Skip to main content

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

See other formats



Jks ^J 






Year Book 72 


Library of Congress Catalog Card Number 3-16716 

The Maple Press Company, York, Pennsylvania 

Issued December 1973 



Officers and Staff v 

Report of the President 1 

Reports of Departments and Special Studies 1 

Department of Embryology 3 

Hale Observatories 97 

Department of Terrestrial Magnetism 165 

Department of Plant Biology 319 

Geophysical Laboratory 415 

Bibliography 735 

Administrative Reports 737 

Report of the Executive Committee 739 

Abstract of Minutes of the Seventy-Fifth Meeting of the 

Board of Trustees 741 

Report of Auditors 743 

Articles of Incorporation 757 

By-Laws of the Institution 761 

Index 767 


Digitized by the Internet Archive 

in 2012 with funding from 

LYRASIS Members and Sloan Foundation 

President and Trustees 


Philip H. Abelson 


Garrison Norton 

William McChesney Martin, Jr. 

William T. Golden 

Rt. Hon. Lord Eric Ashby 
Lewis M. Branscomb 
Michael Ference, Jr. 
Carl J. Gilbert 
Crawford H. Greenewalt 
Caryl P. Haskins 
William R. Hewlett 
Keith S. McHugh 
Henry S. Morgan 
William I. Myers 
Walter H. Page 
Robert M. Pennoyer 
Richard S. Perkins 
William M. Roth 
William W. Rubey 
Frank Stanton 
Charles P. Taft 
Charles H. Townes 
Juan T. Trippe 
James N. White 

Alfred L. Loomis 

Trustee Emeritus 

Former Presidents and Trustees 


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

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

Vannevar Bush, 1939-1955 Caryl P. Haskins, 1956-1971 


Alexander Agassiz 
George J. Baldwin 
Thomas Barbour 
James F. Bell 
John S. Billings 
Robert Woods Bliss 
Amory H. Bradford 
Lindsay Bradford 
Omar N. Bradley 
Robert S. Brookings 
Vannevar Bush 
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 















































Robert A. Lovett 1948-71 

Seth Low 1902-16 

Wayne MacVeagh 1902-07 

Andrew W. Mellon 1924-37 

John Campbell Merriam 1921-38 

Margaret Carnegie Miller 1955-67 

Roswell Miller 1933-55 

Darius O. Mills 1902-09 

S. Weir Mitchell' 1902-14 

Andrew J. Montague 1907-35 

William W. Morrow 1902-29 

Seeley G. Mudd 1940-68 

William Church Osborn 1927-34 

James Parmelee 1917-31 

Wm. Barclay Parsons 1907-32 

Stewart Paton 1916-42 

George W. Pepper 1914-19 

John J. Pershing 1930-43 

Henning W. Prentis, Jr. 1942-59 

Henry S. Pritchett 1906-36 

Gordon S. Rentschler 1946-48 

David Rockefeller 1952-56 

Elihu Root 1902-37 

Elihu Root, Jr. 1937-67 

Julius Rosenwald 1929-31 

Martin A. Ryerson 1908-28 

Henry R. Shepley 1937-62 

Theobald Smith 1914-34 

John C. Spooner 1902-07 

William Benson Storey 1924-39 

Richard P. Strong 1934-48 

William H. Taft 1906-15 

William S. Thayer 1929-32 

James W. Wadsworth 1932-52 

Charles D. Walcott 1902-27 

Frederic C. Walcott 1931-48 

Henry P. Walcott 1910-24 

Lewis H. Weed 1935-52 

William H. Welch 1906-34 

Andrew D. White 1902-16 

Edward D. White 1902-03 

Henry White 1913-27 

George W. Wickersham 1909-36 

Robert E. Wilson 1953-64 

Robert S. Woodward 1905-24 

Carroll D. Wright 1902-08 

Under the original charter, from the date of organization until April 28, 1904, the following 
were ex officio members of the Board of Trustees: the President of the United States, the Presi- 
dent 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. 



1530 P Street, N.W., Washington, D.C. 20005 

Philip H. Abelson President 

Edward A. Ackerman 1 Executive Officer 

James W. Boise Bursar; Secretary -Treasurer, Retirement Trust; 
Executive Secretary to the Finance Committee 

Montgomery S. Bradley 2 Administrative Officer 

Marjorie H. Walburn Assistant to the President 

Sheila A. McGough Editor 

Kenneth R. Henard Assistant Bursar; Assistant Treasurer, 

Retirement Trust 

Joseph M. S. Haraburda Assistant to the Bursar 

Marshall Hornblower Counsel 


Alfred D. Hershey 
Barbara McClintock 
Tatiana Proskouriakoff 


Harry E. D. Pollock 

1 Died March 8, 1973. 

2 Resigned June 30, 1973. 


Report of 
the President 

The year 1973 has been one of those periods in which it is necessary for a 
nation to reevaluate its outlook and policies. For generations we have acted 
as if we enjoyed inexhaustible natural resources. Suddenly we have run headlong 
into crippling limitations. The energy crisis is the more painful because it comes 
relatively soon after the successful Apollo Program. A few years ago people every- 
where said "If we can go to the moon, we can do anything." Now suddenly we 
are unable to provide enough fuel to keep warm. Such a primitive technology as 
the mining of coal is temporarily beyond us. 

We Americans have been smug and careless. We have assumed that our high 
standard of living was a just reward for our inventiveness and organization of 
mass production. While these components had a role, the real key to prosperity 
in the United States was a superior combination of natural resources. We had 
abundant supplies of oil, coal, and iron ore, and many other resources. It did not 
require great ingenuity to exploit them. 

Encountering the new scarcities comes as a shock to many people, but more 
than a decade ago it was already obvious that we were headed for trouble. This 
was particularly evident with respect to energy. 

We were consuming energy, including hydrocarbons, at an ever increasing rate 
but not discovering corresponding amounts of petroleum. While possessing huge 
reserves of oil shale and coal we made only token efforts to develop potential for 
utilizing these reserves to meet our future needs. With domestic production of 
petroleum level or declining and consumption growing, we soon fell into depend- 
ence on foreign sources. 

The environmental movement sharply hastened the growth of dependence. 
Fuel oil replaced coal, and automobiles became more inefficient. The use of 
imported petroleum, which had been increasing slowly, spurted to an annual 
growth rate of as much as 50 percent in September 1973. 

The huge added imports of the United States created a tremendous seller's 
market for the petroleum exporters. In the course of a year, the producers were 
able to roughly double their return on a barrel of oil. A continuation of the growth 
of imports into 1974 would have been financially unsustainable. It would have 
entailed a cost of $15 billion or more. 

The growing dependence of the United States placed a strong weapon in the 
hands of the Arabs, and they used it. In the process they gained in many ways. 
They conserved a nonrenewable natural resource. By creating scarcity they drove 
the price up further. The immediate diplomatic consequences of their act were 
obvious — the long term results are not so clear. As for the United States, we will 
have a winter and likely much longer to reflect on the necessity of more prudence 
in preparing for future needs. 

During the past several decades in which we enjoyed a higher standard of 
living than other countries, our economy was largely insulated from the rest of the 
world. We could build a prosperous economy largely on items derived from 
within the United States. In addition, we enjoyed a superiority in technology that 
enabled us to compete in world trade and to pay for imports of raw materials. 

Our technical superiority has faded at a time when we have become more 
dependent on imports of raw materials, particularly of petroleum and petroleum 
products, but also such commodities as iron ore, bauxite, chromium, nickel, silver, 
and platinum. At the same time, other countries throughout the world are also 
seeking to increase their consumption of materials. The buyer's market charac- 



teristic of petroleum is likely to spread to many other items. If we base our 
economy too heavily on imports, we shall have no end of troubles — financial and 

How can we pay for the oil and other materials? Over the long haul we must 
depend on increased exports. But this need collides with the fact that Western 
Europe and especially Japan can now produce most goods more cheaply than we 
can. Such strength as we have lies in a few items, such as computers, jet aircraft, 
chemicals, and agricultural products. In attempting to pay for imports, we face 
the need to export huge quantities of agricultural products, leading to an increase 
in the cost of food for all our citizens. 

Since our economy has become vulnerable to the actions of others, it seems 
worthwhile to explore a few trends in other countries. 

During the past 15 years there have been profound changes all over the world. 
For better or worse, other peoples are trying to adopt what they consider the 
best elements of our life style. Fifteen years ago there were few automobiles in 
the countries of Europe and Latin America and in Japan. Now many of these 
countries manufacture and export automobiles. Today their major cities have 
traffic jams rivaling our own. 

Traffic jams are an annoying and scarcely desirable manifestation. They are, 
however, an indication of something more fundamental — the consumption of 
energy generally in an industrial society. Neither we nor the peoples of Western 
Europe or Japan can maintain a high standard of living without the use of 
substantial quantities of energy. All these peoples are heavily dependent on 
oil as a principal source of energy. Moreover, they are much less well endowed 
than we in terms of fossil fuels, including oil. Thus our effort to import oil 
is a threat to the well-being of many other peoples. They need the oil even more 
than we do, and they are likely to strive harder than we will to get it. 

The advanced countries are not the only ones that have aspirations to utilize 
more energy and to enjoy amenities that technology can bring. The urge to 
share some of our way of life is widespread. 

Recently I visited Mexico, which is a curious mixture of the completely modern 
and the primitive. Many Indians in the mountains continue to live as their 
forebears did a thousand years ago. A Mexican friend told me of an incident 
which seems significant. The Mexican government decided to try to improve the 
standard of living of some of the tribes. The government thought that the 
building of a highway would bring many economic benefits — better markets for 
Indian handicraft products and cheaper costs of goods the Indians then might 
buy. In building the highway it was decided to avoid use of heavy machinery 
and instead to employ Indian labor. Thus for the first time in their lives a group 
of Indians received a substantial pay-off after the first week's work. Naturally 
on Saturday night the Indians went to town. My informant did not reveal all 
that happened in town, but he did tell me that when the Indians returned to the 
labor camp each brought with him three things : a new sombrero which is essential 
to protect against the tropical sun, a string tie which catered to a man's need for 
finery, and a transistor radio. It is difficult for us to guess what went on in the 
minds of those Indians as they listened to their transistor radios. But one thing 
seems certain. An irreversible change occurred. They will never return com- 
pletely to their pre-Columbian life. Their appetite for our gadgets has been 
whetted, and the advertising and other material that they will receive on their 
radios will stir additional desires. What I have just described is only one example 


of what has been happening in many lands. The rest of the world wants to enjoy 
our standard of living. 

As a world leader of the great scientific technological revolution, we have set 
patterns that have led to the weakening of centuries-old social systems. Our 
example has stirred longings and raised hopes and expectations that probably 
cannot be fulfilled. 

There are not on this planet sufficient resources to permit everyone to live in 
the style in which we have been living. Among other things, it would not be 
possible for the world to maintain for long a civilization based on oil. Were all 
the world to consume petroleum and its products at our rate, the fabulous re- 
sources of the Near East would vanish in about two decades. If the world is to 
enjoy for long a comfortable standard of living, substantial adjustment in the 
procurement and uses of energy must occur. These adjustments will be slow, 
as our own experience shows. 

In the coming years we are going to witness a global scramble for oil, other 
raw materials, and food. We will witness frustration, starvation, and disorder. 
Since we are blessed with great natural resources, a highly productive agriculture, 
and a highly competent scientific and technological establishment, we have the 
potential of being able to avoid much of the trauma that others will face. Our 
most vulnerable feature is our dependence on imports of petroleum. 

At present we obtain 78 percent of our energy consumption from hydrocarbons 
— natural gas and petroleum. Liquid hydrocarbons are especially important, for 
they furnish the energy for agriculture and transportation. We could not exist to- 
day without a highly mechanized agriculture and our transportation system — air- 
planes, trains, trucks, and automobiles. Without ample supplies of petroleum, 
our economy would crumble and we would face the most drastic changes in our 
way of life. We need to move to utilize more effectively our reserves of oil shale 
and coal. We need to conserve energy. 

Liquid hydrocarbons are so useful that we should explore vigorously various 
alternatives for obtaining them besides importation. In principle, enormous 
quantities could be obtained from coal and from oil shale. Use of either of these 
sources is certain to raise some environmental problems. Our best way to identify 
and to minimize those problems is by actual experience. Moreover, if we were 
to design and build a few large plants quickly, we would make some expensive 
engineering mistakes, but we would learn fast how to improve our performance. 

The cost of building major plants would be substantial. We should be thinking 
in the range of $1000 million for each of several plants. The government 
must be prepared to share part of the costs. One method might be for the 
government to invite bids to supply over a ten or twenty year period a fixed but 
large quantity of hydrocarbons taken from domestic coal or shale. The genius 
of the research and engineering staffs of the large chemical and petroleum com- 
panies would be turned loose on the problem, and great progress would soon 

Our economy is also vulnerable in its supply of raw materials. We need to make 
a determined effort to lessen dependence on imported materials by designing 
components that utilize locally available substances, and by developing our own 

Given good organization, discipline, and thoughtful use of the tools that 
science and technology furnish, the world could probably move toward an en- 


during economy — one based on abundant and renewable sources of energy, 
materials, and food. However, in much of the world there is lack of organization 
and discipline as exemplified by the continuing population explosion in many 
lands. Perhaps the major contribution we can make is one of example. If we 
set our house in order and behave with a modicum of discipline, we can enjoy a 
prosperous life while leaving our descendants a proper heritage. In future we are 
going to continually bump against limitations, the need to weigh trade-offs, for 
example, between environmental considerations and energy needs. The public 
and the politicians must come to realize the strengths and the limitations of 
science and technology when responding to public needs. Given steady support 
and encouragement, science can give humanity additional power and tools to 
meet needs. But everyone should be aware of the long time span between dis- 
covery and application — in the range of ten years. Moreover, even when the 
technology of an application is established, considerable time must elapse before 
large-scale production can be achieved. 

At the moment our greatest weakness lies in lack of ability to mesh our political 
institutions with technological facts of life. This weakness has been apparent for 
a long time under many administrations. We have shrunk from realistic plan- 
ning, perhaps out of ideological considerations. Of course, rigidity in planning 
may well be a greater handicap than no plan at all. But we have carried our 
reluctance in this matter to extremes and have seemed unable to prepare for the 
future. Our political system is poorly equipped to deal with long term undramatic 

The behavior of politicians necessarily reflects the demands of the public. In 
large measure we have become a society sharply focused on the here and now. A 
generation ago it was necessary for the individual to give thought to the future. 
Individuals had to consider not only their long-term future but also how they 
would survive the coming winter. But social security and other developments 
have removed the necessity for thought and self-discipline. The future will be 
taken care of by others. Public attention is on the here and now, and politicians 
act accordingly. 

As a resident of Washington for more than 30 years and an observer of the 
actions of politicians and bureau heads, my experience can be boiled down to an 
acid comment which though an exaggeration is not very wide of the mark. The 
key officials follow closely press, radio, and TV. They watch "Meet the Press," 
"Face the Nation," and other such shows on Sunday, digest the weekend news- 
papers, glance at Monday morning headlines, and then make national policy for 
the week. Unfortunately, many of these short term decisions have long term 
consequences; for example the Apollo Program. In that case, a momentary 
political necessity led to a decade-long commitment and a program that continues. 

All of us find it difficult to avoid immersion and entrapment in the excitement 
of the moment. But there are a few elements in society whose careers depend on 
being future oriented; for example, some engineers, utilities executives, and other 
corporate planners. Another group is the scientists. They are continually seeking 
new discoveries which will be significant not only to this but to future generations. 
Ultimately, our way of life must be transformed from one dependent on non- 
renewable resources to one with a more enduring basis. The foundation knowledge 
for such a transformation will come from science. 

While we are working to solve immediate energy problems and those of the 
next decade, we should also be moving toward longer range goals. We should 


seek ultimately to derive our energy and materials from renewable resources. We 
should arrange to obtain a substantial fraction of our energy from the sun 
because that energy is enduring and because the use of solar energy would cause 
less damage to the environment than the use of fossil fuels and less danger than 
nuclear reactors. 

In time the federal government will be geared to overcoming our energy prob- 
lems. Past experience, though, is that the government will emphasize a few 
aspects of the many long range problems we face while largely neglecting most. 
It is important that some future-oriented people be given the freedom, and the. 
funding, to identify promising opportunities not in the immediate spotlight. 

When viewed in the context of federal research and development expenditures, 
the annual budget of the Carnegie Institution is minuscule. Nevertheless, in the 
scheme of things we can have a significant function. We are accustomed to 
taking a long term view, to looking for leverage and areas that the fashion of 
the day is neglecting. The record of past achievements and contributions arising 
from the Institution's efforts supports our view that society makes no better invest- 
ment than the kind of education and basic research being carried out in the 
Institution. Our training and research activities relate to aspects of the great 
human needs — energy, materials, and food. Our activities also relate to two other 
needs: health and, for many, the greatest need of all, the need to know. 

At our Department of Plant Biology we are expanding studies of man's 
greatest source of renewable materials. We are studying the process of photo- 
synthesis with new and more effective tools. Using a highly instrumented mobile 
laboratory, we are investigating the ability of plants to thrive under extreme 
conditions, including those in Death Valley. These studies could well lead to 
better management of our fields and forests and perhaps to more productive 

Our studies of geochemistry and geophysics relate to such natural hazards as 
earthquakes and volcanism, to geothermal energy sources, and to the better 
understanding of the evolution of the crust of the earth. These studies also 
relate to economic geology and to the materials sciences. The history of the 
Institution includes many examples of work that has had practical application in 
the steel, cement, ceramics, and glass industries. 

Our research in biology is directed to understanding fundamental processes in 
growth and development, and this effort too has a proud history of contributing 
to the solution of important practical problems while advancing basic knowledge. 
In the pages that follow, Ebert provides a summary of some of the contributions 
of the Department of Embryology, including a study this year of myasthenia 

We cannot foresee practical applications from astronomy. But it is apparent 
that humans have an urge to know that affects them deeply. Each year hundreds 
of thousands of visitors come to the Observatories. In addition, many amateur 
astronomers grind their own lenses and conduct observations of the skies. These 
people are touched by the mystery and grandeur of the universe. Who can look 
out at the stars on a clear cool night and not be moved? 

In the pages that follow, some highlights of the Institution's work are presented. 
The reader can note many studies that relate to long term societal problems, 
such as food, materials, and health. He can also see interwoven our concerns 
with education, growth of fellows and staff, and everywhere, an appeal to the 
human need to know. 

The Year in Review 


Astronomy is in the midst of continuing evolution. Once astronomers were 
principally concerned with optical measurements of visible light. Now the 
heavens are being explored using virtually the entire electromagnetic spectrum 
from gamma rays to multicentimeter waves. This has uncovered new phenomena 
such as quasars, pulsars, and x-ray emitting stars. The vocabulary of astronomers 
has been enriched with words like neutron stars, black holes, and synchrotron 

Much of the new excitement has originated outside the realm of optical 
astronomy, but, as reported by Horace W. Babcock, Director, Staff Members 
at the Hale Observatories have found stimulus and challenge from the new 

Quasars were first identified by a combination of optical and radio telescopes. 
In those days the resolution of radio telescopes was poor. Thomas Matthews 
and Allan Sandage were able to pinpoint the location of a quasar, using the 
California Institute of Technology facility at Owens Valley and the optical 
facilities of the Hale Observatories. The resolution of radio telescopes has since 
been greatly improved, using long base lines. Today their resolution is comparable 
to that of the large telescopes. We are now entering an era in which there will be 
even more intense interaction between optical and radio astronomy. Similar 
comments are likewise applicable to other parts of the electromagnetic spectrum. 
The x-ray satellite Uhuru has detected more than 200 x-ray emitting objects. 
Optical examination of some of these has led to new puzzles. 

The Hale Observatories conduct studies of the sun and other objects in our 
solar system. But most attention is devoted to other stars and to galaxies, 
their structure, evolution, and relationship to quasars. Star formation, stellar 
evolution, radio emission, and high energy physical processes are subjects of 

The controversy concerning origin of the universe continues. Those who ad- 
vocate the big bang hypothesis seem to be able to marshal the most evidence. 
Nevertheless, some believers in a steady state cosmology continue unconvinced. 
This matter is fundamental, and further evidence is important. 



A contribution by Maarten Schmidt provides additional weight against a 
steady state cosmology. 

Schmidt has developed a theoretical analysis of the abundance and flux densities 
of radio sources to be expected from steady state cosmology. He has applied 
this analysis to observational data on strong radio sources provided by the 
Third Cambridge Catalogue. Optical identification of many of these sources has 
been made and their redshifts determined. The actual number of remote radio 
sources observed is considerably greater than can be accounted for by steady 
state cosmology. 

Sandage's investigations of N galaxies show that a typical object of this type 
is a giant elliptical galaxy with a miniquasar in its nucleus, and, furthermore, 
that the magnitude-redshift relationship for the N galaxies leaves little, if any, 
margin for a possible noncosmological component of the redshift. 

Jerome Kristian has extended the search for galaxies underlying quasars by 
identifying the conditions under which such galaxies should be detectable. When 
the galaxies are relatively distant, one would not expect to resolve the light from 
the quasar as against that from the remainder of the galaxy. In contrast, it should 
be possible to differentiate between the quasar and its accompanying galaxy 
when they are at an intermediate distance. Critical examination of 200-inch 
photographs of 26 quasars gave results in good agreement with the predictions. 
The evidence supports the increasingly general view that quasars occur in the 
nuclei of galaxies. 

Increasing attention is being given to the history of star generation in galaxies 
and to the processes that trigger large-scale bursts of star formation. Studying 
ring galaxies, for example, Leonard Searle and Wallace Sargent have obtained 
from multichannel scans evidence that the energy distributions of the rings are 
closely similar to those of star clusters a few hundred million years old. The 
same age is indicated for these galaxies from the measured expansions of the 
rings. Searle and Sargent believe that these galaxies are the first truly young 
extragalactic systems to be recognized. They also conclude that the outer reaches 
of certain paired galaxies, distorted by mutual interaction, consist of predom- 
inantly young stars. 

The concept that density waves or compressional effects in galaxies accelerate 
the formation of stars from the interstellar material is finding increased support 
not only in situations such as the foregoing, but in the spiral structure and 
nuclear regions of galaxies in general. Pieter van der Kruit, examining the high- 
resolution radio brightness distributions within galaxies, obtained at Westerbork, 
The Netherlands, found that the radio brightness distributions were displaced to- 
ward the inner edges of the optical spiral arms, thus confirming the concept of com- 
pressed regions in the density wave picture of spiral structure. The radio spiral 
structure is believed to result from compression of the magnetic field and rela- 
tivists electrons in the large-scale "base disk" of a galaxy. Star formation is 
triggered in the density wave compression regions, in a manner governed by the 
degrees of compression. Van der Kruit concluded that the relationship of specific 
angular momentum to mass is the governing parameter. His conclusions were 
reinforced by observations at Palomar of the rotation curve for NGC 4321. 

Numerous attempts have been made to identify optical counterparts of recently 
discovered galactic x-ray sources. One such object is the binary BD + 34°3815, 
which is a weak variable radio source and is identified with Cygnus X-l. The 
star is a 5.6-day spectroscopic binary of which the primary optical component 


is of spectral type BO lb, with a velocity amplitude of 74 km s" 1 . Robert Brucato 
and Kristian, observing the star spectroscopically at Mount Wilson, discovered 
that the weak emission feature at A4686 (He II) had a velocity amplitude nearly 
twice as great as that of the B star and that it varied in antiphase; no other 
evidence of a secondary spectrum is seen. After analyzing the evidence, Brucato 
and Kristian concluded that one possibility is that the secondary component 
of the binary is a black hole, the x rays being produced by the thermalization 
of gravitational energy of matter falling in the vicinity of the black hole. 

The pulsating x-ray source Her X-l associated with the variable star HZ 
Herculis is being studied by J. Beverley Oke and Jesse Greenstein with the 
multichannel spectrophotometer and the coude image-tube spectrograph. This 
fascinating star shows an optical variation which requires that the side facing 
the x-ray pulsar be quite hot, and the opposite side near 8000°. Resolution into 
two, or possibly three, different major components on the surface of this optical 
object is required by the multichannel observations. 

Spectroscopy by Greenstein with the coude image-tube at the 200-inch telescope 
has continued to show that the hydrogen-line white dwarfs have quite small 
rotational velocities. At present, five of the six stars successfully observed show 
cores, formed in nonlocal thermodynamic equilibrium, that are barely wider than 
the instrumental resolution. If these stars are viewed as the central regions of 
former red giants, an efficient mechanism must operate to reduce the angular 
momentum. This occurs in the interiors of whatever stars undergo slow evolution 
into the white dwarf region. 

A large number of the infrared sources that have been detected in the past 
several years have been interpreted in terms of the so-called dust shell model. 
In this model energy from a central object at relatively short wavelengths is 
absorbed by particles in a surrounding envelope and reradiated at the much 
lower temperature characteristic of the grains. In this way the maximum of the 
energy distribution can be shifted far into the infrared, perhaps totally obscuring 
the original energy distribution of the central star. Unfortunately, direct obser- 
vational evidence for these shells other than the observed energy distributions 
themselves is largely lacking. For objects sufficiently near the ecliptic, however, 
lunar occultations provide a useful method of measuring angular extent of infrared 
sources and may provide direct evidence for a shell. 

IRC + 1011 (= CIT 3) is a bright infrared source that also emits microwave 
lines of the OH molecule. The object has recently undergone a series of lunar 
occultations. Three of these events could be observed at Palomar and Mount 
Wilson and attempts were made by Robert Zappala, Eric Becklin, and Gerry 
Neugebauer to observe these simultaneously in two wavelengths with the 200-inch 
and 101-inch telescopes. Successful observations were obtained at 2.2 and 10.1 n 
on one occasion, and at 20 n on another. 

These observations provide clear evidence for a shell of material around a 
smaller central object. If the total luminosity of the object is about 10 4 L Q , the 
distance to the star is about 500 pc and the observed extent of the outer dust shell 
corresponds to a diameter of about 60 A.U. The temperature of the dust shell 
is about 600 °K. 

In discussions of scientific accomplishments, emphasis is commonly placed 
on concepts, ideas, and results. Somehow, the means of achieving new insights 
gets short shrift. This is unfortunate, for practically every field of science is 
heavily dependent on new instrumentation or techniques as a means of creating 


new opportunities for examining nature. This is especially true in astronomy. 
Over the years, one of the strengths of the Hale Observatories has been their 
facilities and instrumentation. That tradition is being maintained. Construction 
of the 101-inch Irenee du Pont Telescope at Las Campanas is proceeding and will 
result in a splendid observatory in the southern hemisphere. 

Under the general supervision of Bruce Rule as Project Officer and with the 
collaboration of many members of the scientific, engineering, and administrative 
staffs, substantial progress was made in meeting scheduled design requirements, 
guidance of manufacturing operations, procurement, and delivery of site con- 
struction components for the du Pont Telescope and dome building. Most 
mechanical components are in advanced stages of inspection and completion in 
preparation for shipment to Chile, with assembly planned at the Las Campanas 
site in 1974, after the dome and building are ready. 

Following tests at the Optical Sciences Center of the University of Arizona 
of the primary mirror support cell and grinding table supports with the mirror 
at a spherical figure, the operation of polishing to achieve the final aspheric figure 
was started in early summer 1972 and continued throughout the year. The work 
is under the supervision of Don Loomis, Chief Optician for the telescope optics, 
with close cooperation in testing by Arthur Vaughan and Eugene Fair. Design 
and fabrication of optical test equipment and development of computational 
procedures were pursued at the O.S.C. and by Vaughan in Pasadena. By April 
1973 all test procedures required for evaluation of the final optical figure of the 
primary were in operation. The mirror figure was brought to within 0.5 micron 
of the final surface by June 1973. 

Following a number of general meetings to discuss auxiliary instrumentation, 
several working groups were set up by Oke to supervise the design and construc- 
tion of specific instruments and facilities. It is the hope that most of the initial 
set of instruments will be ready for use when the telescope is completed. The 
groups (with the chairman) are listed below: Direct Plate Holder and Darkroom 
Processing Equipment (Sandage) ; Instrument Mounting Base (Oke) ; TV Guid- 
ing System (Edwin Dennison) ; Image-Tube Spectrograph (Searle) ; Broad-band 
Photometer (Kristian) ; Infrared Photometer (Neugebauer) ; 90-mm Image-Tube 
(Brucato) ; Coude Spectrograph (Vaughan) ; Cassegrain Chair (Rule) ; Computer 
Interfaces (Neugebauer) . 

Substantial progress has been made by the Astroelectronics Laboratory under 
Dennison toward the design and completion of the control system for the du 
Pont Telescope. A large fraction of the effort has gone into ensuring that the 
design is satisfactory in two fundamental interface areas: between the control 
system and the mechanical parts of the telescope and between the control system 
and the operator. This initial study phase was essential to minimize the possibility 
of human error in operating the telescope and to eliminate the necessity for 
on-site modification of the system when it is installed with the telescope in Chile. 

The Hale Observatories are fortunate in benefiting from multiple initiatives in 
development and use of electronic instrumentation. Caltech has been the source 
of a number of new approaches. The SIT-Vidicon photometer developed by 
James Westphal is working out very well, and it provides increased opportunities 
for study of far distant or dim objects (magnitude 23). 

Paul Harvey, Michael Werner, Jay Elias, and Ian Gatley of Caltech have 
constructed equipment for use with the 200-inch telescope that has permitted 
observations at 1 mm wavelength. 


Staff of the Hale Observatories have also been active. Oke and Dennison have 
been carrying on a joint effort with Princeton University to develop, build, and 
test a digital image recorder. The instrument has been tested successfully at 

Vaughan has been developing equipment that has enhanced capability of 
detecting and measuring magnetic fields of stars. 

The Astroelectronics Laboratory under Dennison emphasizes quality and de- 
pendability in equipment and brings to the Observatories engineering skill. 

Of considerable significance to astronomy generally has been a new technique 
in the use of photographic plates developed in part by William Miller of the 
Photographic Laboratory of the Hale Observatories. Using new IHa-J plates 
provided by Eastman Kodak, combined with a bake-out procedure conducted 
in dry nitrogen, sensitivity to weak light sources has been increased by a factor 
of 3. 


Students of the earth sciences have been enjoying a most stimulating period. 
At a time of increased emphasis on relevance, they can take pride in the fact that 
their science is central to the search for scarce materials. Simultaneously, many 
aspects of their intense exploration of the unknown have broad public appeal. 
As scholars they are fortunate, for they can contemplate many challenges. New 
opportunities and frontiers have been opened by a combination of big science 
projects, the concept of plate tectonics, and a host of advances in techniques and 
instrumentation, as indicated by Hatten S. Yoder, Jr., Director, in the Geo- 
physical Laboratory report. 

The Apollo program has been completed, but the astronauts brought back 
invaluable samples from the moon, providing some of the best opportunities 
to probe the early history of the solar system. 

The Joint Oceanographic Institutions for Deep Earth Sampling (JOIDES) has 
made available many cores and new information and questions. The concept of 
plate tectonics has illuminated many old puzzles while creating new ones. New 
instrumentation, largely electronic, has vastly increased the effectiveness of data 
gathering and data handling, and in some instances made accessible phenomena 
hitherto unexplored. Research at the Geophysical Laboratory emphasizes funda- 
mental understanding of the earth and solar system rather than applications. 
However, the history of the Laboratory is rich with examples of practical applica- 
tions arising out of basic work performed there. 

An excellent example of the interplay of the concept of plate tectonics and 
the opportunity created by instrumentation may be seen in Francis Boyd 
and P. H. Nixon's study of kimberlites of Lesotho, South Africa. The kimberlites 
are in themselves very interesting rocks. They are the major source of the 
world's gem grade and industrial diamonds. A number of lines of evidence point 
to an origin deep in the earth's mantle, and a sudden, violent propulsion to the 
surface through a pipe extending from the deep mantle through the crust of the 
earth. The rock, the diamonds, and especially the enclosed fragments of ultramafic 
rock, usually rounded into nodules, are of great interest to petrologists because 
they provide samples and record processes within the mantle. The depths from 
which these materials come have been unknown, yet a remarkable variety of 



rocks is incorporated and transported upward in the explosive ascent of kimber- 
lite to the earth's surface. Boyd and Nixon have been attempting to establish 
the depths from which the nodules come by examining the properties of the 
coexisting minerals in these rocks. The compositions of the minerals reflect the 
temperature and pressure that they last experienced before being torn loose 
and carried up in the explosion. By carefully measuring these compositions and 
relating them to those determined on similar minerals in the laboratory at 
measured temperatures and pressures, the record can be read. Boyd and Nixon 
argue that the nodules reflect a series of temperatures and depths that essentially 
define an ancient geothermal gradient down to about 250 km (Fig. 1). The 
geotherm was disturbed, probably by shearing that resulted from the plate move- 
ments accompanying the break-up and dispersal of Gondwanaland. The nodules 
that plot below the inflection are sheared (see Fig. 1), and those represented by 
points above are granular. They suggest that the point of inflection locates the 
top of the low-velocity seismic zone in Cretaceous times (100 million years ago) 
and also marks the top of a crystal-mush zone from which the kimberlite magma 
originated. This reconstruction is the first experimental attempt to define a geo- 
therm at a particular moment in the evolution of the mantle. 



o* 1200 











Depth, kilometers 

Fig. 1. An experimental estimate of the Cretaceous (100 m.y. ago) geotherm compared with 
the present calculated shield geotherm (dashed) in northern Lesotho, South Africa. Based on 
temperatures and depths of equilibration of pyroxenes from lherzolite nodules from kimberlite 
pipes calibrated with experimental data on synthetic systems. 


The new conclusions rest on a solid foundation — accumulated knowledge of 
silicate systems that provided the necessary geobarometers and geothermometers 
and a highly effective electron microprobe automated by L. W. Finger that 
facilitates collection of accurate data on many samples of kimberlite. 

Another example of new knowledge gained through powerful equipment and 
excellent technique is provided by Ho-Kwang Mao's studies of the behavior of 
minerals at very high pressures. He has examined materials thought to be 
especially abundant deep in the earth's mantle. The experiments are conducted 
in a diamond-faced pressure cell. Optical examination and determination of 
electrical conductivity of materials under pressures as high as 300 kbars (« 1000 
km depth) at temperatures up to 800° C can be carried out in this equipment. 

The measurement of properties of crystals such as compressibility, thermal 
expansion, density, electrical resistivity, and index of refraction yields clues 
to the nature of the mantle of the earth. For example, the compressibility, thermal 
expansion, and density give a measure of the geothermal gradient, thermal con- 
ductivity, and seismic velocity. The index of refraction can be directly related 
to the dielectric constant, which together with electrical resistivity defines the 
behavior of the earth in an electric field and reveals the origin of the earth's 
magnetic field. Mao has measured the optical and electrical properties of crystals 
believed to be important in the earth's mantle — olivine, spinel, and mag- 
nesiowiistite. The crystals absorb an increasing amount of light in the visible 
and near infrared wavelength region and thus darken as the pressure is raised 
to 150 kbar. The frequency of the light absorbed is directly related to the 
radiation transfer and indirectly to the electrical conductivity. His observation 
thus suggests that no significant heat transfer in the mantle is likely to be by a 
radiative process, as previously thought. In other experiments Mao made direct 
measurement of the electrical conductivity, which was sufficiently high to suggest 
that heat transfer in these substances may be mainly the result of electronic 
thermal conductivity. The new data coupled with other parameters are essential 
in constructing models of temperature distribution and of the thermal history 
of the earth. Simple physical measurements on crystals are indeed of great value 
in ascertaining the properties of the earth. 

The nuclear age has produced a wealth of techniques for studying rocks and 
minerals. One technique, particle-track mapping, has opened a new approach 
to determining partition coefficients of trace elements between crystals and be- 
tween crystals and surrounding melt. Martin Seitz has developed a method 
whereby uranium, thorium, and boron can be measured in amounts of 0.1 ppb, 
1.0 ppm, and 1.0 ppm, respectively. He found, for example, that the activity 
coefficients of uranium in the crystal and melt are independent of concentration, 
and the results support the view that silicate melts are dilute with respect 
to uranium at concentrations found in natural systems. In addition, the parti- 
tioning between crystal and melt does not seem to depend appreciably on the 
composition of the melt or the temperature and pressure of the system. Those 
properties simplify the application of experimental data to studies of geologic 
processes. From the very low concentrations of uranium and boron, it appears 
that ultramafic rocks and basalts from the oceanic ridge may have formed in the 
same fractionation event. Radioactive elements such as uranium and thorium 
potentially date the time of their synthesis in stellar environments. Similar 
partitioning studies of these elements in meteorites are essential for a better 
understanding of the chronology of nucleosynthesis. 


Radioactive isotopes are particularly useful not only for measuring the time of 
formation of a rock but also for dating major changes in rocks as they take part 
in various geological processes. The limitations are no longer imposed by ana- 
lytical techniques but by one's ability to unravel the complex response of the 
isotopic systems in the constitutents of the rock to the processes. Thomas Krogh 
and Gordon Davis realized that much of the confusion in dating geological events 
is due to the different response of the various minerals and to the different be- 
havior of the elements of the dating systems used during metamorphism. Much 
effort was put into comparing these responses in two rock layers about 1850 m.y. 
old that underwent metamorphism about 1000 m.y. ago. Each layer was 
initially formed with a different Sr-Rb ratio, and large differences developed 
in the Sr 87 /Sr 86 ratio during the long time between formation and metamorphism. 
The limits of isotopic mixing were uniquely determined: Chemical equilibrium 
had not been obtained in distances of even a few centimeters during the meta- 
morphism. It was shown for the first -time that a single outcrop can yield two 
vastly different whole-rock Rb-Sr isochron ages. The zircons displayed an added 
complication. By analyzing the ends of 200 micron grains, it was possible to 
prove that overgrowths of new zircon formed during the metamorphism. These 
overgrowths are not always evident optically, and therefore the U-Pb dating 
method may yield confusing results. Their analyses of these rock layers cast 
considerable doubt on the attainment of isotopic equilibrium during regional 

The bulk of the organic matter that accumulates in a sediment or a soil 
bears little resemblance to the materials in living organisms from which it is 
formed. A significant fraction of this material, extracted by dilute alkaline solu- 
tions and precipitated by acidifying the extract, is called humic acid. The chem- 
istry of this organic matter has been difficult to study, and an adequately established 
theory for its origin, chemistry, and geological fate does not exist. Over 60 years 
ago Maillard noted the similarities between natural humic substances and the 
products of the reaction between sugars and amino acids. He proposed that these 
products, called melanoidins, were the precursors to humic acids. However, most 
workers consider that lignin of woody plants is the source of humic acid. Because 
the lignin hypothesis entails difficulties in explaining natural occurrences, Thomas 
Hoering has followed up on some studies of the melanoidin model initiated 
earlier by Abelson and Edgar Hare. Hoering has found a number of additional 
chemical and physical properties of synthetic melanoidin that are comparable 
to those of natural humic acid and has concluded that the Maillard hypothesis 
forms a logical basis for the origin of humic acids. He has further shown that the 
small amounts of aromatic and phenolic acids, derived from humic acids, which 
are key pieces of evidence for the lignin hypothesis, may arise from natural 
diagenetic processes and are not necessarily indicative of lignin. The melanoidin 
model is a base for new insights into the structure and reactions of humic acids. 


The Department of Terrestrial Magnetism carries a name which ill describes 
its activities. Yet if one insisted that the designation describe the functions, one 
would be faced with the unhappy choice of frequent name changes to reflect the 
evolution of programs or of freezing the program to preserve the preciseness of 
the name. 


Over the years most of the Department's activities have been in geophysics, 
with substantial contributions in biophysics, nuclear physics, and radio and optical 

By its l are, geophysics implies international activities, and the work of the 
Department since its inception has included a strong international component. 
In this year's report of Ellis T. Bolton, Director, both geophysics and its inter- 
national aspects are highly evident. 

The Geophysical Laboratory and the Department of Terrestrial Magnetism 
are located about a mile apart in Washington. There is considerable interaction 
of the two Departments in research in the earth sciences, with the staffs sharing 
major equipment, exchanging ideas, and often joining together in projects. 

Much of the geophysical and geochemical research at Terrestrial Magnetism 
has been stimulated by the concept of plate tectonics. The rim of the Pacific 
basin is restlessly undergoing tectonic activity. Together with colleagues in Japan 
and in South America, staff scientists are examining some of the features of plates 
and plate motions. 

Selwyn Sacks and Hiromu Okada of Hokkaido University, now a predoctoral 
Fellow at the Department, have undertaken a comparison of the Q structure 
beneath the South American continent and the Japanese Archipelago. Similar 
spectral ratio techniques were used for Q determinations in the two regions. 
When waves travel through a medium, they are attenuated, and the more attenu- 
ated the lower the Q. Low frequency waves are attenuated less than those of high 
frequency. A solid rock tends to have a high Q ; a partially melted viscous mate- 
rial has a low Q. In South America, Q p in the somewhat seismically active wedge 
bounded by the seismic plane, the trench, and volcanoes is 1000 to 2000. Beneath 
Japan in the corresponding region, which is aseismic, Q p is 400 to 500; this value 
is identical to that in the asthenosphere below the trench. In South America, the 
high Q p values extend down to 350 km, but there do not seem to be any paths 
from deep earthquakes (600 km) with Q p greater than 1000. In Japan, however, 
the high Q v ' values in the Benioff zone (^3000) persist down to the deep earth- 
quakes. Above the deep earthquake zone in Japan and Fiji, very low Q p values 
(100) were found, in contrast to a minimum value of 350 in South America. The 
tectonic implications of these data are that lithospheric plates have varying thick- 
nesses and that continental plates are several times thicker than oceanic plates. 

In another study of events accompanying tectonic processes, Stanley Hart and 
J. G. Schilling of the University of Rhode Island have examined basalts from 
Iceland and the adjacent track of the mid-Atlantic ridge with respect to trace 
element composition and Sr 87 /Sr 86 isotope ratios. Tholeiites from a 500 km sec- 
tion of the Reykjanes ridge and its extension on Iceland were analyzed for K, Rb, 
Cs, Ba, Sr, and Sr 87 /Sr 86 Samples most distant from Iceland show 87/86 ratios 
similar to typical ridge basalt (0.7026 to 0.7028) but are more highly depleted in 
K and Sr (K < 400 ppm, Sr < 80 ppm) and show less fractionated element ratios 
(K/Rb < 900, K/Ba < 65). Concentrations of the trace elements increase pro- 
gressively as Iceland is approached (factors of 2 to 5) ; 87/86 ratios increase 
stepwise to 0.7030 about 150 km from Iceland. Tholeiites from Iceland show a 
considerable range in concentration (K, 700 to 2000 ppm; Sr, 90 to 200 ppm) 
which overlaps and is somewhat higher than the nearest ridge samples. Ratios 
of 87/86 on Iceland show little spread (0.7030 to 0.7031). The isotopic data show 


that the mantle sources under Iceland are chemically different from those under 
the southern Reykjanes ridge. The concentration variations between the two 
provinces are not well described by mixing models having only two melt types. 
A possible model involves variable partial melting and mixing of two mantle 

Nobumichi Shimizu and Hart have developed a sensitive chemical procedure 
for the analysis of trace elements in a two-phase system in which one phase is 
crystalline and the other, a quenched liquid, is a glass. They call their procedure 
the "Differential Dissolution Technique," or "DDT." The procedure takes ad- 
vantage of the fact that the glass portion is dissolved in dilute hydrofluoric acid 
much more readily than the highly ordered crystalline portion. Thus, in a series 
of model experiments, it was found that in a mixture of clinopyroxene (from a 
spinel lherzolite from Salt Lake Crater, Hawaii) and glass (tholeiitic basalt glass 
from the Juan de Fuca Ridge) the glass dissolved, leaving pure clinopyroxene as 
a residue. The technique permits accurate determination of the solubility of very 
small amounts of an element in the crystalline phase, for the crystalline phase 
can be freed of contaminating glass. 

Nuclear physics studies have been part of the program at the Department for 
more than four decades. The venerable Van de Graaff accelerator has been up- 
dated, and it has features that facilitate experiments such as those with polarized 
beams. This year Louis Brown and colleagues have been examining an interesting 
phenomenon known as electron promotion. When target atoms are bombarded by 
heavy ions, the simple theory which assumes an interaction of two point charges 
during Coulomb excitation must be modified to account for the presence of the 
projectile ion within the K shell of the target atom. The incoming ion resides 
within the target atom for long periods of time compared with the oscillation 
times of the K electrons. If the electron shells of ion and target atom overlap, 
then a "hyperatom" consisting of the combined nuclei and a new K shell is formed. 
During the time in which the two nuclei are close, two of the four electrons which 
made up the two originally independent K shells are forced out, or "promoted." 
When the two nuclei later separate, there is a high probability that each will have 
a K shell vacancy, or even vacancies in the L shell. In turn, there is a high 
probability of ionization and a consequent high cross section for x-ray production. 
In cooperation with H. A. Van Rinsvelt of the University of Florida, studies were 
undertaken to bombard copper with beams of alkali ions which are readily pro- 
duced in the DTM Van de Graaff accelerator and offer a wide range of atomic 
numbers. It has been found that the characteristic x-ray lines of copper are 
shifted in energy as a result of the high probability of ionizing the L shell when 
the K shell is ionized. 

Activities in Optical Astronomy at the Department of Terrestrial Magnetism 
were initiated in about 1960. The impetus came from Merle Tuve and his initia- 
tive in pushing development of the Carnegie Image Tube. To implement this 
effort, Kent Ford was brought to the Department. He was very helpful in the 
successful development of a series of tubes. As part of the development program, 
testing of the devices was carried out at Lowell Observatory and elsewhere. Ulti- 
mately Ford was joined by Vera Rubin and the two have maintained a modest 
but effective observational program. One of their recent discoveries is the aniso- 
tropic distribution on the celestial sphere of a class of galaxies known as Sc I — 
a special class of spiral galaxies of high luminosity showing photographic qualities 


of being quite blue in color and having reasonably clearly defined arms in the 
spiral structure. 

From an all-sky sample of 208 such galaxies, radial velocities, that is, measure- 
ments of the speed at which these objects are receding from the sun, have been 
determined for 70-odd; 50 velocities range from 4000 to 7000 kilometers per 
second. Strikingly, 28 of these have velocities tightly grouped (4966 ± 122 km/ 
sec) while 22 have a similar grouping (6431 ± 160 km/sec). Each group is posi- 
tioned on a different region of the sky (approximately one-third of the total for 
each) and there is virtually no overlap in the positions of the two groups. These 
observations raise important problems; the most unorthodox is to question the 
validity of the widely accepted idea that the Hubble expansion (which relates 
recessional velocity to distance) is isotropic throughout the universe. 

Much of the work of the Biophysics Section during the past decade has been 
devoted to study of nucleic acids and especially to the matching of DNA strands 
derived from the same and from different organisms. These studies have illumi- 
nated questions about organization of chromosomes and evolution of the species. 

Many mechanisms are at play in establishing genetic diversity even among 
bacteria. Ultimately, according to present understanding, diversity is traceable 
to the message encoded in the nucleic acids. Some portions of bacterial or viral 
genomes are highly conserved, as would be required for the production of enzymes 
and enzyme systems necessary for accommodation to the environment or changes 
in the environment and hence to survival. Enzymes, being proteins and generally 
antigenic in foreign species, are subject to immunologic analysis. Modern organic 
chemistry has made it possible to couple appropriate antibodies to insoluble sub- 
strates such as the polysaccharide sepharose and thus permit antibodies to select 
enzymic antigens from crude bacterial extracts with high efficiency. Throughout 
the Enterobacteriaceae the distinctive features of the regulation of the enzyme 
aspartokinase I-homoserine dehydrogenase I (AKI-HSDI) are conserved even 
though the compositions of the DNAs vary widely. It is accordingly possible to 
investigate immunologic affinities of the enzyme, the conformational states of the 
enzyme as ligands are introduced into the system, and to draw inferences con- 
cerning the systematic relationships of specific proteins as contrasted with the 
similarities and differences between the total genomes indicated by nucleic acid 
interactions among the bacteria. These studies are described in detail by Dean 
Cowie and his collaborators, Georges N. Cohen and Paolo Truffa-Bachi of the 
Pasteur Institute in Paris. 

Roy Britten has continued his studies on the organization of chromosomes. 
These studies are being carried out at the Kerckhoff Marine Laboratory in Cali- 
fornia where he collaborates with a group of California Institute of Technology 


The interplay of biological and medical research has been a feature and a 
strength of the Department of Embryology since its inception. The field of human 
embryology was long virtually synonymous with the Department, in which much 
of the anatomic basis was established for today's worldwide thrust toward an 
understanding of congenital defects. In his report this year, James D. Ebert, 
Director, enumerates some of the many interrelations of the Department's work 


with medical research and practice. Much of modern reproductive physiology, 
one of the fields contributing to the technology of population control, sprang from 
the Department and its pioneering monkey colony. In their day, the Lewises 
were pioneers in the then budding technology of cell culture. It is on such a base 
that much of today's animal virology, vaccine production, and cytogenetic screen 
for genetic defects rests. The Department's contributions in the field of develop- 
mental immunobiology, notably its role in helping to elucidate the graft-versus- 
host reaction, have influenced the course of clinical research in tissue transplanta- 
tion. Although it is still too early to predict accurately, it seems likely that 
Masako Yoshikawa-Fukada and Ebert's pioneering demonstration of the viral 
oncogenic sequences may help in understanding tumorigenesis. 

The fundamental research now being carried on in the Department is concen- 
trated on (1) the isolation of nuclear and mitochondrial genes and the study of 
their structure, evolution, and control, and (2) the development of membranes. 
Most of this research at present is obviously not related' closely to medical prob- 
lems, though most of the studies will ultimately contribute toward better practice. 
Already one of the lines of research on membranes has developed a close relation 
to medicine. This is work of Douglas Fambrough and his colleagues on my- 
asthenia gravis. This disease is a neuromuscular disorder characterized by mus- 
cular weakness and fatigue. It has been thought to involve the neuromuscular 
junction. In collaboration with Daniel Drachman and S. Satyamurti of The 
Johns Hopkins University School of Medicine, Fambrough has conducted experi- 
ments designed to determine whether muscle fibers from patients with the disease 
show changes in the number and distribution of acetylcholine receptors. The 
answer is unequivocal: The neuromuscular junctions of patients with typical 
myasthenia gravis contain only 11 to 32 percent of the control number of ACh 
receptors. This finding permits more critical questions to be asked. Is the number 
of receptor molecules reduced? Or are many receptor molecules blocked or altered, 
resulting in a decrease in functional receptors? 

Fambrough and his colleagues will be able to examine these and other questions 
pertaining to acetylcholine receptors. During the last few years he has developed 
and evaluated an excellent method of measurement of the receptors, based on the 
use of a-bungarotoxin combined with 125 I. 

One of the great miracles that we may possibly never fully penetrate is develop- 
ment of the embryo from an egg. The process is perhaps more accurately a whole 
series of miracles, one of the most dramatic of which is the beginning of beating 
in the embryonic heart and the tendency of embryonic heart cells to beat in 
unison. This has long fascinated Robert DeHaan, and he and his associates have 
conducted a series of investigations on the topic. 

DeHaan and Howard Sachs had demonstrated that spheroidal aggregates pre- 
pared from dissociated heart cells beat at rates which were inversely proportional 
to aggregate volume; that is, the bigger an aggregate, the slower it would beat. 
In the past year, working with an able group of students, DeHaan launched a 
multifaceted analysis of this unexpected finding. 

Two sets of properties that might, influence pulsation rate change with increas- 
ing size of such aggregates are (1) those properties that are related strictly to 
number of cells in the system and are independent of spatial configuration, and 
(2) those properties that depend on spatial organization of the parts of the system. 

To perform experiments designed to distinguish between these broad classes 
of rate-controlling phenomena, DeHaan has worked out methods for manipulating 


spontaneously beating aggregates one at a time under direct visual observation. 
When brought into contact, aggregates beating at different rates adhere, and soon 
(usually within 30 to 60 min) synchronize to a common rhythm. Thus, pairs or 
groups of aggregates of selected rates and sizes can be joined to form systems of 
any desired spatial configuration and allowed to synchronize. 

One approach to factors governing beat rate was carried out by Linda Moyzis, 
an undergraduate student, who built up linear chains of 6 to 20 aggregates and 
demonstrated that the synchronous pulsation rate of these chains was only slightly 
reduced from the average rate of the component aggregates prior to synchroniza- 
tion. In this case, the rate was not a function of total synchronized cell number. 

A second approach was taken in experiments performed by Caroline Krueger, 
also an undergraduate student. When sheets of cells were prepared at known 
densities (in cells/mm 2 on the plate) and cut into rectangles of different areal 
sizes, Krueger found that the synchronous pulsation rates of the rectangles did 
not differ from each other in a systematic way over a 60-fold range of areas 
(i.e., number of synchronized cells). Again, rate was found not to be a function 
of synchronized cell number. When, however, sheets were prepared at two differ- 
ent thicknesses (4 cells or 6 cells deep) , the thicker rectangles beat only half as 
fast as the thin ones. The controlling factor appeared to be not cell number but 
diffusional restriction between the inside of the mass and the outside, or S/V 

A major fraction of the Department's activities has been focused on study of 
nucleic acid structures. This area of effort is one of the most fundamental and 
dynamic of all biological research today. Peter Wellauer and Igor Dawid have 
come up with a most interesting series of discoveries relating to structure of RNA. 
They initially focused their attention on the secondary structure of 28S ribosomal 
RNA. Their electron micrographs showed a reproducible secondary structure 
which permits the mapping of these molecules — different regions of each molecule 
and their polarities becoming recognizable. They next applied the technique to 
the study of the "processing" of the ribosomal RNAs. It has been known for 
some time that the 28S and 18S rRNAs originate by the "processing" (or splitting) 
of a larger 45S precursor. Wellauer and Dawid have compared the structures of 
the 45S, 28S, and 18S molecules and have clarified their relationships. The 28S 
region is located at the 5'-end of the 45S precursor. It is followed by a "spacer" 
region, then an 18S region, and another "spacer" at the 3'-end. In "processing," 
the "spacers" are removed. This technique, which holds great promise for the 
elucidation of nucleic acid structures, comes at an opportune time when the ques- 
tion of chromosome structure is coming in for careful attention in laboratories 
throughout the world. 

Dawid has also been deeply immersed in another investigation of unusual 
interest. In collaboration with Ivan Horak, a second-year Carnegie Fellow 
recently returned to Wurzburg, and Hayden Coon of the National Cancer Insti- 
tute, Dawid has been studying the mitochondrial DNAs (mtDNAs) of somatic 
hybrid cells (human-mouse and human-rat). Many readers of last year's Report 
will recall that mtDNAs of both species can replicate in the same cell for extended 
periods. That much can be said unequivocally. However, a more crucial question 
is this: Do they simply "coexist" in the same cytoplasm, or do they interact in 
some way? Even more difficult to answer is the further question, do they recom- 
bine, forming a hybrid molecule? Molecular hybridization studies, detailed by 
Dawid, make it very likely that mouse mtDNA segments become covalently 


linked to human mtDNA during growth in the hybrid cells. A number of hybrid 
cell strains have been examined. Thus far, five strains have both parental mtDNAs 
without recombination, but there are 13 strains in which recombination has 
apparently taken place. 

Another experimental system in which molecular hybridization techniques have 
revealed significant gene recombination is found in Donald Brown and E. Jordan's 
study of the origin of ribosomal DNA homogeneity. Brown's earlier studies had 
shown that the genes for 28S and 18S rRNA in two species of Xenopus, laevis 
and mullein, are quite similar, if not identical, indicating that these sequences 
have not evolved significantly. However, the associated spacer regions differ by 
as much as 10 percent of their some 7000 base pairs. How do these sequences 
evolve — some genes being maintained in a homogeneous state, others changing 
rapidly? It has been proposed that unequal crossing-over can maintain homo- 
geneity in a family of genes. Brown and Jordan have been examining a number 
of possibilities. They have reared hybricl frogs derived from matings of X. laevis 
and X. mulleri, and backcrossed one such animal with both laevis and mulleri. 
The progeny are now sexually mature and the extent to which the ribosomal 
genes have been recombined has been determined. If there were extensive recom- 
bination, the backcrossed animals should have laevis and mulleri genes adjacent 
to each other. The further question may then be asked, when meiosis occurs, will 
there be a "correction"? Brown and Jordan present the several possibilities clearly 
in their account. Here we need only say that 20 individuals from each group of 
backcrossed frogs have been analyzed. All animals tested in the backcross with 
a laevis male contain some mulleri rDNA. Twenty progeny of the mother's back- 
cross with a mulleri male were also examined. Although none had as much 
laevis rDNA as the mother, 18 out of 20 had some laevis rDNA. Brown and 
Jordan believe that this result, while not yet conclusive, is strong evidence for 
extensive recombination in the oocyte between the homologus chromosomes 
containing rDNA. 

Ebert has maintained a personal research program designed to help bridge the 
intellectual and biological hiatus between nuclear material and membranes. To- 
gether with Peter Stambrook and Sachs, he has been examining the close relation- 
ship that exists between cell membrane potential and cell proliferation. In earlier 
experiments the membrane potential was manipulated by varying the external 
potassium concentration, and the results suggested that the membrane potential 
might participate in the initiation or perpetuation of DNA synthesis. The ex- 
periments this year were designed to determine whether an increase in membrane 
potential is required for cells to begin DNA synthesis and to traverse the S phase 
and G 2 to their next mitotic division. A second question examined was whether 
the high K + medium blocks cells randomly throughout the cell cycle, or whether 
it blocks cells at specific time points in the cycle. To gain experimental evidence 
in these matters, Chinese hamster cells were exposed to various external potassium 

Three conclusions emerged from a series of experiments. (1) Cells can initiate 
and continue DNA synthesis while exhibiting a low membrane potential. Syn- 
chronized cells incubated with control medium but transferred to 115 rnikf K + 
medium prior to the initiation of DNA synthesis can traverse the S phase and G 2 
to their next division with little delay. (2) The high potassium medium impedes 
the passage of cells through mitosis. (3) There is an event (or events) during Gi 
that is particularly sensitive to the 115 mM K + medium and that may be required 


for the initiation and perpetuation of DNA synthesis. Mitotic cells suspended 
in 115 ml K + medium exhibit impaired DNA synthetic activity. However, if 
115 ml K + medium is added 2 hours after mitosis, the cells proceed through 
DNA synthesis and the remainder of the cell cycle almost normally. Presumably 
the high K + -sensitive block is passed during the 2-hour incubation in control 
medium, allowing subsequent traverse of the remainder of the cycle in high K + 


The end of the report year was a time of transition for the Department of Plant 
Biology. C. Stacy French, who guided the Department for 26 years, retired from 
his post as Director and has been succeeded by Winslow Briggs. The new Director 
will conserve and build on the present program which contains elements central 
to plant biology, including study of the process of photosynthesis and investiga- 
tions in physiological ecology. In both these areas the Department has made 
distinguished contributions and has helped train many of the leaders. The new 
Director will take new initiatives. He will broaden the program to include more 
emphasis on plant biochemical studies, particularly on membrane processes. 

During French's tenure as Director, the frontiers of plant biology evolved, and 
the research program of the Department changed with them. In the past quarter 
of a century the idea that the mechanism of photosynthesis could be deduced 
solely from kinetic measurements of such effects as oxygen evolution, fluorescence 
intensity changes, or rates of carbon dioxide uptake was slowly abandoned. In 
place of that procedure, the detailed examination of the overall photosynthetic 
process into its individual chemical steps was found to yield more definitive infor- 
mation. In addition to the various forms of chlorophyll, preparative isolation 
biochemistry has shown the presence of a number of highly reactive compounds 
such as cytochromes, plastoquinones, high-energy phosphates, pyridine nucleo- 
tides, plastocyanin, and nonheme iron compounds in the photosynthetic system. 
Current biochemical work centers on the interactions of such substances when 
light is absorbed by the chlorophyll complexes of plant chloroplasts. Kinetic 
measurements now usually refer specifically to changes in known intermediate 
substances rather than to hypothetical components. Many of these known inter- 
mediate electron carriers of the photosynthetic system change color with their 
oxidation state, so their participation in photosynthesis can be followed by meas- 
uring their absorption changes at appropriate wavelengths. 

Close observation of the plant pigments and their changes on illumination has 
been and continues to be a fundamental area for investigation. French and other 
members of the Department of Plant Biology have long been active in such studies. 

In recent years much of French's efforts have been devoted to identifying and 
measuring the various forms of chlorophyll in plant material. One of his strengths 
has been the development of increasingly powerful and sensitive instrumentation 
for such work. This tradition of a flare for instrumentation is being carried on 
by David Fork, who has been fortunate in having the collaboration of Tetsuo 
Hiyama. They are exploiting the latest techniques in tunable lasers and in fast 
responding circuitry to develop instrumentation ultimately capable of observing 
photobiological effects occurring in the order of 10" 7 seconds. Equipment capable 
of meeting their objectives is not commercially available. They have been able to 
utilize some component circuits but have found it necessary to design and build 


others. While they are developing their equipment, they have been using it. and 
have already been able to measure cytochrome changes in times as short as several 

Olle Bjorkman and his colleagues have created a very important new facility 
at the Department for studies of physiological ecology. This effort is in part a 
joint program with the Department of Biological Sciences of Stanford University. 

Knowledge of the physiological and biochemical mechanisms that enable plants 
to photosynthesize efficiently in the great diversity of environments that exist on 
earth is fundamental to understanding primary productivity in different ecosystems 
and of plant evolution and distribution. Such knowledge, inevitably derived from 
experiments on wild plants occupying extreme environments, is also critical to any 
assessment of the potential possibilities of breeding crop plants that can be suc- 
cessfully cultivated in areas presently considered unsuitable. 

The mechanisms underlying the responses and adaptation of plants to contrast- 
ing environments continue to be in the focus of the physiological ecology investi- 
gations. Most of the group's efforts this year have been devoted to the establish- 
ment of new field transplant gardens, design and construction of new measuring 
equipment, controlled growth facilities, and extensive remodeling and modifica- 
tion of already existing mobile and stationary laboratory equipment and facilities. 

A specially constructed, self-contained mobile laboratory unit will permit 
measurements of photosynthetic gas exchange characteristics on the whole 
plant and single leaf levels to be performed with the same instrumentation in the 
field, in the transplant gardens, and at the controlled growth facilities at Stanford. 
As now equipped, the unit houses all of the measuring systems needed for con- 
tinuous measurements of the exchange of carbon dioxide, oxygen and water vapor, 
and conductance to the diffusive gas transport by whole plants and single leaves 
as well as for measurement of external parameters such as radiant energy 
and quantum fluxes, air flow rates and pressures, air, tissue and soil temperatures, 
and water vapor, C0 2 and oxygen concentrations. This integrated system also 
incorporates an extremely compact computerized data acquisition, data processing 
and control system which should greatly improve the speed of data acquisition, 
the flexibility and convenience of operation, and the range and accuracy of the 
measuring system. 

The new transplant gardens, the controlled growth facilities, the mobile labo- 
ratory, and other instrumentation will certainly provide an excellent and in several 
respects a unique tool in the comparative study of plant adaptation to ecologically 
diverse environments and of the underlying physiological and biochemical 

One of French's happy characteristics as a Director has been an ability to 
attract very capable visiting investigators to the Department. They have come in 
part because of the reputation and opportunities of the place. They have also 
come because of the atmosphere that has been maintained there. Over the years 
scores of leading plant biologists from many countries have worked at the labo- 
ratory. In his report this year, French provides a list of those who have worked 
at the laboratory during the past decade. The presence of these guest investiga- 
tors has been stimulating to staff, and the guests have generally been successful 
in their research efforts. This year was not an exception, as can be seen from a 
brief description of some of the results obtained this year at the laboratory. 

The accessory pigments of red algae, phycoerythrin and phycocyanin, function 
in photosynthesis by transferring their absorbed light energy to chlorophyll. 


Ulrich Schrieber of Aachen, Germany, who spent part of the summer of 1972 at 
the Department, found that hydrostatic pressure greatly reduces the efficiency 
of energy transfer from the phycobilin pigments to chlorophyll in red algae. 

Atusi Takamiya, former head of the Department of Biochemistry and Bio- 
physics at the University of Tokyo, spent five months at the Department on leave 
from Toho University. He found that a water soluble and phototransformable 
chlorophyll protein is far more widely distributed in wild plants than had previ- 
ously been realized. 

Ralphreed Gasanov, Assistant Professor of Plant Physiology at Baku, Azer- 
baijan, USSR, visited the laboratory for eight months. He found that the photo- 
system fraction 1 isolated from chloroplast grana by digitonin differs in its pig- 
ment composition from the corresponding material in the chloroplast stroma 
particles located between the grana stacks. 

With the hope that eventually a more direct use of solar energy may be 
possible, Jeanette Brown and Gasanov are attempting to understand the state 
of chlorophyll in vivo and how this essential pigment functions in photosynthesis. 
They have approached this problem by dividing chloroplasts into the smallest 
units that still retain photochemical activity. Successfully applying an improved 
and extended fractionation procedure to an alga that has unusually distinct 
absorption bands of chlorophyll a, they were able to obtain fractions of Dunaliella 
that had activity only for photosystem 1 or only for photosystem 2, the reducing 
and oxidizing light reactions of complete photosynthesis. 

Curve analyses of the absorption spectra of these fractions revealed the relative 
proportions of the forms of chlorophyll present in each photosystem. Chlorophyll 
a-694 is lacking in the photosystem 2 fraction and is present in all of the fractions 
with photosystem 1 activity. The longer wavelength forms of chlorophyll (>700 
nm) seen in spinach chloroplast particles are not found in this alga, suggesting 
that these forms are not essential for photosynthesis but may function as addi- 
tional light harvesting pigment. 

During his return visit to the laboratory last summer, Norio Murata, from the 
Department of Biochemistry and Biophysics of the University of Tokyo, collabo- 
rated with Fork on a continuation of their study on the function of the copper 
protein plastocyanin in photosynthesis. Their previous work, reported in Year 
Book 70, showed that little plastocyanin remained in particles prepared by ex- 
trusion through a needle valve. These particles nevertheless showed rapid electron 
transfer reactions from cytochrome / to the reaction center of system 1 (P700) — 
as fast as those seen in chloroplasts in their native state. These observations 
suggested that cytochrome /, rather than plastocyanin, functions as the "second- 
ary electron donor" to P700. However, the lack of plastocyanin in these particles 
was questioned by some workers, particularly by those who used the so-called 
bioassay method to determine plastocyanin. Murata and Fork found that, although 
the bioassay method is highly sensitive for plastocyanin, it is not necessarily 
specific for this substance. After doing extensive studies on the contents of plasto- 
cyanin in chloroplasts and in subchloroplast particles prepared by a variety of 
methods to release plastocyanin, they confirmed their earlier conclusions that 
plastocyanin is absent from particles derived by needle valve extrusion and that 
it does not function between cytochrome / and P700. It would seem from these 
results that cytochrome / thus plays the role of the secondary electron donor to 



The Institution has two principal, related objectives. One is to advance knowl- 
edge. The other is to foster intellectual growth through educational activities. 

The young people who are trained by the Institution are a highly select group. 
They receive special individualized tutelage from senior staff and they have 
unusually good access to advanced facilities. Predoctoral and postdoctoral- fellows 
do not serve lengthy apprenticeships as members of research teams. Rather, they 
are encouraged to engage in carefully monitored but independent scholarship. In 
consequence both of the major objectives of the Institution are advanced, for as 
the individuals grow they are creative in research. 

The academic orientation of the Institution is particularly obvious in the three 
departments that are closely associated with particular universities. 

In the joint operation of the Hale Observatories by the Institution and the 
California Institute of Technology, teaching and research are inextricably inter- 
twined. Staff Members of the Observatories serve as thesis advisers for Ph.D. 
candidates and help train students in observational astronomy by supervising 
various research projects. During this report year, of the 120 major papers listed 
in the Observatories' bibliography, 28 were either single author papers by post- 
doctoral Fellows or joint efforts with Staff Members, and an additional 19 were 
written by or were in collaboration with Caltech graduate students. 

The training in, and the teaching of, research is a major activity of Staff 
Members. Particularly active at the Hale Observatories is the infrared studies 
group under Neugebauer. Fellows during the recent past include Harry Hyland, 
now returned to Australia to begin his own infrared group, and David Allen and 
C. G. Wynn-Williams of the United Kingdom. As is usually the case, these 
Fellows provide the seed nuclei for the spread of science in their own countries. 
Becklin, a Staff Associate of the Observatories, also a member of the infrared 
group, discovered the Galactic Center in infrared radiation while he was a gradu- 
ate student at Caltech, preparing his thesis using telescopes on Mount Wilson. 

During the current report year, notable student activities within the Carnegie 
Institution at the Observatories include (1) a pre-thesis project in which Caltech 
student W. G. Bagnuolo is collaborating with Searle and Sargent on the history 
of star formation in galaxies as related to questions of chemical evolution of the 
galaxies as a whole and the age distribution of blue systems; (2) the mastery by 
van der Kruit (Carnegie Fellow) of techniques of spectroscopy of extended 
galaxies, which has led him to develop further the analysis of noncircular (ex- 
plosive?) motions in some galaxies with active nuclei; (3) the studies of Robert 
Kirshner (a predoctoral student at Caltech) in collaboration with Searle and 
Sargent in which he developed the interpretation of supernova spectra to an extent 
that fundamental distances to these stars might be determined from their spectra 
alone, leading perhaps to an independent calibration of the scale of the universe; 
(4) the studies of Michael Hart (Carnegie Fellow) with Robert Howard that have 
shed light on the mystery of the solar limb redshift of spectral lines, perhaps 
solving an enigma of 60 years; (5) the investigations of Ermanno Borra (Carne- 
gie Fellow) studying with Vaughan, that helped to develop new methods of 
measuring magnetic fields in white dwarfs — a fundamental problem in stellar 
evolution; and (6) the finding by Eduardo Hardy (Carnegie-Chilean Fellow, 
under Sandage's guidance), that Abell's hypothesis — that there exists a unique 


feature of the luminosity function of clusters of galaxies that can be used as a 
general distance indicator — is untenable. 

In addition to the graduate programs, the Observatories have a summer program 
for the training of selected undergraduates in methods of research by apprentice- 
ships to Staff Members. Particularly noteworthy are the studies of Robert Harri- 
son, a Caltech undergraduate in astronomy, under Kristian and Westphal in the 
laboratory tests and development of reduction of techniques for the silicon-vidicon 
area photometer. Harrison's exposure to research problems at a professional level 
and his fine ability have been a benefit both to the Institution and to himself. 

The Institution's Department of Plant Biology occupies ten acres in the middle 
of the Stanford University campus. Naturally, there is intense interaction be- 
tween Staff Members and Fellows of the Department and faculty and students at 
Stanford. Joint seminars, joint projects, and a sharing of facilities are customary. 

The educational ties of the Institution to Stanford University were reinforced 
recently when Briggs, the new Director at Plant Biology, was named Professor 
of Biological Sciences at Stanford. Briggs' new appointment includes membership 
in the academic senate of Stanford. 

Graduate students as well as postdoctoral students benefit from the educa- 
tional experiences provided by the Department. William Hagar, a postdoctoral 
Fellow, commented recently that his experience was immeasurably enriched by 
associating with visiting scholars of such renown as Professor Atusi Takamiya, 
formerly from the University of Tokyo, and Professor Hemming Virgin of the 
University of Goteborg, and that the opportunity for exchange of ideas between 
young and mature scholars is one of the many exciting opportunities available 
in the Institution. Charles Weiss, who recently completed his undergraduate 
studies in the Biology Department of Whitman College, Washington, was associ- 
ated with the physiological ecology group for several weeks during the summer 
and found that the Department serves as a testing ground for students who 
want to experience research in the plant sciences before committing themselves 
to further graduate studies. 

The sharing of facilities and programs with Stanford is particularly evident 
in the physiological ecology studies. The mobile laboratory owned by Stanford 
is docked at the Department where it keys into experimental facilities. The 
total enterprise, including field stations in Death Valley and at Bodega Bay, is 
operated jointly as a cooperative venture by Bjorkman of Plant Biology and 
Professor Harold Mooney of Stanford. The studies involve participation of many 

Other facilities maintained by the Department both at Stanford and at the 
field stations provide a wealth of educational opportunities not only to those 
working in the plant sciences but also to those working in other fields. 

The mountain stations, for example, provide a valuable teaching and research 
resource. This summer Professor Theodosius Dobzhansky along with Professor 
Francisco Jose Ayala and associates from the University of California at Davis 
continued their experiments on the genetics of the fruit fly Drosophila at the 
Mather Station. Richard Papp, an entomology graduate student from the 
University of California at Berkeley, used the Timberline Station as a base f( ' 
his research on the ecology and distribution of snowfield insect populations. 
Another entomology student, John Horn, has used the Mather Station in his 


study of insects associated with the Sierra Corn Lily (Veratrum). David Burdin, 
a student of Professor Larry Swan from California State University at San 
Francisco, was at the Mather Station also working on a research project in 

Stanford University makes continued use of the mountain stations as part of 
its teaching program. Each year Professor John Thomas and Professor Mooney 
from the Department of Biological Sciences take field trips with their students 
to these stations. Professor Craig Heller and his graduate students are using the 
Timberline Station for their study of the distribution of rodents across the Sierra 

Professor Clifford Schmidt and associates from the California State University 
at San Jose have used the Mather and Timberline Stations in their study of the 
effects of overuse by people and livestock on the fragile mountain ecosystems. 
This study may be useful in defining what limits must be placed on mountain 
wilderness use. 

The Department of Embryology, located on the campus of The Johns Hopkins 
University, interacts strongly with that institution while maintaining its own 
fellowship program. When asked to comment on educational activities at Em- 
bryology, Ebert submitted the following brief essay. 

How does one evaluate an educational program, one in which no examina- 
tions are set and no grades assigned, one whose purpose is to aid in the produc- 
tion and training of investigators? There are many yardsticks, of course — the 
placement of graduates and fellows and their ultimate rise to positions of 
responsibility, their imprint, in turn, on others, and most importantly the 
impact of their research on the field. One measure of the last criterion is the 
citation index: How frequently does a given article, or the output of a given 
individual, reappear in the literature? The approach can be extended to the 
output of a given department or institution. 

The Twenty-Fifth Anniversary volume of the American Institute of Biolog- 
ical Sciences, Challenging Biological Problems, 1 was designed to kindle thought 
and spark discovery. This collection of articles by distinguished biologists was 
meant to be not a series of commonplace reviews, but a series of blueprints for 
the future, in which only the most critical building stones were to be used as the 
basis for projection. It is especially interesting to see that in F. H. Wilt's 
chapter on developmental biology, "The impact of molecular biology on study 
of cell differentiation," 23 of the 98 articles cited were the products of Carnegie 
students and fellows, and of their students in turn. It is a small sample, to be 
sure, but it is reassuring nonetheless to see how significant an impact the Insti- 
tution has had in this field in the past decade. 

The Institution's Staff Members are teachers, but not in Louis Agassiz's 
words "simple machines of study." In fact Professor Agassiz's address on July 
8, 1873, the opening day of the Anderson School of Natural History, 2 in which 
he described his own relationship with his students, epitomizes some of the 
relations between preceptor and student in the Department of Embryology. 
Agassiz said, in part, "Every morning when I can be here I propose to give you 
a piece of advice about the way to employ your time. In proportion as you 
have made some progress it will be of different character. I shall constantly 
ask you to tell me what you have seen to see how your mind works. I shall never 
make you repeat what you have been told, but constantly ask you what you 

1 J. A. Behnke, editor, Oxford University Press, New York, 1972. 

2 A. H. Wright and A. A. Wright, The American Midland Naturalist, 43(2) : 503-506. 1950. 


have seen yourselves. What I want you to do in order to profit by this is to 
work yourselves." 

Today the spirit is much the same, but today's students have a further 
advantage. They are constantly challenged not just by one preceptor, but by 
a close-knit staff and student group, all anxious to see that every investigation 
moves forward, being shaped by criticism at every step. 

Thus it should not be surprising that of the highlights in this year's report of 
the Director of the Department of Embryology, four research projects involved 
fellows or students — three postdoctoral fellows, Horak, Sachs, and Wellauer, 
and one undergraduate, Camilla Velez. Moreover, as already observed, these 
represent only a sampling of the Department's activities. One sees the influence 
of fellows and students at every turn. 

One final striking similarity to Agassiz's philosophy should be noted. Agassiz 
insisted that his students be his equals in their access not only to problems but 
to equipment as welL Speaking of aquaria and related items, he wrote "I have 
ordered one for every person admitted to the school, so that each of you will 
have means to make these investigations. I have never had in my own labora- 
tory better opportunities for work than I place at your disposal." So it is 
today: outstanding studies are supported, of Staff Member and student alike. 

Though not situated on university campuses, the Geophysical Laboratory and 
the Department of Terrestrial Magnetism cooperate with many universities both 
here and abroad. 

Over the years students from more than a dozen different American universities 
have conducted their doctoral thesis work at the Geophysical Laboratory. Post- 
doctoral Fellows have come from most of the major universities of this nation 
as well as from 16 foreign lands. During the report year, 3 predoctoral and 12 
postdoctoral Fellows were in residence. In the current climate of restricted federal 
support for fellowships, openings throughout the Institution and, of course, at the 
Geophysical Laboratory, are eagerly sought. 

Yoder says that the success of the predoctoral and postdoctoral instructional 
program hangs primarily on the selection of worthy candidates. All of the 
numerous applicants submitting research proposals are carefully reviewed by 
a Selection Committee composed of Staff Members and resident Fellows. There 
are no known predictive measures of motivation, innovation, aggressiveness, 
patience, determination, and all the other personality factors which contribute to 
the success of a researcher in the laboratory environment. Yet, the Selection 
Committee attempts to advise the Director of those individuals who are most 
likely to grow and achieve their goals. How does one compare the overworked 
lecturer burning with desire to return to research, the bright-eyed graduate 
student bubbling with new ideas, and the established professor craving critical 
discussions with his peers. Each wants to achieve that feeling of satisfaction of 
contributing new knowledge to the world, and each must do it in his own way. 

The selection process, although based mainly on intuition, must single out the 
imaginative researcher from the limited facts available. By means of the re- 
search proposal and aided by a biography, bibliography, curriculum vitae, personal 
interviews with the staff, and three letters of recommendation from persons of 
the applicant's choice, the growth potential and capabilities of the applicant are 
judged. In spite of the large amount of time required for the selection process, 
each applicant is given a fair evaluation by the Selection Committee. Their 
selections are judged further and in the light of the scientific goals and economic 
factors within the Institution by the Director and finally by the President. The 


Geophysical Laboratory has enjoyed an unusual measure of success in choosing 
candidates who have grown academically, have been productive, and have become 
leaders in their field. 

At the Geophysical Laboratory as elsewhere in the Institution predoctoral 
Fellows enjoy the same privileges as postdoctoral Fellows and Staff Members. 
There is no ranking among the scientists except that which accrues from the 
accumulation of successful endeavors. One predoctoral Fellow expressed con- 
siderable enjoyment in not being treated as a student. Another found the 
environment of a research laboratory more conducive to rapid advancement than 
that obtained from the occasional personal contacts with university professors. 
Still another predoctoral Fellow had the unusual experience of setting up his 
laboratory from scratch — an intensive, and somewhat disconcerting, educational 
process. The growth observed in predoctoral Fellows is unusually large and most 
gratifying to the other Fellows and Staff who have contributed to their maturation. 

The Department of Terrestrial Magnetism has continued its instructional and 
research efforts in accord with the decades old pattern of close student-Staff 
Member relationships. This year 8 predoctoral and 12 postdoctoral Fellows re- 
ceived training at the Department. These Fellows had a variety of backgrounds 
and experiences. Three Ph.D.'s whose graduate training was in high energy 
nuclear physics or theoretical physics have been broadening their horizons in 
a way that will make them especially versatile. Two young men have joined 
the Biophysics Section where they are becoming acquainted with and are con- 
ducting research in molecular biology. One has already published a significant 
report in the Journal of Molecular Biology. The young theoretical physicist has 
made effective progress and has submitted for publication a new analytical 
approach in the study of strain release in a prestressed medium. With the paucity 
of academic positions presently available in physics and the massive turning 
away of students from the study of physics, it is highly desirable that young 
men such as these be encouraged. Years of experience at the Department of 
Terrestrial Magnetism have proved that a solid training in physics provides an 
excellent background for the individual to branch out into other areas and to 
bring fresh and novel ideas to other subjects. Recent endeavors strongly support 
this contention. 

The Department has also continued to train postdoctoral Fellows in areas 
where the Fellow brought to the Department considerable related background 
and expertise. These Fellows are studying in close cooperation with Staff Members 
in each of the three major areas of the Department's activities: Astrophysics 
(including nuclear and atomic physics), Biophysics, and Geophysics (includ- 
ing geochemistry) where they are mastering new techniques and unique instru- 
ments, and applying them to problems of mutual interest. 

The Department has undertaken to train and advise a considerable number 
of predoctoral students both at the Department and off campus. Three (in bio- 
physics, nuclear physics, and seismology) study in the DTM laboratories and 
under staff guidance will be preparing theses to be submitted to Purdue, Howard, 
and Hokkaido Universities, respectively. Three South American students (on 
leave from the University of Chile, the Servicio Geologico y Mineria, Peru, and 
the Instituto Geofisico del Peru) are being supported for formal graduate course 
work at the University of Alberta, McGill University, and the University of 
Edinburgh. Close contact is maintained with these students and cooperative 


studies in geomagnetics, geochemistry, and seismology are planned or are under 
way. One of the Department's Staff Members is supervising in part the research 
and thesis preparation of a predoctoral student in astronomy at Harvard. While 
interim Director of the joint CIW-IAR Observatory in Villa Elisa, Argentina, 
Turner taught a formal course in radio astronomy and also supervised the studies 
of ten students who used the radio telescope facility and are degree candidates 
at the University of La Plata and the University of Buenos Aires. Additionally, 
four students from the Department of Physics and Astronomy of the University 
of Maryland, in an ongoing cooperative educational program, were taught the 
essentials of operating the 60-foot radio telescope at Derwood, Maryland, and 
were given the opportunity to carry on a modest research project. Under Aldrich's 
sponsorship and guidance, three students (University of Buenos Aires, Cuyo 
University at San Juan, Argentina, and the University of Chile in Santiago) were 
introduced to field work, data reduction, and interpretation in terrestrial electrical 
conductivity studies along the Argentine-Chilean border. This effort is one of a 
long series of interactions of the Department with predoctoral and postdoctoral 
students of Latin America. Seeds planted earlier are now evident in a flourishing 
and first-class geophysical capability of the Andean countries. 


The President's report must of necessity represent only a sampling of the 
activities of the Institution which are described more fully in the reports of the 
various Directors. Even there, however, the accounting is somewhat uneven. 
Were the total activities and contributions to be set forth fully, as much as 3000 
printed pages might be required instead of the 700 pages currently employed. 
In reporting their activities, the various Departments have differing audiences 
and differing objectives. 

The Geophysical Laboratory uses the Annual Report for two principal pur- 
poses — an accounting to the Trustees of the Institution and an accounting to 
scientific colleagues throughout the world. Thus some 2650 copies of the Labora- 
tory's section of the report are sent to colleagues in a total of 68 countries. The 
recipients tell us that they especially value this collection of the work of the 

In contrast, some of the other Departments report more briefly highlights of 
their activities and depend heavily on communication to colleagues by means of 
the scientific literature. 

Another way in which the report is uneven is in describing some of the 
activities that are essential to effective research and to the health of the scientific 
and educational enterprise but do not lead directly to scientific publications. 

For example, our facilities, such as those of the Hale Observatories, are used 
by many visiting students and investigators. Ministering to such guests and 
maintaining facilities for them places a substantial burden on staff. 

Educational activities outside the Institution's facilities are also only briefly 
treated. Our staff annually gives numerous lectures in many universities and 
serves as thesis advisers and on Visiting Committees. 

Another facet of our activities that is accorded minor attention is public 
service. The staff contributes far more than its proportionate share to such efforts. 
These include service on various government advisory groups and responsibilities 
as officers of scientific societies. 


LOSSES . . . 

This year the Institution lost the distinguished astronomer Ira S. Bowen, who 
died February 6, 1973. Dr. Bowen, Director of the Hale Observatories from 1946 
to 1964, was a pioneer in the 1920s in the application of quantum theory to 
spectroscopy, and he applied this knowledge to an insightful interpretation of 
stellar spectra. Some of his greatest scientific contributions were in instrument 
and telescope design. Among many astronomers he is regarded most highly for 
his achievement in placing the 200-inch Palomar Mountain telescope into splendid 
working order. When delivered to the mountain from the optical shop in 
Pasadena the 200-inch mirror was not in finished form. In order to bring it to 
final figure, a temporary polishing machine was set up on the floor of the dome. 
Then the shape of the mirror was gradually completed. This entailed first taking 
star photographs through a Hartmann screen placed over the mirror, careful 
measurement of the resulting plates, removal of the mirror from the telescope to 
the polishing machine, some judicious polishing, replacing the mirror in the 
telescope, and then repeating the whole cycle until no further work was required. 
This tedious and demanding job was completed in about a year. The final result 
was a telescope of excellent optical quality. 

A native of Seneca Falls, New York, Dr. Bowen received his A.B. degree in 
1919 from Oberlin College. Shortly thereafter he joined the faculty of the newly 
reorganized California Institute of Technology, where he did some of the work 
for which he will be best remembered. 

