 155
publication, 159
Arp, Halton C, vi, 49, 254, 278, 279, 282,
283, 284, 287, 290, 296, 314
Asada, T.
publications, 125, 126
Ashby, Eric, iv, 27, 74, 75, 693
Ashworth, J. M., 641
Atkins, F. B., 348, 349, 356, 358, 456, 457.
560, 571
Ator, Claudine C, 130
Austin, C. R.
publications, 641, 643
Axelrod, David, 130
Baade, Walter A., 58, 277, 278
Babcock, H. W., vi, 54, 8, 124, 314
report of the Director, Mount Wilson and
Palomar Observatories, 249-316
Bahcall, J. N., 294
705
06
CARNEGIE INSTITUTION
Bailey. J. L.
publication. 189
Baker, Clara K.. 245
Baker. H. G.
publication. 244
Baker. P. E.. 461, 462. 463, 560
Baker. Robert M.. 245
Baldwin. George J., xi
Ballard. Ray L.. 315
Bancroft. A. M., 27
publication. 124
Bancroft. G. M., 360, 560
Barber. J. T.
publication. 189
Barbon. Roberto, 296
Barbour. Thomas, xi
Barnes, James M., 245
Baron, L. S.
publication, 125
Bartlett, E. P., 540, 560
Barton, Lena R., 245
Barton. Mary N., 644
Barton. P. B\, 437, 561, 568
Baschek, Bodo, 270
Bateman, P. C, 478, 561
Batten. C. A., 570
Battle, Raymond L., 130
Bauer, M., 446, 561
Baum, W. A., 135
publications, 125, 126, 127, 128
studies, 133-140
Beach, Liselotte, 130
publications, 124
Beaman, D. R., 329, 561
Bear, Phyllis, 672
studies, 650-657
Beck, A. J.
publication, 641
Beck, F.
publication, 641
Becker, W., 296
Becklin, Eric, 52, 253, 278, 279, 294, 314
Beermann, W., 578
Belin, H. Lowell, 130
Bell, James F., xi
Bell, Peter M., vi, 319, 320, 324, 326, 369,
433, 463, 480, 548, 551 561 568, 570
studies, 442, 541-544, 545-547
Bendich, Arnold J., 130
publication, 125, 126
Benirschke, K.
publication, 641
Ben-Menahem, A.
publication, 125
Bensch, J. J., 569
Benson, Donald W., 633
Berggren, T., 565
Bergh, Sidney van den, 300
Berkebile, Wilma J., 315
Berlin, Linda, 644
Berry, M. J.
publication, 124
Bertiau, F. C, 58
publication, 125, 127
Bertola, F., 136
Beugeling, T., 151
publication, 154
Beuzenberg, E. J., 239
publications, 242, 243
Biehler, S.
publication, 124
Biggers, John D., 64, 580
publications, 641, 643
Billings, John S., xi, 695, 696, 697
Billingsley, Lynn, 644
Bilpuch, E. G.
publication, 125
Binns, R. A, 357, 561
Birch, F., 21, 52, 561, 562
publication, 125
Bishop, David W., vii, 76
publications, 641, 642, 643
studies, 621-626
Bjorkman, Monika, 148
studies, 216-220
Bjorkman, Olle, vii, 66, 148, 245
publication, 244
studies, 214-216, 216-220, 220-228, 228-
233, 233
Blackwell, James, 644
Blackwell, Paul, 644
Blakee, Lawrence E., 314
Bliss, Robert Woods, xi
Blows, J. H., 561
Bocskay, Elizabeth M., 672
Boesgaard, Ann Merchant, 265, 268, 269,
314
Bohm-Vitense, E., 297
Boise, James W., viii
Bolton, Ellis T., vi, 71, 68, 101, 111, 129
publications, 125, 126
report of the Director, Department of
Terrestrial Magnetism, 3-131
Bolton, John G., 254, 287, 290, 296
Bonica, John, 580, 633, 644
Bonini, W. E.
publication, 124
Borcherdt, R. D.
publication, 124
Borgen, Fern V., 315
Bovenkerk, H. P., 561
Roving, Bent G., vii, 643
publications, 641, 642
studies, 626-632, 638-639
Bowen, Ira S., vii, 52, 57, 135, 137, 140, 302,
303
Bowen, N. L., 362, 364, 365, 366, 371, 460,
477, 479, 561, 567, 568
Boyce, P. B.
publication, 126
Boyd, Francis R., vi, UO, 1*2, 319, 320, 322,
326, 347, 348, 352, 369, 370, 381, 383,
384, 389, 449, 480, 484, 542 548, 551,
554, 561, 570
INDEX
707
publication, 560
studies, 327-334, 353-359, 545-547
Braccesi, Alessandro, 293, 296, 314
Bradford, Amory H., iv, v, 693
Bradford, Lindsay, xi
Bradley, Omar N., iv, v, 693
Bradshaw, N., 565
Brenner, Don J., 68, 129
publication, 125, 126
studies, 106-118
Brett, R., 443, 444, 559, 561
publication, 560
Bridgman, P. W., 543, 545, 561
Briggs, Winslow R., 147
studies, 203-207
Bril, C.
publication, 188
Brink, R. A.
publication, 642
Britten, Roy J., vi, 54, 55, 56, 57, 68, 106,
107, 129
publications, 126, 127, 128
studies, 68-73, 73-88
Brody, M., 196
Brody, S. S., 194, 196
Brookings, Robert S., xi
Brown, Donald D., vii, 58, 59, 76, 576, 577,
578, 642, 643
studies, 580-589
Brown, Jeanette S. vii, 146, 181, 184, 185,
186, 187, 188, 197, 243, 245
publication, 244
studies, 192-196, 196-197
Brown, Louis, vi, 54, 57, 129
publications, 126, 127, 128
studies, 62-66, 66-68
Brown, M. G., 349, 357, 561
Brown, G. Malcolm, 39, UO, U2, 319, 320, 322,
323, 348, 349, 351, 352, 364, 547, 548,
551, 561, 569, 570
studies, 347-353, 353-359, 359-363, 427-
429, 460-567, 467-471
Brown, W. L., 561
Brueckel, Frank J., 314
Bruinsma, Jan A., 315
Bruinsma, Maria J., 315
Buchanan, Jennie S., 672
publication, 673
Buchans, R., 436, 561
Buck, Donald L., 315
Buddington, A. F., 561, 562
Buerger, M. J., 371, 561
Bundy, F. P., 449, 561
Burbidge, E. M., 287, 290, 296
Burd, Sylvia, 314
Burgi, Elizabeth, vii, 6k, 660, 672
studies, 650-657, 657-659
Burke, Bernard F., 58, 130
publication, 125
Burnham, Charles W., vi, 386, 487, 488, 494,
551, 556, 557, 563, 567, 570
publication, 500
Buuren, Hendrika E., van, 315
Buuren, John E., van, 315
Burrhus, Kenneth D., 130
Burns, R. G., 561
Bush, Vannevar, iv, 693
Busing, W. R., 488, 562
Butler, B., 672
Butler, R. L., 570
Butler, W. L., 162, 173, 175, 204, 207
Buynitzky, Stephen J., 130
Byard, Paul L., 136
Cabre, Ramon, 7, 28, 130
Cadmus, Robert, 130
Cadwalader, John L., xi, 695
Caherty, Francis J., 130
Cahn, M. B., 602
Cahn, R. D., 600, 602
publication, 642
Cairns, R. B., 564
Calder, J., 570
Campbell, A. M., 62, 63
Campbell, William W., xi
Canter, Dorothy, 130
Carew, Bayne, 130
Carmichael, I. S. E., 452, 457, 462, 463, 562
Carnegie, Andrew, 697
Carr, T. D., 300
Carrison, L. C, 567
Carty, John J., xi
Casanova, Manuel, 315
Casaverde, Mateo, 7, 130
studies, 12
Case, James F., 579, 643
studies, 618
Castor, John, 314
Cederstand, C. N., 186
Chamberlin, Margaret E., 130
Changeux, Jean-Pierre, 641
Chao, E. C. T., 562
Chase, John, 644
Chase, Richard A., ix
Chayes, Felix, vi, 320, 325, 326, 548, 551,
554, 562, 570
publication, 560
studies, 505-510
Chen, W. T., 566
Chiba, Shigeru, 327, 242, 243
Chinner, G. A., 562, 567
Christian, R. S., 64, 65, 66
publication, 125
Chuadze, A. D., 59
Chugainov, P. F., 263
Citarella, R. V.
publication. 125
Clark, A. H., 562
Clark, L. D., 561, 562
Clark, Maynard K., 314
Clark, S. P., 561, 562
70S
CARNEGIE INSTITUTION
Clausen. Jons C, vii, 66, 67t 149, 213 245
publications. 242. 243, 244
studies, 284-243
Clayton. Roderick, EL, 155, 159, 566
publication. 244
Cleary, William. t>3S. 644
Cole, Whitefoord R.. xi
'.nan. L. R.. 336, 338. 551, 562
Conti. Peter S., 265, 266, 268. 270, 271, 301,
814
Coombs. J.. 228
publications. 228
Coon. Hayden G., 579, 580, 643
publication. 642
studies. 602-606
Cooper. B.
publication, 159
Cortex, Rolando, 315
Cosslett, V. E.. 566
Couch. Hugh T., 315
Cowie, Dean B.. vi, 68, 129
publications, 126
studies, 88-95, 95-106, 106-118
Crafts, W., 564
Cragg, Thomas A., 255, 314
Craig, J. R.. 42, 319, 325, 440, 548, 551, 559,
562, 566, 570
publications, 560
studies, 413-417, 417-419, 431-434, 434-
436, 436-440, 440-441
Critchfield, William B., 235
publication, 242
Crowley, M. S„ 492, 562
Cruickshank, D. W. J., 565
Cupples, H. L., 561
Czaplicki, Helen S., 315
Czyzak, Stanley J., 295
Dalhvitz, W. B., 347, 348, 352, 562
Daly, R. A., 562
Dan, Katsuma, 641
Danforth, D. N.
publication, 642
Danziger, Ivan J., 271, 297, 314
Darken, L. S., 420, 423, 424, 562
Dattner, A., 10
publication, 124
Davis, B. T. C, 554, 560, 561
Davis, B. L., 562
Davis, Gordon L., vi, 48, 7, 42, 320, 326, 347,
353, 543, 570
studies, 44-52 528-536
Dawid, Igor B., vii, 60, 577, 578, 643
publications, 642, 643
studies, 592-597
Day, Floyd E., 314
Deer, W. A., 383, 569
DeHaan, Robert L., vii, 579, 615, 643
publication, 642
studies, 606-617, 618-619
Dehn, J., 565
DeLanney, Louis E., 644
Delano, Frederic A., xi
Demerec, M.
publication, 673
Denis, H.
publication, 642
Dennison, Edwin W., vi, 289, 304, 314
Deutsch, Armin J., vi, 51, 52, 265, 270, 271,
272, 275, 276, 297, 314
Dicke R. H., 52, 264, 275
Dickens, Robert J., 241, 275, 276, 297, 314
Difley, John A., 314
Dillin, Dorothy B., 130
Dimroth, Erich, 538, 562
Doak, John B., 130
Dodge, Cleveland H., xi, 695, 696
Dodge, William E., xi
Dolley, Madeleine, 314
Donahue, R. P., 64
Donath, F. A., 538
Donnay, Gabrielle, vi, 320, 325, 424, 549,
551, 555, 558, 562, 564, 570
publication, 560
studies, 485-487, 497-498, 498-502, 503-
504
Donnay, J. D. H., 320, 325, 558, 564, 570
publication, 560
studies, 497-498, 498-502, 503-504
Donnelly, T. W., 466, 562
Donner, Martin W., 580, 644
publication, 624, 643
studies, 632-634, 634-635
Doring, G., 165
Dougherty, Cary M., 644
Dove, William N., 130
Dreiling, Raymond, 314
Drever, H. I., 456, 457, 562
Duggal, S. P., 11
publication, 126, 127
Dugre, D. H., 57
Duncan, William H., 644
Duncumb, P., 327, 562
Dundon, Robert, 319, 323
studies, 366-369
Duysens, L.N.M., 143, 151, 155, 166, 170,
172, 194
publications, 154, 159, 160, 171, 175, 196
Dyson., J. E., 300
Earl, John B., 672
Ebert, James D., vii, 57, 60, 76, 259, 573,
579, 643
publication, 642
studies, 598-602
report of the Director, Department of
Embryology, 573-644
Echols, Harrison, 64, 660
Eckermann, H. von, 346, 562
publication, 672
Ecklund, Everett T., 130
Edgar, A. D., 566
Eggen, Olin J., 261, 267
INDEX
709
Egle, K., 221, 224
publication, 228
Ehring, Ruth, 101, 672
Eisma, E., 564
Elam, Loyal, Jr., 314
Eldredge, Marylee, 245
El Goresy, A., 570
Ellison, Eleanor G., 315
El-Sharkawy, M. A., 224
publication, 228
Emerson, R., 143, 155, 197, 202
publications, 159, 203
Emery, K. O., 510, 515
Emmel, John M., 245
Emslie, R. F., 320, 324, 570
studies, 479-480
Eng, Marlene, 580, 633, 644
Engel, A. E. J., 561, 562
England, Joseph L., vi, 319, 320, 326, 347,
348, 369, 370, 381, 383, 384, 433, 551,
561, 568, 570
studies, 442, 541-544, 545-547
Engstrom, A., 566
Epstein, Elizabeth R., viii
Epstein, S., 563
Ernst, W. G., 364, 365, 487, 492, 563
Eugster, H. P., 563, 569
Evans, B. J., 563
Evans, B. W., 563
Evans, H. T., Jr., 560, 565, 568
Ewing, J. I., 566
Extermann, P., 65
publication, 125
Fairall, A. P., 287
Falkner, F.