When the Carnegie Institution and California Institute of Technology began 
joint operation in 1948, Dr. Bowen became the first Director of the Mount Wilson 
and Palomar (now Hale) Observatories. After his retirement in 1964, he con- 
tinued to apply himself to problems of optical design. He was responsible for the 
optical design of the recently completed Palomar 60-inch telescope, the wide- 
field Las Campanas 40-inch telescope, and the 101 -inch Irenee du Pont telescope 
for the observatory under construction at Las Campanas in Chile. 

C. Stacy French retired on June 30, 1973, after twenty-six years as Director of 
the Department of Plant Biology. Under his direction the Department has added 
to our knowledge of the ways plants adapt to different environments and it has 
increased our understanding of that complicated and vitally important chemical 
process called photosynthesis. 

Dr. French's own work has centered around the role of plant pigments in 
photosynthesis. He has measured thousands of fluorescence, absorption, and 
action spectra of chlorophyll and other pigments. From one series of experiments 
on fluorescence spectra he was able to establish that energy is transferred from 
accessory pigments to chlorophyll during photosynthesis. In the course of his 
investigations Dr. French invented several useful laboratory devices, such as the 
derivative spectrophotometer for studying chlorophyll in vivo. With this machine, 
French and his associates confirmed the existence of various forms of chlorophyll 
in the living plant and discovered a new form of chlorophyll that absorbs light 
at 695 millimicrons. 

Another of his inventions automatically plots the rate of photosynthesis 
of algae illuminated at different wavelengths, making it possible to obtain action 
spectra and enhancement spectra for photosynthetic reactions and to study their 
interactions. The French pressure cell, which he designed to rupture plant cells 


without the damaging effects caused by detergents or sonication, is also widely- 
used in the study of yeasts and bacteria and for the preparation of vaccines. 

Dr. French has also made a distinct contribution to botanical research in his 
influence on the development of younger scientists. About one fourth of the 
people working in photosynthesis today have studied at the Department of Plant 
Biology, most of them as Carnegie Fellows. 

It is with the deepest sorrow that I report the death on March 8, 1973, of 
Edward A. Ackerman, Executive Officer. Dr. Ackerman served the Carnegie 
Institution for fifteen years. His wise counsel in helping to shape its policies 
was invaluable. We have lost a distinguished scholar, and a warm and loyal 

Dr. Ackerman came to the Carnegie Institution in 1958 as Assistant Executive 
Officer and became Executive Officer in 1960. The perspective and good judgment 
with which he carried out his responsibilities were derived in part from his 
broad background in academia, science administration, and consulting work. 

Dr. Ackerman was especially interested in environmental systems, particularly 
population and water resources. He authored or co-authored eight books and 
more than sixty articles on these subjects, and was consultant to the Departments 
of Interior and State, the House Foreign Affairs Committee, the Senate Committee 
on Natural Resources, the President's Panel on Water Resources Research, the 
National Water Commission, the United States Geological Survey, and the United 
Nations Economic Commission for Europe. 

Dr. Ackerman will be remembered for his patience, unfailing good humor, and 
extraordinary human sensitivity. He embodied the spirit and ideals of the 
Institution, and he will be missed by everyone who had the privilege of knowing 

I must also report the death on April 18, 1973, of Dr. Alfred H. Joy, who was 
associated with the Institution for fifty-eight years. Dr. Joy had been a Staff 
Member of the Mount Wilson Observatory from 1915 to 1948. He was especially 
interested in the chemical spectra of stars, and he participated in the spectral clas- 
sification and radial velocity surveys that provided the data upon which much of 
the later progress in astrophysics and galactic structure was built. In 1945 he 
pointed out a new class of recently formed variable stars designated "T Tauri"; 
it is thought that these objects may be the key to the understanding of how stars 
are formed. 

In addition to his scientific work, Dr. Joy assisted in the administration 
of the Observatory by serving as Secretary from 1920 until his retirement in 
1948. He carried on an active program of postretirement research, and at the 
age of 90 still walked two miles a day to study stellar spectra. 

Harlow Shapley, who died on October 20, 1972, changed the way all of us 
picture the universe. Shapley gave us a truer, more useful model of what lies 
beyond our world, and he gave modern astronomy its most important tool for 
determining celestial distances, the period-luminosity relationship of Cepheid 
variable stars. 

It was early in his career, while he was at the Mount Wilson Observatory, 
that Shapley pursued the study of Cepheid variables that led to his most im- 
portant discoveries. By measuring changes in the temperature and spectra of 


the Cepheid variable stars, he was able to calculate their rates of pulsation and 
from those learn their intrinsic brightness. Their observed brightness could then 
be measured as an index of distance. Using this relationship between the period 
and the luminosity of the Cepheids as a kind of measuring stick for the rest of the 
galaxy, Shapley plotted the location of swarms of stars called globular clusters. 
From the distribution of the globular clusters he calculated that our solar system 
is nowhere near the center of its galaxy, as had been supposed, and that the 
galaxy is larger than astronomers had believed. 

. . . AND GAINS 

One of the year's notable gains was the addition of Lewis McAdory Branscomb 
to the Board of Trustees. Dr. Branscomb's career has combined outstanding 
accomplishments in both science and administration. 

He was graduated summa cum laude from Duke University in 1945 and received 
the M.S. and Ph.D. degrees in physics from Harvard University in 1947 and 
1949. He has taught physics at Harvard, University College (London) , the 
University of Maryland, and the University of Colorado. 

Branscomb joined the National Bureau of Standards in 1951. He served in 
the atomic physics division from 1954 to 1962. For the next seven years he was 
chairman of the Joint Institute for Laboratory Astrophysics in Boulder, Colorado, 
a cooperative venture with the University of Colorado, until President Nixon 
appointed him Director in 1969. As Director, Branscomb led the Bureau to a 
major role in stimulating industrial technology. 

In May 1972 he was appointed chief scientist of the International Business 
Machines Corporation; the following month he was elected vice-president and is 
responsible for planning the research program. 

Dr. Branscomb's career has included service on a number of boards and com- 
mittees, including the President's Science Advisory Committee, the President's 
Committee for the Medal of Science, the Council of the National Academy of 
Sciences, the Board of Directors of the American Association for the Advancement 
of Science, and the Board of Trustees of the Rand Corporation. 

Dr. Branscomb has received the Rockefeller Public Service Award, the Gold 
Medal for Exceptional Service from the U.S. Department of Commerce, and the 
Procter Prize from the Scientific Research Society. He is a Fellow of the American 
Academy of Arts and Sciences, the American Philosophical Society, and the 
American Physical Society. He holds honorary doctor of science degrees from 
Duke, Western Michigan, and Rochester Universities. 

To succeed Dr. French as Director of the Department of Plant Biology, the 
Institution is fortunate in having one of the leaders in modern developmental 
biology, Winslow R. Briggs. Dr. Briggs plans to continue his work in plant 
growth and development and to establish this as one of the areas of research 
at the Department. At the same time, he has considerable interest in physiological 
ecology and the biochemistry of photosynthesis, and he intends to maintain and 
strengthen the Department's programs in these fields. 

Dr. Briggs comes to the Institution from Harvard, where he has been professor 
of botany since 1967. From 1955 to 1967 he was instructor and later associate 
and full professor of biology at Stanford University. While at Stanford Dr. Briggs 
visited our Department of Plant Biology and participated in several studies there. 


Dr. Briggs is well known for his research on the pigment phytochrome, the 
photoreceptor that conveys the required information to the plant about light in 
the environment. Among many other effects phytochrome mediates seed germina- 
tion, stem elongation, and the orientation of leaves to best capture sunlight for 
photosynthesis. Phytochrome also ensures that flowering, seed formation, and 
dormancy in many plants occur at the appropriate times of the year. 

Briggs' work is concentrated on the elucidation of the molecular mechanisms 
that allow phytochrome to regulate plant metabolism. He has greatly advanced 
the techniques for the isolation of this chromoprotein in a highly purified state. 
His studies on the protein chemistry, structure, and the photochemistry of phyto- 
chrome have aided in the understanding of the mechanism of action of this 
substance. Currently he is investigating the possibility that phytochrome trans- 
formations may give rise to alterations of membrane properties, which in turn 
can lead to profound biochemical changes. 

A graduate of Harvard University, Briggs received his bachelor's degree in 
1951 and his doctorate in 1956. He is a member of the American Association for 
the Advancement of Science, the Botanical Society of America, and the American 
Society of Plant Physiologists. Briggs is the author of more than sixty published 
articles on such topics as growth hormone physiology, plant tropisms, and plant 
pigment function, and he is editor of the Annual Review of Plant Physiology. 

I am pleased to announce the following honors to Staff Members. 

Donald D. Brown of the Department of Embryology was elected to the 
National Academy of Sciences. At the annual meeting of the Academy, Dr. 
Brown received the U.S. Steel Foundation Award in Molecular Biology. 

James D. Ebert, Director of the Department of Embryology, received the 
honorary degree of Doctor of Science from Yale University. Dr. Ebert was 
also awarded the President's Medal of the American Institute of Biological 
Sciences and was named as one of the first Eminent Scientists in an exchange 
program sponsored by the Japan Society for the Promotion of Science. 

Vera C. Rubin of the Department of Terrestrial Magnetism was elected to the 
Board of Directors of the Association of Universities for Research in Astronomy, 
Inc., as a Director-at-Large, to serve until 1976. Dr. Rubin was also elected a 
member of the Nominating Committee of the American Astronomical Society for 
a three-year term, and was appointed a member of the Space Astronomy Com- 
mittee of the Space Science Board of the National Academy of Sciences from 
January 1973 to March 1975. 

Hatten S. Yoder, Jr., Director of the Geophysical Laboratory, was given the 
Abraham Gottlob Werner Medal of the Deutsche Mineralogische Gesellschaft, 
the highest scientific award of the society. The American Academy of Achieve- 
ment selected him as one of the recipients of the Golden Plate Award for his 
pioneering work in the study of rock and mineral synthesis in the earth's crust 
and upper mantle. 

C. Stacy French, Director of the Department of Plant Biology, received the 
Award of Merit from the Botanical Society of America. 

Barbara McClintock, Distinguished Service Member of the Institution, had a 
laboratory building at Cold Spring Harbor Laboratory named for her in recogni- 
tion of her contributions to genetics research. 


Stanley R. Hart of the Department of Terrestrial Magnetism was appointed 
to the Panel on Orientations for Geochemistry of the U.S. National Committee 
for Geochemistry, National Academy of Sciences— National Research Council. 

Ronan O'Rahilly and John Tucker of the Department of Embryology won the 
silver medal at the annual meeting of the American Academy of Ophthalmology 
and Otolaryngology for their exhibit on the embryology of the human larynx. 

Fritz Zwicky, who is engaged in postretirement studies at the Hale Observa- 
tories, was awarded the Gold Medal of the Royal Astronomical Society. 





Baltimore, Maryland 


James D. Ebert 

Staff Members 

Donald D. Brown 
Igor B. Dawid 
Robert L. DeHaan 
Douglas M. Famb rough 
Richard E. Pagano 
Ronald H. Reeder 


Joseph F. Albright 

Dana Carroll 

Tasuku Honjo 

Ivan Horak 

Paul Lizardi 

Terence F. McDonald 

Keiko Ozato 

Aileen K. Ritchie 

Howard G. Sachs 

Kazunori Sugimoto 

Yoshiaki Suzuki 

Peter K. Wellauer 

Robert Williamson 

Guang-Jer Wu 

Masako Yoshikawa-Fukada 

Assistant Investigator 
Peter J. Stambrook 

Students or Predoctoral Fellows 

Sandra L. Biroc 
Peter Devreotes 
Thomas E. Durr 
Scott Gilbert 
H. Criss Hartzell 
Carol Kaushagen 
Caroline R. Krueger 
Dennis E. Leister 
Linda Moyzis 
Calvin E. Plitt 
Ralph H. Stern 
Camilla Velez 

Washington, D.C. 


Hatten S. Yoder, Jr. 

Staff Members 

Peter M. Bell 
Francis R. Boyd, Jr. 
Felix Chayes 
Gordon L. Davis 
David H. Eggler 1 
Larry W. Finger 
P. Edgar Hare 
Thomas C. Hoering 
T. Neil Irvine 2 
Thomas E. Krogh 
Ikuo Kushiro 
Ho-Kwang Mao 3 
Robert N. Thompson 
David Virgo 
Alain Weisbrod 

Emeritus Research Associate 
Emanuel G. Zies 

Postdoctoral Fellows 

Jagannadham Akella 
Donald M. Burt 
John S. Dickey, Jr. 
John D. Frantz 
Bastiaan J. Hensen 
Floyd N. Hodges 
Kenneth King, Jr. 
James M. Mattinson 
Karlis Muehlenbachs 
Yoshikazu Ohashi 
Michael G. Raymond 
Martin G. Seitz 

Predoctoral Fellows 

John I. Hedges 
D. James Misener 
Bjorn 0. Mysen 

1 Temporary Staff Member from September 1, 

2 Temporary Staff Member from August 16, 

3 Temporary Staff Member from September 1, 
1972; Postdoctoral Fellow to August 31, 1972. 



Pasadena, California 


Horace W. Babcock 

Associate Director 
J. Beverley Oke 

Distinguished Service Member 
Ira S. Bowen 1 

Staff Members 

Halton C. Arp 
Edwin W. Dennison 
Jesse L. Greenstein 
James E. Gunn 
Robert F. Howard 
Jerome. Kristian 
Robert B. Leighton 
Guido Munch 
Gerry Neugebauer 
George W. Preston 
Bruce H. Rule 
Allan R. Sandage 
Wallace L. W. Sargent 
Maarten Schmidt 
Leonard T. Searle 
Arthur H. Vaughan, Jr. 
Olin C. Wilson 
Harold Zirin 

Staff Associates 

Eric E. Becklin 
Robert J. Brucato 
Michael W. Werner 
James A. Westphal 


Ermanno F. Borra 
Michael M. Dworetsky 
Michael H. Hart 
Pieter van der Kruit 
Andrew Michalitsanos 
Ronald Moore 
Jay M. Pasachoff 
Stephen W. Prata 
William J. Quirk 
Katsuo Tanaka 
Richard E. White 
Robert R. Zappala 

1 Died February 6, 1973. 

Carnegie-Chilean Fellows 

Eduardo Hardy 
Maria Teresa Ruiz 

Student Observers 

Marc Aaronson 
William G. Bagnuolo 
Ian Gatley 
Richard Green 
Paul M. Harvey 
Paul Hickson 
James A. Howell 
John P. Huchra 
Robert P. Kirshner 
John Kormendy 
Elliot C. Lepler 
Steven J. Loer 
Jorge Melnick 
Augustus Oemler, Jr. 
Valdar Oinas 
Glenn S. Orton 
Paul L. Schechter 
Stephen A. Shectman 
Barry E. Turnrose 
Glenn J. Veeder 
Bruce Waddington 
William E. Westbrook 
Richard A. Wickes 
Steven P. Willner 


Stanford, California 


C. Stacy French 

Staff Members 

Joseph A. Berry 
Olle Bjorkman 
Jeanette S. Brown 
David C. Fork 
Malcolm A. Nobs 

Staff Member Emeritus 
William M. Hiesey 

Guest Investigators 

Georgi Detchev 
Ralphreed A. Gasanov 
Ulrich Schreiber 
Atusi Takamiya 




George Bowes 
William G. Hagar 
Tetsuo Hiyama 
Norio Murata 
Helga I. Ninnemann 
John H. Troughton 

Stanford Students and 
Postdoctoral Fellows 

Sherry DeRemer 
Louis Pitelka 

Washington, D.C. 


Ellis T. Bolton 

Associate Director 
L. Thomas Aldrich 

Section Chairmen 

L. Thomas Aldrich, Geophysics 
Louis Brown, Astrophysics x 
Dean B. Cowie, Biophysics 
W. Kent Ford, Jr., Astrophysics 2 

Distinguished Service Member 
Merle A. Tuve 

1 From January 1, 1973. 

2 To December 31, 1972. 

Staff Members 

George E. Assousa 
Roy J. Britten 
Louis Brown 
Dean B. Cowie 
W. Kent Ford, Jr. 
Stanley R. Hart 
Albrecht W. Hofmann 
Bill H. Hoyer 
David E. James 
Alan T. Linde 
Nancy R. Rice 
Richard B. Roberts 
Vera C. Rubin 
I. Selwyn Sacks 
Diglio Simoni 
Norbert Thonnard 
Kenneth C. Turner 


Jesus A. Berrocal 
Tom I. Bonner 
Michael B. Davis 
Borwin Grauert 
Cidambi K. Kumar 
Allan T. Leffler II 
Hiromu Okada 
Urs C. Rohrer 
Nobumichi Shimizu 
J. Arthur Snoke 
Neil A. Straus 
Thangasamy Velusamy 
Donald G. Wallace 
John W. Warner III 

Trainee Fellows 

Arturo Cuyubamba 
Douglas M. Haefele 
Godwin C. Igiri 
Jack M. Klitzman 
Ricardo Quiroga 
Enrique Triep 

Reports of Departments 
and Special Studies 

Department of Embryology 

Hale Observatories 

Department of Terrestrial Magnetism 

Department of Plant Biology 

Geophysical Laboratory 

Department of Embryology 

Baltimore, Maryland 

James D. Ebert 

Carnegie Institution of Washington Year Book 72, 1972—1973 


Introduction 7 

The Isolation of Genes and the Study of Their Structure, Evolution, and Control . 10 

The origin of ribosomal DNA homogeneity 11 

Transcription of 5S DNA in vitro 14 

Analysis of 5S DNA heterogeneity 15 

The use of antibiotics in buoyant density gradients 15 

Studies on silk fibroin mRNA and its genes 18 

A polyadenylic acid sequence in silk fibroin mRNA 19 

The Regulation of Fibroin Genes 20 

Quantitative measurements of fibroin mRNA synthesis 20 

Mutants for fibroin production 22 

Fibroin mRNA from Bombyx mandarina 23 

The Molecular Weight of Mouse Globin Messenger RNA 24 

Control of Gene Transcription in Eukaryotes 25 

Transcription of rRNA and 5S RNA genes in amphibian chromatin by homologous 

RNA polymerase 26 

Transcription of chromatin by bacterial RNA polymerase 28 

Repression of Xenopus mulleri rDNA in X. laevis X X. mulleri hybrid frogs . 31 

The Time of Ribosomal Gene Replication during the S Phase of Synchronized 

Chinese Hamster Cells 35 

Secondary Structure Maps of RNA Molecules Derived by Electron Microscopy and 

Their Use for the Analysis of the Processing of Hela Ribosomal RNA ... 38 

Biogenesis of Mitochondria 39 

Recombination of mitochondrial DNA in somatic hybrid cells 40 

Protein composition of cytoplasmic and mitochondrial ribosomes of Xenopus laevis 42 

Mitochondrial RNA of Drosophila 43 

Measurement of mitochondrial RNA and DNA-RNA hybrids in the electron 

microscope 45 

Transcription of mitochondrial DNA by mitochondrial RNA polymerase ... 46 

Structure and Function in Cell Membranes 47 

Interactions between lipid bilayers and cell surface membranes 47 

Viscoelastic properties and surface tension of membranes 49 

Membrane dynamics 49 

Studies on Muscle Fiber Membranes 50 

Acetylcholine receptors in skeletal muscle fiber membranes 50 

Interaction of ACh receptors with a-bungarotoxin 50 

Turnover of ACh receptors in myotube plasma membranes 51 

Incorporation of new ACh receptors into myotube membranes 55 

The neuromuscular junctions in myasthenia gravis: decreased acetylcholine 

receptors 56 

Electrophysiological properties of the membrane and cholinergic receptor of 

developing myotubes 60 

Membrane potential in developing myotubes 61 

Ionic selectivity of the cholinergic receptor during development 62 

Some effects of Triton X-100 on acetylcholine receptors 63 

The Effect of Potassium on the Cell Membrane Potential and the Passage of Cells 

through the Cell Cycle: A Block in G x 64 

A Cellular Response to Dibutyryl Cyclic AMP: A Cell Cycle Block in G 2 . . . 69 

Functional Differentiation of the Embryonic Heart 73 

Modulation of sodium current channels in embryonic heart cells by tetrodotoxin . 73 

Myofibrillar organization during the first beats of the chick heart 80 

Control of synchronized pulsation rate in cell groups 83 

Cytochalasin B and embryonic heart muscle: Contractility, excitability, and 

ultrastructure 86 

The Collection of Human Embryos 87 

Developmental Stages in Human Embryos 87 

Development of the conducting system of the heart 88 

Development of the nervous system 88 

Development of the larynx 89 

Development of bones and joints 89 

Staff Activities 89 

Bibliography 91 

Personnel 93 


The Department continued its work recognition by their peers. To the tech- 
during the year without serious distrac- nician, who plays an integral part in re- 
tions or interruptions. It has always been search, there is the personal satisfaction 
a friendly and pleasant environment in of a job well done, but other forms of 
which to work; in fact, it is one that is recognition are infrequent. Friends of the 
not infrequently inspiring as the day-to- Department know how much all of us, 
day regimen of research is punctuated by established Staff Members and incoming 
new findings and ideas. Of course many Fellows alike, have relied on trusted col- 
— if not most — of them are short-lived, leagues like Thomas Malooly, Margaret 
but enough stand the test of criticism by Proctor and John Wiser. However, very 
our peers so that our younger Staff Mem- few have appreciated the remarkable ex- 
bers and Fellows pursue their ideas with tent to which many members of the tech- 
the reassuring feeling that if they, too, nical staff have developed over the years, 
persevere, the pieces of their puzzles will It is always difficult to single out exam- 
fall into place. pies for special mention, but biologists 

If any one factor can be singled out to like to use models to illustrate their argu- 
account for the confident attitude and ments, so perhaps I can make my point 
reflective mood that pervades the Depart- by saying that Eddie Jordan, Delores 
ment, it is the continuity and stability of Somerville and Ernestine Flemmings, 
our financial support. It is true that the among others, have shown remarkable 
departmental budget has barely kept personal growth during their association 
pace with inflation over the past several with the Department, 
years, yet we have been able to proceed I began by saying that the year was 
in the confidence not only that we have a one of uninterrupted work in the Depart - 
stable "floor" for our research, but also ment. Significant progress is reported in 
that the Institution has sufficient re- both of our principal streams of activity, 
serves, sufficient flexibility, and, above (1) the isolation of nuclear and mito- 
all, a sufficiently deep understanding of chondrial genes and the study of their 
the scientific process to enable changes in structure, evolution, and control and (2) 
course when they are demanded, either the development of membranes. A glance 
by unexpected new findings or by the at the Bibliography reveals that the year 
"drying up" of what had appeared to be was marked by a tide of published find- 
promising leads. The fact that there have ings, but as I remarked on an earlier oc- 
been no abrupt swings, no radical budget- casion, to say this gives a false impression 
ary ups and downs, has enabled us to of the state of the Department, for the 
continue growing in stature without vitality of the Laboratory is reflected less 
growing in size and administrative com- in published and already "cold" data 
plexity. than in the flow of new ideas and in the 

I would mention a second key ingredi- constant search for new approaches to 
ent, one that is too often overlooked. I long-resistant problems. In offering the 
refer to the rich human resources in the following samples of our activities, I 
Department's supporting staff. Investi- shall stress the novel findings and the ex- 
gators derive deep personal satisfaction ploratory pathways, cautioning the 
from a successful experiment or the distil- reader whose interest is whetted to look 
lation and synthesis of a new concept. In carefully at the detailed accounts that 
addition, however, they gain the added follow, 
satisfaction of national and international No one approach to the structure of 



RNA has generated more interest than 
Peter Wellauer and Igor Dawid's analy- 
sis of secondary structure maps of RNA 
molecules by electron microscopy. Their 
electron micrographs of 28S ribosomal 
RNA show a reproducible secondary 
structure which allows the mapping of 
these molecules — different regions of each 
molecule and their polarities becoming 
recognizable. The technique has been ap- 
plied to a study of the "processing" of the 
ribosomal RNAs. It has been known for 
some time that the 28S and 18S rRNAs 
originated by the "processing" (or con- 
trolled fragmentation) of a larger 45S 
precursor. Now the structures of the 45S, 
28S and 18S molecules have been com- 
pared in the electron microscope, and 
their relationships clarified. The 28S re- 
gion is located at the 5' end of the 45S 
precursor. It is followed by a "spacer" 
region, then an 18S region and another 
"spacer" at the 3' end. In "processing" the 
"spacers" are removed. Wellauer, an In- 
ternational Fellow of the U.S. Public 
Health Service, will continue these highly 
promising studies for a second year. 

Dawid has also been deeply immersed 
in another investigation of unusual inter- 
est and potential importance. I say "po- 
tential" because, as he observes, "explicit 
proof" of his conclusions is still forth- 
coming. In collaboration with Ivan 
Horak, a second-year Carnegie Fellow 
recently returned to Wiirzburg, and Hay- 
den Coon of the National Cancer Insti- 
tute, Dawid has been studying the mito- 
chondrial DNAs (mtDNAs) of somatic 
hybrid cells (human-mouse and human- 
rat) . It was found, first, that mtDNAs 
of both species can replicate in the same 
cell for extended periods. This much can 
be said unequivocally. However, a more 
crucial question is this: Do they simply 
"coexist" in the same cytoplasm or do 
they interact in some way? Even more 
difficult to answer is the further question, 
do they recombine, forming a hybrid 
molecule? Molecular hybridization stud- 
ies, detailed later in the Report, make it 
very likely that mouse mtDNA segments 

become linked to human mtDNA during 
growth in the hybrid cells. A number of 
hybrid cell strains have been examined. 
Thus far, five strains have both parental 
mtDNAs without recombination but 
there are 13 strains in which recombina- 
tion has apparently taken place. 

Another experimental system in which 
molecular hybridization techniques have 
revealed significant gene recombination is 
found in D. D. Brown and E. Jordan's 
study of the origin of ribosomal DNA 
homogeneity. The genes for 28S and 18S 
rRNA in two species of Xenopus, laevis 
and mulleri, are quite similar, if not 
identical, indicating that these sequences 
have not evolved significantly. However, 
the associated spacer regions differ by as 
much as 10% of their 7000 base pairs. 
How do these sequences evolve — some 
genes being maintained in a homogeneous 
state, others changing rapidly? It has 
been proposed that unequal crossing over 
can maintain homogeneity in a family of 
genes. Brown and Jordan have been ex- 
amining a number of possibilities. They 
have reared hybrid frogs derived from 
matings of X. laevis and X. mulleri and 
backcrossed one such animal with both 
laevis and mulleri. The progeny are now 
sexually mature and the extent to which 
the ribosomal genes have been recom- 
bined can be determined. If there is ex- 
tensive recombination, the backcrossed 
animals should have laevis and mulleri 
genes adjacent to each other. When 
meiosis occurs, will there be a "correc- 
tion"? Brown and Jordan pose the sev- 
eral possibilities clearly in their account. 
It will suffice in this summary to say that 
twenty individuals from each group of 
backcrossed frogs have been analyzed. 
All animals tested in the backcross with a 
laevis male contain some mulleri rDNA. 
Twenty progeny of the mother's back- 
cross with a mulleri male were also ex- 
amined. Although none had as much 
laevis rDNA as the mother, 18 out of 20 
had some laevis rDNA. Brown and Jor- 
dan conclude that this result, while not 
yet conclusive, is strong evidence for ex- 



tensive recombination in the oocyte be- 
tween the homologous chromosomes con- 
taining rDNA. 

Progress in the Department's second, or 
membrane, "track" has been equally stim- 
ulating. Of the several impressive continu- 
ing studies of Douglas Fambrough and his 
colleagues, I have selected one for special 
attention. Myasthenia gravis is a neuro- 
muscular disorder characterized by mus- 
cular weakness and fatigability.' It has 
been thought to involve the neuromus- 
cular junction. In collaboration with 
Daniel Drachman and S. Satyamurti of 
the Johns Hopkins University School of 
Medicine, Fambrough has conducted ex- 
periments designed to determine whether 
the disease involves changes in the num- 
ber and distribution of acetylcholine re- 
ceptors in muscle fibers from dystrophic 
patients. The answer is unequivocal: The 
neuromuscular junctions of patients with 
typical myasthenia gravis contain only 
11 to 32% of the control number of ACh 
receptors. This finding represents only a 
beginning, of course, but it permits more 
critical questions to be asked. Is the 
number of receptor molecules reduced? 
Or are many receptor molecules blocked 
or altered, resulting in a decrease in func- 
tional receptors? The writer cannot re- 
frain from showing his"immunobiological 
bias" to raise a further question: The 
older literature speculated (with some 
reason) that myasthenia gravis might be 
an autoimmune disease, somehow related 
to aberrant function of the thymus. Is 
it possible that the two ideas might be 
brought together? Might the blocking of 
ACh receptor sites be accomplished by an 
autoimmune reaction? 

The Department's capacity to deal 
with the structure and function of cell 
membranes was enhanced by the arrival 
during the year of Staff Member Richard 
E. Pagano, who has undertaken the de- 
velopment of physical and chemical tech- 
niques for examining the surface proper- 
ties of natural membranes. His initial 
exploratory findings are summarized in 
the body of the Report. 

The cell membrane has also been a 
focal point in Peter Stambrook's research. 
With Howard Sachs and Ebert, he has 
continued the analysis of the effect of 
potassium on the cell membrane potential 
and the passage of cells through the cell 
cycle, demonstrating that Chinese ham- 
ster cells are blocked in mitosis and in Gi. 

With a Johns Hopkins undergraduate, 
Camilla Velez, he has explored other 
techniques of altering the cell cycle by 
changes at the membrane, in a further 
effort to probe the means whereby mem- 
brane changes impinge upon the genetic 
apparatus. He has succeeded in showing 
that the application of dibutyryl cyclic 
AMP [(but) 2 cAMP] blocks cells pri- 
marily in G 2 and to a lesser extent in 

Fellows continue to play an important 
role in the life of the Department. In 
Donald Brown's laboratory Dana Car- 
roll, Paul Lizardi, Kazunori Sugimoto, 
and Senior Carnegie Fellow Robert Wil- 
liamson contributed significantly. Car- 
roll, in his first year as a Carnegie Fellow, 
has begun to analyze the heterogeneity of 
5S DNA; Lizardi, completing his two- 
year stay as a Jane Coffin Childs Fellow, 
has synthesized a radioactive reverse 
transcriptase product of silk fibroin 
mRNA ("anti-messenger" DNA) and 
has established its specificity and sensi- 
tivity. It should be useful as a tool in de- 
tecting fibroin messenger nucleotide 
sequences. Lizardi will return to Rocke- 
feller University late in 1973. Second- 
year Carnegie Fellow Sugimoto, who will 
return to Nagoya University in August 
1973, has begun to study the fidelity with 
which 5S DNA is transcribed in vitro. 

Williamson's contribution was an im- 
portant one, not only in his specific 
investigations (A polyadenylic acid se- 
quence in silk fibroin mRNA ; with Well- 
auer, the molecular weight of mouse 
globin messenger RNA; plus a study in 
collaboration with Stambrook to be re- 
ported in a future Report) but also in his 
generous sharing of ideas and techniques 



and his lively exchanges here and 
throughout the Nation. 

Second-year Carnegie Fellow Tasuku 
Honjo, working with Reeder, also focused 
his attention on the thorny question of 
the control of gene transcription in eu- 
karyotes. Substantial progress in defin- 
ing the experimental system is reported. 
On July 1, 1973, Honjo will move to the 
National Institutes of Health to work in 
association with Philip Leder. 

Douglas Fambrough's laboratory was 
enhanced by the addition of Aileen 
Ritchie, a Fellow of the Muscular Dys- 
trophy Association of America ; and Max 
Springer of Zurich completed his second 
year with Robert DeHaan. Springer's 
artistry with the electron microscope, by 
both scanning and transmission tech- 
niques, is exhibited in Figures 50 and 51. 

Among the graduate students, two 
completed their studies. H. C. Hartzell 
has taken up a fellowship appointment in 
the Department of Neurobiology at Har- 
vard Medical School, and Dennis Leister 
will hold a fellowship in the laboratory of 
Hans Kiintzel in the Max-Planck Insti- 
tute for Experimental Medicine in Got- 
tingen. Ralph Stern will transfer to The 
Jackson Laboratory; and Carol Kaus- 
hagen, working with Dawid, has begun to 
look at the mitochondrial RNAs of Dro- 

Losses — and Gains. Robert DeHaan 
joined the Department on July 1, 1956. 
For seventeen years he has been a de- 
voted colleague and effective contributor 
who has established himself as a leading 
student of cardiogenesis — of both mor- 
phogenesis and functional differentiation, 

including the origin and differentiation of 
the pacemaker. In addition, he has con- 
tributed importantly to the local com- 
munity and to the scientific community 
at large. I report his departure on June 
30, 1973, with regret, tinged with pleasure 
— regret at the loss of a valued colleague, 
but pleasure in the fact that in his new 
role as Professor of Anatomy at Emory 
University he will be able to provide 
leadership to a new research group em- 
phasizing the developmental physiology 
of the heart specifically, and the proper- 
ties of developing membranes more gen- 

I am pleased to report that Yoshiaki 
Suzuki, currently an Extramural Fellow 
of Carnegie Institution working in the 
National Institute of Health of Japan 
(for a report of his studies, see pp. 20-24) , 
is expected to rejoin the Department as a 
Staff Member in the fall of 1973. 

In addition, plans are under way to 
bring Masako Yoshikawa-Fukada, also 
an Extramural Fellow, currently working 
in the Institute for Virus Research in 
Kyoto (see Bibliography for her most 
recent contribution), back to the Depart- 
ment early in 1974 to resume her studies 
of cell transformation. 

It is also good to announce that Ernest 
Gardner, Professor of Neurology at the 
University of California School of Medi- 
cine at Davis, has accepted our invitation 
to serve as Associate Curator of the Car- 
negie Collection when it is moved to the 
Davis campus in the summer of 1973. 
Professor Ronan O'Rahilly remains 
Curator of the Collection. 




D. D. Brown, D. Carroll, P. Lizardi, R. Stern, K. Sugimoto, and R. Williamson 

Previous Year Books have outlined our their controls in vitro. Experiments in 

approach to developmental genetics. We the earlier phase of this project have in- 

have isolated genes of known function volved methods for isolating these genes 

with the long-range goal of reconstructing from complex mixtures of DNA and anal- 



yses of their structure and arrangement 
along the DNA molecule. More recently 
we have undertaken studies of the evolu- 
tion of these genes. The repetitive genes 
termed rDNA and 5S DNA have been 
purified from Xenopus laevis and Xeno- 
pus mulleri. Studies in this report de- 
scribe analyses of the arrangement and 
homogeneity of these highly repetitive 
genes. These genes appear to evolve as a 
family and the means by which this "par- 
allel" or "horizontal" evolution takes 
place are under experimental test. 
Whereas the genes for 18S and 28S RNA 
and their spacers appear to be homo- 
geneous within a species, the multiple 
genes for 5S RNA are similar to each 
other but nonetheless heterogeneous even 
within a single animal. What mecha- 
nisms cause multiple DNA sequences to 
evolve together, and when heterogeneity 
exists, how is it arranged in the genome? 
A second area of our research deals 
with silk fibroin messenger RNA and its 
genes. Two kinds of problems are being 
pursued. We are studying the length of 
DNA encoding the structural fibroin 
genes, and methods are being developed 
to isolate this DNA component. Secondly, 
the structure and metabolism of the 
mRNA itself are under investigation. We 
are developing an assay system with 
which we hope to analyze transcriptional 
control of fibroin messenger RNA syn- 
thesis. Studies of polyA content of the 
messenger RNA are also summarized. 

The Origin of Ribosomal DNA 

D. D. Brown and E. Jordan 

The DNA containing the genes for 18S 
and 28S RNA is clustered at a single 
chromosomal locus in Xenopus and con- 
tains about 450 extremely similar if not 
identical repeating nucleotide sequences. 
Half of each repeat includes the genes for 
18S and 28S rRNA and these sequences 
have not evolved detectably between two 
species of Xenopus — mulleri and laevis. 
The spacer regions, however, are very 

different and must differ by more than 
10% of their approximately 7000 base 
pairs. How do hundreds of DNA se- 
quences evolve together? One group of 
explanations, referred to as gene "rectifi- 
cation," have adjacent sequences being 
corrected to conform to one of a limited 
number of these repetitious sequences. 
Several mechanisms for this have been 
proposed — notably the "master-slave" 
model of H. G. Callan and the "expan- 
sion-contraction" model. These kinds of 
models require unusual molecular mecha- 
nisms for which we do not now have evi- 
dence. In contrast to this, frequent un- 
equal crossing over can maintain a family 
of genes homogeneous, as has recently 
been pointed out by George Smith of the 
University of Wisconsin. This could oc- 
cur at meiosis between homologues and/ 
or between sister chromatids. In addition, 
"scrambling" of sequences can occur at 
mitosis of germ cells by sister chromatid 

We have begun a series of experiments 
designed to test these different possibil- 
ities. It is implicit in these studies that 
the mechanism which maintains sequence 
homogeneity must correct the DNA 
faster than the mutation rate will cause it 
to diverge. The gene regions remain 
homogeneous due in part to selective re- 
straints on rRNA divergence. In order to 
observe the correction mechanism experi- 
mentally, the process will have to occur 
at a relatively high frequency — at least 
once per animal per generation. If cor- 
rection is less frequent our experiments 
will not detect it. 

We have raised hybrid frogs derived 
from matings of X. laevis and X. mulleri 
(Fi) and backcrossed one such animal 
with both X. laevis and X. mulleri. The 
animals are now sexually mature. We 
wish to measure the extent to which 
ribosomal and 5S genes have recombined 
in the progeny. If this occurs at high fre- 
quency, these backcrossed animals should 
have genes from one species adjacent to 
those of the other. When such a recom- 
bined locus goes through meiosis, will it 



be "corrected"? If not, can we at least 
assess the extent to which recombination 
occurs within these gene clusters? 

When an Fi animal termed Im — 
the progeny of a laevis female and a 
mulleri male) is backcrossed with a laevis 
male (termed 11) , the predicted segrega- 
tion of progeny rDNA genotypes is ex- 
pected to be 50% Im and 50% 11. The 
extent to which the animals deviate from 
this ratio is a measure of several possible 
factors: selection of certain genotypes in 
the progeny, chromosomal abnormalities, 
or prior recombination within the rDNA 
locus. Selection is always possible, since 
most of the progeny from these matings 
die from abnormalities due to problems 
introduced by the hybrid nature of the 
chromosomes. We must, therefore, study 
the chromosome composition of each 
animal to show that it still contains only 
two homologues each with a single nucle- 
olar organizer. In addition, a careful 
study of the number of ribosomal genes in 
each animal together with the relative 
abundance of X. laevis and X. mulleri 
rDNA assesses the extent to which un- 
equal crossing over has occurred in those 
animals. If crossing over has occurred in 
somatic cells, different tissues of the same 
animal may have different numbers of 
rRNA genes. However, if exchange oc- 
curred in meiosis, individual animals 
would vary in gene number, but all the 
cells of an individual should have the 
same number of genes. 

Another possible influence on gene cor- 
rection that can be tested experimentally 
is the importance of gene amplification. 
Gene correction might occur by amplifi- 
cation of one or a limited number of 
chromosomal copies followed by exten- 
sive recombination of the extra copies 
with the chromosomal copies. This would 
effectively remove any divergence among 
chromosome copies. In its extreme form, 
this possibility is also testable in hybrid 
animals. We know from previous experi- 
ments that Fi females only amplify X. 
laevis rDNA. If there were efficient re- 
combination of all chromosomal rDNA 

copies with amplified rDNA molecules, 
then progeny of the backcrosses would 
never receive any X. mulleri rDNA from 
the female; those copies would have been 
replaced with X. laevis rDNA. 

The DNA of each hybrid animal has 
been analyzed for its content of rDNA 
from both parents. This is carried out by 
molecular hybridization of complemen- 
tary RNA mixtures which have been syn- 
thesized with E 4 coli RNA polymerase 
from purified rDNA of X. laevis and X. 
mulleri. The two cRNAs are synthesized 
with different isotopic precursors ( 3 H and 
32 P-labeled nucleoside triphosphate), 
mixed together, and hybridized with 
somatic DNA from individual animals. 
The ratio of isotopes bound to each DNA 
is a measure of the abundance of the two 
kinds of rDNA spacers that are present. 

One experiment is shown in Fig. 1. 
Twenty individuals from each of the two 
groups of backcrossed frogs have been 
analyzed for the rDNA content of their 
somatic DNA. The results show that the 
progeny have not followed the predicted 
Mendelian segregation. All animals tested 
in the backcross with an X. laevis male 
contain some X. mulleri rDNA whereas 
half of them should have only X. laevis 
rDNA. About half of the animals have 
approximately equal amounts of the two 
kinds of rDNA like the mother (Im) . The 
other half have about two-thirds X. laevis 
rDNA. The question then is raised 
whether the mother passed on any X. 
laevis rDNA to her progeny. This is 
answered by her other backcross with an 
X. mulleri male. Twenty of these progeny 
were examined. None of the twenty prog- 
eny had as much X. laevis rDNA as the 
mother. As can be seen in Fig. lb, ten of 
them have 10 to 30% X. laevis rDNA. 
The remainder of the animals which ap- 
peared to have pure X. mulleri rDNA by 
this test were examined by a more sensi- 
tive technique. The somatic DNA from 
each animal was centrifuged to equilib- 
rium in an actinomycin CsCl gradient. 
The X. mulleri and X. laevis rDNAs 
separate by this method. Hybridiza- 














• / 


n o 

1 1 1 



20 40 60 80 

Percent X.laevis rDNA 










-Em x mm 



^m s' 







n o 

i 1 

i i 

20 40 60 

Percent X.laevis rDNA 



Fig. 1. The relative abundance of X. mulleri and X. laevis rDNA in backcrossed hybrid animals. 
Somatic DNA was purified from each animal and hybridized with a mixture of cRNAs. ( 3 H)cRNA 
and ( 32 P)cRNA were transcribed from denatured X. laevis and X. mulleri rDNA, respectively, with 
E. coli polymerase. They were mixed with a 100-fold excess of nonradioactive rRNA and hybrid- 
ized. A standard curve was prepared by including filters with known mixtures of X. laevis and X. 
mulleri rDNA. The DNA of the mother and father of the progeny was measured also in each ex- 
periment. (A) Im X 11. (B) Im X mm. 



tion of fractions across each gradient 
showed that 18 out of the 20 progeny had 
some X. laevis rDNA. This result, while 
not yet conclusive, is strong evidence for 
extensive recombination in the oocyte be- 
tween the homologous chromosomes con- 
taining rDNA. 

Experiments that measure the num- 
bers of each kind of gene and their chro- 
mosomal location are in progress. Satu- 
ration hybridization experiments are be- 
ing used to determine the numbers of 
each kind of gene; while their chromo- 
somal location is being studied by in situ 
hybridization of metaphase chromosomes, 
in collaboration with Mary Lou Pardue 
at Massachusetts Institute of Technol- 

If we assume that selection of these 
genotypes in the backcrossed progeny has 
not occurred and they truly resemble the 
rDNA genotype of the germ cells in the 
F 1 female, then there appear to be two 
kinds of rDNA-containing chromosomes 
in the female which are transmitted to 
her progeny — those with pure X. mulleri 
rDNA and those with about equal 
amounts of each kind of rDNA. This 
suggests but does not prove extensive re- 
combination between the rDNA of homo- 
logues. It further demonstrates that the 
amplified rDNA copies in the oocyte, 
which in this animal are only X. laevis, 
do not recombine with the chromosomal 
A", mulleri rDNA extensively. On the 
contrary, the progeny inherit X. mulleri 
rDNA more abundantly from the female 
Fi than they do her A", laevis rDNA. 

In summary, we now have populations 
of hybrid animals which exhibit unusual 
transmission of rDNA genes to their 
progeny. The nature of this non-Mendel- 
ian segregation is not yet known, but one 
possibility is that these genes have under- 
gone extensive recombination during 
meiotic pairing. Experimental schemes 
for testing this and other possibilities 
have been detailed. 

If extensive recombination exists 
within this locus, it prompts us to specu- 
late about the origin and evolution of 

other highly repetitive regions of the 
genome. Frequent unequal crossing over 
might account for the process referred to 
as "saltatory" replication, which has been 
used to account for the origin of multiple 
repeats within satellite DNAs. This kind 
of mechanism can not only drastically 
change the number of tandemly arranged 
repeats but also cause them to evolve to- 
gether as a family. 

Transcription of 5S DNA in vitro 

K. Sugimoto 

Experiments were begun to study the 
fidelity with which 5S DNA is tran- 
scribed in vitro by added RNA polymer- 
ases. 5S DNA from X. laevis is particu- 
larly advantageous for these studies for 
several reasons: (1) Since the repeat 
length is short (about 0.5 to 0.6 X 10 6 
daltons), individual 5S DNA molecules 
contain many repeats and therefore pre- 
sumably many initiation and termination 
sequences. (2) The exact region of 5S 
DNA which is transcribed in vivo is 
known. It consists of about one-seventh 
of one strand ; the remainder of the strand 
and the opposite strand are not tran- 
scribed in vivo. Our evidence suggests 
that there is no precursor to 5S RNA in 
vivo, so initiation occurs exactly at the 5'' 
end of the molecule. (3) 5S RNA itself 
has been sequenced so that proper initia- 
tion would be recognized by sequence 
techniques. (4) A variety of Xenopus 
RNA polymerases have been purified by 
R. Roeder and are available for tran- 
scription of 5S DNA. 

Our studies to date have assayed the 
transcripts made by E. coli polymerase 
from purified 5S DNA, and recently we 
have done similar experiments with X. 
laevis RNA polymerase III in collabora- 
tion with Roeder. Both kinds of enzymes 
transcribe both strands of 5S DNA as 
well as the spacer sequences. Various salt 
concentrations have been tried with the 
same nonspecific results. 

Another approach to transcriptional 
fidelity has been to study isolated nuclei. 



Endogenous RNA synthesis proceeds for 
only about ten minutes in frog nuclei. 
Techniques have not yet been perfected 
in which reinitiation of RNA synthesis 
by endogenous polymerase can take 

Analysis of 5S DNA Heterogeneity 

D. Carroll 

We are using bacterial restriction en- 
zymes to investigate the heterogeneity 
and organization of the genes for 5S RNA 
in Xenopus. Here we describe experi- 
ments designed to ask whether adjacent 
repeats on a single molecule of 5S DNA 
are always identical in sequence. Let's 
construct some hypothetical repeat units 
(one gene and one spacer) , use them in 
various combinations to represent the 
total 5S DNA, and see what sort of pat- 
terns of restriction fragments might re- 
sult. Suppose that repeat type A has no 
sequence which is cleaved by a certain 
restriction endonuclease, repeat B has one 
such sequence, C has two, D has three, 
etc. If the 5S DNA is made up entirely 
of repeats of a single type, the enzyme 
will produce a homogeneous pattern of 
fragments. The number of different frag- 
ments will equal the number of cuts per 
repeat, and the sum of each kind of frag- 
ment will equal the repeat length. For 
example, multiple repeats of C would lead 
to only two different DNA fragments. If, 
on the other hand, the DNA contains sev- 
eral types of repeats, and repeats of each 
type are clustered, i.e., A's next to A's, B's 
next to B's, etc., the overall pattern of 
fragments will be a sum of the patterns 
for each type alone. DNA made up of 
clusters of A's and clusters of B's would 
lead to a mixture of long, uncut molecules 
(A) and repeat length-sized fragments 
(B). If repeats of different sequence are 
interspersed randomly, a very heteroge- 
neous collection of fragments will result. 
But if the interspersion is regular and re- 
peating— e.g., . . . BCDBCDBCD . . . — 
a unique pattern of fragments will be ob- 

served, and the sum of the fragment 
lengths (if they can all be distinguished) 
will equal the length of the larger repeat 
(B + C + D). 

If, in addition to differing in sequence, 
the repeats differ in length, the distribu- 
tion of fragment length will be very 
heterogeneous. If there is one cut made in 
each repeat, the distribution of fragments 
will be identical to the distribution of re- 
peat lengths; but this may be practically 
indistinguishable from the case in which 
both the number of cuts and the repeat 
lengths are heterogeneous. If there are 
internal repetitions within the spacer re- 
gions, the distribution of fragments might 
be more uniform than the repeat lengths 

In our initial experiments, we have 
examined in the electron microscope frag- 
ments produced by treatment of 5S 
DNAs with various restriction enzymes. 
The Ri enzyme from E. coli apparently 
makes no cuts in either X. laevis or X. 
mulleri 5S DNA. Digestion of X. laevis 
5S DNA with the H. influenzae restriction 
enzyme preparation endo R leads to the 
production primarily of repeat length- 
sized pieces (approximately 0.5 X 10 6 
daltons) . X. mulleri 5S DNA shows a 
heterogeneous distribution of fragments 
on treatment with endo R. Since the 
known restriction sequences of these en- 
zymes are not found within the sequence 
of 5S RNA itself, we presume that any 
cuts which do occur are made in spacer 
sequences. We are now investigating in 
detail the exact size distribution of pieces 
produced by endo R. This plus other 
kinds of experiments should elucidate the 
way in which heterogeneity of sequences 
is arranged in 5S DNA. 

The Use of Antibiotics in Buoyant 
Density Gradients 

R. Stem 

The utility of antibiotics as ligands in 
buoyant density gradient centrifugation 
has been investigated and actinomycin D 
has proven to be a great aid in fractionat- 



















15 5 10 

Fraction Number 

Fig. 2. CsCl buoyant density gradients of X. mulleri DNA. About 400 fig of DNA was centri- 
fuged in 20 ml gradients for 2 days at 35,000 rev/min in a 50.1 fixed angle Spinco rotor. The buoy- 
ant density of each fraction was determined from refractive index measurements. (A) control (B) 
200 fig of actinomycin. 

ing DNA. Actinomycin D binds to DNA thus the highest density bind the most 
and makes it less dense. Usually those actinomycin D and become the lightest 
DNAs with the highest GC content and fractions in a gradient. In Figure 2A a 













* > 







1 \i 

\ 1 



Fraction Number 

Fig. 3. Hybridization of CsCl buoyant density gradients of X. mulleri DNA with cRNA tran- 
scribed from 5S DNA and rDNA. About 100 fig of DNA in 5 ml gradients was centrifuged for 2 
days at 35,000 rev/min in a 40-angle rotor. (A) control (B) 50 fig of actinomycin added. Dashed 
line, cRNA from 5S DNA; open circles, cRNA from rDNA. 



20-ml gradient containing 400 fig of 
Xenopus mulleri DNA is shown. Two 
satellite DNAs are seen in analytical 
gradients at 1.683 and 1.686 g/cm 3 . They 
are not resolved here, but their approxi- 
mate positions are indicated by the 
arrows. In Figure 2B a similar gradient 
containing 200 fig of actinomycin D is 
shown. The satellite DNAs are now well 
resolved from each other as well as from 
the main-band DNA. 

Figure 3A shows a 5-ml CsCl gradient 
containing 100 ^g of X. mulleri DNA. The 
fractions have been placed on filters and 

hybridized with radioactive cRNA tran- 
scribed from 5S DNA and rDNA, and the 
hybridization patterns are indicated. In 
Fig. 3B a similar gradient containing 50 
fig of actinomycin D is shown. The satel- 
lite DNAs are pelleted at the bottom of 
the tube. The ribosomal DNA is found in 
a very low density region of the gradient. 
Unexpectedly the 5S DNA is also less 
dense than the main-band DNA. Figure 
4 shows the results from a series of exper- 
iments in which 400 fig of X. mulleri 
DNA in 20-ml gradients are banded in 
the presence of increasing amounts of 






50 IOO 

Actinomycin jsq 

Fig. 4. Effect of actinomycin D concentration on buoyant density of X. mulleri DNA compo- 
nents. About 400 fig of A", mulleri DNA in 20 ml gradients was centrifuged for 2 days at 35.000 
rev/min in a 50.1 fixed angle rotor. The abscissa indicates the amount of actinomycin D per 100 
fig of DNA. Solid circles, satellite I; open squares, satellite II; open circles with solid line, main- 
band DNA; open circles with dashed lines, 5S DNA; open triangles, rDNA. 



actinomycin D. The abscissa indicates 
the amount of actinomycin D per 100 /xg 
of DNA in each gradient. The buoyant 
densities of rDNA, 5S DNA, main-band 
DNA, and the two satellite DNAs are 
plotted at different actinomycin D levels. 
No additional separation of components 
is obtained with concentrations of actino- 
mycin above 50 /xg. The large buoyant 
density differences show the utility of this 

With separation of main-band DNA 
from the ribosomal and 5S RNA genes of 
X. laevis and X. mulleri as an assay sys- 
tem, other DNA-binding antibiotics 
studied originally in density gradients by 
Kersten et at. {Biochemistry 5, 236, 1966) 
were evaluated. Nogalamycin (a gift of 
Dr. Paul F. Wiley, The Upjohn Co.), 
daunomycin, chromomycin, olivomycin, 
mithramycin (a gift of Dr.T. J. McBride, 
Chas. Pfizer and Co.), and cinerubin (a 
gift of Dr. H. Zahner, Institut fur Mikro- 
biologie, Tubingen) all proved inferior to 
actinomycin D. 

Studies on Silk Fibroin mRNA 
and Its Genes 

P. Lizardi 

As a continuation of our work on silk 
fibroin mRNA, we have developed meth- 
ods for studying the structure and ex- 
pression of the genes which code for this 
mRNA in Bombyx mori. 

As mentioned in last year's Report 
{Year Book 71, pp. 21—22) , we have syn- 
thesized a radioactive reverse tran- 
scriptase product of fibroin mRNA 
("anti-messenger" DNA) . Experiments 
were designed to test whether this a- 
mDNA could be used as a specific probe 
for detection of fibroin genes. We took 
advantage of the fact that fibroin genes 
have a characteristic buoyant density, 
different from that of main-band DNA, 
when fractionated in gradients of cesium 
salts (Suzuki, Gage, and Brown, J. Mol. 
Biol., 70: 637). Using three types of 
equilibrium gradients (neutral CsCl, 
silver-Cs 2 S04 and actinomycin-CsCl) , we 

have fractionated sheared Bombyx DNA 
and have localized the position of indi- 
vidual genes by molecular hybridization 
assays of each gradient fraction. Profiles 
were obtained using ribosomal RNA, 
fibroin mRNA, and a-mDNA as hybrid- 
ization probes. In each of the three types 
of fractionation, the a-mDNA hybridiza- 
tion exhibited the same unique distribu- 
tion as the mRNA hybridization. There 
was little overlap with ribosomal genes 
or with main-band DNA. 

These results demonstrate the specific- 
ity of a-mDNA and imply that it can be 
used as a tool for the detection of fibroin 
messenger nucleotide sequences, whether 
these are in the form of an RNA or a 
DNA molecule. Furthermore, the assay 
is highly sensitive, being able to detect 















10 15 

Fraction No. 

Fig. 5. Hybridization of anti-messenger DNA 
(a-mDNA) with B. mori DNA which had been 
fractionated in an actinomycin-CsCl gradient. 
The gradient contained 90 fig of sheared carcass 
DNA and was centrifuged to equilibrium in a 
50.1 Spinco rotor. The immobilized DNA frac- 
tions were hybridized together with 3 H-a-mDNA 
(solid circles) and 32 P-ribosomal cRNA from X. 
laevis (open circles) ; absorbance at 260 nm 
(solid line). 


fibroin genes in the DNA from a single scribed previously but with DNase and 

silk gland. pronase treatment sometimes omitted. All 

A technique which proved of consider- preparations used were centrifuged 
able value in the aforementioned experi- through at least one sucrose gradient fol- 
ments was the fractionation of Bombyx lowed by one formamide-sucrose gradi- 
DNA in actinomycin-CsCl gradients, ent. Between 20 and 40% of the fibroin 
Figure 5 shows a typical gradient in mRNA was retained by oligo-dT cellulose 
which the position of ribosomal and fi- or polyU Sepharose, which specifically 
broin genes has been determined by mo- bind polyA regions of messenger RNA. 
lecular hybridization. The remarkable The material bound to the oligo-dT cellu- 
separation between main-band DNA and lose and that which is not retained both 
fibroin genes shows that the technique is have the base composition reported for 
potentially useful for studies on the co- fibroin mRNA. The "fingerprints" of a 
valent length of fibroin DNA, and ulti- two-dimensional separation of the oligo- 
mately, for the isolation of these genes, nucleotides from a Ti digest of the bound 
We are presently using the actinomycin- and nonbound mRNA were identical, and 
CsCl technique in a study of buoyant the oligonucleotides obtained were those 
density changes of fibroin genes as a predicted from the known Ti digest pro- 
function of DNA size. Another current file ( Year Book 70, pp. 35-39) . 
project involves the use of poly lysine After digestion with Ti plus pancreatic 
precipitation and actinomycin-CsCl gra- RNase at high salt, 1.4% of the bound 
dients for the isolation of fibroin genes, mRNA and 0.3% of the unbound mRNA 
using a-mDNA hybridization as a puri- could then bind to oligo-dT cellulose, 
fication assay. This retained fragment consisted of 82% 

Other present projects involve struc- A residues. The fragment migrated on 

tural studies of fibroin mRNA using elec- polyacrylamide gel at a size of approxi- 

tron microscopy, and the development of mately 6-7S, or 200 nucleotides in length, 

new methods for fibroin mRNA isolation The proportion of fibroin mRNA which 

using oligonucleotide-cellulose affinity bound to polyU-Sepharose rose from 30% 

columns. to 40% when the initial salt concentra- 
tion was increased from 0.1 to 0.3 M. 

A Polyadenylic Acid Sequence in Messenger RNA labeled for 6 hours and 

Silk Fibroin mRNA 72 hours in vivo had the same proportion 

of molecules which bound to the column. 
The inability to detect polyA in fibroin 

In the initial studies on the isolation of mRNA resulted from the very large size 
silk fibroin messenger RNA from Bombyx f the mRNA and the consequent diffi- 
mori, the oligonucleotides were analyzed culty in finding the relatively small pro- 
after digestion with Ti ribonuclease and portion of tfte sequence which is polyA . 
no significant amount of material larger R ig unclear why Qnl t of the mRNA 

than 10 residues in length was found , , ■, , ■, , ., 

t\7 r> 7 n,n or o^x mi r ., (whose purity can be assessed bv its 

{Year Book 70, pp. 35-39). Therefore, it ' . , J , . ... * 

„„„„ 4. i A i i A . fingerprint and base composition at 

was suggested that a polyA sequence is , ,, «„,* s , . 

not present. In view of the recent interest S reater than 80%) contams P ol ^ A se " 
in polyA sequences, their frequent pres- Quences. The slze of the mRNA whlch 
ence in animal mRNAs, and the priming stlcks and that whlch remains unbound 
capacity of silk fibroin mRNA for reverse to dT-cellulose are similar on a sucrose 
transcriptase with an oligoT primer, this gradient. No functional correlation has 
question was reexamined. been found between those mRNA mole- 
Fibroin mRNA was isolated from pos- cules which contain polyA and those 
terior silk glands of Bombyx mori as de- which do not. 

R. Williamson 




Quantitative Measurements of 
Fibroin mRNA Synthesis 

Y . Suzuki 
Assisted by E. Suzuki and N. Tsuchida 

During the past year, we have studied 
the following questions: (1) Are the fi- 
broin genes expressed not only in the 
fifth instar but also in earlier larval 
stages? (2) Is the induction of extensive 
fibroin synthesis due to increased fibroin 
mRNA content during the fifth instar or 
due to translational controls of a large 
preexisting population of fibroin mRNA 
which had accumulated before the fifth 
instar? (3) Is there repeated switching 
on and off of fibroin genes during larval 

These questions have been answered by 
quantitative measurements of fibroin 
mRNA at varying stages. Bio-Gel A-50, 
originally introduced by Lizardi and 

6 - 

Brown, efficiently separates 2-3 mg of 
fibroin mRNA (Fig. 6). The mRNA 
fraction thus obtained is generally about 
80% pure without further purification. A 
further sucrose gradient centrifugation 
step gives an mRNA fraction which is 
greater than 90% pure. 

During the feeding stage of the fourth 
instar both 32 P-labeled and nonlabeled 
RNAs were extracted from the posterior 
silk glands of several hundred larvae. 
About 1 mg of fibroin mRNA was iso- 
lated easily by the two-step purification 
described above. The mRNA was di- 
gested by Ti RNase and the digest was 
fractionated on DEAE-Sephadex A25 
column (Fig. 7). The optical density 
and radioactivity profiles are character- 
istic for fibroin mRNA (see Year Book 
' 70, p. 35) . From this experiment we con- 
clude that the mRNA can be measured 

u - 


2 - 

- 2 




1 ^ 



1500 ml 

Fig. 6. Large-scale preparation of fibroin mRNA on a Bio-Gel A-50m column. During the feed- 
ing stage of the fourth instar both 32 P-labeled RNA and nonlabeled RNA were extracted from the 
posterior silk glands of several hundred larvae. About 65 mg of the RNA was applied to a Bio-Gel 
column (5 X 80 cm) and was fractionated by an elution buffer containing Tris (20 ml, pH 7.5), 
NaCl (0.15 M), EDTA (1 ml), SDS (0.5%). Solid circles, 32 P cts/min; solid line, A 260 . 



0.3 - 

0.2 - 



0.1 - 

- 400 


200 ^ 

40 80 

Fraction number 

Fig. 7. DEAE-Sephadex fractionation of the oligonucleotides produced by Ti RNase digestion 
of fibroin mRNA from the fourth feeding stage. The mRNA fraction was obtained by the Bio-Gel 
column shown in Fig. 6, and was further purified by a sucrose density gradient centrifugation. About 
520 ^g of the pure mRNA was digested with Ti RNase, and the oligonucleotides produced were 
fractionated on DEAE-Sephadex column A25. Solid circles, 32 P cts/min; open circles, A 26 o. 

quantitatively and comprises about 2% 
of cellular RNA by weight. Therefore, 
mRNA comprises about the same fraction 
of cellular RNA in the fourth instar 
larvae as it does in the fifth instar (Table 
1). This result also answers the second 
question. The mRNA present in the 
fourth instar comprises only 2% (0.1 mg) 
of the maximum amount which accumu- 
lates in the fifth instar (5 mg) (Table 1) . 
These results led us to the third ques- 
tion: Are the fibroin genes switched on 
once in the beginning of larval life, and is 
the transcription going on continuously 
throughout the entire larval life? Or, is 
there repeated switching on and off of 
fibroin genes during larval life? In order 
to answer the question, studies were car- 
ried out on RNA preparations from the 
fourth moulting stage. Using the tech- 
nique described above, an RNA fraction 

which should contain fibroin mRNA was 
obtained. Its GC content was 42% DNA- 
like in base composition, and its oligo- 
nucleotide profile after Ti RNase diges- 
tion, both by radioactivity and optical 
density, was quite distinct from that of 
mRNA. The results suggest two alterna- 
tive possibilities. (1) Both the mRNA 
which accumulated in the fourth instar 
and the newly synthesized mRNA were 
mostly degraded. In this case "degrada- 
tion" would be selective for fibroin 
mRNA because synthesis and accumula- 
tion of rRNA, sRNA and DNA-like RNA 
are detectable at this time. (2) The tran- 
scription of the fibroin genes is abolished 
and the preexisting mRNA is degraded 
due to its natural half-life in the gland. 
At present, we cannot decide between 
these possibilities. If the fibroin genes 
are repeatedly switched on and off, this 



TABLE 1. Fibroin mRNA Content during the Fourth and Fifth 
Larval Instars of Bombyx mori 

Amount of RNA in 

a Pair of Posterior 

Silk Glands 


% of Fibroin mRNA 





RNA, mg 


by Weight 


4th instar (feeding 






Middle of 4th 

moulting stage 





5th instar (early 






5th instar (middle) 





5th instar (spinning 






K A11 the data for fibroin mRNA have been corrected for its purity. 

adds another dimension to the control 

In the fifth instar synthesis and ac- 
cumulation of fibroin mRNA begin again 
as early as the second day and continue 
as late as the spinning period. At the be- 
ginning of the fifth instar the mRNA 
comprises about 1% of the cellular RNA 
by weight and 1% by radioactivity 
(Table 1). These values increase grad- 
ually up to 3% (by weight) and 30% (by 
radioactivity) of cellular RNA toward 
the end of larval life (Table 1). A new 
question is raised if all of the fibroin 
mRNA is functional. The results also in- 
dicate that the transcriptional and/or 
posttranscriptional control for fibroin 
mRNA is quite different from that of 

Mutants for Fibroin Production 

Y . Suzuki and H. Chikushi 

Several spontaneous mutants of B. 
mori are affected in silk production. 
Larvae of these mutants metamorphose 
into pupae without cocoons or with un- 
usual cocoons composed of only sericin 
(Nd, naked pupa mutants). Other mu- 
tants spin a very thin cocoon containing 
both fibroin and sericin and are desig- 
nated flc (flimsy cocoon) . 

We have analyzed the Nd mutants. 
The mutant gene is dominant over its 
normal allele. Thin silk glands have the 
same number of cells as normal glands, 
and the anterior and middle portions of 
the mutant silk glands look almost iden- 
tical to those from normal animals. The 
posterior gland of the Nd animal, how- 
ever, is quite different from that of the 
normal animal. It is markedly inhibited 
in its growth, is very small in its size 
especially in the fifth instar, and does 
not contain appreciable amounts of 
fibroin in its lumen. The mutant cocoon 
(or often just fibers that do not form the 
typical cocoon shape) weighs a quarter 
of the normal and consists mainly of 
sericin with only 0.6% of the cocoon as 
fibroin-like material. The sericins from 
the normal and the mutant are indis- 
tinguishable and the mutant sericin is 
suitable for sericine chemistry. Genetic 
and biochemical experiments support the 
idea that the Nd gene acts only on the 
posterior silk gland and not on the an- 
terior or middle silk gland. In the fifth 
instar the posterior gland is partly de- 
generated, its wet weight and nucleic 
acid contents being about one-fifth of 
the normal. In the cells some usually 
large blocks of fibroin-like material are 



deposited. The deposition might be due 
to disturbed secretion of the fibroin-like 
material from cell to lumen, which in 
turn might cause degeneration of the cell. 
Recently, we analyzed 32 P-labeled 
RNA from this mutant. The RNA 
looked intact in spite of the slight de- 
generation of the posterior silk gland and 
contained about 1% of its total radioac- 
tivity as fibroin mRNA. Again the 
mRNA was identified by the typical 
oligonucleotide profile after RNase Tl 
digestion. We conclude that the gene 
Nd is not a mutant affecting fibroin gene 
transcription, but rather a mutant af- 
fecting some translational or posttrans- 
lational control mechanism. 

Fibroin mRNA from 

Bombyx mandarina 

Y. Suzuki and E. Suzuki 

A wild silkworm, Bombyx man- 
darina, which inhabits China, Japan, 

and Korea is presumed to be the ancestor 
of Bombyx mori. According to Chinese 
literature Bombyx mori was domesti- 
cated in China for the purpose of silk 
production more than 4000 years ago. 

If this presumption is correct, then B. 
mori and B. mandarina should have in- 
distinguishable fibroin genes. The argu- 
ment stems from the following fact. 
Fibroins are secreted and have an "ex- 
tracellular function" as cocoons, webs, 
and various attachment threads. Once 
produced they do not have contact with 
intracellular components of the animal. 
It is proposed that this lack of contact 
enables animals to tolerate a variety of 
these proteins. Actually, a surprising 
variety of silk fibroin is produced by 
various insects and arachnids, and fi- 
broins are classified into several groups 
by their extremely different amino acid 
compositions and x-ray diffraction pat- 
terns, in marked contrast to the con- 
served homology of amino acid sequence 

12 - 

'2 8 



20 40 60 

Fraction number 

Fig. 8. The oligonucleotide profile of Ti RNase digest of fibroin mRNA from Bombyx man- 
darina. 32 P-labeled fibroin mRNA was purified and digested with Ti RNase, and the digest was 
fractionated on a DEAE-Sephadex A25 column. 



or nucleotide sequence of functionally 
related proteins or RNAs in different 

The amino acid composition of fibroins 
from B. mori and that of B. mandarina 
are indistinguishable. Although the 
amino acid sequence of the latter fibroin 
has not been analyzed, x-ray diffraction 
patterns of the two fibroins strongly sup- 
port the idea that they have similar 

The fibroin mRNA from B. mandarina 
has been isolated and partly sequenced. 
The mRNAs from B. mori and B. man- 
darina are indistinguishable in molecular 

size under denaturing conditions. Their 
nucleotide sequences are the same ex- 
cept for some modifications in the third 
nucleotides of some codons (compare 
Fig. 8 with Year Book 70, p. 35). These 
modifications should provide a good ex- 
perimental system for the analysis of 
functional adaptation of the tRNA pop- 
ulation to fibroin synthesis in the pos- 
terior silk gland. Mutations in the third 
nucleotides of codons of fibroin mRNA 
should have been accompanied by a 
change in regulation of the tRNA popu- 



R. Williamson and P. Wellauer 

The accurate determination of the 
molecular weight of mouse globin mes- 
senger RNA has been complicated by the 
lack of standard reference RNA mole- 
cules; ribosomal RNA may migrate dif- 
ferently in polyacrylamide gels or su- 
crose gradients because of its extensive 
secondary structure. Since the electron 
microscope visualization techniques de- 
veloped by Norman Davidson and now 
used in this Department utilize spread- 
ing in formamide, which eliminates most 
secondary structure, the molecular 
weight of the globin mRNA can be 
measured directly. Moreover, as in ribo- 
somal RNA, any regions of resistant 
secondary structure can be seen and 

Current estimates of the molecular 
weight of globin mRNA vary from 150,- 
000 to 230,000. The coding region of the 
mRNA has a calculated molecular weight 
of 143,000 (a-chain) or 148,000 (0-chain). 
There is a polyA sequence approximately 
60 nucleotides in length (20,000) and a 
noncoding but identified sequence 93 nu- 
cleotides in length, at least in human a- 
globin mRNA, as shown by study of the 

abnormal hemoglobins Constant Spring 
and Wayne (30,000). A more accurate 
estimate of molecular weight should per- 
mit an analysis of the number of nucleo- 
tides at the 5'-end of the molecule (prior 
to the initiation triplet) and between the 
final termination triplet (that for mutated 
hemoglobin Constant Spring) and the 
polyA region. 

Three different globin mRNA prepara- 
tions, prepared either by gradient centri- 
fugation of EDTA-dissociated polysomes 
followed by isolation of the 14S mRNP 
or by polyU-Sepharose affinity chroma- 
tography, were spread from 80% form- 
amide— 4M urea and the lengths measured 
using E. coli 16S rRNA as a standard. 
The molecular weight determined was 
200,000 ± 50,000. This is in agreement 
with the values obtained by gel electro- 
phoresis but not those from gradient sedi- 
mentation or end group analysis. It al- 
lows a prediction that for human a-chain 
mRNA there is a maximum of 60 nucleo- 
tides both at the 5'-terminus and between 
the termination triplet for hemoglobin 
Constant Spring and the polyA region. 





R. H. Reeder and T. Honjo 
with the assistance of R. Handler 

A long-term goal of this laboratory 
has been the establishment of an in vitro 
system in which the control mechanisms 
acting on animal cell genes could be re- 
produced. The problem of gene control is 
one of the central problems in develop- 
mental biology and its solution is almost 
certain to require reproduction of those 
controls in vitro. The first chance for a 
truly molecular attack on this problem 
was provided by the isolation of a single 
eukaryotic gene of known function, the 
gene coding for ribosomal RNA from the 
frog Xenopus laevis. In order to use this 
isolated gene in studying transcription, it 
was first necessary to learn more about 
its structure and function in vivo and 
then to develop assays that would accu- 
rately measure the correctness of its 
transcription in vitro. Therefore, con- 
siderable time was expended in develop- 
ing methods for measuring strand selec- 
tion, spacer-versus-gene transcription, 
and specific initiation. In the course of 
setting up these assays, Escherichia coli 
RNA polymerase was used because of its 
ready availability to transcribe the ribo- 
somal genes in vitro. It was shown that 
this prokaryotic enzyme is unable to rec- 
ognize the proper transcription initiation 
signals in a eukaryotic ribosomal gene 
{Year Book 68 and Year Book 69). The 
bacterial enzyme does transcribe these 
genes in a nonrandom manner, but the 
pattern of transcription is not the same 
as occurs in vivo. At about this time, the 
first reliable procedure for isolating eu- 
karyotic RNA polymerase was devel- 
oped, making it possible to study in vitro 
transcription of the ribosomal genes using 
their own homologous polymerase. This 
work was done in collaboration with R. 
Roeder. At first it was not clear which 
polymerase to use, however, since Xen- 
opus has at least three distinct types of 
nuclear RNA polymerases {Year Book 

69). Studies on isolated nuclei using the 
fungal inhibitor a-amanitin showed that 
the most likely candidate was polymerase 
I and showed furthermore that ribosomal 
RNA was by far the major product of 
that enzyme {Year Book 70). Methods 
were developed during that study for 
measuring rRNA synthesis in the pres- 
ence of large amounts of non-rRNA syn- 
thesis and to assess its fidelity of tran- 
scription. A major disappointment came 
when Xenopus polymerase I was used to 
transcribe purified ribosomal genes. Poly- 
merase I (and later polymerases II and 
III) transcribed rDNA just as aber- 
rantly, if not more so, as E. coli poly- 
merase had done. The obvious implica- 
tion was that some factor (s) essential for 
specific transcription had been lost dur- 
ing purification of the genes, the poly- 
merase, or both. The first efforts were 
directed toward the enzyme itself to see 
if some subunit or cofactor had been lost 
during purification. Enzymes were tested 
at various stages of purification, crude 
fractions were added back to purified en- 
zymes, and enzymes from several differ- 
ent sources (embryos, liver, cultured 
cells, oocytes) were all tested. None 
showed any improved fidelity of tran- 
scription over the purified enzyme. At- 
tention was then shifted to the possibility 
that some essential substance had been 
lost from the gene itself during purifica- 
tion. T. Honjo began experiments in 
which partially purified enzyme was 
added to crude chromatin to see if this 
could stimulate correct transcription of 
the ribosomal genes. The results of this 
investigation are presented in this year's 

AVhile the experiments on ribosomal 
gene transcription were in progress, 
Brown isolated another gene, that coding 
for 5S RNA of ribosomes. Assays have 
also been developed for measuring the 



fidelity of its transcription under various 

Corollary to the experiments with Xen- 
opus RNA polymerase and chromatin, a 
second study has been under way for 
about two years in which the ability of 
E. coli RNA polymerase to transcribe 
chromatin faithfully has been explored. 
This work was undertaken largely be- 
cause there is a considerable literature 
stating that E. coli RNA polymerase 
transcribes chromatin faithfully, i.e., it 
copies the same genes in vitro as were 
transcribed in the living cell from which 
the chromatin was derived. Since we had 
at hand techniques for examining this 
question in much greater detail than had 
been done before, we felt it was worth pur- 
suing. The findings are also presented in 
the present account. 

During the year, another, more biolog- 
ical, study of ribosomal gene control was 
also undertaken by Hon jo — a biochem- 
ical description of the dominance of 

Xenopus laevis rRNA synthesis over that 
of Xenopus mulleri rRNA in Fx hybrids 
between these two species. 

Transcription of rRNA and 5S RNA 

Genes in Amphibian Chromatin by 

Homologous RNA Polymerase 

T. Honjo and R. H. Reeder 

As described previously (Year Book 
71, pp. 15-16), we have been testing the 
possibility that the use of intact chroma- 
tin and RNA polymerase purified from 
the homologous animal would provide us 
with accurate transcription. We were un- 
able to get conclusive results earlier, since 
exogenous RNA polymerase stimulated 
RNA synthesis only 2- to 3-fold over 
endogenous RNA synthesis. 

We have overcome this difficulty by 
concentrating RNA polymerase up to 40- 
fold and isolating chromatin from liver 
instead of tissue-culture cells. The chro- 
matin preparation from the latter cells 

TABLE 2. 5S RNA Synthesis by Chromatin and RNA Polymerases I and II 

RNA Hybridized to 5S DNA 2 , 




Total RNA 

Experimental System 1 











(A) Chromatin alone 




Chromatin + 

polymerase I 






(B) Chromatin alone 






Chromatin + 

polymerase II 






1. Each reaction mixture contained 50 ml Tris-Cl, pH 8.0, 5 ml MgCl 2 , 
1 mikf MnCl 2 , 50 ml (NH 4 ) 2 S0 4 , 0.63 ml each of ATP, UTP, and GTP 
50 ^Ci of CTP- 32 P (specific activity 20 Ci/mmole), chromatin (47 fig of DNA 
for exp. A and 31 £ig of DNA for exp. B) and enzymes as indicated (54 units 
of polymerase I and 40 units of polymerase II). The reaction mixtures were 
incubated at 30° for 60 minutes. 

2. Hybridization was performed in 4X SET and 50% formamide at 40° 
for 18 hours using 0.1 /".g each of strand-separated 5S RNA bound to a Milli- 
pore filter in the presence or absence of 30 £ig of cold 5S RNA. The efficiency 
of hybridization was monitored by the addition of ( 3 H)-RNA transcribed from 
either H or L strand of 5S DNA by E. coli RNA polymerase. The efficiency 
ranged from 14% to 19% for L strand and from 15% to 18% for H strand. 















in vivo r RNA 

in vitro RNA 


*>—A ^...O 




1000 — 





Fraction Number 

Fig. 9. Fidelity of ribosomal gene transcription by RNA polymerase I. 32 P-labeled RNA syn- 
thesized by polymerase I with chromatin as template was mixed with 3 H-labeled rRNA synthesized 
in vivo. The RNA mixture was hybridized to strand-separated rDNA in the presence and absence 
of a 100-fold excess of unlabeled rRNA. 

has a very high endogenous activity for 
RNA synthesis and does not respond well 
to exogenous RNA polymerase. As shown 
in Table 2, using liver chromatin both 
polymerases I and II stimulate total 
RNA synthesis at least 10-fold above 
endogenous RNA synthesis. The RNA 
from each reaction was purified and hy- 
bridized to the L and H strands of 5S 
DNA in the presence and absence of an 
excess amount of cold 5S RNA. Chroma- 
tin alone synthesized only a negligible 
amount of 5S RNA, while polymerase I 
transcribed 5S genes in chromatin in sig- 
nificant quantity. However, the products 
of 5S genes transcribed by polymerase I 

hybridized to both L and H strands. 
Since 5S RNA synthesized in vivo is 
complementary only to the L-strand, it 
is clear that transcription in vitro by 
polymerase I is aberrant. Spacer se- 
quences were also transcribed by poly- 
merase I because only 30% of RNA 
hybridized to L strand was competed by 
native 5S RNA. Essentially similar ob- 
servations were made with polymerase II. 
Polymerase II seems not to be responsible 
for 5S RNA synthesis in vivo but it tran- 
scribes 5S genes in chromatin in vitro. A 
similar experiment measuring the fidelity 
of transcription of the genes for 18S and 
28S rRNA is shown in Fig. 9. In this case 



also, RNA made by polymerases I and II 
hybridized to both strands and to both 
gene and spacer sequences. 

From the results just described, we can 
conclude that (1) both RNA polymerases 
I and II can initiate RNA chains in vitro 
using chromatin as template; (2) RNA 
transcripts from this system contain 
RNA sequences not found in living cells. 
For both 5S and rRNA genes, therefore, 
eukaryote RNA polymerases transcribe 
chromatin no more accurately than does 
E. coli RNA polymerase. 

Transcription of Chromatin by 
Bacterial RNA Polymerase 

R. H. Reeder 

There are a number of experiments re- 
ported in the literature stating that bac- 
terial RNA polymerase can only tran- 
scribe a restricted set of DNA sequences 
when given eukaryotic chromatin as a 
template. Furthermore, the sequences it 
does transcribe are believed to be very 
similar, if not identical, to those tran- 
scribed in the living cell from which the 
chromatin originated. These conclusions 
are largely based on the results of RNA- 
DNA hybridization experiments that de- 
tected only sequences which are highly 
repetitious in the DNA. 

Our own studies of the transcription of 
specific sequences in chromatin seemed to 
lead to the opposite conclusion, i.e., that 
bacterial RNA polymerase transcribes 
eukaryotic genes nonspecifically and the 
presence or absence of the chromatin pro- 
teins makes little difference. Therefore, 
additional experiments have been done to 
explore this matter further. 

We began with chromatin from mouse 
liver, since fairly extensive studies have 
already been reported by others using 
this tissue. In the first experiment un- 
labeled nuclear RNA was tested for its 
ability to compete with (1) radioactive 
RNA transcribed from chromatin by E. 
coli RNA polymerase, (2) RNA tran- 
scribed from deproteinized bulk DNA, 
and (3) RNA synthesized by the endoge- 

nous RNA polymerase in nuclei. The re- 
sults are shown in Fig. 10. 

In this experiment RNA from chroma- 
tin and endogenously labeled RNA from 
nuclei were both competed by unlabeled 
nuclear RNA in an identical manner. To 
this extent, the results agree with previ- 

100 150 

Unlabeled nuclear RNA (jd.g) 

Fig. 10. Hybridization of various RNAs to 
mouse DNA and competition with unlabeled 
mouse liver nuclear RNA. In two separate ex- 
periments a different 3 H-labeled RNA was 
mixed with 32 P-labeled RNA transcribed from 
deproteinized mouse bulk DNA, hybridized to 
filters containing unfractionated mouse DNA 
(15 ^g), and competed with unlabeled mouse 
liver nuclear RNA. Since the ( 32 P)-RNA com- 
petition curves were superimposable, both ex- 
periments have been combined in one graph. 
Mouse liver chromatin transcript (25,000 3 H 
cts/min added, 100% represents 246 cts/min 
bound), solid circles. Deproteinized mouse DNA 
transcript (61,200 32 P cts/min added, 100% rep- 
resents 1,100 cts/min bound), open circles. En- 
dogenous nuclear RNA (44,000 3 H cts/min 
added, 100% represents 492 cts/min bound), 
solid squares. In a separate experiment mouse 
satellite L-strand transcript (18,000 3 H cts/min) 
was hybridized to niters containing mouse satel- 
lite L-strand DNA (3 fig,) and competed with 
unlabeled mouse liver nuclear RNA (100% rep- 
resents 572 cts/min bound) ; shown by solid 
triangles. All points are the average of two 
duplicate filters. 



ous observations. The unexpected result 
is that the bulk DNA transcript is also 
strongly competed by unlabeled nuclear 
RNA. This strong competition disagrees 
with previous reports and, if true, greatly 
reduces the sensitivity of the assay to 
detect differences between RNA and 
chromatin transcripts in vivo. 

A trivial explanation for the strong 
competition of bulk DNA transcripts 
would be the presence in the unlabeled 
nuclear RNA of some contaminant which 
nonspecifically prevented hybrid forma- 
tion. To test this possibility 32 P- 
labeled RNA was transcribed from the 
L-strand of purified mouse satellite DNA 
and hybridized to mouse satellite L- 
strand DNA on filters. Addition of the 
same amount of unlabeled nuclear RNA 
as used previously caused no reduction in 
hybrid formation (Fig. 10). This shows 
that the competition between nuclear 
RNA and bulk DNA transcript is due to 
base sequence homology and not some 
nonspecific artifact. This experiment is 
additional evidence that, as reported by 
other workers, mouse satellite L-strand 
is not transcribed in liver nuclei or, at 
least, the transcript is not accumulated. 
The experiment in Fig. 10 has been re- 
peated using the hybridization conditions 
described by Smith et al. (1969) with and 
without RNase. Under all conditions 
tested, nuclear RNA competed more than 
50% of the hybridization of bulk DNA 
transcripts. We conclude that this assay 
procedure is a very insensitive method for 
demonstrating that chromatin is a re- 
stricted template for bacterial RNA 

Mouse satellite sequences comprise 
about 10% of mouse DNA and are 
known to reassociate very rapidly. There- 
fore, it seemed possible that the entire 
difference between chromatin and DNA 
transcripts in Fig. 10 could be explained 
by a failure of E. coli RNA polymerase 
to transcribe satellite sequences in chro- 
matin. As a test of this possibility, mouse 
DNA free of satellite sequences was im- 
mobilized on filters and hybridized to a 

mixture of 3 H-labeled chromatin tran- 
script and 32 P-labeled deproteinized 
DNA transcript exactly as done for the 
experiment in Fig. 10. The hybridization 
was competed by increasing amounts of 
unlabeled liver nuclear RNA. The results 
(Fig. 11) show that after removal of 
satellite sequences from the filter-bound 
DNA (but not from the DNA used for 
template) chromatin and deproteinized 
DNA transcripts behave identically in 
this type of test: 

The ability of E. coli RNA polymerase 
to transcribe mouse satellite sequences in 
chromatin was examined more directly in 
the following experiment. 3 H-labeled 
RNA was transcribed from liver chroma- 

100 < 


remaining (% ) 

O* 00 

o o 
i i 



^ 1 

i, 40- 

Chromatin ""^ 


£ 20- 






l l 

50 100 150 

Unlabeled nuclear RNA (jj.g ) 

Fig. 11. Hybridization of mouse liver chroma- 
tin and deproteinized DNA transcripts to satel- 
lite-free mouse DNA. Mouse liver chromatin 
transcript (100,000 32 P cts/min) was mixed with 
deproteinized DNA transcript (487,000 3 H cts/ 
min), hybridized to mouse DNA from which 
the satellite has been removed (6 //.g/ filter), and 
competed with unlabeled mouse liver nuclear 
RNA. Chromatin transcript (100% represents 
1575 cts/min bound), solid circles. Deprotein- 
ized DNA transcript (100 c /c represents 14,445 
cts/min bound), open circles. All points are the 
average of two duplicate filters. 