publication, 641
Falkow, S., 68, 110, 116
publications, 125, 126
studies, 88-95, 95-106
Faulkner, John, 274, 275
Fawcett, J. J., 380, 385, 386, 563
Fenner, Charles P., xi
Ferguson, Homer L., xi
Ferguson, M., 570
Fermor, L. L., 449, 563
Fernandez, Luis, 7, 28, 130
Fisher, Agnes C, 672, 673
Fisher, G. W., 570
Fitzgerald, M. J. T., 644
Flamm, W. G., 74
publication, 125
Flexner, J. B., 118, 122, 130
publication, 126
Flexner, Louis B., ix, 118, 122, 130, 643
publications, 126, 127
Flexner, Simon, xi
Florin, Rudolf, 239
publication, 242
Fock, H., 221, 224
publication, 228
Forbes, W. Cameron, xi
Forbush, Scott E., vi, 58, 54, 7, 120
publications, 124, 126
studies, 8-12, 12
Ford, W. Kent, Jr., vi, 52, 5/,, 57, 129, 135,
136, 137, 140
publications, 126, 127
studies, 57, 66-68
Fork, David C, vii, 67, 68, 143, 144, 145,
147, 182, 183, 245
publications, 165, 243, 244
studies, 149-155, 155-160, 160-165, 165-
171, 203-207
Forrestal, James, xi
Forrester, M. L., 221, 224
publication, 228
Forsbacka, Allen, 130
Foshag, W. F., 452, 563
Foster, P. K., 563
Foster, W., 389, 424
Foster, W. R., 562
Fowler, B. C.
publication, 124
Francis, G. H., 389, 563
Frank-Kamenetsky, Va. A., 498, 563
Franz, H., 567
Freed, S., 169
publication, 186
French, C. Stacy, vii, 66, 186, 245
publications, 186, 188, 197, 244
studies, 175-177, 177-186
report of the Director, Department of
Plant Biology, 141-248
Frenkel, A. W., 144, 163
publication, 165
Frew, William N., xi, 695, 696
Fry, R. M., 644
Fudali, R. F., 427, 466, 563
Fukada, Masako, 579, 643
Fulder, Stephen J., 193, 245
Fuller, Henry, Jr., 644
Fulling, W., 569
Furshpan, E. J., 641
Futergendler, S. L, 446, 449, 498, 563
Fyfe, W. S., 563
Gabrielson, Fanny G., 315
Gaddy, V. L., 540, 569
Gaffron, H., 143, 147, 198, 202
publication, 203
Gage, Lyman J., xi, 695, 696
Gajardo, Enrique, 7, 28, 130
Gammel, J. L., 64, 65, 66
publication, 125
Garde, Wilbur F., 644
Garlick, G. D., 563
Garrett, Patricia S., 570
Garrison, Robert F., 271, 297, 314
Garver, L. C, 570
Garzoli, S., 62
Gates, Howard, 287, 314
Gauhl, Eckard W., 245
studies, 233-234
10
CARNEGIE INSTITUTION
Gawrys, Stanley, 130
Gay, Helen, vii
Gay. P.. 561
Geiser, P., 570
Gelderman, Albert, 130
Georgen, Robert P., 314
Gerrity. Marline. 314
Ghose, S.. 563
Giacconi, R.. 2S8
Giardini. A. A., 565
Giesecke, Alberto A., 7, 130
Gifford, Walter S., xi, 693
Gilbert. Carl J., iv. v. 693
Gilbert. Cass, xi
Gilbert. M. C. 319, 323, 324, 549, 570
studies, 374-375
Giles. Solomon, 314
Gill. Richard P., 245
Gillett. Frederick H., xi
Gilman, Daniel C, xi, 695, 696, 697
Giumarro, C.
publication, 232
Glass. G.
publication, 154
Godovikov, A. A., 570
Goedheer, J. C, 194, 195
publications, 154, 165, 196
Goetz, Alexander F. H., 261, 314
Goldberg, E. B.
publication, 672
Gonzalez, R., 563
Good, R. A.
publication, 642
Goodenough, Gerald K., 644
Goodwin, T. W., 196, 244
Gorsline, D. S., 515
Gottlieb, Sheldon H., 612, 643
studies, 606-617
Govindjee
publication, 186
Graham, John, Jr., 643, 644
studies, 606-617
Gram, K., 237
publication, 242
Green, D. H., 562
Green, Howard, 578, 590
Green, T. H., 479, 480, 563
Greenewalt, Crawford H., iv, v, 694
Greenlee, Beulah, 315
Greenlee, Frank K., 315
Greenhill, J. P.
publication, 641
Greenstein, Jesse L., vi, U9, 51, 249, 254,
261, 262, 266, 268, 269, 270, 283, 290,
305, 314
Greenwood, Hugh J., 388, 389, 551, 563, 566
Gregory, R. P. F.
publication, 189
Greig, J. W., 424, 563
Griffin, Roger, 297
Griggs, D. T., 568
Grill, Richard D., 644
Grossman, A. L.
publication, 672
Gruenwald, Peter, 644
Grygar, J., 266
Guidotti, C. V., 563
Gurdon, J. B., 11, 577
publication, 642
Gurry, R. W., 562
Giiven, N., 320, 325, 549, 551, 563, 570
studies, 487-492, 494-497
Haber, Leo J., 130
Hadidiacos, C. G., 570
Haeberli, Willy, 130
publication, 126
Hafner, S., 563
Hagg, Gunnar, 498, 563
Hair, J. B., 239
publication, 242
Hales, A. L.
publications, 125, 126
Hall, John S., 136
publications, 125, 126
report of the Chairman of the Committee
on Image Tubes for Telescopes, 133-140
Hallberg, Elizabeth L., 644
Hallberg, R. L., 643
studies, 589-590
Hallimond, A. F., 348, 349, 494, 495, 563
Hancock, Eugene L., 315
Handwerker, D. S., 560
Hanneman, R. E., 545, 563
Hanse, Edward C, 320, 326, 420, 549, 570
studies, 538-539, 536-538
Hansen, M., 563
Haraburda, Joseph M. S., viii
Harcourt, G. A., 568
Hare, P. Edgar, vi, 54, 55, 320, 326, 549,
570
studies, 526-528
Harker, A., 563
Harker, R. I., 563
Harper, Ernest, 644
Harris, Bennie, 130
Harris, J. W., 446, 498
Harris, R. L., Jr., 450, 570
Hart, Lawrence W., 245
Hart, P. J.
publication, 124
Hart, Richard W., 245
studies, 228-233
Hart, Stanley R., vi, A6, 7, 42, 129, 320, 558
publications, 125, 126, 127, 560
studies, 52-57, 528-536
Harvey, J. W., 255
Harvey, John, 130
Harwood, H. F., 469, 564, 568
Haskell, P. T.
publication, 673
Haskins, Caryl P., iv, v, viii, 693
publications, 673
INDEX
711
Hast, M. H., 644
Hatch, M. D., 228
publication, 228
Hauck, Lilo, 315
Haul, R., 498, 563
Hawley, J. E., 417, 420, 429, 563
Hay, John, xi, 695, 696, 697
Hayward, C. R., 563
Heard, H. C, 568
Heber, Ulrich W., 245
Heintze, J. W. R., 266, 272, 314
Helms, Anna, 237
publication, 242
Henard, Kenneth R., viii
Henderson, E. P., 564
Henderson, Mark D., 315
Henriques, R. C, 563
Henry, Barklie McKee, xi, 72, 73, 7 A, 693
Henry, N. F. M., 354, 563
Hernandez, Frank, 315
Herrick, Myron T., xi
Herrin, E. T.
publication, 124
Hershey, Alfred D., vii, 61, 62, 63, 76, 94,
101, 645, 650, 672
studies, 659-660
report of the Director, Genetics Research
Unit, 645-672
Hertig, Arthur T., 643
Herzen, R. von, 55, 125
Herzog, Emil, 285, 314
Hess, H. H., 348, 349, 357, 564
Hess, J. L., 221
publication, 228
Hett, Victor A., 315
Hewitt, Abram S., xi
Hey, M. H., 563
Hicks, Virginia, 644
Hiesey, William M., vii, 148, 229, 245
publications, 232, 244
studies, 208-214, 214-216, 228-233, 233
Higginson, Henry L., xi, 695, 696
Highkin, Harry R., 163, 181, 183
publication, 165
Hilgeman, T., 260
Hill, Byron, 315
Hill, E. Kathleen, 130
Hilty, D. C, 420, 421, 424, 425, 564
Hitchcock, Ethan A., xi, 695, 696
Hitchcock, Henry, xi
Hoch, G., 149
publication, 154
Hoering, Thomas C, vi, 42, 320, 326, 512,
513, 514, 549, 551, 570
studies, 510-514, 515-526
Hoffmaster, Robert, 130
Hogness, D. S.
publication, 125
Holborn, L., 540, 564
Holgersson, S., 498, 564
Holland, H. D.
publication, 127
Holloway, Helen D., 315
Holmes, A., 460, 56.°,
Holmgren, Paul, 218, 220
publications, 243, 244
Holser, W. T., 541, 564
Hoover, Herbert, xi
Hopper, Anne, 315
Hornblower, Marshall, viii
Howard, Robert F., vi, 255, 256, 257, 258,
314
Howe, William Wirt, xi, 695, 696
Howell, B. F.
Howie, R. A., 561, 562
publication, 124
Hoyer, Bill H., 80, 81, 82, 106, 130
publication, 125
Hubble, Edwin P., 280
Huber, N. K., 561
Huber, P., 126
Huckenholz, H. G., 319, 322, 549, 570
studies, 335-347
Huffaker, E. C., 570
Humphrey, Samuel, 130
Humphreys, Joyce, 314
Hunger, K., 269
Hunziker, Juan H., 239
publication, 242
Hurst, J. V., 499, 562, 564
Hutchinson, Charles L., xi, 695
Hutchinson, R. W., 429
Iben, Icko, 264
Idzinga, Fred, 314
Imlay, Marjorie E., 570
Ingamells, C. O., 330, 565
Ingerson, Earl, 538, 564
Ingraham, Laura J., 64, 101, 663, 672
studies, 650-657, 659-660, 663-664
Inoue, Shinya, 641
Irwin, John B., 306, 314
Ito, T., 485, 564
Jaccarino, V., 569
Jacob, F., 115
publication, 125
Jacques, Rienaldo, 315
Jahns, R. H., 565
James, David E., 7, 43, 44, 129
publication, 127
studies, 13, 13-24
James, H. L., 562
Jamieson, J. C, 543, 564
Jeffery, Doris, 315
Jensen, E., 564
Jessup, Walter A., xi
Jewett, Frank B., xi
Johnson, H. L., 274, 276
Johnson, H. S., 228
Johnson, Hugh M., 297
Johnson, Melvin W., 314
Johnson, Paul A., 130
12
CARNEGIE INSTITUTION
Johnston, Glenda J.. ISO
Johnston, J.. 568, 564
Johnston. R.. 562
Joliot. A.. 153, 165, 167
publications. 154, 159
Joliot. P.. 153, 155, L57
publications. 154, 159
Jones. Henry S.. 673
Jordan. Eddie, 644
Jordan, John D., 315
Jergensen, C, A.. '2X~
publication, 242
Joy. Alfred H.. 314
Jugaku, Jun, 51, 265, 269, 289, 314
Jungwirth. C. 642
Jurg, J. W., 512, 564
Kaighn. M. Edward, 579. 580, 603, 643
publication, 642
studies, 598-602
Kaiser. A. D., 101, 115
Kaiser. I. H., 642
Kahan, E., 126
Kamb, W. Barclay, 551
Kameda, M., 566
Kamitsuki, Akira, 129
Kanamori, H., 28
publication, 125
Kaplan, Norman, 673
Kapp, H. E., 568
Karle, Isabella L., 485, 564
Karle, J., 485, 564
Karpowicz, M., 285
Kasche, V., 204, 207
Kashkai, M.-A., 346, 564
Kasinsky, Harold H., 579, 643
Katem, Basil, 314
Katoh, S., 149
publication, 154
Katz, Margaret, 314
Kaufmann, B. P.
publication, 673
Kawano, Y., 564
Ke, C. L., 165, 171
publication, 171
Kearns, Charles E., 315
Keenan, Philip C, 268, 272, 297
Keith, M. L., 389, 564
Kennedy, G. C, 359, 481, 484, 564
Key worth, G. A., 125
Khorana, Gobind, 663
Kilgore, Willis, Jr., 130
Kimmel, C. B., 642
Kinman, T. D., 49, 275, 290, 296
Knopf, E. B., 538, 564
Knopoff, L., 124
Knorring, O. von, 566
Kocmaneck, J. F., 570
Koelbloed, David, 270, 314
Koerper, Marion A., 245
K off man, A. R., 429, 564
Kohne, David E., vi, 54, 55, 56, 57, 68, 106,
107, 129
publications, 126, 127
studies, 73-88
Kok, Bessel, 67, 143, 149
publications, 154, 159
Kondo, N., 57
Konigsberg, Irwin, R., 602, 643
publication, 642
Koranda, John, 233
Kouchkovsky, Yaroslav de
publication, 244
Kouvo, O., 553
publications, 127, 560
Kowal, Charles T., 285, 286, 314
Kowalik, Jan, 245
Kowallik, W., 147, 198, 202, 203
publication, 203
Kraft, Robert P., vi, 76, 263, 264, 265, 276,
314
Krause, O., 498, 564
Krieghoff, N. G., 325, 551
Kristian, Jerome, 298
Krogh, Thomas E., vi, 58, 7, 42, 129, 320,
326, 453, 549, 570
studies, 44-52, 528-536
Krotkov, G., 221
publication, 228
Kruskal, W., 554, 562
publication, 560
Krzeminski, Wojciech, 263, 264, 265, 276,
314
Kullerud, Gunnar, vi, 40, 42, 319, 320, 325,
371, 419, 420, 424, 429, 430, 434, 436,
549, 551, 554, 556, 559, 561, 562, 564,
565, 566, 569, 570
publication, 560
studies, 413-417, 431, 434, 440-441, 442,
442-446, 497-498, 503-504
Kushiro, I., 472, 474, 475, 476, 555, 560
publication, 560
Kyker, G. C, Jr.