TABLE 3. Measurement of Mouse Satellite L-strand Transcription 

Template for ( 3 H) RNA 

Fraction of ( 3 H) cts/min % of ( 3 H) 
( 3 H) cts/min ( 3 H) cts/min ( 32 P) cts/min in L-strand in L-strand 
Added Hybridized 1 Hybridized 2 Sequences 3 Sequences 

Endogenous nuclear DNA 






Deproteinized DNA 







polymerase 4 (units) 

























1 Each reaction contained a filter with 3 /ug of mouse satellite L-strand DNA. 

2 As an internal standard, 32 P-labeled L-strand transcript (10,000-20,000 cts/min) was added to 
each reaction. 

3 Calculated by dividing ( 3 H cts/min hybridized) by (fraction of 32 P cts/min hybridized). 

4 Each reaction contained mouse liver chromatin (57 ng of DNA) plus the indicated amount of 
E. coli RNA polymerase. After 30 minutes at 30° the total RNA from each reaction was isolated 
and its content of satellite sequences was measured. 

tin and mixed with a small amount of 
( 32 P)-RNA transcribed from purified 
satellite L-strand. The mixture was hy- 
bridized to a filter containing satellite L- 
strandDNA. The amount of ( 32 P)-RNA 
hybridized measures the efficiency of 
satellite sequence hybridization in this 
particular reaction. Knowing this effi- 
ciency and the amount of ( 3 H) chroma- 
tin RNA that hybridized, one can calcu- 
late the fraction of the total ( 3 H) 
chromatin RNA consisting of satellite se- 
quences. The results of such measure- 
ments when made on several types of 
RNA are shown in Table 3. RNA made 
in isolated mouse liver nuclei by endoge- 
nous polymerases contained no satellite 
sequences. When E. coli polymerase was 
used to transcribe deproteinized mouse 
DNA, 6.6% of the CMP incorporated 
was found in satellite sequences. Since 
it is known that the mouse satellite L- 
strand is about 5% of the total DNA and 
has a CMP content of 14% versus 20% 
for bulk DNA, one would expect 3.5% of 
the CMP incorporated to be in L-strand 
sequences if transcription were com- 
pletely random. The fact that almost 
twice this amount was incorporated sug- 
gests that E. coli RNA polymerase tran- 

scribes L-strand sequences preferentially 
in bulk DNA. 

Table 3 also shows the amount of satel- 
lite transcription that occurred when E. 
coli polymerase was used to transcribe 
mouse liver chromatin at varying chro- 
matin-to-polymerase ratios. At all poly- 
merase concentrations, ranging from be- 
low saturation to above saturation of the 
chromatin, transcription of satellite se- 
quences was uniformly low. The low 
level of satellite transcription observed in 
chromatin is significant and reproducible 
but is about an order of magnitude lower 
than that observed for deproteinized 

Previous workers were unable to detect 
any differences between chromatin- 
primed RNA and RNA made in vivo in 
the same tissue. Therefore, they con- 
cluded that the two types of RNA were 
largely identical and the bacterial poly- 
merase was transcribing chromatin in a 
faithful manner. The experiments re- 
ported here suggest that the major reason 
for not finding differences earlier was the 
insensitivity of the hybridization test em- 

There are several reasons for this lack 
of sensitivity. (1) The reiterated DNA 
sequences in eukaryotic genomes are 



composed of large families of related se- 
quences, each family cross-hybridizing 
within itself. Transcription of only one 
member of a family in vivo is sufficient to 
cause nuclear RNA to compete with the 
hybridization of all members of that 
family whether they are transcribed in 
vivo or not. The finding that unlabeled 
nuclear RNA competes effectively with 
deproteinized DNA transcripts (Fig. 10) 
suggests that in the mouse at least one 
member of most reiterated sequence fam- 
ilies is transcribed in vivo. (2) Hybridi- 
zations done at low RNA concentrations 
tend to be dominated by the most reiter- 
ated and rapidly reassociating sequences, 
such as the satellite in mouse DNA. Re- 
moval of this one sequence caused mouse 
chromatin transcript and deproteinized 
DNA transcript to behave identically in 
hybridization-competition (Fig. 11). It 
is likely that previous hybridization ex- 
periments were dominated similarly and 
no conclusions can be drawn from them 
concerning transcription of the rest of the 
mouse genome. (3) The previous experi- 
ments were also insensitive, since they 
were not designed to distinguish wrong- 
strand from correct-strand transcription. 
It has been shown that unlabeled RNA 
homologous to only one strand of the 
DNA will compete efficiently for the hy- 
bridization of labeled RNA homologous 
to both stands. Presumably competition 
on one strand occurs through the usual 
isotope dilution and on the other through 
formation of RNA-RNA hybrids. There- 
fore, E. coli polymerase can (and prob- 
ably does) transcribe both DNA strands 
in chromatin, and the transcripts will still 
be efficiently competed by unlabeled nu- 
clear RNA. 

A reasonable interpretation of these 
results, which could also accommodate 
earlier observations, is that transcrip- 
tional control in eukaryotes occurs on at 
least two levels. One level of control 
would be a gross control in which certain 
blocks of DNA would be complexed with 
protein so as to make it inaccessible to 
any RNA polymerase, bacterial or eu- 

karyotic. It is possible that in many cells 
the most highly reiterated sequences 
(satellites) are subject to this type of 
control and could thus account for the 
previously observed similarity between 
chromatin-primed and in vivo transcripts. 
This possibility is suggested by the occur- 
rence of many satellite DNAs in hetero- 
chromatic regions of the chromosomes. 
Hybridization of such satellite sequences 
might be expected to contribute dispro- 
portionately to hybridization experiments 
that measure only reiterated sequences. 
It is clear, however, that a second level of 
fine transcriptional control must also 
exist in vivo. The endogenous RNA poly- 
merases are able to recognize initiation 
and termination signals and to select the 
correct DNA strand. The evidence shows 
that E. coli RNA polymerase recognizes 
none of these fine controls in eukaryotic 
DNA whether or not chromatin proteins 
are present. Thus, it transcribes spacer 
as well as gene regions and incorrect as 
well as correct DNA strands. 

Repression of Xenopus mulleri rDNA in 
X. laevis X X. mulleri Hybrid Frogs 

T. Hon jo and R. H. Reeder 

Wild-type frogs of the species Xenopus 
laevis and Xenopus mulleri contain two 
prominent nucleoli per cell throughout 
most of embryogenesis. Blackler and co- 
workers have recently shown, however, 
that Fi hybrids between these two species 
have only one nucleolus in the large 
majority of their cells, suggesting that 
the nucleolar organizer of one of the two 
species is somehow inactivated by the 
presence of the other. Since it is well 
established that the nucleolus is the site 
of rRNA synthesis in interphase cells, it 
has been inferred that the rDNA of one 
species is repressed in the hybrid cells. If 
this inference is correct, such cells may 
provide a useful system in which to ana- 
lyze the control of rRNA synthesis in 

We have undertaken a biochemical de- 
scription of the phenomenon of nucleolar 



dominance in hybrids between laevis and 
mulleri. These hybrids have an obvious 
advantage for this type of study, since 
the rDNA from each species has been 
isolated in pure form and their structures 
have been studied extensively. Both the 
rDNA and the rRNA precursors of the 
two species can be distinguished. 

In describing the crosses and genotypes 
in this report, we follow the convention of 
giving the species of the female first. The 
diploid genotypes of the rDNA of wild- 
type laevis and mulleri are expressed by 
l/l and m/m, respectively. For example, 
l/l~ indicates laevis heterozygous for the 
nucleolar mutation which deletes the 

Brown and Blackler originally showed 
that hybrid frogs contain rDNA from 
both parents, as expected. Our results 
have further confirmed this finding. 
Knowing that both types of rDNA were 
present, we developed a method of deter- 
mining which rDNA was being tran- 
scribed in hybrid frog nuclei. The rRNAs 
of both species of Xenopus are initially 
transcribed as large precursor molecules 
with a size of about 2.5 X 10 6 daltons. 
About 2.2 X 10 6 daltons of this molecule 
are accounted for by sequences for 18S 
and 28S rRNA which cannot be distin- 
guished between laevis and mulleri. The 
remainder of the precursor molecule 
however, differs considerably in sequence 
between the two species and hybridizes 
about 10-fold more efficiently to homolo- 
gous than to heterologous rDNA. To 
determine which species of rRNA pre- 
cursor is being synthesized in a hybrid 
animal, authentic laevis 3 H-labeled pre- 
cursor was mixed with 32 P-labeled pre- 
cursor from the hybrid animal and hy- 
bridized to purified laevis rDNA and to 
purified mulleri rDNA in the presence of 
excess unlabeled 18S and 28S rRNA. 
From the isotope ratios binding to the 
two types of rDNA, the ratio of laevis to 
mulleri rRNA in the hybrid animal can 
be determined. 

To label precursor rRNA molecules to 
a sufficiently high specific radioactivity, 

nuclei were isolated either from the liver 
of a single animal or from a group of 
embryos, and the endogenous RNA poly- 
merase was allowed to elongate rRNA 
chains in vitro. This rRNA has been 
shown to be transcribed from the H- 
strand of the rDNA, to consist largely of 
18S and 28S rRNA sequences, and to 
contain additional sequences which are 
not 18S or 28S. It thus contains nucleo- 
tide sequences similar to those found in 
authentic precursor rRNA molecules. 

Using rRNA labeled in isolated nuclei, 
we constructed a standard curve relating 
the isotope ratios bound by the two 
rDNAs to the proportion of laevis and 
mulleri rRNA in a mixture. In separate 
reaction mixtures nuclear RNAs from 
embryos of laevis and mulleri were 
labeled with 32 P. The 32 P-labeled RNAs 
were then mixed in several proportions. 
Nuclei from laevis tissue culture cells 
were used to synthesize 3 H-labeled nu- 
clear RNA as a standard. To each sam- 
ple of 32 P-labeled RNAs was added an 
equal amount of 3 H-labeled standard 
RNA and lOO^g of unlabeled 18S and 28S 
rRNA. Each doubly labeled RNA mix- 
ture was hybridized to a filter containing 
mulleri rDNA and to a filter containing 
laevis rDNA in the same vial. 

The result of such an experiment is 
that for each artificial mixture of 32 P- 
labeled mulleri and laevis RNA one ob- 
tains a 3 H/ 32 P ratio bound to laevis 
rDNA (L ratio) and a 3 H/ 32 P ratio 
bound to mulleri rDNA (M ratio) . When 
the ( 32 P)-RNA is 100% laevis, the L 
ratio should be the same as the M ratio 
and therefore the ratio of the L ratio to 
the M ratio (L/M ratio) should equal 
unity. At the other extreme, when the 
( 32 P)-RNA is 100% mulleri, the L ratio 
is greater than the M ratio. We have 
determined empirically that the L/M 
ratio is close to 158 at this extreme. A 
standard curve obtained by the above 
method is shown in Fig. 12. 

The sensitivity of the method was 
checked by making measurements on dif- 
ferent batches of pure laevis and pure 






i 1 1 r 

0' 20 40 60 80 100 

mulleri rRNA(% 

10 20 30 40 
mulleri rRNA (%) 

Fig. 12. Standard curve for the measurement 
of mulleri rRNA synthesis in hybrid frogs. Con- 
struction of the curve is described in the text. 
The points are from individual measurements 
on artificial mixtures of laevis and mulleri nu- 
clear RNA. Each symbol represents a separate 

mulleri nuclei. From the measured L/M 
ratio and the use of Fig. 12, we calculated 
the apparent percentage of mulleri rRNA 
synthesized in each batch. The values 
range from to 3 for pure laevis and 96 
to 100 for pure mulleri nuclei. The lower 
limit of detection is therefore about 5% 
mulleri rRNA in the presence of 95% 
laevis rRNA. 

Using this method, we first measured 
rRNA synthesis in Fi hybrid embryos 
from the cross laevis (I /I) X mulleri 
(m/m). The results are shown in Table 
4. From the earliest stage tested, stage 
15, and thereafter until about stage 43, 
the mulleri rDNA was repressed com- 
pletely. At stage 43 and beyond there 
was detectable mulleri rRNA synthesis in 
the population of hybrid embryos. 

We also measured rRNA synthesis in 
adult tissues from several hybrid frogs 
(Table 5) . In liver nuclei from three of 

TABLE 4. rRNA Synthesized by laevis X 
mulleri Hybrid Embryos 


mulleri rRNA, % 




100 Mg of 18S and 28S laevis rRNA and 3 H- 
labeled standard RNA (191,000 cts/min) syn- 
thesized by laevis nuclei were added to ( 32 P)-RNA 
of each stage except stage 15-17 where 382,000 
cts/min of ( 32 H)-RNA was employed. The RNA 
mixtures were hybridized to filters containing 
approximately 1 /u-g each of laevis and mulleri 

the animals mulleri rRNA synthesis was 
either absent or very low. In the other 
four animals between 17% and 50% of the 
rRNA synthesized was mulleri. The 
second type of cross we examined was the 
cross laevis (l/l~) X mulleri (m/m). 
This type of mating results in an embry- 
onic population with a one to one mixture 
of the two genotypes, l/m and I'/m. Since 
the two genotypes cannot be distinguished 
morphologically, mixed populations of 
embryos were analyzed for rRNA syn- 

TABLE 5. rRNA Synthesized by laevis X 
mulleri Hvbrid Adults 




rRNA, % 























A separate animal was used for each analysis 
except for cross no. 3, in which both liver and 
kidney were taken from one animal. Cross no. 
1 was l/l~ X m/m. Somatic DNA of the animals 
used from cross no. 1 was shown to have the 
genotype l/m. Crosses no. 2 and no. 3 were 
I /I X m/m. RNA from hybrid nuclei was labeled 
with 32 P and standard laevis RNA was labeled 
with 3 H. 



thesis. It was assumed that any mulleri 
rRNA synthesis would be derived from 
the l~/m genotype, since the l/m geno- 
type makes no mulleri rRNA prior to 
stage 43 (Table 3) . The results are shown 
in Fig. 13. At stage 13-14 there was al- 
most no detectable mulleri rRNA synthe- 
sis. After stage 13-14 mulleri rRNA syn- 
thesis gradually increased relative to 
laevis until stage 43 when mulleri rRNA 
accounted for about 50% of the rRNA 
synthesized in the total embryo popula- 

In both of the crosses described so far, 
the laevis rDNA was donated by the fe- 
male parent, which also donated the cyto- 
plasm of the zygote. We therefore tested 
animals from the reverse cross, mulleri 
(m/m) X laevis (l/l) to assess the influ- 



i 30h 

°= 20 



/ — @— 

l/l x m/m / 





□ ,' l/l x m/m 

i i i i i 

20 40 60 80 100 
Time after fertilization (hr ) 


10 1519 26 32 35 
_u l L 



40 42 4445 46 

_l I LJ I 


^ t 

Gastrula Neurula | Hatching Feeding 
Heart beat 

Nieuwkoop-Faber stages 

Fig. 13. Synthesis of mulleri rRNA during 
early development in Xenopus hybrids. Nuclei 
were isolated from embryos at various stages 
and analyzed for their ability to synthesize 
rRNA as described in the text. Embryos from 
the same mating have the same symbol. Open 
symbols with dotted line, embryos from the 
cross lll~ X m/m. Solid symbols with solid line, 
embryos from the cross l/l X m/m. Data for 
this cross were taken from Table 4. 

ence of egg cytoplasm on rDNA expres- 
sion. The results were essentially the 
same regardless of which species was the 
female parent. 

In one experiment, we also analyzed 
embryos from the cross mulleri (m/m) X 
laevis (l/l~). As early as stage 15-20 
these embryos were making a significant 
amount of mulleri rRNA (30%) in con- 
trast to the reverse cross shown in Fig. 
13. The embryos from this cross are a 
mixture of two genotypes, m/l and m/l". 
Since at the same stage (16-19) the m/l 
genotype embryos make no mulleri 
rRNA, it is clear that the m/l' embryos 
are engaged in active synthesis of mulleri 

The results of these matings can be 
summarized as follows. When the hybrid 
receives both laevis and mulleri rDNA 
(l/m and m/l) transcription of the laevis 
rDNA is usually dominant and the mul- 
leri rDNA is repressed. Repression of 
mulleri rDNA occurs as early as rRNA 
synthesis can be measured in embryos 
(stage 15) . In most cases the repression 
extends throughout embryogenesis and is 
found in adult tissues as well. When 
laevis rDNA is present, the maternal 
cytoplasm has little effect on the severity 
of the repression. 

When the hybrid receives only mulleri 
rDNA and laevis rDNA is deleted, two 
different results can occur. (1) If the 
zygote receives laevis ooplasm (hybrid 
genotype l~./m) } mulleri rRNA synthesis 
is transiently repressed and then is grad- 
ually turned on. (2) If the zygote re- 
ceives mulleri ooplasm (hybrid genotype 
m/l~) , turn-on of mulleri rRNA synthesis 
is not delayed. 

These results suggest that mulleri 
rRNA synthesis can be repressed in two 
apparently different ways. One type of 
repression is due to the presence of laevis 
rDNA, is not maternally inherited, and 
can be permanent. The second type of 
repression is due to some substance in the 
laevis cytoplasm, is maternally inherited, 
and is transient. 

Our biochemical measurements on hy- 



brid embryos are in agreement with the disappearance of one nucleolus in the hy- 

cytological observations by Blackler and brid cells. Synthesis of both species of 

co-workers. In general, repression of rRNA correlates with the appearance of 

mulleri rDNA correlates well with the two nucleoli per cell. 




P. J . Stambrook, assisted by D. Somerville and B. Smith 

The S phase, during which nuclear 
DNA is replicated, occupies a discrete 
interval of the cell cycle. Experiments 
were initiated to examine the time of 
replication of specific genes within the 
S phase. The genes investigated in this 
study were those which code for ribo- 
somal 18S and 28S RNA, and were 
selected for two reasons: (1) In eukary- 
otic cells, ribosomal genes are repre- 
sented by multiple copies and this facili- 
tates their identification by RNA-DNA 

hybridization. (2) Ribosomal genes from 
Xenopus laevis, which have previously 
been isolated and purified, can be used 
as a template for the in vitro synthesis 
of radioactive RNA with very high 
specific activity that is complementary 
to ribosomal genes (cRNA). Since re- 
gions of the ribosomal genes which code 
for 18S and 28S ribosomal RNA have 
been highly conserved through evolution, 
the X. laevis ribosomal cRNA can be 
used as a very sensitive probe for detec- 











2.0 3.0 


Fig. 14. Competition of X. laevis ribosomal cRNA with Chinese hamster ribosomal RNA. Nitro- 
cellulose filters containing 50 fig of Chinese hamster DNA were incubated with X. laevis 3 H ribo- 
somal cRNA (120,000 cpm/ml) in the presence of Chinese hamster ribosomal RNA. 



tion of ribosomal genes in heterologous 
systems, such as the Chinese hamster 
cells used in these experiments. 

The specificity of hybridization be- 
tween X. laevis ribosomal cRNA and 
Chinese hamster DNA was established 
by two criteria. Hybridization of the 
cRNA was performed in the presence of 
increasing amounts of unlabeled ribo- 
somal RNA from Chinese hamster cells 
which acted as an effective competitor 
of the cRNA (Fig. 14). When the 
cRNA was hybridized to Chinese ham- 
ster DNA which had been fractionated 
by CsCl equilibrium density centrifuga- 
tion, the cRNA hybridized with DNA of 
high G-C content, characteristic of ribo- 
somal genes, in the dense region of the 
gradient (Fig. 16). 


50 - 

TIME (hrs.) 

Fig. 15. Cell synchrony during the mitotic 
cycle. Mitotic cells were collected by selective 
detachment and seeded into a series of plastic 
60-mm culture plates. The cells were labeled for 
10 minutes with 3 H thymidine (5 /Wml) at 
intervals after synchronization and coated with 
emulsion for autoradiography. Percentage of 
labeled cells, open circles; mitotic index, dashed 



500 _ 







15 20 25 




Fig. 16. Hybridization of X. laevis ribosomal cRNA to BrdUra-substituted Chinese hamster 
DNA from different intervals of the S phase. Synchronized cells were labeled with BrdUra (10~ 5 
M ) during successive 2-hour intervals of the S phase. The DNA was extracted, banded in CsCl, 
denatured, and immobilized on filters. Filters from all 4 gradients were combined in a single vial 
and hybridized with X. laevis ribosomal cRNA (250,000 cpm/ml). Absorbance at 260 nm (open 
circles with solid line); radioactivity (dashed line). 



A population of mitotic cells was col- 
lected and their synchronous traverse 
through the cell cycle was followed. The 
degree of synchrony and the time of 
DNA synthesis in each experiment was 
monitored by pulse-labeling replicate 
plates for 10 minutes with 3 H-thymidine 
at intervals after synchronization and 
then determining by autoradiography 
the percentage of cells incorporating 3 H 
thymidine at each interval (Fig. 15). 
The percentage of cells in mitosis at the 
start of each experiment, usually 95 to 
98%, was always recorded, as was the 
mitotic index at each time point after 
synchronization. Cells were density 
labeled with 5-bromodeoxyuridine 
(BrdUrd) for 2-hour intervals extending 
over the duration of the S phase so as to 
separate DNA replicated within discrete 
periods of the S phase from DNA which 
has replicated or will replicate at an 
earlier or later time. BrdUrd, which is 
incorporated into DNA in place of thy- 
midine, increases the density of newly 
replicated DNA which appears as a 
heavy peak after CsCl equilibrium den- 
sity centrifugation and which is easily 

TABLE 6. Replication of DNA Containing 

Ribosomal Genes at Intervals during the 

S Phase 









300 J 


200 => 


100 5 




6 8 

TIME (hrs. 


Percent Radio- 





Percent DNA 

to Newly 



Replicated DNA 













Synchronized cells were density labeled with 
BrdUrd at 2-hour intervals during the S phase. 
The DNA was extracted, banded in CsCl, 
immobilized on filters and hybridized with X . 
laevis ribosomal cRNA. The percentage of the 
DNA replicated at each interval was calculated 
from the O.D. profiles in Fig. 16, and the per- 
centage of the radioactivity associated with 
newly replicated DNA was determined from 
the same figure. 

Fig. 17. Replication of ribosomal genes dur- 
ing the S phase. The relative specific activities of 
newly replicated DNA at different times during 
the S phase after hybridization with X. laevis 
3 H ribosomal cRNA. Percentage of labeled cells 
(open circles). 

separable from bulk DNA. DNA ex- 
tracted from cells labeled with BrdUrd 
from 3-5 hours, 5—7 hours, 7—9 hours 
and 9-11 hours was subjected to iso- 
pycnic centrifugation in CsCl. The gradi- 
ent was fractionated and the DNA in 
each fraction was denatured, trapped on 
nitrocellulose filters, and hybridized with 
A", laevis ribosomal cRNA. 

The relative amount of DNA which is 
replicated during each time interval 
varies (Fig. 16) and is presented in Table 
6. The proportion of ribosomal genes 
which replicate during each interval is 
reflected by the fraction of the radioac- 
tivity which hybridizes to DNA in the 
heavy peak (Fig. 16) . By the time repli- 
cation of the first 23% of the DNA has 
occurred (3-5 hours), almost half of 
the ribosomal genes have replicated (Fig. 
16 and Table 6). Most of the remaining 
ribosomal DNA is replicated during the 
next time interval (5-7 hours), indicat- 
ing that the multiple ribosomal gene 
copies (between 200 and 300) replicate 
during a restricted period of the S phase. 

The results are summarized in Fig. 17. 
The width of each bar represents the time 
interval during which cells were labeled 
with BrdUrd, and their relative height 
indicates the proportion of newly repli- 
cated DNA in ribosomal genes. 








P. K. Wellauer and I. B. Dawid 

We observed that 28S rRNA from 
HeLa cells showed a reproducible sec- 
ondary structure of hairpin loops when 
spread for electron microscopy from 80% 
formamide and 4 M urea. This second- 
ary structure allowed the mapping of 
these molecules in the sense that different 
regions of each molecule and their polar- 
ity could be recognized. Therefore, we 
used such secondary structure maps to 
investigate the processing pathway of 
HeLa cell rRNA. Many aspects of this 
processing pathway have been character- 
ized previously by the work of Darnell, 
Penman, Weinberg, Maden, and others. 

Figure 18a shows an electron micro- 
graph of the 45S rRNA precursor, which 
is known to be the primary transcrip- 
tion product of the ribosomal genes in 
HeLa. A number of loops can be seen 
along the molecule. Starting from the 
left end (which is the 5' end, as we shall 
show below) we see an extended region 

leading to a characteristic double loop. 
An extended region follows, then a series 
of multiple loops, and a longer extended 
region. The other end (3' end) of the 
molecule is occupied by a large cluster 
of loops. We compare to this 45S mole- 
cule the 28S rRNA molecule shown in 
Figure 18b. It starts with an extended 
region, followed by the characteristic 
double loop, another extended region, 
and finally several small loops. This 
series of structures can be found in the 
45S RNA molecule in one region only, 
at its 5' end, i.e., the left end as shown 
in the figure. We conclude that the 28S 
RNA must be derived from the 5'-ter- 
minal region of the 45S RNA. The 18S 
RNA molecule (Fig. 18c) shows no sec- 
ondary structure when examined under 
these conditions. There is only one ex- 
tended region of sufficient length in 45S 
RNA to accommodate the 18S molecule 

Fig. 18. Electron micrographs of HeLa cell rRNA molecules, (a) 45S RNA; (b) 28S RNA; (c) 
18S RNA. 



in the right half of the molecule as 

The conclusions regarding the location 
of the 28S and 18S regions in the 45S 
RNA were supported by measuring and 
plotting extended and looped regions in 
40 to 60 separate molecules of each of 
these RNAs. Secondary structure maps 
also were prepared for a number of in- 
termediates in the processing pathway. 
Furthermore, the molecular weights of 
precursor, intermediate, and mature 
molecules were measured by comparing 
their length to the length of E. coli 
rRNA (Table 7). 

To determine the 5'-to-3' polarity of 
the rRNA molecules, we studied the 
structure of 45S and 28S rRNA partially 
digested with the 3'-exonuclease from 
ascites nucleoli. This enzyme was gener- 
ously supplied by Dr. Robert Perry. 

TABLE 7. Molecular Weight of HeLa rRNA 

Molecules as Determined by Electron 




Mol. Wt. ± S.D. X 10- 6 , 
number of molecules 


0.661 ± 0.07 (166) 


1.757 ± 0.14 (100) 


2.266 ± 0.20 ( 95) 


3.301 ± 0.33 ( 87) 


4.714 d= 0.45 (100) 

The molecular weights were determined from 
length measurements of the RNA after treat- 
ment with 10% formaldehyde (3% for 18S RNA) 
at 63° for 15 minutes, followed by spreading 
from formamide-urea. E. coli rRNA was used 
as internal standard. 







24 S 


Cx\\\VV\\VV3 ( 







Fig. 19. Processing pathway of HeLa rRNA. 
The molecules are drawn to scale. The 28S and 
18S regions are hatched. 

Secondary structure maps of the partially 
digested molecules clearly showed loss 
of material at one end, which was there- 
by identified as the 3' end. These ex- 
periments placed the 28S RNA region 
at the 5' end of the 45S RNA molecule. 

From a consideration of secondary 
structure maps and the lengths (or molec- 
ular weights) of these molecules an un- 
ambiguous processing pathway was de- 
rived (Fig. 19). This model is in good 
agreement with the conclusions from 
earlier work but adds considerable de- 

The experiments described above dem- 
onstrate the usefulness of secondary 
structure mapping of RNA for the study 
of the arrangement of sequences in 
RNAs with a precursor-product relation- 
ship. The method is also likely to be 
useful for evolutionary studies: we have 
analyzed the secondary structure in 
Xenopus 28S rRNA and found it to be 
very similar to that of HeLa 28S RNA, 
even though the X en opus molecule is 
about 15% smaller. 


/. B. Dawid, I. Horak, C '. Kaushagen, D. E. Leister, and P. K. Wellauer 

Our studies on mitochondrial bio- 
genesis (Year Book 71, p. 30) have been 
continued in several directions. One 
aspect is the study of hybrid somatic 
cells in collaboration with H. G. Coon 
at The National Cancer Institute, which 
aims to develop an approach for the 
application of genetic tools to the study 

of mitochondrial biogenesis in animal 
cells. The other aspect of our work in- 
cludes several projects on the character- 
ization of various macromolecular com- 
ponents of mitochondria which are likely 
to be involved in the biogenesis of these 
particles. We have investigated the com- 
position of mitochondrial ribosomes, the 



properties of mitochondrial RNA poly- 
merase and the interaction of this enzyme 
with mitochondrial DNA (mtDNA) , and 
the properties of mitochondrial RNA 
molecules and the arrangement and 
evolution of the genes on mtDNA which 
code for some of these RNAs. 

Recombination of Mitochondrial 
DNA in Somatic Hybrid Cells 

/. Horak and I. B. Dawid, in collaboration with 
H. G. Coon* 

We reported in Year Book 71, p. 24, 
that hybrid cells of human and mouse 

* Laboratory of Cell Biology, The National 
Cancer Institute. 

or human and rat origin contained mito- 
chondrial DNA (mtDNA) sequences 
from both parental species. Since that 
time, additional examples were studied, 
confirming our conclusion that mtDNA 
from two animal species can replicate 
in the same cell for extended periods of 

One question concerning the state of 
mtDNA in hybrid cells asks whether the 
molecules derived from the two different 
species coexist separately in the same 
cytoplasm or whether they interact with 
each other. One possible interaction 
would be recombination, an event in 
which parts of, say, human mtDNA 
would attach covalently to mouse 








50 60 

Fraction number 

Fig. 20. Separation of human and mouse mtDNA by banding in a CsCl gradient. Open circles, 
3 H; solid circles (human cRNA) ; 32 P (mouse cRNA). For explanation see text. 









30 40 

Fraction number 

Fig. 21. CsCl gradient centrifugation of the DNA from a hybrid cell strain. Symbols have the 
same meaning as in Fig. 20. 

mtDNA, forming a hybrid molecule. We 
have tested this possibility in the follow- 
ing way. 

The density of human mtDNA is 1.707, 
that of. mouse mtDNA 1.699, and rat 
mtDNA 1.701 g/cm 3 (assuming the den- 
sity of E. coli DNA as 1.710). This 
density difference can be exploited to 
separate the various mtDNA in CsCl 
gradients. Figure 20 shows a gradient in 
which a mixture of human and mouse 
mtDNAs was banded; the gradient was 
fractionated and each fraction adsorbed 
to a membrane filter. The entire series 
of filters was then hybridized with a 
mixture of human 3 H-cRNA (that is, 
( 3 H)-RNA transcribed in vitro with E. 
coli RNA polymerase from human 
mtDNA) and mouse 32 P-cRNA. Each 
cRNA hybridized with its own template 
DNA and gave very little cross-hybridi- 
zation with the other mtDNA. Gradi- 

ents of this type were used to assay for 
recombinant molecules of mtDNA in hy- 
brid cells. Figure 21 shows such a gradi- 
ent with DNA from a cell strain that 
contained about 20% mouse mtDNA 
(and 80% human mtDNA) . The human 
3 H-cRNA detected a broad peak of 
human mtDNA sequences. The mouse 
32 P-cRNA hybridized with DNA mole- 
cules which had banded at the same 
density as the molecules detected by the 
human cRNA; in addition, there were 
DNA molecules containing mouse se- 
quences which had banded at a lower 
density. This result strongly suggests 
that a physical link existed between at 
least some of the mouse and human 
mtDNA sequences. An interpretation of 
this gradient implies that the denser 
peak consists of molecules which are 
predominantly human mtDNA and con- 
tain pieces of mouse mtDNA. These 



molecules would band at or close to 
human mtDNA density, but they con- 
tain sequences which hybridize with 
mouse cRNA. In addition, the dense peak 
may also contain "clean" human mtDNA 
molecules. The lower-density peak which 
binds mouse cRNA could consist of 
"clean" mouse mtDNA molecules or of 
hybrid molecules containing more mouse 
than human sequences. This question 
cannot be decided. Human cRNA does 
bind to DNA molecules which have 
banded at the lower density, but it is 
not clear whether this is due to "smear- 
ing" of the large peak of human mtDNA 
or to the presence of recombinant mole- 
cules with a density close to that of 
mouse mtDNA. 

The experiment in Fig. 21 shows that 
the DNA molecules containing mouse 
sequences, and possibly those containing 
human sequences, are heterogeneous in 
the hybrid cell. Furthermore, the ex- 
periment makes it very likely that mouse 
mtDNA segments became covalently 
linked to human mtDNA during growth 
in the hybrid cell strain. However, the 
latter conclusion is not explicitly proven: 
Other interpretations of the observation 
are possible, e.g., the presence of con- 
catemers or of mouse mtDNA molecules 
modified in such a way as to make their 
density similar to that of the human 
mtDNA. These alternate interpretations 
are much less likely than recombination 
of human and mouse mtDNAs. 

Similar experiments were carried out 
with several mouse-human and rat- 
human hybrid cell strains. We have 
found five examples of cell strains which 
contained both parental mtDNAs with- 
out detectable recombination and 13 cell 
strains in which recombination had ap- 
parently taken place. Therefore, recom- 
bination of mtDNA in hybrid cells is a 
frequent event. These results suggest 
that recombination of mtDNA molecules 
may occur frequently in the "normal" 
cells of animals but would not be de- 
tectable by biochemical analysis. 

Protein Composition of Cytoplasmic 
and Mitochondrial Ribosomes of 

Xenopus laevis 

D. E. Leister 

In Year Book 71, p. 29, we reported the 
fractionation of Xenopus laevis cyto- 
plasmic ribosomal proteins by two- 
dimensional (2D) polyacrylamide gel 
electrophoresis. The large number of 
spots (at least 60) detected on 2D gel 
slabs indicated that greater resolution 
would be obtained if ribosomal subunits, 
rather than monosomes, were the source 
of proteins. Such analyses have now been 
carried out with both cytoplasmic and 
mitochondrial ribosomal subunits. 

Similar strategies were used to isolate 
cytoplasmic and mitochondrial ribosomal 
subunits. First, undissociated monosomes 
were isolated. Then, the monosomes were 
dissociated into their constituent sub- 
units by exposure to buffers containing 
high concentrations of monovalent salts; 
subunits were subsequently separated by 
centrifugation in sucrose gradients. Iso- 
lated cytoplasmic large and small sub- 
units were largely free of cross-contami- 
nation and, when mixed, were as active 
as undissociated monosomes in polyU- 
directed polypeptide synthesis. Isolated 
mitochondrial large and small subunits 
were completely free of contamination by 
cytoplasmic ribosomes (RNA extracted 
from the mitochondrial particles con- 
tained less than 2% cytoplasmic ribo- 
somal RNA, as measured by hybridiza- 
tion competition) . In addition, these 
subunits were free of cross-contamination 
and, when mixed, were as active as undis- 
sociated mitochondrial monosomes in 
polypeptide synthesis. 

The composition and other properties 
of the ribosomal subunits were studied by 
physical and chemical means, in particu- 
lar by the separation of the proteins by 
two-dimensional gel electrophoresis. 
Table 8 summarizes the results. The 
mitochondrial rRNA molecules are much 
smaller than the cytoplasmic rRNAs, but 



TABLE 8. Composition of Cytoplasmic and Mitochondrial Ribosomes 

of X. laevis 

Ribosomal Subunit 

Mean Mol. 
Number of Wt. of RNA, mol. Particle Wt., 

Proteins Proteins wt X 10~ 6 10 6 daltons 























the mitochondrial ribosomal subunits 
contain a larger number and a larger 
total mass of ribosomal proteins. Con- 
sequently, the particle weight of the 
mitochondrial ribosome is only slightly 
lower than the weight of the cytoplasmic 
ribosome. This conclusion is unexpected 
in view of the low sedimentation coeffi- 
cient of the mitochondrial ribosome 
( Year Book 69, p. 576) . It appears on the 
basis of the data in Table 1 that this low 
sedimentation coefficient is related more 
to the lower density of the protein-rich 
mitrochondrial ribosomes than to their 
slightly lower mass. 

Figure 22 shows the molecular weight 
histogram of mitochondrial ribosomal 
proteins. Comparison to histograms of 
X. laevis cytoplasmic and E. coli ribo- 
somal proteins {Year Book 71, p. 30) re- 
veals that the difference in the mean mo- 
lecular weights of mitochondrial and 
cytoplasmic proteins is not the result of a 
bias introduced by a few large mitochon- 
drial proteins. On the contrary, all three 
patterns are distinct from each other. 




^ 2 

o c _ 





MW x 10 


Fig. 22. Molecular weight distribution of 
mitochondrial ribosomal proteins from X. laevis. 

In all physical and chemical properties 
known so far, including the size and com- 
position of rRNA and the number and 
molecular weight distribution of proteins, 
mitochondrial ribosomes are clearly dis- 
tinct from both bacterial and cytoplasmic 
ribosomes. Therefore, these studies nei- 
ther support nor speak against any of the 
current hypotheses regarding the evolu- 
tionary origin of mitochondria. 

Mitochondrial RNA of Drosophila 
C. Kaushagen 

Mitochondrial ribosomal RNAs (mt- 
rRNAs) from Xenopus laevis, mouse, 
and humans have been characterized, but 
the detailed characterization of mtRNA 
from an invertebrate has not been re- 
ported. We have initiated a study of 
mtRNA of Drosophila melanog aster. 

Mitochondria were isolated from adult 
flies and the RNA extracted by methods 
very similar to those used for the isola- 
tion of Xenopus mtRNA. The nucleic 
acids were analyzed on 2.4% polyacryl- 
amide gels at room temperature (Fig. 
23). Peaks corresponding to the 28S and 
18S cytoplasmic rRNA were present, as 
well as two peaks with mobilities similar 
to those of E. coli rRNA. When similar 
gels were run at 10°C (Fig. 23), the mo- 
bilities of the two latter peaks were in- 
creased such that both ran faster than 
18S. The mobilities at 10 °C correspond 



Fig. 23. Separation of nucleic acids from a 
mitochondrial preparation of D. melanog aster. 
The temperature of gel electrophoresis in 2.4% 
polyacrylamide gels was room temperature 
(top) and 10° (bottom). Lg and Sm refer to 
the large and small mt-rRNAs, respectively. 

to molecular weights of about 0.55 X 10 6 
and 0.3 X 10 6 MW, respectively. Be- 
cause of the location in the mitochondrial 
fraction and the variability of electro- 
phoretic mobility with temperature, these 
two RNA species were tentatively identi- 

10 15 20 25 




Fig. 24. Sucrose gradient sedimentation of 
mitochondrial RNA of D. melanogaster. Direc- 
tion of sedimentation is to the left. 

fled as the large and the small mitochon- 
drial rRNAs of Drosophila. 

Several preparations of Drosophila 
mtRNA were run in sucrose-SDS gradi- 
ents (Fig. 24). Two RNA components 
occur which sedimented at 13S and 10.5S, 
in addition to 28S and 18S rRNA, which 
contaminated the preparation. These 
gradients were fractionated and peak 
fractions pooled to separate the two 
mtRNA species. Samples of these pools 
were analyzed by gel electrophoresis ; the 
13S peak contained the large mt-rRNA; 
the 10. 5S peak, the small mt-rRNA. 

Aliquots of each of the two mtRNA 
species so separated on a sucrose-SDS 
gradient were spread by the Kleinschmidt 
technique and analyzed by electron mi- 
croscopy. Molecular weights of 560,000 
and 280,000, respectively, were deter- 
mined for the two mt-rRNA molecules 
(Table 9) . These values are similar to 
those measured for HeLa mtRNA by 
Robberson and his colleagues and are 
slightly but significantly smaller than the 
molecular weights of X. laevis mtRNA. 

Drosophila mt-rRNA is similar in size 
to vertebrate mt-rRNA and also shares 
its property of variable electrophoretic 
mobility depending on temperature and 
gel composition. Drosophila mt-rRNA 
appears even more variable in this re- 
spect than Xenopus RNA. Furthermore, 
the sedimentation rate of the Drosophila 
RNAs is also unusual: the Xenopus mt- 
rRNAs, which are only slightly larger, 
sediment at 13 and 17S, and the HeLa 
mt-rRNAs sediment at 12 and 16S. The 
Drosophila RNA sedimentation coeffi- 
cients of 10.5 and 13S are distinctly lower 
and suggest that these molecules assume 
an unfolded configuration which slows 
their sedimentation rate. We suggested 
previously that Xenopus mt-rRNA has a 
more unfolded structure than cytoplasmic 
or bacterial rRNAs (Year Book 71, p. 
30) . Drosophila mt-rRNA appears to be 
a more extreme example of such an un- 
folded structure. 



Measurement of Mitochondrial RNA 

and DNA-RNA Hybrids in the 

Electron Microscope 

P. K. Wellauer and I. B. Dawid, with the 
assistance of M . Rebbert 

Major products of mitochondrial DNA 
(mtDNA) are the two RNA components 
of the mitochondrial ribosome, the mt- 
rRNAs (see Year Book 68, p. 514; Year 
Book 69, p. 575; Year Book 70, p. 42). 
The physical properties of these mole- 
cules from ovaries of Xenopus laevis have 
been characterized to some extent. We 
have continued these studies in the past 
year using electron microscopic tech- 

When different RNA molecules are 
spread by a modified Kleinschmidt tech- 
nique from 80% formamide and 4 M 
urea, most types of RNA appear as fully 
extended strands (for important excep- 
tions, see below) . The length of RNAs of 
unknown size can be compared to an 
internal standard, i.e., a different RNA of 
known molecular weight. This measure- 
ment allows the accurate determination 
of molecular weights of various RNA 

We have measured the length of Xen- 
opus mt-rRNAs under these conditions 
and calculated their molecular weights, 
using E. coli rRNA and HeLa cell 18S 
rRNA as standards. The results (Table 
9) confirm the previous finding that the 
mt-rRNAs are the smallest ribosomal 
RNAs known. However, the values ob- 
tained are larger by about 15% than the 
molecular weights determined previously 
by sedimentation (Year Book 70, p. 43). 

Table 9 also shows the molecular 
weights determined by electron micros- 
copy of the mt-rRNAs from Drosophila 
melanog aster. These RNAs are slightly 
smaller than the corresponding molecules 
from Xenopus and are very similar in size 
to the mt-rRNAs from HeLa cells which 
were analyzed by the same technique by 
Robberson and his colleagues. 

Each circular molecule of mtDNA car- 

TABLE 9. Molecular Weight of Mitochondrial 

rRNA Molecules as Determined by 

Electron Microscopy 


Moi. wt, x i(r 5 ± 

S. D., number 
of molecules 

X. laevis 

large mt-rRNA 
small mt-rRNA 

D. melanogaster 
large mt-rRNA 
small mt-rRNA 

6.15 =fc 0.74 (318) 

3.38 d= 0.47 (289) 

5.22 ± 0.81 (154) 

2.98 ± 0.53 (153) 

E. coli 23S rRNA was used as internal stand- 
ard. Its molecular weight was taken as 1.07 X 
10 6 . 

ries one region, or gene, coding for the 
small and the large mt-rRNA {Year 
Book 69, p. 576) . The arrangement of the 
two gene regions with respect to each 
other was investigated. A similar study 
had been done previously by Robberson, 
Aloni, Attardi, and Davidson on HeLa 
cell mtDNA. These authors found that 
the two rRNA genes are next to each 
other, separated by a small gap. We ob- 
tained similar results with Xenopus 
mtDNA and RNA in the following way. 
The two strands of mtDNA were sepa- 
rated with the aid of polyUG by a 
method introduced by Szybalski. The H- 
strand, which is known to carry the 
rRNA genes, was then annealed with mt- 
rRNA and the hybrid examined in the 
electron microscope. When both rRNAs 
were annealed with the DNA, two double- 
stranded regions were seen which we in- 
terpret as DNA-rRNA hybrids. These 
regions were separated by a small gap. 
The length of each region and of the gap 
was measured and the results are pre- 
sented as a histogram in Fig. 25. The 
sharp distributions of double-stranded re- 
gions correspond to the hybrid of the H- 
strand with the small and the large mt- 
rRNA, respectively. The length ratio of 
these two regions is very similar to the 
length ratio of the two rRNAs measured 
individually (Table 9) . Exact molecular 



0.2 0.4 

LENGTH (/xm ) 

Fig. 25. Length of the duplex region of the 
H-strand of X. laevis mtDNA with the large 
and small mt-rRNA and of the gap which sepa- 
rates the two heteroduplex regions. The length 
of the gap is 0.046 fxm. and has been converted 
into number of nucleotides. The length of the 
duplex with the large mt-rRNA is 0.53 ^m, and 
the length of the small rRNA is 0.30 /mi. 

weights could not be calculated from 
these lengths, since the linear density of 
RNA-DNA hybrids is not accurately 
known. The length of the gap between 
the two rRNAs could be translated into 
numbers of nucleotides by the use of an 
internal standard of single-stranded 
DNA (<f> X 174 DNA) . The length of the 
gap of 120 nucleotides (Fig. 25) is very 
similar to the corresponding gap between 
the mt-rRNA genes of HeLa cells. 

It is clear from preliminary experi- 
ments (Year Book 71, p. 22) that the nu- 
cleotide sequences in mtDNA of animals 
of different species vary extensively. 
Relatively little homology exists between 
the mtDNAs of animals from different 
classes of vertebrates. Nevertheless, the 
arrangement of the rRNA genes on the 
mtDNA has been conserved for a long 
period of evolution. This fact suggests 
that this arrangement has important 
functional implications. 

Transcription of Mitochondrial DNA 
by Mitochondrial RNA Polymerase 

/. B. Dawid and G.-J . Wu* with the assistance 
of M. Rebbert 

In Year Book 71, pp. 25-29, we re- 
ported on the isolation and some proper- 
ties of RNA polymerase from ovarian 
mitochondria of X. laevis. The transcrip- 
tion of mtDNA by this polymerase has 
been studied further in the past year. 

We observed previously that nascent 
RNA chains were associated with the 
template in such a way that they traveled 
together in SDS-sucrose gradients (Year 
Book 71, p. 28). When the template was 
digested with DNase or when the com- 
plex was heated, the RNA was released. 
These facts suggested that an RNA-DNA 
hybrid may exist between the nascent 
RNA (or part of it) and the template. 
This interpretation has now been sup- 
ported by the finding that a fraction of 
the nascent RNA, amounting to 25 to 
30%, is resistant to digestion with 
RNase; this material becomes sensitive 
to RNase when the DNA is digested or 
when the complex is dissociated by heat- 

We investigated the size of the RNase- 
resistant region in nascent RNA. For this 
purpose the complex was digested with 
RNase, the RNase destroyed with pro- 
nase, and the resistant RNA released by 
digesting the DNA with DNase. This 
resistant RNA was then characterized by 
chromatography on Sephadex G-100 and 
by gel electrophoresis. The resistant 
RNA was very heterogeneous in size, from 
short fragments of about 50 nucleotides 
up to long chains of about 500 nucleotides. 
These results suggest that mitochondrial 
RNA polymerase "opens" the DNA helix 
during transcription in vitro in such a 
way that a duplex is formed between the 
template strand of the DNA and the 
nascent RNA. 

* Now at the Department of Biological Sci- 
ences, Columbia University. 



We also investigated the problem of 
initiation of RNA chains by mitochon- 
drial polymerase on the mtDNA tem- 
plate. This was done with nucleoside tri- 
phosphates labeled in the gamma-position 
with 32 P. It is known that the 5'-terminal 
nucleotide in any RNA chain retains its 
gamma-phosphate while all internal nu- 
cleotides lose it. By assaying the incor- 
poration of 32 P-phosphate from each of 
the four nucleoside triphosphates we 
found that mitochondrial RNA poly- 
merase initiates predominantly with ATP 
and at a frequency about two times lower 
with GTP. Initiation with CTP to UTP 
is very low or absent. This result is in 
agreement with previous studies on other 
RNA polymerases which also showed 
initiation with purine nucleotides only. 

We wondered whether initiation with 
ATP and GTP each takes place at a par- 
ticular site of the DNA or whether initia- 
tion could take place at many different 
sites. To answer this question, RNA was 

synthesized with gamma-labeled ATP 
(or GTP), the product digested with 
RNase, and the resulting oligonucleotides 
separated on DEAE-Sephadex columns. 
If initiation had occurred on a single site, 
only a single oligonucleotide should have 
arisen from the terminal region of the 
RNA chain. We found, however, several 
oligonucleotides labeled with 32 P, both 
when ATP and when GTP was used as a 
terminal label. These results suggest that 
chain initiation of mitochondrial RNA 
polymerase on mtDNA occurs in vitro on 
several different sites. 

It is not known how many initiation 
sites exist in vivo for mitochondrial RNA, 
but one might expect that there would be 
only a few. If this is true, we would have 
to conclude that mitochondrial polymer- 
ase in vitro does not faithfully reproduce 
the in vivo transcription of mtDNA. 
Faithful in vitro transcription has not 
been obtained so far with any system de- 
rived from animal cells. 



R. E. Pagano 


The fundamental problem facing the 
membrane biologist is how to unravel the 
relationships between the molecular 
structure of the membrane and its varied 
functions and properties. In order to ex- 
plore some of these relationships in bio- 
logical membranes, we have undertaken 
the development of several physical and 
chemical techniques for examining some 
of the surface properties of natural mem- 
branes. Initially this work will empha- 
size the use of phospholipid model mem- 
branes, both as an analytical tool for 
studying natural membranes and as a test 
system for some of the physical tech- 
niques under development. 

Interactions between Lipid Bilayers 
and Cell Surface Membranes 

Since numerous cellular properties 
(e.g., adhesion, vesicle formation, endo- 

cytosis, and secretion) depend upon the 
phenomena of adhesion and coalescence 
between two continuous membranes, un- 
derstanding the interactions between 
membranes in terms of their molecular 
composition and possible structural varia- 
tions is a problem of great interest to the 
developmental and the cell biologist. As 
one approach to this problem, we have 
begun to look at the interaction of arti- 
ficially generated lipid membranes of bi- 
molecular dimension with each other and 
with the surface membranes of cells 
grown in culture. A preliminary finding, 
and basis for this approach, is outlined 
schematically in Fig. 26. Chinese ham- 
ster V79 cells were cultured on nylon 
fibers strung across petri dishes, using a 
modification of the technique described 
by Edelman and co-workers. The culture 
medium was then exchanged with a sim- 
ple balanced salt solution, and a lipid bi- 



nylon fiber 

*////////////////////////* arched ?/////////////////?7?///A 





Fig. 26. Schematic representation of the interaction between a lipid bilayer vesicle and a Chi- 
nese hamster cell grown on a nylon fiber, (a) Bilayer vesicle is manipulated up to a single cell by 
means of a micropipette. (b) Several seconds after contact the cell was observed to detach from 
the nylon fiber and spread over the surface of the lipid vesicle. 

layer vesicle, generated according to 
standard techniques, was manipulated up 
to a single cell (Fig. 26a). Several sec- 
onds after contact, the cell'was observed 
to spread over the surface of the bilayer 
and spontaneously detach itself from the 
nylon fiber (Fig. 26b). Since the molec- 
ular composition of the lipid vesicle is 
subject to direct experimental control, 
and since the distribution of certain of 
the macromolecules within the cell sur- 
face can be varied by derivatization of 
the fiber to which the cells are bound, it 
should be possible to gain some insights 
into some of the factors responsible for 
cell recognition, adhesion, and possibly 
membrane fusion. 

Elucidation of the structure of the 
junction which forms between the bilayer 
vesicle and cell membrane, and the 
mechanism by which it forms, must await 
careful ultrastructural studies. It is ap- 
parent, however, that the attractive forces 
in this system are considerable, and it is 
possible that a partial fusion or coales- 
cence of the membranes may exist over 
the region of contact. This notion gains 
some support from the early studies of 

Chambers and Kopac on the spontaneous 
coalescence of oil droplets with the sur- 
face membranes of various echinoderm 

If the junction that forms does involve 
a partial fusion of the two membranes, or 
if the junction is a relatively low resist- 
ance one compared to that of the bilayer 
vesicle (approx. 10 10 ohm-cm 2 ), then 
several new kinds of electrophysiological 
and biochemical investigations of the cell 
membrane may be carried out because of 
the unusual intrinsic properties of the bi- 
layer. One of the most exciting possibil- 
ities will be to look for and study the 
electrical properties of individual con- 
ducting channels, which by virtue of the 
low ionic permeability of the bilayer 
vesicle should be detectable if such chan- 
nels are free to diffuse across the junc- 
tional complex that forms in this system. 
Such experiments have already been car- 
ried out in lipid bilayers modified by the 
addition of chemically well-defined 
ionophores, and the statistical properties 
of the resulting conducting channels have 
been used to investigate and delineate the 
possible mechanisms of ion permeability 



through membranes. A second, and very 
intriguing idea, is that the generation of 
such hybrid membranes as described here 
could form the basis of a procedure for 
the isolation and purification of mem- 
brane proteins which obviates the use of 
detergents or organic solvents. This tech- 
nique would have the advantage that the 
protein components never leave the mem- 
brane phase but would be diluted in two 
dimensions as the hybrid is formed, be- 
cause of the fluid nature of the mem- 
branes. Final purification of the mem- 
brane components would take place by 
chromatographic methods which have 
already been successfully applied to lipid 
dispersions and natural membrane prep- 

Viscoelastic Properties and Surface 
Tension of Membranes 

When a fluid drop is placed in a liquid 
of higher density contained in a rotating 
horizontal tube, it will tend to find an 
equilibrium position on the axis of rota- 
tion because of the pressure caused by 
the centrifugal force. As the speed of 
rotation is increased, the droplet will 
elongate along the axis until finally it is 
in the form of a long narrow cylinder 
with rounded ends (Fig. 27). For each 
speed of rotation, the droplet will assume 
an equilibrium shape determined by its 
density, volume, and interfacial tension. 
The quantitative relationships between 
these parameters were first discussed by 
Lord Rayleigh and have subsequently 
been treated in greater detail by others. 
To date, this technique has only been 
used to measure the interfacial tensions 
of simple liquids. 

In collaboration with Dr. Norman L. 
Gershfeld of the Laboratory of Physical 
Biology, National Institutes of Health, 
we are constructing an apparatus based 
on the principles outlined above, which 
can be used to observe the deformation of 
particles of cellular dimension rotating in 
a horizontal tube at angular velocities up 
to about 100,000 rpm. While the immedi- 

Jm&m oj 

gu 2 

mmmmm go 

cu 4 

Fig. 27. Schematic representation of the de- 
formation of a spherical droplet in a rotating, 
horizontally oriented tube, at increasing values 
of the angular velocity, a>. The axis of rotation 
is given by the dashed line. 

ate goal of these experiments is to meas- 
ure the surface tensions of a variety of 
phospholipid vesicles which are to be 
used in the bilayer membrane-cell surface 
interaction studies outlined above, it is 
also planned to use this method to exam- 
ine the viscoelastic properties of cell sur- 
faces. A final application of this tech- 
nique will be to determine whether, under 
the deformation forces of a rotational 
field, cells can be induced to undergo 
fusion to give viable hybrids. 

Membrane Dynamics 

One of the central problems in any dis- 
cussion of the physical properties of a 
membrane is assessment of its thickness. 
For artificially generated lipid bilayers 
separating two aqueous phases, the trans- 
verse dimension of the membrane can be 
determined from a measurement of the 
intensity of light reflected from the mem- 
brane and hence the difference in optical 
path length between light rays reflected 
from the two surfaces of the membrane. 



This technique has not been successfully 
applied to cell membranes, largely be- 
cause a flat membrane of not too small 
dimension is needed for the experiment 
and because scattering from extraneous 
layers of tissue complicates the interpre- 
tation of the experimental data. Re- 
cently, in collaboration with Cherry and 
Chapman (Department of Chemistry, 
University of Sheffield), we have shown 
that the reflectance technique can be used 
not only to determine the transverse di- 
mensions of a lipid bilayer but also to 
observe membrane phase transitions and 
surface inhomogeneities. Furthermore, 
we found that even in bilayers formed 
from a single, chemically well-defined 
lipid, local fluctuations in the intensity of 
reflected light could be observed. This 

raises the interesting possibility that the 
local molecular organization of a mem- 
brane may be time dependent, and it sug- 
gests that any discussion of the relations 
between membrane structure and phys- 
ical properties should also include this 
dynamic aspect of its organization. 

In order to examine this possibility in 
greater detail, we are constructing the ap- 
paratus described by Cherry and Chap- 
man for measuring membrane reflectiv- 
ity. Recent advances in the theory of 
light scattering from fluctuations in mem- 
branes and monolayers make us hopeful 
that this type of measurement may also 
eventually be applicable to the detection 
of structural changes associated with 
chemical reactions or other processes oc- 
curring in a membrane. 


D. M . Fambrough, H. C. Hartzell, and A. K. Ritchie, in collaboration with 
S. Satyamurti* and D. B. Drachman* and with the technical assistance of 

N. Joseph 

Acetylcholine Receptors in Skeletal 
Muscle Fiber Membranes 

Acetylcholine (ACh) receptors are of 
special interest to us both as membrane 
proteins and as elements figuring promi- 
nently in neuromuscular interactions. We 
have reported (Year Book 70, p. 39) on 
the appearance of ACh receptors during 
muscle development, the production and 
incorporation of ACh receptors into the 
surface membranes of myotubes, the clus- 
tering of ACh receptors at newly formed 
neuromuscular junctions both in vitro 
and in vivo, the number and distribution 
of ACh receptors in adult rat skeletal 
muscle fibers, and the time course of 
change in these parameters following de- 
nervation. During the past year, we have 
continued investigation along these lines, 
focusing upon physiological aspects of 
receptor function, ACh receptor turnover, 
and ACh receptors in pathological states. 
We have continued to use two very differ- 

* Department of Neurology, The Johns Hop- 
kins University Medical School. 

ent assays — iontophoresis, described in 
Year Book 68, p. 53, and a-bungarotoxin 
binding (Year Book 70, p. 62) — to meas- 
ure receptor function and the number of 
receptors, respectively. 

Interaction of ACh Receptors with 
a-Bungaro toxin 

D. Fambrough 

oi-bungarotoxin (BGT) , a component 
of the venom of the Formosan banded 
krait, forms very stable complexes with 
ACh receptors. We have used a radio- 
active derivative of BGT in which 
125 iodine is covalently linked to the toxin 
(I-BGT). For practical purposes the 
complexes may be considered irreversible 
because (1) when toxin is bound to mus- 
cle fibers it cannot be competed off with 
excess cold toxin or with cholinergic 
antagonists such as tubocurarine and (2) 
solubilized toxin-receptor complexes are 
extremely stable. Our experiments dem- 
onstrating this latter point are described 



Free BGThas a molecular weight of ap- 
proximately 8000 daltons. Toxin-receptor 
complexes have a molecular weight be- 
tween 200,000 and 350,000 daltons. Thus, 
toxin-receptor complexes are easily dis- 
tinguished from and separated from free 
toxin. We prepared toxin-receptor com- 
plexes by incubating chick muscle cultures 
in a medium containing 0.2 /xg/ml I-BGT. 
The unbound toxin was then removed by 
repeatedly rinsing the cells with culture 
medium, and the toxin-receptor com- 
plexes were solubilized by flooding the 
cultured muscle with a mixture of 1% 
Triton X-100 and 10 ml Tris buffer at 
pH. 7.8. This extraction is quantitative, 
is complete in about one minute, and re- 
sults in substantial purification of the re- 
ceptors, since nuclei and much of the 
cytoplasm of the cultured cells remain 
attached to the petri plate. To the solu- 
bilized toxin-receptor complexes we added 
a 10,000-fold excess of unlabeled BGT 
and, in some experiments, 10" 4 M Tubo- 
curarin or 10~ 4 to 10" 2 M hexametho- 
nium. These solutions were then steril- 
ized by passage through a Millipore filter 
and stored at controlled temperature 
until analysis. 

Toxin-receptor complexes were identi- 
fied either as approximately 10S material 
in sucrose gradient centrifugation or as 
material excluded from Bio-Gel P-60 
during column chromatography. Dis- 
sociation of BGT from receptor was ob- 
served as the appearance of free I-BGT, 
which has a sedimentation constant ap- 
proximately 2S and is partially included 
in Bio-Gel P-60 column chromatography. 
Proteolytic degradation of BGT did not 
occur in any of these experiments. 

At 4°C, dissociation of toxin-receptor 
complexes was immeasurably slow; after 
30 days all of the material still sedi- 
mented as a 10. 6S peak. At 25 °C, toxin- 
receptor complexes dissociated about 7% 
in 6 days, suggesting a half-time for dis- 
sociation greater than 6 weeks. At 37 °C 
the half-time was 53 hours and at 48 °C it 
was 2.5 hours. As shown in Fig. 28, com- 
plexes are very rapidly dissociated at 

70 °C. All of these measured dissociation 
rates most probably reflect the rate of 
denaturation of receptor or of BGT, with 
concomitant dissociation of complexes. 
Neither 10~ 4 M tubocurarine nor 10~ 2 M 
nor 2 X 10" 4 M hexamethonium had any 
large effect upon dissociation rate. In 
one set of experiments there was less than 
15% dissociation in 5 days in the pres- 
ence of 10~ 4 M tubocurarine and less than 
4% in the presence of 2 X 10" 4 M hexa- 
methonium. Short-term experiments with 
10" 2 M hexamethonium suggest a slight 
acceleration of dissociation, amounting to 
approximately 0.6% per hour compared 
with 0.25% for the control. 

Samples of BGT and of I-BGT were 
sent to Drs. B. Fulpius and E. Reich 
(The Rockefeller University) . They re- 
port that complexes of BGT with eel 
electroplax ACh receptors are similarly 
stable. The homologous toxin from the 
cobra (Naja naja siamensis) forms com- 
plexes with ACh receptors which readily 
dissociate (half-time about 160 minutes 
at 25 °C). The difference in stability of 
the complexes is surprising in view of the 
high degree of homology between the 

Turnover of ACh Receptors in Myotube 
Plasma Membranes 

In rapidly growing and differentiating 
chick skeletal muscle cultures, we have 
described the increase in ACh receptors 
(Year Book 71, p. 39). There is an ap- 
proximately linear increase in receptors 
per culture dish with time for several 
days. During this period, the receptor 
density (receptors per unit area of myo- 
tube surface) increases, and the amount 
of myotube surface membrane also in- 
creases as the result of both myotube en- 
largement and new differentiation. At 
early stages in the course of receptor in- 
crease, the rate of appearance of new re- 
ceptors is so large that turnover (that is, 
the degradation of receptors and their 
replacement by new ones) is negligible. 
During the past year, we have studied 




n ' 



Bio-Gel P-60 


A \ 

15 x 1 cm 



! a 
' t 

i \ 

i \ 
i \ 

1 i 

/ a-BGT I 



i i 

/ \ 

i i I i 1 

i T i I 





Fig. 28. Demonstration of the practically irreversible interaction of a-BGT with the ACh receptor 
from embryonic skeletal muscle. Primary cultures of chick muscle were incubated in 125 I a-BGT at 
0.15 /ttg/ml culture medium for 30 min at 37 °C. Then the unbound toxin was removed by extensive 
washing as described in text. The cultures were immediately extracted with 1% Triton X-100 in 
10 ml Tris buffer, pH 7.8, to solubilize the toxin-receptor complexes. Unlabeled a-BGT was added 
to the extract to a final concentration of 100 /Ag/ml and this solution was sterilized by Millipore 
filtration and stored at 25 °C for 5 days. Then an aliquot was applied to a Bio-Gel P-60 column 
(15 X 1 cm) and eluted with 1% Triton X-100 in 100 ml Tris buffer, pH 7.8. 0.5-ml fractions 
were collected and the radioactivity in each fraction determined by scintillation counting. Elu- 
tion profile is indicated by open circles and solid lines. The unheated complexes are excluded from 
the column matrix and appear in the void volume. This material was further analyzed by sucrose 
gradient velocity sedimentation and shown to move as a single peak with sedimentation coeffi- 
cient about 10S as judged by internal catalase and gamma-globulin markers. Another aliquot of 
a-BGT-receptor complex was heated to 70 °C for 5 minutes and chromatographed as described 
above (open triangles). Heating to 70°C caused release of free, undegraded 125 I a-BGT, suggesting 
that the toxin was not covalently bound to the receptor. 

more developed cultures at which stage 
turnover becomes a significant aspect of 
new receptor appearance. 

Part of the evidence for ACh receptor 
turnover comes from measurements of 
the release of radioactivity from myo- 
tubes after formation of I-BGT-receptor 
complexes in chick muscle cultures. In 
these experiments the cultures were in- 
cubated in medium containing 0.1 to 0.2 
/xg/ml I-BGT for 30 min at 37 °C and 
then given eight 2-minute rinses with 
medium to remove unbound toxin. The 

cultures were gently swirled, usually for 
6 to 8 hours at 37 °C. Then the medium 
was removed, centrifuged, and an aliquot 
counted to obtain a determination of 
radioactivity released from the cells. The 
bound I-BGT was determined by count- 
ing a 0.1% Triton X-100 extract of the 
culture. When the released radioactivity 
was expressed as a fraction of the total 
(released plus bound) , the reproducibility 
of values for a given set of cultures was 

The released radioactivity was ana- 



lyzed by Bio-Gel P-60 column chroma- 
tography (Fig. 29). The great majority 
of counts were found in the low-molec- 
ular-weight included peak and thus rep- 
resent proteolytic degradation of the 
toxin. Some radioactive material ap- 
peared in the excluded peak and may 
reflect the liberation of toxin-receptor 
complexes from the cells. Only a very 
small amount of free toxin is present. 
This pattern is independent of incubation 

Thus, the release of radioactivity from 
the cells does not reflect dissociation of 
toxin-receptor complexes but rather re- 
flects the degradation of toxin which was 
bound to ACh receptors. The degrada- 
tion could occur right at the cell surface 
(if proteolytic enzymes are present on the 
exterior of the surface membrane) or 
could occur, as more conventionally 
imagined, within the cell cytoplasm after 
internalization of small bits of surface 
membrane. The destruction of toxin 

could result in the reactivation of the re- 
ceptors to which the toxin had been bound 
or could be accompanied by destruc- 
tion of the receptor also. We have per- 
formed experiments which help to distin- 
guish between these possibilities. These 
experiments suggest that toxin-receptor 
complexes are internalized, that both 
toxin and receptor are destroyed, and 
that newly synthesized receptors are con- 
tinuously replacing older ones. The data 
are presented in Table 10 and summarized 

As indicated in Table 10, the release of 
radioactivity after I-BGT binding is very 
temperature sensitive, is largely inhibited 
by dinitrophenol and fluoride, and is par- 
tially inhibited by colchicine. The release 
is more rapid in the absence of serum and 
embryo extract. Release is slightly 
affected by cycloheximide and not signifi- 
cantly affected by cytochalasin B, 0.01% 
Triton X-100, nor is it affected by 10 -4 M 
tubocurarine or by the following com- 


t 30 


Bio-Gel P-60 
15x1 cm 





Fig. 29. Analysis of the radioactive material released into the culture medium for 8 hours fol- 
lowing the binding of 125 I a-BGT to cultured chick myotubes. Binding and column assay are as 
described in text and legend to Fig. 28. Note that very little of the released material chromato- 
graphs like free, undegraded toxin. Some material is excluded from the Bio-Gel beads, but the 
major fraction is included, thus representing low molecular weight material, the result of enzyma- 
tic degradation of a-BGT. 



TABLE 10. Release of Radioactivity from Myotubes after 
125 Iodine-labeled a-Bungarotoxin Binding 

Experimental Treatment 

126 Iodine Released 
per hour 

Inhibition of 

to Control 




















Complete medium* 
Complete medium, 26°C 
Complete medium, 18°C 

Complete medium containing 

10~ 3 M fluoride and 2 X 10~ 4 

M dinitrophenol 
Complete medium containing 

100 /wg/ml colchicine 
Complete medium containing 

fluoride, dinitrophenol and 

Complete medium containing 

0.01% Triton X-100 
Complete medium containing 

100 /ig/ml cycloheximide 

BSA medium* 
BSA medium, 26°C 
BSA medium, 18°C 

BSA medium containing 

20 Aig/ml BGT 
BSA medium containing 2.5 

/Ag/ml cytochalasin B 
BSA medium containing 

100 Mg/ml colchicine 
BSA medium containing 10~ 3 M 

fluoride and 2 X 10~ 4 M 



*Complete medium is a modified Ham's F12 containing 15% horse 
serum, 2% chick embryo extract, and buffered with 18 ml N-2- 
hydroxyethylpiperazine-N'-2-ethanesulfonate. BSA medium lacks 
serum and embryo extract but contains all other components of 
complete medium, including 0.5% bovine serum albumin. All in- 
cubations were at 37°C unless otherwise indicated. 

pounds which inhibit certain proteases: 
400 ixg/m\ soybean trypsin inhibitor and 
saturated solutions of tosylarginylmeth- 
ylester, tosylchlorophenyl ketone, and 
phenylmethanesulfonyl fluoride. This 
spectrum of inhibitor sensitivity is con- 
sistent with internalization of toxin- 
receptor complexes followed by proteoly- 
tic degradation and rapid release of 
iodinated proteolytic fragments, but it is 

not consistent with extracellular prote- 
olytic activity. 

Since cycloheximide does not drastic- 
ally reduce the rate of release of 125 iodine 
after I-BGT binding but does block the 
appearance of new receptors (after sev- 
eral hours treatment) , this inhibitor was 
used in experiments designed to distin- 
guish between the reactivation of recep- 
tors following I-BGT destruction and the 



replacement of such receptors by turn- 
over. In these experiments one group of 
cultures was pretreated 3 hours in 100 
/xg/ml cycloheximide and incubated in 
unlabeled BGT to block all existing re- 
ceptors. Then the BGT was removed and 
the cells were incubated for 6 more hours, 
still in the presence of cycloheximide. 
Control cultures were treated in the same 
manner except that no cycloheximide was 
present. At the end of the 6-hour period, 
the number of I-BGT binding sites was 
determined. In other controls the rate of 
release of 125 iodine in the presence and 
absence of cycloheximide was determined. 
The control cultures lost an average 2530 
cpm by release of 125 iodine during the 
6-hour period and the BGT-pretreated 
controls averaged 2660 cpm of new 
I-BGT binding after 6 hours. The cyclo- 
heximide-treated cultures lost an average 
1760 cpm in 6 hours, but the BGT-pre- 
treated, cycloheximide-treated cultures 
averaged only 570 cpm new I-BGT bind- 
ing after 6 hours. In the presence of cyclo- 
heximide release of 125 iodine continues 
but many fewer new BGT binding sites 
appear. Thus, the appearance of new 
sites must represent new synthesis rather 
than reactivation of previously blocked 

Incorporation of New ACh Receptors 
into Myotube Membranes 

H . C. Hartzell and D. M . Fambrough 

When all the ACh receptors on myo- 
tubes in culture are blocked with BGT, 
the appearance of new BGT binding sites 
is the result of the insertion of recently 
synthesized ACh receptors into the sur- 
face membrane. Whether or not a net 
increase in receptor sites occurs depends 
upon the relative rates of insertion and 
removal of receptors. The removal proc- 
ess was described above as a process 
wherein I-BGT is destroyed concomit- 
antly with receptor removal, and the sen- 
sitivity of that process to various inhibi- 

tors was described. The insertion process 
can likewise be studied using BGT, and 
some data on the sensitivity of the proc- 
ess to compounds which interfere with 
ATP synthesis, to temperature, and to 
cycloheximide have been published (De- 
velopmental Biology 30, 155-166, 1973). 
We have subsequently examined the sen- 
sitivity of this process to other condi- 
tions: calcium ion concentration, trans- 
membrane potential, and osmolarity of 
the medium. As shown in Table 11, the 
incorporation of new receptors into the 
surface membrane is not greatly affected 
by any of these conditions. 

TABLE 11. Effect of Various External Solu- 
tions on ACh-Receptor Incorporation 
into Membranes 



Rate of 




of New 

External Medium 



Standard medium 



Na 152 ml 

K 6 ml 

Low potassium 




Na 158 mM 

K 0.3 mM 

350 mM sucrose 





Na 15 mM 

K .06 mM 

Low sodium medium 



Na3 mM 

K 155 mM 

High calcium medium 


Ca 100 mM 

Standard medium plus 


150 mM sucrose 

Standard medium plus 


350 mM sucrose 

Low calcium medium 


0.001 mM Ca ++ 

Standard medium plus 


100 yug/ml 


(3 hr pretreatment) 

Standard medium plus 


2 X 10~ 4 M DNP 

and 2 X 10" 3 M 





The cycloheximide inhibition of new 
receptor appearance develops slowly be- 
cause there exists within the myotubes 
sufficient material to allow continued con- 
struction of chemosensitive surface mem- 
brane for several hours after protein syn- 
thesis is blocked. By blocking all of the 
surface receptors with BGT and then 
measuring the binding of I-BGT to re- 
ceptor sites in cell homogenates, we have 
identified a set of internal binding sites 
associated with material which sediments 
like small mitochondria. During rapid 
myotube growth, this pool of internal 
binding sites contains about one-third as 
many sites as do the surface membranes. 
Consistent with the hypothesis that the 
internal sites are indeed ACh receptors is 
the observation that the internal pool of 
toxin binding sites is diminished by about 
30% by a 3-hour exposure of the cultures 
to cycloheximide. During this treatment, 
the loss of internal toxin binding sites is 
quantitatively balanced by the appear- 
ance of new sites in the surface mem- 
brane. Another measure of the decay of 
the intracellular BGT-binding site pool is 
the specific activity of this pool (in I-BGT 
binding sites per unit weight protein) in 
control and in cycloheximide-treated cul- 
tures. BGT-binding material in cell 
homogenates was centrifuged at 15,000 X 
g for 20 minutes and washed four times 
by resuspension and centrifugation. The 
final pellets were dissolved in 0.1 iV 
NaOH, and aliquots were used for pro- 
tein determination and for counting. In 
this experiment the specific activity of 
the toxin-binding fraction was 14,530 
cpm/mg protein for control cultures and 
8300 cpm/mg protein for 3-hour cyclo- 
heximide-treated cultures — a 43% reduc- 
tion in specific activity after cyclohexi- 
mide treatment. Presumably the internal 
toxin-binding material is preformed 
membrane, and the insertion of this mem- 
brane into the surface involves an exo- 
cytotic mechanism. 

The Neuromuscular Junctions in 

Myasthenia Gravis: Decreased 

Acetylcholine Receptors 

D. Fambrough, in collaboration with 
D. B. Drachman and S. Satyamurti 

Last year we described techniques for 
determining the number and distribution 
of ACh receptors at neuromuscular junc- 
tions and in extra junctional membranes 
in rat diaphragm. During the past year, 
we have put these techniques to use in the 
analysis of muscle biopsy material from 
humans with various muscle disorders in- 
cluding severe cramps, congenital myo- 
tonic dystrophy, myotonia congenita, 
myokymia, Eaton-Lambert syndrome, 
fascio-scapulo-humeral dystrophy and 
myasthenia gravis. The ACh receptor 
analytical techniques should allow us to 
detect the involvement of denervation in 
these muscle disorders and to assess the 
state of the receptive postsynaptic mem- 
brane at the neuromuscular junction. So 
far, we have identified several disorders 
in which no gross receptor-related ab- 
normality is present, we have identified 
denervated fibers in neuropathies and in 
congenital myotonic dystrophy and, most 
interesting, we have discovered a severe 
reduction in the number of functional 
ACh receptors at the neuromuscular 
junctions of patients with myasthenia 

Myasthenia gravis is a neuromuscular 
disorder manifested by muscular weak- 
ness and fatigability. Although the ab- 
normality is thought to involve the neuro- 
muscular junction, the precise nature of 
the defect is uncertain. We measured the 
number and distribution of ACh recep- 
tors in biopsy material from eight pa- 
tients with typical myasthenia gravis. In 
each case, the diagnosis was established 
by a typical history and physical 
findings of muscular weakness and fa- 
tigue. The diagnosis was confirmed by a 
decline of more than 15% in the third or 
fourth muscle action potential in response 
to a train of supramaximal stimuli ap- 





» M4t1 


Fig. 30. Autoradiograms of human muscle fibers after incubation in I-BGT and staining for 
acetylcholinesterase to reveal neuromuscular junctions. (A) and (B) Normal fibers from patients 
HH and NA. Note dense accumulation of silver grains over the junctional area and paucity of 
grains elsewhere. (C) Myasthenic fiber (patient AN) with decreased gram density but grains 
localized over acetylcholinesterase-stained areas. (D) Myasthenic fiber from patient FS showing 
reduced grain density at neuromuscular junction and high density of grains over adjacent extra- 
junctional region. (E) Myasthenic fiber from patient FS. There was an elevated grain density 
extending for at least 3 mm from the neuromuscular junction on this fiber. Phase contrast micro- 
graph in which nuclei appear as dark objects and neuromuscular junction is labeled J. Magnifica- 
tion bars represent 50 ixm in (A) and (E), 20 /xm in (B), (C) and (D). Specific activity of the 
I-BGT ranged from 2.4 X 10 4 to 3.4 X 10 4 Ci/M. Exposure times were 4 days for (A), (D) and 
(E) ; 5 days for (B) and (C). On the scale of relative grain density described in the text, (A) and 
(B) are 4, (C) is 1 and (D) is 2. 



plied to a rate of 2 to 5 per second. All 
eight patients showed improvement in 
muscle strength after injection or oral ad- 
ministration of anticholinesterase drugs. 
Six of the eight had been treated with 
such drugs for a year or more, while the 
other two were not under treatment at 
the time of biopsy. The control group 
consisted of a series of patients with a 
variety of problems. Two of these pa- 
tients can be considered normal. 

Several bundles of muscle fibers were 
removed from each patient, stretched to 
approximately resting length, and pinned 
in dishes in a nutrient medium. Small 
groups of fibers were dissected from the 
bundles, separately pinned and incubated 
for 2 hours at 37 °C in fresh medium con- 
taining 0.1 /xg/ml a-bungarotoxin labeled 
with 125 iodine. After extensive rinsing 
and fixation in glutaraldehyde, the fibers 
were stained for acetylcholinesterase to 
reveal neuromuscular junctions. For 
autoradiographic analysis, lengths of sin- 
gle muscle fibers including the neuro- 
muscular junction were dissected out, 
mounted on microscope slides, and proc- 
essed for autoradiography. Representa- 
tive autoradiographs are shown in Fig. 
30. In order to evaluate the autoradio- 
graphs without bias, two of us studied all 
the autoradiographic slides after they 
had been coded in random order, and 
graded the grain density over the junc- 
tions as follows: 4, normal; 3, possibly 
reduced; 2, definitely reduced; 1, mark- 
edly reduced. Our estimates were in 
close agreement; the grade indicated in 
Table 12 is the mean of the grades given 
to all the slides by both observers. While 
this technique does not yield precise 
quantitative data, the basic features of 
acetylcholine receptor number and dis- 
tribution in myasthenic muscle fibers are 
clear: The number of ACh receptors at 
neuromuscular junctions is strikingly 
subnormal. In most cases, there was no 
increase in extra junctional receptors. 

For quantitative analysis of ACh re- 
ceptors (performed on 4 myasthenic and 
4 normal samples), short segments of 

muscle bundles containing up to 50 neu- 
romuscular junctions were dissected out 
and the exact number of junctions counted 
at high magnification. Then each group 
was hydrolyzed in 100 /xl of 6 N HC1 in a 
sealed tube at 115°C for 48 to 72 hours. 
The tubes were cut open and inverted in 
scintillation counting vials to which was 
added a miscible scintillation cocktail. 
Samples were counted for at least 40 
minutes at 60% efficiency. The results 
are summarized in Table 12. The neuro- 
muscular junctions of patients with 
typical myasthenia gravis contained only 
11 to 32% of the control number of ACh 

Of the eight myasthenic patients, six 
were currently being treated with the 
anticholinesterase drug pyridostigmine 
bromide. One patient had received no 
anticholinesterase medication for more 
than four months and one had never had 
such medication. Since these last two pa- 
tients did not differ markedly from the 
others in ACh number or distribution and 
since two of the "control" patients who 
had taken anticholinesterase medication 
for more than a year had a normal re- 
ceptor number, these results cannot be 
attributed solely to an effect of medica- 
tion. Nevertheless, to obtain some idea 
of the effects of anticholinesterase treat- 
ment on BGT binding sites, we gave the 
anticholinesterase inhibitor neostigmine 
bromide to rats. The doses were 10 to 20 
times those recommended for humans. 
The data are summarized in Table 13. 
Acute administration of neostigmine had 
no effect upon measured receptor number. 
Chronic administration for 14 to 34 days 
led to a decrease in receptor number al- 
though not as large a decrease as seen in 
myasthenic patients. Due to species and 
dosage level differences, we cannot be cer- 
tain whether chronic anticholinesterase 
treatment of humans can lead to reduced 
numbers of ACh receptors, although it 
did not in the two "normal" cases re- 
ported here. This matter warrants addi- 
tional investigation, especially since a re- 
duction in ACh receptors would seem 































■3 G 

03 03 

d o3 

° d 


d <J 


Oh 03 







■+j ■+■= 

a 2 

cu d 
o ^ 






cu 73 

bfi aj 

d m 
a o 

-d > 



d "53 

° d 

CK r, 

03 " 
























O TjH <N 

lO OS oq p 

i— i H r— ( <N 

ON H lO H 
iO W iO CO N 

o d d d o 

-H -H -H -H -H 

o o o o o 


d d d d 
-H -H -H -H 

!*• r- r*- r- 

o o o o 


O CO M lO © 

-^ co id co oi 




lO ^h tJH GO 

t— i rt^ lO CO 

T-H d o d 




l^-COO^uOOOO _, o 

I I I I I I I In I 

















'co "go "x 'x "on "cq "od "as 






rK O co 

-d >» 


TfH O 









































Eaton sy 



































£*-} ^5 r*~s ^i K^5 d K*i ^*i 

■> '> 'S '> '> ■> *> ■> 


"3 "S 'd '£ '£ 'S '5 "S . 



K*"} K*^ P^ r^ P^i r^i r^ K*^ 


^^^^ fe pq ^ fe fe fe fe fe 

O C3i lC t^ 00 GO 
lO ^ lO iO H CO 

H »-S W >-9 »-3 <I 




TABLE 13. Effect of Neostigmine on Number of Acetylcholine Receptors 

(a-Bungarotoxin Binding Sites) at Neuromuscular Junctions 

in Rat Diaphragm 


Control (untreated) 
Acute prostigminef (24 hrs) 
14 days prostigminej 
34 days prostigminef 

No. of Receptor 

No. of 

Sites per Junction 


(Mean db SEM) 



2.81 X 10 7 ± 0.07 



3.27 X 10 7 ± 0.17 



1.48 X 10 7 ± 0.05 



1.83 X 10 7 ± 0.10 


*Number of observations. 

t50 to 75 Mg injected subcutaneously. One or two injections per animal. 
J100 /wg injected subcutaneously twice a day. For the first 5 days, 50 to 75 
Mg doses are given, to avoid excessive mortality. 

highly undesirable for myasthenic pa- 

The finding of a reduced number of 
ACh receptors in patients with myas- 
thenia gravis raises questions about the 
arrangement of functional receptors at 
myasthenic neuromuscular junctions. One 
possibility is that the surface area of the 
postsynaptic surface is reduced by 70 to 
90% in myasthenic junctions. This seems 
very unlikely in view of the light and 
electron microscopic data currently avail- 
able. These data suggest that the folding 
of the postsynaptic surface is simplified 
but that the junctions are extremely elon- 
gated, so that the total surface area may 
not be greatly abnormal. In our biopsy 
material the mean length of neuromuscu- 
lar junctions in myasthenic fibers was 
66.5 ^m, while the mean length for con- 
trols was 27.7 //.m. A more likely correlate 
of reduced ACh receptor number is that 
the packing density of functional ACh 
receptors in the postsynaptic membrane 
is reduced. This could be due to a re- 
duced number of receptor molecules or, 
alternatively, many of the ACh receptors 
may be blocked or altered in some way to 
prevent a-bungarotoxin binding. (The 
clinical symptoms of myasthenia gravis 
suggest the presence of a circulating neu- 
romuscular blocking agent and the lim- 
ited electrophysiological findings are con- 
sistent with the possibility.) A reduction 

in the density of functional ACh recep- 
tors in the postsynaptic membrane, as 
suggested by our results, would be suffi- 
cient to account for the clinical manifes- 
tations and physiological correlates of 
myasthenia gravis. 

Electrophysiological Properties of 
the Membrane and Cholinergic 
Receptor of Developing Myotubes 

A. K. Ritchie 

Two important specialized functions of 
the mature skeletal muscle fiber mem- 
brane are the maintenance of a large rest- 
ing transmembrane potential and the 
ability to generate an electrical signal 
upon activation of surface receptors by 
the chemical transmitter acetylcholine. 
Both the magnitude of the membrane po- 
tential and the electrical signal depend 
upon the presence of ion selective chan- 
nels in the membrane and of ion gradi- 
ents, largely sodium and potassium, 
across the membrane. 