publication, 125
Lachenbruch, A. H., 54
publication, 125
Lackner, Dora Russo, 314
Langley, Samuel P., xi, 695
Lara, J. Luz, 315
Larsen, C. Muhle, 237
publication, 242
Larsen, C. Syrach, 237
publication, 242
Larsen, E. S., 345, 346, 565
Larson, Richard, 295
Lasker, Barry M., 298, 314
Lawrence, Ernest O., xi
Lawrence, Mark C, 230, 245
studies, 175-177
Lederberg, Esther, 652, 653
Lederberg, J., 647, 648, 649, 652, 653
Lee, Duk Hee, 314
INDEX
7 1 3
Lee, Harold H., 579, 643
studies, 598-602
Legum, Elizabeth, 644
studies, 606-617
Leighton, Robert B., vi, 249, 255, 259, 279,
306, 314
Leonard, B. F., 562
Levy, H. A., 562
Lewis, A. T., 570
Lewis, Brian T. R., 17, 130
publication, 124
Lewis, Charlton M., 197, 202
publications, 203, 244
Lewis, J. F., 348, 462, 565
Leymann, A., 567
Lichtenstein, Edna G., 644
Liddle, Ralph R., 245
Lihl, F., 498, 565
Lindbergh, Charles A., xi
Lindsay, William, xi, 695, 696
Lindsley, Donald H., vi, 38, 39, 319, 320,
323, 324, 348, 350, 353, 363, 479, 480,
484, 550, 551, 565, 558, 561, 570
publication, 560
studies, 359-363, 363-366, 366-369, 479-
480
Linschitz, H., 204, 207
Little, Charles A., 130 .
Little, Elbert L., 235, 242
Littna, E.
publication, 642
Livingston, R., 165, 171
publication, 171
Livingston, William C, 257
Locanthi, Dorothy D., 269, 314
Locke, M.
publication, 642
Lodge, Henry Cabot, xi
Long, A., 570
Long, J. V. P., 561
Loomis, Alfred L., iv, v, 675, 693, 694
Loomis, R. S., 227
publication, 228
Lorz, John G., 130
Lovett, Robert A., iv, v, 675
Low, Seth, xi, 695, 696
Lowen, A. Louise, 314
Luckett, W. P., 644
Lumry, R., 169
publication, 171
Lundqvist, D., 434, 441, 565
Lupanova, N. P., 346, 565
Luth, W. C, 320, 477, 551, 565, 567, 570
studies, 480-484
Lutz, H. J., 570
Luyten, Willem J., U9, 253, 279, 284, 291,
292, 293, 298, 299
Lynch, Patricia, 315
Lyndon, R. F.
publication, 189
Lynds, R., 293, 307
Lyons, W. H., 570
MacGregor, I. D., 449, 570
Mackay, Andrew, 314
Macres, V. G., 327, 565
MacVeagh, Wayne, xi, 695, 696
Madsen, A., 191
Madsen, B. A., 562
Magruder, Louise E., 130
Maio, J. J., 74
publication, 125
Makita, Mitsugu, 259, 314
Makover, Shraga, 672
studies, 650-657
Malooly, Thomas F., 644
Mandy, Juanita, 644
Mantuani, Mary, 243
Marconi, H., 7, 28
studies, 37-42
Mark, Gerda, 612, 613
Marsh, Fayette, 644
Marshall, R. E., 641
Martin, C. B., Jr., 580, 644
publications, 642, 643
studies, 632-634, 634-635, 636-638
Martin, K. O., 562
Martin, Malcolm A., 80, 81, 82, 106, 130
publication, 125
Martin, R. D., 644
Martin, William McC, Jr., iv, 74, 75, 693
Marton, L. L.
publications, 125, 127
studies, 133-140
Marzalek, Ethel, 315
Marzolf, J., 366, 367, 369, 565
Marzullo, Giovanni, 603, 604, 605
Mason, B., 330, 498, 565
Matousek, B. H., 565
Matthes, Z., 570
Maxwell, A. E., 55
publication, 125
May, L., 568
McCallum, M., 74
publication, 125
McCammon, D., 259
McCamy, K., 35
publication, 125
McCarthy, B. J.
publication, 125
McCarthy, Martin F., 58, 101, 111, 130
McClintock, Barbara, vii, 20, 60, 61, 7U, 664,
672
studies, 664-672
McClunin, O. R., 570
McCord, Thomas B., 299, 314
McCracken, K. G., 10
publication, 124
McDonald, W. S., 496, 565
McDonell, Sharon, 315
McElroy, W. D., 154
McGaughey, Harry S., Jr., 644
publication, 642
McGee, J. D., ix
publication, 126
714
CARNEGIE INSTITUTION
McGough. Sheila A., viii
McHugh. Keith $.. iv, v. 675, 698, 69 I
McKie, P.. 376, 878, 880, 565
McLaren. Ann. 74, 82
publication, 125
McMahon. Daniel. 217
McManus, Philip J.. Jr., 315
McMullen, P.. 126
Means. Kay 0., 644
Meenakshi. V. R., 570
Mellon. Andrew \\\, xi
Merwin, H. E., 558, 563, 564
Mover, Henry 0. A., 319, 320, 324, 325, 326,
334. 448. 5^S. 565, 563, 570
studies. 446-450. 541-544, 498-502
Meyer. K. K.
publications, 642, 643
Meyer. R. P., 35
publications, 124, 125
Michel, Jean-Marie, 134, 146, 173, 185, 245
studies, 186-189
Michel-Wolwertz, Marie-Rose, 146, 181, 245
publications, 191, 244
studies, 189-191
Miescher, P. A., 642
Miki. Shigeru, 239
publication, 242
Milledge, H. J., 449, 565
Miller, Leonard, 130
Miller, Margaret Carnegie, iv, 74, 694
Miller, Roswell, xi
Miller, William C, 314
Mills, Darius O., xi, 695, 696
Milner, Harold W., 229
publication, 232
Minkowski, R., 59
Mitchell, R. S., 446, 448, 498, 565
Mitchell, S. Weir, xi, 695, 696, 697
Mitchell, Walter E., Jr., 299
Mitchell, William, 233
Mitterer, R. M., 43, 320, 326, 550
studies, 510-514
Moffat, A. F. J., 300
Moh, G., 319, 325, 417
Moldave, K., 672
Montague, Andrew J., xi
Morev, G. W., 566
Morgan, Henrv S., iv, v, 675, 693, 694
Morimoto, X., 320, 495, 554, 564, 566, 570
publication, 560
studies, 494-497
Moore, J. G., 561
Moore, Sandra, 58, 130
publications, 125, 127
Morrill, Michael, 315
Morrow, William W., xi, 695, 696
Morse, S. A., 570
Mosier, James R., 315
Mosig, G., 672
Mossor, Doy E., 130
Muan, A., 341, 363, 566, 567
Mudd, Seeley G., iv, v, 693
Muecke, E. C, 644
publications, 641, 642
studies, 621-626
Muir, I. D., 139, 319, 323, 566
studies, 359-363
Miiller, A., 169
publication, 171
Miiller, K., 566
Munch, Guido, vi, 51, 259, 260, 270, 272,
276, 277, 278, 300, 302, 304, 314
Munizaga, Fernando, 7, 42, 129
Munoz, J. L., 319, 323, 348, 570
studies, 363-366, 369-370, 370-374
Murray, Bruce C, 260, 261, 314
Murthy, V. R., 43
publication, 125
Musselman, A. D., 644
publications, 641, 642
studies, 621-626
Myers, William I., iv, v, 693
Nagamori, M., 426, 427, 566
Nagashima, K., 565
Naldrett, Anthony J., 42, 319, 325, 430, 437,
550, 551, 566, 570
studies, 417-419, 419-427, 427-429, 431-
434, 436-440, 440-441
Naray-Szabo, I., 381, 566
Narbaitz, Roberto, 638, 644
Nation, Bula H., 315
Nelson, C. D., 221
publication, 228
Ness, N. F., 8
publication, 124
Neugebauer, Gerry, 52, 253, 259, 260, 272,
279, 289, 294, 314
Newson, H. W., 125
Newton, Mildred, 315
Newton, R. C, 564, 566
Nicholson, Frank, 245
studies, 228-233
Nishimura, M., 150, 151, 155
publication, 154
Nixon, P. H., 330, 331, 448, 449, 566
Nobes, John C, 245
Nobs, Malcolm A., vii, 148, 245
publication, 244
studies, 208-214, 214-216, 228-233, 233
Nolan, J., 566
Norton, Garrison, iv, v, 675, 693, 694
Novacheck, F., 124
Numata, Makoto, 243
Nur, Uzi, 641
Nuttall, C, 610, 643
studies, 606-617
Obata, F., 188
Oberhoffer, P., 420, 566
O'Brien, Peter N. S., 7, 129
publications, 124, 125, 127
studies, 13, 24-25, 25-27
INDEX
715
O'Dell, C. R., 136, 278
Oersted, A. S. 240
publication, 242
Officer, C. B., 463, 566
Ogawa, T., 186
publication, 188
O'Hara, M. J., 466, 552, 566
Ohmoto, H.
publication, 127
Okada, T. S., 600, 604
Oke, J. Beverley, vi, 50, 51, 254, 265, 267,
271, 272, 273, 282, 283, 287, 291, 305,
314
Olmstead, Alfred H., 315
Omer Guy C, 299
O'Neil, J. R., 566
O'Neill, Frederick, 315
Oppenheimer, Jane, 575
O'Rahilly, Ronan, 644
publications, 642
studies, 639
Orville, C. M., 566
Osawa, K., 136
Osborn, A., 238
publication, 242
Osborn, E. K, 460, 467, 477, 557, 565, 567
Osborn, William Church, xi
Osmer, Patrick S. 288, 289, 314
Osterbrock, D. E., 300
Otsuka, Michio, 129
publications, 127
Otto, J., 40, 564
Padget, Dorcas H., 644
Page, Thornton, 136
Palm, Dennis, 315
Palmer, D. R., 301
Pannbacker, Richard, 641
Parker, E. N., 8
publication, 124
Parker, R. B., 538, 562
Parmelee, James, xi
Parmelee, Suzanne, 245
Parsons, Wm. Barclay, xi
Paton, Stewart, xi
Pattee, H. H., 566
Patton, Donald J., viii
Paul, John, 641
Pavich, Milan, 130
Pazdernik, John, 644
Peach, John V., 289, 293, 314
Peacor, D. R., 346
Pearson, A. A., 644
Peck, D. L., 568
Peck, Eleanor F., viii
Pepper, George W., xi
Perkins, Richard S., iv, v, 693
Perrotta, A. J., 347, 566
Perry, Robert, 641
Pershing, John J., xi
Petersen, Ulrich, 551
Peterson, Bruce A., 283, 314
Petit, Charles, .'il4
Petitjean, Claude, 57, 129
studies, 62-66
Petry, C. B., 570
Philibert, J. A., 327, 566
Phillips, B. 341, 566
Philpotts, A. R., 479
Pickett, James M., 147, 172, 24u
publications, 175, 244
studies, 197-203
Pierri, R. de, 567
Podolsky, T., 568
Poe, Glenn R., 130
Polinger, Iris S., 610, 643
studies, 606-617
Pollock, Harry E. D., ix
Pomerantz, Martin A., 10
publications, 126, 127
Poppel, W. G., 62
Popper, Daniel M., 300
Posnjak, E., 561, 563
Pouillard, E., 498, 566
Pozo, Salvador del, 7, 130
studies, 12
Praagh, Richard van, 616, 644
Prentis, Henning W., Jr., xi
Presnall, D. C, 320, 326, 343, 550, 557, 567,
570
publication, 560
studies, 539-541
Press, F., 124
Preston, J., 348, 567
Prewitt, C. T., 374, 495, 496, 556, 557
publications 560
Price, R. L., 564
Pritchett, Henry S., xi
Proctor, Margaret J., 644
Proskouriakoff, Tatiana, viii
Prothro, Lelah J., 130
Pulfrey, W., 346, 567
Purgathofer, Alois Th., 130, 136
Putnam, H.-M., 346, 567
Quade, Elliott M., 130
Quareni, S., 341, 567
Rabideau, G. S., 175
Rabinowitch, E., 186
Radford, Pamela, 245
Radoslovich, E. W., 488, 494
Rake, Adrian V., 68, 129
studies, 118-124
Ramsay, J. G., 567
Ramsey, Elizabeth M., vii, 580, 643
publication, 642, 643
studies, 632-634, 634-635, 638-639
Rao, V. R., 124
Ratcliffe, E. H., 55
Raven, Peter, 243
Rebbert, Martha, 644
716
CARNEGIE INSTITUTION
Reed, S. J. B., 384. 561, 567
Rentschler, Gordon S., xi
Renzoni, L. S., 568
Rever, Arthur G., 644
Reynolds, Kathleen. 314
Reynolds. Samuel R. M., 618
Rhodes. William T.. 245
Ribbens, Rudolf EM 315
Richardson. S. \\\, 319, 324, 381, 550, 570,
studies. 4Ji>-431
Rice, Marilynne, 315
Riddle G. 0.. 456
Riecker. R. E., 348, 369, 370, 567
Ried, August, 147, 198, 200, 202
Riggs, Suzanne, 644
[lighter, Francis I., 235
publication, 242
Riley, Malcolm S., 314
Rinehart. Carl M., 130
Rinehart, C. O., 561
Roberts, Richard B., vi, 68, 129
publications, 126, 127
studies, 118-124
Robertson, D. W., 163
publication, 165
Robertson, E. C, 561, 562
Robson, D., 62, 63, 64, 130
publications, 126, 127
Robson, G. R., 461, 462, 567
Rockefeller, David, xi
Roddy, John, 130
Roddy Martin, 130
Roddy, Paul, 130
Rodriguez, A., 7, 28
Rogers, J. J. W., 562
Rojas, Herman, 315
Roller, J. C, 124
Rooke, J. M., 566
Rooney, T. P., 567
Root, Elihu, xi, 695, 696, 697
Root, Elihu, Jr., xi, iv, 71, 72, 693, 694
Roseboom, E. H., Jr., 414, 554, 565
Rosen, L. A., 232
Rosenfeld, David, 644
Rosenquist, Glenn C, 615, 644
Rosemvald, Julius, xi
Rowe, F. A., 570
Roy, R. F., 52, 54, 372, 373, 380, 390, 562,
567
publication, 125
Rubey, William W., iv, 693
Rubin, Vera C, vi, 52, 54, 57, 58, 129, 135,
137, 138, 140
publications, 127
studies, 57-59, 66-68
Rudnick, D.