While these properties have been 
widely studied in the adult skeletal mus- 
cle, little is known about the development 
of these properties during cell differentia- 
tion. The ion channels responsible for the 
passive membrane permeability and the 
channel associated with the activated 
acetylcholine receptor may conceivably 
change, not only in number but in char- 



acter, as the membrane composition 
changes to accommodate the specialized 
function of the differentiated muscle 
fiber. Furthermore, depending on the 
activity and the work required of the 
sodium pump, the steady-state distribu- 
tion of ions across the membrane may 
also be changing. Using electrophysio- 
logical techniques, we have examined the 
development of these membrane proper- 
ties during the maturation of myotubes 
maintained in culture. 

Membrane Potential in Developing 

The large resting potential in adult 
muscle is mainly due to the high concen- 
tration of K + ions inside the cell and to 
the cell membrane's large K + permeabil- 
ity relative to its Na + permeability. 

In contrast to the mature skeletal mus- 
cle fiber, the membrane potential of im- 
mature fibers is initially very low and 
progressively increases as the fibers ma- 
ture. During this period of change, it has 
been reported that the internal K + con- 
centration (Ki) remains fairly constant. 
Since the transmembrane potential (Em) 
is largely determined by the K + ion dis- 
tribution, it seems unlikely that differ- 
ences in ion distribution could account for 
the large changes in resting potential. It 
is more likely that changes in specific 
membrane permeability are responsible 
for the increase in Em with age. There- 
fore, we have attempted to measure rela- 
tive ionic permeabilities of the develop- 
ing muscle fiber membrane from experi- 
ments in which the ionic composition of 
the bathing medium was altered. 

Using cultured cells derived from em- 
bryonic rat muscle, we measured the 
membrane potential with an intracellular 
microelectrode and found that the Em 
increases from a mean low value of — 8 
mV in myoblasts to a mean of — 54 mV 
in mature fibers (Fig. 31A). These re- 
sults thus confirm the data published by 
others. The intracellular K + concentra- 
tion and the relative permeability of the 

membrane to Na + and K + ions, pNa/pK, 
were estimated in the following manner. 
The Em of a single myotube was con- 
tinually monitored through an intracellu- 
lar microelectrode while the culture dish 
was perfused with an ion gradient in 
which the Na + content in the medium was 
completely replaced by K + within a few 
minutes. The extracellular K + concentra- 
tion could be measured within 15 /x of the 
myotube by simultaneously recording 
with an extracellular K + selective ion 
electrode. A curve of Em vs log K can 
thus be generated. According to the Gold- 
man equation, extrapolation of such a 
curve to zero Em provides an estimate of 
the intracellular K + concentration. From 
such extrapolation K ; is initially 154 mM 
and declines to 131 mM as the myotubes 
mature (Fig. 31C). If the same data are 
plotted as e EmF /RT vs K , a straight line 
is obtained and an estimate of pNa/pK 
can be calculated from the y intercept of 
such a plot, assuming that Nai is small 
enough to be disregarded. Our results 
produced such straight lines, or lines 
which were only slightly concave at low 
K,). The calculated pNa/pK are shown 
in Fig. 31D as a function of age. The 
data indicate that the low resting poten- 
tial in young myotubes is due to a mem- 
brane with a large pNa/pK ratio and 
that the membrane becomes relatively 
more permeable to K + as the fibers ma- 
ture. We hope to confirm these observa- 
tions by using flame photometry to meas- 
ure the internal Na and K ion content. 

Due to the irregular shape of these cul- 
tured myotubes, precise measurements of 
the specific membrane resistance have 
not been made. Therefore, we do not 
know if a decrease in pNa or an increase 
in pK is responsible for the declining 
pNa/pK ratio with age. Similar changes 
in Em and permeability ratios have been 
reported by others to occur in developing 
heart cells and in the fertilized sea urchin 
egg. In these cells, it is believed that an 
increasing potassium permeability ac- 
counts for the change in permeability 



-60 r 


^ -40 










iM, 4 A 

a— §• 

II 6 


■s-i-f B 

12 5 9 

H c 




4 6 8 10 12 

Days in culture 

Fig. 31. Electrophysiological properties of rat 
myotubes as a function of developmental age. 
(A) The resting transmembrane potential of pre- 
sumptive myoblast (open triangles) and myo- 
tubes (solid circles). (B) The acetylcholine re- 
versal potential determined by measuring the 
ACh response in myotubes held at different 
membrane potentials with an intracellular cur- 
rent passing electrode. (C) The intracellular po- 
tassium concentration and (D) The pNa/pK 
ratios estimated by bathing myotubes in media 
of increasing external potassium concentrations 
while simultaneously recording the membrane 
potential. Each point represents the mean ± 
S.E.M. The number adjacent to each point 
refers to the number of myotubes examined. 

The influence of temperature on the 
resting membrane potential was also ex- 
amined. These studies were performed 
on well-developed cells, usually after 7 to 
12 days in culture. A marked tempera- 
ture dependence of the membrane poten- 
tial, greater than could be accounted for 

by the Goldman equation, was sometimes 
observed. Two likely explanations for 
the phenomena are the presence of an 
electrogenic Na pump or temperature- 
dependent permeability changes. We are 
investigating both possibilities. 

There is another phenomenon which 
may indicate the presence of an electro- 
genic sodium pump in developing chick 
muscle. After young chick myotubes are 
depolarized with a large dose of acetyl- 
choline for 20 to 30 seconds, their trans- 
membrane potentials, instead of return- 
ing to pre-ACh levels, increase to much 
larger values. The increase in transmem- 
brane potential is accompanied by a de- 
crease in input resistance, indicating an 
increase in membrane permeability. The 
ACh-induced hyperpolarization occurs 
when the extracellular medium contains 
only NaCl but does not occur in LiCl, 
sucrose, or other low-sodium media. The 
effect is reversibly blocked at 25 °C and 
is blocked by long-term treatment with 
10" 4 M ouabain (although the blockade is 
slow to develop). Curiously, the ACh- 
induced hyperpolarization has rarely 
been observed in young rat myotubes, al- 
though these myotubes also possess low 
transmembrane potentials. Our tentative 
explanation for the phenomenon in chick 
myotubes is that the ACh by depolarizing 
the cells allows an increased influx of Na 
ions which activate an electrogenic meta- 
bolic mechanism for Na ion extrusion. 

Ionic Selectivity of the Cholinergic 
Receptor during Development 

The cholinergic receptor appears as one 
of the earliest signs of skeletal muscle 
differentiation, and throughout the early 
stages of cell development there is con- 
tinual insertion of new receptors into the 
membrane. Activation of these acetyl- 
choline receptors results in a transient 
depolarization due to the opening of 
channels permeable to both Na + and K + 
ions. A measure of the ionic selectivity of 
such channels, that is, the relative change 
in conductance to Na and K ions (AgNa/ 



AgK), can be estimated from the acetyl- 
choline reversal potential {Er) . This re- 
lationship is given by the equation 

Er = [E K + (AgNa/AgK)# Na ] 

[1+ (AgNa/AgK)] 1 , 

where E K and E NSL refer to the K + and Na + 
equilibrium potentials (Takeuchi, J. 
Physiol. 167: 128-140, 1963). Er was 
measured in cultured myotubes by using 
two intracellular microelectrodes, one to 
pass a polarizing current and the other 
to record membrane potential. Using a 
third microelectrode, acetylcholine was 
ejected over the surface of the myotube 
during the 200 to 300 milliseconds of 
polarizing currents. In all cases a re- 
versal in the acetylcholine response was 
obtained and plots of acetylcholine re- 
sponse vs Em produced straight lines. 
The membrane potential at which the 
straight line intercepted the y axis was 
taken as the reversal potential. The re- 
sults are shown in Fig. 31B for rat myo- 
tubes in different stages of development. 
While it is conceivable that a change in 
ionic selectivity could have occurred 
which exactly compensated for a change 
in ion distribution, the constancy of Er 
with age suggests that neither the ionic 
selectivity of the receptor nor the distri- 
bution of Na + and K + ions is changing 
appreciably with development. 

Our results thus far suggest that the 
ionic selectivity of the acetylcholine re- 
ceptor is a fairly stable property which 
does not seem to be influenced by changes 
in the composition of the differentiating 
membrane. However, the "passive" mem- 
brane permeability characteristics of the 
myotubes appear to be a variable prop- 
erty which is influenced not only by sur- 
face membrane differentiation but also 
by metabolic state. 

Some Effects of Triton X-100 on 
Acetylcholine Receptors 

D. M . Fambrough 

Effect on ACh-induced desensitization : 
The non-ionic detergent Triton X-100 is 

widely used for extraction of ACh recep- 
tors from tissues. In the case of cultured 
chick and rat muscle, 0.05% (w/v) Triton 
X-100 will solubilize receptors whereas 
0.02% or less will not. However, at lower 
concentrations Triton X-100 has some in- 
teresting effects upon ACh receptor func- 
tion and on the distribution of ACh re- 
ceptors on cells. In 0.01% Triton, 
myotubes fail to respond to acetylcholine 
although the myotubes can maintain a 
substantial transmembrane potential 
(—20 to —40 mV). In 0.001% Triton 
ACh sensitivity is reduced by about two 
orders of magnitude compared with con- 
trols, while the resting membrane poten- 
tials will remain in the — 40 to — 60 mV 
range for several hours. Even though 
ACh receptors are not active in the pres- 
ence of Triton, they retain the ability to 
form complexes with BGT. 

It is perhaps not surprising that recep- 
tor function is perturbed by a detergent. 
Another aspect of the Triton effect, how- 
ever, makes it of potentially great useful- 
ness in studying ACh receptor function. It 
has long been known that ACh receptors 
can become desensitized by high concen- 
trations of acetylcholine. The process has 
been difficult to study, since desensitiza- 
tion is generally a minor phenomenon. In 
the presence of 0.001% Triton X-100, de- 
sensitization is extremely rapid and com- 
plete after approximately 500 msec of 
iontophoretic application of ACh. This 
desensitization is also readily reversible 
with a half-time of resensitization on the 
order of 5 seconds. The rapidity and 
completeness of desensitization in this 
case may allow much more detailed anal- 
ysis of the phenomenon. 

Effect of ACh receptor distribution: 
Triton X-100 also has the effect of over- 
coming associative forces at neuromuscu- 
lar junctions where ACh receptors are 
closely packed and at points of ACh re- 
ceptor aggregation on old chick myotubes 
in culture. In both cases treatment of the 
cells for 4 hours with 0.01 % Triton X-100 
results in a redistribution of receptor sites 








. ,: J* 




Fig. 32. Bright field micrographs of autoradiographs of chick muscle cultures after 125 I a-BGT 
treatment. Binding conditions as described in text. Autoradiographs exposed 7 days. (A) Control 
culture, 13-day chick muscle culture. Note clusters of grains, suggesting nonrandom distribution 
of ACh receptors on myotubes. (B) Culture as in (A) except treated for 2 hours at 37°C with 
0.01% Triton X-100 in culture medium after 125 I a-BGT binding. Note the relatively uniform dis- 
tribution of grains over myotubes (B) compared with control cultures (A). 

as though the individual receptors were and that their redistribution is a result of 

now free to diffuse (Fig. 32). Since the their movement in a fluid membrane. 

Triton effect on ACh sensitivity is re- This Triton effect is not complete and 

versible, it seems likely that the receptors thus far there is no satisfactory method 

remain a part of the plasma membrane for quantification of the effect. 





P. J. Stambrook, H. G. Sachs, and J. D. Ebert, assisted by D. Somerville and B. Smith 

A close relationship exists between the 
cell membrane potential and cell prolifer- 
ation. Earlier experiments which manip- 
ulated the membrane potential by vary- 
ing the external potassium concentration 
suggested that the membrane potential 

might participate in the initiation or per- 
petuation of DNA synthesis. In Year 
Book 71, p. 34, we described normal 
changes in the membrane potential of 
Chinese hamster cells as a function of 
their position in the cell cycle. Cells in 



-10 - 


Fig. 33. Cell membrane potential as a func- 
tion of external potassium concentration. Ex- 
ponentially growing cells were transferred to 
media containing increasing concentrations of 
potassium. Five minutes after addition of the 
potassium media, cells were impaled and their 
membrane potentials recorded. Abscissa: potas- 
sium concentration (ml) ; ordinate: membrane 
potential (mV). 

Gi had a characteristically low mem- 
brane potential which increased abruptly 
at the onset of DNA synthesis and de- 
clined only when the cells entered their 
subsequent mitotic division. The coinci- 
dent rise in membrane potential and on- 
set of DNA synthesis have prompted fur- 
ther experimentation. The experiments 
were designed to determine whether an in- 
crease in membrane potential is required 
for cells to begin DNA synthesis and to 
traverse the S phase and G 2 to their next 
mitotic division. A second question ex- 
amined was whether the high K + medium 
blocks cells randomly throughout the cell 
cycle or whether it blocks cells at specific 
times in the cycle. 

Chinese hamster cells respond to high 
external potassium concentrations with a 
decrease in cell membrane potential (Fig. 
33) and a reduction in rate of cell prolif- 
eration (Fig. 34) . At 115 ml K + there is 

little, if any, increase in cell number. 
Although prolonged exposure to 115 ml 
K + is eventually lethal to the cells, there 
is complete recovery of cell growth after 
12 hours when high K + medium is re- 
placed by fresh control medium (Fig. 35) . 
When the mitotic index of the cultures 
in various K + media was determined, we 
observed an inverse relationship between 
the growth rate and the number of cells 
in mitosis (Table 14). With increasing 
K + concentrations there is a reduction in 
growth rate yet an increase in the fre- 
quency of mitotic cells, suggesting that 
high exogenous K + levels impede the 
progress of cells through mitosis. 

20 40 60 

TIME (hrs) 


Fig. 34. Effect of external potassium on cell 
growth. Cells were seeded into replicate 60-mm 
culture plates at 10 5 cells per plate. After 24 
hours the cells were placed into media contain- 
ing increasing concentrations of potassium, and 
cell counts on duplicate plates were made at 
subsequent intervals. 1 raM K + , crosses ; 10 mM 
K + , triangles; 30 mM K + , solid circles; 60 rmV 
K + , open squares; 100 mM K + , solid squares. 



TIME (hrs) 

Fig. 35. Recovery of cell growth. Cells were 
seeded into replicate 60-mm culture plates at 10 5 
cells per plate. After 24 hours, media containing 
various potassium levels was introduced. Cells 
incubated with 115-mM potassium were divided 
into two groups after 12 hours. One series was 
transferred to fresh medium containing 10 mM 
potassium and the other was incubated further 
in 115 mM potassium. The number of cells per 
plate was determined in duplicate at different 
times during the culture period. 10 mM K + 
(open circles, solid line) ; 100 mM K + (squares, 
solid line) ; 115 mM K + (open triangles, solid 
line) ; 115 mM K + transferred to 10 mM K + 
(solid triangles, dashed line). 

To further examine the effect of high 
levels of potassium on the cell cycle, cells 
were synchronized by collection of mi- 
totic cells, and their passage through the 
cell cycle was followed in the presence or 
absence of high K + medium. A popula- 
tion of mitotic cells was divided into four 
series of aliquots, each series containing 
18 replicate culture plates. The first 

series contained cells seeded directly into 
medium containing 10 mM K + and served 
as the control (Fig. 36a). In the second 
series the cells were placed directly into 
medium containing 115 mM K + (Fig. 
36b). The third group consisted of cells 
first suspended in 115 mM K + and after 2 
hours transferred to 10 mM K + (Fig. 
36c). The last group was first placed in 
10 mM K + and after 2 hours transferred 
to 115 mM K + prior to the onset of DNA 
synthesis. In each series, duplicate plates 
of cells were labeled with 3 H thymidine 
for 10 minutes at varying times after 
mitosis, and at every time interval the 
percentage of labeled cells was deter- 
mined by autoradiography. The mitotic 
index was calculated from the same plate. 
The control series displayed an approxi- 
mate 10-hour generation time with about 
a 3-hour Gi and a 6-hour S phase (Fig. 
36a). Cells placed directly into 115 mM 
K + medium were retarded in their exit 
from mitosis and exhibited a greatly im- 
paired DNA synthetic activity (Fig. 
36b) . At no time interval examined were 
more than 60% of the cells labeled, and 
the density of label over nuclei was con- 

TABLE 14. The Relationship between Exo- 
genous Potassium Concentration and 
Mitotic Index 

Potassium Concentra- 

tion (mM) 

Mitotic Index 













Cells seeded at 10 5 cells per 60 mm culture 
plate were grown for 24 hours in control medium 
and transferred to media containing 1.0, 10, 
30, 60, 100, or 115 mM K + . After 60 hours of 
additional culture (24 hours for the 115 mM K + 
sample) the media were aspirated and the cells 
washed and fixed in methanol :ace tic acid (3:1). 
One thousand cells were counted from several 
random fields from each plate and the percent- 
age of mitotic cells determined. 









100 -¥ 

~1 — 


TIME (hrs) 

X I000 




iooo A 
h200 <3 



[j 1000 






j;l000 CL 

200 <f) 





Fig. 36. Response of synchronized cells to different exogenous potassium regimens. A synchro- 
nized cell population was prepared by selective detachment of mitotic cells and divided into four 
different aliquots. Each series consisted of 18 replicate 60 mm culture plates with equal numbers 
of cells per plate, (a) Cells incubated in 10 ml K + . (b) Cells incubated in 115 mM K + . (c) Cells 
incubated for 2 hours in 115 mM K + then transferred to 10 mM K + . (d) Cells incubated for 2 
hours in 10 mM K + then transferred to 115 mM K + . In each series, duplicate plates were labeled 
with 3 H thymidine (1 fic/ml) for 10 minutes at successive intervals after synchronization. The 
plates were aspirated and washed, and the cells fixed with methanol: acetic acid (3:1). The plates 
were then coated with emulsion for autoradiography. Percentage labeled cells (open circles, solid 
line); mitotic cells per 1000 cells (crosses, dashed line). 

siderably lower than in the control series. 
By 14 hours, 4 hours after the mitotic 
peak in the control series, no mitotic cells 
were evident. However, at least a frac- 
tion of the population was able to tra- 
verse the entire cycle, albeit abnormally, 
since at 24 hours nearly 20% of the cells 
were in metaphase, apparently blocked at 
that stage. Cells that were first sus- 
pended in high K + medium and two hours 
later transferred to control medium (Fig. 

36c) were retarded in their exit from 
mitosis. Once fresh medium was added, 
the cells progressed normally through the 
remainder of the cycle, although all 
marker events were delayed by 2 hours 
compared to the control series. The 
marker events included the times of onset 
of DNA synthesis, the minimum in the 
DNA synthetic curve, and the mitotic 
peak. In the last series (Fig. 36d), the 
cells, which were first suspended in con- 



trol medium and transferred to 115 ml 
K + medium at 2 hours, exited mitosis 
normally and began DNA synthesis on 
schedule. The duration of the S phase 
appeared slightly lengthened and the cells 
arrived at mitosis with minor delay. A 
block in mitosis was again evident, since 
cells accumulated in mitosis at 12 and 14 
hours, but by 24 hours the mitotic index 
had dropped below control values. 

Three conclusions can be drawn from 
the above series of experiments. (1) Cells 
can initiate and continue DNA synthesis 
while exhibiting a low membrane poten- 
tial (Figs. 33 and 36d). Synchronized 
cells incubated with control medium but 
transferred to 115 mM K + medium before 
the initiation of DNA synthesis can tra- 
verse the S phase and G 2 to their next 
division with little delay. (2) The high 
potassium medium impedes the passage 
of cells through mitosis (Table 14 and 
Figs. 36b and 36d) . (3) There is an event 
(or events) during Gi that is particularly 
sensitive to the 115 mM" K + medium and 
that may be required for the initiation 
and perpetuation of DNA synthesis. Mi- 
totic cells suspended in 115 mM K + me- 
dium exhibit impaired DNA synthetic 
activity (Fig. 36b). However, if 115 mM 
K + medium is added 2 hours after mitosis, 
the cells proceed through DNA synthe- 
sis and the remainder of the cell cycle 
almost normally (Fig. 36d) . Presumably 
the high K + -sensitive block is passed dur- 
ing the 2-hour incubation in control me- 
dium allowing subsequent traverse of the 
remainder of the cycle in high K + medium. 

As demonstrated in Table 14 and in 
Figs. 36b and 36d, the high K + medium 
interferes with mitosis. The effect of the 
high K + medium on the chromosomes 
themselves was examined by making 
metaphase chromosome preparations. In 
stained chromosome spreads of cells from 
the 24-hour time point of Fig. 4b, up to 
80% of the metaphase figures consisted of 
condensed chromosomes that were dis- 
tinctly banded (Fig. 37a). Furthermore, 
a number of the metaphase figures dis- 


Fig. 37. Effect of prolonged exposure to 115 
mM K + on chromosome morphology. Mitotic 
cells were collected by mitotic selection and ex- 
posed to 115 mM K + for 24 hours (Fig. 36b). 
All cells were trypsinized, washed, and sus- 
pended in 1% Na citrate for 15 minutes. After 
fixation in methanol: acetic acid (3:1), meta- 
phase chromosomes were spread on slides, (a) 
Metaphase chromosomes from a control culture, 
(b) Metaphase chromosomes from a potassium- 
treated culture showing heavy banding, (c) 
Metaphase chromosomes from a potassium- 
treated culture showing distinct coiling. 



played a marked supercoiling (Fig. 37b) banding and supercoiling suggest that 

that presumably contributes to the ob- aside from the Gi block, one effect of the 

served banding. A metaphase figure from high K + may be to modify the chromo- 

control cultures is presented for compari- some architecture and thereby impede 

son (Fig. 37c). The induced chromosome the mitotic event. 


P. J. Stambrook and C. Velez, assisted by D. Somerville and B. Smith 

The above report shows that a high 
level of external potassium not only ex- 
erts an effect on the cell membrane po- 
tential but also blocks Chinese hamster 
cells at specific junctures in the cell cycle, 
namely, in mitosis and Gi. In an effort 
to uncover other cell cycle-specific phe- 

nomena, cells were treated with dibutyryl 
cyclic AMP [(but) 2 cAMP] which is 
also known to reduce cell growth rate and 
the density to which cells will grow. The 
response of exponentially growing cells to 
increasing (but) 2 cAMP concentrations 
in the medium is shown in Fig. 38. Cells 









I0 7 







5xl0 6 



/ x/ 

I0 6 


5xl0 5 






TIME (hrs.) 



Fig. 38. Effect of (but) 2 cAMP on cell growth. Replicate culture plates (60 mm) were seeded 
with 2 X 10 5 cells. After 24 hours (time = hr), the medium was replaced with media containing 
increasing amounts of (but) 2 cAMP, and cell counts were made at subsequent intervals. Control 
(solid circles, solid line) ; 10" 4 M (but) 2 cAMP (crosses, solid line) ; 5 X 10" 4 M (but) 2 cAMP 
(squares, solid line) ; 10" 3 M (but) 2 cAMP (open circles, solid line) ; 10" 3 M (but) 2 cAMP changed 
to control medium at 48 hours (inverted triangles, dashed line). 



Fig. 39. Cell culture response to removal of (but) 2 cAMP. Cells were seeded at 2 X 10 5 cells per 
60 mm culture plate and 24 hours later transferred to medium containing 10 -3 M (but) 2 cAMP. 
After 48 hours in (but) 2 cAMP the drug was removed and fresh medium added, (a) A typical field 
in the culture plate one hour after removal of the drug, (b) A field in the culture plate 2 hours 
after removal of the drug. 



Fig. 40. Release and progress of a population 
of cells blocked by (but) 2 cAMP. Exponentially 
growing cells in replicate culture plates were ex- 
posed to 10 -3 M (but) 2 cAMP for 48 hours. At 
that time, the drug was removed and duplicate 
plates were labeled with 3 H thymidine (1 /xc/ 
ml) for 15 minutes at successive hourly inter- 
vals. The cells were fixed with methanol: acetic 
acid (3:1) and the plates were coated with 
emulsion for autoradiography. Percentage 
mitotic cells (open circles, solid line) ; percent- 
age labeled cells (solid circles, dashed line). 

treated with 10 -4 M (but) 2 cAMP display 
only a slightly altered growth rate, 
whereas at concentrations of 5 X 10 4 M 
and 10" 3 M (but) 2 cAMP, growth rates 
and saturation density are visibly re- 
duced. After 48 hours in 10" 3 M (but) 2 
cAMP there is little further increase in 
cell number. The cells, however, are 
viable since normal growth resumes as 
soon as the (but) 2 cAMP is removed 
(Fig. 38). 

To determine whether (but) 2 cAMP 
blocks cells randomly in the cell cycle or 
at a particular stage of the cell cycle, cells 
were incubated in (but) 2 cAMP for 48 
hours. At that time, the drug was re- 
moved and replaced with fresh medium. 
Subsequently, at hourly intervals, repli- 
cate cultures were labeled for 15 minutes 

with 3 H thymidine. The cells were fixed 
with methanol: acetic acid (3:1) and 
coated with emulsion for autoradiogra- 
phy. Within one hour after withdrawal 
of the drug, between 15% and 20% of the 
cell population was in metaphase (Figs. 
39a, 40, and 41). The normal mitotic 
index for these cells is roughly 4% and 
after extended exposure to 10" 3 M (but) 2 
cAMP is elevated to about 7% (Fig. 41). 
The sudden burst of mitotic activity rep- 
resents the release of a viable population 
of cells which had accumulated at a point 
prior to the onset of cell division. The 
released cells proceed through metaphase 
so that 2 hours after removal of (but) 2 
cAMP a large fraction of the mitotic 
population had progressed from meta- 
phase to anaphase and telophase (Fig. 
39b). Although there is variability from 
one experiment to the next, the mitotic 
index drops roughly 4-fold over the next 
2 hours (Figs. 40 and 41 ) . The percentage 
of cells which incorporate 3 H thymidine 
at each time interval reflects the progres- 
sion of this partially synchronized popu- 
lation of cells through the cell cycle. 
There is a 40% decrease in the fraction of 
labeled cells as the synchronized sub- 
population enters and traverses Gi. This 
decline in labeled cells is followed by a 
greater than 3-fold rise in the fraction of 
cells synthesizing DNA as the parasyn- 
chronous subpopulation enters the S 
phase (Fig. 40). 

The rapid appearance of mitotic cells 
after removal of the (but) 2 cAMP indi- 
cates that under these incubation condi- 
tions a block exists shortly before the 
onset of cell division, either in G 2 or at 
the end of the S phase. To distinguish 
between these two alternatives, cells were 
again grown for 48 hours in the presence 
of (but) 2 cAMP. After 48 hours the drug 
was removed and the cells exposed for 
15 minutes to 3 H thymidine so as to label 
only those cells which were in the S phase 












TIME (hrs ) 

Fig. 41. Localization of the (but) 2 cAMP inhibition as a block in G 2 . Exponentially growing cells 
in replicate plates were grown for 48 hours in the presence of 10" 3 M (but) 2 cAMP. After 48 hours 
the drug was removed and the cells pulse labeled for 15 minutes with 3 H thymidine (1 ^c/ml). 
The radioactive medium was aspirated and the cells were quickly washed. Fresh medium contain- 
ing a 100-fold excess of unlabeled thymidine was added to the plates. At subsequent hourly inter- 
vals, the cells were fixed with methanol: acetic acid (3:1) and the plates coated with emulsion for 
autoradiography. Percentage mitotic cells (open circles, solid line) ; percentage labeled mitotic cells 
(crosses, dashed line). 

at the time (but) 2 cAMP was withdrawn. 
The plates were washed and fresh me- 
dium containing a 100-fold excess of un- 
labeled thymidine was added to the cells. 
At subsequent hourly intervals the cells 
were washed and fixed with methanol: 
acetic acid (3:1) and the plates coated 
with emulsion for autoradiographic ex- 
amination. For the first 2 hours following 
the removal of the drug, approximately 
20% the cells were in mitosis (Fig. 41). 
None of the mitotic cells at these early 
time intervals, however, was labeled (Fig. 
41), which indicates that between 30% 
and 40% of the cells had accumulated in 
G 2 during their incubation with (but) 2 
cAMP. At late time intervals, as the 
number of mitotic cells declines, the pro- 
portion of mitotic cells that are labeled 

increases up to 100%. Since mitotic cells 
at later intervals are labeled, they must 
have been distributed throughout the S 
phase at the time of exposure to 3 H thy- 
midine. This observation agrees with the 
finding that at the time (but) 2 cAMP is 
removed, 30% of the cells incorporate 
radioactivity when labeled with 3 H thy- 

In summary, the reduction in cell 
growth rate and cell density effected by 
incubation with (but) 2 cAMP is not due 
to random lengthening of the cell cycle 
but is the result of an accumulation of 
cells in G 2 and to a lesser extent in mi- 
tosis. The remainder of the cells are 
scattered throughout the S phase and 
maybe Gi, although this latter possibility 
has not been pursued. 




R. L. DeHaan, T. E. Durr, C. R. Krueger, T. F. McDonald, L. Moyzis, C. E. Plitt, 

H. G. Sachs, and M . Springer 

During the past year, DeHaan and his 
colleagues have focused most of their at- 
tention on three aspects of the physiolog- 
ical development of heart cells: (1) the 
differentiation of tetrodotoxin-sensitive 
fast sodium channels in the early heart, 
(2) the rate-setting mechanism in syn- 
chronized groups of cells, and (3) the 
state of myofibrillar organization at the 
time of the first heartbeats. 

Modulation of Sodium Current 

Channels in Embryonic Heart 

Cells by Tetrodotoxin 

The action potential in adult cardiac 
tissue is associated with a rapid increase 
in conductance to inwardly flowing ions. 
This inward current is known to be car- 
ried mainly by sodium ions moving 
through specific sodium channels at high 
velocity. A second component of the in- 
ward current appears to be carried 
through different channels, which are less 
specific. These permit either sodium or 
calcium ions to flow through them, at 
slower velocities than the specific sodium 

It has been known for some time that 
the fast sodium channels are blocked by 
the Puffer-fish poison tetrodotoxin 
(TTX). In contrast, the slow calcium 
channels are insensitive to TTX, but 
these may be blocked instead by manga- 
nese ions or by the drug D-600. 

In work described in these pages last 
year, McDonald, Sachs, and DeHaan 
showed that the heart of the early chick 
embryo was virtually insensitive to TTX, 
unlike adult cardiac tissue, suggesting 
that the action potential in the early 
heart involved only a slow channel 
mechanism. Intact hearts from 2- to 4- 
day-old embryos beat spontaneously in 
TTX at concentrations up to 10" 5 g/ml. 
Tetrodotoxin sensitivity increased pro- 

gressively with development, however. 
Hearts from 7-day embryos were blocked 
at a concentration of 10" 7 g/ml TTX. 
Furthermore, they found that spheroidal 
aggregates formed from reassociated 
heart cells behaved in similar fashion, 
increasing in TTX sensitivity with em- 
bryonic age. During the past year, these 
workers have shown that the fraction of 
such aggregates blocked by TTX also 
changes with time after exposure to the 
drug. Those aggregates, prepared from 
cells of 4-day hearts, whose spontaneous 
beat is suppressed upon treatment with 
TTX gradually lose their sensitivity to 
the drug over a 2—3 hour period of ex- 
posure, resume beating in its presence, 
and are not blocked by a fresh dose at the 
same concentration (10~ 5 gm/ml). Using 
aggregates and analytical techniques 
similar to those described previously, De- 
Haan and his colleagues have therefore 
designed experiments to test whether this 
loss of TTX response: (1) resulted from 
desensitization of the aggregate cells or 
from an inactivation of the drug in their 
presence, (2) was associated with changes 
in electrophysiological parameters of the 
treated cells, and (3) was prevented by 
inhibitors of protein synthesis. 

Figure 42 shows the effect of TTX on 
the beating of aggregates from hearts 
aged 4 to 7 days. Fifteen minutes after 
the addition of the drug 26% of the 4-day 
aggregates were beating, while less than 
10% of the 6- and 7-day aggregates re- 
mained active. During the next 2 hours 
many of the 4- and 5-day aggregates 
resumed beating, while there was little or 
no increase in the number of active 6-day 
and 7-day aggregates. 

The pattern of suppression and recov- 
ery in 4- and 5-day aggregates treated 
with TTX was apparent whether aggre- 
gates were formed from ventricular cells, 






100 r 




g 60 




TIME (hr) 

Fig. 42. The response of heart cell aggregates to TTX (10~ 5 g/ml). Aggregates were formed from 
dissociated heart cells of chick embryos aged 4 to 7 days and maintained in medium 818A con- 
taining 1.3 mM/1 K + . Data for 4-day and 7-day aggregates are mean values from 6-8 plates in 4 
different cultures; data for 5- and 6-day are mean values of 2 plates each. 

atrial cells, or cells from the whole heart. 
The response was also essentially the 
same in medium containing 1.3 or 4.3 
mM/1 K + . The percentage of 4-day ag- 
gregates beating after 15 minutes in TTX 
varied from culture to culture but was 
usually 15% to 35%. Within the next 2 
to 3 hours, the percentage of beating ag- 
gregates increased to between 50% and 
80%. During prolonged incubation with 
TTX, the beating percentage after 6 
hours was usually the same as after 3 
hours. Aggregates from 7-day heart 
failed to show a similar suppression- 
recovery response even at lower doses of 
TTX. For example, at 2.5 X 10" 8 g/ml 
40% of 7-day aggregates continued beat- 
ing. Over a 4-hour incubation at that 
concentration none of the suppressed ag- 
gregates resumed activity. 

To confirm that the toxin was not in- 
activated in the culture plates, TTX was 
added to one of a pair of plates contain- 
ing 4-day or 5-day aggregates. After 3 
hours, the TTX-medium was transferred 
to the companion plate. Such an experi- 
ment is shown in Fig. 43. The two plates 

had essentially identical suppression- 
recovery responses, indicating that the 
toxin remained active. Further, when the 
companion plate (Fig. 43B) was washed 
and incubated for another 45 minutes 
with fresh TTX, the percentage of beat- 
ing aggregates did not substantially de- 

The electrical activity of 4-day aggre- 
gates was monitored at intervals during 
the 3-hour incubation with TTX (Fig. 
44). Normal pacemaker activity was 
blocked about 3 minutes after the addi- 
tion of the drug, resulting in a stable 
resting potential (Fig. 44A, arrow) . Ap- 
proximately an hour later, in those ag- 
gregates that resumed beating, the stable 
resting potential gave way to small oscil- 
lations and subsequent resumption of 
full-scale pacemaker action potentials 
(Fig. 44B). 

To assess the changes in electrical 
properties of cells that might be corre- 
lated with desensitization, action poten- 
tials were recorded in some 60 aggregates 
(in three experiments) before treatment 
with TTX and during the second and 




(A) TTX IO"°g/ml 

(B) TTX medium from (A) 

100 r 

TIME (hr) 

Fig. 43. The response of 5-day aggregates to TTX (10~ 5 g/ml). (A) One plate of aggregates was 
incubated for 3 hours with medium containing TTX. (B) After the 3-hour incubation, the medium 
from plate A was transferred to duplicate plate B for 3 hours. Plate B was subsequently washed 
three times and immediately treated with fresh TTX. 

third hours of incubation with the drug. 
The majority of impalements during the 
TTX incubation were from aggregates 
which had stopped and then resumed 
beating. The results are shown in Table 

15. There was no important difference in 
the maximum diastolic depolarization 
(MDP) or overshoot (OS) of aggregates 
before and during TTX. However, in the 
course of TTX-desensitization, the 

Fig. 44. Electrical activtiy in a 4-day heart aggregate before and during incubation with TTX 
(10~ 5 g/ml). (A) Superimposed records showing control action potentials before TTX and a stable 
resting potential without action potentials (arrow) 3 minutes after treatment with the drug. (B) 
Resumption of pacemaker activity during TTX incubation. Record B was obtained about one 
hour after record A. Voltage calibration is 100 mV. The upper horizontal line in both records in- 
dicates mV. Time calibration is 400 msec in (A) and 4 sec in (B). 



TABLE 15. Action Potential Parameters of Beating 4-day Aggregates before and during 

Incubation with TTX 



TTX (g/ml) 
K+ (mAf/1) 
MDP (mV) 
OS (mV) 
V max (V/sec) 

1 X 10~ 5 

1 x 10- 5 





67.8 ± 2.4 

66.4 ± 1.9 

72.2 dz 1.5 

70.9 ± 1.6 

16.8 ± 1.0 

21.2 d= 1.3 

20.3 ± 1.0 

23.2 d= 1.4 

22.8 d= 1.5 

33.4 d= 2.1 

12.1 =b 0.9 

12.4 d= 1.1 





The maximum diastolic potential (MDP) overshoot (OS) and maximum rate of rise of the action 
potential (V max) were measured in aggregates before TTX and during the second and third hours °f 
incubation with the drug in medium containing 1.3 or 4.3 milf/1 potassium. Data represent mean 
dz S. E. for N number of impalements. 

maximum rate of rise of the action poten- 
tial (V max ) was reduced on the average 
by more than 50%. Before TTX, V max 
ranged from 8 to 45 V/sec and exceeded 
30 V/sec in about one-third of the im- 
palements, while in the presence of the 
drug the highest V max was 18 V/sec. 

The appearance of TTX-insensitive 
action potentials in skeletal muscle fol- 
lowing denervation has been found by 
other workers to be dependent on a pro- 
tein synthetic process and was sensitive 
to inhibition by cycloheximide. When 4- 
day aggregates were pretreated for 4 

100 r 








TIME (hr) 

Fig. 45. Inhibition of TTX-desensitization in 4-day heart aggregates by treatment with cyclo- 
heximide (5 jUg/ml). Control aggregates treated with TTX (10 -5 g/ml) (solid circles) ; 15 minutes 
pretreatment with cycloheximide followed by cycloheximide and TTX (open triangles) ; 4 hours 
pretreatment with cycloheximide followed by cycloheximide and TTX (solid triangles) . Aggre- 
gates were washed three times after 3Mj hr in TTX. Values are the means of 2 plates each. 



hours with cycloheximide at a concentra- 
tion sufficient to block 90% of amino acid 
incorporation (Fig. 45, solid triangles), 
the aggregates continued to beat normally 
in control medium and did not become 
desensitized in TTX. After 3% hours the 
percentage of beating aggregates was still 
only 17. In contrast, those not exposed to 
cycloheximide (solid circles) or treated 
for only 15 minutes (open triangles) be- 
fore TTX rebounded from 16% to about 
75% beating. Fifteen minutes after wash- 
ing out the TTX, the percentages of beat- 
ing aggregates in the control and those in 
the 4-hour cycloheximide plates returned 
to 100 and 98, respectively. To test 
whether 4-day aggregates responded to 
manganese or to D600, which block slow 
calcium channels, preparations in tricine- 
buffered medium were treated with TTX, 
allowed to undergo desensitization for 2.5 
hours, and then exposed to manganese ion 
(Fig. 46). At a concentration of 0.5 mf, 
which had no appreciable effect on the 
activity of control aggregates, the per- 
centage of desensitized aggregates that 
remained beating decreased from 64% to 
6%. In one experiment, D600 (1/xg/ml) 


_a A 

Fig. 46. Response of 4-day aggregates to 
manganese. Aggregates were treated with 10 -5 
g/ml TTX (circles) or an equal volume of 
diluent (triangles), and the percentage of beat- 
ing aggregates (% BA) was determined 15, 75, 
and 150 minutes later. At that time, 20 /x\ of 
50 mM MnCl 2 was added to each dish, bringing 
the final concentration of 0.5 mM. Thirty min- 
utes later % BA was again determined. Each 
point represents the mean of 4 plates. 

was found to have a similar effect. Im- 
palement of nonbeating aggregates in the 
presence of either agent showed them to 
be electrically inactive. 

To determine whether desensitization 
is a time-dependent process that might be 
triggered by even a brief exposure to 
TTX, the experiment illustrated in Fig. 
47 was performed. Aggregates were 
treated with TTX for 15 minutes, washed, 
incubated witl} drug-free medium for 2% 
hours, and then exposed to TTX again 
(Fig. 47A). The percentage of suppressed 
aggregates was similar after both doses. 
When companion aggregates were desen- 
sitized by a 3-hour exposure to TTX, 
washed, and again treated with the toxin, 
the percentage of beating aggregates 
counted 15 minutes later was similar to 
that present at the end of the 3-hour 
TTX densensitization period (Fig. 47B). 
The response of desensitized aggregates 
to the second dose of TTX was in marked 
contrast to that of control aggregates, 
indicating that a brief exposure to TTX 
does not produce desensitization and that 
desensitized aggregates remain in that 
state after the removal of TTX. 

To determine how long the desensitiza- 
tion lasts after removal of TTX, DeHaan 
and his colleagues employed the following 
experimental protocol: aggregates were 
incubated with TTX for 3 hours, washed, 
and then incubated for 0—4 hours in drug- 
free medium before exposure to a second 
dose of TTX. After the initial 3-hour 
incubation with TTX, the aggregate pop- 
ulation was divided into three groups (see 
Fig. 47B) : (1) Sensitive aggregates — 
those not beating after 3 hours in TTX. 
(2) Desensitized aggregates — aggregates 
which resumed beating after between 15 
minutes and 3 hours in TTX. (3) Insen- 
sitive aggregates — those which had con- 
tinued beating after the initial 15 minutes 
in TTX. In the cycloheximide experi- 
ment (Fig. 45, solid triangles) the per- 
centage of insensitive aggregates re- 
mained nearly constant throughout the 
3-hour incubation period with TTX. As- 
suming that the percentage of sensitive 







lOOr ♦ n 





tt 60 



|= 40 








Fig. 47. The response of 4-day heart aggregates to successive doses of TTX (10~ 5 g/ml). (A) 
Aggregates were incubated with TTX for 15 minutes, washed 3 times, and then incubated in drug- 
free medium for 2% hours before the second dose of TTX. (B) Aggregates were incubated for 3 
hours with TTX, washed 3 times, and then treated with TTX 15 minutes later. Values are the 
means of 2 plates each. Also illustrated is the method used to classify aggregates into 3 groups — 
sensitive, densensitized, and insensitive — for the experiments described in Fig. 48. 

aggregates also remained constant during 
experiments similar to that depicted in 
Fig. 47B, any decrease in the total per- 
centage of beating aggregates after the 
second dose of TTX may be attributed 
to a decline in the number of desensitized 

The percentage of desensitized aggre- 
gates after the second dose of TTX was 
expressed as a fraction of the percentage 
of desensitized aggregates following the 
initial 3-hour incubation with TTX. This 
fraction was plotted against elapsed time 
between washing and the second dose of 
TTX (Fig. 48). The fraction of densen- 
sitized aggregates remained nearly con- 
stant for 1 hour, declined sharply during 
the next hour, and stabilized thereafter at 

about zero ; that is, 4 hours after removal 
of TTX, just as many aggregates were 
sensitive to the drug as had been prior to 
the first exposure. 

The curve in Fig. 48 illustrates the re- 
appearance of response to TTX following 
withdrawal from a prolonged exposure to 
the drug at 10" 5 g/ml. In earlier experi- 
ments, during which the dose of TTX was 
raised in increments gradually from 10" 8 
g/ml to 10" 5 g/ml, desensitization was 
sometimes apparent by the time (about 
IV2 hours) the final dose of 10" 5 g/ml was 
administered, i.e., the large dose did not 
provoke as great a depression of beating 
as had a smaller dose earlier. Since de- 
sensitization occurred upon exposure to 
lower doses of TTX, it seemed possible 





£ 0.6 




TIME (hr) 

Fig. 48. Time course of the reappearance of 
response to TTX by 4-day aggregates. Aggre- 
gates were incubated with TTX (10~ 5 g/ml) for 
3 hours, washed 3 times, and incubated in drug- 
free medium for a specified period of time be- 
fore receiving a second dose of TTX. The 
abscissa denotes the time in drug-free medium 
before the second dose. The percentage of de- 
sensitized aggregates beating 15 minutes after 
the second dose of TTX was compared to the 
percentage of desensitized aggregates at the end 
of the initial 3-hour TTX incubation and ex- 
pressed as a fraction thereof. For example, the 
fraction 0.2 indicates that only one-fifth of the 
original population of desensitized aggregates 
were beating after the second dose of TTX. 
Fractions greater than 1.0 or less than oc- 
curred when the calculated percentage of de- 
sensitized aggregates after the second dose of 
TTX either exceeded the original 3-hour per- 
centage or was lower than the 15 minute count 
of the 3-hour incubation. (For further details 
refer to the text and Fig. 47B). Each point 
represents one plate containing 100-200 aggre- 
gates taken from six different cultures. 

that the reappearance of sensitivity might 
be prevented by concentrations of the 
drug below 10 -5 g/ml. After the standard 
3-hour desensitization with 10" 5 g/ml, 
TTX aggregates washed and incubated 
for 4 hours with 10" 8 g/ml TTX were 
found to be still desensitized, as shown by 
a second challenging dose of 10~ 5 g/ml; 
in two experiments the mean fraction of 
desensitized aggregates was 0.93, com- 
pared with 0.05 in the control plates. 
Thus, once cells are desensitized to TTX, 
they do not regain the ability to respond 
as long as they are exposed to the drug, 
even at low doses. 

From these results, McDonald, Sachs, 
and DeHaan conclude that aggregates of 
4-day heart cells respond to TTX in one 
of three ways. Some (20%— 25%) stop 
beating immediately and remain electric- 
ally inactive as long as they are exposed 
to the drug. In a second group (usually 
about 50%), electrical activity ceases 
upon exposure to the toxin, but beating 
is gradually resumed over the next 2—3 
hours in its continued presence. These 
aggregates are shown to have been ren- 
dered unresponsive to a second dose of 
TTX (i.e., desensitized) . The third group 
of aggregates appears to be unaffected by 
TTX (up to 10" 5 g/ml) and continues 
beating throughout the period of expo- 
sure. With increasing embryonic age a 
progressively greater fraction of such ag- 
gregates is sensitive to the drug, and 
fewer undergo desensitization in its pres- 
ence. A reasonable hypothesis to explain 
these differences is that action potential 
generation in sensitive aggregates in- 
volves a "fast" sodium-specific channel, 
while those aggregates which remain ac- 
tive in the presence of TTX, as well as 
those which become desensitized, utilize a 
slow channel mechanism. 

In last year's report, DeHaan and his 
co-workers suggested that the TTX-insen- 
sitive, slowly rising action potentials of 
the early embryonic heart are produced by 
currents carried exclusively through these 
slow cation channels and that the transi- 
tion with development to rapidly rising 
TTX-sensitive action potentials repre- 
sents the appearance in the heart cell 
membrane of a fast-channel mechanism. 
Desensitization of 4-day aggregates by 
TTX appears to involve a similar transi- 

By determining the fluxes of several 
organic cations across the frog myelin- 
ated sciatic nerve membrane, Bertil Hille 
has recently been able to deduce certain 
properties of the ion-selective channel 
that carries this fast inward current. The 
model he proposed for the sodium chan- 
nel is of a small pore 3 X 5 A wide, lined 



with oxygens able to form hydrogen 
bonds, and haying one anionic site capa- 
ble of interacting with cations. The 
known structure of TTX, with a posi- 
tively charged guanidinium group and 
hydroxyls capable of forming hydrogen 
bonds, is consistent with the hypothesis 
that the drug acts by bonding specifically 
in the opening of the fast sodium channel. 
This binding has been measured, the den- 
sity of binding sites on the membrane has 
been estimated, and the site has been 
isolated and partially characterized — all 
attesting to the reality of the channel as 
a definable structure and the specificity 
of its association with TTX. 

Desensitization was blocked by inhibi- 
tion of protein synthesis with cyclohexi- 
mide, as long as the aggregates were ex- 
posed to the inhibitor for a long enough 
time (4 hours) before being treated with 
TTX. This suggests that a pool of pro- 
tein precursors is somehow involved in 
the TTX desensitization and must be 
exhausted before the consequences of the 
inhibition of amino acid uptake can be 
observed. Hartzell and Fambrough have 
reported that the incorporation of acetyl- 
choline receptors into tissue-cultured 
skeletal muscle myotubes is also depend- 
ent on protein synthesis, and that recep- 
tor incorporation or activation is blocked 
only after inhibition of protein synthesis 
for 2 to 3 hours by cycloheximide. 

It is unlikely that the prevention of 
desensitization by cycloheximide re- 
sulted from general deleterious effects of 
that inhibitor. It is known that cyclo- 
heximide does not interfere with oxygen 
consumption or oxidative phosphoryla- 
tion in chick cardiac cells. Moreover, ag- 
gregates continued to beat vigorously 
even after 24 hours in cycloheximide with 
no significant changes in action potential 
parameters. It remains to be resolved 
whether desensitization of cardiac aggre- 
gates involves the modification of exist- 
ing fast channels, activation of dormant 
slow channels, or synthesis and/or inser- 
tion of new slow channels into the mem- 

Myofibrillar Organization during the 
First Beats of the Chick Heart 

During the present report year, one of 
several studies in which Max Springer 
and DeHaan have collaborated has been 
an ultrastructural analysis of the state of 
myofibrillar organization in the heart at 
the time it makes its first few contrac- 
tions. The question addressed is whether 
the heart begins to beat at a given devel- 
opmental stage because it is at that time 
that contractile myofibrils first become 
organized. It has been known for many 
years that the first beats begin in a small 
region along the right margin of the form- 
ing ventricle. Experiments were there- 
fore designed to analyze the level of or- 
ganization of myofibrils after the first 
few heart beats in these areas on the right 
side, at a time when the left side of the 
heart tube had not begun to beat. 

Chick embryos were incubated for 30- 
36 hours to a stage (stage 9) just prior to 
the onset of the heartbeat and were then 
explanted to paper ring cultures, endo- 
derm side up. The embryos were covered 
with nontoxic mineral oil to prevent 
evaporation of the medium and maintain 
optical clarity, and transferred to a mic- 
roscope warm stage (37.5 °C) for contin- 
uous observation. The first pulsations of 
the heart were observed and recorded 
continuously with the closed-circuit video 
photoelectric system described in previ- 
ous reports. Each beat was displayed on 
an oscilloscope and recorded on magnetic 
tape. Immediately after the first spon- 
taneously beating region of the heart was 
identified and its contractions recorded, 
the embryos were fixed rapidly and pre- 
pared for electron microscopy. After the 
embryos were embedded in Epon, they 
were cut with a razor blade into six pieces 
in order to divide the heart into three left 
and three right longitudinal segments. 
Each embryo fragment was mounted on 
double-stick cellophane tape, re-em- 
bedded in Epon, and sectioned perpendic- 
ular to the long axis of the embryo. To 
determine the percentage of cells contain- 



ing myofibrils at a given degree of differ- 
entiation, the ultrathin sections were ori- 
ented at low magnification, and counts 
were performed directly on the screen of 
the electron microscope at a primary 
magnification of 30,000. 

Springer and DeHaan confirmed that 
contraction generally starts along the 
right margin of the tubular heart at about 
stage 10. By exploring the parts of the 
heart with photoelectric sensors on the 
video screen over the next 40—60 minutes, 
they were able to observe the gradual 
spread of contraction to the left side of 
the ventricle and the onset of rhythmic 
synchronized beats throughout the entire 
primitive heart. 

Examination of sections of these tubu- 
lar hearts, fixed just before they had 
started beating (stage 10"), showed that 
the myocardium was composed of two to 
three layers of cells. These cells were 
characterized by abundant free ribo- 
somes, mitochondria, an extensive Golgi 
system, and sparse granular reticulum 
distributed widely. Wavy intermediate 
filaments and a few randomly oriented 
thick filaments were found throughout 
the whole cell, while straight thin fila- 
ments seemed to be more frequent near 
the cell surface. Long helical chains of 
polyribosomes and randomly oriented 
bundles of 4-5 microtubules ran through 
the cytoplasm. 

At stage 10, after the first few beats 
along the right side of the heart, the num- 
ber of filaments in those cells had sig- 
nificantly increased. Single thick fila- 

ments were abundant and were now often 
surrounded by parallel thin filaments, 
forming irregular but identifiable hexa- 
gonal arrays. Primitive myofibrils with a 
few sarcomeres were occasionally present. 
The primitive sarcomeres showed a longer 
Z-to-Z spacing than in older stages and 
consisted of relatively few, imperfectly 
aligned myofilaments. 

By stage 11 (13 somites) when each 
rhythmic beat involved the entire heart, 
some relatively well-organized myofibrils 
and bundles of thick and thin filaments 
were found in virtually all cells. Z-bands 
were sharper, better aligned, and more 
prominent. Myofibrils contained more 
but shorter sarcomeres. 

In order to make a quantitative com- 
parison of beating and nonbeating regions 
of the heart, 17 of the 42 embryos from 
which cardiac contractions were recorded 
(Table 16) were cut into six pieces ac- 
cording to a pattern designed so that the 
right midpiece would include the tissue 
that exhibited the first contractions in 
each heart. Every cell in a thin section 
through each of these segments was ex- 
amined at high magnification (30,000X) 
and tabulated in three categories: (1) 
Cells without detectable thick filaments. 
(2) Cells with one or more thick fila- 
ments but without obvious myofibrils. 
Those showing cross sections through 
one or a few primitive myofibrils but 
which exhibited no Z-band material were 
also counted in this category. (3) Cells 
with easily recognizable primitive myo- 
fibrils showing distinct sarcomeres and Z- 

TABLE 16. Characterization of Embryos Used for Quantitative Study 
of Myofibrillar Organization 


al Activity 

Morphological Features of 
the Heart 

Number of 



Left Side 

Right Side 




straight primitive heart tube 

not yet beating 




heart tube slightly bent to 

the right side 


first contraction 

i 10 



distinct heart loop 

both sides contract vigorously 




inverse heart loop; heart tube 

bent to the left side 

first contractions 





lines. These data are summarized in the 
histogram of Fig. 49. 

Before the hearts started to beat (stage 
10") no myofibrils could be observed in 
any of the six regions examined. The 
right wall of the ventricle showed the 
highest percentage of cells with thick fila- 

ments. However, even in this region, 
many cells contained no myofilaments at 
all (Fig. 49-1). About one hour later, 
after beating began, the percentage of 
cells with thick filaments was about 
double that at the previous stage. The 
highest degree of differentiation (i.e., the 





50 7< 

A B C D E F 


A B C D E F 






50 % 


Fig. 49. Quantitative comparison of six defined regions of the embryonic heart. The regions are 
A rostral right side of ventricle, B middle right side, C caudal right side, D caudal left side of ven- 
tricle, E middle left side, F rostal left side. The white areas represent cells without detectable 
thick filaments; cross hatched areas, cells with one or more filaments, but without obvious myo- 
fibrils ; dotted areas, cells with easily recognizable primitive myofibrils showing distinct sacromeres 
and Z-lines. The interval of confidence of 95% indicated. I. Heart at stage 10", not yet beating. 
II. Heart at stage 10; first contractions in region B. III. Heart at stage 11; both sides contract 
vigorously. IV. Inverse heart loop at stage 10; the first contractions are in region E. 


greatest number of myofibrils) was still (histogram 49-11) , visible contraction of 
found in the right wall of the ventricle, the heart may require not only a certain 
coincident with the site where contrac- degree of differentiation but also a mini- 
tions started. By this time primitive mal number of myofibril-containing cells, 
myofibrils could be found in all six re- One of the primary questions the pres- 
gions of the heart but were most frequent ent study set out to answer was whether 
in the area that was already beating (Fig. the primitive heart starts to beat at the 
49-IIB) . Four to five hours later, at stage stage it does because myofibrils first be- 
ll, thick filaments were abundant in come organized at that time. In prin- 
every cell in all six regions of the heart, ciple, heart tissue must differentiate at 
Myofibrils were frequent in all areas but least three systems before it can exhibit 
somewhat more abundant in the right visible contractions: (1) organized con- 
ventricle (Fig. 49-IIIB, C). In the two tractile myofibrils, (2) spontaneously ac- 
hearts with the situs inversus the highest tive pacemaker membranes in some or all 
percentage of myofibrils and thick fila- cells, and (3) low-resistance junctional 
ments were found at stage 10 in the left contacts between cells to mediate a con- 
wall of the ventricle, where the first con- ducted stimulus. The nexus represents 
tractions were also visible. Thus, the the morphological site for the low-resist - 
histogram produced (Fig. 49-IV) is a ance pathway between cells, 
mirror image of the normal (Fig. 49-11). DeHaan and Springer found that focal 

When the first contractions occurred nexal junctions were already present be- 

(histograms II and IV), the beating re- fore the first heartbeat could be observed 

gions showed a significantly higher de- and that no structural changes or striking 

gree of cytodifferentiation than the quies- differences in the distribution of the spe- 

cent regions. Development in the quies- cialized junctions were observed between 

cent regions was very similar to that in beating and quiescent hearts. It has also 

the best-developed regions at stage 10" been reported by other workers that elec- 

(histogram I) which were soon to beat, trical activity can be recorded from the 

Thus, differentiation started in the right heart at stage 9-10, before active contrac- 

ventricle and spread out to the other side tions begin. Therefore, the data presented 

of the ventricle and to the rest of the here suggest that the final critical event at 

heart according to the same pattern as stage 10 is indeed the assembly of myo- 

the contractions, which also spread over filaments into functional myofibrils. A 

the heart from their site of origin in the reasonable, though still tentative, conclu- 

right side of the ventricle. sion seems to be that the embryonic heart 

The analysis of myofilament and myo- starts beating as soon as a certain degree 

fibril distribution suggests that visible of myofibrillar organization is established 

contraction requires a degree of organiza- in a critical number of localized cells, 
tion greater than that found in a few cells 

at stage 10". The best differentiated cells Control of Synchronized Pulsation 

at this stage showed scattered myofila- Rate in Cell Groups 
ments and small foci of Z-body densities 

similar to cells in the nonbeating areas DeHaan and Sachs had demonstrated 
(A, C, D, E, and F) of stage 10 embryos, that spheroidal aggregates prepared from 
Primitive myofibrils with sarcomeres, dissociated heart cells beat at rates in- 
distinct Z-bands and at least poorly versely proportional to aggregate volume ; 
aligned thick and thin filaments seem to that is, the bigger an aggregate, the 
be the minimal condition for visible con- slower it would beat. In the past year, 
traction of a heart cell. Since a few cells working with an able group of students, 
which met these conditions were present DeHaan has launched a multifaceted 
in the nonbeating regions of the heart analysis of this unexpected finding. 



Fig. 50. (A) Scanning electron micrograph of two synchronized cardiac myocytes in tissue cul- 
ture. X 1700. (B) Contour map of cell pair shown in Fig. 50A; courtesy of Mr. F. S. Baxter, U.S. 
Geological Survey. Contour interval 1.2 fiM, X 2400. 



Two sets of properties that might influ- 
ence pulsation rate change with increas- 
ing size of such aggregates: (1) those 
properties that are related strictly to 
number of cells in the system and are 
independent of spatial configuration, and 
(2) those properties that depend on spa- 
tial organization of the parts of the 
system. In the first category would be 
such characteristics as total aggregate 
volume, total amount of cell membrane, 
and total aggregate capacitance. The 
second category of properties would in- 
clude ratio of external surface to volume, 
and intercellular diffusional considera- 

To perform experiments designed to 
distinguish between these broad classes of 
rate-controlling phenomena, DeHaan has 
worked out methods for manipulating 
spontaneously beating aggregates one at 
a time under direct visual observation. 
When brought into contact, aggregates 
beating at different rates adhere and soon 
(usually within 30—60 minutes) synchro- 
nize to a common rhythm. Thus, pairs or 
groups of aggregates of selected rates and 
sizes can be joined to form systems of any 
desired spatial configuration and allowed 
to synchronize. 

According to the data of Sachs and 
DeHaan small spherical aggregates (1 X 
10 5 yu, 3 in volume) have a mean pulsation 
rate of about 160 B/M, whereas aggre- 
gates 20 times that size (2 X 10° /x 3 ) beat 
only half as fast. Is the difference in rate 
in this case due to the 20-fold difference 
in total cell number, or does it result from 
other factors only secondarily related to 
volume, such as surface-to-volume (S/V) 
ratio? One approach to this question was 
carried out by Linda Moyzis, who built 
up linear chains of 6-20 aggregates and 
demonstrated that the synchronous pul- 
sation rate of these chains was only 
slightly reduced from the average rate of 
the component aggregates prior to syn- 
chronization. In this case, the rate was 
not a function of total synchronized cell 

A second approach was taken in experi- 

ments performed by Caroline Krueger. 
When sheets of cells were prepared at 
known densities (in cells/mm 2 on the 
plate) and cut into rectangles of different 
areal sizes, Krueger found that the syn- 
chronous pulsation rates of the rectangles 
did not differ from each other in a sys- 
tematic way over a 60-fold range of areas 
(i.e., number of synchronized cells). 
Again, rate was found not to be a function 
of synchronized cell number. However, 
when sheets were prepared at two differ- 
ent thicknesses (4 cells or 6 cells deep) , 
the thicker rectangles beat only half as 
fast as the thin ones. The controlling fac- 
tor appeared to be not cell number but 
diffusional restriction between the inside 
of the mass and the outside, or S/V ratio. 

This possibility was tested directly in a 
series of experiments carried out by 
Thomas E. Durr. Joining pairs of aggre- 
gates of different sizes (and therefore of 
different rates) Durr determined that the 
change in rate experienced by each mem- 
ber of the pair upon synchronization con- 
formed to that predicted by the new S/V 
ratio rather than to the total volume of 
the pair. 

The consistency of this finding using 
tissue-like multicellular systems has en- 
couraged DeHaan and Springer to apply 
innovative methods to ask whether at the 
single-cell level, pulsation rate is also 
controlled by S/V ratio. In collaboration 
with Durr, the independent pulsation 
rates of pairs of single cells have been 
recorded photoelectrically before and im- 
mediately after synchronization. The 
cells were then fixed and prepared for 
scanning electron microscopy, and stereo- 
scopic scanning EM photographs were 
taken of each pair (Fig. 50A) . Using 
photogrammetric techniques (with the 
aid of Messrs. F. Baxter, F. Doyle, and 
H. Loving of the U.S. Geological Survey) 
contour maps of the cell surfaces have 
been drawn (Fig. 50B) on a Kern plotter 
(model PG2). From these contour maps, 
surface areas and volumes of the cells can 
be determined with a computerized digit- 
izer. At the time of this writing, these 



calculations are in progress, and thus it is 
too early to report the results. If the 
hypothesis is borne out, however, it may 
lead to a profound generalization con- 
cerning the rate-setting mechanism in 
heart cells. 

Cytochalasin B and Embryonic Heart 

Muscle: Contractility, Excitability, 

and ultrastructure 

M. Springer, T. F. McDonald, and H. G. Sachs 

The effect of cytochalasin B (CB) on 
spontaneous contraction, electrical activ- 

ity, and cytological ultrastructure of em- 
bryonic hearts and tissue-cultured heart 
preparations was investigated. The in- 
hibition of spontaneous pulsation and the 
subsequent recovery after drug washout 
were found to be dose dependent and to 
be a function of the type of preparation 
used. In the presence of CB, spontaneous 
action potentials can be recorded from all 
three (intact hearts, isolated myocytes, 
reaggregated cells) after cessation of all 
visible contraction, although the action 
potential duration is reduced. CB at 0.5 
to 10 /Ag/ml was found neither to elec- 

f ( r t *. » 


■ ; - •■*:>■■■■ 

■ . * 

;k , L 


,....,-. ... 

: ■ - ;> ■ ■ : : . ....■: \- ... . 

Fig. 51. (A) Electron micrograph of 7-day heart cell aggregate. After 24 hours in culture, CB 2 
jug/ml was added for 30 hours. The myofibrils are disrupted in bundles of thick and thin filaments 
of approximately sarcomere-length, scattered throughout the cytoplasm. Dense clumps of Z-mate- 
rial (Z) seem to be out of alignment with the bundles. X 42,000, calibration bar 0.5 /xM. (B) Con- 
trol. Myofibrils with well-developed sarcomeres in an aggregate after 24 hours in culture. X 30,000, 
calibration bar 0.5 /xM . 



trically uncouple heart cells nor to pre- 
vent cells from aggregating and estab- 
lishing new electrical coupling between 
cells. In all three systems, however, CB 
does result in disruption of myofibrillar 
organization. After 30 hours in cytocha- 
lasin B (2 /xg/ml) bundles of myofila- 
ments are found to be randomly oriented, 
with Z-band material condensed into 
amorphous bodies, no longer aligned with 
the myofibril but remaining in register 
between the A bands (Fig. 51). Within 
the bundles the normal hexagonal cross 

section of thick and thin filaments is 
maintained. The unaligned Z bodies, 
however, demonstrate attached thin fila- 
ments of very reduced length. Following 
washout of the drug, normal sarcomeric 
organization is restored in parallel with 
resumption of spontaneous beating. The 
present data, therefore, are consistent 
with the hypothesis that one major effect 
of CB is an alteration in myofibril (myo- 
filament) stability and that this effect 
alone is sufficient to account for the ob- 
served inhibition of contractility. 


R. O'Rahilly and E. Gardner 

Owing to the imminent move of the 
Carnegie Collection, the material has un- 
fortunately been "in mothballs" for most 
of the year. However, it is now scheduled 
to leave Baltimore during July 1973* for 
its new home in the University of Cali- 
fornia School of Medicine, Davis, Cali- 
fornia 95616. Professor Ronan O'Rahilly 
will continue his stewardship as Curator 
of the Collection; he has been joined by 
Professor Ernest Gardner as Associate 
Curator. It is anticipated that the Col- 
lection will be "open for business" by 
October 1, 1973. It will once again be 
available to qualified investigators. In- 
quiries may be directed to either Profes- 
sor O'Rahilly, whose address is Wayne 
State University, Scott Hall of Basic 
Medical Sciences, 540 East Canfield Ave- 
nue, Detroit, Michigan 48201, or to Pro- 
fessor Gardner in the Department of 
Human Anatomy at Davis. 

Developmental Stages in 
Human Embryos 

The famous Blechschmidt Collection of 
human embryos, housed in the Anatom- 
ical Institute of the University of Gott- 
ingen, has been given Carnegie numbers 
(10,304 to 10,434). Professor Blech- 

* Note added in proof : The move was com- 
pleted as scheduled. 

schmidt is in the process of assigning 
these embryos to their appropriate de- 
velopmental stages. Because of the ex- 
istence of a particularly important col- 
lection of reconstructions made over 
many years under Professor Blech- 
schmidt's supervision, it is believed that 
a joint cataloguing of staged embryos 
will be advantageous. The Blechschmidt 
reconstructions cover 22 embryos and one 
fetus (40 mm). 

O'Rahilly's account of stages 1 to 9, 
that is, embryos of the first three weeks, is 
now in press. In addition to the detailed 
descriptions of staged embryos, the 
monograph will include a specimen index, 
75 photographs and drawings, and a bib- 
liography of some 325 references. 

Although the first three stages (1, uni- 
cellular stage; 2, stage of segmentation; 
3, free blastocyst) follow Streeter's de- 
velopmental horizons I to III, the pro- 
posed stages 4 to 9 do not correspond to 
horizons IV to IX, respectively. Stages 
4 and 5 cover implantation, and stage 6 
is characterized by the initial appearance 
of chorionic villi, both branched and un- 
branched. Stage 7 is marked by the pres- 
ence of the notochordal process, and stage 
8 by one or more of the following: primi- 
tive pit, notochordal canal, neurenteric 
canal. The neural folds form during 
stage 8. Stage 9 is characterized by the 



appearance of one to three pairs of 

Progress continues in the tabulation 
of the timing and sequence of develop- 
mental events in staged human embryos. 
Data for the heart, nervous system, and 
urinary system have already been pub- 
lished. The limbs (O'Rahilly and Gard- 
ner) and the respiratory system 
(O'Rahilly and Boyden) are now being 
readied for publication. 

At the end of the embryonic period 
proper (stage 23), the human embryo is 
approximately 30 mm in C.-R. length and 
is about 8 postovulatory weeks in age. It 
is surprising that articles and books are 
still appearing in which an age (7 weeks) 
based on macaque data is being cited. 

Development of the Conducting System 
of the Heart 

Gardner and O'Rahilly are studying 
the nerve supply of the human heart and 
the conducting system at stage 23 (8 
postovulatory weeks). 

It has been found that the nerves on 
the right-hand side arise from cervical 
sympathetic and from cervical and tho- 
racic vagal filaments. Out of their inter- 
connections a single vagosympathetic 
nerve emerges, which (1) sends a branch 
in front of the trachea to the aorticopul- 
monary ganglion, and (2) as the right 
sinal nerve, supplies the sinu-atrial and 
atrioventricular nodes and pulmonary 

The nerves on the left-hand side arise 
similarly from cervical sympathetic and 
from cervical and thoracic vagal filaments. 
The cervical filaments form a vagosym- 
pathetic nerve, which descends to the 
right of the arch of the aorta and enters 
the aorticopulmonary ganglion. Fila- 
ments leaving the ganglion supply the 
pulmonary trunk, ascending aorta, inter- 
atrial septum, pulmonary veins, and, as 
the left sinal nerve, the fold of the left 
vena cava. The thoracic vagal filaments 
descend to the left of the arch of the aorta 

and supply chiefly the arterial end of the 

No thoracic sympathetic cardiac fila- 
ments were found. 

The sinu-atrial node is a crescentic 
mass in front of the lower part of the 
superior vena cava. The left limb of the 
node curves downward as the anterior 
internodal tract. The right limb curves 
downward behind the vena cava and, to- 
gether with the crista terminalis, forms 
the middle and posterior internodal 
tracts. An additional internodal connec- 
tion by way of the left venous valve was 

The atrioventricular node is a conspic- 
uous mass in the anterior and lower part 
of the interatrial septum, from which a 
clearly defined bundle leaves to enter the 
interventricular septum. Right and left 
limbs were observed, the former being a 
rounded bundle that passes immediately 
in front of the root of the aorta. 

Development of the Nervous System 

Work by Gardner and O'Rahilly on the 
embryonic brain continues, and O'Rahilly 
has studied the early development of the 
hypophysis cerebri. 

The site of the craniopharyngeal in- 
vagination is detectable immediately 
rostral to the oropharyngeal membrane 
at stage 10. The overlying neuroecto- 
derm is the infundibular area, and the 
hypophysis arises at a single locus, not 
from two distinct evaginations which ap- 
proach each other and fuse. At stage 13 
the basement membrane of the cranio- 
pharyngeal evagination and that of the 
brain are clearly in contact. The cranio- 
pharyngeal pouch becomes elongated by 
stage 14, and a prechordal strand, rather 
than the notochord itself, is inserted into 
its dorsal wall. From stage 14 onward, 
blood vessels grow in between the afore- 
mentioned basement membranes (c/. the 
lentiretinal space) and are later incor- 
porated in the stroma (stage 20). At 



stage 15 mitotic figures are common in 
both walls of the pouch and are located 
next to the cavity. The infundibular re- 
cess, a slight indication of which may be 
seen in some embryos at stage 16, dis- 
plays a characteristically folded wall, 
namely, the neurohypophysis, by stage 
17. The juxtacerebral wall of the cranio- 
pharyngeal pouch is the thicker. The 
lateral lobes (future infundibular, or 
tuberal, part) and the anterior chamber 
(Vorraum) are clearly visible by stage 17. 
The adenohypophysial portion of the 
pouch expands during stage 18, when its 
pharyngeal stalk may become closed. 
The adenohypophysial epithelium adja- 
cent to the neurohypophysis constitutes 
the beginning pars media by stage 20. 

Development of the Larynx 

Studies by O'Rahilly and Tucker on 
laryngeal differentiation continue, and a 
preliminary account was published dur- 
ing the year. An exhibit on the embryol- 
ogy of the human larynx was prepared 
for the annual meeting of the American 
Academy of Ophthalmology and Oto- 
laryngology, 1972, and a silver medal was 
awarded to Tucker and O'Rahilly "for 
excellence of presentation, originality of 
work, and teaching value." 

Development of Bones and Joints 

An article on early ossification and an 
abstract on the development of the shoul- 
der joint appeared during the year. 


A Symposium on Developmental As- 
pects of Membrane Biology was held at 
the Laboratory on April 12 and 13, 1973. 
Organized by Robert DeHaan and Doug- 
las Fambrough, the program focused on 
four major topics: structure and dynam- 
ics of cell membranes, electrical coupling 
between cells, physiological differentia- 
tion of membranes, and interaction be- 
tween cell surface and genome. The roster 
of participants included Michael Aber- 
crombie (Strangeways Research Labora- 
tory), Max Burger (Biozentrum der Uni- 
versitat, Basel), Philippa Claude (Harvard 
Medical School), Philip Dunham (Syra- 
cuse Urliversity) , G. M. Edelman (Rocke- 
feller University), Alan Finkelstein (Al- 
bert Einstein College of Medicine), 
Norton Gilula (Rockefeller University) , 
Victor Ling (The Ontario Cancer Insti- 
tute), Philip Nelson (National Institutes 
of Health) , J. R. Sheppard and Judson 
Sheridan (both of the University of Min- 
nesota), S. J. Singer (University of Cali- 
fornia at San Diego) , and Hermann Wolf 
(Dalhousie University) . A number of 
colleagues at The Johns Hopkins Univer- 
sity also played leading roles, among 
them L. Brand, R. A. Cone, Pedro Cuatre- 
casas, M. Edidin, M. Jacobson, R, T. 

Johnson, W. Kundig, Saul Roseman, and 
Stephen Roth. 

International conferences and symposia 
in which members of the Department 
participated during the year included the 
following: Fourth International Confer- 
ence "De la Physique Theorique a la 
Biologie" (Versailles, France) ; Japanese 
Society for Developmental Biology 
(Tokyo) ; Fourth Lepetit Colloquium on 
the Search for Episomes in Eucaryotic 
Cells (Cocoyoc, Mexico) ; Eighth Meet- 
ing of the Federation of European Bio- 
chemical Societies (Amsterdam) ; Work- 
shop of the European Molecular Biology 
Organization (Paris) ; and the Fourth In- 
ternational Congress of Biophysics (Mos- 

Among the other conferences engaging 
the interest of members of the group, the 
following should be mentioned : a Sympo- 
sium on Synaptic Transmission and Neu- 
ronal Interaction (Woods Hole) ; one deal- 
ing with Trophic Functions of the Neuron 
(New York) , and two dealing with recep- 
tors: the American Society for Neuro- 
chemistry's Symposium on Cholinergic 
Receptors and a related program on the 
Current Status of Pharmacological Re- 
ceptors, sponsored by the Amercian So- 



ciety of Pharmacology and Experimental 
Therapeutics. The staff found several 
Gordon Conferences attractive, among 
them conferences on developmental biol- 
ogy, heart muscle, molecular pathology, 
and nucleic acids. The annual symposia 
at Cold Spring Harbor and Gatlinburg 
included speakers from the Department, 
as did both the University of Minnesota's 
program on Repetitive DNA, Chromo- 
some Defects and Neoplasia, and the 
University of Colorado's Molecular Neu- 
robiology Symposium (Given Institute of 
Pathobiology, Aspen) . 

Lectures were presented on a number 
of campuses, including Case Western Re- 
serve University, Columbia University, 
Cornell University, Duke University, 
Georgetown University, Harvard Med- 
ical School, Harvard University, Indiana 
State University, Kansai Medical School, 
Kyoto University, Kyushu University, 
Nagoya University, Ohio State Univer- 
sity, Rensselaer Polytechnic Institute, 
Rockefeller University, Rutgers Uni- 
versity, Temple University, Tohoku 
University, Tokyo Metropolitan Uni- 
versity, The Universities of Cali- 
fornia (Berkeley, San Diego, Santa 
Barbara), Chicago, Cincinnati, Connecti- 
cut, Georgia, Oregon, Osaka, Paris, 
Tokyo, Washington, Wisconsin and Wurz- 
burg, Vanderbilt University, Wheaton 
College, Washington University School of 
Medicine, and Yale University. 

Special presentations included the 
American Institute of Biological Sciences 
Silver Anniversary Lecture (Pennsyl- 
vania State University), the Carter- 
Wallace Lectures (Princeton University) 
the Centennial Lecture (Upstate Medical 
Center, State University of New York) , 
the Distinguished Scientist Lecture (Tu- 
lane University) , the H. V. Wilson Lec- 
ture (University of North Carolina), and 
the Tamaki Memorial Lecture (Kyoto 
University) . 

Members of the group also spoke at a 
number of hospitals and research centers, 
including the Boston University Marine 
Program; Max-Planck Institut, Tubin- 

gen, Germany; M. D. Anderson Hospital; 
National Cancer Center, Tokyo; Na- 
tional Institutes of Health; Roche Insti- 
tute for Molecular Biology ; Roswell Park 
Memorial Institute ; and Scripps Institute 
for Medical Research. 

Members of the Department took part 
in meetings of a number of learned so- 
cieties, including, in addition to those al- 
ready mentioned, the American Associa- 
tion of Anatomists, American Association 
for the Advancement of Science, Amer- 
ican Society of Biological Chemists, 
American Society for Cell Biology, Amer- 
ican Society of Zoologists, Biophysical 
Society, Federation of American Societies 
for Experimental Biology, Japanese So- 
ciety for Molecular Biology, Maryland 
Academy of Sciences, National Academy 
of Sciences, New York Heart Association, 
Society for Developmental Biology, So- 
ciety of General Physiologists, Society 
for Neuroscience, and Tissue Culture 

Advisory and consultative services in- 
cluded membership on the editorial 
boards of Anales del Desarrollo, Devel- 
opmental Biology, In Vitro, Journal of 
Biological Chemistry, Journal of Cell 
Biology, Journal of Embryology and Ex- 
perimental Morphology, Journal of Ex- 
perimental Zoology, Excerpta Medica 
(section on Human Developmental Biol- 
ogy), Current Topics in Developmental 
Biology, and Quarterly Review of Biol- 

Members of the staff also acted in 
these capacities: Member of the Board of 
Scientific Overseers, Jackson Laboratory; 
Trustee, President, and Director, Marine 
Biological Laboratory; Member of the 
Corporation, Woods Hole Oceanographic 
Institution; Vice President, Maryland 
Academy of Sciences ; and Member of the 
Executive Committee for the Fourth In- 
ternational Conference on Congenital 

Other posts occupied by members of 
the Department include the following: in 
the American Institute of Biological Sci- 
ences, member, Council of Past Presi- 



dents; in the American Society for Cell 
Biology, member of the Council; in the 
Institut de la Vie, Chairman, World 
Committee on the Formative Weeks of 
Human Life ; in the International Society 
of Developmental Biologists, Secretary; 
in the National Academy of Sciences, 
member, Committee on Science and Pub- 
lic Policy, and member, Report Review 
Committee; in the National Institutes of 
Health, member, Board of Scientific 
Counselors, National Institute of Child 
Health and Human Development, mem- 
ber, Gerontology Research Center Re- 
sources Advisory Committee, member, 
Cell Biology Study Section, and member, 
Physiological Chemistry Study Section; 
at the National Science Foundation, 
member of the Advisory Committee for 
Biological and Medical Sciences; in the 
Society for Developmental Biology, mem- 
ber of Executive Committee; and in the 
Tissue Culture Association, Chairman, 
Honor B. Fell Division. Members of the 
Department served on a number of visit- 
ing committees including those for the 
Departments of Biology at Brookhaven 
National Laboratory, Harvard Univer- 
sity, Massachusetts Institute of Technol- 
ogy, and Princeton University. 

Formal teaching was again largely con- 
fined to the Johns Hopkins Department 
of Biology, but during the year lectures 
were offered in other departments of the 

University as well, among them Biophys- 
ics, Engineering, Laryngology and Otol- 
ogy, and Pediatrics. 


The roster of speakers at the seminars 
organized by the Department to serve all 
those working in developmental biology 
in the region included Joyce Benjamins 
(Johns Hopkins University School of 
Medicine) ; George Brownlee (Medical 
Research Council, Cambridge, England) ; 
Annette W. Coleman (Brown Univer- 
sity) ; Norman Davidson (California In- 
stitute of Technology) ; Gary Felsenfeld 
(National Institutes of Health) ; Richard 
Gelinas (Harvard University) ; K. Gross 
(Massachusetts Institute of Technol- 
ogy) ; Richard Hallberg (Cornell Univer- 
sity) ; Richard Hendler (National Insti- 
tutes of Health) ; W. Hennig (Zurich) ; 
Thomas Hollinger (Purdue University) ; 
Philip Leder (National Institutes of 
Health) ; Harvey Lodish (Massachusetts 
Institute of Technology) ; John J. Mar- 
chalonis (Walter and Eliza Hall Institute 
of Medical Research, Melbourne, Austra- 
lia) ; Reiji Okazaki (Nagoya Univer- 
sity) ; Ernest Page (University of Chi- 
cago) ; John Paul (Beatson Institute for 
Cancer Research, Glasgow, Scotland) ; 
and Allan Schneider (National Institutes 
of Health). 


Barkley, D. S., see Edds, M. V., Jr. 

Bladder, A. W., see Dawid, I. B. 

Brown, D. D., The evolutionary solution to the 
antibody dilemma. In Molecular Genetics 
and Developmental Biology, M. Sussman ed., 
Prentice-Hall, Englewood Cliffs, New Jersey, 
pp. 101-125. 1972. 

Brown, D. D., see also Gage, L. P.; Sugimoto, 
K.; Suzuki, Y. 

Dawid, I. B., Evolution of mitochondrial DNA 
sequences in Xenopus. Develop. Biol., 29, 
139-151, 1972. 

Dawid, I. B., and A. W. Blackler. Maternal and 
cytoplasmic inheritance of mitochondrial 
DNA in Xenopus. Develop. Biol., 29, 152-161, 

Dawid, I. B., Mitochondrial protein synthesis. 
In Mitochondria I Biomembranes, S. G. van 
den Bergh et al., eds. North Holland. Amster- 
dam, pp. 35-51, 1972. 

Dawid. I. B., see also Wu. G.-J. 

DeHaan, R. L., and H. G. Sachs, Cell coupling 
in developing systems: the heart-cell para- 
digm. Current Topics in Develop. Biol., 7, 
193-228, 1972. 

DeHaan, R. L., Embryology of the heart. In 
The Heart, J. W. Hurst and R. B. Logue, eds., 
McGraw-Hill, New York, pp. 7-21. 1973. 

DeHaan, R. L., see also McDonald, T. F. ; 
Nason, A.; Sachs, H. G. 

Ebert, J. D., Report of the Director of the Ma- 



rine Biological Laboratory for 1971. Biol. 

Bull, US, 7-13. 
Ebert, J. D., The molecular basis of gene ex- 
pression during devlopment (in Russian). 

Yearbook, Science and Humanity, 1971-1972, 

pp. 240-253. 
Ebert, J. D., Conclusions. Strategie Institution- 

nelle. Les Cahiers de L'Institut de la Vie, 

number 36, 297-312, 1973. 

Ebert, J. D., A. G. Loewy, R. S. Miller, and 
H. A. Schneiderman, Biology, Holt, Rinehart 
and Winston, Inc. New York, 798 pp., 1973. 

Ebert, J. D., see also McDonald, T. F. 

Edds, M. V., Jr., D. S. Barkley, and D. M. Fam- 
brough, Genesis of neuronal patterns. Neuro- 
sci. Res. Prog. Bull, 10, 254-367, 1972. 

Edidin, M., and D. Fambrough, Fluidity of the 
surface of cultured muscle fibers: rapid lateral 
diffusion of marked surface antigens. J. Cell 
Biol, 57, 27-37, 1973. 

Fambrough, D. M., see Edds, M. V., Jr.; Edidin, 
M.; Hartzell, H. C; Rash, J. E. 

Gage, L. P., Y. Suzuki, and D. D. Brown, Spe- 
cific hybridization of the silk fibroin genes in 
Bombyx mori. In Molecular Genetics and 
Developmental Biology, M. Sussman, ed., 
Prentice-Hall, Englewood Cliffs, New Jersey, 
pp. 127-137, 1972. 

Gage, L. P., see also Suzuki, Y. 

Gardner, E., The early development of the 
shoulder joint in staged human embryos. 
Anat. Rec, 175, 503, 1973. 

Gardner, E., see also O'Rahilly, R. 

Gershfeld, N. L., and R. E. Pagano, Physical 
chemistry of lipid films at the air-water inter- 
face. I. Intermolecular energies in single- 
component lipid films. /. Phys. Chem., 76, 
1231-1237, 1972. 

Gershfeld, N. L., and R. E. Pagano, Physical 
chemistry of lipid films at the air-water inter- 
face. III. The condensing effect of cholesterol. 
A critical examination of mixed-film studies. 
/. Phys. Chem., 76, 1244-1249, 1972. 

Gershfeld, N. L., see also Pagano, R. E. 

Hartzell, H. C, and D. M. Fambrough, Acetyl- 
choline receptors: distribution and extrajunc- 
tional density in rat diaphragm after denerva- 
tion correlated with acetylcholine sensitivity. 
J. Gen. Physiol, 60, 248-262, 1972. 

Hartzell, H. C, and D. M. Fambrough, Acetyl- 
choline receptor production and incorporation 
into membranes of developing muscle fibers. 
Develop. Biol, 30, 153-165, 1973. 

Honjo, T., and R. H. Reeder, Xenopus mulleri 
ribosomal genes are repressed in X. laevis X 
X. mulleri hybrid frogs. Fed. Proc, Fed. 
Amer. Soc. Exp. Biol, 32, 656 (Abstract), 1973. 