publication, 643
Rudnicki, Konrad, 261, 286, 287, 314
Ruegg, F. C, 568
Rule, Bruce H., vi, 302, 303, 305, 314
Rundles, Charlotte, 644
Russell, Carol, 315
Russell, Helen E., 130
Russell, J. H., 296
Rust, David M., 258, 314
Rust, Philip, 314
Ryerson, Martin A., xi
Saa, German, 7, 28, 129
Sachs, Howard, 315
Sacks, I. Selwyn, vi, 7, 28, 129
publications, 127
studies, 28-35
Sadanaga, R., 564
Sager, D. A.
publications, 642, 643
Saleuddin, A., 570
Salgueiro, Reynaldo, 7, 130
studies, 12
Salmon, J., 565
Salow, Robert S., 315
Samis, C. S., 565
Sancier, K. M., 186
Sandage, Allan R., vi, 6', 49, 50, 51, 76, 249,
253, 254, 274, 275, 280, 281, 287, 288,
289, 291, 292, 293, 294, 296, 298, 314
Sanduleak, N., 307
Sanger, Glen, 315
Sargent, Wallace L. W., vii, 265, 267, 270,
274, 275, 282, 283, 289, 294, 314
Saxinger, W. C, 57
Scarfe, C. M., 475
Scargle, Jeffrey D., 278, 314
Schadee, Aert, 269, 314
Schaefer, Henry F., 315
Schairer, J. Frank, vi, 319, 320, 322, 323,
327, 328, 329, 330, 339, 347, 351, 352,
361, 362, 364, 365, 366, 371, 373, 457,
458, 476, 477, 479, 481, 550, 552, 561,
564, 567, 570
studies, 335-347, 450-457, 460-467, 467-
471
Schenk, R., 567
Schild, Rudolf E., 273, 305, 314
Schildkraut, C. L., 74
publication, 125
Schlegel, H., 567
Schmidt, Maarten, vi, 49, 50, 61, 254, 287,
289, 296, 303, 314
Schmidt-Mende, U. Siggel
publication, 165
Schoon, Th., 563
Schopper, H., 126
Schrank, Winslow W., 644
publication, 641, 643
studies, 621-626
Schreyer, W. F., 319, 324, 340, 387, 388, 389,
551, 567, 570
studies, 376-380, 380-392
Schmucker, U., 12
INDEX
717
Schuetz, Allen W., 580
publications, 641, 643
studies, 619-621
Schwartz, Merry C, 578, 644
studies, 590, 592
Schweinitz, Alan de, 245
Schwerdtfeger, K., 567
Schuller, A., 567
Schwartz, C. M.r 567
Sclar, C. B., 567, 568
Scoon, J. H., 459
Scott, William H., 320, 501, 538, 550, 570
studies, 538-539
Searle, Leonard, 267, 270, 272
Seely, Gilbert, 171
publications, 171, 244
Seemann, Michael, 130
Segnit, E. R., 343, 567
Sellman, Adolph H., 644
Senftle, F., 485
Severn, C. B., 644
Seyler, Richard G., 62, 130
Sharp, W. E., 450, 568
Shaw, H. R., 540, 541, 568
Shea, M. A., 124
Sheahan, Mary T., 130
Shepard, Anna O., viii
Shepley, Henry R., xi .
Sheppard, Carol, 644
Shibata, K., 175
publication, 188
Shields, Oakley, 245
Shields, P. K., 327, 562
Shipley, E. J., 570
Shoemaker, E. M., 562
Shuey, Elizabeth M., 315
Siegel, S. M., 228
publication, 228
Siegelman, H. W., 204
publication, 207
Signer, E. R., 95
publication, 125
Simpson, G. C., 10, 87
publication, 125
Simpson, Louis E., 299
Sinclair, John H. 58, 59, 579, 643
studies, 580-589
Singer, A. D., 570
Singleton, Robert, 130
Sinha, K. P., 498, 568
Sironval, C., 146, 189
publication, 191
Sitarski, G., 261
Skalka, Anna Marie, 64, 94, 658, 660, 672
studies, 650-657, 657-659, 660
publication, 672
Skinner, B. J., 341, 371, 417, 568
Skippen, George, 551
Slack, C. R., 228
Smak, J., 265
Smart, D. E., 124
Smith, Benny W., 315
Smith, Bessie, G44
studies, 598-602
Smith, C. A., 189
Smith, C. H., 429, 456, 457, 508
Smith, F. G., 568
Smith, Harlan J., 136
Smith, James H. C., vii, 144, 245
Smith, J. V., 561, 568
Smith, M. G.
publication, 672
Smith, R. T.
publication, 642
Smith, T. Jefferson, vi, 43, 44, 46, 7, 35, 129
publications, 124, 125, 126, 127
studies, 13, 13-24, 52-57
Smith, Theobald, xi
Sollins, Jeff, 644
Somerville, Delores, 644
studies, 598-602
Sosman, R. B., 563
Souch, B. E., 417, 429, 568
Speicher, G. E., 570
Spencer, L. J., 561
Spijkerman, J.J., 367, 568
Spinrad, Hyron, 269
Spooner, John C., xi, 695, 696
Sproule, K., 414, 468
Stalsberg, Helge, 579, 615, 643
studies, 606-617
Stanton, Frank, iv, v, 693
Stanton, R .L., 461, 462, 463, 553, 568
Staples, Barrett A., 315
Stebbins, Ledyard
publication, 244
Steiger, R. H., 558
publication, 560
Stein, Robert, 294, 314
Steinhart, John S. vi, 43, 44, 46, 7, 129
publications, 124, 125, 126, 127
studies, 13, 13-24, 52-57
Stephenson, D. A., 353, 566, 568
Steward, F. C, 186
publication, 189
Stewart, D. B., 371, 552, 568
Stiehl, H. H., 165
Stiles, Robert G., 315
Stishov, S. M., 449, 568
Stockwell, C. H., 44
publication, 125
Stockwell, Palmer, 235
publication, 242
Storey, William Benson, xi
Storm, C. B., 555, 560
Strauss, F., 53, 61
Strittmatter, Peter, 254, 267, 276, 291
Strom, Stephen E., 265, 267, 268, 314
Strong, A. M., 545, 561
Strong, H. M., 563
Strong, Richard P., xi
Strother, G. K., 173
publication, 175
71S
CARNEGIE INSTITUTION
Strux, H., 568
r, Allan M., T. 42. L29, 320
publication, 125
>. 4:2—44. 528-536
Sumner. R. P.. 29
publication. 125, 127
ton, J. R., 568
Suyehiro, Shigeji, 7. 28, 120
publications, 127
studies. 35-36
Swamy, Krishna, 27S
Swope, Henrietta H.. 274, 281, 314
Szafranski. P.. 110
Taft. Charles P.. iv, v, 693
Taft. William H., xi
Taggart, J.
publication, 124
Tait. D. R.. 566
Taljaard, M. S., 459, 568
Tammann, G. A., 51, 296
Tank, E. S., 644
Tatel, H. E.. 124
Tatewaki, Misao, 235, 243
publication, 243
Taylor, L. A., 420, 570
Tennyson, C, 568
Thayer. William S., xi
Thewlis, J., 498
Thiel, W., 498, 564
Thomas, Dolores M., 570
Thomas. J. B.
publications, 154, 165
Thomas, J. R., 570
Thomas, R., 661
Thomason, Carole E., 672
Thompson, David, 315
Thompson, J. E., 562
Thornber, J. P., 186
publication, 189
Tillev, C. E., ix, 320, 323, 347, 352, 361,
362, 444, 459, 460, 476, 547, 568, 570
studies, 450-457, 457-460, 467-471
Tilton, G. R., 553
publications, 126, 127, 560
Title, Alan, 255
Titus, Doris, 130
Toews, Helen A., 644
Tolbert, X. E., 221
publication, 228
Tomblin, J. F., 461, 462, 463, 562, 568
Toulmin, P., 437, 508,501
Townes, Charles H., iv, 693
Trachslin, Walter, .5/,, 57, 129
publications, 126, 128
studies, 62-66, 66-68
Traxler, Benjamin B., 315
Tregunna, E. B., 221, 224, 228
-lad, R. L., 641
Tr^mearne, T. H., 561
Trimble, Virginia L., 262, 277, 314
Trippe, Juan T., iv, v, 675, 693, 694
Tsuji, Takashi, 266, 268, 269, 314
Turner, F. J., 348, 385, 568
Turner, Kenneth C, vi, 57, 58, 59, 129
publications, 125
studies, 59-60, 61-62
Tuton, Gary M., 315
Tuttle, O. F., 481, 561, 563, 565, 567, 568
Tuve, M. A., 53, 57 58, 59
publications, 124, 125, 128
studies, 133-140
Ubbelohde, A. R., 504. 569
linger, Anne, 130
Urbach, W.
publications, 188
Ursprung, H., 575, 576
Utter, Merwyn G., 255, 314
Van de Velde, Neltje W., 130
Van Hook, William C, 315
Varsavsky, Carlos M., 53, 57, 59, 129
studies, 61-62
Vaughan, Arthur H., Jr., vi, 272, 276, 277,
304,314
Vaughan, Virgal Z., 315
Velde, B., 492
publication, 560
Veliz, Thomas, 315
Venkatesan, D., 10
publication, 124
Verhoogen, J., 369, 563, 568, 569
Vernon, Leo P.,
publications, 171, 244
Vess, Grace D., 314
Vincent, E. A., 561
Vincent, E. C, 548
Visvanathan, Natarajan, 267, 273, 314
Vogel, R., 420, 425, 569
Volponi, Fernando, 7, 28, 130
studies, 37-42
Vredenberg, W. J., 155, 159, 164
publications, 160, 165
Wadsworth, James W., xi
Wager, L. R., 351, 359, 360, 361, 364, 463,
569
Wagner, Manfred, 315
Walburn, Marjorie H., viii
Walcott, Charles D., xi, 695, 696, 697
Walcott, Frederic C, xi
Walcott, Henry P., xi
Walker, L. R., 369, 569
Walker, Peter M. B., 74, 82
publication, 125
Wallace, Betty A., 315
Waller, J. O., 570
Wallerstein, George, 265, 297, 301
Walter, L. S., 449, 457, 474, 569
Wampler, E. J., 50, 291
Waring, M. J., 55, 74
publications, 126 128
INDEX
719
Wasserburg, G. J., 564
Watanabe, T.
publication. 125
Watterson, Clinton, 644
Weber, Carl S., 58, 59, 576, 579, 643
studies, 580-589
Weed, Lewis H., xi
Weedman, D. W., 300
Weiss, L. S., 536, 537, 538, 569
Weistrop, Donna E., 314
Welch, A. J. E., 563
Welch, William H., xi
Wentorf, R. H., 561
Werner, Rudolf, 64, 661, 663, 672
studies, 650-657, 661-663
Wertheim, G. K., 366, 569
Wessels, J. S. C, 169
publication, 171
West, G. F.
publication, 124
Westergaard, M., 237
publication, 242
Westphal, James A., 50, 254, 272, 289, 291,
303, 307, 314
White, Andrew D., xi, 695, 696
White, Edward D., xi
White, E. W., 563, 565
White, Henry, xi
White, James N., iv, v, 693
Whittingham, C. P., 228
publication, 228
Whittingham, David G,, 6U, 65, 580
publications, 641, 643
studies, 626
Wholhieter, R. V.
publication, 125
Wickersham, George W., xi
Wiebe, R., 540, 569
Wilcox, J. M., 8, 257, 258
publication, 124
Wilcox, R. A., 452, 457, 540, 569
Wildey, Robert L., 274
Wilkinson, J. F. G., 548
Willemse, J., 474, 475, 569
Williams, A. F., 569
Williams, Isabelle P., 602, 644
Williams, K. L., 570
Williams, Leroy, 644
Williams, W. A., 227
publication, 228
Williamson, E. D., 558, 564
Willis, D. E.
publication, 124
Wilmoth, David, 644
Wilson, Edward O.
publication, 673
Wilson, Robert E., xi
Wilson, Olin C, vi, 249, 262, 263, 314
Wilson, Ralph W., 315
Wimmenauer, W., 346, 569
Winfree, Arthur, 644
Winkler, H. G. F., 569
Wiser, John L., 644
Witt, H. T., 164
publication, 165
Wittry, D. B., 333
Woese, C. R., 57
Wolken, J. J., 173, 175
Wolstenholme, David E., 60, 577
publication, 642, 643
Wolstenholme, G. E. W.
publication, 641
Wones, W. R., 319, 324, 540, 563, 568, 560
Wood Stephen G., 245
Woodson, Frederick P., 315
Woodward, Robert S., xi
Woodworth, Felice, 315
Woollard, G. P., 21, 23
Woolley, Richard, 136
Worden, Clare, 315
Wright, Barbara, 641
Wright, Carroll D., xi, 695, 696, 697
Wright, F. E., 450
Wright, Jonathan W., 236
publication, 243
Wrubel, M., 276
Wylie, Ann P., 240
publication, 242
Wynn, Ralph, 638, 644
publication, 642
Wynne, C. G., 303
Wys, E. C, de, 562
Yagi, K., 569
Yang, L.
publication, 159
Yoder, Hatton S., Jr., vi, 42, 319, 320 322,
323, 324, 347, 349, 352, 361, 380, 429,
430, 454, 459, 460, 464, 465, 467, 468,
479, 494, 547, 550, 551, 552, 555, 563,
565, 567, 569, 570
publication, 560
studies, 335-347, 376-380, 442-446, 450-
457, 457-460, 471-477, 477-478
Young, M. Wharton, 580, 621, 644
publication, 643
Young, R. A., 504, 569
Younkin, Robert L., 301, 302
Yund, R. A., 416, 417, 433, 434, 440, 556,
565, 569
publication, 560
Yuyama, Shuhei, 579, 643
Zelitch, I., 221, 222
publication, 228
Zies, E. G., 330, 570
Zimmerman, Lorenz, 639
Zirin, Harold, vi, 258, 259. 272, 314
Zussman, J., 562
Zverev, S. M.
publication 127
Zwicky, Fritz, vi, 282, 283, 284, 285, 286,
287, 314
Live Music Archive
Librivox Free Audio
Metropolitan Museum
Cleveland Museum of Art
Internet Arcade
Console Living Room
Books to Borrow
Open Library
TV News
Understanding 9/11