Loewy, A. G., see Ebert, J. D. 

McDonald, T. F., and R. L. DeHaan, Ion levels 
and membrane potential in chick heart tissue 
and cultured cells. J. Gen. Physiol, 61, 89-109, 

McDonald, T. F., H. G. Sachs, C. W. Orr, and 
J. D. Ebert, Multiple effects of ouabain on 
BHK cells. Exp. Cell Res., 74, 201-206, 1972. 

McDonald, T. F., see also Sachs, H. G. 

Miller, I. R., see Pagano, R. E. 

Miller, R. S., see Ebert, J. D. 

Muecke, E. C, see O'Rahilly, R. 

Nason, A. and R. L. DeHaan, The Biological 
World, John Wiley and Sons, New York, 736 
pp. 1973. 

Nojima, T., see Yoshikawa-Fukada, M. 

O'Rahilly, R., Guide to the staging of human 
embryos. Anat. Anz., 130, 556-559, 1972. 

O'Rahilly, R., The early development of the 
hypophysis cerebri in staged human embryos. 
Anat. Rec, 175, 511, 1973. 

O'Rahilly, R., and E. Gardner, The initial ap- 
pearance of ossification in staged human em- 
bryos. Amer. J. Anat., 134, 291-307, 1972. 

O'Rahilly, R., and E. C. Muecke, The timing 
and sequence of events in the development of 
the human urinary system during the embry- 
onic period proper. Z. Anat. Entwicklungs- 
gesch., 138, 99-109, 1972. 

O'Rahilly, R., see also Tucker, J. A. 

Orr, C. W., see McDonald, T. F. 

Ozawa, E., The role of calcium ion in avian 
myogenesis. Biol. Bull, 143, 431-439, 1972. 

Pagano, R. E., and N. L. Gershfeld, Physical 
chemistry of lipid films at the air-water inter- 
face. II. Binary lipid mixtures. The princi- 
ples governing miscibility of lipids in surfaces. 
/. Phys. Chem., 76, 1238-1243, 1972. 

Pagano, R. E., and N. L. Gershfeld, A Milidyne 
film balance for measuring intermolecular 
energies in lipid films, /. Colloid Interface 
Sci., 41, 311-317, 1972. 

Pagano, R. E., J. M. Ruysschaert, and I. R. 
Miller, The molecular composition of some 
lipid bilayer membranes in aqueous solution. 
/. Membrane Biol. 10, 11-30, 1972. 

Pagano, R. E., and T. E. Thompson, Spherical 
lipid bilayers: Surface tension and contact 
angle measurements. /. Colloid Interface Sci., 
43, 209-210, 1973. 

Pagano, R. E., see also Gershfeld, N. L. 

Polinger, I. S., Identification of cardiac myo- 
cytes in vivo and in vitro by the presence of 
glycogen and myofibrils. Exp. Cell Res., 76, 
243-252, 1973. 

Polinger, I. S., Growth and DNA synthesis in 
embryonic chick heart cells, in vivo and in 
vitro. Exp. Cell Res., 76, 253-262, 1973. 



Rash, J. E., and D. Fambrough, Ultrastructural 
and electrophysiological correlates of cell 
coupling and cytoplasmic fusion during myo- 
genesis in vitro. Develop. Biol., 30, 166-186, 

Reeder, R. H., Transcription of chromatin by E. 
coli RNA polymerase. Biophys. J., 13, 189a, 

Reeder, R. H., and R. G. Roeder, Ribosomal 
RNA synthesis in isolated nuclei. J. Mol. 
Biol, 67, 433-441, 1972. 

Reeder, R. H., see also Honjo, T. 

Roeder, R. G., RNA polymerases during am- 
phibian development, In Molecular Genetics 
and Developmental Biology, M. Sussman, ed., 
Prentice-Hall, Englewood Cliffs, New Jersey, 
pp. 163-172. 1962. 

Roeder, R. G., see also Reeder, R. H. 

Ruysschaert, J. M., see Pagano, R. E. 

Sachs, H. G., and R. L. DeHaan, Embryonic 
myocardial cell aggregates: Volume and pulsa- 
tion rate. Devel. Biol, 30, 233-240, 1973. 

Sachs, H. G., and T. F. McDonald, Membrane 
potentials of BHK (Baby Hamster Kidney) 
cell line: Ionic and metabolic determinants. 
J. Cell Physiol, 80, 347-358, 1972. 

Sachs, H. G., T. F. McDonald, and R. L. De- 
Haan, Tetrodotoxin sensitivity of cultured 
embryonic heart cells depends on cell inter- 
actions. /. Cell Biol, 56, 255-258, 1973. 

Sachs, H. G., M. Springer, and T. F. McDonald, 
Cytochalasin B and embryonic heart cells. 
Anat. Rec, 175, 433 (Abstract), 1973. 

Sachs, H. G., see also DeHaan, R. L.; Mc- 
Donald, T. F. 

Schneiderman, H., see Ebert, J. D. 

Springer, M., see Sachs, H. G. 

Sugimoto, K., and D. D. Brown, Evolution of a 
family of genes: Comparison of 5S DNA 
from Xenopus laevis and Xenopus mulleri. 
Fed. Proc, Fed. Amer. Soc. Exp. Biol, 32, 663 
(Abstract), 1973. 

Suzuki, Y., The messenger RNA and the genes 
for silk fibroin — A model system for the study 
of differential gene action in eukaryotes. Ab- 
stracts of papers presented at the 1972 Molec- 
ular Biology Meeting of Japan, pp. 62-63, 

Suzuki, Y., Studies on the mechanisms of gene 
expression in development. I. Isolation and 
identification of the messenger RNA for silk 
fibroin from Bombyx mori. Protein, Nucleic 
Acid and Enzyme, 17, 79-91, 1972. 

Suzuki, Y., The genes in development, Science 
oj Japan, 42, 590-599, 1972. 

Suzuki, Y., Studies on the mechanisms of gene 
expression in development. II. The genes for 
silk fibroin in Bombyx mori. Protein, Nucl 
Acid, Enzyme, 18, 673-685, 1973. 

Suzuki, Y., L. P. Gage, and D. D. Brown, The 
genes for silk fibroin in Bombyx mori. J. Mol. 
Biol, 70, 637-649, 1972. 

Thompson, T. E., see Pagano, R. E. 

Tucker, J. A., and R. O'Rahilly, Observations 
on the embryology of the human larynx. Ann. 
Otol Rhinol Laryngol, 81, 520-523, 1972. 

Wu, G.-J., and I. B. Dawid, Purification and 
properties of mitochondrial deoxyribonucleic 
acid dependent ribonucleic acid polymerase 
from ovaries of Xenopus laevis. Biochemistry, 
11, 3589-3595, 1972. 

Yoshikawa-Fukada, M., and T. Nojima, Bio- 
chemical characteristics of normal and virally 
transformed mouse cell lines. J. Cell Physiol, 
80, 421-430, 1972. 


Year Ended June 30, 1973 
(including those whose services began or ended during the year) 

Research Staff 

Donald D. Brown, Biochemistry 

Igor B. Dawid, Biochemistry 

Robert L. DeHaan, Experimental Embry 

James D. Ebert, Director 
Douglas M. Fambrough, Biochemistry 
Richard E. Pagano, Biophysics 
Ronald H. Reeder, Biochemistry 

Assistant Investigator 
Peter J. Stambrook 

Research Associates (extramural) 

Bent G. Boving, Detroit, Mich. 

Louis B. Flexner, Philadelphia, Pa. 

Ernest Gardner, Davis, Cal. 

Arthur T. Hertig, Boston, Mass. 

Irwin R. Konigsberg, Charlottesville, Va. 

H. Robert Misenhimer, Chicago, 111. 

Ronan O'Rahilly, Detroit, Mich. 


Joseph F. Albright, Senior Fellow of Car- 
negie Institution of Washington 



Dana Carroll, Fellow of Carnegie Institu- 
tion of Washington 

Robert J. Hay, Fellow of Carnegie Institu- 
tion of Washington 

Tasuku Honjo, Fellow of Carnegie Institu- 
tion of Washington 

Ivan Horak, Fellow of Carnegie Institution 
of Washington 

Paul Lizardi, Fellow of Jane Coffin Childs 
Memorial Fund for Medical Research 

Terence F. McDonald, Fellow of Carnegie 
Institution of Washington 

Keiko Ozato, Fellow of Carnegie Institution 
of Washington 

Aileen K. Ritchie, Fellow of Muscular Dys- 
trophy Association of America, Inc. 

Howard G. Sachs, Fellow of Carnegie Insti- 
tution of Washington 

Kazunori Sugimoto, Fellow of Carnegie In- 
stitution of Washington 

Yoshiaki Suzuki, Fellow of Carnegie Insti- 
tution of Washington 

Peter K. Wellauer, International Fellow of 
the U.S. Public Health Service 

Robert Williamson, Senior Fellow of Car- 
negie Institution of Washington 

Guang-Jer Wu, Fellow of Carnegie Institu- 
tion of Washington 

Masako Yoshikawa-Fukada, Fellow of Car- 
negie Institution of Washington 


Sandra L. Biroc, Graduate, Johns Hopkins 

Peter Devreotes, Graduate, Johns Hopkins 

Thomas E. Durr, Undergraduate, Johns 
Hopkins University 

Scott Gilbert, Graduate, Johns Hopkins 

H. Criss Hartzell, Graduate, Johns Hopkins 

Carol Kaushagen, Graduate, Johns Hopkins 

Caroline R. Krueger, Undergraduate, Johns 
Hopkins University 

Dennis E. Leister, Graduate, Johns Hop- 
kins University 

Linda Moyzis, Undergraduate, Johns Hop- 
kins University 

Calvin E. Plitt, Undergraduate, Johns Hop- 
kins University 

Ralph H. Stern, Graduate, Johns Hopkins 

Camilla Velez, Undergraduate, Johns Hop- 
kins University 

Visiting Investigators and Extramural 

Antonie W. Blackler, Ithaca, N.Y. 
George Brownlee, Cambridge, England 
John W. Chase, Boston, Mass. 
A. Chikushi, Tokyo, Japan 
Hayden G. Coon, Bethesda, Md. 
Daniel B. Drachman, Baltimore, Md. 
Michael Edidin, Baltimore, Md. 
J. W. S. Harris, London, England 
Edward C. Muecke, New York, N.Y. 
T. Nojima, Kyoto, Japan 
Mary Lou Pardue, Cambridge, Mass. 
Robert Roeder, St. Louis, Mo. 
Glenn C. Rosenquist, Baltimore, Md. 
S. Satyamurti, Baltimore, Md. 
J. E. Sisken, Lexington, Ky. 
Max Springer, Zurich, Switzerland 
Kenneth D. Tartof, Philadelphia, Pa. 
John A. Tucker, Philadelphia, Pa. 
Ifermann K. Wolf, Halifax, Nova Scotia, 

Clerical and Technical Staff 

Aldon L. Alston, Recorder 

Elaine S. Asch, Technician 

James H. Blackwell, Custodian 

Julia L. Certain, Librarian (part time) 

William J. Cleary, Recorder 

Laura L. Costello, Technician 

William H. Duncan, Senior Technician 

Ernestine V. Flemmings, Laboratory Helper 

Richard D. Grill, Photographer 

Richard F. Handler, Technician 

Norman Handy, Custodian 

Virginia S. Hicks, Laboratory Helper 

James B. Iglehart, Recorder 

Eddie D. Jordan, Senior Technician 

Nancy J. Joseph, Senior Technician 

June M. Lam, Secretary 

Alice H. Mabin, Laboratory Helper 

Thomas F. Malooly, Business Manager 

Thomas F. Miller, Custodian 

Joyce I. Oaks, Laboratory Helper 

Alex Ovcharek, Custodian 

John Pazdernik, Jr., Building Engineer 

Betty Lou Phebus, Bookkeeper-Clerk 

Diane M. Prather, Technician 

Margaret J. Proctor, Secretary 

Martha L. Rebbert, Technician 

Adrienne Robinette, Secretary 

Phillip E. Rose, Custodian 

Carroll Sears, Custodian 


John J. Sexton, Custodian Student Assistants 

Bessie H. Smith, Laboratory Helper Jan c Culbertson, Southampton College 

Delores V. Somerville, Technician Andrew Scott Klein, Duke University 

Leroy G. Williams, Recorder Nancy Kolzak, Smith College 

John L. Wiser, Machinist Andrew Cader, Park School 

Hale Observatories 

Operated by Carnegie Institution of Washington 
and California Institute of Technology 

Pasadena, California 

Horace W. Babcock 

J. Beverley Oke 
Associate Director 


Horace W. Babcock, Chairman 

J. Beverley Oke, V ice-Chairman 

Robert B. Leighton 

Guido Munch 

George W. Preston 

Allan Sandage 

Leonard Searle 

Harold Zirin 

Carnegie Institution oj Washington Year Book 72, 1972—1973 


Introduction 101 

Observing Conditions 103 

Physics of the Sun 104 

Active regions ........ 104 

Rotation and large-scale velocity fields 105 

Large-scale magnetic fields .... 105 

Polar magnetic fields 105 

Solar limb shift 105 

Solar flares 106 

Solar System Studies 106 

Saturn 106 

Titan 107 

Uranus 107 

Comet 1972h 108 

Stellar Spectroscopy and Photometry . 108 

White dwarfs . 108 

Halo population 108 

Emission-line star MWC 349 .. . 109 

Spectrophotometry of B-type stars . . 109 

Mercury lines in Si stars 110 

Helium-weak B stars 110 

Early A stars 110 

Photometry of Pleiades stars ... 110 

Stellar magnetism Ill 

Photoelectric measurements of stellar 

magnetic fields 112 

Magnetic stars 112 

Stellar Chromospheres 112 

Stellar cycles 112 

Absolute magnitudes of late-type giants 112 

Variable Stars 112 

Infrared Sources 113 

Occupations of infrared stars ... 113 
Size measurements of extragalactic ob- 
jects 113 

2-20 micron band 114 

Millimeter-Wave Observations ... 115 

1-mm photometry 115 

Interstellar Gas and Gaseous Nebulae . 116 
Interferometry of nebulae with low sur- 
face brightness 116 

Pulsars 116 

X-Ray Sources 116 

Cygnus X-l 116 

Hercules X-l 117 

Supernovae 118 

Supernova search 118 

Astrophysical studies 118 

Review of spectra 119 

Galaxies 119 

Redshifts of fainter galaxies in the 

Shapley-Ames Catalogue ... 119 
Relative distance to nearby E and SO 

galaxies 119 

Motion of the sun relative to the 

nearby galaxies 120 

Markarian galaxies 120 

History of star formation in galaxies . 121 

Bright spiral galaxies 121 

Synthesis of the stellar population of 

galaxies 123 

Photometry of OJ 287 and M82 . . 123 

Chains of galaxies 123 

Radio sources and spiral galaxies . . 125 
Multiple interacting galaxies . . . 126 
Deep photography of galaxies with con- 
nected companions 126 

Superposition printing 126 

Nebulosity near 3C120 and BL La- 

certae 126 

Redshifts in small groups of galaxies . 126 

Compact galaxies 126 

Radio Sources 127 

Identification of 3C sources .... 127 

Radio source structure 127 

Search for the radio source near 

Cygnus X-3 127 

Quasars 127 

Quasar surveys 127 

Spectroscopic observations .... 127 

Quasars as events in nuclei .... 128 
Association of quasars and clusters of 

galaxies 128 

Small quasars at the center of N 

galaxies 128 

Observational limits to nonvelocity 

redshifts for the mini-quasars in N 

galaxy nuclei 129 

Properties of quasars 129 

Angular diameters of quasars . . . 130 

Observational Cosmology 131 

Colors of E galaxies as an indication 

of age 131 

Absolute magnitude of first-ranked E 

galaxies in clusters and groups . . 132 

Distant clusters 133 

Radio galaxy source counts .... 134 

Theoretical Studies 134 

Chemical history of galaxies . . . 134 

Diffuse x-ray background 135 

Magnetic field of the Crab Nebula . 135 
Penumbral waves and umbral flashes in 

sunspots 135 

Photospheric oscillations 135 

White dwarfs 136 

Magnetic field of 53 Camelopardalis . 136 

Zeeman-broadened line profiles . . . 136 

Instrumentation 136 

Prime-focus corrector 136 

SIT-vidicon area photometer . . . 137 

200-inch Cassegrain spectrograph . . 137 

Prime-focus image tube 137 

Coude image tube 137 

Data system for 150-foot tower ... 138 

Vacuum telescope 138 

Guest Investigators 138 

Las Campanas 149 

Astroelectronics Laboratory .... 151 

Digital image recorder 151 

Computer systems , . . 151 

Electronic instrument utilization . . 152 

Photographic Laboratory 152 

100-Inch du Pont Telescope .... 153 

Mechanical parts and building . . . 154 

Optics 154 

Auxiliary instruments 154 

Control system 155 

Computer and data system .... 155 

Las Campanas Observatory .... 155 

Site development 156 

Bibliography 156 

Staff and Organization 161 


Twenty-five years have elapsed since for the radio galaxy 3C295; of the first 

the 200-inch Hale Telescope was dedi- optical identification of a quasar by 

cated on Palomar Mountain. This great Sandage and Matthews in 1960; and of 

instrument has been an unqualified sue- the elucidation of quasar spectra by 

cess. It has fully justified the hopes of Schmidt in 1963, with recognition of the 

its planners in expanding man's concep- enormous redshifts of some of them. 
tion of the universe and it has partici- The faith of George E. Hale in the 

pated in opening a new array of challeng- future of astronomy as he gauged it in 

ing problems, some of which are of a kind 1928 and the generosity of the Rocke- 

and magnitude quite unimagined at the feller sources that made possible the 

time the telescope was built. construction of the 200-inch telescope 

If this instrument had done nothing have thus been amply justified. The 
else, its construction would have been present generation of astronomers can 
abundantly justified by the progress it take satisfaction in the fact that 45 years 
has made possible on the cosmological after the 200-inch was funded and 25 
problem: the determination of the scale years after it was dedicated, the optical 
of the universe, the refinement of the exploration of the discernible universe 
Hubble constant that defines the dis- seems to offer more opportunities for new 
tance-redshift relationship, and the plac- knowledge than ever before. This tele- 
ing of limits on the deceleration param- scope remains our most powerful instru- 
eter. In the hands of capable observers, ment for conducting optical investiga- 
te 200-inch telescope has produced tions, although it will soon be joined, at 
thousands of direct photographs, tens of various other observatories throughout 
thousands of photometric measurements, the world, by a number of other tele- 
and more than 20,000 spectrograms of scopes that approach it in light-gathering 
stars and extragalactic objects. Such power. 

data have provided the basis for the In Hale's 1928 article in Harper's Mag- 

refinement of the distance scale by Baade azine, in which he proposed the 200-inch 

and his associates, and for correcting the telescope, he referred first to the great 

expansion rate of the universe (the Hub- light-gathering power that would pro- 

ble constant) by a factor of 10 from its vide an immense gain for exploring re- 

1935 value to approximately 50 km s" 1 mote space and the structure of the uni- 

Mpc" 1 , as recently determined by Sand- verse, then to the value of the instru- 

age and Tammann. While the decelera- ment for studying stellar evolution; 

tion parameter of the universal expan- thirdly he mentioned the "greatest of 

sion, q , cannot yet be specified with any these problems: that of determining with 

certainty, the extension of observations certainty the successive stages in the 

to still fainter clusters of galaxies, to- development of spiral nebulae [galaxies] 

gether with the application of evolution- . . . ." It is no accident that this year's 

ary corrections, promises steady progress Report shows the results of increasing 

toward this goal as well. attention to problems of the large subject 

In this brief introduction, one can of galaxies — their structure, evolution, 

scarcely even attempt to list the other and relationship to quasars. Star forma- 

major investigations that the 200-inch tion within galaxies, age effects, radio 

telescope has made possible. Mention emission, and high-energy physical proc- 

must be made, however, of Minkowski's esses are increasingly rewarding subjects 

1959 measurement of a redshift of 0.46 of study. 




Investigations of the distribution of 
galaxies and clusters of galaxies in the 
universe are proceeding. Advances in our 
understanding of the general expansion 
are promised by new observations of red- 
shifts and more precise distance cali- 
brations, as well as through statistical 
studies. In terms of steady-state cos- 
mology, Schmidt has analyzed counts of 
strong radio sources in the Third Cam- 
bridge Catalogue for which optical iden- 
tifications have been made and redshifts 
observed. His method allows the indi- 
vidual properties of the identified sources 
to be represented in the predicted source 
counts. The predicted curve falls below 
the observed curve by a factor of 5 at the 
lower radio flux densities. His conclusion 
is that radio source counts are incom- 
patible with the steady-state theory. 

Kristian has extended the search for 
galaxies underlying quasars by identify- 
ing the conditions under which such ga- 
laxies should be detectable. Critical 
examination of 200-inch photographs of 
26 quasars gave results in good agreement 
with the predictions. The evidence sup- 
ports the increasingly general view that 
quasars occur in the nuclei of galaxies. 

Sandage's investigations of N galaxies 
show that a typical object of this type is 
a giant elliptical galaxy with a mini- 
quasar in its nucleus; furthermore, that 
the magnitude-redshift relationship for 
the N galaxies leaves little, if any, margin 
for a possible noncosmological component 
of the redshift. 

Increasing attention is being given to 
the history of star generation in galaxies 
and to the processes that trigger large- 
scale bursts of star formation. Studying 
ring galaxies, for example, Searle and 
Sargent have obtained from multichannel 
scans evidence that the energy distribu- 
tions of the rings are closely similar to 
those of star clusters a few hundred mil- 
lion years old. The same age is indicated 
for these galaxies from the measured ex- 
pansions of the rings. Searle and Sargent 
believe that these galaxies are the first 
truly young extragalactic systems to be 

recognized. They also conclude that the 
outer reaches of certain paired galaxies, 
distorted by mutual interaction, consist 
of predominantly young stars. 

The concept that density waves or 
compressional effects in galaxies acceler- 
ate the formation of stars from the in- 
terstellar material is finding increased 
support not only in situations such as the 
foregoing, but in the spiral structure and 
nuclear regions of galaxies in general. 
Van der Kruit, examining the high-reso- 
lution radio brightness distributions 
within galaxies, obtained at Westerbork, 
found that the radio-brightness distribu- 
tions were displaced toward the inner 
edges of the optical spiral arms, thus 
confirming the concept of compressed re- 
gions in the density-wave picture of 
spiral structure. The radio-spiral struc- 
ture is believed to result from compres- 
sion of the magnetic field and relativistic 
electrons in the large-scale "base disk" 
of a galaxy. Star formaton is triggered 
in the density-wave compression regions 
in a manner governed by the degree of 
compression. Van der Kruit concluded 
that the relationship of specific angular 
momentum to mass is the governing pa- 
rameter. He found that, for a particular 
mass, galaxies with strong compressional 
effects have a lower specific angular mo- 
mentum than galaxies with weak com- 
pression. Statistically, this results in a 
larger central concentration of mass for 
strong compression galaxies, which ex- 
plains the result that such galaxies 
generally have better developed nuclei. 
Other evidence suggests that the char- 
acteristics of a galaxy such as the visi- 
bility of its spiral structure and the de- 
gree of concentration toward the nucleus 
are not determined by the details of the 
density generating mechanism but rather 
by the characteristics of the disk, that is, 
by the specific angular momentum and 
the mass. Given these governing param- 
eters, the compressional strength is de- 
termined just as is the form of the 
rotation curve as measured by optical 
spectroscopy. These conclusions were re- 



inforced by observations at Palomar of 
the rotation curve for NGC 4321. 

Compact chains of galaxies, particu- 
larly apparently linear chains of a few 
members, have been under investigation 
for some time by Sargent. Studies of the 
colors of member galaxies provide no 
indications that these are young systems, 
although typical dissolution times com- 
puted by others are about 5 X 10 s years 
— a relatively short interval on the cos- 
mic scale. A new and comparatively 
simple explanation advanced by Sargent, 
which has now found support from com- 
puter analysis by himself and Turner, 
is the idea that apparent linear forma- 
tions in compact groups of galaxies repre- 
sent states (nearly coplanar) in the 
normal dynamical evolution of small 
clusters that on the average are spherical. 

Numerous attempts have been made to 
identify optical counterparts of recently 
discovered galactic x-ray sources. One 
such object is the binary BD+34°3815, 
which is a weak variable radio source and 
is identified with Cygnus X-l. The star 
is a 5.6-day spectroscopic binary of which 
the primary optical component is of 
spectral type BO lb, with velocity ampli- 
tude of 74 km s" 1 . Brucato and Kristian, 
observing the star spectroscopically at 
Mount Wilson, discovered that the weak 
emission feature at A4686 (He II) had a 
velocity amplitude nearly twice as great 
as that of the B star and that it varied 
in anti-phase; no other evidence of a 
secondary spectrum is seen. After ana- 
lyzing the evidence, Brucato and Kristian 
concluded that one possibility is that the 
secondary component of the binary is 
a black hole, the x rays being produced 
by the thermalization of gravitational 
energy of matter falling in the vicinity of 
the black hole. Other interpretations, 
however, still remain open. 

The brightest supernova in many 

years, 1972e, was discovered by Kowal in 
NGC 5253 as reported in Year Book 71, 
p. 677. Collaboration by Kirshner, Oke, 
Penston, and Searle has resulted in a 
report on the spectra of supernovae that 
describes and interprets the scans of 
several such objects of both Types I and 
II and includes a comprehensive series of 
observations of SN 1972e covering 230 
days. Spectrum scans of this object are 
expected to be obtainable for at least one 
additional year. Important new conclu- 
sions regarding supernova spectra have 
been reached from a study of the new 

Observations with the telescopes at 
Mount Wilson and Palomar have never 
been limited to the visible region of the 
spectrum. Significant infrared observa- 
tions were conducted locally, long before 
the advent of sensitive modern detectors. 
C. G. Abbot used the 100-inch Hooker 
reflector as early as 1924 for determining 
infrared energy curves of stars with his 
special radiometer. P. W. Merrill made 
important observations of stellar spectra 
in the infrared, using Eastman Z plates 
in the 1930s. The infrared solar spectrum 
was measured in detail, and its photo- 
graphic limit was extended to 13,495 A in 
1935. The past decade has seen remark- 
able advances in the 1 to 20 micron range 
through the application of lead-sulfide 
and mercury-doped germanium detectors 
in the hands of Neugebauer, Becklin, and 
their associates. Now with the 200-inch 
an extension of photometry at a wave- 
length of 1 millimeter has been effected 
by Harvey, Werner, Elias, and Gatley. 
They have found it feasible to observe at 
this wavelength during extended periods 
of twilight, thus utilizing the telescope 
for six hours or more per day when it 
would otherwise be idle. Objects studied 
included planets, galactic H II regions, 
and the QSO 3C273. 


At Mount Wilson the 100-inch tele- 
scope was used for observations on 203 

complete nights and 91 partial nights 
for a total of 2566 hours. Sixty-nine 


TABLE 1. 200-Inch Observations 

Number of 




























































nights were lost because of the weather. 
Total precipitation was 40.5 inches, with 
63 inches of snow. 

The 200-inch Hale Telescope at Palo- 
mar Mountain was used for a total of 
2917.6 hours, as shown in Table 1. 

The figures include a substantial amount 
of twilight time that was used for obser- 
vations not requiring a dark sky. Total 
precipitation at Palomar was 38 inches, 
with 85 inches of snow. 


Observations of the sun on a daily 
basis continue at Mount Wilson under 
the supervision of Howard. Between 
June 1, 1972, and May 31, 1973, the fol- 
lowing observations were obtained at 
Mount Wilson: 

Direct photographs 293 

Ha spectroheliograms, 30-foot focus 527 

K2 spectroheliograms, 30-foot focus 535 

Full-disk magnetograms 389 
Integra ted-light magnetic-field 

measurements 190 

Sunspot drawings 290 

Observations were made on a total of 
318 days. 

The analyses of the digital data from 
the magnetograms and the integrated- 
light magnetic-field measurements are 
carried out on the Raytheon 704 digital 
computer. The magnetograms continue 
to be published monthly in the National 
Oceanic and Atmospheric Administration 
bulletin Solar Geophysical Data, and the 
computer-plotted synoptic charts con- 

tinue to be published in the Quarterly 
Bulletin on Solar Activity of the Interna- 
tional Astronomical Union. Partial sup- 
port for these observing programs comes 
from the National Aeronautics and Space 
Administration, the Office of Naval Re- 
search, and the Air Force Cambridge 
Research Laboratories. 

Active Regions 

From a study of active regions of the 
last few years, Howard and S. Edberg 
have concluded that chance eruptions of 
magnetic flux can explain the occasional 
renewal of activity in an active region. 
It is not necessary to assume that such 
resurgences of activity represent a prop- 
erty of an active region; magnetic flux 
erupts at the solar surface randomly with 
respect to existing active regions. This 
does not bear on the phenomenon of ac- 
tive longitudes, but it relates to the 
problem of the origin of solar active re- 



gions because it appears that whatever 
is the mechanism to produce the eruption 
of magnetic flux to form an active center, 
it does not affect the probability that 
another region will form at nearly the 
same location. 

Rotation and Large-Scale Velocity Fields 

The velocity data from more than 300 
full-disk magnetograms made in various 
spectrum lines within the last year have 
been studied by Howard. The Fe I and 
Fe II lines show identical equatorial ro- 
tational velocities, but the Fe II line 
shows a more rapid decline of angular 
velocity with latitude than does the neu- 
tral line. The redshift parameters for the 
two lines at the limb differ by nearly 
a factor of 2. The ionized line has the 
larger value. Variations with time of the 
equatorial velocity measured with the 
two lines are essentially identical. The 
average equatorial velocity measured 
during the year was significantly greater 
than for a comparable period five years 
ago. This may be associated with a lower 
level of solar activity, but further obser- 
vations are required to be certain of this 

Large-Scale Magnetic Fields 

From a study of the solar magnetic 
flux data from September 1966 to Sep- 
tember 1972, Howard has been able to 
draw conclusions about the average in- 
clination of magnetic lines of force and 
related matters. These observations con- 
sist of about 1200 individual digitized 
magnetograms. Over the whole period, 
the average distribution of total mag- 
netic fluxes (\F + \ -\- \F.\) over the vari- 
ous latitude zones into which the data 
were divided shows that in all latitude 
zones, and in the north and south sepa- 
rately, the fluxes west of the central 
meridian were greater than those in the 
east by roughly 1%. This indicates an 
average inclination of field lines of a few 
degrees in a direction to trail the rota- 
tion. About 90% of the magnetic flux is 

confined to latitudes below 40°, not 
taking account of projection effects. If 
one assumes radial field lines, 85% of the 
flux is equatorwarcl of 40°. In both 
hemispheres the preceding and following 
polarities are inclined toward each other 
by a few degrees. In the north, the 
following-polarity flux is inclined slightly 
to lead the rotation, and the preceding- 
polarity flux trails the rotation. In the 
south, where the total magnetic flux is 
significantly weaker than in the north, 
both polarities trail the rotation. 

Polar Magnetic Fields 

Over the past decade, the polar mag- 
netic fields of the sun have been weak 
and variable. At times the fields at the 
poles have shown a tendency to vary in 
unison. Howard has found that near the 
middle of 1971 the north heliographic 
polar fields changed sign to become con- 
sistently positive, first at lower latitudes, 
then above 70°. The south pole remained 
weak but predominantly positive until 
about April 1973, when it appeared to 
change sign. The north polar field is 
measured to be about 0.5 gauss. This 
polar field polarity reversal has occurred 
about three years later in the phase of 
the cycle than the polarity reversal of the 
preceding cycle. 

Solar Limb Shift 

With the 150-foot tower telescope, 
Hart made measurements of the center- 
to-limb variation in the wavelengths of 
spectral lines. Most lines, when observed 
at the limb of the sun, are shifted slightly 
to the red compared with their wave- 
lengths at the center of the disk. These 
shifts, which are generally of the order of 
a few milliangstroms, vary greatly from 
line to line. 

Hart then suggested a theoretical ex- 
planation for this effect, whose cause 
has been unknown since the shift was first 
observed over 60 years ago. According 
to his hypothesis, the apparent redshift 
toward the limb is the result of pressure- 



induced blue shifts of spectral lines com- 
ing from the center of the disk. (These 
pressure shifts are much smaller at the 
limb.) Detailed computations based on 
this theory predict a shift of 3.39 mA 
for the Na I doublet at AA6154, 6161, and 
a shift of 2.24 mA for the Na I doublet 
at AA5683, 5688. These results are in 
good agreement with observed values. 

Solar Flares 

A series of unusually intense flares took 
place in August 1972 and were exten- 
sively observed at Big Bear. Zirin and 
Tanaka completed an extensive study of 
the different observations, based on 
multiwavelength filter cinematograms, 
spectra, and radio and x-ray data. They 
found the principal cause of the flares to 
be the extreme shearing, due to sunspot 
motions, of magnetic fields along the 
neutral field line. The field on the sur- 
face was generally force-free, running 
parallel to the boundary of opposite 
polarities, while after each flare a loop 
system appeared above the neutral line, 
tracing out a lower energy potential sys- 
tem. The energy in the Ha radiation of 
the great flares of August 4 and August 
7, 1972, was found, on the basis of ex- 
tensive spectra reinforced by focal-plane 
observations, to be only 2 X 10 30 ergs, 
approximately equal to the energy ob- 
served in cosmic rays above 10 Mev from 
these flares. This number is substantially 
lower than previous estimates of optical 
flare emission but is probably a better 
value because of the high quality of the 
observations. Most of the emission from 
the flares came from small intense ker- 
nels, and most of the bright flare areas 
showed much weaker emission than these 
kernels. Just after the August 7 flare, a 
sharp discontinuity of radial velocity was 

detected near the neutral line by means 
of spectra, corresponding to a shear flow 
of several kilometers per second. This 
flow appears quite important in the flare 
phenomenon. On August 2, two flares 
were observed with a broad-band (15 A) 
filter centered on A3835. This filter shows 
phenomena in the low chromosphere very 
well. An impulsive flare was detected 
which produced a number of very rapid 
brightenings of individual points along 
the neutral line. Each point was about 
1 arc sec in diameter and had a lifetime 
of 5 or 10 sec. These brightenings are 
interpreted by Zirin and Tanaka as due 
to the dumping of energetic electrons at 
the feet of different lines of force coming 
from the flare, which is somewhat higher 
above the atmosphere. The actual emis- 
sion is in the Balmer line H 9 . This is the 
first time such rapid and small phe- 
nomena have been observed in a flare. 
Emission in this wavelength has been ob- 
served in other, less implusive flares, but 
in these it corresponds to a slower, 
broader band of emission, also produced 
by conductive heating from the flare. 

Zirin and Bernard Lazareff, a gradu- 
ate student, concluded another study of 
activity in a large sunspot group in 
August 1971. It was found that the 
major activity in this group coincided 
with the growth and westward motion of 
a new sunspot. So long as this new sun- 
spot plowed ahead through the solar at- 
mosphere, a number of flares occurred 
at its leading edge. A careful comparison 
was made between these flares and Type 
III radio emission observed at the Uni- 
versity of Colorado, showing that Type 
III bursts occurred only after flares in 
this part of the active region, whereas 
other flares in the sunspot group were un- 
able to produce the fast electrons that 
give rise to Type III bursts. 




The brightness of the equatorial belt 
of Saturn was measured by Neugebauer 

and Munch in the M-band (4.6-5.2 /mi) 
and through interference filters centered 
at Ai = 4.805 and A 2 = 4.720 ^m, with 



widths at half-peak transmission of 0.11 
/Am. An absorption band due to CH 3 D 
appears in the spectrum of Jupiter 
centered at A 2 (Year Book 70, p. 399; 
Beer et al, Science, 175, 1360, 1972), 
while the pass band at Ai is essentially 
free of the CH 3 D absorption. Thus it 
was hoped that a color Ci 2 = m(X 1 ) = 
m(A 2 ) would provide evidence for the 
presence or absence of CH 3 D in Saturn. 
If the transmission curves of the Ai and 
A 2 filters are folded into spectral scans 
(A/AA = 180) of Jupiter and Mars 
(Year Book 71, p. 657), a color excess 
C 12 (Mars) — d 2 ( Jupiter) = +0.16 mag 
is obtained. The photometry of Saturn 
and of the moon, a Tauri, a Orionis, and 
(3 Pegasi was carried out at the Casse- 
grain focus of the Hale telescope on 
November 23, 1972. The brightness tem- 
perature of Saturn in the M-band, 
measured over a central circular area of 
7.5 arc sec diameter, is 180° K, and its 
color excess with respect to that of the 
moon is Ci 2 (moon) — Ci 2 (Saturn) = 
+0.08 mag. This value equals that ex- 
pected for blackbodies at the observed 
M-brightness temperatures, namely, 180° 
K and 245° K for Saturn and the moon, 
respectively. The color measurement 
therefore does not provide evidence for 
the presence of CH 8 D absorption in 
Saturn. However, it should be noticed 
that the observed mean color excess 
C i2 (stars) — Ci 2 (moon) = 0.02 mag 
does not agree with the value +0.20 mag 
calculated for blackbodies. It follows 
that either the moon or the reference 
stars are not radiating as blackbodies in 
the (Ai,A 2 ) range. The effect of stellar 
CO absorption in its fundamental band 
is a possible explanation for the apparent 


The spectrum of the brightest satellite 
of Saturn has been studied by Munch on 
material obtained with an image intensi- 

fier (S25 cathode) at the 72-inch camera 
of the 200-inch coude spectrograph. With 
resolutions of 0.3 and/or 0.6 A, over the 
range AAO.63-0.88 /mi, no proper spectral 
feature that could not be attributed to 
CH 4 has been detected. In particular, 
the NH 3 bands at AA0.65 and 0.79 /mi 
and the H 2 quadrupole lines of the (3, 0) 
band have not been seen on these plates. 
In comparison with the CH 4 bands in 
Saturn and Uranus, however, those in 
Titan show a noticeable curve-of-growth 
differential effect. The weak absorption 
lines in Titan (equivalent widths smaller 
than about 0.1 A) are as strong as cor- 
responding lines in Saturn, but the strong 
absorption features are relatively weaker 
in Titan. This effect can be qualitatively 
understood if the effective column den- 
sity in Titan is as large as in Saturn but 
the gas pressure is lower. 


The quadrupole spectrum of H 2 in 
Uranus has been under continued study 
by Munch. The strength of the SO and 
SI lines of the (3, 0) band has been 
measured photoelectrically with the 
Perot-Fabry scanner of the 100-inch 
coude spectrograph. Under a resolving 
power of 41,000, the equivalent widths 
of these two lines, obtained from scans 
involving 3000 counts at sampling in- 
tervals of 0.087 A, are 0.166 and 0.160 
A, respectively. An uncertainty of about 
10% in these values arises from difficul- 
ties in locating the reference continuum. 
The equivalent width of the SO line in 
the (4, 0) band, also measured inter- 
ferometrically with a resolving power of 
44,000, is 0.023 A. These measurements 
are in essential agreement with those re- 
ported by Trafton {Bull. Amcr. Astron. 
Soc, 5, 290, 1972) but do not confirm the 
larger values given by Price (Bull. Am. 
Astron. Soc, 5, 291, 1972) for the (4, 0) 

The spectrum of Uranus in the region 



of the (5, 0) band of H 2 has also been 
explored by Munch on image-tube spec- 
tra obtained with the 200-inch coude at 
6.8 A mm -1 . The measurement of two 
Uranus spectrograms has revealed the 
presence of two nonsolar lines at 
AA5327.37 and 5286.36, which are only 
about 1.0 A redward of the predicted 
wavelengths for the SO and SI lines, re- 
spectively, of the (5, 0) Ho vibrational 
band. The identification of these lines 
appears certain, but because of blends 
with weak lines in the Fraunhofer spec- 
trum their equivalent widths can be esti- 
mated to be no more than 20 mA. 
Further work at higher resolving powers 
will be carried out to make a precise 
measurement of their strengths. 

Comet 1972h 

This new comet was found on June 9, 
1972, by Sandage [Intern. Astron. Union 
Circular No. 2413) during a patrol of 
quasar fields for variable objects with 
the Palomar 48-inch schmidt. The comet 
was of magnitude 13 at the time of dis- 
covery and remained near that magni- 
tude for most of the report year because 
of the unusual nature of its orbit. The 
comet has a large perihelion distance 
of 4.28 A.U. according to elements cal- 
culated by B. G. Marsden (Intern. 
Astron. Union Circular A r o. 4^6) , and the 
distances between the comet, earth, and 
sun changed slowly for several months. 
Perihelion passage occurred near Novem- 
ber 15, 1972. 


White Dwarfs 

Spectroscopy by Greenstein with the 
coude image tube at the 200-inch tele- 
scope has continued to show that the 
hydrogen-line white dwarfs have quite 
small rotational velocities. At present, 
five of the six stars successfully observed 
show cores, formed in nonlocal thermo- 
dynamic equilibrium, that are barely 
wider than the instrumental resolution. 
If these stars are viewed as the central 
regions of former red giants, an efficient 
mechanism must operate to reduce the 
angular momentum. This occurs in the 
interiors of whatever stars undergo slow 
evolution into the white dwarf region. 

Halo Population 

In Year Book 69 (p. 84), Figure 1 dis- 
played the schematic Hertzsprung-Rus- 
sell diagram of the faint blue stars in 
the halo of our Galaxy. The large num- 
ber of high-temperature model atmos- 
pheres computed by various investigators 
has made it necessary to reanalyze the 
observed energy distributions, line pro- 
files, and line intensities in the approxi- 
mately 200 stars that Greenstein and 

Anneila Sargent have been discussing. 
Particularly the high-temperature end of 
the halo population has been extensively 
studied. The multichannel spectrophoto- 
meter provides reliable temperature cali- 
bration to approximately 40,000°, the 
line widths provide a spectroscopic esti- 
mate of the surface gravity, and the 
Balmer discontinuity at a given tempera- 
ture provides some indirect evidence of 
low hydrogen-to-helium ratio in a few 
stars. Fundamentally, the general fea- 
tures of the H-R diagram for halo stars 
as newly rederived by Greenstein and 
Anneila Sargent are the same as pictured 
earlier. A nearly horizontal group of 
stars exists over a range of temperatures 
from below 10,000° to at least 40,000°. 
If it is assumed that all these stars have 
identical masses, the fit to the horizontal 
branches of globular clusters permits de- 
termination of the mean mass of these 
stars ; this is approximately 0.50 9TZ . The 
large extension to higher temperatures, 
not normally seen in globular clusters, 
and the extremely high temperatures in- 
dicated, requires that the helium core 
of such stars be surrounded by a hydro- 
gen envelope of negligible mass. Above 



a temperature of 30,000° the surface 
helium abundance appears to be high in 
certain stars, which will be classified as 
subdwarfs. If they have the same 
mass as the horizontal-branch stars, they 
form a connecting link between the hot 
end of the horizontal-branch sequence 
and the hotter white dwarfs. Hot O-type 
subdwarfs of high luminosity, related to 
the nuclei of planetary nebulae, are rare 
among the halo population. The wide 
variety of strengths of absorption lines 
of other elements (ionized carbon and 
silicon) in the halo stars suggests a sur- 
face-composition variation like that in 
the globular-cluster giants at lower tem- 
peratures, which have a large diversity 
of abundances of the metals. 

Emission-Line Star MWC 349 

Greenstein has studied the Be star 
MWC 349, which proves to be one of the 
most highly reddened stars known, with 
an absorption of over 9 mag at visual 
wavelengths. The star shows variations 
from night to night in the hydrogen 
emission lines, which are formed in an 
extremely compact, dense H II region. 
D. Allen had observed this object in his 
infrared survey of Be stars. MWC 349 
is particularly notable in its nonthermal 
radio-frequency emission spectrum. At- 
tempts to fit the spectrum, from 0.4 to 
1.0 microns, with model atmospheres re- 
sult in a redetermination of the reddening 
law which lies within 5% of the Whit- 
ford reddening law. The total energy 
emitted by the B star is sufficient to heat 
interstellar grains and provide the in- 
frared emission observed by Allen. If 
the radio-frequency source is as small as 
the optical He II region, the star pro- 
vides sufficient energy to maintain that 
radiation also. MWC 349 thus becomes 
one of the most completely observed 
radio stars. It shows many resemblances, 
over the range from 21 cm to 0.4 microns, 
to the nuclei of Seyfert galaxies and to 
some quasars. Typical electron densities 
of 10 6 to 10 7 per cubic centimeter are 

required in a region only a light-day in 

Spectrophotometry of B-Type Stars 

Preston is engaged in an extensive spec- 
trophotometric study of some 400 bright 
B5— A0 stars. The coude scanner of the 
100-inch telescope is being used to meas- 
ure relative intensities in 14 contiguous 
4 A bands in the longward wing of HS. 
The measurements provide a profile of 
HS and equivalent widths of Si II A4128- 
30 and He I A4144. If temperature esti- 
mates (as from Q) are available, a fit to 
D. Peterson's theoretical HS profiles 
yields a gravity and, in principle, mass 
and age estimates for stars of normal 
composition. Low-dispersion spectro- 
scopists have reported numerous "mild" 
or marginal Si stars. High-dispersion 
studies have found a few of these to be 
normal (e.g., 21 Aquilae), but C. J. Dur- 
rant recently has confirmed the existence 
of such marginally peculiar stars and also 
has found evidence for stars with ab- 
normally weak silicon lines. The purpose 
of this investigation is to provide a sub- 
stantial body of quantitative information 
on the distribution of silicon and helium 
abundances among stars of different tem- 
peratures, gravities, and rotational ve- 
locities in the temperature domain of the 
Si stars. 

Rotational velocities, required for the 
comparison of observations with theo- 
retical HS profiles, are not known for the 
great majority of stars in this program. 
They are being determined from coude 
spectrograms obtained concurrently, and 
these will greatly improve the statistics 
of rotation for upper main-sequence 
stars, particularly in the range V sin i 
< 50 km s" 1 . These spectrograms provide 
a wealth of additional information. A 
number of new Hg-Mn stars have been 
discovered as well as occasional double- 
lined spectroscopic binaries. Finally, 
some 70 of the 200 spectrograms obtained 
to date possess interstellar K lines. These 
will provide new information about the 



motion and distribution of interstellar 
gas within 200 parsecs of the sun. 

Mercury Lines in Si Stars 

The presence of the Hg II line A3984 
has come to be regarded as a common 
characteristic of the spectra of the Mn 
stars, but the line has been identified in 
a few Si stars as well. Preston has ex- 
amined coude spectrograms (mostly 4 A 
mm" 1 ) of some 70 Si stars with T e (Q) 
> 10,000° K and V sin i < 50 km s' 1 for 
the presence of Hg II A3984. He con- 
cludes that the line is definitely present 
in 10 stars and undetectable in 15 others. 
No conclusion can be drawn from visual 
inspection of the remainder of the sample 
because of the likely contributions due 
to Cr I A3983.90 and Fe I A3983.96. Thus 
it appears that the Hg anomaly is an 
occasional but not ubiquitous character- 
istic of the Si stars, and it is not accom- 
panied by Mn enhancement in this group. 

The spectrum of HD 200311, however, 
contains the three strongest lines of PT 
II in addition to Hg II A3984, and re- 
solved Zeeman patterns indicate the pres- 
ence of a mean surface field of 10 kilo- 
gauss. To judge from its colors (B — T 7 = 
0.11, U — B — —0.50), it is easily the 
hottest of all known Pt stars. 

Helium-Weak B Stars 

Dworetsky has continued the survey 
of B8-B9 III stars listed in the catalog 
of Cowley, Cowley, Jaschek, and Jaschek 
{Astron. J., 74, 375, 1969), and has ex- 
tended it to include A0 III stars as well. 
The survey is being conducted with the 
100-inch coude spectrograph; the 60-inch 
x-spectrograph is being used to observe 
stars too far north for the coude con- 
figuration. The purpose of the survey is 
to see whether or not additional peculiar 
stars (especially Hg and Hg-Mn types) 
can be found in this manner. So far about 
10% of the stars observed are peculiar 
in some way. Among the more interest- 
ing objects found is HD 191110, a double- 
lined spectroscopic binary Hg star. Al- 

though both stars share the Hg anomaly, 
only the primary has enhanced Mn lines, 
while the secondary is overabundant in 
platinum. The system closely resembles 
another DLSB Hg star, 46 Draconis (P. 
S. Conti, Astron. Astrophys., 7, 213, 1970) . 

Early A Stars 

Dworetsky has completed a 40 A mm -1 
survey of all A0 main-sequence stars in 
the Cowley, Cowley, Jaschek, and Jas- 
chek catalog with the 60-inch x-spectro- 
graph. All normal stars of MK types 
B9.5-A1, IV-V, were included. The data 
will be used to obtain an improved rota- 
tional velocity distribution and to search 
for new sharp-lined bright early A stars 
to be observed with higher dispersions. 
It is hoped that new information about 
the compositions of AO-type stars with 
subtle peculiarities (e.g., early Am stars 
and cooler Hg stars) will help in under- 
standing the processes that cause these 
stars to have anomalous compositions. 

Photometry of Pleiades Stars 

Zappala has now obtained observations 
at 1.6 and 2.2 /x for about 30 Pleiades 
main-sequence stars of spectral types K 
through early M with reliable UBV 
photometry. Most of these stars have 
been observed at the two wavelengths on 
at least two good nights with the 100- 
inch Mount Wilson reflector. Although 
many of these objects are known to be 
flare stars, no obvious variations were 
detected either visually at the telescope 
or in the photometry. A standard rela- 
tion between the V — [2.2 /x] and B — V 
colors has been derived for the lower 
main sequence through observations of 
field and Hyades late-type dwarfs. 

The result suspected earlier that the 
Pleiades stars are systematically redder 
than the standard relation has been con- 
firmed. Between B — V colors of about 
+ 1.00 and + 1.50, the Pleiades have an 
excess averaging nearly 0.2 mag in V — 
[2.2 p] and as large as 0.7 mag in one 
case, independent of interstellar redden- 



ing. Furthermore, the excess seems to be 
independent of the position of the star in 
the reflection nebula surrounding the 
cluster, which might be held responsible 
for producing erroneous UBV measure- 

Interpretation of these results is cur- 
rently unclear. Cool companions could 
produce excesses in V — [2.2 /x] ap- 
proaching 0.1 mag in favorable cases, 
but this would be inadequate even in the 
unlikely event that all the stars are 
binaries. If young stars with consider- 
able convective activity are heavily cov- 
ered with cool regions similar to ordi- 
nary sunspots, an apparent excess in 
V — [2.2 [x~\ could be produced; how- 
ever, the excess would be no larger 
than 0.2 mag even for a star whose photo- 
sphere is half spotted. The most likely 
explanation is that these comparatively 
young objects are still surrounded by 
circumstellar materials of low tempera- 
ture and moderate optical depth, perhaps 
remnants of the period of star formation. 
There is support for this hypothesis in 
the fact that the excess in T 7 — [2.2 /a] 
seems to correlate with a deficiency in 
the visual magnitude, i.e., the distance 
the star appears to lie below the standard 
zero-age main sequence. However, there 
is no general excess in the observed 
[1.6 /x] — [2.2 fx] color index as large 
as 0.05 mag, so that shells, if present, 
must have temperatures less than ^500° 
K. The actual thermal emission of such 
shells would be observed at wavelengths 
longer than 3 [x] such observations will 
be very difficult, since these stars are all 
fainter than magnitude 10 at 2.2 (x. 

The K-type star BD— 10°4662 (FK 
Serpentis), recently discovered to have 
flared on several occasions, has been 
studied by Zappala both spectroscopic- 
ally and in the infrared. Both com- 
ponents of this close visual binary have 
been shown to have high lithium abun- 
dances by Herbig, who suggests that their 
similar spectroscopic and photometric 
properties place them in the post— T 
Tauri phase of evolution near the bottom 

of the convective portion of their Ha- 
yashi tracks. The combined object has 
been observed on several occasions in 
the infrared with the Mount Wilson tele- 
scopes with the following results: [2.2 
/x] = 6.55, [1.6 fx] —[2.2 fxl = +0.27, 
[2.2^] — [3.5/*] =+0.59. These colors 
approach those of conventional late-type 
T Tauri stars and are quite different 
from ordinary K stars (both indices 
about +0.1). A single 40 A mm -1 spec- 
trogram has also been obtained at the 
100-inch coude. This blue plate shows 
absorption features similar to those of 
an ordinary K4 dwarf, with the B aimer 
series weakly in emission and Ca II (H 
and K) moderately strong in emission. 
Thus — 10°4662 indeed appears to be 
closely related to the T Tauri stars and 
is perhaps in the final part of that stage 
of stellar evolution. 

Stellar Magnetism 

Borra and Vaughan continued their 
development of high-resolution photo- 
electric polarimetry for investigation of 
the Zeeman effect in magnetic Ap stars. 
The instrumentation, used with a Fabry- 
Perot interferometer in the coude spec- 
trum scanner of the 100-inch telescope, 
employs a KDP electro-optic crystal 
modulated at 0.5 kHz between two ortho- 
gonal states of birefringence and yields 
simultaneously the degree of polari- 
zation (Zeeman signature) and the rela- 
tive spectral intensity (line profile) with 
a resolution of about 0.08 A. Observa- 
tions of circular polarization due to the 
longitudinal component of the magnetic 
field have been carried out at various 
phases on a 1 Canum Venaticorum, (3 
Coronae Borealis, 78 Virginis, and 52 
Herculis during the first quarter of 1973. 
So far the line of Fe II at 4520.2 A has 
been observed exclusively. The Zeeman 
signatures, which are revealed in con- 
siderable detail by these observations, 
show many of the attributes generally 
predicted on the rotating dipole model. 
Further observations, including the at- 



tempt to study linear polarization, are 
being pursued as the basis for detailed 
comparison with theoretical models of 
these and other magnetic Ap stars. 

Photoelectric Measurements of Stellar 
Magnetic Fields 

Borra has interfaced polarizing optics 
and electronics with the Mount Wilson 
low-resolution scanner. The instrument 
has been used with the 60-inch telescope 
to measure circular polarization in the 
wings of Ha in bright, rapidly rotating 
Ap stars whose lines are too broad to be 
measured photographically for magnetic 
fields. None of the stars so far observed 
has been found to show detectable cir- 
cular polarization or a magnetic field 
within the observational errors (typical 
standard deviation is a few hundred 

Magnetic Stars 

Borra and Dworetsky have examined 
all of the Zeeman-analyzed plates of /? 
Coronae Borealis in. the plate files of the 
Observatories. Many of the plates have 

been remeasured or measured for the first 
time. There appears to be a correlation 
between the measured longitudinal mag- 
netic field and the photographic density 
of the plates at negative polarity ex- 
tremum but not at positive polarity. 
Numerical models, based on the mag- 
netic-field configuration of (3 CrB de- 
termined by Wolff and Wolff (Astrophys. 
J., 160, 1049, 1970), show simple sym- 
metrical Zeeman-analyzed line profiles 
at the positive extremum and complicated 
asymmetrical profiles at the negative ex- 
tremum. Heavily exposed plates tend to 
emphasize the core of the spectral lines, 
while lightly exposed plates tend to em- 
phasize the wings. On the basis of those 
profiles, one can expect a dependence of 
the measured longitudinal field strength 
on photographic density at the negative 
extreme but not at the positive extreme. 
This effect might explain the large scat- 
ter seen at negative polarity when ob- 
servations of many years are plotted 
in the 18.5-day period of the magnetic 
variation. This scatter had previously 
been interpreted as a secular variation 
in the field strength. 


Stellar Cycles 

Wilson has continued his measurement 
of fluxes at the center of the H and K 
lines in a number of main-sequence stars, 
using the coude scanner at the 100-inch 
telescope. The earlier types, F5-G5, con- 
tinue to show no real evidence of cyclical 
behavior. However, among the later- 
type objects several have reversed the 
direction of change of the flux values 
during the past year. For most or all of 
these stars it now appears that another 

two to four years' observation may yield 
completed cycles with adequate overlap 
to render the results convincing. 

Absolute Magnitudes of Late-Type 

Photographic observation of the spec- 
tra of late-type giants at 10 A mm -1 is 
being continued by Wilson to determine 
absolute magnitudes from measures of 
the widths of the chromospheric K lines. 
About 100 spectrograms were obtained on 
this program during the year. 


Douglas Duncan and Preston have (e '— ' 0.7) of the orbit, the lines are double 

completed a study of the double-lined for only 3 days preceding periastron. 

spectroscopic binary 8 Delphini (F orb = Series of spectrograms (Lick 120-inch 

40.58 days). Because of the eccentricity coude) obtained during those phases in- 



dicate that both components have small 
amplitude velocity variations character- 
istic of 8 Scuti-type variables. The pul- 
sational periods of the primary and 
secondary are Pi = 0. d 158 and P 2 = 
0. d 134, respectively. The mass ratio 

(M x /M 2 = 1.2) and the luminosity ratio 
(L 1 /L 2 = 1.05) inferred from the equiva- 
lent widths of Fe II lines do not seem 
to be compatible with either pulsation 
theory or stellar evolution theory. 


Occultations of Infrared Stars 

A large number of the infrared sources 
that have been detected in the past 
several years have been interpreted in 
terms of the so-called "dust shell" model. 
In this model energy from a central ob- 
ject at relatively short wavelengths is 
absorbed by particles in a surrounding 
envelope and reradiated at the much 
lower temperature characteristic of the 
grains. In this way the maximum of the 
energy distribution can be shifted far into 
the infrared, perhaps totally obscuring 
the original energy distribution of the 
central star. Unfortunately, direct ob- 
servational evidence for these shells other 
than the observed energy distributions 
themselves is largely lacking. For objects 
sufficiently near the ecliptic, however, 
lunar occultations provide a useful method 
of measuring angular extent of infrared 
sources and may provide direct evidence 
for a shell. 

IRC+10011 (= CIT 3) is a bright 
infrared source that also emits micro- 
wave lines of the OH molecule. On the 
basis of CO and H 2 absorption features 
in its 2.0-2.5 /x spectrum and its vari- 
ability at 2.2 /x, it has been classified as 
a late-type Mira star. The object has 
recently undergone a series of lunar oc- 
cultations. Three of these events could 
be observed at Palomar and Mount Wil- 
son and attempts were made by Zappala, 
Becklin, and Neugebauer to observe 
these simultaneously in two wavelengths 
with the 200-inch and 100-inch tele- 
scopes. Successful observations were ob- 
tained at 2.2 and 10.1 /x on one occasion 
and at 20 ^ on another. The 2.2 /x ob- 
servation is particularly interesting, for 

it cannot be interpreted in terms of a 
simple single source. A good fit to the 
data is provided by the Fresnel pattern 
of a uniform disk ^0."013 in angular 
diameter, contributing 65% of the total 
luminosity, superimposed on a disk 
~O."070 in diameter, contributing the 
remaining 2.2 /x flux. The 10 /x observa- 
tion is consistent with a single uniform 
disk slightly larger than 0."120 in extent. 
At 20 jx the observed diameter is roughly 
equivalent to that at 10 /x. 

These observations thus provide clear 
evidence for a shell of material around 
a smaller central object. If the 0."013 
source is interpreted as the central star 
seen directly through the obscuring ma- 
terial, an effective temperature of about 
2100° K may be derived, since the total 
flux of the object has been measured. 
This value will be somewhat smaller if a 
limb-darkened model is assumed and is 
in satisfactory agreement with the tem- 
perature derived from the observed spec- 
tral features. If the total luminosity of 
the object is about 10 4 L Q , the distance 
to the star is about 500 pc and the ob- 
served extent of the outer dust shell 
corresponds to a diameter of about 60 
A.U. The temperature of the dust shell 
is about 600° K. 

Size Measurements of Extrag atactic 

It is by now well established that 
several extragalactic sources show anom- 
alously strong infrared emission. Al- 
though the effect is most noticeable in 
peculiar galaxies of the Seyfert and 
Markarian types, large excesses have 
also been observed in some apparently 



normal galaxies such as NGC 253. How- 
ever, the physical mechanisms producing 
the infrared emission have not been es- 
tablished. In particular, observations in 
the visible band of many of the galaxies 
in question show evidence for nonthermal 
emission. Therefore, a central question 
is whether or not the infrared emission 
is also nonthermal in origin. 

One means of deciding this question is 
to search for rapid variability in the in- 
frared flux, since rapid variations would 
denote small sizes which, coupled with 
the observed luminosities, would pre- 
clude thermal emission. A continuing 
part of the long-term program of Becklin 
and Neugebauer has been to monitor the 
intensity of about 10 extragalactic 
sources. Another method of discrimi- 
nating between thermal and nonthermal 
emission is to find the size of the emitting 
region. During the last two years, the 
sizes of two strong infrared sources have 
been measured directly. The infrared 
energy from NGC 253 from 2 to 20 p has 
been shown to come from a core 10" 
(about 150 pc) in diameter. In addition, 
slit scans of the very strong infrared 
source NGC 1068 show that it consists of 
a core about 0."5 (or 50 pc) in diameter. 
The data are not sufficiently good to de- 
termine the profile of the core; models 
with both gaussian and cylindrical geo- 
metries fit the data. For both NGC 253 
and NGC 1068 these emitting regions 
are too large to be single nonthermal 
sources. It seems, therefore, that the 
radiation is probably thermal. It is 
possible the radiation arises from heated 
dust, as in many galactic infrared 

2-20 Micron Band 

During the past two years Wynn-Wil- 
liams, Becklin, and Neugebauer have 
been studying the 1-20 ^ radiation from 
regions where star formation is thought 
to be taking place. In particular, studies 
of the infrared emission (1.6 fx < A < 20 
jx) from H II regions have been continued 

and extended; 12 regions have so far 
been studied in detail (W3, W49, Orion 
Nebula, M8, G333.6— 0.2, H2-3, W51, 
NGC 7538, OH 0739—14, Sh 269, W75, 
K3-50), leading to the discovery of a 
large number of new infrared sources. 
The H II regions studied have in general 
been those that have displayed fine struc- 
ture at radio wavelengths and are asso- 
ciated with OH/H 2 masers. Insofar as 
it is possible to generalize at this stage, 
it is found that: 

1. Almost all compact (<20") H II 
regions detected at radio wavelenghs 
have strong infrared emission at 20 /x. 
This emission probably arises from dust 
mixed with the ionized gas and heated 
to 100-200° K by the hot central star 
of the condensation. Abnormal physical 
properties or abundances of the dust do 
not appear to be necessary to explain 
the emission. 

2. Most of the same objects show 
emission at 2 /a, weaker than that pre- 
dicted from the radio emission, a result 
which we interpret as indicating the pres- 
ence of very large amounts of extinction 
between the sources and the sun. A 
value of A v = 50 mag has been found in 
one case. 

3. Most, but not all, OH or H 2 
maser sources in H II regions are infra- 
red sources. There seems to be a ten- 
dency for these infrared sources, whether 
or not they show radio continuum emis- 
sion, to have a hotter energy distribution 
and a stronger 10 /x silicate absorption 
feature than those sources without maser 
emission. There is no obvious relation- 
ship between the flux densities in the in- 
frared continuum and in the microwave 
maser lines. 

4. Some H II regions contain infrared 
sources without radio counterparts. Of 
these, two may be protostars. These 
objects, in the Orion Nebula and in W3, 
both have approximately blackbody 
energy distributions, strong silicate ab- 
sorption features, and OH or H 2 maser 
emission; they are less than 3" in di- 




1-mm Photometry 

A liquid-helium-cooled bolometer sys- 
tem designed for 1-mm photometry at 
the prime focus of the 200-inch telescope 
has been developed. Successful observa- 
tions with this system have been carried 
out by Paul Harvey, Michael Werner, 
Jay Elias, and Ian Gatley during twilight 

A number of objects, including planets, 
galactic H II regions, and the QSO 3C 
273, have been observed. The most in- 
teresting results to date are extensive 
maps of the H II regions W51 and M42 
(the Orion Nebula) . The map of the 
Orion region is shown in Fig. 1. 

The data show a ridge of enhanced 
brightness extending some 3' to the north 
and east from the central position. This 





— i 1 1 1 r 

I mm Continuum Emission 
Relative To K L Nebula 






Fig. 1. One-mm continuum emission from the 
Orion Nebula region, measured at the prime 
focus of the 200-inch Hale Telescope. Positions 
and fluxes are given relative to position M, 
which coincides with both the Kleinmann-Low 
infrared nebula and the center of the molecular 
cloud. Position H is the center of the H II 

feature is similar in shape, size, and 
position to features seen in maps made 
of the Orion region in the emission lines 
of H 2 CO (2 cm) and 13 C 1G C (3 mm). 

The flux density at the central position 
of the map is — 250 ± 100 f.u. at 1 mm. 
From hydrogen recombination within the 
H II regions alone, a flux of only ^25 f.u. 
is expected at 1 mm — much less than is 
seen. Furthermore, the distribution of 
1-mm emission resembles more closely 
that of the molecular cloud than of the 
H II region. It is suspected, therefore, 
that the 1-mm radiation is associated 
with the prominent molecular cloud in 
the Orion region. 

Several mechanisms for producing this 
1-mm radiation from the molecular cloud 
have been explored. Molecular emission 
from the cloud in the 1-mm region prob- 
ably fails by at least an order of magni- 
tude to produce this high a density within 
the bandwidth used. On the other hand, 
it seems very likely that the dust ex- 
pected to be associated with the large 
column density of gas in the cloud is 
capable of thermally emitting the ob- 
served energy. 

Identification of this 1-mm radiation 
with dust in the molecular cloud has 
some interesting consequences. The in- 
tensity observed at 1 mm is proportional 
to the product of the grain temperature 
and the column density of matter; its 
distribution across the source presumably 
represents the spatial variation of these 
quantities. These two parameters of the 
cloud are not accessible to the line spec- 
troscopists, whose results, because of the 
necessity of accounting for the excitation 
of the molecular emission, reflect some 
complicated function of the gas density 
and temperature. The 1-mm observations 
are therefore complementary to those 
of the radio spectroscopists and should 
advance our understanding of these ob- 

The objects studied at 1 mm are not 
visible against the bright twilight sky. 



A technique has been developed for morning and evening twilight periods, 

pointing the telescope by offsetting from the 200-inch telescope can be utilized six 

bright stars that can be seen through an to seven hours per day in this manner, 

eyepiece on the photometer. Including thus effectively extending its usefulness. 


Interferometry of Nebulae with Low 
Surface Brightness 

Emission nebulae with low surface 
brightness have been observed interfero- 
metrically with the 40-inch telescope of 
Las Campanas Observatory by Munch 
and P. Cruvellier. The interferometer, 
constructed for the purpose at the Ob- 
servatoire de Marseille, produces Fabry- 
Perot fringes of an area in the sky 25 
arc min in diameter. The fringes are 
focused on the cathode of an image in- 
tensifler tube by means of an f/1.5 focal 
reducer. With an S25 image tube, Ha 
interferograms of faint emission regions 

in directions close to the galactic center 
of the Vela-Puppis nebular complex and 
of some extended planetary nebulae were 
obtained. The material is yet to be meas- 
ured for detailed radial velocity studies. 
However, visual inspection of the fringes 
reveals large internal motions in the 
planetaries NGC 246, 5189, 6164-65, and 
6853. In contrast, the planetaries Abell 
35 and 36 show small velocity disper- 
sions. Interferograms in the line He I 
A10830-29 were also obtained with an 
SI image intensifier, but only of the 
bright galactic nebulae Messier 8, 16, 17, 
and 20 and that around -q Carinae. 


Kristian is continuing a program of 
monitoring the optical intensity of the 
Crab pulsar at the 0.3% level to search 

for possible secular changes due to the 
slowing down of the spin. 


Cygnus X-l 

Over the past year, a number of in- 
dividual stars have been suggested as 
candidates for the optical counterparts 
of the galactic x-ray sources. A signif- 
icant number of these are binary systems, 
generally containing a luminous early- 
type star. An important member of this 
class is BD+34°3815 (HDE 226868), 
which is a weak variable radio source 
and is identified with Cyg X-l. 

The star is a spectroscopic binary with 
a period of 5.6025 days in which the 
primary optical component has a spec- 
tral type of BO lb and a velocity ampli- 
tude of 74 km s -1 . Brucato and Kristian 
noted that the radial velocity of a weak 
emission feature of He II at A4686 did 

not have the same radial velocity as the 
stellar absorption lines. During the past 
year, observations made at the coude 
focus of the Mount Wilson 100-inch in- 
dicated that the emitting region is nearly 
180° out of phase with the B star with a 
velocity amplitude of around 140 km s" 1 . 
Brucato and Kristian estimated the 
mass of the unseen secondary (presumed 
to be the x-ray source) by assuming that 
the B star is normal for its spectral type 
( — 22 371©). Combined with the mass 

= 971/sinH = Q 2256 

(Wl x + SfTCo) 2 

the lower limit for the mass of the sec- 
ondary is 5.5 gn (i — 90°). If it is 



assumed that the He II emission arises 
in the vicinity of the secondary and thus 
provides an estimate of its velocity, then 
3TC.r "-' 16 9TC . These estimates are sig- 
nificant, since several workers have sug- 
gested that x rays are produced by the 
thermalization of gravitational energy of 
matter falling onto the surface of a star. 
To produce temperatures sufficiently high 
to generate x rays, the star must be a 
collapsed object. Because the lower limit 
of 5.5 3TZ exceeds the maximum mass 
of a white dwarf or a neutron star, four 
possibilities exist: (1) The binary sys- 
tem is the source of the radio emission 
but not of the x rays; (2) the x rays are 
produced by the mechanism outlined 
above and the collapsed object is a black 
hole; (3) the x rays are produced by 
some means, as yet unspecified, which 
does not require a collapsed secondary; 
and (4) the basic assumption that the 
mass of the B star is normal is incorrect. 
To evaluate this last point, Brucato 
and Kristian assumed that the x rays 
are produced by infall to a white dwarf 
or neutron star with mass less than 
2 9frc . In this case, the mass of the 
primary is extremely peculiar, for it has 
a mass of 4 2fH or less and yet success- 
fully mimics the spectrum of a BO lb star. 
The absence of eclipses in the x-ray band 
and the assumption that the the B star 
fills its Roche lobe further limits its 
mass to ^3.15 9TC . At the present time, 
there is no evidence favoring any of these 
models over the others. 

Hercules X~l 

The identification of the pulsating 
x-ray source Her X-l with the vari- 
able star HZ Herculis, made by 
several observers at Princeton and 
Harvard, was based on the 1.7-day 
period observed at both optical and 
x-ray wavelengths. Shortly thereafter, it 
was reported that the 1.24-sec pulses 
were also detected in the optical band. 

Using the 60-inch reflectors at Palomar 
and Mount Wilson, Zappala and Brucato 
engaged in a program to verify this re- 
sult. Fourier analysis of four nights of 
photometric data did not show the pulsa- 
tion with upper limits as low as 0.11% 
of the total signal (at the 2a confidence 
level ) . Two of these nights corresponded 
to orbit phase 0.2, reported to be pre- 
ferred for the appearance of the optical 
pulsation. It is now known that, even 
at this preferred phase, the variations are 
only irregularly present. The nature of 
this object is still obscure, and further 
observations (particularly spectroscopic) 
have been and are being obtained. 

HZ Her is also being studied by Oke 
and Greenstein with the multichannel 
spectrophotometer and the coude image- 
tube spectrograph. This fascinating star 
shows an optical variation which requires 
that the side facing the x-ray pulsar be 
quite hot, and that the opposite side be 
near 8000°. Resolution into two, or pos- 
sibly three, different major components 
on the surface of this optical object is 
required by the multichannel observa- 
tions. The amplitude of the light curve 
in the ultraviolet is largest because of 
the increased strengths of the Balmer 
jump on the far side of the star. Coude 
spectra by Greenstein show a spectrum 
characterized almost completely by ab- 
sorption lines. The lines are of diverse 
widths, and the profiles of the hydrogen 
lines seem complex and variable. Broad- 
ening corresponding to velocities of from 
300 to 500 km s" 1 is indicated by He I 
and Mg II at maximum light. However, 
this may not, in fact, be due to rotation 
but rather to stream motions. The spec- 
trum at minimum is that of a late A- 
type star and has relatively sharper lines. 
This object will provide major astro- 
physical puzzles, since the heating of the 
side facing the x-ray pulsar may be 
expected to have drastic results, such as 
winds along the surface. 




Supernova Search 

Twelve supernovae were discovered at 
Palomar during the report year in the 
course of the continuing search, which 
is directed by Sargent and Searle. Ten 
of the objects were found by Kowal and 
two by graduate student J. Huchra, all 
with the 48-inch schmidt telescope. In 
addition, F. Zwicky discovered 18 super- 
novae by comparing accepted and re- 
jected plates obtained in the early 1950s 
for the Palomar Sky Survey, making a 
total of 68 found in this way. He found 
that the Sc spiral NGC 5668 produced a 
supernova in 1952 in addition to, that of 
1954, which was discovered by P. Wild. 

The survey of Sc I galaxies, which are 
thought to have about one super- 
nova every ten years per galaxy, con- 
tinued with the 18-inch schmidt tele- 
scope. No supernovae in the Sc I galaxies 
have been discovered during the 2 x /> 
years since this program was begun. A 
paper by Sargent, Searle, and Kowal on 
the statistical results of the Palomar 
supernova search in the interval 1958- 
1971 was presented by Searle at the In- 
ternational Conference on Supernovae 
held in Lecce, Italy, in May 1973. The 
supernova rate in bright galaxies is about 
one every 30 years. 

Substantial progress was made during 
the report year on the project to set up 
UBV photoelectric sequences in each of 
the supernova search fields in order to 
enable light curves to be determined from 
old plates. 

Astrophysical Studies 

Searle and Oke continued their pro- 
gram of observations of the spectra of 
supernovae and, in particular, continued 
to obtain spectrum scans of SN 1972e 
in NGC 5253, the brightest such object 
of recent years. Kirshner, Oke, Penston, 
and Searle have prepared for publication 
a report on the spectra of supernovae that 

describes and interprets the spectrum 
scans of several supernovae of both 
Types I and II and includes a comprehen- 
sive set of observations spanning 230 
days of SN 1972e. These observations 
represent the first major extension of 
knowledge regarding spectra of super- 
novae since Minkowski's great series of 
observations on SN 1937c in IC 4182, 
which were obtained 35 years ago. Three 
general conclusions have been reached 
from a study of the new data which pro- 
foundly modify earlier conceptions of 
the supernova spectrum. These are: (1) 
that supernovae of Types I and II have 
a continuum which changes slowly and 
uniformly with time and which carries 
the bulk of the radiated flux; (2) that 
there exist lines in the spectra of both 
Types I and II which have P Cygni pro- 
files, i.e., broad and strong emission lines 
flanked on the violet edge by broad and 
deep absorptions; and (3) that lines exist 
which persist through the evolution of 
the spectrum and which are common to 
Types I and II. 

The most recent scans of SN 1972e 
were obtained by Searle and Oke more 
than one year following maximum light. 
They show that while the supernova con- 
tinues to fade the rate of spectral evolu- 
tion is now very slow. The observations 
cover the entire spectrum without gaps 
from A3200 to Al0,500 in 40 A bands 
shortward of A5800 and in 80 A bands 
longward of this. Over the last 130 days 
of the supernova's first year, the light 
curve in each of these band passes can be 
described, within the accuracy attained, 
as an exponential decline at a steady rate 
of 0.017 mag per day. This decline rate 
is typical of other Type I supernovae, but 
this is the first evidence that the decay 
rate characterizes the entire spectrum 
rather than one dominant spectral fea- 
ture or an arbitrary band pass. It is 
anticipated that scans of SN 1972e can 
be obtained for at least one more year. 



Review of Spectra 

The rapid developments in the spectros- 
copy of supernovae, using the multi- 
channel spectrophotometer, as reported 
by Oke and others, made it desirable to 
provide as complete a review as possible 
of the earlier photographic data on spec- 
tra. Such a review of the best available 
data has just been published by Green- 
stein and Minkowski; more than 20 dia- 
grams show typical spectra, at various 
phases, of supernovae of Types I, II, and 
III, and IV. In general, line profiles are 

of the P Cygni character, showing both 
emission and absorption. No narrow fea- 
tures are present except in Type V. 
There is a small amount of fine structure 
within the absorption dips, notably that 
near A3800 ascribed by Oke et al. to Ca 
II. The spectra of supernovae of Type I 
seem to be remarkably uniform; Types 
II and III are essentially gigantic ordi- 
nary novae with high expansion velocity. 
Some lines are common to Types I, II, 
and III, but there are substantial differ- 
ences, notably in the absence of hydro- 
gen in Type I. 


Redshifts of Fainter Galaxies in the 
Shapley-Ames Catalogue 

In a continuation of the program to 
map the velocity field for local galaxies, 
Sandage spent seven weeks in Australia 
during October and November 1972. He 
observed redshifts of southern galaxies 
with the Mount Stromlo 74-inch reflector 
using a Carnegie image-tube spectro- 
graph. Exceptionally good weather and 
the cordial hospitality and cooperation 
of the Mount Stromlo staff resulted in 
completion of the southern galaxy pro- 
gram for the remaining objects in the 
&hapley-Ames Catalogue south of $ = 

During the three sessions on this pro- 
gram (14 months in 1968-1969, six weeks 
in February-March 1972, and the ses- 
sion in this report year) , about 330 meas- 
urable spectrograms on 290 different 
galaxies were obtained. These have now 
been measured and reduced and are being- 
analyzed for the uniformity of the ve- 
locity field. 

A similar redshift program for bright 
galaxies in the north that have no data 
was carried on by Sandage as standby, 
using the image-tube spectrograph at the 
200-inch during partly cloudy nights. 
At the end of the report year, there re- 

mained 230 galaxies north of 8 = — 30° 
to be observed before the Catalogue is 
complete with at least one plate at- 
tempted for each galaxy. 

Relative Distance to Nearby E and 
SO Galaxies 

After six months of preparation, San- 
dage and Visvanathan began routine 
photometric observations with the Palo- 
mar 60-inch of the energy distributions 
of all bright E and SO galaxies to obtain 
relative distances. More detailed cali- 
bration of the color— absolute magnitude 
effect than was available last year was 
secured by observing most E and SO gal- 
axies in the Virgo cluster brighter than 
m pg = 13, using the Visvanathan auto- 
matic scanner described in Year Book 
71, p. 691. The color change with abso- 
lute magnitude becomes steeper as one 
proceeds farther toward the ultraviolet, 
changing from A(U — V)/&M V = 0.08 
mag per mag for (\ v ) = 3650 A to 
A (3460 — V)/AM V = 0.15 mag per mag 
and using a narrow-band interference 
filter at (A) = 3460 A. 

The goal of the investigation is to 
combine distances determined in this way 
with the new redshifts to test how nearly 
the local flow is Hubble-like. 


TABLE 2. Space Densities* Log *(Af p ) 


All Markarian 



M p 





-1.74 (2) 

-0.42 (1) 


-2.32 (1) 

-1.12 (2) 


-2.32 (5) 

-1.13 (4) 


-2.53 (14) 

-1.58 (6) 


-2.97 (17) 

-1.84 (15) 


-3.34 (31) 

-4.43 (3) 

-1.98 (41) 


-3.53 (74) 

-4.52 (6) 

-2.28 (77) 


-4.20 (59) 

-4.80 (16) 

-3.17 (33) 


-5.35 (12) 

-5.75 (5) 

-4.46 (6) 


-6.56 (4) 

-6.92 (2) 


*Space densities are in Mpc~ 3 mag -1 . H is assumed to be 75 km 
s _1 Mpc -1 . The number in parentheses is the number of galaxies on 
which each estimate is based. 

Motion of the Sun Relative to the 
Nearby Galaxies 

In a related program, G. A. Tammann 
spent four weeks at the Observatories 
working with Sandage's new southern 
redshifts for supergiant Sc galaxies in 
the first step for a new solar motion 
solution of the sun relative to all nearby 
galaxies. The results are promising and 
show that the kinematic field for spirals 
can be mapped in this way once the 
necessary direct plates are available to 
classify the galaxies into the van den 
Bergh luminosity classes. Brucato and 
later Babcock began to obtain direct 
plates of the desired southern galaxies 
using the Las Campanas 40-inch reflector 
with a Carnegie image tube mounted at 
the f/7.5 Cassegrain focus. 

Markarian Galaxies 

Sargent and graduate student John 
Huchra obtained spectra of 26 galaxies 
in Markarian's fourth list of galaxies 
with an ultraviolet continuum, continu- 
ing work done by Sargent in previous 
years. All of the galaxies had emission- 
line spectra. They included two new 
Seyfert galaxies, Markarian Nos. 315 and 

358. Sargent and Huchra then used data 
obtained by various spectroscopists on 
242 galaxies brighter than m p = 15.55 
mag in Markarian's first four lists in 
order to obtain an estimate of the space 
density of these objects as a function 
of absolute magnitude. This work is an 
improvement over the estimates reported 
earlier by Sargent (Astrophys. J., 173, 7, 
1972) , which were based on data for only 
about half as many objects. At the same 
time, Sargent and Huchra redetermined 
the space density of normal field galaxies, 
using a list of galaxies brighter than 11.0 
mag in the Reference Catalogue of Bright 
Galaxies that had been prepared by grad- 
uate student Paul Schechter. The results 
are given in Table 2. 

Markarian's fourth list is the only one 
that includes a region of the south 
galactic hemisphere. This region, unlike 
those covered by lists I— III, does not 
contain any nearby clouds or groups of 
galaxies. There are indications in the de- 
tailed statistics (not presented in Table 
2) that Markarian galaxies fainter than 
M p = — 18 are preferentially concen- 
trated in clouds and groups of galaxies 
and that, in consequence, the high value 
for their space density may be an over- 

Bright Spiral Galaxies 


History of Star Formation in Galaxies from both regions is starlight, but evi- 

,.,,,. t , ,, dently the history of star formation has 

Very little is yet known about toe been gi ificantl different in the tails 

history of star formation in galaxies nor frQm ^ [q ^ m&in fa R 
about what processes trigger the arge- ^ ^ formation of interaction tails 
scale bursts of star formation that give inyolveg mQre than & A h d ica] 
rise to such diverse phenomena as mi- perturbation of stars that exlsted before 
clear hot spots the patchy appearance ^ laxieg interacted It ig h d that 
of irregular galaxies, and to the very gtudies Qf guch turbed gtemg wiu 
blue galaxies that have been extensively ;de & ugefu , meang Qf lon the 
discussed in recent Year Books. In an gituationg that tri gtar formation in 
attempt to throw some light on these mope near , norma] laxieg 
problems, Searle has continued to obtain 
multichannel scans of star clusters, blue 
galaxies, and of the nuclear, disk, and 
halo regions of spiral galaxies. Inde- Van der Kruit started his appointment 
pendent parameters of the energy distri- as Carnegie Fellow with the completion 
butions are being measured to study the of the discussion of his survey of the 
stellar content and history of star forma- radio continuum brightness distribution 
tion in these sample stellar populations, of bright spiral galaxies that he obtained 
In connection with this work, Searle ob- with the Westerbork Synthesis Radio 
tained at Las Campanas multicolor Telescope. In a number of galaxies the 
photometry of 40 open clusters in the radio observations had shown that the 
Large Magellanic Cloud. This work is radio ridges in the brightness distribu- 
an extension of the study by Searle, Sar- tions were displaced toward the inner 
gent, and Bagnuolo which was reported edges of the optical spiral arms, thus con- 
in Year Book 70 (p. 423) and Year Book firming the concept of compressed regions 
71 (p. 679). in the density-wave picture of spiral 

In a related program, Searle and Sar- structure. This permitted determination 
gent have obtained multichannel scans of of the strength of the density-wave corn- 
ring galaxies. They find that the energy pression in a dozen galaxies by compari- 
distributions of these are closely similar son of the radio-flux density of the 
to those of star clusters a few hundred large- and small-scale structure, since 
million years old. The same age is indi- the radio-spiral structure results from 
cated for these galaxies from the meas- compression of the magnetic field and 
ured expansions of the rings. Searle and relativistic electrons in the large-scale 
Sargent believe that these galaxies, of "base disk." This property was related to 
which VII ZW 466 is the best studied other integral properties and a few cor- 
example, are the first truly young extra- relations were established and discussed, 
galactic systems to be recognized. Although other relations were also found, 

Searle and Sargent are applying similar these appear to be most fundamental, 

techniques to the study of tidally inter- Strong compression galaxies turned out 

acting galaxies and have found that the to be early (bright) van den Bergh lumi- 

outlying fragments of interacting pairs nosity classes and to have relatively low 

are systematically bluer than the main mass-luminosity ratios, which indicates 

bodies. A good example is NGC 4676 that strong compression results in for- 

(Arp 242) ; this interacting pair of gal- mation of either more stars or stars of 

axies has been named "The Mice" by lower M/L. Also, these galaxies had a 

Vorontsov-Velyaminov. Searle and Sar- small radius of the maximum in the ro- 

gent find that the mice have red bodies tation curve compared to the radius of 

and blue tails! They find that the light the outermost H II region, which natur- 



ally results in a relatively high velocity 
between gas and wave pattern. 
Three other results are: 

1. Seyfert nuclei are the strongest 
radio emitters at 1415 MHz (10 20 to 10 22 
WHz -1 ster -1 ), while most normal spirals 
have nuclei of 10 18 to 10 20 WHz" 1 ster 1 . 

2. A strong correlation between the 
powers at 1415 MHz and 10/x for the Sey- 
fert nuclei may persist (more weakly) 
for nuclei of normal galaxies. 

3. All disks with brightness tempera- 
tures above 1°K have nuclei stronger 
than 10 19 WHz -1 ster" 1 , which may be 
interpreted as meaning that although 
generally cosmic rays in the disks are 
accelerated by supernovae, the stronger 
nuclei may also contribute substantial 

Having related the compression 
strength to rotational properties, the next 
step was to investigate those galaxies 
with well-determined optical rotation 
curves. From a sample of 26 spirals, the 
relation with the luminosity class was 
confirmed and a possible correlation with 
the development of the nucleus ( Byurakan 
type) was indicated. It was then shown 
that the compressional strength did not 
depend on the total mass or specific 
angular momentum (angular momentum 
per unit mass) , but on the ratio of these 
properties. For a 'particular mass, strong 
compression galaxies have a lower spe- 
cific angular momentum than weak com- 
pression galaxies. Statistically, this re- 
sults in a larger central concentration of 
mass for strong compression galaxies, 
which explains the result that such gal- 
axies have generally better developed 
nuclei. If the nuclei are more active, 
there may be important consequences in 
the context of the possibility that nuclear 
explosions regenerate density waves. This 
suggests a closed cycle because strong- 
compression results in transport of more 
angular momentum, with enhancement of 
nuclear activity. 

To continue this line of research, spec- 
tra were taken of NGC 4321 (M100) 
with the Cassegrain spectrographs of the 

Palomar 60-inch and 200-inch telescopes 
for determination of the rotation curve. 
This could serve as a confirmation of the 
relations mentioned above, since the 
Westerbork data suggested that NGC 
4321 most likely has strong density-wave 
compression. The characteristics of the 
rotation curve and the derived mass- 
luminosity ratio plus the luminosity class 
did confirm the relations and the strong 
compression in this galaxy. Evaluation 
of all possible integral properties re- 
vealed a close similarity with M51. In 
view of the possibility that the spiral 
structure of M51 is regenerated by the 
passage of NGC 5195 and the absence of 
equally massive companions to NGC 
4321, this similarity is truly remarkable. 
This means that the details of the gen- 
erating mechanism, or possibly even the 
mechanism itself, do not determine the 
characteristics of the galaxy but that 
they are determined by the character- 
istics of the disk. Since apparently gal- 
axies can be specified completely by only 
two parameters (P. Brosche, Astron. 
Astrophys., 23, 259, 1973), we then have 
the following scheme to explain such 
similarity. The parameters, mass and 
specific angular momentum, initially 
specify a galaxy. This means that the 
compression strength is determined just 
as is the form of the rotation curve. 
Through star formation triggered in the 
density-wave compression regions, the 
development of other properties is now 
governed by the compression strength or 
is in that way determined by the same 
two initial conditions independently of 
the generating mechanism of the density 
waves. Such a scheme explains the pres- 
ence of only two dimensions in many- 
dimensional galaxy-parameter space, as 
well as the close similarity of NGC 4321 
and M51. 

An extensive spectroscopic study was 
started by van der Kruit of those galaxies 
for which the Westerbork survey indi- 
cated strong recent activity in the nu- 
cleus, in particular NGC 3310 (extremely 
bright radio nucleus), NGC 4258 (com- 



plicated radio structure and Ha arms) , 
and NGC 4736 (triple radio source in the 
central region). 

Synthesis of the Stellar Population 
of Galaxies 

Gunn has obtained high-resolution 
scans of four giant galaxies (NGC 4889, 
4874, 6166, and 4472) and some sixty 
stars to be used as material for stellar 
population synthesis of these systems, in 
collaboration with Dr. Beatrice Tinsley 
of the University of Texas. The observa- 
tional program is nearing completion, but 
virtually no analysis has been done as 

Photometry of OJ 287 and M82 

The absolute-energy-distribution ob- 
servations of OJ 287 were obtained by 
Visvanathan with 20 A bandwidth in blue 
and 40 A in red from 3300 A to 10,000 
A on January 18, 19, 20, and 21, 1972. 
The continuum is featureless without any 
absorption or emission lines. Further, 
the continuum can be represented by a 
straight line with a slope of 1.25 in log 
F v — logv plot. Polarization observations 
on January 19, 1972, in the wavelength 
range 3300 A to 8000 A show that the 
continuum is linearly polarized and P 
and are independent of wavelength. 
The above observations are interpreted 
to indicate that the continuum of OJ 287 
is of synchrotron origin. A theoretical 
synchrotron spectrum based on a simple 
model of high-energy electrons having a 
single power-law energy distribution is 
computed and found to fit the observed 
spectrum of OJ 287 from 3.5 mm to 
visible wavelengths. 

Visvanathan has obtained narrow- 
band photoelectric scans (20 A in blue 
and 40 A in red) of the region RD in 
the filaments of M82 with the 200-inch 
telscope. The intensity of emission lines 
[O II], H 7 , H/?, [O III], Ha, [N II], 
and of some absorption features has been 
obtained. From the Balmer emission 

lines of hydrogen, an interstellar absorp- 
tion of A v ~ 2.0 mag is derived for the 
region RD. 

Chains of Galaxies 

For the past several years, Sargent has 
done observational work on compact 
chains of galaxies. During the report 
year work was completed on Markarian 
6, a chain of six galaxies of 15 to 16 mag 
at R.A. (1950) 16 h 55 m 0, Decl. — 81°39'. 
All the galaxies in this chain have rough- 
ly the same redshift; the mean is 
cAA/A = 11,660 km s _1 with a dispersion 
of 300 km s" 1 . Markarian 6 is about 340 
kpc in length. Its dissolution time, if it 
is really a linear formation in space, is 
about 5 X 10 8 years, a typical value for 
compact chains. 

The apparent existence of chains of gal- 
axies has in the past led to several more or 
less bizarre explanations of their nature 
and origin. For example, Burbidge, Bur- 
bidge, and Hoyle (Astrophys. J., 138, 
873, 1963) concluded that the chains 
must be young systems which condensed 
within the past 10 8 years from tubelike 
magnetic formations in intergalactic 
space. Both Vorontsov-Velyaminov (As- 
iron. J., 66, 551, 1961) and Arp (Publ. 
Astron. Soc. Pacific, 80, 129, 1968) have 
expressed the view that certain chains 
result from the ejection of galaxies from 
the nuclei of parent galaxies. The idea 
that true chains of galaxies are unstable 
and therefore must have ages which are 
less than the Hubble time seems un- 
doubtedly correct. However, studies of 
the UBV colors of member galaxies by 
Sargent over the past few years have 
led to no indications that they are young 
systems. (Galaxies that have ages of 
a few times 10 8 years should be recog- 
nizable from their integrated colors, ac- 
cording to the calculations by Searle, 
Sargent, and Bagnuolo, mentioned in 
Year Book 71, p. 679.) 

A fairly mundane hypothesis which 
does not seem to have been considered in 
the past is the idea that linear formations 





Arp 330 


. D 

f C 
* B 

, A 



Fig. 2a. A sketch of the chain Arp 330. It is 
comprised of six member galaxies lying along a 
somewhat broken line. The elongated form of 
the large central galaxy (C) connects the AB 
and DEF subchains. 

in compact groups of galaxies represent 
transient states in the normal dynamical 
evolution of small clusters of galaxies 
that are, on the average, spherical. This 
hypothesis has presumably been ignored 
because, intuitively, the chance that an 
apparent chain of galaxies on the plane 
of the sky results from such an accidental 
alignment seems very small. However, 
it is in practice hard to calculate this 
probability because one's notion as to 
what kind of formation constitutes chains 
is very subjective. Accordingly, Sargent 
and Edwin Turner, graduate student in 
astronomy at Caltech, have investigated 
the problem by carrying out extensive 
numerical calculations of the evolution 
of gravitationally bound clusters com- 
posed of six galaxies represented as mass 
points. In these calculations the initial 
conditions were defined by randomly dis- 
tributing 6 mass points in a sphere; each 
mass point was randomly allocated 3 
components of velocity arranged so that 
the cluster was gravitationally bound. 
The dynamical evolution of several such 

systems was studied, using a computer- 
printed picture of the cluster as seen from 
three fixed, mutually orthogonal direc- 
tions each crossing time. (This ensures 
that the configuration of the "galaxies" 
changes considerably between pictures.) 
The resulting pictures were then in- 
spected for chainlike formations in a 
manner similar to that in which astron- 
omers examine pictures of the sky. Con- 
sequently, although the judgment as to 
what constitutes a chain is still subjec- 
tive, at least the same subjective criteria 
are applied to the real and the simulated 
pictures (see Fig. 2a, Fig. 2b). 

Several series of computations pro- 
duced a total of 297 pictures of 6-body 
clusters: chains similar to the well-known 
examples in the sky (a good collection is 

6 D. 4 3Z (SP) 



D o 






100 KPC 


VIEW 3 l_^2 

o fc 




Fig. 2b. Views of an assimilated cluster of 
galaxies as it would appear from three ortho- 
gonal directions. The very striking chain visi- 
ble when the system is viewed along the second 
axis disappears completely in the other two 
orientations. A sixth object (No. 2) was ejected 
to a large distance during previous dynamical 
activity and is not shown. 


contained in Arp's A tlas of Peculiar Gab This work was inspired by Sargent's 
axies) were found in 8% of these pictures, study of the cluster ZwCl 0152+33, 
The chains arise by chance during the which is a highly elongated system on the 
clusters' evolution. They appear and plane of the sky. Sargent suggested 
dissolve in a fraction of a crossing time {Astrophys. J., 176, 581, 1972) that this 
and from this point of view they are un- cluster might be a flattened system seen 
stable, as astronomers have supposed, edge-on, but rejected the idea because 
although the clusters are not unstable, no cluster with abnormally small velocity 
The three-dimensional views produced in dispersions, which would be expected of 
the calculations showed that all the simu- such systems if seen face-on, had been 
lated chains were configurations of gal- discovered. However, almost immedi- 
axies instantaneously in a plane; no ately Robinson and Wampler {Astrophys. 
genuine three-dimensional chains were J. [Lett.], 179, L135, 1973) reported that 
found. the compact cluster Shakhbazian I has 
The proportion of projected chains the smallest known radial velocity dis- 
found in these calculations is much higher persion, less than 60 km s" 1 . Sargent and 
than one would intuitively expect. As a Turner's computer simulations show that 
result, Sargent and Turner conclude that this cluster and ZwCl 0152+33 can 
the most likely explanation of the well- barely exist as flattened systems for one 
known chains of galaxies in the sky is Hubble time, if that is indeed their shape, 
that they are due to transient projection In arriving at this conclusion, they found 
effects in clusters that are on average it necessary when treating close en- 
spherical. More exotic phenomena need counters to perform calculations making 
not be invoked. a crude allowance for the fact that gal- 
Sargent and Turner have also used the axies are not mass points, 
computer simulations to investigate the 

accuracy with which the virial theorem Radio Sources and Spiral Galaxies 
can be applied to small clusters of gal- 
axies. It is not possible to account for Tt has lon g been known that radio 
the abnormally high mass-to-light ratios sources are often associated or paired 
inferred for many small clusters. They with elliptical galaxies. Arp is now con- 
are also investigating the extended dy- ducting investigations of the neighbor- 
namical evolution of clusters that eject hood s of nearby spiral galaxies with the 
one or two members because of close en- aim of cataloging all radio sources and 
counters. In such systems the remaining optical objects that might be associated 
objects assume a more compact configur- with the spirals. He used the 210-foot 
ation. This process may form a basis Goldstone Telescope* for mapping areas 
for explaining systems such as Seyfert's of U P to 4 square degrees around each of 
Sextet, in which the galaxies are so close six s P iral galaxies in the magnitude in- 
together that they should be torn apart terval 10.0 to 12.7. He reports that there 
by mutual encounters on a time scale appears to be an excess of radio sources 
much less than the Hubble time. The over normal background counts in the 
galaxies in such systems show no spectro- 01 to °- 2 flux-unit range around these 
scopic or other indication of being young galaxies. The excess is about 80% at a 
systems. 70% confidence level. To confirm the 
Finally, Sargent and Turner have suggested association of radio sources 
studied the dynamical evolution of flat- 
tened clusters of galaxies in which the * The NASA^L Deep Space Network is 
, ., ,. • • ii i. •• * operated by the Jet Propulsion Laboratory, 
velocity dispersion in the direction of California Institute of Technology, under Con- 
flattening is initially considerably less tract NAS 7-100 sponsored by the National 
than that in the perpendicular plane. Aeronautics and Space Administration. 



with spiral galaxies, two team investiga- 
tions have been organized with collabora- 
tion from the Jet Propulsion Laboratory 
and from the National Radio Astronomy 

Areas around the six spirals referred to 
are being analyzed by Arp on 48-inch 
schmidt plates for the purpose of cata- 
loging all peculiar galaxies therein. 

Multiple Interacting Galaxies 

Arp has listed all systems of three or 
more bright galaxies that appear to in- 
teract strongly on the basis of mutual dis- 
tortion. He believes that these tend to 
occur in aligned configurations, some- 
times near galaxies of much larger ap- 
parent dimensions and redshift. 

Deep Photography of Galaxies with 
Connected Companions 

Arp continued the study of luminous 
connections between galaxies. Plates of 
NGC 4319 and NGC 4151 were analyzed 
by means of the automatic isodensity 
tracing machine (GALAXY) at the 
Royal Observatory in Edinburgh, Scot- 
land, in collaboration with N. M. Pratt. 
Addition of density information from 
three separate plates of NGC 4139 led 
Arp and Pratt to suspect a connection of 
uniform width between the galaxy and 
the quasar Markarian 205, even though 
the connection is not visible on any one 
of these three moderately short-exposure 

Additional objects analyzed in the 
same way were the suspected galaxy- 
quasar association of NGC 7413 and 3C 
455, as well as 3C 33 and 4C 31.32. 

Superposition Printing 

In a joint project connected with John 
Kormendy, a graduate student at Cal- 
tech, Arp has obtained extremely deep 
baked Illa-J exposures of the NGC 
7331-Stephan's Quintet area of the sky 
with the 48-inch telescope at Palomar. 
The superposition printing of six such 

plates produced a composite photograph 
that reached farther than any previous 
photograph with the 48-inch schmidt 
telescope and about as faint as even deep- 
limiting 200-inch exposures. 

A network of luminous filaments was 
revealed in the region between the large 
Sb spiral NGC 7331 and the small inter- 
acting group of galaxies in Stephan's 
Quintet. Similar plates are being ac- 
quired by Arp in the region north of NGC 
7331 in an attempt to determine the fre- 
quency of occurrence of such faint fila- 

Nebulosity Near 3C120 and BL Lacertae 

A project that, like the preceding one, 
involves very low light levels of nebu- 
losity as recorded by the 48-inch schmidt 
is a study of the peculiar radio galaxies 
3C 120 and BL Lac. Faint, fairly linear, 
luminous filaments have been recorded 
in the vicinity by Arp and Barbieri for 
3C 120 and by Arp and Penston for BL 
Lac. For 3C 120, spectra with the 200- 
inch show emission nebulosity at galactic 
redshift. It is desirable, therefore, to ob- 
tain direct photographs in Ha emission 
of 3C 120 in order to study the form of 
this peculiar nebulosity. 

Redshifts in Small Groups of Galaxies 

As part of a continuing program, Arp 
has obtained additional spectra and red- 
shifts of galaxies in small groups and 
associations. Most recent results are for 
companions to Markarian 94, 459 (com- 
panions E and W) , and 474. Spectra of 
companions to NGC 4151, 2974, 2146, 
3166, 7448, 157, 210, 1199, and 864 were 
also obtained. 

Compact Galaxies 

Inspection of National Geographic So- 
ciety— Palomar Observatory Sky Survey 
plates by Zwicky has resulted in the 
identification of some hundreds of com- 
pact galaxies and "post-eruptive" gal- 




Identification of 3C Sources 

Kristian, Sandage, and Katem are con- 
tinuing a program begun in 1970 to es- 
tablish the nature of all 3CR sources to 
the 200-inch photographic limit, using 
the very accurate radio positions being 
measured by several groups. The pro- 
gram involves position measurements of 
optical candidates on 48-inch and 200- 
inch direct plates, and photometry and 
spectroscopy of new identifications. A 
review by Kristian at the end of 1972 
shows that one-third of all 3CR sources 
are still not identified. 

Radio Source Structure 

Gunn and Malcolm Longair of Cam- 
bridge have begun a program of image- 
tube photography of all high-latitude 3C 
radio sources, both to obtain identifica- 
tions to the limit of the 200-inch tele- 
scope and to make an effort to correlate 
radio structure on the scale of a few 
seconds of arc obtained with the 5-kilo- 
meter Cambridge radio telescope with 
optical structures on the same scales. 
The most difficult part of the program 
turned out to be the necessary astrom- 
etry; serious distortion in the image-tube 

plates and measuring difficulty with the 
48-inch plates used to obtain secondary 
standards limited the overall accuracy to 
slightly less than 1 arc sec, whereas some- 
thing better than twice as good had been 
expected. A serious component in the 
error is lack of good positions for the 
primary standards. The AGK3 or some- 
thing equivalent is badly needed for this 

Search for the Radio Source Near 
Cygnus X-3 

On 2 September 1972, this source un- 
derwent an unprecedented outburst, in- 
creasing its radio brightness by a factor 
of 1000 within 2 days. During the next 
week, a search was made for a visible 
counterpart of the source by Westphal, 
Kristian, Huchra, Shectman, and Bru- 
cato, using the silicon vidicon photo- 
meter at the 200-inch prime focus and 
direct plates at the 48-inch. No visible 
object was detected in the range AA3500- 
8000 to a limit of 23 mag. A month after 
the outburst, however, an infrared search 
by Becklin, Kristian, Neugebauer, and 
Wynn-Williams yielded an identification 
at 1.6 and 2.2 /a. 


Quasar Surveys 

Spectroscopic studies reported by 
Schmidt in previous annual reports have 
established the redshift distribution of 
quasars of 18th magnitude. Such infor- 
mation is lacking for brighter quasars 
since no substantial complete samples are 
available. Richard Green, a graduate 
student, has started a survey with the 
18-inch schmidt telescope that will 
eventually cover about 10,000 square de- 
grees. Blue and ultraviolet exposures on 
the same film are used to search for ob- 
jects brighter than 16 or 17 mag with an 
ultraviolet excess. Spectroscopic work 

will be undertaken to eliminate the white 
dwarfs and hot blue stars that constitute 
a large fraction of the candidates. 

Spectroscopic Observations 

Spectroscopic observations of quasi- 
stellar sources identified by D. Shaffer in 
the 6-cm National Radio Astronomy Ob- 
servatory survey are carried out by 
Schmidt. This work will yield a com- 
plete sample of quasi-stellar sources 
selected at high radio frequency. Such 
a sample is essential for the interpreta- 
tion of high-frequency radio source 
counts that contain a large fraction of 
quasi-stellar sources. 



Quasars as Events in Nuclei 

Kristian has extended the search for 
galaxies underlying quasars and has 
studied the general problem of detecting 
underlying galaxies on direct plates. If 
a galaxy is present, it will not be seen 
if its image is smaller than that of the 
quasar. The image size of the quasar, as 
of any point source, depends only on its 
brightness, while the image size of a 
galaxy depends on the actual size and 
distance of the galaxy and is known to 
have a well-defined upper limit at a given 
redshift. By measuring stellar image 
sizes as a function of brightness and using 
Sandage's measurements of bright gal- 
axy sizes, Kristian divided the Hubble 
diagram (magnitude versus redshift) for 
quasars into regions in which a quasar 
image is larger or smaller than a giant 
galaxy image, and he was able to predict 
in individual cases whether or not an 
underlying galaxy should be seen if it 
were present. An examination of 200- 
inch plates of 26 quasars gave results in 
good agreement with the predictions: 
Those quasars predicted to show a galaxy 
did so, and those quasars predicted not 
to show a galaxy did not. Galaxy images 
concentric with the quasars were clearly 
seen in five cases, and some evidence was 
seen in six additional cases. Four other 
quasars showed off-center structure. 

The available evidence is consistent 
with the hypothesis that all quasars occur 
in the nuclei of giant galaxies, but the gal- 
axies are not seen in most cases because 
their light is overwhelmed by that of the 
much brighter quasar. 

The original definition of a quasar in- 
cluded the requirement that it have a star- 
like image on a direct plate. This require- 
ment, however, was intended at the time 
to distinguish quasars from large nearby 
radio galaxies, and it appears that 
whether or not it is met literally depends 
on such accidental factors as atmospheric 
seeing and telescope focal length. 

Association of Quasars and Clusters of 

Forty-eight-inch plate material is es- 
sentially complete for the sample of 
quasars with redshifts z less than 0.36. 
Of the sample of 28, good plates now exist 
for 26. For more than half of these, 
"suspect" galaxies exist that are candi- 
dates for redshift determination. Thus 
far, six objects are known to be associ- 
ated with groups or clusters of the same 
redshift: B264 (0.06), Ton 256 (0.13), 
PKS 2251+11 (0.32), 3C323.1 (0.27), 
PHL 1093 (0.27), and PKS 1049—69 
(0.33). All have been published except 
the last two, which were measured in late 
1972. Gunn and French, an undergrad- 
uate, have undertaken a statistical sur- 
vey of the plates to determine whether 
QSOs are significantly correlated with 
the general distribution of galaxies to the 
plate limit, using randomly selected con- 
trol fields. The counting is virtually com- 
plete, but the analysis has not yet begun. 

Small Quasars at the Center of 
N Galaxies 

N galaxies, as a class, are intermediate 
between Seyfert galaxies and quasars in 
the strength of their central bright blue 
nucleus relative to the total light. The 
spatial distribution of light was studied 
in a number of N galaxies during the 
report year in an effort to separate the 
two luminous components of these sys- 

Sandage made UBVR photoelectric 
color measurements of 12 N galaxies, 
using a number of different sky-blocking 
apertures at the 200-inch telescope with 
its large focal-plane scale. The measure- 
ments showed a marked reddening and an 
increase in intensity as the apertures 
were enlarged on each N galaxy. The 
color and intensity gradients permitted 
a separation to be made of the two com- 
ponents to the light. It was found that 
the central blue source had the photo- 



metric properties of a quasar, while the 
extended source had colors and the in- 
tensity distribution of a normal giant 
elliptical galaxy. 

This composite model of N systems, 
with a mini-quasar in the nucleus of the 
giant E galaxy, resembles the model of 
quasars themselves that is required from 
the data reported last year that all 
quasars lie on the bright side of the Hub- 
ble line for giant E galaxies in the red- 
shift-magnitude diagram, and from 
Kristian's direct observations of the un- 
derlying galaxies. Calculations were 
made by Sandage for the intensity and 
color contamination by the central 
quasar on the combined light of quasar 
and galaxy. Good agreement was ob- 
tained between the observations and the 
predictions for the place in the two-color 
U — B, B — V diagram of both quasars 
themselves and for N galaxies as the 
contamination ratio is changed. The re- 
sults support the conventional interpre- 
tation that quasars are the nuclei of 
some galaxies. 

Observational Limits to Nonvelocity 

Redshifts for the Mini-Quasars in 

N Galaxy Nuclei 

The separation of the galaxy and 
quasar components in the total light of 
N galaxies permits a unique test for a 
limit to any nonvelocity redshifts for 
the compact quasarlike central source. 

The redshifts that are measured in 
N-galaxy spectra are always emission- 
line shifts from the central mini-quasar; 
the spectrum is not from the underlying 
E galaxy. Hence, if there were a large 
nonvelocity component to this redshift, 
the redshift-apparent magnitude dia- 
gram for the underlying galaxy, plotted 
using the measured redshift for 'the cen- 
tral mini-quasar, would show a displace- 
ment relative to the points for normal 
giant E galaxies, whose absorption-line 
redshifts are clearly cosmological. San- 
dage made the test with the 12 N galaxies 
previously studied, using the apparent 

magnitudes of the underlying E galaxy 
that were separated out. The Hubble 
diagram, given as Fig. 3, shows no devia- 
tion of the N galaxy E components from 
the mean relation of similar E galaxies 
in the field. Any nonvelocity redshift is 
smaller than the intrinsic dispersion of the 
sample. The limit to such a nonvelocity 
effect AZ, put by analysis of the scat- 
ter, is v(AZ/Z) < 0.1, which means that 
any such postulated component is less 
than 10% of the cosmological redshift, 
and from these data it could be zero. 

Sandage concludes that there is no evi- 
dence from N galaxies to support non- 
cosmological redshifts in quasars, in 
view of their absence in the compact 
mini-quasars of the N systems. Details 
of the test were published during the 
report year (Astrophys. J., 180, 687, 

Properties of Quasars 

Simultaneous observations of blue 
magnitude, linear polarization, wave- 
length dependence of polarization, abso- 
lute energy distribution of the con- 
tinuum, and strength of the emission lines 
obtained in the years 1968-1971 for the 
variable quasars 3C279, 3C345, 3C446, 
and 3C454.3 were discussed by Visva- 
nathan. He found that the continuum 
was linearly polarized in all these objects 
during outburst and nonoutburst periods 
and when the objects reached minimum 
brightness. AVhenever the total flux in- 
creased, an increase in polarized flux was 
observed. The position angle of the elec- 
tric vector changed during the burst. The 
ratio between the polarized flux and the 
total flux does not exceed a value of about 
0.16. This low value is interpreted as 
being due to the polarization caused by 
several active spots in the continuum- 
emitting region. The absolute energy 
distribution of the continua both in active 
and quiet periods shows that the power- 
law representation of the continuum is 
valid. This fact and the finding that the 
polarization is independent of wave- 




O 4.6 







3C 287.1 

3C 390.3 
3C37I — 

3C 445 
3C 120 





V 2 6-K v -A v 

Fig. 3. A comparison of the Hubble diagram for the underlying galaxy component of N sys- 
tems with aormal radio galaxies. The agreement of the two sets of data indicates that the quasar 
components of the N systems have no nonvelocity components of redshift to better than 10%. 

length support the view that the radia- 
tion in these objects in the observed fre- 
quency range is of synchrotron origin 
in both active and quiet periods. Varia- 
tions in the spectral index of the continua 
were observed, but no simple relationship 
between color and brightness was found. 
During the period of the observations, 
the absolute intensity of the emission line 
Mg II in 3C345 and 3C454.3 and the C 
IV line in 3C446 remained constant 
within 10% to 20%. The emission line 
Mg II in 3C345 was unpdlarized, indicat- 
ing that the regions from which the 
emission line originates are randomly ar- 

Angular Diameters of Quasars 

Van der Kruit discussed the inferences 
drawn from the distribution of (radio) 
angular diameters of quasars. It was found 
that the distribution with redshift is en- 
tirely compatible with evolving universes 
and with the Rees and Setti (Nature, 
219, 127, 1968) evolution of physical di- 
ameters ; it was further indicated that an 
apparent gap in the linear size distribu- 
tion (Reinhardt, Astrophys. Lett., 12, 
135, 1972) can very well be caused by an 
observational effect, namely, the distri- 
bution of the interferometer baselines 
used in the observations. Rowan-Robin- 
son's {Nature, 236, 112, 1972) proposition 



that some quasars are local, derived from 
the uniform V/V m distribution of small 
diameter 3CR quasars and their associa- 
tions with peculiar galaxies and Zwicky 
clusters, was examined and it was found 
that "compactness" of the spectra is a 
better criterion than linear size. The uni- 
form V/V m distribution is a property of 
only a 3CR sample with respect to its 
optical and radio detection limits. All 
associations between 3CR sources and 
Zwicky clusters on the sky are examined. 
From the number of such associations of 

radio galaxies with clusters of a different 
redshift, it is shown that the agreement 
in position between 3CR quasars and 
Zwicky clusters at a different redshift 
is accidental. It is also shown that there 
is no correlation between compactness of 
the spectra and membership of Arp's 
pairs of radio sources across peculiar 
galaxies. It is concluded that the angular 
diameters of quasars in no way contra- 
dict the cosmological interpretation of the 


Colors of E Galaxies as an Indication 
of Age 

It has been known since the early 
work of Stebbins and Whitford at Mount 
Wilson in the late 1940s that the giant E 
galaxies show only a small dispersion in 
color distribution. This fact was used by 
Baade and others as an argument in 
favor of nearly equal ages for E galaxies, 
in contrast to the continuous age distri- 
bution required by a steady-state cosmol- 
ogy where galaxies of all ages would 
be found. 

New BVR observations of giant E gal- 
axies in clusters and groups were begun 
by Sandage in 1967 for a study of the 
redshift-apparent magnitude relation ; 
the distribution of colors is a by-product 
of the photometry. Analysis of the data 
was completed during the report year, 
and comparison has been made of the ob- 
served color distributions with three 
models for the age distribution of giant 
E galaxies. 

The observations, made with the 
Mount Wilson 100-inch Hooker reflector 
for the brighter galaxies in groups and 
with the Palomar 200-inch for the fainter 
clusters, were corrected for the effects 
of latitude reddening due to our Galaxy, 
redshift effects on the colors, and meas- 
uring errors. An upper limit to the dis- 
persion of the color distribution for giant 
E galaxies is a{B — V) = 0.028 mag, 

with a distribution about the mean that 
is closely gaussian. 

To what extent does this mean similar 
ages for E galaxies? Stellar aggregates 
become redder as they age. An earlier 
study (Astrophys. J., 138, 863, 1963) of 
galactic clusters of known age and known 
integrated B — V color shows that the 
rate of reddening for these simple struc- 
tures is &(B — V) ~ 0.3 mag between 
ages of 10° and 10 10 years. Rates for E 
galaxies are expected to be less but not 
smaller than '—0.1 mag in the same time 
interval. The variation of color index 
with time will follow closely the form 
B — V csl a log t -\- b, where a is the 
reddening rate. 

Using this reddening rate, Sandage pre- 
dicted the shape of a color distribution 
curve for E galaxies in the situations 
where (1) E galaxies have formed con- 
tinuously over the past 10 billion years 
at a constant rate, so that there would be 
old and young galaxies in the sample; 
(2) the age distribution is that required 
by the steady-state model where new 
galaxies are born at the rate needed to 
keep the space density of galaxies con- 
stant in the presence of expansion that 
moves galaxies of all ages out of the 
volume; and (3) E galaxies formed over 
a small interval of time near 10 10 years 
ago, as in a Friedman-Lemaitre big-bang 

The predicted color distributions of 



models (1) and (2) for reddening rates 
between 0.3 mag/dex and 0.1 mag/dex 
are strongly nongaussian and have a 
much wider spread than a(B — V) = 
0.028 mag of the observations. Conse- 
quently, in agreement with earlier con- 
clusions of Baade, Sandage believes that 
the similarity of colors of giant E gal- 
axies requires that they formed over 
some small but definite interval early 
in the life of a Friedman-like big-bang 
universe. The results are given in Paper 
V of the redshift-distance series (Astro- 
phys. J., 183, No. 1, 1973). 

Absolute Magnitude of First-Ranked 
E Galaxies in Clusters and Groups 

Richness effect. The effect of the rich- 
ness of an aggregate of galaxies on the 
absolute magnitude of the first-ranked 
E galaxies was further investigated by 
Sandage, extending the results reported 
last year. Photometry of brightest gal- 
axies in certain small groups that were 
isolated in the 1956 study of redshifts 
by Humason, Mayall, and Sandage was 
reduced and corrected for aperture effect, 
iv-dimming, and galactic absorption. The 
data were plotted in the Hubble diagram 
for first-ranked E galaxies in great 
clusters (Year Book 71, pp. 689-690) and 
the deviations from the mean line were 
analyzed as a function of group or cluster 
richness. Surprisingly, there was no sig- 
nificant correlation: The absolute magni- 
tude of first-ranked galaxies in groups as 
small as 5 to 10 members is the same 
(to within la of the distribution) as in 
great clusters such as those in Coma or 
Virgo, which have hundreds of members. 
From this it would appear that there is 
something special about the formation of 
first-ranked group members: A galaxy 
grows to a maximum size (luminosity) 
initially, limited by some unknown effect 
that produces an upper bound in mass 
that is sharp at the <t(M v ) ~ 0.3 mag 
level (i.e., = 30% in mass). This upper 
limit is independent (or nearly so) of the 

total amount of matter (richness of the 
cluster) available at formation. 

Effect of Bautz-Morgan cluster form. 
Sandage and Hardy investigated the ab- 
solute magnitude of the first three- 
ranked galaxies in 97 groups and clusters 
of galaxies as a function of the form of 
the cluster given by the Bautz-Morgan 
classification scheme. This classification 
is based on the luminosity contrast of the 
brightest cluster member with the rest of 
the cluster. 

The absolute magnitude of the first- 
ranked galaxy is well correlated with 
Bautz-Morgan class. In Class I clusters, 
where the contrast is greatest, the first- 
ranked galaxy averages 0.4 mag brighter 
than the mean Hubble line in the magni- 
tude-redshift line. However, startlingly, 
the second- and third-ranked members in 
Class I clusters are fainter than normal. 
This inverse effect appears to be well 
determined and suggests that the domi- 
nance of first-ranked galaxies in clusters 
occurs at the expense of the fainter mem- 

The Bautz-Morgan contrast classes are 
not correlated with cluster richness. From 
this, Sandage and Hardy argue that the 
dominance of a first-ranked galaxy was 
not acquired by growth during the life- 
time of the cluster by accumulation of 
matter from the less massive members 
(by tidal encounters, for example) be- 
cause the frequency of such encounters 
would depend on richness, contrary to the 
observations. They believe, therefore, 
that the Bautz-Morgan effect is more 
probably related to an initial condition 
of cluster formation rather than to later 
evolutionary processes. 

Corrections for the variation of <M r > 
with Bautz-Morgan cluster type were 
made to all data, and new correlations of 
<Af v c > with cluster richness again showed 
that <M v c y for first-ranked cluster gal- 
axies had no significant variation (there 
was a la effect) with richness as the 
cluster size changes from 5 to 250 mem- 
bers in the first three magnitudes of the 
cluster. However, <M v c y for second- 



and third-ranked cluster members was 
strongly dependent on richness. 

The effect of the Bautz-Morgan cor- 
rection, and the slight (lo-) richness cor- 
rection on the redshift-magnitude dia- 
gram, was analyzed by making a new 
solution for the apparent value (no 
evolutionary correction) of the decelera- 
tion parameter after applying the new 
corrections. The change of q due to 
these corrections was negligible. The new 
value is q () = 1 ± 1 (the quoted error is 
a 2<r value) , which is the same as reported 
last year. Only the grossest alternatives 
to this value (such as q () = — 1 or 
g > 3) can be discarded by the pres- 
ent data. New data for many clusters 
with large redshift (z > 0.4) are needed 
for a finer solution, but Sandage and 
Hardy believe that the prediction of 
steady-state cosmology {q = — 1) is 
clearly at variance with the present data. 

Distant Clusters 

As part of the collaborative effort by 
Kristian, Westphal, and Sandage to ex- 
tend the redshift-magnitude diagram to 
very distant clusters with Westphal's 
new Silicon Intensified Target vidicon 
system, the first part of a catalog of 
distant cluster-candidates was completed 
by Sandage during the report year. A 
deep Illa-J photographic survey with 
the Palomar schmidt has been in progress 
over both the north and south galactic 
polar caps since 1969. To date, some 25 
fields have eben obtained with 2^/2-hour 
exposures that reach 2 mag fainter than 
the red Sky Survey plates. Six fields have 
been inspected for candidate clusters that 
have estimated redshifts larger than z ^ 
0.25. Many faint groupings are present. 
Only the best candidates have been re- 
tained in the working list for the SIT 
vidicon. The number of these candidate 
clusters averages 15 per field. Test 
plates on 3C295 show that the fainter 
one-third of the candidate clusters should 
have redshifts larger than z ~ 0.46. 

Routine observations using the SIT 
vidicon were begun on the candidate 
clusters by Kristian, Westphal, and 
Sandage during the report year. Other 
clusters for which conventional photo- 
metry exists were also observed in order 
to connect the program to existing data 
(cf. Instrumentation). 

The ultimate aim of the program is to 
obtain magnitudes, redshifts, colors, and 
diameters of galaxies in about 100 faint 
clusters in order to distinguish among 
world models. The first part of the pro- 
gram, which is now under way, is broad- 
band (BVR) photometry, yielding mag- 
nitudes, diameters, and colors directly. 
In addition, the V — R color provides 
a rough redshift as a guide for the second 
stage of the program, which calls for ob- 
taining line spectra for redshifts and 
energy distributions to investigate pos- 
sible evolutionary effects. 

Gunn and Oke have pursued an inde- 
pendent program to find faint clusters 
of galaxies. Their photographic search, 
made with the 48-inch schmidt and deep 
Illa-J plates, in practice appears to be 
capable of finding clusters with values of 
z up to about 0.30. The 90-mm image 
tube has been used at the prime focus of 
the 200-inch telescope to photograph 
about 5 to 6 square degrees of sky with 
deep Illa-J plates. This involves about 
100 plates so far. A number of very 
distant clusters of galaxies have been 
found on these plates. 

Oke and Gunn have continued to ob- 
tain spectral energy distributions and 
redshifts of galaxies in distant clusters 
discovered in the survey referred to 
above. Objects observed during the year 
that have redshifts z > 0.20 are given in 
Table 3 below. The program has pro- 
duced data for a total of 18 galaxies with 
redshifts greater than 0.20. Energy dis- 
tributions have also been obtained of 
galaxies in nearer clusters, including sev- 
eral Abell clusters and the Hydra cluster. 
These data will be used for synthesis 


TABLE 3. Magnitudes and Redshifts 




1600+42, 48" field 

Cluster 46 



Galaxy Gl near PKS 1049-09 



Galaxy Gl near PHL 1093 




Galaxy G1+G2 



Abell 2317 

Galaxy Gl 



Abell 1961 

Galaxy Gl 



Cluster 1319+30 

Galaxy G2 



Analysis of the new redshift material 
has begun; as yet no firm results have 
emerged because of the difficulty of esti- 
mating corrections for the large selection 
effects apparent in the data. 

Radio Galaxy Source Counts 

Schmidt has studied radio source 
counts in steady-state cosmology on the 
basis of identifications and redshifts of 
strong 3CR sources. The source counts 
corresponding to each of the identified 
sources are derived from the relation be- 
tween volume distance and flux density 
generated if the source is hypothetically 

moved to different redshifts. The method 
allows all the individual properties of the 
identified sources to be represented in the 
predicted source counts. Quasars were 
assumed to be local, since a cosmological 
interpretation of their redshifts leads to 
<y /y m y = 0.75 in steady-state cosmol- 
ogy, a value incompatible with steady 
state in which no evolutionary effects are 
admitted. The predicted total source- 
count curve is a factor of 5 below the 
observed counts at low radio flux densi- 
ties. The conclusion is that radio source 
counts are incompatible with steady- 
state cosmology. 


Chemical History of Galaxies 

Searle continued theoretical studies of 
element enrichment in galaxies. It is gen- 
erally believed that galaxies collapse 
from clouds of primordial material and 
that the heavy elements are formed in 
massive stars and are ejected from them 
into the interstellar gas. Each succeeding 
generation of stars is formed from pro- 
gressively more enriched interstellar gas. 
Ten years ago Schmidt found that if the 
distribution over mass of newly formed 
stars is the same at all times, it is difficult 
to reconcile this hypothesis with the ob- 
served scarcity of metal-poor stars in our 
own Galaxy. He proposed that in early 
times the stars that formed were mainly 
short-lived massive ones. 

Searle has reexamined this conclusion 
in an attempt to account for the observa- 

tion that the interstellar gas in all normal 
galaxies has a composition very similar 
to that in the solar neighborhood. He 
finds that this observation is very hard 
to understand unless the distribution over 
mass of newly forming stars is every- 
where nearly the same, averaged over 
sufficiently large volumes and times. 
Searle has shown that Schmidt's conclu- 
sion follows only if it is assumed that the 
interstellar gas of the collapsing galaxy 
is homogeneous at all times. He has 
proposed as an alternative model for the 
early evolution of galaxies that they are 
chemically inhomogeneous during their 
collapse and that the stars form prefer- 
entially in regions where the metal con- 
tent is unusually large. With models of 
this type, Searle has been able to ac- 
count both for the scarcity of metal-poor 



stars in the Galaxy and for the nearly 
uniform chemical compositions found in 
the interstellar gas of galaxies that have 
otherwise grossly different integral prop- 

Diffuse X-Ray Background 

Gunn has constructed mathematical 
models of the distribution of clusters of 
galaxies and their evolution in order to 
compute the expected contribution of 
clusters to the diffuse x-ray background. 
The conclusion is that clusters must con- 
tribute between a quarter and a half of 
the total observed background in the 0.5- 
10 keV range and may account for essen- 
tially all of it. A satisfactory theory of 
the fluctuations has also been developed, 
and definite predictions can be made that 
can be checked by x-ray detection of 
moderate spatial resolution ; existing data 
are not quite good enough. 

Magnetic Field of the Crab Nebula 

Gunn and Martin Rees of Cambridge 
have developed a model for the genera- 
tion of the magnetic field of the Crab 
Nebula and other supernova remnants 
which combines the wave ideas of 
Ostriker, Gunn, and Pacini and the uni- 
polar inductor picture of Goldreich and 
Julian. Explained in a natural way are 
the field geometry as inferred from polar- 
ization and the cavity -wisp structure ob- 
served near the pulsar. The work is 

Penumbral Waves and Umbral Flashes 
in Sunspots 

Observed properties of umbral flashes 
and running penumbral waves were re- 
ported last year by Zirin, Tanaka, and 
Stein. During the present report year, 
Moore concluded that a likely source of 
these periodic phenomena is the oscilla- 
tory convection which is expected to oc- 
cur in the superadiabatic subphotospheric 
layers of sunspot umbras due to the 
presence of a strong (2000-3000 gauss) 

vertical magnetic field. To test this idea, 
Moore computed the periods and growth 
rates for oscillatory modes arising in a 
simple two-layer model umbra. The re- 
sults indicate that umbral flashes result 
from disturbances produced by oscilla- 
tory convection occurring in the upper 
subphotospheric layer of the umbra 
where the superadiabatic temperature 
gradient is much enhanced over that in 
lower layers, while running penumbral 
waves are due to oscillations in a layer 
just below this upper layer. The oscil- 
lations in the lower layer may be driven 
by the oscillations in the upper layer, 
which would explain the phase-locking 
between umbral flashes and running pen- 
umbral waves. 

Photospheric Oscillations 

The well-established cause-and-effect 
notion for the 5-minute oscillations ob- 
served in the solar photosphere is that the 
oscillations are excited by disturbances 
generated by the motions in the convec- 
tion layer just below the photosphere. 
Moore has investigated the photospheric 
oscillations by analyzing the steady re- 
sponse of an isothermal atmosphere to 
pressure fluctuations at its base. This 
work was stimulated in part by the work 
of G. Worrall (Astrophys. J.\ 172, 749, 
1972), who concluded that the maximum 
response of the atmosphere occurs at a 
fixed frequency (o g , the maximum cutoff 
frequency for gravity waves (buoyancy 
waves), regardless of the horizontal scale 
of the perturbation applied to the base 
of the atmosphere. Moore has found that 
this behavior occurs only for downward 
traveling waves and hence does not apply 
to the case at hand, since causality re- 
quires the response of an atmosphere per- 
turbed from below to be due to waves 
traveling up from the disturbance rather 
than downward from infinity. When 
downward traveling waves are excluded, 
Moore finds that the response does de- 
pend on the horizontal scale of the ap- 
plied perturbations. For horizontal scales 



much longer than about 10 H, where H 
is the scale height of the atmosphere, 
the maximum response occurs at w a (^ 
o) g ) , the minimum cutoff frequency for 
compression (acoustic) waves. Only for 
perturbations of horizontal scale much 
shorter than 10 H does the maximum re- 
sponse occur at <o g . These results imply 
that the photospheric oscillations of ob- 
served horizontal scale ^ 4000 km are 
compression modes oscillating at <o a 
rather than buoyancy modes oscillating 
at w g . Either w a or u> g gives an oscillation 
period in the vicinity of 5 min, so it is the 
horizontal scale rather than the period 
that indicates the physical nature of the 

White Dwarfs 

Borra has computed models for the 
Balmer lines of DA white dwarfs with 
strong magnetic fields. It appears that 
the interpretation of quadratic Zeeman 
wavelength shifts to obtain, the mag- 
netic field is not simple. Different parts 
of the line give significantly different 
shifts. In particular, emphasizing the 
core of the line leads one to underestimate 
the violet shifts and therefore the mag- 
netic field. The line profiles and the 
shifts depend also on the geometry of the 
magnetic field. 

Magnetic Field of 53 Camelopardalis 

Borra has computed models for the 
magnetic field of the star 53 Cam. It is 
found that the linear polarization ex- 
pected from the transverse Zeeman effect 
in H/3 is two orders of magnitude lower 
than the linear polarization reported re- 

cently in H/3 by Kemp and Wolstencroft 
(Astrophys. J. [Lett.], 179, 153, 1973). 
It is unlikely that errors in the magnetic 
field configuration used or the approxi- 
mations in the models can explain the 

Zeeman-Broadened Line Profiles 

Borra has done extensive computations 
of the Stokes parameters (intensity, 
linear, and circular polarization) as a 
function of wavelength across a spectral 
line formed in the presence of a magnetic 
field. The models are oblique rotator 
models with dipolelike configurations of 
the magnetic field. Rotation is taken into 
account. The polarization profiles will be 
used in the interpretation of the photo- 
electric observations currently being 

Theoretical Zeeman-analyzed line pro- 
files are obtained from the Stokes param- 
eters. They are used to simulate photo- 
graphic observations of longitudinal mag- 
netic fields. It is found that the longi- 
tudinal fields inferred from those profiles 
can differ substantially from the true 
longitudinal field when only parts of the 
lines are used to compute the Zeeman 
displacements. In particular, very de- 
centered dipolar fields give profiles such 
that photographic spectroscopy will not 
reveal a reversal of polarity. This might 
explain the excess of stars that do not 
reverse sign of the longitudinal field. 

Many of the magnetic curves obtained 
from the models derived from only the 
line-profile cores mimic very well the 
curves of known magnetic stars, while the 
true curves from the models utilizing 
complete line profiles do not. 


Prime-Focus Corrector 

For the prime focus of the 200-inch 
telescope, a new field correcting lens de- 
signed by C. G. Wynne has become avail- 
able for use. The lens, having four ele- 

ments all of UBK-7 glass, was con- 
structed by the Optical Shop. It provides 
a field of excellent definition 25 minutes 
of arc in diameter. Between the limits 
3650 and 10,000 A, the aberrations of the 
lens were computed to be less than 0."3. 



With either photographic plates or the 
90-mm image tube, it provides very sig- 
nificant gains. 

SIT-Vidicon Area Photometer 

The SIT-vidicon photometer, described 
in Year Book 71 (p. 705), has been fur- 
ther developed by Westphal, Kristian, 
and Sandage and has been advanced from 
a developmental to an operational stage. 
Much of the effort this year has gone into 
adapting it for use at the prime focus of 
the 200-inch, where it is being used regu- 
larly, and in developing techniques for 
handling the large amount of data. Raw 
data (65,000 elements per exposure) are 
collected and stored digitally at the tele- 
scope and reduced later. The reduction 
ultimately requires the use of a large 
computer, but special hardwired equip- 
ment has been designed and built to 
facilitate and simplify early stages of the 
reduction. A playback system reads and 
stores the digital data tapes, either com- 
plete frames or selected parts of frames, 
processes them, and displays them in 
analogue form on an oscilloscope from 
which photographs of conventional ap- 
pearance can be made, intensity profiles 
examined, star positions located, etc. The 
playback system has proved especially 
valuable in early stages of examination 
and editing of the data and has increased 
the efficiency of using the large computer. 

200-Inch Cassegrain Spectrograph 

A new spectrograph has been designed 
and built for use at the Cassegrain focus 
of the 200-inch telescope. It is designed 
primarily for use with the television-type 
area sensors and will be capable of ob- 
taining slit-type spectra or absolute 
spectrophotometric data. The spectro- 
graph has both sky and object apertures 
and uses a cross-dispersing quartz prism 
to separate grating orders. The spectro- 
graph is being tested and is expected to 
be ready for use on the telescope in the 
fall of 1973. 

Prime-Focus Image Tube 

A 90-mm diameter single-stage mag- 
netically focused image tube has been 
put into operation at the prime focus of 
the 200-inch telescope with either the 
Ross or new Wynne corrector and at the 
Cassegrain focus of the new Palomar 60- 
inch telescope. The cathode of the tube 
is an extended S-20 type and the output 
is via a fiberoptic plate. Limiting expo- 
sures with 500 to 1000 A bandpasses on 
baked Illa-J plates are obtained in 5 
minutes or less with the 200-inch and in 
40 minutes or less with the 60-inch. The 
image quality does not depend on the 
image tube except for the faint honey- 
comb pattern produced by the fiberoptic 
output plate. 

Coude Image Tube 

A new image-tube system has been put 
into operation by Zappala at the 32-inch 
camera of the Mount Wilson 100-inch 
coude spectrograph. This system uses a 
40-mm diameter single-stage electrostat- 
ically focused image intensifier with 
fiberoptic input and output faceplates 
manufactured by Varo Electron Devices. 
The faceplates have been found to be of 
high quality and are relatively free of 
the shear and dislocations commonly 
found in such tubes. An excellent ex- 
tended-red-response S-20 photocathode 
allows spectra to be obtained in the near- 
infrared to beyond A8800. During opera- 
tion the cathode is cooled to approxi- 
mately 0°C to allow exposures longer 
than 3 hours without objectionable back- 
ground fog. 

Preliminary observational tests with 
unbaked Ila-D emulsion pressed into con- 
tact with the output screen show that 
speed gains of about 90 can be expected 
at A8400 compared to ammoniated IN 
emulsion, while gains at Ha are about 
10 and 20 compared to 098 and 103a-F 
emulsions, respectively. Although the 
tube can be used in the AA4000-9000 re- 
gion, little real information gain can be 



expected in the blue because of degrada- 
tion in resolution by the intensifier. 

Data System for 150-Foot Tower 

The digital data system for the 150- 
foot solar tower magnetograph was 
nearly completed during the report year. 
This system and the Raytheon 704 com- 
puter will be moved to the 150-foot tower 
telescope sometime during the late sum- 
mer or autumn of 1973. The computer 
will operate the telescope and magneto- 
graph and do the data reduction. 

Vacuum Telescope 

Through a grant to Zirin from the Na- 
tional Aeronautics and Space Adminis- 
tration for support of the Skylab pro- 
gram, construction was begun and largely 

completed on a new 26-inch vacuum 
telescope for the Big Bear Solar Observa- 
tory, to be placed in the existing fork 
mount in conjunction with two smaller 
vacuum telescopes. Final assembly is 
scheduled for the summer of 1973, with 
the principal remaining problem the com- 
pletion of the vacuum window. It is 
hoped that the new instrument will make 
possible full exploitation of the seeing at 
Big Bear. 

The new telescope will utilize two new 
birefringent filters recently purchased: 
one 0.3 A wide for the A10830 helium line, 
and a Universal Birefringent Filter capa- 
ble of use at any wavelength between 
A4000 and A7000, with bandpass ranging 
from 0.1 to 0.25 A in that range. The 
latter is being fabricated by Zeiss and 
will be delivered shortly. 


During the report year, the Hale Ob- 
servatories accommodated 56 guest in- 
vestigators representing 34 different in- 
stitutions. Fourteen guest investigators 
were from countries other than the 
United States. 

Dr. Saul J. Adelman of the NASA 
Goddard Space Flight Center has con- 
tinued his studies of sharp-lined magnetic 
Ap stars and of suitable main-sequence 
comparison standards. His current pri- 
mary interests are silicon and B5-B7 
stars. In March 1972, he obtained for 
several of these stars 4.3 A mm -1 IIa-0 
spectrograms with 100-inch coude and 
energy distributions in the region AA3200- 
7400 with the 60-inch Cassegrain scanner. 

Dr. J. R. P. Angel of the University of 
Texas and Dr. J. D. Landstreet of the 
University of Western Ontario used the 
multichannel spectrophotometer on the 
200-inch telescope together with a Pock- 
els cell modulator as a narrow-band 
multichannel polarimeter during two 
nights in August 1972. Observations of 
the circular and linear polarization spec- 
trum of the magnetic white dwarf Grw 
+70° 8247 were obtained. The new cir- 

cular polarization spectrum, obtained 
with 80 A resolution in the blue, is simi- 
lar to the one reported by Angel, Land- 
street, and Oke (Astrophys. J. [Lett.], 
171, Lll, 1972) but is more accurate. 
The linear polarization spectrum, ob- 
tained with 160 A resolution in the blue, 
has never been studied at this resolution 
before. In the blue some structure is ap- 
parent, with a polarization maximum of 
about 4% at 4000 A. The polarization 
falls to near zero at 5000 A, rotates 
smoothly through 90°, and rises again to 
a value of 3-4% near 1/x. This material 
has not yet been published. Preliminary 
measurements were obtained with 20 A 
resolution of the wavelength dependence 
of linear polarization of the Be star y Cas 
and of <f> Cas, which has strong inter- 
stellar polarization. Further observa- 
tions of this sort are planned in future 
observing runs. 

Dr. K.-H. Bohm and Dr. R. Schwartz 
of the University of Washington, using 
the multichannel spectrometer of the 
200-inch telescope, made observations of 
the condensations G and H (designations 
according to G. H. Herbig in Nonperiodic 



Phenomena in Variable Stars, Reidel, 
1969, p. 75) in Herbig-Haro object No. 2, 
as well as observations of the brightest 
part of Herbig-Haro object No. 1 and of 
the small nebula near T-Tauri. A band- 
width of 20 A was used in the whole spec- 
tral range from —3200 to 11,000 A. The 
line intensities are being used to deter- 
mine the electron density, electron tem- 
perature, and the degree of ionization in 
individual condensations in order to find 
out whether the physical parameters in 
the condensations depend on their pre- 
vious brightness changes. The ratio of 
the intensities of the infrared [S II] lines 
AAl0317/10336tothe [S II] AA4068/4076 
lines will be used to determine the red- 
dening of the spectra of the individual 
condensations. The energy distribution 
of the continuum (or quasi-continuum) 
and the B aimer decrement in individual 
condensations are also being studied. 

Dr. Clark G. Christensen of Brigham 
Young University, making use of the 
Cassegrain scanner at Mount Wilson, has 
obtained energy distributions in the 
wavelength range AA3448-8050 for the 
nuclei of globular clusters NGC 6229, 
NGC 6934, and M13, and of the Irr I 
galaxy NGC 4449. He also obtained 
energy distributions of the metal-defi- 
cient stars HD 122563, HD 151937, HD 
170153, M92 111-13, and M92 VII-18. 
These data will be used in globular clus- 
ter and galaxy population syntheses. 

In July 1972, Dr. S. V. M. Clube of 
the Royal Greenwich Observatory ob- 
tained direct plates, with the Mount Wil- 
son 60-inch telescope, of twelve Kapteyn 
selected areas. These are matched by 
the old photometric plates taken in 1911- 
1912 and will be used to determine proper 
motions of faint red stars. 

Observations for studying the be- 
havior of optical interstellar lines in dark 
clouds have been obtained by Dr. Judith 
G. Cohen of the University of California 
at Berkeley. The equivalent widths and 
radial velocities of the interstellar ab- 
sorption lines of Ca II, Na I, CH, and 
CH + were measured in 30 stars. The 

stars were selected to be within or be- 
hind dense clouds with up to 3 magni- 
tudes of absorption. The interstellar lines 
in the cloud stars are compared with those 
in a group of highly reddened supergiants 
whose reddening arises from a large 
distance, i.e., a long path length through 
relatively thin material. The atomic in- 
terstellar lines of the cloud stars are 
weak compared to those of supergiants 
with the same color excess; the molecular 
lines are of comparable strength in the 
two groups, except that the ratio 
ncu/^cn is larger in the cloud stars. The 
ionization equilibrium in the two cases 
can be approximated, and it is concluded 
that the deficiency of atomic Ca and Na 
relative to hydrogen in the clouds is 
about a factor of 100 larger than in the 
supergiants. A new series of observations 
is planned to study interstellar lines in 
stars at high galactic latitudes. 

Dr. Peter S. Conti of the Joint Insti- 
tute for Laboratory Astrophysics of the 
University of Colorado has employed the 
coude spectrographs of both the 100-inch 
and the 200-inch telescopes to study the 
spectra of O and Of stars. Spectra ob- 
tained of HDE 226868, the supergiant 
O-type star in the binary system which 
contains the x-ray source Cyg X-l, have 
been discussed in association with Mar- 
gon and Smith of Berkeley (Astrophys. 
J. [Lett.], 179, L125, 1973). They find 
the emission at A4686 of He II to be vari- 
able in intensity, and they suggest it is 
coming from material flowing in the sys- 
tem rather than from the primary star or 
secondary object. They also observe 
slight changes in spectral class in the 
supergiant, probably due to distortion 
effects in this close binary system. Conti 
obtained spectra of two anomalous O 
plus Of binary systems. In both cases, 
emission lines had previously been de- 
tected in one component, and absorption 
lines in the other; the mass ratios indi- 
cated that the Of star was the less mas- 
sive. For BD+40° 4220 the spectra ex- 
hibit absorption lines in both components, 
indicating that the previously derived 



mass ratio was incorrectly estimated. 
The lines are very broad and shallow and 
it has not proved possible to measure 
them accurately on a Grant machine. 
However, attempts are being made to 
obtain velocities, using microphotometer 
tracings that have sufficiently accurate x 
recording. For the other system, HDE 
228766, no certain absorption lines are 
detected in the Of component, but there 
are certain variations in line profiles in 
the star's absorption lines. Studies of 
this system are continuing with tracings. 

One excellent spectrogram of the 06f 
star, HD 153919, associated in a binary 
system with the x-ray source 2U 1700-37, 
has been obtained. Its spectral appear- 
ance suggests a high luminosity and a 
large mass, but this cannot be proven 
with certainty. Emission lines of Si IV, 
N III, C III, and A4686 He II are seen, 
although they are not unique to this Of 
star. A few spectra in the blue of some 
early Of stars in the h and x association 
have also been secured. These will be 
used to determine accurate line profiles 
to compare with theoretical models, 
which will include the physical effects of 
extended atmospheres. The reduction 
procedure from the microphotometer in- 
volves minimizing the noise due to the 
grain fluctuations by inverse Fourier 
transform filtering. The profiles are 
thereby smoothed without bias. The 
computer programs for this method are 
presently being developed. 

Conti also obtained two very good 
spectrograms of the supernova in NGC 
5253, which was detected just before his 
first coude run at Mount Wilson com- 
menced. Both spectra show the inter- 
stellar H and K lines in our own Galaxy, 
and in NGC 5253 itself. This has been 
discussed by Conti, Greenstein, and Wal- 
lerstein in Astrophysical Letters, 12, 101, 

In February 1973 Dr. Kyle Cudworth 
of the University of California at Santa 
Cruz used the 48-inch schmidt at Palo- 
mar to obtain second-epoch plates to be 

compared with the Sky Survey to deter- 
mine proper motions in the Praesepe 
cluster and the galactic anticenter. This 
study of Praesepe is in collaboration 
with Dr. Burton Jones of Greenwich. 

Dr. D. G. Currie, Mr. S. L. Knapp, and 
Mr. K. M. Liewer of the University of 
Maryland have used the amplitude inter- 
ferometer on the 60-inch and the 100- 
inch telescopes on Mount Wilson and the 
200-inch telescope on Palomar Mountain 
to conduct observations of the apparent 
angular size and the structure of late- 
spectral-type giant stars. These observa- 
tions were conducted using a modern 
version of the Michelson stellar inter- 
ferometer. The measurements are made 
in a sufficiently short interval of space, 
time, and wavelength to guarantee that 
any atmospheric degradation of the 
measurements is less than some predeter- 
mined value. Initial tests were conducted 
on the 100-inch at Mount Wilson in June 
1972, followed by a series of measure- 
ments in June and July. During Decem- 
ber 1972, a second series of measurements 
was conducted at Mount Wilson and 
Palomar Mountain. The diameters of 
four stellar disks have been measured: 
a Ori (Betelgeuse) , a Tau (Aldebaran) , 
a Boo (Arcturus), and ft Peg (Scheat). 
The measurements, their accuracy, and 
the wavelength regions studied have been 
reported elsewhere. The better sets of 
measurements have an internal precision 
of 0.002 arc seconds (standard deviation 
of the mean). The wavelength depend- 
ence of the effective diameter (which de- 
pends upon both the geometric diameter 
of the stellar disk and the degree of limb 
darkening at that wavelength) was also 
measured. Finally, preliminary measure- 
ments were made on the apparent oblate- 
ness of a Ori and on the orbital param- 
eters of a Aur (Capella). Further obser- 
vations on a Boo, a Her and a Sco were 
conducted in June 1973. A continuing 
program of observations and instrument 
improvement is planned. In particular, 
a new instrument, the multi-aperture 



amplitude interferometer, is planned 
which will increase the sensitivity by 
about ten magnitudes. 

Dr. S. J. Czyzak of Ohio State Uni- 
versity and Dr. L. H. Aller of the Uni- 
versity of California at Los Angeles have 
continued their spectrophotometry of 
planetary nebulae. Observations were 
secured of 29 nebulae in the range 
AA4400-7800. They also observed thfe 
diffuse nebula NGC 604 in M33. Reduc- 
tions have been completed for 12 of these 
objects. For most planetary nebulae, 
photoelectric measurements give the 
monochromatic fluxes integrated over the 
object. Because of the necessity of using 
wide apertures, the spectral resolution is 
often low. In some instances, it has been 
possible to secure higher resolution data. 
These higher resolution data invariably 
refer to regions of small angular size in 
the nebula and can be compared with 
photoelectric measurements only if the 
relative spatial energy distribution in the 
monochromatic images is known. A spe- 
cial effort has been directed to NGC 7027. 
Here the photoelectric scanner measure- 
ments must serve as fundamental cali- 
bration for other, higher spectral resolu- 
tion measurements. Reductions covering 
the spectral range from A3100 to A8500 A 
are now completed. The spectrograms of 
NGC 6543 secured at Palomar in 1971 
have been reduced and utilized to inves- 
tigate the physical conditions (electron 
density, temperature, and level of excita- 
tion ) at various points in the filamentary 
pattern of this object. Results for NGC 
40 {Astrophijs. J.] 172, 361, 1972), NGC 
6778 (Astrophijs. J., 181, 817, 1973) have 
been published; those for NGC 6445 are 
in press. The discussion of NGC 6803 
has been written and is ready to be sub- 

Dr. R. J. Dickens of the Royal Green- 
wich Observatory used the Palomar 60- 
inch telescope to obtain spectra of gal- 
axies in the clusters Abell 1367 and 
I311 53m _j_ 390 In con j unc tion with 

spectra taken on other telescopes, the 
measured redshifts will be analyzed to 

yield velocity dispersions and estimates 
of the virial masses of the clusters. 

Mr. Dainis Dravins of the Lund Ob- 
servatory carried out solar observations 
simultaneously with the magnetograph 
at the Mount Wilson 150-foot tower and 
with a photographic telescope at Big 
Bear. These studies aim at determining 
the magnetic features of network ele- 
ments in quiet regions on the sun. At Big 
Bear an attempt was made to detect 
possible transient electric fields in the 
energy buildup phase preceding a flare 
by measuring linear polarization in the 
wings of a line sensitive to the Stark 
effect. However, no such fields were seen. 

Dr. Yoshio Fujita of the University of 
Tokyo is continuing his study of carbon 
stars. In late carbon stars the shorter 
wavelength region below A4380 is masked 
by strong absorption considered to be 
due to C 3 . However, carbon stars of 
comparatively high temperature do not 
show such a remarkable feature and well- 
exposed spectrograms are extended to the 
violet region far beyond the H and K 
lines. The purpose of this investigation 
is to identify the spectral feature short- 
ward of A4380 in such a star as HD 
182040 (RO, CL>). The comparative 
study in connection with the problem 
described above was carried out using 
spectrograms of HD 182040 (RO, CL>), 
HD 156074 (RI, CD) and RU Cam (RO, 
COi) obtained with the 100-inch and the 
200-inch telescopes. The coude spectro- 
graph of the 100-inch telescope has been 
used to obtain spectrograms of carbon 
stars with a dispersion of 13.6 A mm" 1 in 
the spectral region from A6400 to A8900. 
One purpose is to estimate the intensity 
of the forbidden [CI] A8727 line in some 
carbon stars. A preliminary investiga- 
tion (Y. Fujita, Proc. Jap. Acad., Ifi, 
295, 1970) revealed the presence of this 
line in some late carbon stars. It is de- 
sirable to extend this investigation to 
some early carbon stars and to establish 
the intensity variation of atomic carbon 
along the spectral sequence; this should 
give some knowledge on the dissociative 



equilibrium of carbon species in stellar 
atmosphere. The second purpose is to 
estimate the abundance ratio 12 C/ 1:5 C by 
comparing the intensity of 12 CN satellite 
lines with some isolated rotational lines 
of 1H CN in the spectral region from A7750 
to A8060. This was discussed in a pre- 
liminary paper (Y. Fujita, T. Tsuji, and 
H. Maehara, Proc. Jap. Acad., 4-5. 484, 
1969). Following this method, the abun- 
dance ratio of 12 C to 13 C will be deter- 
mined with regard to late carbon stars as 
well as early carbon stars. The value of 
12 C/ 13 C in the spectral sequence of car- 
bon stars may give information on the 
chemical elements' synthesis in these 

Dr. Tom Gehrels of the University of 
Arizona used the Palomar 48-inch 
schmidt for a survey of asteroid and 
comet populations; the limiting magni- 
tude was near 21.0. New Mars-orbit 
crossing asteroids were discovered; the 
survey is continuing for statistical de- 
termination of the magnitude-frequency 
relation. A high density was found for 
the Hungaria asteroids, which have near- 
circular orbits immediately beyond that 
of Mars. Trojan asteroids in both La- 
grangian points of Saturn were searched 
for, but none was found; the same pre- 
liminary conclusion was reached for 
Neptune, but that search is continuing. 
A preliminary result also has been 
reached regarding the occurrence of dis- 
tant comets (perihelion distance q > 3 
AU). These comets are of particular in- 
terest because of their relative freedom 
from nongravitational effects and be- 
cause they may be original condensations 
on their first passage into the solar sys- 
tem. Three such comets were discovered, 
one of which was subsequently identified 
by Dr. B. G. Marsden at the Smithsonian 
Astrophysical Observatory as Comet 
Swift 1889 IV. 

Dr. Alexander F. H. Goetz of the Jet 
Propulsion Laboratory of the California 
Institute of Technology is using a silicon 
vidicon photometer and the Leighton 
Image Motion Compensator to produce 

high quality, high dynamic range pic- 
tures of Saturn. These digital pictures 
are further analyzed and enhanced by 
the JPL Image Processing Laboratory. 
During this spring's run he was plagued 
by TV interference and poor seeing con- 
ditions. The TV problems have been cor- 
rected and he is looking forward to ob- 
taining suitable data in the fall. 

Dr. R. F. Griffin of the Cambridge 
Observatories and Gunn used the radial 
velocity spectrometer at the 200-inch 
coude focus on three nights during the 
report year. Observations were made of 
stars in the regions of the Hyades, M67, 
and M3. A photographic spectrogram of 
BD+12°1919, newly demonstrated by 
radial velocity to be the brightest mem- 
ber of M67, was obtained with the 36- 
inch camera of the coude spectrograph. 

Dr. F. D. A. Hart wick of the Univer- 
stiy of Victoria and Dr. Robert D. Mc- 
Clure of Yale University Observatory 
have done intermediate-band photoelec- 
tric photometry of giant stars in several 
outlying old stellar systems, using the 
200-inch Hale telescope. The ultimate 
goal is to intercompare the surface gravi- 
ties, effective temperatures, and metal 
abundances of the giant stars in these 
systems and some of the bright halo 
globular clusters. Giants in the remote 
globular cluster NGC 7006 were found to 
have a cyanogen excess compared to 
giants in other clusters with the same 
ultraviolet excess. Previously, Sandage 
and Wildey found this cluster to be 
anomalous in having too red a horizontal 
branch for its metal abundance. These 
results suggest that the CNO abundances 
may be the second parameter affecting 
the color-magnitude diagrams of globular 
clusters. The dwarf spheroidal galaxy in 
Draco, which has a similarly anomalous 
horizontal branch, was also observed. 
The giants in this system exhibit an ultra- 
violet excess comparable to that of the 
most metal-poor globular clusters in the 
Galaxy. The observations also suggest 
that the surface gravities of giants in 
this system may be significantly different 



from globular cluster giants in the Gal- 
axy. Further observations, using the 
multichannel spectrometer, of giants in 
the Ursa Minor dwarf galaxy are in 

During the year Dr. T. V. Johnson and 
Dr. D. L. Matson of the Jet Propulsion 
Laboratory of the California Institute of 
Technology used the Mount Wilson 60- 
inch telescope to carry out spectropho- 
tometry of asteroids in the visual and 
near infrared (0.3-2.5 /xin) wavelengths. 
During July 1972 the earth-crossing 
asteroid (1685) Toro came unusually 
close to the earth and was bright enough 
for spectrophotometric study in the 0.4 
to 0.8 fxm wavelength range. This was 
the first detailed spectral information ob- 
tained for a member of this important 
asteroid group which is believed to be 
the immediate source of some meteorites. 
In October 1972, observations of asteroid 
(43) Ariadne were made in five colors 
from 0.34 to 0.95 tim. It was discovered 
that this asteroid has a much larger light- 
curve amplitude (~0.7 mag) than had 
been reported previously and that it also 
has a significant solid state absorption 
band near 0.95 /xm. The spectral reflec- 
tance of this asteroid matches that of or- 
dinary chondrite meteorites and that of 
Toro, but is not similar to other asteroids 
in the main belt. Thus, it may be part of 
the sought-after source or reservoir of 
chondritic material. Observations at 1.2, 
1.6, and 2.2 /xm of four large asteroids 
(Ceres, Pallas, Juno, and Vesta) were 
made on several occasions in cooperation 
with G. Veeder and S. Loer of the Cal- 
tech Infrared Group. This is the first 
time observations of asteroids have been 
made at these wavelengths, and the re- 
sults will extend the wavelength cover- 
age of their spectral reflectivity by more 
than a factor of 2. Preliminary results 
have indicated large differences between 
these four asteroids and have suggested 
that the surface composition of Ceres, 
and perhaps that of Pallas, resembles 
Type-I carbonaceous chondrites. If so, 

then the surfaces of these two asteroids 
may be very old and unmodified by endo- 
genic forces sinCe the time of accretion. 

In one run with the 200-inch coude 
spectrograph at Palomar, Dr. Philip C. 
Keenan of the Perkins Observatory ob- 
served several very luminous red super- 
giants with 098-01 plates. The purpose 
was to test the effectiveness of the widths 
of metallic absorption lines as measures 
of luminosity for G, K, and M stars of 
luminosity classes lab and brighter. At 
the moderate scale of 13.5 A/mm it was 
found possible to confirm the high lumi- 
nosity of four supergiants by comparison 
with BS 8752 (G5 0) and RW Cep (KO 
0-Ia) . Most interesting was the very red 
star, no. 1 in the cluster Tr 27, which 
appears to be CoD— 33°12241. In the 
study of this cluster by The and Stokes 
they had measured the color of no. 1 as 
B — V = 2.83, and since its position co- 
incides with a very heavily absorbed re- 
gion at one edge of the cluster, they 
thought no. 1 might be an enormously 
reddened OB star. Albers, however, ob- 
tained an IR objective-prism spectro- 
gram on which he could see weak CN 
bands, showing the star to be of late 
type. In spite of the southern declina- 
tion, several degrees below the galactic 
center, the 098-01 spectrogram was well 
exposed and shows greatly widened lines 
and indicates a type of about K5 and a 
luminosity class near la. Although no. 1 
is less than 4' distant from the four 
brightest blue stars in the cluster, its 
membership is not certain and it may lie 
behind the cluster. 

Mr. Richard G. Kron and Air. David 
C. Koo of the University of California at 
Berkeley have used the 48-inch schmidt 
to search for highly obscured stars in the 
Taurus clouds. Because of poor weather, 
the only plate obtained of the region 
reached only m = 16 at A ( , ff = 8300 A and 
was badly affected by differential refrac- 
tion. A plate of the dark clouds in Ophiu- 
chus was also obtained, but the exposure 
was limited by dawn. A very red star was 
found at 16 i; 22»\3, — 24° *22'.5 (1950). 



for which Luis Carrasco has observed 
about 15 magnitudes of extinction in the 
visible. At the request of Francois 
Schweizer, Kron and Koo obtained a 1-N 
plate of the interacting galaxies NGC 
4038-39. At this wavelength the nuclei 
stand out above the emission regions, and 
the identification agrees with that de- 
duced spectroscopically by Rubin et al. 
(Astrophys. J. 160, 801, 1970). 

Dr. A. Labeyrie of the Paris Observa- 
tory has developed the speckle interfer- 
ometry program at the 200-inch tele- 
scope, which has a unique resolution 
advantage in addition to the luminosity 
advantage. For efficient use of the ob- 
serving time, a television system was 
specially built and used at the prime 
focus in June, October, and April. Ap- 
proximately two million images were 
recorded. This has created a processing 
problem requiring some technological de- 
velopment, which is being carried out at 
Meudon. However, preliminary process- 
ing work has evidenced color-dependent 
limb-darkening features in o Ceti, a 
Orionis (Bonneau and Labeyrie, Astro- 
phys. J. [Lett.], 181, LI, 1973) and a 
Scorpii (to be published) , as well as pre- 
viously unknown companions in 8 Scorpii, 
o- Hercules, i Serpentis, BSC 6927, (3 
Corona Borealis and R Lyrae. The limb- 
darkening observations suggest the pres- 
ence of a blue-scattering envelope around 
cool stars. It was recently found that 
speckle interferometry observations can 
in principle be extended to objects much 
fainter than previously assumed. Cur- 
rent plans are to add an image-intensifier 
stage to the present recording system; 
this should permit a limiting magnitude 
of the order of m v = 22 to be reached. 

Mr. Howard H. Lanning of California 
State University at San Diego has been 
studying the photometric variability of 
the short-period binary Humason- 
Zwicky 22 (VX Canum Venaticorum). 
Differential UBVr and no-filter observa- 
tions of the 13th magnitude object have 
been obtained using an S-20 photocath- 
ode on the Mount Wilson 60-inch tele- 

scope. Poor weather conditions severely 
limited the total number of acceptable 
nights. The light curve was, however, 
successfully obtained and will be ana- 
lyzed in detail shortly. Lanning expects 
to continue the survey of the two-color 
48-inch schmidt plates of the galactic 
plane begun recently at Hale Observa- 
tories as part of the program on optical 
identification of x-ray sources. Arrange- 
ments have been made for loan of the 
plates, and equipment is presently being 
constructed at San Diego to aid in their 

Dr. Craig D. Mackay and Mr. R. J. E. 
Kraft of the Institute of Astronomy, 
Cambridge, England, have used the 48- 
inch schmidt telescope to observe fields 
centered on the clusters of galaxies Abell 
1367, Abell 2151, and Abell 2199 in six 
colors. The plates have been calibrated 
using Gunn's wedge sensitometer, and 
the reduction will be done with the GAL- 
AXY mk. 2 measuring machine at Edin- 
burgh. These data will allow a detailed 
study of the color structure and morphol- 
ogy of the cluster over a range of nearly 
7 mag. 

Dr. T. B. McCord of the Massachusetts 
Institute of Technology employed the 
60-inch and 100-inch telescopes with the 
MIT silicon vidicon to record images of 
Jupiter, Saturn, and Uranus through 
narrowband interference filters. Images 
were obtained in methane and in some 
cases ammonia-absorption bands and at 
nearby continuum wavelengths. Maps of 
methane absorption were obtained. The 
same system was used to obtain visible 
and 1.0 /iiii images of several galaxies. 
Despite bad weather, a series of high- 
resolution images of geologically inter- 
esting regions of the moon was obtained 
using interference filters. These images 
are extremely important, for they show 
the boundaries of compositional units 
and are in fact geologic maps of a sort. 
The work is described in a publication in 
preparation. The spectrometer was first 
used with the vidicon during the run. A 
series of calibration and sensitivity tests 



were run. Spectra 0.3-1.2 jum were ob- 
tained for a selection of standard stars 
and for Saturn and several areas of the 
moon. The devices worked quite well. 

Dr. Walter E. Mitchell, Jr., of the 
Perkins Observatory utilized the com- 
puting facilities at Caltech and to a 
lesser extent the Raytheon 703 computer 
at Santa Barbara Street to reduce data 
obtained at the Snow Telescope during 
1971. These data were obtained for two 
programs: (1) the construction of a 
spectrum map from 3000 A to the atmos- 
pheric limit, and (2) the limb-darkening 
measurements for the determination of 
the run of solar opacity in the wave- 
length range 4000 A to 3000 A. In the 
first program, the digitally recorded 
spectra were "spliced" and combined into 
a single high-resolution intensity map. 
In the second, the numerical corrections 
for the instrumental profile were worked 
out and the limb-darkening coefficients 
for the gray case were calculated for a 
first approximation. 

During the past year, two types of 
television sensors were used at Palomar 
by Dr. D. C. Morton of Princeton Uni- 
versity Observatory. In October 1972 the 
SEC vidicon integrating system was 
mounted on the coude spectrograph to 
obtain further high-resolution spectra of 
quasars and galaxies. The objects ob- 
served included PHL 957 for a detailed 
analysis of the absorption lines, and the 
central regions of M31 and M32 for fur- 
ther studies of velocity dispersions and 
rotation curves. An earlier TV spectrum 
of M31 (Astrophys. J., 180, 705, 1973) 
showed that the line-of-sight dispersion 
in the nucleus was only 120 ± 30 km s" 1 . 

In March and May 1973, in collabora- 
tion with Oke, a new photon-counting 
television system was tested for direct 
imagery at the Cassegrain focus of both 
the 60-inch and 200-inch telescopes. 
Amplifications of 2 X 10 5 are possible 
from a two-stage tube that has a silicon 
target, so that individual photoelectron 
events can be detected and counted by a 
fast digital memory with a dynamic 

range of 2 16 for each pixel in a 256 X 256 
format. The system was operated suc- 
cessfully for several nights, but further 
work is needed to calibrate spatial irreg- 
ularities in the tube efficiency and reduce 
significant losses of counts in the circuits. 

Dr. Thomas W. Noon an of the State 
University College at Brockport, New 
York, continued his study of counts of 
galaxies in clusters of galaxies. He ob- 
served two clusters (Abell 234 and Abell 
732) and obtained data from existing 
plates of five other clusters (1306+2716; 
1448+2617; 0024+1654A; 0024+1654B; 
and 1410+5224) . The radii of the seven 
clusters, which range in redshift from 
0.17 to 0.46, are being studied, along with 
radii from published counts of other 
clusters, for use in the angular-size 
cosmological test. 

Using the coude spectrograph of the 
100-inch telescope, Dr. John Norris of 
Yale University Observatory has ob- 
tained high-dispersion spectra of the sub- 
dwarf B-type star HD149382 and the 
high-velocity B star HD125924. Five 
analyses of these spectra are under way 
to study the heavy element abundances 
in the groups of objects to which the 
stars belong. 

Dr. Valdar Oinas of the University of 
Nebraska at Lincoln employed the Casse- 
grain scanner at the Mount Wilson 60- 
inch telescope to obtain energy distribu- 
tions for a selection of late-type stars. 
The purpose of the scans was twofold: 
(1) to provide temperatures for an anal- 
ysis of a discrepancy between ion and 
neutral line strengths that appeared dur- 
ing his thesis work at Caltech; (2) to 
delineate more accurately a suspected 
wide absorption feature near 8800 A. 
Two of the three nights assigned were of 
photometric quality and good results 
were obtained. 

Dr. B. E. J. Pagel of the Royal Green- 
wich Observatory used the coude of the 
100-inch telescope during April and May 
1973 to secure several 2.9 A mm -1 spectra 
in the yellow and red of the "super- 
metal-rich" candidates fx Leonis and /? 



Oph and standard stars e Vir and a Ser. 

Dr. Sidney B. Parsons of the Warner 
and Swasey Observatory observed the 
classical Cepheid rj Aquilae and several 
nonpulsating supergiants of similar tem- 
perature at the coude focus of the 100- 
inch telescope. Five out of the seven 
scheduled nights were clear and gave a 
good sampling of phases for the 7.2-day 
variable. Observations were made with 
the 32-inch and 73-inch cameras on am- 
moniated I-N, Ila-D, and baked IIa-0 
emulsions at reciprocal dispersions of 20, 
4.5, and 3.0 A mm" 1 , respectively. The 
spectra are intended primarily for work 
on the effective temperatures and surface 
gravities of F- and G-type supergiants 
and Cepheids, although radial velocities 
will also be measured. In the near- 
infrared, profiles of Ca II and Paschen 
lines are sensitive primarily to tempera- 
ture and gravity and are not blended 
badly into other lines. Lines of N I are 
also present and will be a test of the in- 
creased abundance expected after mixing 
in the red-giant phase of evolution. In 
the near-ultraviolet, blending of numer- 
ous absorption lines seriously hampers 
the study of the Balmer jump, which is 
quite sensitive to temperature and grav- 
ity for these stars. The new observations 
down to about 3300 A will give a much 
better calibration of synthetic spectra 
being computed in collaboration with Dr. 
R. A. Bell (at the University of Mary- 
land), and the computations will allow 
good model-atmosphere analysis of pho- 
tometric measurements of the Balmer 
jump. H/3 profiles for a few stars were 
photographed in order to provide a check 
on the temperatures; H/3 is not sensitive 
to surface gravity in cool stars, is not 
blended too badly, and is not as sensitive 
as Ha to the presence of a chromosphere. 

Dr. S. Eric Persson and Dr. Jay A. 
Frogel of Harvard College Observatory 
began a spectrophotometric study with 
the Mount Wilson 100-inch and Palomar 
60-inch telescopes of northern H II re- 
gions that are known to be infrared 
sources (Sh 138, 156) . The planetary 

nebula BD-|-30 o 3639 was shown to have 
an optical continuum at H/3 slightly in 
excess of that expected, although the re- 
sult is quite sensitive to the electron 
temperature assumed. The explanation 
is that the particles believed responsible 
for the known large infrared excess also 
scatter a small fraction of the light from 
the central star. They used the 200-inch 
multichannel spectrometer in September 
1972 to measure emission lines in the 
planetary nebulae NGC 7027 and NGC 
7662 at 2 and 3.5 arc sec resolution. A 
complete study of the level of ionization, 
temperature, reddening, and He I-line 
ratio variations from point-to-point is 
under way. One interesting result for 
NGC 7027 is that the interstellar extinc- 
tion at H/3 varies by a factor of 3 across 
the face of the nebula, consistent with the 
very different appearance of the object 
at optical and radio frequencies. Some 
time was spent measuring the energy dis- 
tributions of two H II regions, K3-50 and 
Sh 228, which are infrared sources. Meas- 
urements of several suspected young 
planetary nebulae having large 10 ^ ex- 
cess emission (IC 4997, MHa 328-116, 
VV 8, HBV 475) were obtained. The 
data show evidence for high densities, 
^lO 6 to 10 7 cm -3 , and will be compared 
with previous spectrographic data. 

During the summer of 1972, Dr. Daniel 
M. Popper of the University of Cali- 
fornia at Los Angeles carried out UBV 
photometry of eclipsing binaries with the 
20-inch reflector on Palomar Mountain. 
The purpose of the observing program 
was to obtain orbital inclinations, frac- 
tional radii, and colors of systems for 
which he had spectrographic orbits but 
for which adequate photometry was not 
available. These systems are double- 
lined binaries, and complete spectro- 
graphic and photometric analyses will 
provide fundamental masses and radii of 
some precision. The systems observed 
most intensively were V805 Aql, TV Cet, 
V478 Cyg, BS Dra, and BK Peg. The 
observations have not yet been analyzed. 



Dr. Diane M. Pyper of the University 
of Wisconsin at Parkside has made a 
study of periodic magnetic stars, using 
the scanner on the Mount Wilson 60-inch 
telescope. The low-resolution spectrum 
scans (20 A band pass) are accurate 
enough to reproduce the variations pre- 
viously published for the program stars 
(this was expected, of course) . So far, 
scans of t Cas, 56 Ari, 53 Cam, a CVn, 
HD 153882, 73 Dra and HD 224801 
(CGAnd) have been studied. The varia- 
tions in M(5556) for these stars have 
been compared with published V varia- 
tions and in all cases the amplitudes of 
variation and the scatter in the observed 
points agree with the published data. In 
some cases it may be possible to improve 
upon the published periods. There is 
evidence for variation of the Balmer 
jump index [M(4255) — M(3636)] in 
several stars; the Paschen continuum 
index [M(4255) — M(6055)] also seems 
to change in some stars. The amplitudes 
of these variations are from about 0.08 
mag to 0.03 mag from the present data. 
Line blocking effects have not yet been 

Dr. Paul H. Richter of California State 
University at Northridge made observa- 
tions at the Cassegrain focus of the 
Mount Wilson 60-inch telescope to study 
the effects of atmospheric turbulence on 
the optical contrast transfer function for 
the atmosphere as a function of integra- 
tion time and zenith angle, and to at- 
tempt some short-exposure photography 
of Jupiter and Ganymede through the 
use of a magnetically focused, single- 
stage image tube. This program is part 
of an ongoing effort to utilize digital 
image processing techniques to enhance 
atmospherically degraded images of arbi- 
trary, extended objects and so resolve 
detail not visible in photographs ob- 
tained by conventional means. The data 
gathered during the above observing 
schedule has aided in the understanding 
of the effect of turbulence on astronom- 
ical images and has permitted some very 
substantial improvements to be made in 

the resolution of selected images of 

Dr. W. L. Sanders of New Mexico 
State University has a program to use the 
100-inch Newtonian nebular spectro- 
graph to obtain classification spectra of 
faint Be stars (B = 12-14.5). The ob- 
ject is to identify the la Supergiants 
among those stars for use as spiral arm 
tracers. Due to bad weather no results 
have been obtained so far. 

A search program for new Apollo- 
Amor asteroids was initiated in January 
1973 by Dr. Eugene M. Shoemaker and 
Mrs. Eleanor Helin of the California In- 
stitute of Technology, using the 18-inch 
schmidt telescope at Palomar Mountain. 
Although adverse weather has reduced 
the anticipated sky coverage, 174 fields 
had been obtained by early July. As 
each field is photographed twice, the 
number represents approximately 87 in- 
dependent fields. An exceptional Apollo 
asteroid has been discovered (Int. Astron. 
Union Circular No. 2555) within the 
statistically predicted number of fields 
necessary for discovery of a new object. 
Preliminary results of the new discovery, 
1973 NA, are the following: it had the 
greatest apparent motion at discovery of 
any known Apollo asteroid (greater than 
ll°/day); it has an inclination of ap- 
proximately 70°, the highest of any 
known asteroid; and, at discovery mag- 
nitude of 10, it has an absolute magni- 
tude of 14, which represents one of the 
brightest Apollos. It is anticipated that 
this unique Apollo will provide new in- 
formation for the analysis of the dynam- 
ical evolution of these objects. A second 
set of observations of 1973 EA (a new 
Apollo asteroid) was obtained by Mrs. 
Helin, which permitted still further ob- 
servations to be secured in Europe and 
led to the establishment of its orbit. 
Helin 's observations and positions were 
announced in the Int. Astron. Union Cir- 
culars, Nos. 2507 and 2509. In addition, 
several objects were detected but were 
unconfirmed because of periods of in- 



clement weather and the fast motion of 
the objects. 

Dr. R. W. Shorthill of The Boeing Com- 
pany and Dr. T. F. Greene of The Boeing 
Company and the University of Washing- 
ton have continued their study of the at- 
mosphere of Jupiter and of the Galilean 
satellites by observing an ingress and 
egress of Callisto in late 1972. This event 
enabled probing of the extreme south 
polar region of Jupiter. No refractive tail 
was observed, indicating a significant 
aerosol content in the south polar Jovian 
atmosphere at the 10 18 — 10 19 cm -3 level. 
This result stands in contrast to the pres- 
ence of a refractive tail (reduced rate of 
light intensity falloff) on the latter por- 
tion of several other observed satellite 
light curves. A complete discussion of 
the nature of the aerosol haze at the 
levels probed during the observed eclipses 
is in preparation. The aerosol haze is 
stratified and apparently not uniformly 
distributed in longitude even on the size 
scale of the region sampled by the beam 
illuminating the satellite. The variation 
of the integrated aerosol extinction cross 
section with altitude has been determined 
at 1.05 fi. At the time of this writing, 
other shorter wavelength data are also 
being interpreted. The wavelength de- 
pendence of the aerosol extinction cross 
section and a number of other properties 
of the atmosphere are being sought in 
this study. 

Dr. Sidney van den Bergh of the David 
Dunlap Observatory reports that the 
dwarf spheroidal galaxy And III, which 
appears to be a companion to the Androm- 
eda nebula, has been resolved into stars 
with the 200-inch Hale Telescope. Since 
resolution in this galaxy occurs at the 
same magnitude level as does the resolu- 
tion of NGC 185, it follows that these 
two objects are at approximately the 
same distance. In collaboration with 
Herbst and Pritchet, an area of 6.2 
square degrees near M31 has been 
searched for variable objects down to the 
limit of the 48-inch schmidt telescope. 
Thirteen faint variable objects were dis- 

covered close to the plate limit. The ob- 
jects detected in this survey might be 
either (a) quasars of intermediate color, 
(b) very distant galactic or possibly 
intergalactic variable stars, or (c) in 
some cases very distant galaxies that 
increased their total luminosity because 
of a supernova outburst. The ten-year 
program on the study of slow variables 
in galaxies of the Local Group and some 
members of the M81 group and South 
Pole Group was continued with the 48- 
inch schmidt telescope on Palomar 

An optical atlas of galactic supernova 
remnants, based in part on plate mate- 
rial obtained with the 200-inch and 48- 
inch telescopes, has been produced by 
van den Bergh, Marscher, and Terzian 
and will appear as Astrophys. J. Supp. 
No. 227. The atlas presents the best 
available photographs of all known op- 
tical supernova remnants. Comparison 
of Baade's 1950 plate of Kepler's super- 
nova with 200-inch plates taken twenty 
years later shows that the fan-shaped 
remnant of this object is expanding with 
a proper motion of ^- O."03 per year. 
Combining this result with Minkowski's 
radial velocity observations and assum- 
ing a distance of 10 kpc yields an expan- 
sion velocity of ^1400 km s _1 for the 
optical remnant of Kepler's supernova. 
This value is very similar to the expan- 
sion velocity of the Crab Nebula. Since 
Kepler's supernova was almost certainly 
a supernova of Type I, the low expansion 
velocity of the Crab can no longer be 
used as an argument against the hypoth- 
esis that the Crab Nebula was produced 
by a supernova of Type I. Additional 
spectrograms and direct plates were ob- 
tained of the optical remnant of Cassi- 
opeia A. It it hoped that these ob- 
servations will lead to an improved 
understanding of the physical processes 
taking place in this supernova remnant. 

A study was done on the hard x-ray 
and microwave components in the impul- 
sive phase of flares by Dr. Joan Vorpahl 
of Sacramento City College. In 36 ran- 



domly chosen events the value for the 
slope in the differential electron power 
spectrum, E~ 8 electrons cm -2 sec -1 keV" 1 , 
was related to the 20—32 keV spike rise- 
time as t rise = 0.6 exp (0.878), i.e., as 
the rise-time becomes shorter, the frac- 
tion of higher electron (and photon) en- 
ergy increases. A good correlation was 
also found between the peak flux at 8800 
MHz in an event and the number of 
electrons above 100 keV: F(8800) aN 1 - 7 , 
when thin target theory was used. Only 
13 out of 173 impulsive x-ray events had 
no microwave flux above 2800 MHz; 
N(>100) for these 13 flares was less than 
10 33 electrons. Conversely, N(>100) 
was greater than 10 33 electrons for 18 
random events with flux at 8800 MHz or 
above. We conclude from this that 10 33 
determines the lower threshold for detec- 
tion of impulsive microwave emission 
(assuming thin target models hold). 

Vorpahl and Zirin examined two multi- 
ple impulsive events in May 1969 and 
compared their impulsive H-alpha, hard 
x-ray and microwave components with 
the observed Type III emission. No good 
correlation was observed between meter 
intensity and the degree of simultaneous 
Ha outflow, although almost every Type 
III burst had some associated Ha activ- 
ity. Correspondingly, an inverse rela- 
tionship existed between x-ray and Type 
III emission, i.e., a strong meter burst 
accompanied weak x radiation. This, 
along with the fact that sufficient ener- 
getic electrons were normally present to 
produce the x rays, implies that the tra- 
jectory, rather than electron spectrum or 
supply, determines whether Type III 
noise is observed during an impulsive 

Dr. George Wallerstein of the Univer- 
sity of Washington completed a program 
of searching for interstellar Ti II, using 
the resonance line at A3383 A. From 
spectrograms of 15 stars combined with 
data for 8 stars reported by others, Wal- 
lerstein and D. Goldsmith have deter- 
mined column densities for Ti II for 11 

stars and upper limits for 12 others. Ob- 
servations and theory show that little or 
no Ti I or Ti III is expected in inter- 
stellar gas. The derived Ti/H ratios 
show a deficiency of from 10 to 200 with 
respect to the sun and solar system. By 
investigating the correlations among col- 
umn densities of Ca II, Ti II, H I and 
color excess, they can show that the 
missing titanium (as well as calcium) is 
tied up in the grains. 

An abundance analysis by Wallerstein, 
Snow, Roberts, and Baird of two low- 
velocity bright giants with very strong 
CH, HR 7606 and HR 8626, yields a 
small metal deficiency and a deficiency 
of oxygen but normal carbon and nitro- 
gen. HR 8626 has a substantial lithium 
line while HR 7606 does not. These 
abundances are difficult to understand in 
terms of stellar evolution and nucleo- 
synthesis. Wallerstein's program of ob- 
servation of OH- and H^O-emitting stars 
is near completion. Spectra near mini- 
mum of R Cas, R Aql and RT Cyg were 
obtained in October 1972. Extensive 
radial velocity measurements of most of 
the spectra have been completed and the 
data are being analyzed in terms of mass 
loss from these outermost layers. Ob- 
servations of stars behind the Monoceros 
Supernova remnant have been started. 
One star shows displaced Ca II com- 
ponents at ±70 km sec -1 , thus yielding 
an expansion velocity at one point. Ad- 
ditional stars will be observed in the 
autumn of 1973. 

Las Campanas 

Mr. Gonzalo Alcaino of Santiago, 
Chile, has obtained six blue and six yel- 
low plates of the globular clusters NGC 
362, 1904, 3201, 5927, and 6809. Expo- 
sures range from 40 sec to 40 min. Photo- 
electric sequences of 15 stars between 
visual magnitudes 11 and 16 have been 
set up for the globular clusters NGC 288, 
1904, 3201, 5927, 6121, and 6809. 

Mr. Eduardo Hardy of the University 
of Chile, using the 40-inch telescope, 



started work on a photographic atlas of 
the Large Magellanic Cloud, a joint proj- 
ect with Sandage. He also began a study 
of the "old" stellar content of the LMC. 
A number of deep B and V plates were 
obtained on the Bar and selected areas of 
the Cloud. Photoelectric UBV integrated 
photometry was performed on special 
regions of the Bar. 

Dr. Edward P. Ney of the University 
of Minnesota and Dr. Roberta M. Hum- 
phreys of the University of Arizona used 
the 40-inch telescope to observe super- 
giants identified by Roberta Humphreys. 
Approximately 24 supergiants in the 
Norma-Sagittarius region were measured 
at wavelengths from 1.2 to 18 microns. 
Four very unusual stars with large in- 
frared excesses were identified. Measure- 
ments of Caltech Anonymous Objects in 
the — 10° to — 30° declination range 
were also made. Approximately 250 un- 
identified IRC objects were available 
during nighttime hours. A total of 235 
of these were observed in 92 clear hours' 
observing. The goal was to identify the 
true infrared stars. Ney and Humphreys 
had previously studied 230 objects in the 
northern hemisphere and identified two 
classes of infrared stars, which they 
called Taurids (like NML Tauri) and 
Cygnids (like NML Cygni). About 10% 
of the northern hemisphere objects were 
infrared stars and 20 were Taurids, 8 
Cygnids. In the sample of 235 southern 
hemisphere stars, again 10% (25 objects) 
were infrared stars, but the preliminary 
analysis indicates that they are all 
Taurids. With only two exceptions, the 
infrared stars were not visible in the 40- 
inch telescope and had visual magnitudes 
>16 mag. They were found by scanning 
a 4 square-minute area around the IRC 
coordinates. The two visual objects had 
m v = 13 and m v = 14, and Humphreys 
plans to get spectra of these on the 
Steward Observatory 90-inch telescope. 

Dr. John Norris of Yale University 
Observatory has used the image-tube 
spectrograph of the 40-inch telescope to 

obtain radial velocities of stars which fall 
above the horizontal branch and to the 
left of the giant branch in the HR dia- 
grams of the globular clusters w Centauri, 
M4, and M22. Seven new members have 
been found in w Cen, and five have been 
found in M22. These stars are being in- 
vestigated to determine their atmospheric 
parameters so that a comparison can be 
made with theoretical predictions of 
postasymptotic giant-branch evolution. 

Dr. S. Eric Persson and Dr. Jay A. 
Frogel of Harvard College Observatory 
have continued their infrared studies of 
H II regions in the wavelength range 1.2 
to 20 /x. Thirteen southern thermal radio 
sources having high-emission measure 
were found to be strong 10 and 20 \x 
emitters and were mapped at several 
wavelengths. One of these has been de- 
scribed in detail by Becklin et al. {Astro- 
phys. J. [Lett.], 182, L137, 1973). The 
data show that these sources have the 
following properties: The energy spectra 
from 1.2 to 20 /x are at least as steep as 
F v a v~ 3 ; the observed flux at 20 /* is 
greater by a factor of 100 than that ex- 
pected by extrapolating the radio obser- 
vations, while the flux shortward of 2.2 ^ 
is less than that expected; observations 
with intermediate band-width filters in the 
10 jx region show a spectral feature in 
several of the objects which has been 
associated with silicate absorption; sev- 
eral of the sources are barely resolved 
with a 14. "5 beam at 10 and 20 /x, while 
others (such as RCW 38 and RCW 57) 
are highly resolved and show complex 
structure. At least one of the objects 
(RCW 57) contains an unresolved infra- 
red source with a spectrum similar to 
that of the source IRS 5 in W 3 and the 
point source in the Orion nebula. An 
interpretation which has been advanced 
for these two sources is that they are 
pre-main-sequence objects. The extended 
infrared emission is believed to arise 
from radiating dust particles that are 
heated by the Lyman-a radiation field, 
and perhaps also by direct starlight. In 



addition, Persson and Frogel found the frared study (made in collaboration with 

peculiar optical nebula NGC 6302 to I. J. Danziger) show that NGC 6302 and 

have excess emission at 10 and 20 p. The the peculiar planetary nebula NGC 7027 

results of a combined optical and in- are similar in several respects. 


Digital linage Recorder 

This project has been carried out as a 
joint effort with Princeton University 
Observatory. The instrument was used 
on both the 60-inch and 200-inch tele- 
scopes at Palomar during the spring of 
1973. The Princeton group was respon- 
sible for the high-gain television camera 
and its associated controls; AEL was 
responsible for the digital memory, the 
system controls, the computer interface, 
and the operational software. 

The memory is a 65,536-word, dynam- 
ically shifted Metal Oxide Semiconduc- 
tor array. Each word consists of 16 
binary bits, or 65,536 counts. For its first 
application, the system was used to re- 
cord the data from a 256 X 256 pixel 
(picture element) format at a pixel rate 
of 4.8 MHz. The memory array can be 
set to between 8 and 2,048 pixels per scan 
line and between 8 and 2,048 scan lines 
per picture. The memory size cannot be 
less than 4,096 words or more than 
65,536 words. The memory word rate 
can be adjusted from 1 sec to 9,999 sec. 
Other controls allow the observer com- 
plete flexibility in his data collection 

The memory is scanned in synchro- 
nism with the camera sweep circuits to 
ensure that the memory words and image 
pixels maintain a one-to-one correspond- 
ence with each other. When a photoelec- 
tron pulse is detected in any pixel, a cor- 
responding memory word is incremented. 
Although the memory was designed pri- 
marily for use with photoelectron- 
counting-type detector systems, it can 
also be used with camera tubes that use 
target integration and slow-scan readout. 

The memory data read into the Hale 
Observatories' computer system is for- 
matted and transmitted to an industry- 

compatible magnetic tape recorder. The 
computer is also used for recording time, 
telescope coordinates, and object identi- 
fication data. Other telescope control 
functions are also handled by the com- 

The first observing runs clearly dem- 
onstrated that this type of image recorder 
is successful and that it will be an effec- 
tive new observing tool. The initial data 
tapes are now being analyzed to deter- 
mine the overall quantum efficiency and 
system characteristics that may require 
future improvements. 

Computer Systems 

In addition to the computer systems 
on the Palomar 60-inch and 200-inch 
telescopes, the computer systems for the 
Mount Wilson 60-inch telescope and the 
150-foot solar tower were completed this 
year. These latter two systems are under- 
going software checkout. 

After more than a year of experience 
with the computer systems, it is interest- 
ing and helpful to assess their effective- 
ness. On the positive side, it is very clear 
that the computer systems provide new 
capability to the observers. This is de- 
rived from more flexible control, more 
efficient data collection, and more accu- 
rate data displays. Most observers feel 
that the basic computer concept is good 
and that it offers further potential for 
increases in effectiveness. On the nega- 
tive side, the observers have raised two 
serious objections to the present system. 
These objections are directly related to a 
failure to adhere strictly to two of the 
most important philosophical guidelines 
for astronomical instrument applications. 
These guidelines were not followed be- 
cause limited resources led to an attempt 
to compromise these principles. It is now 



clear that the compromises have caused 
serious problems and that these guide- 
lines are of fundamental importance to 
astronomical instrument design and ap- 

The first principle is that any instru- 
ment must be extremely reliable and free 
from even minor defects. The number of 
components that have failed to give per- 
fect performance has been very small 
and the actual observing time lost due to 
these failures has also been small. But 
these occasional failures have led to frus- 
tration and irritation for the observers. 
This small and infrequent failure rate 
has also increased the problems of 
troubleshooting and of making corrective 
changes in the system design. The AEL 
priorities have been changed to ensure 
early elimination of these problems. 

The second principle that was com- 
promised is that the observer should at 
all times feel that he is fully in control 
of his instruments. This concept was 
completely followed in the design of the 
observer controls, but it was neglected in 
the design of the software system. The 
operational programs were written in 
assembly language to minimize the cost 
of program generation, the program ex- 
ecution time during critical phases of the 
operation, and the required size of the 
computer memory. The software system 
is now being redesigned to permit the 
data collection programs to be written in a 

modified FORTRAN language. This new 
software system will use almost all of the 
existing assembly language subroutines 
as well as the computer manufacturer's 
FORTRAN programs. As a result, this 
effort will require only minimal additional 
cost. It is anticipated that the user pro- 
gram system will be available in the early 
fall of 1973. The computer memory will 
have to be enlarged from 16,000 to 32,000 
words and a card reader will be added to 
facilitate program editing. This new ap- 
proach will restore to the observers full 
control over the operation of the com- 
puter systems. 

Electronic Instrument Utilization 

Electronic instrumentation has con- 
tinued to play an important role for the 
Hale Observatories' telescopes. The fol- 
lowing table indicates the number of 
nights for which AEL had at least some 
degree of responsibility for the observing 
instruments. The actual utilization of 
electronic equipment was somewhat 
higher because a few of the observers 
used their own electronic equipment with- 
out the asisstance of AEL. 

No. of 













Mount Wilson 




Mount Wilson 





Miller continued tests of methods of 
reducing low-intensity reciprocity failure 
by outgassing oxygen and water vapor 
from photographic emulsions. Experi- 
ments using vacuum, reported in Year 
Book 71 (p. 707), were followed by tests 
using dry nitrogen without recourse to 
vacuum. The results closely paralleled 
those found for vacuum, with several 
advantages in favor of the inert gas. The 
nitrogen system was adopted for general 
use on Mount Wilson and Palomar. 

Existing baking equipment is being con- 
verted to the use of nitrogen as rapidly 
as possible. 

For 2-hour exposures the increase in 
efficiency of Ha and 103a emulsions 
baked optimally in nitrogen amounts to 
around 2 times, and for the new Ilia 
emulsions the gain factor ranges from 3 
times to nearly 10 times for different 
batches. Optimal baking times at 65° C 
in nitrogen range from 4 hours for the 
Ilia emulsion to 15 hours for the Ha and 



24 hours for the 103a emulsions. Current 
efforts are concentrated on establishing 
optimum baking times for all types of 
emulsions to facilitate application of the 
system by observers. 

A major plate-salvage program was 
instituted to reverse image deterioration 
in old plates resulting from faulty proc- 
essing in past years before the chemistry 
involved was fully understood. Many 
plates in the files were developing brown 
stain caused by silver sulfide, a product 
of slow decomposition of a silver thio- 
sulfate complex left in the processed 
emulsion by inadequate fixation. The 
treatment applied is a chemical bath 
which not only returns the photographic 
image to neutral silver, but forestalls any 
further deterioration. Improved under- 
standing of the chemistry of fixation has 
resulted in new darkroom techniques that 
will prevent deterioration of plates being 
taken now and in the future. 

The Observatories' long-standing pol- 
icy of close cooperation with the East- 
man Kodak Company in improving 
existing types of astronomical photo- 
graphic plates and devising new types 
was continued. Miller made many labo- 
ratory tests of samples submitted by the 
Kodak Research Laboratories. Tele- 
scope tests of the most promising samples 
were made to confirm laboratory tests 
and to give quantitative results regard- 
ing the astronomical value of the experi- 
mental emulsions. The objective of cur- 
rent experiments is the production of a 

red-sensitive counterpart of the high 
Detective Quantum Efficiency Illa-J 
emulsion that has proved so popular. It 
is hoped that such a red-sensitive emul- 
sion will help significantly in the detec- 
tion of very distant objects whose large 
redshifts leave very little detectable en- 
ergy in the blue region of the spectrum. 

Making use of a greatly improved 
image-registration device, Bedke has 
produced impressive "integration" prints 
that combine the information contained 
in several separate negatives of astro- 
nomical objects in a single print with 
greatly improved signal-to-noise charac- 
teristics. These prints increase visibility 
of very faint details by increasing the 
limiting magnitude beyond that avail- 
able from any one of the negatives alone. 

Miller has continued routine quality- 
control tests on all shipments of 
photographic plates received to ensure 
maximum efficiency of photographic ob- 
servations at the telescopes. 

The Photo Lab had several visitors from 
other observatories during the year, who 
came to receive training in astronomical 
photographic techniques. These were: 
Mr. William Schoening (Kitt Peak Na- 
tional Observatory), Miss Elizabeth Sim 
(Royal Observatory, Edinburgh), Mr. 
Hans Schuster (European Southern Ob- 
servatory, La Silla, Chile), Mr. Hans 
Vogel (European Southern Observatory, 
Santiago, Chile) , and Mr. Robert Wilson 
(Lick Observatory). 


Under the general supervision of Rule 
as Project Officer, and with the collabora- 
tion of many members of the scientific, 
engineering, and administrative staffs, 
substantial progress was made in meeting 
scheduled design requirements, award of 
contracts, guidance of manufacturing op- 
erations, procurement and delivery of 
site construction components for the du 
Pont Telescope and dome building. Some 

15 major fabrication contracts for tele- 
scope and dome items were committed by 
April 1973. These, together with most 
other components, are in advanced stages 
of inspection and completion in prepara- 
tion for shipment to Chile, with assembly 
planned at the Las Campanas site in 
1974, after the dome and the building 
are ready. 



Mechanical Parts and Building 

All telescope heavy parts are complete, 
with subassemblies of drives, bearings, 
and auxiliary mechanisms undergoing 
acceptance tests now for plant trials of 
the complete telescope assembly during 
August and September, after which the 
telescope will be dismantled and packed 
for shipment. 

Dome construction work at the site 
was started in December 1972. Excava- 
tion is complete, with concrete poured for 
pier, foundations, basement, and sub- 
walls. Some contractor delays were ex- 
perienced, but work is progressing to 
complete the building up to the circular 
dome rail to be ready for structural steel 
dome erection by about November 1973. 
Fabrication of the rotating dome was 
started in February, with most heavy 
parts complete in June. Fit-up and as- 
sembly of the completed dome structure 
will be made at the contractor's plant by 
about September. 

Other building equipment, hardware, 
mechanical supplies, and interior em- 
bedment steel have been obtained and 
shipped as needed. Major equipment 
items such as dome crane, hoists, passen- 
ger elevator, central ram platform, and 
all equipment for heating and air con- 
ditioning systems are all in process, 
nearly complete, or ready for shipment. 

The aluminizing system for mirror 
coating to accommodate primary and all 
secondary mirrors was engineered in de- 
tail. Fabrication was started in April 
by High Vacuum Equipment Corpora- 
tion for coating tests to be made by 
December 1973. 


Following tests at the Optical Sciences 
Center of the primary mirror support cell 
and grinding table supports with the 
mirror at a spherical figure, the operation 
of polishing to achieve the final aspheric 
figure was started in early summer 1972 
and continued throughout the year. This 

work is under the supervision of Mr. Don 
Loomis, Chief Optician for the telescope 
optics, with close cooperation in testing 
by Vaughan and Fair. Design and fabri- 
cation of optical test equipment and de- 
velopment of computational procedures 
were pursued at O.S.C. and by Vaughan 
in Pasadena. By April 1973 all test pro- 
cedures required for evaluation of the 
final optical figure of the primary were 
in operation. The mirror figure was 
brought to within % micron of the final 
surface by June 1973, with the work ex- 
pected to continue for several weeks. 
The coude elliptical flat and Cassegrain 
f/7.5 and f/13.5 secondary blanks were 
delivered in late fall to Loomis Custom 
Optics after preliminary polishing in 
Pasadena. By June 1973 the f/13.5 sec- 
ondary had been completed and awaited 
final acceptance tests. Preliminary pol- 
ishing of the Gascoigne-Bowen corrector 
plate and fabrication of a spherical test 
mirror for the corrector were in progress 
in Pasadena. 

Auxiliary Instruments 

Following a number of general meet- 
ings to discuss auxiliary instrumentation, 
several working groups were set up by 
Oke to supervise the design and construc- 
tion of specific instruments and facilities. 
It is the hope that most of the initial set 
of instruments will be ready for use when 
the telescope is completed. The groups 
(with the chairman) are listed below: 
Direct Plate Holder and Darkroom Proc- 
essing Equipment (Sandage) ; Instrument 
Mounting Base (Oke) ; TV Guiding Sys- 
tem (Dennison) ; Image-Tube Spectro- 
graph (Searle) ; Broad-Band Photometer 
(Kristian) ; Infrared Photometer (Neuge- 
bauer); 90-mm Image Tube (Brucato) ; 
Coude Spectrograph (Vaughan) ; Casse- 
grain Chair (Rule) ; Computer Inter- 
faces (Neugebauer). Design work has 
started on direct plate holders, instru- 
ment mounting bases, and the Cassegrain 
observer's chair. 



Control System 

During the report period, substantial 
progress has been made by the Astro- 
electronics Laboratory under Dennison 
toward the design and completion of the 
control system for the du Pont Telescope. 
Much of the effort has gone into ensuring 
that the design is satisfactory in two fun- 
damental interface areas: between the 
control system and the mechanical parts 
of the telescope and between the control 
system and the operator. This initial 
study phase was essential to minimize 
the possibility of human error in operat- 
ing the telescope and to eliminate the 
necessity for on-site modification of the 
system when it is installed with the 
telescope in Chile. 

To increase its reliability and to facil- 
itate telescope checking, the system has 
been designed and built to enable all 
telescope functions to be operated, on oc- 
casion, from portable control boxes that 
are totally independent of the basic tele- 
scope control system. This facilitates 
checking of the telescope drive at the 
fabrication plant and also guarantees 
that all telescope functions can be in- 
dividually operated as soon as the tele- 
scope is erected in Chile. This concept 
also gives an independent means of oper- 
ating the telescope in case of control sys- 
tem failure. The telescope-control sys- 
tem is being designed with solid state 
logic and operational interlocks and with 

mechanical relay safety and limit inter- 
locks, which give maximum operational 
lifetime and protection for the telescope 
and provide an efficient computer inter- 

Computer and Data System 

A substantial beginning has been made 
on the computer and data system. Many 
of the purchased system components have 
been delivered and are operational. The 
individual interface units that are to be 
built by AEL will generally be identical 
with the existing interface units now in 
operation at several telescopes of the 
Hale Observatories. By using these 
existing designs, substantial savings in 
time and cost should be achieved in both 
hardware and software. The software 
savings will be particularly outstanding 
because it should be possible to use most 
of the 250 subroutines that now exist in 
our currently operating system. 

The only remaining major problem is 
to provide the observers with an easily 
modifiable high-level programming sys- 
tem. The user's software system must 
provide for almost complete independ- 
ence between the control and data sys- 
tems. There are many practical solutions 
to this problem, and it is expected that a 
decision, based on both cost and effec- 
tiveness considerations, will be arrived at 


The average temperature for the year 
was 18° C with a maximum of 30° C on 
January 6, 1973, and a minimum of 
— 11° C on August 26, 1972. Total pre- 
cipitation for the year was 76 millimeters, 
of which 36 millimeters occurred in the 
form of rain. Total snowfall was meas- 
ured at 46 centimeters. Approximately 
280 nights were suitable for observing. 

Following the stormy months of July 
and August 1972, the 40-inch telescope 
was in regular use for all dark runs and 

some of the light runs during the year. 
Staff members or staff associates of the 
Hale Observatories who observed with 
the 40-inch during the year ending June 
30, 1973, were Becklin, Brucato, Munch, 
Neugebauer, Searle, Vaughan, and Bab- 
cock. Guest investigators included Mr. 
G. Alcaino of the Catholic University, 
Santiago ; Dr. P. Cruvellier of the Obser- 
vatoire de Marseille; Dr. Jay A. Frogel 
of Harvard College Observatory; Mr. 
Eduardo Hardy of the University of 



Chile; Dr. Roberta M. Humphreys of the 
University of Arizona; Dr. Edward P. 
Ney of the University of Minnesota ; Dr. 
John Norris of Yale University Observa- 
tory; Dr. S. Eric Persson of Harvard 
College Observatory; and Dr. John 
Warner of the Department of Terrestrial 
Magnetism, Carnegie Institution. 

The 24-inch telescope of the Univer- 
sity of Toronto was in regular operation 
throughout the year. 

The working group of approximately 
32 Chilean employees and laborers as- 
sisted in the operation of the telescopes, 
provided logistical services, maintained 
equipment, assisted the contractor in 
various phases of construction, and con- 
tinued the work of site development, all 
under the general supervision of Adkison 
in Pasadena and Wagner in Chile. 

Site Development 

For the water system, an additional 7 
kilometers of pipeline was laid and the 

contractor completed the major reservoir 
of 200,000-gallon capacity on the slope 
of Las Campanas. The structure of the 
new powerhouse was largely completed 
and the new 300 kilowatt diesel genera- 
tor was placed in position. Satisfactory 
electrical service is being provided tem- 
porarily by two gasoline generators. A 
materials storage yard was provided and 
a much needed warehouse was well on 
the way to completion. Construction of 
the Lodge was begun. Three temporary 
dormitory rooms were provided on the 
mountain. In La Serena, three new 
dormitory rooms were built adjacent to 
the El Pino office. 

Regular shortwave communications 
were maintained between Pasadena and 
the El Pino office and between El Pino 
and Las Campanas. Frequent negotia- 
tions between the Pasadena office and the 
shipping lines and air freight carriers 
have achieved an acceptable if not com- 
pletely predictable level of transport 


Adelman, Saul J., A model atmosphere abun- 
dance analysis of the B9V star v Capricorni. 
Astrophys. J., 182, 531-538, 1973. 

Adelman, Saul J., and Wallace L. W. Sargent, 
A peculiar B star in the old disk population. 
Astrophys. J., 176, 671-676, 1972. 

Agnew, Duncan, and Halton Arp ? A list of 
quasi-stellar sources and quasi-stellar source 
candidates from 3C and 4C catalogs between 
Dec —7° and +40°. Publ. Astron. Soc. 
Pacific, 85, 162-173, 1973. 

Allen, David A., Near infrared magnitudes of 
248 early-type emission-line stars and related 
objects. Mon. Notic. Roy. Astron. Soc, 161, 
145-166, 1973. 

Allen, David A., The method of determining 
infrared diameters, in Physical Studies of the 
Minor Planets, NASA SP-276, pp. 41-43, 1972. 

Allen, David A., and F. C. Porter, Infrared 
photometry of southern Wolf-Rayet stars. 
Astron.. Astrophys., 22, 159-160, 1973. 

Allen, David A., P. M. Harvey, and J. P. 
Swings, Infrared photometry of northern 
Wolf-Rayet stars. Astron. Astrophys., 20, 
333-336, 1972. 

Allen, David A., see also Swings, J. P. 

Altschuler, Martin D., see Newkirk, Gordon, Jr. 

Arp, Halton, and Frank Brueckel, Diameters of 
H II regions in M31 and comparison to the 
largest regions in M33. Astrophys. J., 179, 
445-451, 1973. 

Arp, Halton, and John Kormendy, Faint sur- 
face-brightness features between NGC 7331 
and Stephan's quintet. Astrophys. J. (Lett.), 
178, L101-L103, 1972. 

Arp, Halton, see also Agnew, Duncan. 

Babcock, Horace W., Stellar magnetic fields: 
5i-5. Spectral observations, in American In- 
stitute of Physics Handbook, 3d ed., pp. 
5-306-5-310, Dwight E. Gray coord, ed., 
McGraw-Hill Book Company, New York, 

Babcock, Horace W., Ira Sprague Bowen, 1898- 
1973. Eng. Sci., 36, 26-27, 1973. 

Bagnuolo, W. G., see Searle, Leonard. 

Bahcall, J., see Spinrad, Hyron. 

Becklin, E. E., E. B. Fomalont, and G. Neuge- 
bauer, Infrared and radio observations of the 
nucleus of NGC 253. Astrophys. J. (Lett.), 
181, L27-L31, 1973. 

Becklin, E. E., G. Neugebauer, and C. G. 
Wynn-Williams, On the nature of the infra- 
red point source in the Orion Nebula. Astro- 
phys. J. (Lett.), 182, L7-L9, 1973. 



Becklin, E. E., G. Neugebauer, and C. G. 
Wynn-Williams, Infrared emission from the 
OH/H2O sources in W49. Astrophys. Lett., 
13, 147-149, 1973. 

Becklin, E. E., J. Kristian, G. Neugebauer, and 
C. G. Wynn-Williams, Discovery of infrared 
emission from the radio source near Cygnus 
X-3. Nature, 239, 130-131, 1972. 

Becklin, E. E., J. A. Frogel, G. Neugebauer, 
S. E. Persson. and C. G. Wynn-Williams, The 
H II region G333 .6-0.2, a very powerful 1-20 
(x source. Astrophys. J. (Lett.), 182, L125- 
L129, 1973. 

Becklin, E. E., see also Epstein, E. E.; Kirsh- 
ner, R. P.; Neugebauer, G.; Spinrad, Hvron; 
Wilson, W. J.; Woolf, N. J.; Wynn-Williams, 

C. G. 

Bhatnagar, Arvind. W. C. Livingston, and J. W. 
Harvey, Observations of sunspot umbral 
velocity oscillations. Solar Phys., 27, 80-88. 

Bowen, I. S., and A. H. Vaughan, Jr., "Non- 
objective" gratings. Publ. Astron. Soc. Pacific, 
85, 174-176, 1973. 

Braccesi, A., L. Formiggini, I. Gioia, and 
W. L. W. Sargent, Two galaxies dominated 
by bright ultraviolet knots. Publ. Astron. Soc. 
Pacific, 84, 592-593, 1972. 

Brucato, Robert, and Jerome Kristian, Spec- 
troscopic observations of the optical candi- 
date for Cygnus X-l. Astrophys. J. (Lett.), 
179, L129-L133, 1973. 

Brucato, Robert J., and Jerome Kristian, 
BD+34°3815, Int. Astron. Union Circular 
No. 2421, July 6, 1972. 

Brucato, Robert J., see also Westphal, J. A. 

Brueckel, Frank, see Arp, Halton. 

Chaldu, R., R. K. Honeycutt, and M. V. Pen- 
ston. The extinction curve for Cygnus OB2 
No. 12. Publ. Astron. Soc. Pacific, 85, 87-90, 

Conti, Peter S., see Wallerstein, George. 

Cragg, Thomas A., The classical Cepheid pro- 
gram, J. D. 2,440,000-2,441,000. /. Amer. 
Assoc. Var. Star Observers, 1, 9-15, 1972. 

Crutcher, Richard M., see Quirk, William J. 

Dennison, Edwin W., An integrating television 
system for visual enhancement of faint stars, 
in Advances in Electronics and Electron 
Physics. Vol. 33B, pp. 795-800, J. D. McGee, 

D. McMullen, and E. Kahan, eds., Academic 
Press, London. 1973. 

Dravins, Dainis, A possible solar electrograph. 

Astrophys. Lett., 13, 243-245, 1973. 
Dworetsky, Michael M., The orbit of 53 Tauri 

(B9p-Mn). Publ. Astron. Soc. Pacific, 84, 

652-655, 1972. 
Dworetsky, Michael M., and Arthur H. 

Vaughan, Jr., Wavelength shifts and probable 

istotopic structure of Pt II X4046 in mercury 
stars. Astrophys. J., 182, 811-816, 1973. 

Edberg, Stephen J., see Howard, Robert. 

Epstein, E. E., E. E. Becklin, G. Neugebauer, 
and C. G. Wynn-Williams, and 20 coauthors, 
3C120, BL Lacertae, and OJ 287 : coordinated 
optical, infrared, and radio observations of 
intraday variability. Astrophys. J. (Lett.), 
178, L51-L59, 1972. 

Fomalont, E. B., see Becklin, E. E. 

Formiggini, L., see Braccesi, A. 

Foukal, Peter, Measurement of the electron 
temperature of small 3-cm radio bursts. Solar 
Phys., 24, 411-413, 1972. 

Foukal, Peter, and Harold Zirin, Reply to "The 
relations between chromospheric features and 
photospheric magnetic fields" by E. N. 
Frazier. Solar Phys., 26, 148-150, 1972. 

Frogel, J. A., see Becklin, E. E. 

Gillett, F. C, see Woolf, N. J. 

Gioia, I., see Braccesi, A. 

Glackin, David L., Spatial distribution of 
emerging flux regions. Publ. Astron. Soc. 
Pacific, 85, 241-248, 1973. 

Goad, L. E., L. Goldberg, and Jesse L. Green- 
stein, High-n. Balmer transitions in gaseous 
nebulae. Astrophys. J., 175, 117-125, 1972. 

Goldberg, L., see Goad, L. E. 

Gott. J. Richard, III, see Gunn, James E. 

Grandi, Steven, The expansion of NGC 7293. 
Publ. Astron. Soc. Pacific, 85, 200-202, 1973. 

Greenstein, Jesse L., A highly evolved, low- 
mass binary, HZ 22. Astron. Astrophys., 28, 
1-7, 1973. 

Greenstein, Jesse L., The future of astronomy. 
Eng. Sci., 36, 18-23, 1972. 

Greenstein, Jesse L., Preface, in Astronomy 
and Astrophysics for the 1970s, Vol. 2, Re- 
ports of the Panels, pp. vii-x, National Acad- 
emy of Sciences, Washington, D.C.. 1973. 

Greenstein, Jesse L., and R. Minkowski, An 
atlas of supernovae spectra. Astrophys. «/., 
182, 225-243, 1973. 

Greenstein, Jesse L., and Deane M. Peterson. 
Line profiles and rotation in white dwarfs. 
Astron. Astrophys., 25, 29-34, 1973. 

Greenstein, Jesse L., and Virginia Trimble, The 
gravitational redshift of 40 Eridani B. Astro- 
phys. J., 175, L1-L5, 1972. 

Greenstein, Jesse L., see also Goad, L. E.; 
Trimble, Virginia; Wallerstein, G. 

Gunn, James E., and J. Richard Gott, III, On 
the infall of matter into clusters of galaxies 
and some effects on their evolution. Astro- 
phys. J., 176, 1-19, 1972. 

Gunn, James E., see also Oemler. Augustus, Jr. ; 
Spinrad, Hyron. 



Harvey, J. W., see Bhatnagar, Arvind. 
Harvev, P. M., see Allen, David A.; Wilson, 

Honeycutt, R. K., see Chaldu, R. 

Howard, Robert, General magnetic field of the 
sun, in American Institute of Physics Hand- 
book, 3d ed., p. 5-304, Dwight E. Gray, coord, 
ed., McGraw-Hill Book Company, New York, 

Howard, Robert, Sunspot fields, in American 
Institute of Physics Handbook, 3d ed., p. 
5-305, Dwight E. Gray, coord, ed., McGraw- 
Hill Book Company, New York, 1972. 

Howard, Robert, Recent solar research. Science, 
177, 1157-1163, 1972. 

Howard, Robert, Polar magnetic fields of the 
sun: 1960-1971, Solar Phys., 25, 5-13, 1972. 

Howard, Robert, Photospheric magnetic fields, 
in Solar Wind, NASA SP-308, pp. 3-9, C. P. 
Sonett, J. P. Coleman, Jr., and J. M. Wilcox, 
eds., National Aeronautics and Space Ad- 
ministration, Washington, D.C., 1972. 

Howard, Robert, and Stephen J. Edberg, On 
the random nature of the eruption of magnetic 
flux at the solar surface. Solar Phys., 28, 73- 
75, 1973. 

Howard, Robert, see also Newkirk, Gordon, Jr.; 
Schatten, Kenneth H. 

Huchra, John, and S .P. Willner, UBV photom- 
etry of selected Ap stars. Publ. Astron. Soc. 
Pacific, 85, 85-86, 1973. 

Huchra, John, see also Westphal, J. A. 

Hunten, Donald M., and Guido Munch, Helium 
abundance in Jupiter. Space Sci. Rev., 14, 
433-443, 1973. 

Kirshner, R. P., S. P. Willner, E. E. Becklin, G. 
Neugebauer, and J. B. Oke, Spectrophotom- 
etry of the supernova in NGC 5253 from 0.33 
fi to 2.2 /i. Astrophys. J. (Lett.), 180, L97- 
L100, 1973. 

Kormendy, John, Calibration of direct photo- 
graphs using brightness profiles of field stars. 
Astron. J., 78, 255-262, 1973. 

Kormendy, John, see also Arp, Halton. 

Kowal, C. T., Comet Gehrels (1972e), Intern. 
Astron. Union Circular No. 2469, December 
12, 1972. 

Kowal, C. T., W. L. W. Sargent, L. Searle, and 
F. Zwicky, The 1971 Palomar supernova 
search. Publ. Astron. Soc. Pacific, 84, 844-849, 

Kristian, Jerome, Quasars as events in the nu- 
clei of galaxies: the evidence from direct 
photographs. Astrophys. J. (Lett.), 179, L95- 
L99, 1973. 

Kristian, Jerome, A summary of optical pulsar 
observations, in Proceedings of Symposium 
on Pulsars and High Energy Activity in 

Supernova Remnants, pp. 113-120, L. Grat- 
ton, ed., Accademia Nazionale dei Lincei, 

Kristian, Jerome, see also Becklin, E. E.; Bru- 
cato, Robert; Spinrad, Hyron; Westphal, 
J. A. 

Kruit, P. C. van der, Spiral structure and nu- 
clear activity in galaxies. Nature, 243, 127- 
130, 1973. 

Kuiper, T. B. H., and Jay M. Pasachoff, De- 
tailed correlation of type III radio bursts 
with Ha activity, I, Active region of 22 May 
1970. Solar Phys., 28, 187-196, 1973. 

Lanning, Howard, A finding list of faint UV- 
bright stars in the galactic plane. Publ. 
Astron. Soc. Pacific, 85, 70-84, 1973. 

Leighton, Robert B., see Schatten, Kenneth H. 

Livingston, W. C, see Bhatnagar, Arvind. 

McKee, Christopher F., and Wallace L. W. 
Sargent, The absorption-line spectrum of the 
bright QSO Markarian 132. Astrophys. J. 
(Lett.), 182, L99-L101, 1973. 

Merrill, K. M., see Woolf, N. J. 

Minkowski, R., see Greenstein, Jesse L. 

Munch, Guido, see Hunten, Donald M. 

Neugebauer, G., and Eric E. Becklin, The 
brightest infrared sources. Sci. Amer., 228, 
28-40, 1973. 

Neugebauer, G., see also Becklin, E. E.; Ep- 
stein, E. E.; Kirshner, R. P.; Spinrad, Hy- 
ron; Wilson, W. J.; Woolf, N. J.; Wynn- 
Williams, C. G. 

Newkirk, Gordon, Jr., Dorothy E. Trotter, 
Martin D. Altschuler, and Robert Howard, 
Atlas of Magnetic Fields in the Solar Corona, 
35 mm film, High Altitude Observatory, 
Boulder, Colorado, 1972; Research note, 
Solar Phys., 24, 370-372, 1972. 

Oemler, Augustus, Jr., The cluster of galaxies 
Abell 2670. Astrophys. J., 180, 11-23, 1973. 

Oemler, Augustus, Jr., James E. Gunn, and J. 
Beverley Oke, 3C232.1 : A QSO in a rich clus- 
ter of galaxies. Astrophys. J. (Lett.), 176, 
L47-L50, 1972. 

Oke, J. B., and E. M. Shoemaker, Object 
Kowal, Int. Astron. Union Circular No. 2509, 
March 19, 1973. 

Oke, J. B., see also Kirshner, R. P;. Oemler, 
Augustus, Jr.; Shields, G. A. 

Pasachoff, Jay M., see Kuiper, T. B. H. 

Penston, M. J., see Penston, M. V. 

Penston, M. V., Multicolor observations of stars 
in the vicinity of the Orion Nebula, in Stellar 
Ages, International Astronomical Union Col- 
loquium No. 17, pp. IX-l-IX-6, G. Cayrel de 
Strobel and A. M. Delplace, eds., Paris- 
Meudon Observatory, France, 1973. 



Periston, M. V., and M. J. Penston, New obser- 
vations of two compact galaxies. Mon. Notic. 
Roy. Astron. Soc, 162, 109-116, 1973. 

Penston, M. V., see also Chaldu, R. 

Pepin, T. J., see Woolf, N. J. 

Persson, S. E., see Becklin, E. E. 

Peterson, Deane M., and H. L. Shipman, Helium 
abundances in NGC 2264, II Scorpii, and I 
Lacertae. Astrophys. J., 180, 635-645, 1973. 

Peterson, Deane M., see also Greenstein, Jesse L. 

Pope, Thomas, see Vorpahl, Joan. 

Porter, F. C, see Allen, David A. 

Prata, Stephen W., Some aspects of flare prop- 
erties versus magnetic boundary morphology. 
Solar Phys., 25, 136-140, 1972. 

Preston, George W., Stellar astronomy objec- 
tives, in Astronomy from a Space Platform, 
Vol. 28, Science and Technology, 1972, pp. 
95-104, G. W. Morgenthaler, ed., American 
Astronautical Society, Tarzana, California, 

Preston, George, The unique magnetic and 
spectrum variations of HD 24712. Astrophys. 
J., 175, 465^72. 1972. 

Quirk, William J.. On the gas content of gal- 
axies. Astrophys. J. (Lett.), 176, L9-L14, 

Quirk, William J., Do interstellar gas clouds 
exist between spiral arms? Astrophys. J., 
180, 25-30, 1973. 

Quirk, William J., and Richard M. Crutcher, 
Evidence for a shock associated with the local 
spiral arm. Astrophys. J., 181, 359-362, 1973. 

Richstone, Douglas, and Wallace L. W. Sargent, 
A new determination of the mass of M32 
Astrophys. J., 176, 91-98, 1972. 

Sandage, Allan, Absolute magnitude of E and 
SO galaxies in the Virgo and Coma clusters 
as a function of U — B color. Astrophys. J., 
176, 21-30, 1972. 

Sandage, Allan, The redshift-distance relation, 

II, The Hubble diagram and its scatter for 
first-ranked cluster galaxies: a formal value 
for q . Astrophys. J., 178, 1-24, 1972. 

Sandage, Allan, The redshift-distance relation, 

III, Photometry and the Hubble diagram for 
radio sources and the possible turn-on time 
for QSOs. Astrophys. J., 178, 25-44, 1972. 

Sandage, Allan, The redshift-distance relation, 

IV, The composite nature of N galaxies, their 
Hubble diagram, and the validity of the 
measured redshifts as distance indicators. 
Astrophys. J., 180, 687-697, 1973. 

Sandage, Allan, Opening the last frontier. Eng. 
Sci., 36, 4-10, 1973. 

Sandage, Allan, and Basil Katem, Positions of 
comet Sandage 1972h. Int Astron. Union Cir- 
cular No. 2416, June 21, 1972. 

Sandage, Allan, see also Visvanathan, N. 
Sargent, Wallace L. W., The cluster of compact 

galaxies ZW CI 0152+33. Astrophys. J., 176, 

581-587, 1972. 

Sargent, Wallace L. W., Redshifts for 51 gal- 
axies identified with radio sources in the 4C 
catalogue. Astrophys. J. (Lett.), 182, L13- 
L15, 1973. 

Sargent, Wallace L. W., The redshifts of ten 
clusters of galaxies. Publ. Astron. Soc. Pacific, 
84, 281-282, 1973. 

Sargent, Wallace L. W., Lectures on compact 
galaxies, in Scuola di Fisica Cosmica, Sum- 
mer Course in High-Energy Physics, pp. 1-37, 
at the "Ettore Majorana" Center for Scien- 
tific Culture, Erice, Italy, June 1971. 

Sargent, Wallace L. W., The properties of extra- 
galactic X-ray sources from visible light ob- 
servations, in X and Gamma Ray Astronomy, 
Int. Astron. Union Symp. No. 55, pp. 184-198, 
H. Bradt and R. Giacconi, eds., D. Reidel 
Publishing Co., Dordrecht-Holland, 1973. 

Sargent, Wallace L. W ., and Leonard Searle, 
Supernova in NGC 2841, Int. Astron. Union 
Circular No. 2498, February 27, 1973. 

•Sargent, Wallace L. W., see also Adelman, Saul 
J.; Braccesi, A.; Kowal, C; McKee, Christo- 
pher F.; Richstone, Douglas; Searle, Leon- 
ard; Shields, G. A.; Spinrad, Hyron. 

Schatten, Kenneth H., Robert B. Leighton, 
Robert Howard, and John M. Wilcox, Large- 
scale photospheric magnetic field: the diffu- 
sion of active-region fields. Solar Phys., 26, 
283-289, 1972. 

Schmidt, Maarten, Statistical studies of the 
evolution of extragalactic radio sources, I, 
Quasars. Astrophys. J., 176, 273-287, 1972. 

Schmidt, Maarten, Statistical studies of the 
evolution of extragalactic radio sources, II, 
Radio galaxies. Astrophys. J., 176, 289-301, 

Schmidt, Maarten, Statistical studies of the 
evolution of extragalactic radio sources, III, 
Interpretation of source counts and discus- 
sions. Astrophys. J., 176, 303-314, 1972. 

Schmidt, Maarten, Comet Huchra (1973h). 
Int. Astron. Union Circular Nos. 2423, July 
21, 1972; 2533, May 7, 1973; 2534, May 10, 

Schmidt, Maarten, Supernova in an anonymous 
galaxy. Int. Astron. Union Circular Nos. 
2U7," September 28, 1972; 2485, February 1. 
1973; 2487, February 7, 1973. 

Schmidt, Maarten, Supernova in NGC 735. 
Int. Astron. Union Circular No. 2448. October 
3, 1972. 

Schmidt, Maarten, Supernova in NGC 3656. 
Int. Astron. Union Circular No. 2491, Feb- 
ruary 12, 1973. 



Schmidt, Maarten, 1973 EA. Int. Astron. Union 
Circular No. 2532, May 1, 1973. 

Schmidt, Maarten, Radio source counts and 
redshifts in steady state cosmology. Nature, 
240, 399-400, 1972. 

Schmidt, Maarten, Reply to Hoyle. Nature, 
242, 109, 1973. 

Schwartz, P. R., see Wilson, W. J. 

Searle, Leonard, Star formation and the chem- 
ical history of galaxies, in Stellar Ages, Inter- 
national Astronomical Union Colloquium No. 
17, pp. LII-l-LII-15, G. Cayrel de Strobel 
and A. M. Delplace, eds., Paris-Meudon Ob- 
servatory, France, 1973. 

Searle, Leonard, Wallace L. W. Sargent, and 
W. G. Bagnuolo, The history of star forma- 
tion and the colors of late-type galaxies. 
Astrophys. J., 179, 445-451, 1973. 

Searle, Leonard, see also Kowal, C. T. ; Sargent, 

Wallace L. W. 
Sheet-man, Stephen A., Clusters of galaxies and 

the cosmic light. Astrophys. J., 179, 681-698, 

Shectman, Stephen A., see also Westphal, J. A. 

Shields, G. A., J. B. Oke, and W. L. W. Sargent, 
The optical spectrum of the Seyfert galaxy 
3C120. Astrophys. J. t 176, 75-89,* 1972. 

Shipman, H. L., Masses and radii of white 
dwarfs. Astrophys. J., 177, 723-743, 1972. 

Shipman, H. L., see also Peterson, Deane M. 
Shoemaker, E. M., see Oke, J. B. 
Smith, H., see Spinrad, Hyron. 

Spinrad, Hyron, J. Bahcall, E. E. Becklin, 
James E. Gunn, J. Kristian, G. Neugebauer, 
W. L. W. Sargent, and H. Smith, Optical and 
near infrared observations of the nearby 
spiral galaxy Maffei 2. Astrophys. J., 180, 
351-358, 1973. 

Stein, W. A., see Woolf, N. J.; Zirin, Harold. 

Swings, J. P., and D. A. Allen, Photometry of 
symbiotic and W Cephei stars in the near 
infrared (with a note on MWC 56). Publ. 
Astron.. Soc. Pacific, 84, 523-527, 1972. 

Swings, J. P., see also Allen, David A. 

Trimble, Virginia, and Jesse L. Greenstein, The 
Einstein redshift in white dwarfs, III. Astro- 
phys. J., 177, 441-452, 1972. 

Trimble, Virginia, see also Greenstein, Jesse L. 

Trotter, Dorothy E., see Newkirk, Gordon, Jr. 

Vaughan, Arthur H.. Jr., Ira Sprague Bowen, 
Phys. Today, 26, 77-78, 1973. 

Vaughan, Arthur H., Jr., see also Bowen, I. S.; 
Dworetsky, Michael M. 

Visvanathan, N., and Allan Sandage, Linear 
polarization of the Ha emission line in the 
halo of M82 and the radiation mechanism of 
the filaments. Astrophys. J., 176, 57-74, 1972. 

Vorpahl, Joan, and Thomas Pope, Solar bright 
points in 3840 A and Ha. Solar Phys., 25, 
347-356, 1972. 

Wallerstein, George, Peter S. Conti, and Jesse 
L. Greenstein, Interstellar lines in the spec- 
trum of the supernova in NGC 5253. Astro- 
phys. Lett., 12, 101-102, 1972. 

Weart, Spencer, What makes active regions 
grow? Astrophys. J., 177, 271-276, 1972. 

Weistrop, Donna, The luminosity function and 
density distribution of disk population stars. 
Astron. J., 77, 849-862, 1972. 

Westphal, J. A., J. Kristian, J. P. Huchra, S. A. 
Shectman, and R. J. Brucato, Search for a 
visible counterpart of the 2 September 1972 
radio outburst in Cygnus. Nature, 239, 134- 
135, 1972. 

Wilcox, John M., see Schatten, Kenneth H. 

Willner, S. P., see Huchra, John; Kirshner, R. P. 

Wilson, Olin C., Ira Sprague Bowen (1898- 
1973). Sky and Telescope, 45, 212-214, 1973. 

Wilson, W. J., P. R. Schwartz, G. Neugebauer, 
P. M. Harvey, and E. E. Becklin, Infrared 
stars with strong 1665/1667-MHz OH micro- 
wave emission. Astrophys. J., 177, 523-540, 

Woolf, N. J., W. A. Stem, F. C. Gillett, K. M. 
Merrill, E. E. Becklin, G. Neugebauer, and 
T. J. Pepin, The infrared sources in M8. 
Astrophys. J. (Lett.), 179, L111-L115, 1973. 

Wynn-Williams, C. G., E. E. Becklin, and G. 
Neugebauer, Infrared sources in the H II 
region W3. Mon. Notic. Roy. Astron. Soc, 
160, 1-14, 1972. 

Wynn-Williams, C. G., E. E. Becklin, and G. 
Neugebauer, New infrared sources associated 
with H II regions, in Interstellar Dust and 
Related Topics, Int. Astron. Union Symp. No. 
52, J. M. Greenberg and H. C. van de Hulst, 
eds., D. Reidel Publishing Co., Dordrecht- 
Holland, in press, 1973. 

Wynn-Williams, C. G., see also Becklin, E. E.; 
Epstein, E. E. 

Zirin, Harold, Response to Dr. Frazier's re- 
search note. Solar Phys., 26, 145-147, 1972. 

Zirin, Harold, The absence of flares in X3835 in 
the heating of the chromosphere. Solar Phys., 
26, 393-396, 1972. 

Zirin, Harold, High-resolution solar observa- 
tions from space, in Space Research XII, pp. 
1751-1763, S. A. Bowhill, L. D. Jaffe, and 
M. J. Rycroft, eds., D. Reidel Publishing Co., 
Dordrecht-Holland, 1972. 

Zirin, Harold, and Alan Stein, Observations of 
running penumbral waves. Astrophys. J. 
(Lett.), 178, L85-L87, 1972. 

Zirin, Harold, see also Foukal, Peter. 



Zwicky, Fritz, Radiations in outer space. Astro- 
naut. Acta, 17, 129-135, 1972. 

Zwicky, Fritz, Morphology of rapid cosmic 
processes, Astronaut. Acta, 17, 307-314, 1972. 

Zwicky, Fritz, Circular Letter No. 12 on super- 
novae, California Institute of Technology, 
May 1972. 

Zwicky, Fritz, Morphologisches Denken und 
Vorgehen, in Die Neuen Method der Ent- 
scheidungsftndung, pp. 130-143, Verlag Mod- 
erne Industrie, Munich, 1972. 

Zwicky, Fritz, In memory of Milton L. Huma- 
son. Eng. Sci., 36, 30-31, 1972. 

Zwicky, Fritz, 100 supernovae — the reward of a 
40-year search. Eng. Sci., 36, 20-21, 1973. 

Zwicky, Fritz, and Margrit A. Zwicky, Cata- 
logue of Selected Compact Galaxies and of 
Post-Eruptive Galaxies, Guemligen (Berne), 
Switzerland, 1971. 

Zwicky, Fritz, see also Kowal, C. T. 

Zwicky, Margrit A., see Zwicky, Fritz. 


Ira S. Bowen, Director of the Mount 
Wilson and Palomar Observatories from 
1946 to 1964, died suddenly on February 
6, 1973. 

As a graduate student, Bowen came to 
the California Institute of Technology in 
1921 to begin a productive decade in lab- 
oratory spectroscopy, physical optics, 
and the study of cosmic rays. His solu- 
tion of the enigma of the "nebulium 
lines" in 1928 earned him international 
renown; it also helped to swing his re- 
search interests irrevocably to the field 
of astronomy and astrophysics. 

Bowen assumed the Observatory di- 
rectorship at a time when the 200-inch 
telescope was nearing completion and 
when the Carnegie Institution and the 
California Institute were joining forces 
to establish and build a unique observa- 
tory organization for the advancement of 
research and teaching in astronomy. 
During a period of remarkable growth, 
Bowen presided over the expansion of the 
staff, the formulation of Observatory 
policy, and the guidance of research. The 
excellent performance and productivity 
of the 200-inch telescope reflect the high 
scientific and technical standards that 
he set for himself and, by example, for 
the staff. Since putting aside administra- 
tive concerns in 1964, Bowen had been a 
Distinguished Service Member of the 
Carnegie Institution and had continued 
to be active at the Observatories. In re- 
cent years his interests had centered on 
the design of optical systems for astron- 
omy. He made major contributions to 

the optical design of such wide-field 
telescopes as the Palomar 60-inch, the 
Las Campanas 40-inch, and the 100-inch 
Irenee du Pont Telescope that is now 
under construction for Las Campanas. 
Over the years his counsel on the optical 
design and technical aspects of telescopes 
was widely sought and freely given, to 
the benefit of practically all of the 
world's large telescopes of the past 

Alfred H. Joy, a staff member from 
1915 until his formal retirement in 1948, 
died on April 18, 1973, at the age of 90. 
He was the last survivor of the group 
that constituted the scientific staff prior 
to 1930. Internationally known and re- 
spected for his contributions to stellar 
spectroscopy, especially of variable stars, 
he pursued this field with characteristic 
energy and devotion for nearly 60 years. 
Dr. Joy was Secretary of the Mount Wil- 
son Observatory from 1920 until his 

Dr. Robert J. Brucato was appointed 
Staff Associate on September 15, 1972. 

Research Division 

Distinguished Service Member, Carnegie 
Institution of Washington 

Ira S. Bowen 1 

Staff Members 

Halton C. Arp 

Horace W. Babcock, Director 

Edwin W. Dennison 

1 Died February 6, 1973. 



Jesse L. Greenstein 2 

James E. Gunn 3 

Robert F. Howard 

Jerome Kristian 

Robert B. Leighton 4 

Guido Munch 3 

Gerry Neugebauer 5 

J. Beverley Oke, Associate Director 3 

George W. Preston 

Bruce H. Rule, Chief Engineer 

Allan R. Sandage 

Wallace L. W. Sargent 3 

Maarten Schmidt 6 

Leonard T. Searle 

Arthur H. Vaughan, Jr. 

Olin C. Wilson 

Harold Zirin 7 

Members Engaged in Post-Retirement 

Alfred H. Joy 8 
Alexander Pogo 
Henrietta H. Swope 
Fritz Zwicky 

Visiting Associates 

Clark G. Christensen 
N. Visvanathan 

Staff Associates 

Eric E. Becklin 9 
Robert J. Brucato 
Michael W. Werner 10 
James A. Westphal n 

2 Lee A. DuBridge Professor of Astrophysics, 
California Institute of Technology. 

3 Professor of Astronomy, California Institute 
of Technology. 

4 Professor of Physics ; Chairman, Division of 
Physics. Mathematics, and Astronomy, Cali- 
fornia Institute of Technology. 

5 Professor of Physics, California Institute of 

6 Professor of Astronomy ; Executive Officer 
for Astronomy, California Institute of Tech- 

7 Professor of Astrophysics, California Insti- 
tute of Technology. 

8 Died April 18, 1973. 

9 Senior Research Fellow in Physics, Cali- 
fornia Institute of Technology. 

10 Assistant Professor of Physics, California 
Institute of Technology. 

11 Associate Professor of Planetary Science, 
California Institute of Technology. 

Carnegie Fellows 

Ermanno F. Borra 
Michael M. Dworetsky 
Michael H. Hart 
Pieter C. van der Kruit 
Richard E. White 12 
Robert R. Zappala 

Research Fellows 

Andrew Michalitsanos 
Ronald Moore 
Jay M. PasachofT 13 
Stephen W. Prata 13 
William J. Quirk 13 
Katsuo Tanaka 14 

Carnegie-Chilean Fellows 

Eduardo Hardy 13 
Maria Teresa Ruiz 


Eleanor G. Ellison 
Marjorie A. Henderson 
Helen Z. Knudsen 

Senior Research Assistants 

Grace V. Knox 
A. Louise Lowen 

Research Assistants 

John M. Adkins 
John E. Boyden 
Frank J. Brueckel 15 
Ken D. Clardy 
Thomas A. Cragg 
Thomas S. Gregory 
Basil N. Katem 
Margaret Katz 
Charles Kowal 
Howard H. Lanning 16 
Susan B. Mellerup 
Malcolm S. Riley 

Student Observers 

Marc Aaronson 
William G. Bagnuolo 
Ian Gatley 

12 Resigned August 15, 1972. 

13 Resigned August 31, 1972. 

14 Resigned March 31, 1973. 

15 Retired December 31, 1972. 

16 Resigned October 24, 1972. 



Richard Green 
Paul M. Harvey 
Paul Hickson 
James A. Howell 
John P. Huchra 
Robert P. Kirshner 
John Kormendy 
Elliot C. Lepler 
Steven J. Loer 
Jorge Melnick 
Augustus Oemler, Jr. 
Valdar Oinas 
Glenn S. Orton 
Paul L. Schechter 
Stephen A. Shectman 
Barry E. Turnrose 
Glenn J. Veeder 
Bruce Waddington 
William E. Westbrook 
Richard A. Wickes 
Steven P. Willner 

Photographic Department 

John R. Bedke, Photographer 

William C. Miller, Research Photographer 

Instrument Design and Construction 

David A. Bell, Electronics Specialist 
Lawrence E. Blakee, Supervisor, Electronic 

Carroll Parrott Blue, Junior Photographic 

Laboratory Technician 
Robert E. Cadman, Senior Electronics Spe- 
Ertugrul Cepni, Electronics Engineer 17 
Cynthia V. Cable, Laboratory Technician 
Maynard K. Clark, Electronics Operations 

John P. Cowley, Engineering Assistant 
Martin A. Danihel, Junior Programmer 
Floyd E. Day, Optical Shop Superintend- 
ent 18 
Theresa M. DiBella, Junior Electronics 

Specialist 19 
Stephen Doro, Machinist 
Raymond Dreiling, Machinist 
Earle B. Emery, Research Engineer 
Eugene B. Fair, Optician 
Bradley R. Feinner, Senior Electronics 

Wiley P. Fowlie, Electronics Engineer 

17 Resigned August 14, 1972. 

18 Retired June 30, 1973. 

19 Resigned March 16, 1973. 

Jerry T. Fridenberg, Chief Electronics En- 
Robert D. Georgen, Machinist 
Richard M. Goeden, Engineer 
Simon Groesz, Senior Electronics Techni- 
Donn M. Hall, Senior Programmer 
Robert Stephen Hayes, Associate Engi- 
neer 20 
Robert J. Hippard, Photographic Labora- 
tory Aide 21 
Melvin W. Johnson, Optician 
Herman F. Kelderman, Research Engineer 
Leroy M. Kimoto, Senior Electronics Spe- 
John H. Knowles, Junior Electronics Tech- 
nician 22 
Wilfred H. Leckie, Senior Draftsman 
Steve A. Macenka, Senior Engineer 
William B. Mainland, Electronics Engi- 
neer 23 
William H. McLellan, Senior Engineer 
Charles R. McMahan, Electronics Engineer 
Martin J. Olsiewski, Senior Electronics 

Frederick G. O'Neil, Shop Foreman 
Emilio B. Perez, Junior Electronics Tech- 
Kenneth C. Perry, Junior Draftsman 
Rudolf E. Ribbens, Design and Shop Su- 
Alexander Rittenhouse, Junior Electronics 

Robert G. Stiles, Optician 
Merle R. Sweet, Electronics Technician 
David F. Thompson, Technical Assistant 
Virgil Z. Vaughan, Supervisor, Electronics 

Robert V. Webber, Laboratory Techni- 
cian 24 
Madeleine B. Williams, Draftswoman 25 
Felice Woodworth, Draftswoman 

Maintenance and Operation 

Mount Wilson Observatory and Offices 
Fern V. Borgen, Typist-Telephone Opera- 
Clyde B. Bornhurst, Mountain Superin- 

20 Resigned February 9, 1973. 

21 Resigned April 6, 1973. 

22 Resigned October 6, 1972. 

23 Resigned November 3, 1972. 

24 Resigned November 17, 1972. 



Herman E. Carpentier, Carpenter 

Hugh T. Couch, Superintendent, Buildings 
and Grounds 

Helen S. Czaplicki, Typist-Editor 

Sue H. DeWitt, Secretary 

James E. Dittmar, Night Assistant 

Raquel E. Ferrer, Translator-Secretary 

Hazel M. Fulton, Head Stewardess 

Eugene L. Hancock, Night Assistant 

Mary Hark, Stewardess 

John Kondratowicz, Jr., Night Assistant 

Jose Lopez-Tiana, Accounting Clerk 

Peter Mastrosimone, Custodian 

Jane Newcomb, Bookkeeper 26 

Glen Sanger, Senior Custodian 

Clair E. Sharp, Accountant 

William D. St. John, Chauffeur 

Frank Trylko, Custodian 

Warren Weaver, Assistant Mountain Su- 
perintendent 27 

Frederick P. Woodson, Assistant to the 

Palomar Observatory and Robinson 

Ranney G. Adams, Night Assistant 

Mayme L. Adkins, Secretary 28 

Albert R. Andrews, Junior Maintenance 

Ray L. Ballard, Senior Administrative As- 

Mary E. Bauer, Secretary 

Kathleen S. Brown, Secretary 

Jan Adriaan Bruinsma, Painter and Gen- 
eral Maintenance 

Maria J. Bruinsma, Housekeeping Aide 

Juan R. Carrasco, Night Assistant and 
General Mechanic 

Rita M. Duprey, Accounting Clerk 

Gail Fuqua, Accounting Clerk 29 

Beulah Greenlee, Housekeeping Aide 

Frank Greenlee, Senior Custodian 

Delia Harris, Secretary 26 

Liselotte M. Hauck, Secretary 

Victor A. Hett, Maintenance Mechanic 

Helen Holloway, Secretary 

Taras Kiceniuk, Mountain Superintendent 

J. Luz Lara, Mechanic 

25 Terminated June 30, 1973. 

26 Resigned July 31, 1972. 

27 Resigned June 30, 1973. 

28 Resigned February 23, 1973. 

29 Resigned April 4, 1973. 

Patricia McKay, Secretary 30 

Dorothy J. Mulkey, Records Clerk 

Carl D. Palm, Night Assistant 

Marilynne J. Rice, Secretary 

Tamsin S. Schulte, Secretary 

Sidney R. Shannon, Assistant Mountain 

Superintendent 31 
Gabrielle South, Accounting Clerk 32 
Nadia Tesluk, Clerk-Typist 
Elsa B. Titchenell, Secretary 
Gary M. Tuton, Senior Night Assistant 
Paul Van Ligten, Maintenance Mechanic 

and Electrician 
Ruth E. Weaver, Clerk-Typist 
David K. Williams, Relief Night Assistant 

and Maintenance Mechanic 

Big Bear Solar Observatory 

Jack R. Klemroth, Laboratory Assistant 
Eugene H. Longbrake, Superintendent and 

Senior Electronics Specialist 
Charles F. Mason, Custodian 
Walter M. Nagao, Jr., Mechanic 
Thomas P. Pope, Sr., Chief Observer 33 
Georg Schacht, Laboratory Assistant 

Las Campanas Observatory 

Bruce Adkison, Associate for Administra- 
tion, Pasadena 

Wilma J. Berkebile, Secretary 

S. Thomas Couch, Shipping and Receiving 

Wayne Thomas, Accounting Clerk 

Manfred Wagner, Project Supervisor, La 

Irenee du Pont Telescope Project 

Louis E. Beidler, Engineer 

Betty A. Browne, Secretary 

Oscar D. Dubon, Field Engineer, Las 

C. L. Friswold, Instrument Designer 
Richard C. Haskell, Project Engineer 
Harald Scott Jensen,. Engineer 27 
Charles W. Jones, Engineer 31 
Roger L. Minnix, Engineer 
Bruce H. Rule, Chief Engineer and Project 

Edward H. Snoddy, Engineer 
Albert Villanueva, Electrical Engineer 

30 Resigned June 15, 1973. 

31 Resigned April 30, 1973. 



Resigned April 13, 1973. 
Resigned September 22, 1972. 

of Terrestrial Magnetism 

Washington, District of Columbia 

Ellis T. Bolton 

L. Thomas Aldrich 
Associate Director 

Carnegie Institution of Washington Year Book 72, 1972—1973 


Introduction 169 

Astrophysics 173 

Introduction 173 

Optical astronomy 173 

Radio astronomy 179 

Equipment development 179 

1.35-cm water-line observations 180 

Search for 21-cm hydrogen absorption during flare-ups of Cygnus X-3 ... 181 

Survey of neutral hydrogen in the vicinity of 17 galactic supernova remnants . 182 
Accurate radio positions for Ohio radio sources, using the NRAO three-element 

interferometer at 3.7 cm and 11.1 cm 184 

Institute Argentino de Radioastronomia activities 185 

Atomic physics 186 

Nuclear physics 190 

Improvements in the Atomic Physics Observatory 190 

Boron-10 191 

A study of (p, n) reactions 194 

Biophysics 194 

Introduction 194 

Studies of memory 195 

The mouse satellite in two subspecies of Mus musculus 197 

Relatedness among several hamsters 200 

Instantaneous binding fractions of DNA 204 

Enzyme evolution in the Enterobacteriaceae 207 

Sonication of DNA to produce fragments suitable for reassociation experiments . 214 

Expression and divergence of repeated DNA sequences 217 

Labeling of DNA with 125 I 217 

Cooperative research at California Institute of Technology 221 

Interspersion of sea urchin repetitive and nonrepetitive DNA sequences . . . 222 
Size distribution of interspersed repetitive sequences in sea urchin DNA by 

reassociation and single-strand-specific nuclease digestion 222 

Hyperpolymers and another approach to repetitive sequence organization . . 223 

General assessment of advances in the knowledge of DNA sequence arrangement 224 

Geophysics 224 

Introduction 224 

Cosmic-ray research 225 

A comparison of the anelasticity structure between western South America and 

Japan 226 

Empirical models for anomalous high-frequency arrivals from deep-focus earth- 
quakes in South America 233 

Converted P phase from the ScS phase at the inclined deep seismic zone . . . 238 

Source spectral relations for earthquakes 245 

Borehole strainmeter installation in Japan 246 

Project Naririo 247 

Electrical conductivity studies 249 

Geochemistry and geochronology 252 

Geochemical trends in Andean rocks 252 

The geochemistry of basalts from Iceland and the Reykjanes Ridge .... 259 

Geochemistry of ultramafic inclusions from Salt Lake Crater, Hawaii .... 262 
Differential dissolution technique (DDT) : chemical separation of crystals from 

glass 268 

Trace-element content of liquid formed by partial melting of a garnet lherzolite 

at high pressures : a preliminary report 270 

Trace-element contents of clinopyroxenes from garnet lherzolites in kimberlites 272 

Metasomatic theory : some examples of diffusion and infiltration 276 

Ion microprobe analysis of an isotopic diffusion experiment on biotite . . . 280 

Direct determination of rock rheology from wavelength selection in folding . . 283 
U-Pb isotopic studies of zircons from the Baltimore Gneiss of the Towson Dome, 

Maryland 285 

U-Pb isotopic studies of zircons from the Gunpowder granite, Baltimore County, 

Maryland 288 

U-Pb isotopic analyses of zircons from granulite and amphibolite facies rocks 

of the West Chester Prong and the Avondale Anticline, southeastern 

Pennsylvania 290 

Age and origin of zircons from metamorphic rocks in the Manhattan Prong, 

White Plains area, southeastern New York 293 

Old radiogenic lead components in zircons from the Idaho batholith and its 

metasedimentary aureole 297 

Effect of regional metamorphism on whole-rock Rb-Sr systems 299 

Uranium gain of detrital zircons studied by isotopic analyses and fission track 

mapping 302 

Special Activities 304 

References Cited 306 

Bibliography 312 

Personnel 314 


No field of thought can be properly laid out by men who are merely 
measuring with a ruler. Sections of history are liable to be transformed 
— or, even where not transformed, greatly vivified — by an imagination 
that comes, sweeping like a searchlight, from outside the historical pro- 
fession itself. Old hunches are then confirmed by fresh applications of 
the evidence or by unexpected correlations between sources. New matter 
emerges because things are joined together which it had not occurred to 
one to see in juxtaposition. New details are elicited, difficult details be- 
come relevant, because of a fresh turn that the argument has taken. 

Herbert Butterfield 

The Sleepwalkers, by Arthur Koestler, 
Grosset & Dunlap, New York, 1959 

This report illustrates the thrust of the 
quotation from Professor Butterfield, an 
historian of science. It is no accident 
that, though we measure with astonish- 
ingly sophisticated "rulers," new matter 
has emerged. The reader will perceive 
this fact in our efforts this year. 

Rubin and Ford have brought to frui- 
tion their studies of the internal motions 
in Andromeda — a galaxy generally be- 
lieved to be a homologue of our own. It 
can be stated that our "sister" is dynam- 
ically and structurally complex, and still 
full of puzzles. Particularly, old stars 
and (presumably) young gas have similar 
motions. Perhaps the gas and associated 
dust are shed outw r ard from stars which 
are undergoing evolution — losing mass, 
and changing their chemical composition 
through nuclear transmutations. Most 
exciting at present is the recent discovery 
of the anisotropic distribution of veloc- 
ities on the celestial sphere of a class of 
galaxies known as Scl — a special class of 
spiral galaxies of high luminosity. 

From an all-sky sample of 208 such 
galaxies, radial velocities, i.e., the speeds 
at wdiich these objects are receding from 
the Sun, have been determined for 70-odd ; 
50 velocities range from 4000 to 7000 kilo- 
meters per second. Strikingly, 28 of these 
have velocities tightly grouped about 5000 

km/sec (actually, 4966 ± 122 km/sec) 
wdiile 22 have a similar grouping about 
6500 km/sec (actually, 6431 ± 160 km/ 
sec) . Each group is positioned on a differ- 
ent region of the sky ( approximately one- 
third of the total for each) and there is 
virtually no overlap in the positions of 
the two groups. These observations raise 
most important cosmological problems; 
the most unorthodox is to question the 
validity of the w r idely accepted idea that 
the Hubble expansion (which relates re- 
cessional velocity to distance) is isotropic 
throughout the universe. Other alterna- 
tives are discussed in the Astrophysics 
Section report. 

In the report of this section are also in- 
cluded observations and conclusions in 
radio astronomy, atomic and nuclear 
physics. They include the work at the 
NRAO, Greenbank, West Virginia, at our 
Derwood, Maryland, Observatory, and 
the cooperative effort w r ith the joint CIW- 
Argentine installation outside La Plata 
where Dr. Kenneth Turner has served as 
interim director for the past two years. 
The focus at NRAO has been on 21 -cm 
hydrogen line observations of large 
galactic supernova remnants in order 
to study the effect of supernova rem- 
nants on the interstellar medium. The 
South American studies have centered 




principally upon neutral hydrogen radia- 
tion of selected objects. Many Argentine 
students and scientists have engaged in 
this almost unexplored area and they are 
beginning to yield some of the harvest 
envisioned a decade and more ago. It is 
our conviction that continuing coopera- 
tion will prove fruitful. 

At our local Derwood Observatory a 
large investment of talent and time has 
been expended in improving the facility 
for work at higher frequencies than 1420 
MHz in order to study molecules in 
our Galaxy and also to have at hand 
a readily available facility for Fellows 
and local students, especially from the 
University of Maryland and The Johns 
Hopkins University. Nevertheless, the 
H-line work, which has been our main- 
stay, continues with much improved effi- 
ciency, with continuing emphasis on the 
study of high velocity neutral hydrogen 

In the areas of atomic and nuclear 
physics our work, largely under Dr. 
Louis Brown's aegis, has taken on 
new significance. Twenty years ago 
Drs. Norman Heydenburg and Georges 
Temmer studied low-lying excited states 
of nuclei through Coulomb excitation 
with a-particles. In recent years it has 
become possible in this and other labora- 
tories to study atoms by bombardment 
with heavy ions which induce both nu- 
clear and atomic excitations over a wide 
range of energies, and a new phenomenon, 
electron promotion, has been uncovered. 
In the case of bombardment of target 
atoms by heavy ions, the simple theory 
which assumes an interaction of two 
point charges during Coulomb excitation 
must be modified to account for the pres- 
ence of the projectile ion within the K 
shell of the target atom. The incoming 
ion lingers within the target atom for 
relatively long periods of time compared 
with the oscillation times of the K elec- 
trons. If the electron shells of ion and 
target atom overlap, then a "hyperatom" 

is formed consisting of the combined nu- 
clei and a new K shell. During this for- 
mation two of the four electrons which 
made up the two originally independent 
K shells are forced out, or "promoted." 
When the two nuclei later separate there 
is a high probability that each will have 
vacancies in their K and L shells. In turn, 
there is a high probability of ionization 
and a consequent high cross section for x- 
ray production. In cooperation with Dr. 
H. A. Van Rinsvelt of the University of 
Florida, studies were undertaken to 
bombard copper by beams of alkali ions, 
which are readily produced in the DTM 
Van de Graaff accelerator and offer a 
wide range of atomic numbers. It has 
been found that the characteristic x-ray 
lines of copper are shifted in energy as a 
result of the high probability of ionizing 
the L shell when the K shell is ionized. 
The energy shifts and the cross section 
for ionizing the K shell result from two 
effects: Coulomb excitation and electron 

Studies of brain biochemistry, the 
evolution of DNA in rodents, the phys- 
ical chemistry of nucleic acid interactions 
in vitro, and the immunochemistry of en- 
zyme evolution have occupied the atten- 
tion of the Biophysics Section. 

In recent months there have been reports 
which implicate peptides in learning and 
memory. N-acetyl aspartic acid (NAA) , 
which is found only in the brain, stimu- 
lates peptide synthesis. Consequently, 
experiments were set up to test whether 
NAA might affect learning and memory 
in mice. Performance in the avoidance of 
a mild electric shock was improved by 
nearly twofold. The improvement per- 
sisted for at least a week after training 
and NAA withdrawal, indicating that 
NAA may well work on the consolidation 
of memory rather than as a response to 
the continued administration of the drug. 

Studies of the evolution of DNA in 
closely related mice and other rodents 



have continued. This is a most curious 
area in the bio-sciences, and some of the 
potential for understanding evolution at 
the molecular level would seem to reside 
here. It has been known for some time 
that the common laboratory mouse (Mus 
musculus musculus) contains a "satellite" 
DNA which appears upon centrifugation 
in CsCl density gradients and which con- 
sists of highly repeated sequences (about 
a millionfold) 300 or so nucleotides in 
length. Satellite DNAs have been found 
in other Mus species (M. caroli and M. 
cervicolor) , yet these DNAs appear only 
loosely related to those of M. m. musculus, 
while the satellite DNA from a Japanese 
subspecies (M. musculus molossinus) is 
very similar by DNA cross-hybridization 
tests even though the Japanese species 
contains much less of the satellite. 
Phenotypic differences are obvious: The 
species differ in size, coloration, and 
chromosomal rearrangements and they 
interbreed to form fertile hybrids. Thus, 
tracking down the functional difference 
in the DNA of these species with differ- 
ent satellite contents remains a challenge. 
Many mechanisms are at play in 
establishing genetic diversity even among 
bacteria. Ultimately, diversity is trace- 
able, according to present understanding, 
to the message encoded in the nucleic 
acids. Some portions of bacterial or viral 
genomes are highly conserved as would 
be required for the production of enzymes 
and enzyme systems necessary for ac- 
commodation to the environment or 
changes in the environment and hence to 
survival. Enzymes, being proteins and 
generally antigenic in foreign species are 
subject to immunologic analysis. Mod- 
ern organic chemistry has made it possi- 
ble to couple appropriate antibodies to 
insoluble substrates such as the polysac- 
charide sepharose and thus permit anti- 
bodies to select enzymic antigens from 
crude bacterial extracts with high effi- 
ciency. Throughout the Enterobacteri- 
aceae the distinctive features of the regu- 

lation of the enzyme aspartokinase I- 
homoserine dehydrogenase I (AKI- 
HSDI) are conserved even though the 
compositions of the DNAs vary widely. 
It is accordingly possible to investigate 
immunologic affinities of the enzyme, the 
conformational states of the enzyme as 
ligands are introduced into the system, 
and to draw inferences concerning the 
systematic relationships of specific pro- 
teins as contrasted with the similarities 
and differences among the total genomes 
indicated by nucleic acid interactions 
among the bacteria. These studies are 
described in detail by Cowie and his col- 
laborators, Georges N. Cohen and Paolo 
Truffa-Bachi of the Pasteur Institute in 

In his introduction Dr. L. T. Aldrich, 
Associate Director and Chairman of the 
Department's Geophysics Section, ap- 
propriately describes the appreciation 
we have for the opportunity to ex- 
press our creativity on a global scale. 
His words — substituting only the names 
of different sets of fellow scientists — 
apply equally to our activities involving 
cooperation in research in Astrophysics 
and Biophysics and to the generous sup- 
port provided to the Department by the 
National Science Foundation, the U.S. 
Public Health Service, and numerous 
agencies throughout the world. This 
spirit of cooperation in which we have 
most frequently taken the initiative to 
stimulate new ventures on new frontiers, 
wherever these may lie, has been a hall- 
mark of the Department's way of life 
since its very beginning. Indeed, the 
Department's first proposed name was 
"International Magnetic Bureau of the 
Carnegie Institution of Washington." 
A few paragraphs illustrate the wide 
sweep of the work of the Geophysics 

Dr. Sacks and Mr. Hiromu Okada of 
Hokkaido, now a predoctoral Fellow at 
the Department, have undertaken a com- 
parison of the Q structure beneath the 



South American continent and the Japa- 
nese Archipelago. Similar spectral ratio 
techniques were used for Q determina- 
tions in the two regions. In South Amer- 
ica, Q p in the somewhat seismically active 
wedge bounded by the seismic plane, the 
trench and volcanoes, is 1000-2000. Be- 
neath Japan for the corresponding region, 
which is aseismic, Q p is 400-500; this 
value is identical to that in the astheno- 
sphere below the trench, which indicates 
there is no evidence for extra low Q 
values beneath the volcanoes. In South 
America, the high Q p values (up to 
3000) extend to 350 km, but there do not 
seem to be any paths from deep earth- 
quakes (600 km) with Q p greater than 
1000 ; therefore, there must be a Q inver- 
sion below 350 km. In Japan, however, 
the high Q p values in the Benioff zone 
(«3000) persist down to the deep earth- 
quakes. Above the deep earthquake zone 
in Japan and Fiji, very low Q p values 
(100) were found, in contrast to a mini- 
mum value of 350 in South America. 
Ray paths through these very low Q 
zones have positive arrival time anoma- 

The tectonic implications of these 
structures are that lithospheric plates 
have varying thicknesses and continental 
plates are several times thicker than 
oceanic plates. 

In another study relating to massive 
events concerning tectonic processes, Dr. 
Hart, in cooperation with Dr. J. G. 
Schilling of the University of Rhode 
Island, has examined basalts from Ice- 
land, and the southerly track of the mid- 
Atlantic ridge has been analyzed with 
respect to trace element composition and 
87 Sr/ 86 Sr isotope ratios. Tholeiites from 
a 500-km section of the Reykjanes ridge 
and its extension on Iceland were ana- 
lyzed for K, Rb, Cs, Ba, Sr, and 87 Sr/ 86 Sr. 
Samples most distant from Iceland show 
87/86 ratios similar to typical ridge ba- 
salt (.7026-.7028) but are more highly 
depleted in these L1L elements (K < 400 

ppm, Sr < 80 ppm) and show less frac- 
tionated element ratios (K/Rb < 900, 
K/Ba < 65). Concentrations increase 
progressively as Iceland is approached 
(factors of 2-5) ; 87/86 ratios increase 
stepwise to .7030 about 150 km from Ice- 
land. Tholeiites from Iceland show a 
considerable range in concentration (K, 
700-2000 ppm; Sr, 90-200 ppm) which 
overlaps and is somewhat higher than the 
nearest ridge samples. 87/86 ratios on 
Iceland show little spread (.7030-.7031). 
The isotopic data show that the mantle 
sources under Iceland are chemically 
different from those under the southern 
Reykjanes ridge. The concentration 
variations between the two provinces are 
not well described by mixing models in- 
volving only two melt types. A possible 
model involves variable partial melting 
and mixing of two mantle sources. 

Drs. Shimizu and Hart have developed 
a sensitive chemical procedure for the 
analysis of trace elements in two-phase 
systems in which one phase is crystalline 
and the other, a quenched liquid that is 
a glass. They call their procedure the 
"Differential Dissolution Technique," or 
"DDT." The procedure takes advan- 
tage of the fact that the glass por- 
tion is dissolved in diluted hydrofluoric 
acid much more readily than the highly 
ordered crystalline portion. Thus, in a 
series of model experiments, it was found 
that in a mixture of clinopyroxene (from 
a spinel lherzolite from Salt Lake Crater, 
Hawaii) and glass (tholeiitic basalt glass 
from the Juan de Fuca Ridge) the glass 
dissolved, leaving pure clinopyroxene as 
a residue. Milligram amounts of crude 
starting material containing only parts 
per billion of trace elements can be freed 
of contaminating glass and analyzed 

In these vignettes of the work of the 
past year, and especially in the body of 
the Report, it is hoped that the reader 
will find "new matter emerging and new 
details elicited, difficult details becoming 
relevant, because of fresh turns that the 
arguments have taken." 




G. E. Assousa, L. Brown, E. T. Ecklund, W. K. Ford, Jr., C. K. Kumar, C. A. Little, 
U. Rohrer, V. C. Rubin, N. Thonnard, K. C. Turner, M . A. Tuve, and J. W. Warner 


The preliminary results of an investi- 
gation begun two years ago by Rubin and 
Ford have brought into question a crucial 
assumption under which observations of 
distant galaxies are interpreted as meas- 
urements of an expanding universe. They 
have found an anisotropic grouping over 
the sky of the redshifts of galaxies whose 
magnitudes and angular sizes lead one to 
expect them to lie at about the same dis- 
tance. As they point out, this does not 
necessarily imply that the expansion is not 
isotropic at these distances, but it does 
require a more critical interpretation of 
redshift observations than has heretofore 
been made. 

Observations of the water-line signals 
from W-49, of the hydrogen line emission 
near supernova remnants, and of an ob- 
ject whose spatial velocity and structure 
suggest its ejection from the galactic cen- 
ter illustrate the varied activities in radio 
astronomy of the Department's staff at 
the Derwood Experimental Laboratory 
(Derwood, Maryland) , the National 
Radio Astronomy Observatory (Green 
Bank, West Virginia) , and the Institute 
Argentino de Radioastronomia (Villa 
Elisa, Argentina), respectively. Recent 
studies of x rays produced with fast, 
heavy ions point out the versatility of our 
Van de Graaff machine, now 35 years of 
age and the second oldest operating Van 
de Graaff (atom smasher) in the world. 
It has recently been used for optical, 
x-ray, and nuclear spectroscopy, thereby 
spanning microscopic structures with en- 
ergies from less than 1 eV to several MeV. 

Improvements of the instrumental ca- 
pabilities, some of them major, took place 
throughout the section. One can see over 
the years a continuous trail of instru- 
mental improvements, which have ab- 
sorbed a large fraction of the staff's time 
but which are indispensable for observa- 

tional progress. Particularly noteworthy 
this year was the use of nitrogen-baked 
Illa-J plates, a simple procedure that 
greatly enhances the value of the image 
tube— photographic plate combination. 
Work was begun to exploit image tube 
capabilities at infrared wavelengths, and 
major improvements were made at the 
Derwood and the IAR-CIW telescopes, 
resulting in a much improved system 
noise figure. The Van de Graaff machine, 
our Atomic Physics Observatory, re- 
ceived its third accelerator tube and a 
magnetic tape recording system for its 
pulse-height analyzer. 

Optical Astronomy 

W. K. Ford, Jr., C. K. Kumar, Vera C. Rubin, 
M. A. Tuve, and J. W. Warner 

This report year has seen the comple- 
tion of one major program, the study of 
internal motions in the Andromeda Gal- 
axy, and the early stages of two others: 
the investigation of the isotropy of the 
Hubble expansion and a program to ob- 
tain direct photographs and spectra of 
nebulae and galaxies at a wavelength 
near one micron. These are described 

The isotropy of the Hubble expansion. 
Rubin and Ford discussed last year the 
start of an observing program to deter- 
mine radial velocities of Scl galaxies 
(spiral galaxies of high luminosity and 
well-defined arms) with 14.0 < m < 15.0. 
From an all-sky sample of 208 galaxies 
they now have velocities for 75; 50 ve- 
locities are in the range 4000 < V < 7500 
km/sec. There is a striking nonrandom 
distribution of these velocities on the sky, 
as shown in Fig. 1. One-third of the sky 
contains 28 galaxies, with <V 7 j> = 4966 
± 122 km/sec; one-third contains 22 gal- 
axies, with <T n > = 6431 ± 160 km/sec, 
with almost no overlap between the two 
velocity groups. The mean magnitudes 




O 4000 <V< 5400 km/s (28) 
• 5400<V<6100km/s (2 
■ 6100 <V<7500 km/s (20) 

Fig. 1. Distribution on sky of Scl program galaxies, 14.0 < m < 15.0, with observed velocities 
between 4000 km/s and 7500 km/s; equatorial coordinates. Note separations on sky of gal- 
axies with different velocities. 

and mean diameters of galaxies are simi- 
lar on both parts of the sky. 

There are several possible explanations 
for this curious distribution, which they 
are presently investigating while continu- 
ing to gather radial velocities of sample 
galaxies. These include: (a) Significant 
intergalactic or excess galactic absorp- 
tion ( — 0.5-1 mag.) over large regions of 
space, (b) A large (^1000 km/sec) mo- 
tion of our Galaxy and the local group of 
galaxies toward the lower velocity gal- 
axies. Both in velocity and direction, 
such a motion would violate the conclu- 
sions from the observed isotropy of the 
3°K background relict radiation, (c) 
Large-scale clumping of galaxies or dif- 
fering luminosity properties of Scl gal- 
axies over regions of 100-200 Mpc, such 
that the galaxies observed with <Fi> = 
4966 km/sec are actually 30 Mpc closer, 
as their velocities would imply, but in- 
trinsically (0.5 magnitude) fainter than 
those observed with < V n > = 6431 km/sec. 
(d) A difference of ^20% in the value of 
the Hubble constant in regions I and II. 
Such an effect might be due to local gravi- 
tational distortions, remnants of the big 

bang, or second-order effects in the 
Hubble expansion. 

In collaboration with Dr. M. S. Roberts 
of the National Radio Astronomy Ob- 
servatory, Rubin has started a program 
to observe neutral hydrogen in the Scl 
program galaxies, using the NRAO 300- 
foot telescope. They expect to determine 
both HI content and velocity, at least for 
the nearer objects in the sample. 

Internal motions in the Andromeda 
Galaxy. Rubin, Ford, and Kumar com- 
pleted the analysis of stellar motions near 
the nucleus of M31. There is a complex 
pattern of stellar motions, generally simi- 
lar to the complex motions observed 
earlier in the ionized gas. Why the mo- 
tions of the old stars and young gas 
should be similar is a puzzle, perhaps sug- 
gesting that the gas and dust arise from 
the material shed by the evolving stars. 

Since 1967, Rubin, Ford, and Kumar 
have studied the rotation of the outer 
regions of M31, gas motions and stellar 
motions near the nucleus, and the varia- 
tion of chemical abundance or ionization 
processes across the nucleus and spiral 
disk. 1 ' 2 - 3 - 4 



Infrared investigations. Warner is 
working on several programs involving 
the infrared. After testing and evaluating 
14 Carnegie SI image tubes, he built two 
image tube systems, one to be used with 
the DTM spectrograph and the other as 
a direct camera. The spectrographic sys- 
tem is being used to obtain strengths of 
the [S III] 9069, 9532, He I 10830, and 
Paschen lines in external galaxies and 
galactic emission regions for chemical 
abundance studies. 

The direct camera will be used for the 
study of extended objects in highly ob- 
scured regions and for photography in 
the light of infrared emission lines. 

In collaboration with R. F. Wing of 
Ohio State University, Warner is continu- 
ing a program of identification and photo- 
electric photometry of sources from the 
California Institute of Technology Two- 
Micron Sky Survey. To date, 50 identifi- 
cations and spectral types have been ob- 
tained in the —20° and —30° declina- 
tion zones near the galactic center. All 
but one are M stars, the exception being 
IRC 20385 = Terzian 5, a heavily red- 
dened globular cluster at approximately 
9 kiloparsecs distance in the general di- 
rection of the center of our Galaxy. A 
number of supergiants have been found 
as well as one object (IRC 30350) with 
the properties of VX Sagittarius. The 
purpose of this work is to make statistical 
studies of the distribution of cool stars in 
this region of our Galaxy. 

Far ultraviolet variability of quasi- 
stellar objects. Quasi-stellar objects 
(QSOs) are known to vary in the radio, 
optical, and near-ultraviolet spectral re- 
gions. Observations have not yet been 
made in the far ultraviolet (FUV, A res t 
< 912 A) . A knowledge of the variability 
in the FUV is important if one is to 
properly interpret the spectra of QSOs. 
Kumar has interpreted the observed hy- 
drogen line intensities in the spectra of 
quasars to indicate that these objects 
could vary by about 2.5 magnitudes in 
the far ultraviolet. 

The following simple model is adopted: 
A central source emitting the continuous 
radiation is surrounded by the line- 
emitting region (nebula). The nebula 
consists of two parts: one, optically thin 
to Lyman continuum radiation, and the 
other, optically thick. Observations of 
high-redshift objects indicate the pres- 
ence of the optically thin part. Measure- 
ments of 4C 05.34 (Oke, 1970) 5 and PHL 
957 (Lowrance et al., 1972) 6 indicate 
that the optical thickness (to) of the 
nebula at A res t = 912 A is of the order of 
unity. This optically thin part subtends 
a solid angle 4tt£ at the central source. 
The factor e is found to have the value 
1/2 from the observation that nearly one- 
half of the 36 high-velocity (Z = AA -f- A 
> 1.8) objects show Lya in absorption. 
Similarly, observations of low redshift 
objects (0.25 < Z < 0.75) show that only 
4 out of 70 objects have Mg II in absorp- 
tion. Since the Mg II ion is abundant 
mainly in the H I zone, the optically 
thick part of the nebula subtends a solid 
angle of 0.24 it (= 4/70 X 4tt) at the 
central source. Consequently, the two- 
component model is defined to be ei = 



2 and eo = 



Assume that the nebula is primarily 
photoionized, that the spectral index of 
ionizing radiation is given by a power law 
with spectral index a = 1.5, and that the 
line is formed mainly by recombination. 
Then the ratio of the number of Ha pho- 
tons emitted by the nebula r to the num- 
ber of Lyman continuum photons emitted 
by the central source N(Ha)/N(LyC) 
can be computed. 

From the measured H/3 fluxes in a 
number of QSOs (Oke et al, 1970) 7 and 
the predicted Ha/H/3 intensity ratio of 
2.84, Kumar has calculated values of the 
Ha flux. The "observed" N(Ha)/N(LyC) 
ratios are plotted in Fig. 2, along with the 
model prediction shown as a solid line. 
It is clear that in most cases the observed 
values exceed the prediction and the 
average observed value is twice the pre- 
dicted value. This excess emission line 







£ 0.31- 





-O— O- 

i I I 

0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 

Fig. 2. Comparison of observed (circles) and 
predicted (solid line) N(Ha)/N(Ly C) ratios of 
quasi-stellar objects. 

strength is attributed to line formation at 
a time when the FUV intensity was higher 
than the average intensity. 

The time-dependent behavior of line 
intensities under the influence of a vari- 
able flux of ionizing photons has been dis- 
cussed (Bahcall et al, 1972). 8 The work 
of Angione (1970) 9 shows that on time 
scales larger than 10 years the QSOs are 
incessantly active and are above their 
minimum luminosity at least half the 
time. This information and the relevant 
formulae given by Bahcall et al. shows 
that on the average the amplitude of 
variation of QSOs in the far ultraviolet is 
about 2.5 magnitudes for this model. 

The deduced variability of 250 S is 
based on two rather important assump- 
tions. They are (1) that the far UV con- 
tinuum is given by the spectral index 1.5, 
and (2) that interstellar extinction is 
negligible. These are the assumptions 
generally made by investigators attempt- 
ing to explain line intensities in 250 S. It 
is proposed to continue this investigation 

by removing the above assumptions and 
trying to explain the hydrogen and 
helium line intensities in 250 S. 

Anticenter blue stars. A search for blue 
stars near the galactic plane at large dis- 
tances from the sun in the galactic anti- 
center direction is under way by Tuve 
and Rubin using three-color plates taken 
with the 48-inch Palomar schmidt in 
1965. This is a part of our interest in 
determining the rotation of the outer 
parts of the galaxy which can only 
broadly be determined by neutral hydro- 
gen observations. 

Image tube development. The cas- 
caded image tube that has been used so 
successfully in the Carnegie image tube 
systems was designed ten years ago, early 
in the 1960's. In 1969 the same basic type 
of tube was built at RCA with more 
modern fabrication techniques, including 
the use of a ceramic rather than a glass 
envelope. These tubes were operable at 
higher voltages and hence had higher 
brightness gains than the older, low- 
voltage tubes. The general performance, 
however, was inferior to the older tubes 
because of low-quality phosphor screens 
and, in some tubes, excessive ion emis- 
sion. Consequently, with financial sup- 
port from the National Science Founda- 
tion, we instigated a two-year develop- 
ment program at RCA on the RCA 
C33063 image tube. This development 
program has now resulted in an improved 
version of the Carnegie cascaded tube. 

The work at RCA has been directed 
toward (1) reducing the ion scintillations 
and other sources of spurious background, 
(2) reducing the geometrical distortions 
in the tube by redesigning the electron 
optics, (3) improving the quality and 
uniformity of the phosphor screens, and 
(4) improving the contrast in the in- 
tensified image by reducing the amount 
of internally scattered light in the tube. 

The higher gain of these tubes makes 
the ion scintillations a particular prob- 
lem. Unless the system is operated with 



the transfer lens stopped down, each ion 
scintillation is recorded on the photo- 
graphic emulsion as a discrete image. 
After a long exposure, the plate appears 
to be covered with these random faint 
star images as though the plate had been 
sprinkled with black pepper. To reduce 
the number of the scintillations, the RCA 
engineers have relocated the alkali chan- 
nel electrode used in cathode formation to 
a more nearly field-free region. The 
chromoxide coating technique used to 
provide an even voltage distribution in 
the tube was improved. A better vacuum 
technique was introduced to reduce the 
chance of high-energy collisions of photo- 
electrons with free gas molecules. Better 
assembly and clean-room procedures were 
also introduced. These steps have not 
eliminated the ion scintillations but have 
greatly reduced their number per unit 

The geometrical distortions in a mag- 
netically focused tube depend on the uni- 
formity of the externally generated mag- 
netic field and the electric field generated 
internally by accelerating electrodes. To 
improve the electronic imaging, the 
flanges on the tube bulb, or envelope, 
and the accelerator spacers were rede- 
signed, resulting in a reduction of image 
rotation and spiral and linear distortions. 
There are three accelerating electrodes 
per stage in the C33011 and four acceler- 
ating electrodes per section in the C33063. 
The new C33063 is 6.375 inches in overall 
length, or 1.5 inches longer than the old 
Carnegie tube. Hence a longer, lower 
field-strength focusing magnet and a 
modified voltage divider are required to 
operate the new tube. 

The performance of the cascaded image 
intensifiers has been limited by the pres- 
ence of mottling and fine structure in the 
phosphor screens. Attempts to use the 
tube to detect weak spectral lines or 
faint objects in a field photograph were 
often frustrated by the uncertainty intro- 
duced by mottling in the screens. In the 
course of the improvement program at 

RCA, considerable attention was directed 
to identifying and controlling the sources 
of structure in the screens. The average 
particle size of the phosphor material 
is many times smaller than the average 
size of mottle pattern. Many test screens 
were prepared utilizing a variety of depo- 
sition, drying, filming, and aluminizing 
techniques. Microphotographs were taken 
of the screens at each step of processing. 
It was found that the organic filming 
used to isolate the phosphor from the 
aluminum back coating played a much 
more critical role in determining screen 
uniformity than was previously supposed. 
Various procedures have now been intro- 
duced to control the filming and, conse- 
quently, much-improved screens are in- 
corporated in our recent intensifiers. 

An attempt has been made to reduce 
the amount of light internally reflected 
and scattered about by bright surfaces 
within the tube. Also, we are experiment- 
ing with thin cover slides having an effi- 
cient antireflection coating that can be 
oiled to the tube faceplate to reduce re- 
flections at the air-glass surface. It is 
hoped that these techniques will improve 
the contrast in the image tube system. 

One of these improved RCA C33063 
tubes has been mounted with a Bowen 
f/1.4 Cassegrain schmidt camera for use 
on the Kitt Peak 84-inch Cassegrain 
spectrograph. This system is proving to 
be a substantial improvement over the 
original Carnegie system. 

The development of a digitized spec- 
trum scanner based on the cascaded tube 
has continued. The approach we are in- 
vestigating uses the persistence of the 
phosphor screens in the tube to store 
single photoelectron events until they can 
be detected with an image dissector. The 
phosphor screen is, of course, an inefficient 
storage element, as each scintillation is 
decaying exponentially. The problem, 
then, is to determine the optimum scan- 
ning rate to minimize the probability of 
missing photoelectron events. Much of 



the effort has been to develop fast linear- 
sweep amplifiers to drive the deflection 
yoke of the image dissector. This has 
been done by Mr. Scuderi and Mr. Doak. 
Experiments designed to measure the 
counting efficiency of the system are now 
in progress. 

Figure 3 illustrates the impressive in- 
crease in information content of image 
tube spectra of the night sky due to in- 
creasing excellence of the optics and 
photographic plates. The first spectrum 
(1965) is a 60-minute exposure using 
baked IIa-0 plates and taken with the 
69-inch Perkins telescope of Ohio State 
and Ohio Wesleyan Observatory, using a 
Super Faron spectrograph camera and an 
RCA C33011 image tube. The middle 
spectrum (1967) illustrates the increased 
resolution arising from the use of a 
Bowen f/2.25 schmidt camera; the spec- 
trum is a 180-minute exposure on a baked 
IIa-0 plate, taken with the 84-inch Kitt 

Peak telescope. The third spectrum 
(1973) shows the highest resolution; it 
was taken with an f/1.4 schmidt camera, 
an RCA 33063 image tube, and used a 
nitrogen-baked Illa-J plate at an expo- 
sure time of 150 minutes on the 84-inch 

We thank Drs. Babcock, Blanco, Gold- 
berg, Hall, and Heeschen for making 
telescope time available at CARSO, 
Cerro Tololo Inter-American Observa- 
tory, Kitt Peak National Observatory, 
Lowell Observatory, and National Radio 
Astronomy Observatory. 

We also wish to thank Drs. Goldberg 
and Blanco for assistance in travel ex- 
penses of observing trips to Kitt Peak 
and Cerro Tololo. 

The development of the high-gain 
image tube and the work on this digital 
system have been supported in large part 
by a grant from the National Science 

Fig. 3. Increasing resolution and gain of image-tube night sky spectra: (a) RCA C33011 cas- 
caded image tube and f/0.87 Super Faron spectrograph camera. Exposure 60 min on baked IIa-0 
plates, 69-inch Perkins telescope of Ohio Wesleyan and Ohio State Universities at Lowell Observ- 
atory; August 3, 1965. (b) RCA C33011 cascaded image tube and f/2.25 Bowen schmidt camera. 
Exposure 180 min on baked IIa-0 plates, 84-inch Kitt Peak telescope ; March 8, 1967. The in- 
creased resolution from spectrum a is due principally to improved spectrograph optics, (c) RCA 
C33063 cascaded image tube and f/1.4 Bowen schmidt camera. Exposure 150 min on nitrogen- 
baked Illa-J plates, 84-inch Kitt Peak telescope; March 5, 1973. Increased gain from spectrum b 
is due to higher gain of type C33063 tube. Increased resolution is due principally to use of slower, 
finer grained Illa-J plates. 



Radio Astronomy 

G. E. Assousa, E. T. Ecklund, C. A. Little, 
N. Thonrmrd, K. C. Turner, and M. A. Tuve 

Equipment Development 

E. T. Ecklund, C. A. Little, and N. Thonnard 

The old gray trailer, which for many 
years had housed our frequency analyz- 
ing spectrometer, has finally been decom- 
missioned. This was an occasion for us, 
as the trailer had served well in the early 
1960's to make our receiver portable 
when NRAO had no spectrometers. How- 
ever, the increased stability and flexibil- 
ity gained by having all of the electronics 
in a temperature-controlled environment 
has more than compensated for this loss. 
The move out of the trailer allowed us to 
make many modifications to enhance the 
stability and versatility of the multi- 
channel spectrometer. Figure 4 is a block 
diagram of the present spectrometer. 

The solid-state broad-band amplifier 
described in Year Book 71, p. 224, was 
completed this year. The new system has 
27 filter drivers arranged in three sets. 
Each driver is capable of driving at least 
six filters. Therefore, the two present sets 
of filters have independent drive systems, 
eliminating any interaction between filter 
sets and allowing the filter outputs to be 
set independently. The remaining set of 
drivers will soon be used in conjunction 
with the very narrow crystal filters now 
under construction. 

The wide-band filters discussed in 
Year Book 71 are now complete. To fully 
utilize the versatility of a multiple filter 
system" without transferring the cables 
and resetting all the gains, two special 
three-position, 60-pole switches were con- 
structed. One switch connects the R. F. 
output from the filter system in use to the 
R. F. detectors; the second switch selects 
one of three sets of 60 presettable gain 

30 MHz IF 

30 MHz 

&\ amp 

From control room (BW = IO) 

^ Mixer 

[25 MHzh>, 


J28MHz| ~Q 


Video amp 
(BW = 12) 

System A 


System B 

Filter drivers 
(9 in each system) 

System C 




System C 

Crystal oscillators 
(total of 6) 

System B 

System A 





and phase 
(total of 56) 

and gain 

M> «o6\- equalization *-(56)— detectors ^6V 
C,0r ' - V ^ 1 (total of 56) V 

(3 position, 
60 poles' 




[ 3 position, 
60 poles) 

Wide band 

total of 54 K 
spaced 75 



band filters 
(total of 56 
Spaced 19 kHz) 

> Electrometer 

Voltage to 

.to chart recorder 

to frequency counter 
and card punch 

NOTICE - . Encircled numbers indicate number of separate lines in this circuit 

Fig. 4. Block diagram of present frequency analyzing spectrometer at Derwood. 



controls, which are in the feedback loop 
of 60 IC preamplifiers. The presettable 
gain controls were initially adjusted so 
that the detected output level from each 
filter (both narrow-band and wide-band) 
was the same. This enables one to go 
from narrow-band to wide-band observa- 
tions by simply rotating two switches. 
Two additional advantages were obtained 
by using the IC preamplifiers. First, with 
the increased audio gain it was possible 
to reduce the R. F. level in the detectors 
and optimize it for the best power-law 
response. Second, by having higher levels 
in the phase detectors and on the inte- 
grating capacitors, the electrometer gain 
was reduced by a factor of 10, which has 
completely eliminated drifts due to capa- 
citor leakage and diode balance changes. 

This system has been in use for more 
than four months. Its stability has been 
much greater than our old tube system: 
Since the initial gain adjustment, no ad- 
ditional adjustments have been neces- 
sary; the old system required adjustment 
every few weeks. 

A new local oscillator system with bet- 
ter power and frequency stability and 
much wider frequency coverage has been 
completed for the hydrogen-line receiver. 
The system uses a solid-state phase- 
locked source that accepts a reference 
signal of about 106 MHz and delivers ap- 
proximately 220 mW at 1390 MHz. The 
reference source uses an oven-stabilized 
crystal oscillator and a temperature- 
compensated variable frequency oscilla- 
tor so that the local oscillator can be 
varied by ±10 MHz. 

Owing to the much wider bandwidth 
of the new filters, a new gain modulator 
has been designed and constructed. It 
uses recently introduced hybrid inte- 
grated circuit modules having 500-MHz 
bandwidths. We are using a voltage- 
controlled attenuator module and a 15-db 
amplifier module, together with an FET 
switch and IC amplifier, to synchronously 
modulate the system gain by as much as 

6 db. The linearity is better than we can 
measure over a frequency range wider 
than we anticipate using. The gain mod- 
ulator system has also been useful as a 
very flat, synchronously varying signal 
source to adjust the individual filter gains 
in the spectrometer. 

Last August we finally had to return 
the K-band mixer-preamplifier loaned to 
us by NRAO for our 1.2 to 1.5 cm receiver 
front-end. We now have our own mixer 
and IF amplifier, which we placed in 
operation in February 1973. Even though 
the IF amplifier noise figure is 2.0 db, 
the overall mixer-IF noise figure was 5.0 
db at 22 GHz. The overall single side- 
band system noise temperature during 
active observations has been about 
1400 °K, a tremendous improvement over 
the 2400 °K system temperature we were 
obtaining with the NRAO system. The 
IF bandwidth is 500 MHz, which will 
make this system also very useful for 
continuum observations. 

1.35-cm Water-Line Observations 

N . Thonnard 

Owing to the major equipment modi- 
fications and additions to our line spec- 
trometer and the installation of our new 
K-band mixer-IF system, observations of 
H 2 0-line emitters were not resumed until 
this February. It was decided to under- 
take an intense study of W-49, by far 
the brightest and most variable water- 
line emitter. So far, ten continuous weeks 
of observations have been made on this 
source. Most observations were made 
twice a week, and for one week the source 
was observed daily. All observations 
were first taken with the wide-band fil- 
ters and covered a velocity range of at 
least 300 km/sec; then the narrow-band 
filters were used to study the fine struc- 
ture of the central emission features of 
W-49. Some of the features are so nar- 
row that they should be studied with even 
finer resolution than our present 0.25 km/ 








280 - 





I I I I I I I I 

I I I I I I I I 

I I I I I I 

i i i r 




= 2.0°K 

i I , i , i i i i i i i i l i i i i i i i l i i i i i i i l i i i l i i 1 1 i 

-4 4 8 12 16 20 24 28 

FEBRUARY 23, 1973 


W-49 oC -- I9 h 07 m 50 s 8= 9°l.r095O) 

W°- 8 °K 

*&m,* * S^—m+<m*IH——m**w**<m»»** >' » ■*■ 

MARCH 23, 1973 

.■ ^ ' ^/ ^■■ ^ ■llWfl n >i 

J_l I I I I L 

-220 -200 -180 -160 -140 -120 -100 -80 -60 -40 -20 20 40 60 



i l l l I L 

100 120 140 160 180 

Fig. 5. H 2 line emission from W-49 on February 23, 1973, and March 23, 1973. 

sec. In a few instances, the velocity cov- 
erage was from — 450 km/sec to +400 

A typical illustration of water-vapor 
emission from W-49 on two occasions is 
given in Fig. 5. These data were taken 
with the wide-band filter. An expanded 
view of the velocity spectrum taken with 
the narrow-band filter is shown for a 
small section to illustrate the complexity 
of the emission features. Analysis of 
these data and further observations are 
now in progress. It is hoped that by care- 
ful monitoring of these features it will be 
possible to understand the nature of the 
large variation and, in particular, the 
mechanisms giving rise to the large ve- 
locity spread in the observed emission 

Search for 21 -cm Hydrogen Absorption 
during Flare-ups of Cygnus X-3 

G. E. Assousa, N. Thonnard, and M. A. Tuve 
A sudden rise of three orders of mag- 
nitude in the intensity of the radio source 
associated with the x-ray source Cygnus 
X-3 was first noted by Gregory et al. 10 on 
September 2, 1972, at the 46-m radio tele- 
scope of the Algonquin Radio Observa- 
tory. On August 31, 1972, measurements 
taken at NRAO indicated intensities of 
0.015 f.u. at 3.7 cm and 0.016 f.u. at 11.1 
cm (1 f.u. = 10" 26 Wm" 2 Hz" 1 ). The in- 
tensity on September 2-3 rose sharply to 
18.4 f.u. and 9.4 f.u., respectively. B. 
Balick 11 of NRAO notified us of this 
event and asked us to look for 21-cm 
hydrogen-line absorption in the direction 
of the source. 



Unfortunately, our Derwood receiver 
was undergoing major modification when 
the Cygnus X-3 flare-up occurred. After 
very hurried and makeshift reconnect- 
ing of the receiver, observations were 
started on the evening of September 5, 
1972. By that time, the intensities ob- 
served by NRAO had dropped to 5.0 f.u. 
at 3.7 cm and 7.3 f.u. at 11 cm. Observa- 
tions were also in progress during the 
second flare-up of Cygnus X-3 on Septem- 
ber 20. No absorption could be detected 
from our data. French observers at the 
Observatoire de Paris, Meudon (Lauque 
et al., 1972) 12 were successful in detect- 
ing absorption of the source by the inter- 
vening neutral hydrogen at 21 cm and so 
were able to place Cygnus X-3 in our 
Galaxy at a distance between 8 and 11 

Our failure in observing absorption in 
the direction of the source is not too sur- 
prising. At 21 cm our Derwood dish has 
a beamwidth of 50' and hence a sensi- 
tivity of only 0.05°K per f.u. This would 
give an absorption of at most 0.5°K. 
Two methods were used to search for 
absorption. First, data were taken on the 
source, and at four points one-half beam- 
width and four points one beamwidth 
from the source. An attempt was made to 
construct a background hydrogen emis- 
sion for the region by interpolating these 
data, but, owing to the large beamwidth 
and the complexity of the Cygnus region, 
no absorption could be detected. An at- 
tempt was also made to compare data 
taken at the Cygnus X-3 position on days 
when the source was intense with data 
taken when the source was weak. This 
required reproducibility of the receiver 
gain of better than 0.5%. Our typical 
results of 3-5% rendered this method 

Survey of Neutral Hydrogen in the 
Vicinity of 17 Galactic Supernova 

G. E. Assousa, B. Balick, and J. W. Erkes 

Recent 21 -cm observations in the vicin- 
ity of the supernova remnant HB21 (As- 

sousa and Erkes, Year Book 71, p. 240) 
showed evidence of an expanding con- 
centric shell. At the best distance esti- 
mate for HB21, the accreted shell of 
interstellar material has a calculated 
radius of 26 pc and a total mass of 6300 
M . Based on the HB21 results, an ex- 
tensive survey of the neutral hydrogen 
near 17 large galactic supernova rem- 
nants was completed at the NRAO.* 
This included CTA1, HB3, CTB13, HB9, 
VRO 42 05 01, OA 184, S147, P 0607+17, 
W41, 3C3961, W50, CTB 72, CTB 63, 
W56, Cygnus Loop, W63, and CTB1. 
Observations were made at beamwidth 
intervals (0.33°) over and in the imme- 
diate vicinity of each remnant, and at 
0.5° or 0.66° intervals elsewhere. Pre- 
liminary longitude versus velocity maps 
and latitude versus velocity maps for all 
17 remnants have been completed. 

There is evidence for the presence of 
anomalous velocity neutral hydrogen 
near the Cygnus Loop (Fig. 6) and S147 
(Fig. 7). The contour levels in both fig- 
ures are those of antenna temperature in 
°K. The center of the Cygnus Loop is at 
l u = 74.0° and b„ = —8.6°. In Fig. 6 
there is a long feature between — 12° 
< frn < —7° which may well be the 
remnant-associated hydrogen. Bright 
optical filaments in the Loop are known 
to have velocities of up to 115 km/s 
(Minkowski, 1958) , 13 Here, the veloc- 
ities observed extend to roughly the same 
value. In S147 (Fig. 7), a velocity ex- 
tension to — 53 km/s appears at the op- 
tical center, l u = 179.6°, 6„ = —2°; no 
filament velocities are available. In both 
places there is a strong hint that the neu- 
tral hydrogen is associated with the rem- 

Integrated hydrogen maps over fixed 
velocity intervals give the best evidence 
for any shell structure (Assousa and 
Erkes, Year Book 71, p. 240) . Such maps 

* The National Radio Astronomy Observa- 
tory is operated by the Associated Universities, 
Inc., under contract with the National Science 













. (-■ ro 

^ o 

. t» 

^-v CO 

CO °3 

I o3 

,, P 



CO X! 
CO =3 

,, o 

w o 

O xn 

O -- 1 

J 8 

in P 

Q o 

0) i— I 

-P I 

CD W) 

-JS p 

o P 

CD +s 

M X 



CD h 

P -M 

p CD 

O «+H 

'-M CD 

22 P 

P 13 

oj ^ 

2 » 

,P bC 

p— i ^ 

d T3 

03 ao 


-P d 

H > 

fab B 

w s-. 









































are now being prepared for the survey. 
Detailed analysis of the Cygnus Loop, 
S147, and the other remnants must thus 
be deferred until all integrated hydrogen 
maps are available. 

Accurate Radio Positions for Ohio Radio 

Sources, Using the NRAO Three-Element 

Interferometer at 3.7 cm and 11.1 cm 

J. W. Warner, G. E. Assousa, and B. Balick 

Recent studies for the identification of 
the optical counterparts of radio sources 
from the 1415-MHz Ohio survey have 
led to the discovery of unusual astro- 
nomical objects. In several instances 
identifications have been made with op- 
tically distinctive sources which show 
high redshifts of variable luminosity. 
The radio source OH471, recently ob- 
served by Carswell and Strittmatter 


(1973) 14 at Steward Observatory to have 
the largest known redshift, Z = 3.4 


1 + V/C 

[1_ ( y/c)2]i/2 


is a result of such comparative studies. 

Statistical studies of positions of Ohio 
sources, in conjunction with lists of com- 
pact galaxies and "blue" stellar objects, 
have shown a significant number of these 
objects to have radio counterparts. Radio 
source positions, accurate to several sec- 
onds of arc, are essential for positive 
identifications. We have used the 
NRAO three-element radio interferom- 
eter at 3.7 cm and 11.1 cm to obtain 
accurate positions and fluxes at both 
wavelengths. Sources we have already 
observed appear in Table 1. Positive 
identification of optical counterparts is in 

TABLE 1. List of Sources Observed with the NRAO Three-Element Interferometer 









































































Instituto Argentino de Radioastronomia 

E. Arnal, R. Colomb, R. Dugatkin, E. Filloy, 

S. Garzoli, M. Gil, D. Goniadzki, M. Gordon, 

I. Mirabel, R. Morras, W. Poppel, 

R. Quiroga, and K. C. Turner 

This organization, founded to utilize 
the 100-foot radio-telescope and receivers 
provided by the Carnegie Institution, 
continued during the year under the 
temporary direction of Dr. K. C. Turner 
of our Department. As in the past, the 
major emphasis has remained on studies 
of galactic hydrogen distribution. 

Equipment development. The new sur- 
face for our existing 30-meter-diameter 
telescope was completed in September. 
It consists of a solid steel surface from 
the center to a radius of 10 meters and a 
perforated steel surface on the remainder. 
The antenna efficiency at 21 cm is now 
essentially equal to that obtained before 
the hailstorm of last year. 

The new parametric amplifier front 
end, utilizing one of the four paramps 
kindly donated by S. Colgate of the New 
Mexico Institute of Mining and Tech- 
nology, has been operated for several 

5 Km/sec 

-5 Km/sec 
• • • 

$- 35 Km /sec 

25 Km/sec 

20 2 

Scale: 1 =10 at/cm 

Fig. 8. Integrated brightness temperature for Vr > 30 km /sec. The contour units are 10 20 
atoms/cm 2 . The arrow indicates the direction of the galactic center. Small circles represent the 
points observed. The dashed lines are lines of constant radial velocity, corrected for the effect of 
galactic rotation. 



months and is proving much more stable 
and reliable than our old system. The 
measured system noise temperature is 
now 190°K. 

The mount and carriage for the second 
telescope are now finished, and the parab- 
oloid is ready to be mounted. This will 
be done by three large cranes as soon as 
the ground is dry enough to support their 
weight; it should be completed by the 
time this report is published. 

The bank of thirty 100-kHz-wide fil- 
ters has been modified to eliminate a 
serious problem of spurious responses 
well away from the band centers. 

21 -cm neutral hydrogen observation. 
K. C. Turner and I. F. Mirabel have been 
studying an object with peculiar prop- 
erties illustrated in Fig. 8. Its elongation 
suggests the possibility that it was 
ejected from the galactic center, whose 
direction is indicated by the solid arrow. 
However, the intrinsic velocity field of 
the object (indicated by the dashed lines) 
is that which would be expected if the 
object were moving parallel to the galac- 
tic plane in the direction toward the cen- 
ter. Analysis of this material is not yet 

W. Poppel, working with E. R. Vieira 
of the Institute de Fisica of the Universi- 
dade Federal do Rio Grande do Sul, has 
completed the data reduction of their 
very large survey of the region 248° < I < 
12°, 3° < b < 17°. It is hoped that pub- 
lication of this material will begin shortly. 

These investigators have also com- 
pleted an extremely accurate study of a 
set of calibration points for H I observa- 
tions, and have observed, as well, the 
neutral hydrogen distribution near NGC 
2287, NGC 6231, and 6 Carinae. 

S. Garzoli and I. F. Mirabel are cur- 
rently studying H I structures in the outer 
parts of the galaxy between 288° < I < 
310°, —7° < b < 2°. 

I. F. Mirabel and W. Poppel are in- 
vestigating an anomalous velocity dis- 
covered by Cugnon near I = 349°, 6 = 3°. 

F. R. Colomb and D. Goniadzki are 
investigating the neutral hydrogen dis- 

tribution near the supernova remnants 
G 261.9 + 5.5, G 296.3 + 10, G 327.6 + 
14.5, and G 330 + 15. 
F. R. Colomb, M. Gil, and R. Morras 

are observing an extensive region near 
the south celestial pole. 

R. Morras is beginning a program to 
study small-scale angular structure of 
neutral hydrogen in the galaxy using the 
moon as an occulting screen. 

E. Arnal has begun a study of the dis- 
tribution of neutral hydrogen near very 
early type stars. 

F. Strauss and Z. Strauss of the Insti- 
tute de Fisica of the Universidade do Rio 
Grande do Sul visited to search for possi- 
ble "fossil" supernova remnants near 

S. Garzoli has just published her Atlas 
of Galactic Neutral Hydrogen for the 
Region 270° < I < 310°, —7° < b < 2° 
as a Carnegie Institution of Washington 
monograph (Publication No. 629) . 

Atomic Physics 

L. Brown, G. E. Assousa, and 
H. A. Van Rinsvelt 

In Year Book 53, p. 67, Norman Hey- 
denburg and Georges Temmer described 
their pioneer work in studying the low- 
lying excited states of nuclei through 
Coulomb excitation with a-particles. 
These investigations included measure- 
ments of the excitation of the K x-ray 
lines of the target atoms as well as the 
y-rays from its nucleus, since their scin- 
tillation counters recorded both kinds of 
photons for the heavy elements. These 
data are the first in a research field that 
has grown very active in the past few 
years, the study of atoms by bombard- 
ment with heavy ions. The theory of 
Coulomb excitation has proved capable 
of describing not only a wide range of 
nuclear excitations but atomic excitations 
as well. However, a new phenomenon, 
called electron promotion, is known to 
exist; its theory is as yet only qualitative. 

Coulomb excitation assumes a simple 
interaction of two point charges. It can 


be calculated in a straightforward man- work of others who had observed shifts 

ner with quantum mechanics, the theory in photon energies when targets were 

being applicable to atoms or nuclei. For bombarded with particles of tens of MeV; 

the atoms ionized by heavy ion bombard- it thus seemed reasonable to follow this 

ment, the original theory must be modi- to the lowest energies capable of exciting 

fied to account for the presence of the the lines. The shifts result from a high 

heavy ion within the K shell of the target probability of ionizing the L shell when 

atom. Typically, it remains there for the K shell is ionized. Both the energy 

periods that are long compared with the shifts and the cross sections for ionizing 

oscillation times of the K electrons. With the K shell from the experiment can be 

this binding energy correction, Coulomb divided into two groups: effects resulting 

excitation accounts accurately for a wide from Coulomb excitation, and effects 

range of excitation phenomena, provided from electron promotion, 

that the incident particle has no electrons Figure 9A shows the spectrum of the K 

of its own or, if it does, that they do not lines of copper excited by bombardment 

overlap the region occupied by the K with 1.75 MeV Rb + ions. It was observed 

shell of the target. If the electron shells with a lithium drifted silicon detector, 

do overlap, the electrons try to accom- which is a solid-state device made of a 

modate themselves to the structure of a single crystal of silicon with lithium dif- 

new hyperatom, having as its nucleus the fused through certain portions. It func- 

combined nuclei of projectile and target, tions as would an ionization counter of 

Thus a new K shell forms, and two of the high density and low energy loss per ion 

four electrons, which made up the two pair. The open circles are from the rubi- 

independent K shells, are forced out or dium beam; the smooth curves are from 

"promoted." When the two nuclei sepa- copper fluorescence radiation, multiplied 

rate later, there is a high probability that by 0.1, which was used to locate the un- 

each will have a K-shell vacancy. The shifted lines. The intensity is plotted as 

same kind of arguments can be applied to the ordinate, the photon energy as the 

the L shells. Electron promotion gives a abscissa. The K a line, shown at the left, 

very high probability of ionization, hence is made up of two unresolved lines of 

a very high cross section for x-ray pro- energies 8027.8 and 8047.8 eV, and the 

duction. K/3 line, shown at the right, has an energy 

During the past summer we returned of 8902.9 eV. The resolution of the de- 

to the study of the excitation of x rays tector is 220 eV. The shifts in energies 

by heavy ion bombardment. We used the of the lines excited by rubidium are 

beams of alkali ions, used previously in apparent. 

the study of the lifetimes of optical The energy shifts proved to be almost 

states, because they offered a wide range independent of particle energy, as shown 

of atomic numbers. Our first bombard- by the results plotted in Fig. 9B. Lithium, 

ments were the K lines of copper. the incident ion that had electrons with 

The experiment proposed to answer little overlap of the L shell of copper, 

two questions: (1) Are the characteristic produced almost no energy shifts. The 

lines of copper shifted in energy when other alkalis, which have overlap of L 

excited by the heavy ion beams of a few shells, had shifts that were close to what 

MeV, and, if so, how do the shifts vary one expects theoretically from a single 

with the kind of ion and the energy? (2) vacancy produced in the L shell. This 

How do the cross sections for the lines, high probability easily follows from the 

as observed by us through the yields, de- idea of electron promotion, 

pend on the kind of ion and the energy? The yields of the K x rays can be cal- 

The first question was suggested by the culated for Coulomb excitation, so 





Channel no 


Fig. 9 A. Spectrum of the K lines of copper observed with a lithium drifted silicon detector. 
The open circles are from bombardment with a Rb + beam of 1.75 MeV; the smooth curves are 
from copper fluorescence radiation, multiplied by 0.1, which was used to locate the unshifted posi- 
tion of the centroids. 










_ Li 




o o 

N 1 


„• o o o a 





" Na 



■ 1 




■ a 





> ♦ 


" K 



-1 i 





} I M 



~ Rb 



} M ♦ 

♦ < 

f v 



I^ S 


> 6 




♦ ♦ ♦ ' 


2 E (MeV) 4 

Fig. 9B. Shifts in the energies of the copper K a and Kg lines in eV as functions of incident ion 
energy in MeV. Data on nitrogen and cesium were taken at only one energy and are plotted with 
the results for lithium and rubidium, respectively. The horizontal lines shown for sodium, potas- 
sium, and rubidium are calculated values of the shifts produced by a single hole in the 2p shell of 



Fig. 9C. Thick target yields of the K lines of 
copper as a function of atomic number of the 
incident ion at 1.6 MeV. The smooth curve is 
the prediction of Coulomb excitation with bind- 
ing energy correction and shows the enormous 
increase for the ions heavier than sodium. 

comparison with experiment is possible. 
Figure 9C shows the measured values for 
the thick target yields for incident par- 
ticles of 1.6 MeV plotted as a function of 
the atomic number of the incident ion. 
At the left is shown a smooth curve that 
connects values calculated for Coulomb 
excitation with a binding energy correc- 
tion. The agreement is good for lithium 
and sodium but misses by many orders 
of magnitude for potassium, rubidium, 
and cesium. The high probability of ion- 
ization is evident if the K shells of the 
colliding atoms overlap. 

Nuclear Physics 

L. Brown and U. Rohrer 

Improvements in the Atomic 
Physics Observatory 

The Van de Graaff accelerator — in 
spite of its 35 years, which make it the 
second oldest operating atom smasher in 
the world — continues to be a most useful 

instrument for observing nuclear and 
atomic phenomena. The heroic size of its 
high-voltage terminal allows the use of 
the polarized ion source of the University 
of Basel, which gives the machine capa- 
bilities still unmatched. During the past 
year, two changes increased the efficiency 
of this instrument: the accelerator tube 
was replaced, and the pulse height ana- 
lyzer was connected to a magnetic tape 

During the past few years the vacuum 
tube of the Van de Graaff has cost users 
increasing amounts of time for locating 
and sealing leaks. During the last two 
years alone 150 detectable leaks occurred. 
When a decision was made a year ago to 
replace the tube, a new design, which 
had the same electrode geometry as the 
previous tube Norman Heydenburg built 
and several elements common to the tube 
Merle Tuve built for the 1934 machine, 
evolved from studies with Joseph Peoples 
of Potentials, Inc., Austin, Texas, who 
has considerable experience in bonding 
accelerator tubes. The new tube is made 
up of seven identical sections, each made 
of cylinders of Pyrex bonded to the alloy 
Kovar with vinyl acetate. The machine 
is subjected to large changes of tempera- 
ture, which caused leaks in the old tube. 
The near identity of the thermal expan- 
sivities of Pyrex and Kovar should pre- 
vent such leaks. The aluminum elec- 
trodes on the interior of the tube, which 
shape the electrostatic field for accelerat- 
ing the ions, are connected to the high- 
voltage machine through the Kovar. The 
new tube is not strong enough to be self 
supporting, as were its predecessors, the 
tubes of 1938 and 1955, so it is connected 
to the main structure at eight points 
through an assembly designed to isolate 
it from vibration and shock. Selwyn 
Sacks aided in this part of the design by 
measuring the amount of the vibration 
present with a seismometer. 

The new tube first operated in Decem- 
ber 1972 and has surpassed the design 
goals. No leaks have opened after four 



months of operation during a period when 
changes of temperature were severe. Fur- 
thermore, the new tube delivers a much 
steadier and more intense beam than the 
old one. This was not expected but is 
very welcome. It is evidently the result 
of better electrical connection between 
the electrodes in the vacuum and the high 
voltage machine. 

In most experiments with the APO the 
data are acquired in a pulse-height ana- 
lyzer. At the end of an experimental run, 
they are read out of the memory of the 
analyzer for reduction and interpretation. 
In the past this has been done with a line 
printer, and therefore part of the data 
reduction had to follow by hand. This 
was reasonable for the simple spectra of 
the charged particle reactions observed in 
the past. It proved onerous in the x-ray 
work reported elsewhere in this report 
and would have been increasingly so in 
the neutron work just started. An oppor- 
tune purchase by the Department of five 
digital tape recorders at bargain prices 
took place in happy coincidence with 
these new requirements. One of these 
new recorders was interfaced to the pulse- 
height analyzer, so all data reduction now 
takes place with the computer. For the 
experiments under way, this removes 
much tedious labor and improves the 
reliability of the extracted data by allow- 
ing many mathematical operations to test 
the measurements that otherwise would 
not have been made. 


Much of our effort in nuclear physics 
has applied phase-shift analysis to meas- 
urements of elastic scattering made with 
the polarized beam. Last year we re- 
ported the results of our studies of 
7 Li(p,p) 7 Li, our first attempt with a 
spin-1/2 on spin-3/2 interaction which 
gave information about the structure of 
8 Be. A similar study of 9 Be(p,p) 9 Be, 
an identical spin system, has led us to 
results about 10 B. This isotope, consist- 

ing of five protons and five neutrons, has 
both of its Is shells filled with two pro- 
tons in one shell and two neutrons in the 
other, but has only three nucleons in each 
of the lp shells (protons and neutrons fill 
separate shells) . The nuclear shell model, 
our best picture of nuclear structure to 
date, works best when the nucleons fill a 
shell completely, when there is one nu- 
cleon in excess, or when there is a hole in 
the shell. Boron-10, with its lp shells 
only 3/8 full, is difficult to handle but 
offers a relatively simple case for theorists 
to attempt an explanation of intermedi- 
ate structure. Accurate identification of 
the angular momentum and parity of the 
excited states is required, of course. 

The five bound states of 10 B appear to 
have been completely mapped: Their 
energies are accurately determined and 
the quantum number assignments are 
not in dispute. Even the seven states 
between the thresholds for cx-par- 
ticle (4.461 MeV) and proton (6.587 
MeV) emission have but three uncertain 
spin-parity assignments. This harmony 
disappears above the proton threshold, 
however, where the states encountered 
are generally broad, overlapping, and of 
impure iso-spin (refer to Fig. 10) . These 
conditions lead to data that are difficult 
to interpret, and there is as yet no agree- 
ment about the number and location of 
states, much less about their angular mo- 
mentum and parity. 

We have measured the angular dis- 
tribution of the polarization analyzing 
power at 20 energies and have subjected 
them, together with the previously meas- 
ured cross-section data taken at the Cali- 
fornia Institute of Technology and the 
University of Maryland, to a phase-shift 
analysis. Our first attempts to fit the 
data naturally employed the structure 
proposed on the basis of cross-section 
measurements and (p,y) reactions. The 
result was a failure, the discrepancies be- 
ing too great to be resolved by small 
modification of the resonating phase 
shifts. Eventually, the set of phase shifts 





0.038 & 0.085 

2 + 8 3 + 



(It2 + ) 






o+ i 

7.47 a o.n 

l + 82~ 


(l + ) 




(2 + ) 


- 3 




9 B + n 



8 Be+p+n 



- 2 

- I 

'Be + p 



Fig. 10. Limited energy level diagram of 10 B. The only levels shown lie above or near the pro- 
ton threshold in the energy range spanned by this experiment. E x denotes the excitation energy, 
which is the energy of the state above the ground state ; r denotes the width of the state ; J 17 , the 
angular momentum and parity; and T, the iso-spin. At the right is shown the information neces- 
sary to relate the proton energy, E p , to the energies of the excited states; three particle disinte- 
gration thresholds are shown. 

shown in Fig. 11 developed. The exist- 
ence of an excited state is characterized 
by the phase shift of a particular partial 
wave resonating, which in most cases 
shows an increase of about 180°. The 
structure that satisfies the data is readily 
apparent from Fig. 11: there are two 
levels at proton energy 0.930 MeV (7.47 

MeV excitation energy) : a 1 + (designated 
5 Pi in the figure) of width 0.10 MeV and 
a 2- (designated 5 S 2 ) of width 0.11 MeV; 
a third level at 1.37 MeV (7.82 MeV ex- 
citation energy) is 1" (designated 3 Si) of 
width 0.30 MeV. This structure is simpler 
than what has generally been supposed 
necessary to satisfy the cross-section 








'S, - 

-8ooo °ooOo? OOo o_ 

o o o„ 

j0 -^o-o-°-o- 







Fig. 11. Phase shifts as functions of energy. The curves show the phase shifts that were ob- 
tained as the best fits to the experimental data. Closed circles denote energies at which there are 
cross-section and polarization data; open circles indicate cross-section data alone. 

data. The key to the matter came from 
polarization, which forced a change of 
spin assignment of one of the levels at 
0.98 MeV and of the one at 1.37 MeV. 

Reference to Fig. 10 discloses a + state 
at 1.086 MeV (7.56 MeV excitation en- 
ergy) that is not seen in Fig. 11. This 
state is very narrow, about 3 keV, and 
shows up strongly in elastic scattering. 
It is narrower than the energy resolution 
of our experimental set-up, which pre- 
cluded our observing it with the polarized 

beam; more to the point, we carefully 
avoided it so that its effects did not enter 
into our measurements. We did look into 
the spin-parity assignment of this state, 
since we could bring a complete phase- 
shift analysis to bear on the cross- 
section data. The old assignment of + 
withstood all our efforts to change it. 

One aspect of the analysis of 7 Li(p,p) 7 Li 
was the demonstration of the need of 
channel spin mixing in order to fit the 
measurements for the two 1 + states en- 



countered (see Year Book 71, p. 247). 
The same amount of mixing was also 
needed for the 1 + state at 0.98 MeV in 
9 Be(p,p) 9 Be. Information on channel 
spin mixing is difficult to obtain by other 
methods. We have now seen three cases 
of 1 + states requiring the same amount of 
energy independent mixing. The theo- 
retical implications of this are yet to 

Our measurements extended beyond 
the energy limits shown in Fig. 11, but 
the energy region from 1.7 to 2.8 MeV 
had usable cross-section measurements at 
only one angle. This proved to be a seri- 
ous handicap. The polarization data 
alone were insufficient to determine the 
phases; in particular, the two s-phases 
could not be determined, nor could the 
effects of the 5 Pi and 5 P 2 phases be sepa- 
rated. Two levels in this region attracted 
our interest (Year Book 71 , p. 250) orig- 
inally because they are both at the same 
energy, E p = 2.56 MeV. Although we 
could not determine a satisfactory set of 
phases, we were able to test the assign- 
ments, which had surprised us at first, of 
the two levels degenerate in energy to 
within 1 keV. The assignments seem 
quite reasonable as seen through the 
polarization measurements. The two 
levels at 0.98 MeV are also degenerate 
in energy, so our surprise at this phenom- 
enon has been diluted. 

A Study of (p,n) Reactions 

The nuclear physicist, probably more 
than any other kind of scientist, is locked 
to the experiments that can be done with 
a given set of equipment. It is his lot to 

find that the pursuit of an interesting and 
potentially fruitful aspect of his work 
must terminate because further means of 
observation have not been developed. 
The converse requires him to study nu- 
clear interactions because their observa- 
tion is virtually a consequence of his 
equipment. Our interest in the two over- 
lapping 3 + states of 8 Be (see Year Book 
71, p. 247) led to building equipment for 
studying them through the 7 Li(p,n) 7 Be 
reaction. Two pairs of neutron detectors 
are now functioning. One pair uses recoil 
protons to produce pulses in scintillation 
counters that use fast electronics to dis- 
tinguish the recoil protons from fast elec- 
trons produced by y rays, an old and 
serious problem of neutron detection. 
This method allows only about 1% of the 
y-ray photons to be counted as neutrons. 
Another pair of detectors uses the 6 Li 
(n,t) 4 He reaction to produce the needed 
scintillations. This technique is used for 
neutron energies too low for the proton 
recoil method. It is, fortunately, insensi- 
tive to y rays but presents some shielding 
problems against background neutrons. 

Thus our interest in the structure of 
8 Be caused us to build equipment for ob- 
serving neutrons. It is obvious that we 
should observe the polarization effects in 
two other reactions that fit almost per- 
fectly into this experimental setup: 
9 Be(p,n) 9 B and 3 H(p,n) 3 He. The former 
reaction may well furnish information 
about the structure of 10 B, although just 
what information is not immediately evi- 
dent. The latter reaction is a study of the 
first excited states of the a-particle, a 
subject of fundamental importance. 


E. T. Bolton, T. I. Bonner, D. B. Cowie, M. B. Davis, B. H. Hoyer, N. R. Rice. 
R. B. Roberts, and D. G. Wallace 


A very broad range of experimental 
problems has been pursued by members 
of the Biophysics Section in the past 

year, and real progress has been made in 
each area. Immunochemical studies of 
enzyme evolution in bacteria have been 
brought to the Section. This program, 
initiated last year at the Institut Pas- 



teur, Paris, has continued, and a col- 
laborative program, both here and in 
Paris, has been projected. Studies of 
brain biochemistry have also been vigor- 
ously pursued, yielding new and exciting 
findings of a chemically induced en- 
hanced rate of learning in mice. Investi- 
gations concerning nucleic acids, long our 
dominant research interest, have centered 
on satellite and nonrepeated DNAs in 
rodents and on the nature of DNA that 
appears to reassociate instantaneously. 
In addition, we are pleased to be able to 
report technical advances in the produc- 
tion of DNA fragments of defined piece 
size as well as the in vitro labeling of 
DNA with 125 I. Two new postdoctoral 
fellows and one graduate student joined 
us late in 1972 and early 1973, bringing 
backgrounds in nuclear physics, protein 
evolution, and RNA synthesis. Their re- 
search is now well under way. 

Studies of Memory 

R.B. Roberts 

Peptides are synthesized in the brain 
and are known to act in releasing hor- 
mones from the pituitary gland. In ad- 
dition, there are several indications that 
peptides may play some role in learning 
and memory. Lande, Witter, and de 
Wied 15 have shown that hypophysec- 
tomized rats cannot learn the conditioned 
avoidance response in the shuttle-box un- 
less they are injected with one of a num- 
ber of peptides related to vasopressin. 
We have observed that large injections 
of crude ACTH improve the rate of 
learning of normal mice in the shuttle- 
box. De Wied has reported large effects 
of small quantities of ACTH and vaso- 
pressin on extinction of an acquired re- 
sponse. Flexner found that injections of 
vasopressin conferred an immunity to the 
amnesia caused by injection of puro- 
mycin. 16 This amnesia (caused by puro- 
mycin) is probably due to puromycin 
peptides which remain in the brain for 
long periods of time. Finally, in the ex- 
periments which indicate a transfer of an 

acquired behavior pattern after injections 
of extracts of the brains of trained ani- 
mals, the active material seems to be a 

With these thoughts concerning a pos- 
sible role of peptides in memory forma- 
tion in mind, we were intrigued by read- 
ing that N-acetyl aspartic acid (NAA), 
a compound found only in the brain, can 
cause a stimulation of peptide synthesis 
in brain slices. It seemed worthwhile to 
determine whether this compound would 
have any effect on learning and memory. 

Mice were given ten training runs per 
day in the shuttle-box and the training 
was continued for five days. Control 
groups were injected with water or not 
injected and the experimental animals 
were injected (i.p.) with NAA immedi- 
ately after training each day. 

On the first two days the mice learned 
to escape the shock which was delivered 
after 2% seconds of a warning tone. On 
the third day and thereafter the mice 
scored an increasing number of success- 
ful avoidances of the shock by running 
at the sound of the tone. The perform- 
ance of the individual mouse was meas- 
ured by the sum of the avoidances on 
days 3, 4, and 5. 

The dramatic and consistent improve- 
ment caused by NAA injections is shown 
in Table 2. This table includes tests on 
ten different sets of DBA mice (different 
ages, different sex, some purchased, some 
home grown) and 17 groups of control 
animals. In all cases, the performance 
of the NAA-treated mice was better, 
usually by a factor of 2. The improved 
performance persisted when the mice 
were tested again a week after the con- 
clusion of the training runs. Thus, the 
effect seems to be on the consolidation 
of the memory of the training experi- 
ence and is not simply due to improved 
performance in the presence of the drug. 

Dosage experiments indicate the NAA 
is effective down to a level of 10 mg/kg. 
Three mice injected with that dosage 
after each day's training showed an aver- 
age score of 10.3, compared to 5.6 for the 



TABLE 2. Performance of Mice after Treatment with NAA 


Size of 

Average Scores of Groups 








13.7 12.7 









7.0 2.33 4.67 
4.25 4.5 5.5 

Percent Correct Responses 




day 3 



day 4 

day 5 


4.33 10.0 

The upper part of the table shows the average scores of ten groups of NAA-treated animals and 
seventeen groups of control animals; all were injected immediately after training on days 1-5. The 
lower part shows the day-by-day average performance of 37 NAA-treated and 61 control animals. 
All were DBA strain, but differed in age, sex, and source. 

control animals. Animals given daily in- 
jections of 200 mg/kg of NAA immedi- 
ately after training showed marked im- 
provement in performance. Four groups 
of 4 mice each, had average scores of 
10.0, 10.8, 11.7, and 12.7 again compared 
to 5.6 for the controls. A group of 3 
animals injected at the 200 mg/kg level 
showed an average score of 13.7, and 
a group of 5 mice averaged 9.4. The 
normal level of NAA in the brain is re- 
ported to be 5.4 milf; thus a dose of 200 
mg/kg is sufficient to triple the normal 
level in the brain even if it is distributed 
uniformly through the animal. If the 
compound is concentrated exclusively in 
the brain (approximately % of the total 
weight), then 5 mg/kg would suffice to 
triple the concentration in the brain. The 
compound is toxic when given daily at 
400 mg/kg. 

A few experiments were done using 
other related compounds in place of, or 
in addition to, the NAA. Each drug was 
injected for five days immediately after 
training runs, at a dosage equimolar to 
NAA at 200 mg/kg. N-acetyl alanine 
appeared to have similar effects in two 
tests. After its injection in a group of 
3 mice, an average score of 9.3 was ob- 
served, compared to 5.0 for a control 
group; in another group of 3, the average 

score was 9.2, while controls averaged 4.2. 
N-acetyl glutamic acid injections did not 
produce an increase over control group 
scores (average = 4.6 in a group of 3 
animals). Aspartic acid gave ambiguous 
results — increasing average scores to 9.2 
in one group of 4 mice, while not effect- 
ing any improvement in another group 
of 3, which averaged 4.3. L-dopa may 
counteract the effect of NAA ; it reduced 
the average score from 11.7 to 5.5. The 
statistical probability of a score as low 
as 5.5 in a group of four is only 15/1000. 
In some experiments the NAA was in- 
jected at different times before training 
sessions. The interval varied from 3 
hours to 3 days. Injection before the 
training period seems effective even when 
a single dose is given 3 days in advance. 
For two groups of 3 mice each, scores of 
10.7 and 7.3 were observed after such 
an injection. When 3 mice were injected 
on three successive days before the start 
of training, their average score was 8.3. 
For the 4 animals injected 3 hours before 
each training session, the average score 
was 10.4. As these tests were carried out 
with 200 mg/kg (20 times the minimum 
effective dose) and the half-life is re- 
ported to be 12 hours, it is not surprising 
that its effect could persist. 



Some work was also done on the effect 
of NAA injected after the completion of 
the usual training period. When injected 
on days 5, 6, and 7, it seems to have no 
effect on the performance a week after 
the last injection. When injected on days 
11, 12, and 13, however, it has little effect 
on the performance on day 14, but does 
seem to improve the performance of day 
15. This is an indication that NAA af- 
fects not performance but the consolida- 
tion of the additional training of day 14. 

NAA was tested in several other learn- 
ing situations. Dr. Rake found no effect 
of NAA on the time to step into a cham- 
ber where the mouse had previously 
received a shock. Dr. Flexner observed 
that treatment with NAA did not confer 
immunity to the amnesia caused by puro- 
mycin. We found no improvement in 
maze learning in a small number of tests. 
Thus, NAA seems quite limited in its 
effects on various types of learning. 

However, the effect on shuttle-box 
learning is large and reproducible, so it 
offers an opportunity to compare a 
change in behavior with the concurrent 
changes in biochemistry. At present, this 
effort is in its initial stages. No change 
was found in the level of norepinephrine 
in the brain when NAA was injected 
either 1 or 4 hours before the measure- 
ment. Amino acid incorporation into 
brain and liver does, however, seem to 
be altered, but further checks are needed 
before any definite conclusions can be 

The Mouse Satellite in Two 
Subspecies of Mus musculus 

Nancy Rice and Patricia Esposito 

About ten percent of the DNA of all 
tested strains and cell types of the labor- 
atory mouse Mus musculus musculus ap- 
pears as a satellite in CsCl density gradi- 
ents. 17 It has been found to consist of 
a rather short sequence (about 300 nu- 
cleotides) which is highly repeated 
(about a millionfold) , 18 Related satellite 
families have been found in two other 

Mus species, M. caroli and M. cervicolor. 
Each of these identified families displays 
quite limited heterogeneity (i.e., reasso- 
ciates to high thermal stability Te 50 over 
80°C in 0.14 M PB) and yet is only very 
loosely related to the others (Te$ Q of 
cross-product as low as 60°C in 0.14 M 
phosphate buffer). 19 - 20 Satellite DNA 
from a much closer relative of the lab- 
oratory mouse, the Japanese subspecies 
M. musculus molossinus, is virtually in- 
distinguishable from that of M. m. muscu- 
lus (cross-product of very high thermal 
stability). Nevertheless, as demonstrated 
below, there are differences in the satel- 
lites of even these two close relatives: 
M. m. molossinus DNA contains signif- 
icantly less satellite DNA than does M. 
m. musculus DNA. 

Difference in satellite content of the 
two DNAs is clearly seen in preparative 
Ag + -Cs 2 S0 4 density gradients. In the 
examples of two such gradients shown 
in Fig. 12, about 7.5% of the total optical 
density (260 m/*) of M. m. musculus 
DNA is found in the satellite band; the 
M. m. molossinus satellite, of identical 
density to the M. m. musculus satellite, 
comprises about 4.5% of the total optical 
density. As seen in Table 3, this differ- 
ence in satellite content has been ob- 
served consistently in various prepara- 
tions of these two DNAs. Over all prepa- 
rations, a mean of 7.9% satellite was 
found for M. m. musculus DNA and 
5.0% for M. m. molossinus DNA. These 
values are significantly different (t test, 
p < 0.01). 

Difference in satellite content as de- 
fined by a density gradient does not 
necessarily mean difference in sequence 
content. For example, if an altered ar- 
rangement within the genome of many of 
the M. m. molossinus satellite sequences 
resulted in their migrating with main- 
band DNA in the density gradient, the 
percentage of DNA banding as a satellite 
might significantly underestimate the 
total number of satellite sequences pres- 
ent. However, there are two indications 
that M. m. musculus and M. m. molos- 



1 1 






E 1.0 


















i 1 

i I i 


i I i 






Fig. 12. M. musculus DNA in preparative Ag + -Cs 2 S04 gradients. 355 tig of each DNA was in- 
cubated with AgNo 3 at Ag + /DNA — P = 0.3 in 6.9 ml 0.01 M sodium borate buffer, pH 9.1, for 
2 hours at room temperature. 21 Solid Cs 2 S0 4 was added to give a density of 1.5 g/cc, and the sam- 
ples were centrifuged at 33,000 rpm for 64 hours at 21 °C in the L40 rotor. Gradients were collected 
(by tube puncture) into about 60 fractions; optical density was assayed after addition of 0.6 ml 
0.1 M Tris, pH 8.0. The M. m. musculus satellite comprises about 7.5% of the total OD; the 
M. m. molosinnus satellite, about 4.5%. 

sinus DNA do in fact differ in the num- 
ber of their copies of the satellite se- 

First, the percentage of 3 H-M. m. mo- 
lossinus DNA (derived from an adult 
kidney primary cell culture grown in 
3 H-thymidine) which behaves as satel- 

lite in reassociation experiments is lower 
than observed for radioactive 3 H-M. m. 
musculus DNA. For example, when a 
mixture of 3 H-M. m. molossinus DNA 
and 14 C-M. m. musculus DNA is incu- 
bated with M. m. musculus DNA to C t 
5 X 10" 2 (sufficiently low that much of 

TABLE 3. Satellite Content of M . musculus DNA in Preparative 
Ag + -Cs 2 S04 Density Gradients 



Number of 


Source of DNA 




M. m. musculus liver 




M. m. musculus liver 




M . m. musculus liver 




M. m. musculus liver 




M. m. musculus liver 




M. m. musculus brain, 

kidneys, spleen 




M. m. molossinus liver 




M. m. molossinus carcass 




M. m. molossinus liver, 

kidneys, spleen 



*A11 DNAs were prepared by the urea-phosphate method of Britten et at., 
Year Book 68, p. 400, and, except for Prep. No. 5, all were treated identically. 
A modification of the same method was used in Prep. No. 5, resulting in impure 
DNA; it was subsequently RNase treated, ethanol precipitated, and repassed 
through HAP in urea-phosphate. Whether this additional handling accounts 
for the low percent satellite in the purified product is not known. 



the observed reassociation is due to satel- 
lite), about 12% of the 14 C, but only 
about 8% of the 3 H, elutes from HAP 
above 76°C (Fig. 13). This difference is 
also apparent when the same radioactive 
DNAs are allowed to reassociate with M. 
m. molossinus DNA to C t 10~ 2 . Here, 
about 10% of the 14 C, but only 6% of 
the 3 H, elutes from HAP above 76°C. 
Similarly, over 6% of radioactive M. m. 
musculus DNA, but only about 4% of 
3 H-M. m. molossinus DNA, is able to 
reassociate with the isolated L strand of 
the M. m. musculus satellite. In these 
reassociation experiments, there is no de- 
tectable difference (< 1°C) in the 
thermal stability of the heterologous and 

homologous products (T m from HAP in 
0.14 MPB = 82°C). Thus, the only 
observable difference in the satellite in 
these two radioactive DNAs is in its 
relative abundance. 

The second indication that M. m. mo- 
lossinus DNA contains fewer satellite 
sequences than does M. m. musculus 
DNA comes from reassociation kinetics 
experiments: a 3 H-M. m. musculus DNA 
fraction composed primarily of satellite 
reassociates more slowly with M. m. mo- 
lossinus DNA than with M . m. musculus 
DNA. The 3 H-DNA fraction was incu- 
bated with M. m. musculus DNA (a 
preparation which exhibited 7.9% satel- 
lite in density gradients) or M. m. 






1 1 

I IV' 



- / Vl4 c - 



/*\ 1 1 

/t l4 c 


80 90 
TEMP.( C) 


^r \\ 

1 \ \ 

(m ^^ 


1 -^~^ == *f-#-^. 

^\ 1 



60 70 80 90 









Fig. 13. Reassociation of tracers with M . m. musculus DNA. A mixture of sheared U C-M. m. 
musculus and 3 H-M. m. molossinus DNA was incubated with unlabeled M. m. musculus DNA at 
50°C in 0.14 M PB to Cot 5 X 10~ 2 . The sample was applied to HAP at 50°C, and the bound 
DNA (22.1% of 14 C, 18.5% of 3 H) was eluted with increasing temperature. A parallel incubation 
of tracers alone measured their self-reaction (5.4%; Cot 3 X 10" 5 ). Inset: Material eluting above 
76 °C has been corrected for self-reaction and normalized to 100% reaction. 



molossinus DNA (a preparation which 
exhibited 4.6% satellite in density gradi- 
ents) to identical C t at 60 °C and the 
percentage of reassociation of the tracer 
was measured. The difference in reac- 
tion rate of the tracer with these two 
DNAs can be calculated from the data 
shown in Table 4 to be about 1.6-fold, in 
very good agreement with the 1.7-fold 
difference in satellite content seen in the 
density gradients. 

Except for their satellite contents, the 
DNAs of these two subspecies appear 
indistinguishable in all respects tested. 
Their complements of repeated DNA are 
extremely similar ( 3 H-M. m. musculus 
DNA reassociates to identical extent and 
thermal stability with repeated DNA of 
either M. m. musculus or M. m. mo- 
lossinus) , and any sequence differences in 
their single-copy DNAs have been esti- 
mated at less than 1% (Year Book 70, 
p. 366). The mice do differ in at least 
several phenotypic traits (size, colora- 
tion, some chromosomal rearrange- 
ments, 22 etc.) ; nevertheless, they inter- 
breed to form fertile hybrids. Whether 
their differing satellite content has any 
functional significance remains to be 

TABLE 4. Reassociation of a Satellite-Enriched 
3 H-M. m. musculus DNA Fraction 





DNA Added 


of Tracer 



M . m. musculus 

1.9 X10- 3 


M. m. molossinus 

1.9 X10- 3 


M . m. musculus 

7.6XKT 3 


M. m. molossinus 

7.6XIO- 3 


The 3 H-DNA fraction was prepared by in- 
cubating sheared DNA at 60°C in 0.14 M PB to 
C t 10~ 2 and collecting all molecules which eluted 
from HAP above 70°C (13% of total 3 H). 
Instantaneous binding material was eliminated 
from this fraction by heating it to 100°C and 
repassing it through HAP; the nonbinding DNA 
(93%) is the 3 H- tracer used in the subsequent 
experiments. Tracer was incubated with 5 or 
20 p-g nonradioactive sonicated DNA in 11 ml 
0.14 M PB at 60°C for 20 minutes; binding of 
the 3 H-DNA to HAP at 60°C was measured. 

Relatedness among Several Hamsters 

Nancy Rice and Patricia Esposito 

Interspecies comparisons of both 
single-copy and repeated DNA fractions 
are often helpful in determining the de- 
gree of relatedness among the species 
tested. Toward this end, we have com- 
pared the DNAs of four species of ro- 
dents: Mesocricetus auratus (the 
"golden," or "Syrian," hamster) , Cri- 
cetulus griseus (the "Chinese" hamster) , 
Mystromys albicaudatus (the "white- 
tailed rat," or "South African" hamster) , 
and Mus musculus (the "house mouse"). 
M. musculus is classified as a member 
of the family Muridae. Though Eller- 
man 23 expressed uncertainty as to the 
best classification for Mystromys, both 
Simpson 24 and Arata 25 have grouped all 
three hamsters in the Tribe Cricetina 
(true hamsters) of the subfamily Crice- 
tinae, family Cricetidae. Both the Syrian 
and Chinese hamsters are native to parts 
of Europe and Asia ; the white-tailed rat 
is found only in southern Africa. 26 

Relatedness among these four rodent 
DNAs has been assessed according to 
their ability to reassociate with Syrian 
hamster single-copy 3 H-DNA. In these 
experiments, the tracer has been incu- 
bated with a large excess of nonradioac- 
tive DNA to high Cot; both the extent 
of cross-reaction and the thermal sta- 
bility of the cross-product were then 
measured. The results of one such ex- 
periment, in which Syrian hamster single- 
copy 3 H-DNA was allowed to reassociate 
with Chinese hamster DNA, are shown 
in Fig. 14. About 55% of the tracer was 
able to reassociate with the heterologous 
DNA in the time allotted (compared to 
over 70% reassociation of the tracer with 
homologous DNA under similar condi- 
tions), and the thermal stability of the 
cross-product (Te 50 about 70 °C) is much 
reduced relative to the homologous prod- 
uct (Te 50 81-83°C). This Te 50 of the 
cross-product is roughly comparable to 
that observed between single-copy DNAs 
of mouse and rat (Year Book 71, pp. 


50 60 70 80 90 

3 H 






-0.200 > 



o.ioo <3 


20 40 


Fig. 14. Thermal elution profile of Syrian hamster single-copy 3 H-DNA reassociated with Chi- 
nese hamster DNA. Syrian hamster single-copy 3 H-DNA (of average length about 170 nucleo- 
tides) was prepared by incubation to Cot 120 and collection of nonreassociated molecules (52% of 
total) by passage through HAP at 50 °C in 0.14 M PB. For the experiment shown above, about 2 
fig (4000 cpm/Vg) of single-copy 3 H-DNA was incubated with about 2.5 mg denatured Chinese 
hamster DNA (length about 450 nucleotides) and 2.2 /ig denatured E. coli 14 C-DNA (which serves 
as an internal standard for column operation) in 1 ml 1 M PB, 0.005 M EDTA at 63 °C for 7 days. 
One-half of the sample was diluted to 0.14 M PB and applied to a water-jacketed HAP column at 
45°C; 55% of the 3 H, 85% of the OD, and 86% of the 14 C bound. Reassociated DNA was eluted 
during a linear temperature gradient. Column buffer, 0.14 M PB, 0.06% SDS; volume per frac- 
tion, 6.5 ml. 

262-264) and between single-copy DNAs 
of man and New World monkey. 27 

Quite different results are obtained 
when the Syrian hamster single-copy 3 H- 
DNA is allowed to reassociate with either 
white-tailed rat or mouse DNA (Table 
5) . In neither case is there more than 
30% reaction of the tracer (actually an 
overestimate, since some of the observed 
reassociation occurs when tracer and non- 
radioactive DNA are incubated to low 
C t) , and the Te 50 of the cross-product is 
even lower than that observed with 
Chinese hamster DNA (about 65°C). 
Thus, Syrian hamster DNA appears no 

more closely related to white-tailed rat 
DNA than it does to mouse DNA, a re- 
sult which indicates that lines leading 
to Syrian hamster and white-tailed rat 
diverged considerably earlier (perhaps 
up to a factor of 2) than those leading 
to Syrian and Chinese hamsters. 

The same pattern of relatedness is 
seen when repeated DNAs are compared. 
Over 25% of Chinese hamster 14 C-DNA 
(a preparation of 250 nucleotides aver- 
age size) will reassociate with unlabeled 
Chinese hamster DNA at C t 10 (50°C, 
0.14 M PB); roughly one-third of this 
reassociated material is of high thermal 



TABLE 5. Reassociation of Syrian Hamster Single-copy 3 H-DNA with Other Rodent DNAs. 

DNA added 



% bound to HAP 
OD 3 H 


Te 5 ot 

3 H 


Te 50 

Syrian hamster 







Chinese hamster 












White-tailed rat 
















*No correction for the high salt concentration has been made in calculation of Cots. For comparison, 
E. coli DNA of similar size reassociates under these conditions with Coti/i about 1.3. 

fT^o is the temperature at which one-half of the bound material has eluted. Syrian hamster single- 
copy 3 H-DNA was incubated with variable amounts of nonradioactive DNA to the indicated Cots 
in 1 M PB, 0.005 M EDTA at 63°C. Denatured E. coli 14 C-DNA was included in all incubations 
and serves as an internal standard for column operation; at least 65% bound in each case. Details 
in legend to Fig. 14. 











MOUSE (18%) 


+ -- + -- +--+---+--_._+ 



Fig. 15. Thermal elution profile of Chinese hamster 14 C-DNA reassociated with other DNAs. 
Chinese hamster 14 C-DNA (average length about 250 nucleotides) was incubated with each of the 
indicated DNAs (average length about 450 nucleotides) to Cot 6-11 at 50 °C in 0.14 M PB. Sam- 
ples were applied to HAP (percent binding of the tracer is indicated in the figure) and eluted in 
4°C increments. 












MOUSE (14%) 

+ rr.j7rrSELF(2%) 

i i i 







Fig. 16. Thermal elution profile of white-tailed rat "C-DNA reassociated with other DNAs. 
White-tailed rat "C-DNA was incubated with each of the indicated DNAs to Cot 9 at 50° C in 
0.14 M PB. Samples were applied to HAP (percent binding of the tracer is indicated in the fig- 
ure) and eluted in 4°C increments. 

stability (elutes from HAP above 75°C). 
The results of incubating the same tracer 
with other DNAs to comparable Cot are 
shown in Fig. 15. The most extensive 
cross-reaction is observed with Syrian 
hamster DNA (20% net tracer reaction) , 
though less than half as many duplexes 
of high thermal stability are formed as 
with Chinese hamster DNA. Cross-reac- 
tion between the tracer and either mouse 
or white-tailed rat DNA is lower (less 
than 15% net) , and little or no material 
of high thermal stability is observed. 
A reciprocal experiment, reassociation 
of white-tailed rat 14 C-DNA with other 
DNAs, gives the same results: white- 
tailed rat DNA appears no more closely 
related to Chinese hamster DNA than 
to mouse DNA (Fig. 16). The extent of 
cross-reaction is low (about 13% net), 
and there is little or no cross-product of 
high thermal stability. Surprisingly, re- 
association of the tracer with homologous 

DNA to comparable C t (9) gives results 
that differ only slightly from those of 
the heterologous incubations. About 12% 
of the tracer (and of the unlabeled DNA) 

TABLE 6. Repeated Sequence Content of 
Various Rodent DNAs 



of Total 

of Total 







to HAP 

above 75 °C 

Mus musculus 




Rattus norvegicus 














Cricetulus griseus 












All DNAs are of 450 nucleotides average 
length. Incubation in all cases was performed in 
0.14 M PB at 50°C; binding to HAP was mea- 
sured under these same conditions. 



reacts, and only about one-fourth of this 
elutes from HAP above 74°C. Slightly 
more reassociation is observed (about 
17% of total cpm or OD) when the DNA 
is incubated to higher C t (63) ; here, 
about one-third of the product elutes 
from HAP above 75°C. These levels of 
repeated sequences are the lowest we 
have observed in any rodent DNA tested, 
a list which includes various species of 
Mus, Rattus, Mastomys, Cavia, Pero- 
myscus, Oryzomys, and Microtus, as well 
as those employed in the present experi- 
ments (Table 6). The significance, if 
any, of this variable repeated sequence 
content (previously noted in optical re- 
association studies of rodent DNAs by 
Hennig and Walker), 28 as well as the 
identity of a closer relative of Mystro- 
mys, is still unknown. 

Instantaneous Binding Fractions of 

T. I. Bonner 

In most DNA reassociation experi- 
ments using tracers, one finds that even 
at extremely low C t values (e.g., 10 -6 
for mouse DNA) a few percent of the 
labeled DNA binds to hydroxyapa- 
tite, 29 ' 30 - 31 suggesting that it has some 
double-stranded structure. The C t val- 
ues rule out the possibility of reassocia- 
tion by bimolecular collision. There are, 
however, several possible sources of the 
binding of the DNA. Some are of a tech- 
nical nature, such as the possibility that 
the DNA was not completely denatured 
or that the hydroxyapatite randomly 
binds a small amount of DNA independ- 
ently of any double-stranded structure. 
On the other hand, the binding could be 
specific to a fraction of the DNA for 
several reasons. First, some of the pieces 
of DNA could be cross-linked so that 
complementary strands never completely 
separate in the denaturation step. The 
existence of this effect has been demon- 
strated for some bacterial DNAs. 32,33 
The second possible cause of specific bind- 

ing would be the existence of partially 
complementary sequences on a single 
strand of DNA. After denaturation and 
return to reassociation conditions, such a 
strand could rapidly reassociate with it- 
self without requiring a bimolecular col- 
lision. There is reason to believe that 
such sequences may be common. A num- 
ber of RNAs are known to have large 
amounts of secondary structure, e.g., 
tRNA and ribosomal RNA, which is 
about 50% base paired at a normal re- 
association condition (50°, 0.14 M PB). 
The DNA sequences from which these 
RNAs are transcribed would be expected 
to also have large amounts of intrastrand 
base pairing. It also appears that a large 
amount of secondary structure is present 
in some RNA viruses and that the sec- 
ondary structure plays a role in control- 
ling the order of transcription of contigu- 
ous genes. Thus, sequences of this type 
could be present either as the result of 
the integration of viral DNA or of their 
evolutionary advantage as a mechanism 
for the control of gene expression. A 
third possibility is that some fragments 
have an enhanced affinity for HAP which, 
while specific to those fragments, does 
not depend on the presence of secondary 
structure. It is possible, for example, 
that longer fragments are more likely to 
stick to HAP. In order to determine the 
nature of the rapidly binding portion of 
the DNA, we have performed a number 
of experiments using 3 H-thymidine— 
labeled mouse DNA. 

We have measured the "instantaneous 
binding" for two sizes of DNA, sonicated 
DNA with a length of about 450 nucleo- 
tides and DNA which results from the 
MUP extraction process and which is 5- 
10 times as long. Sonicated tracer gives 
an instantaneous binding of about 4% 
with two-thirds of the material having 
high thermal stability, i.e., eluting from 
hydroxyapatite above 75 °C (see Fig. 17). 
The thermal elution profile of sonicated 
native tracer is nearly identical to the 
profile of the instantaneously binding 












55 60 70 80 90 95 .4 

Fig. 17. Thermal elution profiles of sonicated 
(about 450 nucleotides long) tracer (circles) 
and tracer which is estimated to be 5-10 times 
as long (plusses). The tracer was denatured by 
boiling for 5 minutes in 0.14 M PB, chilled in 
ice water for 2 minutes, placed on a 50° C hy- 
droxyapatite column and eluted 2 minutes later 
with 0.14 M PB. The total Cot prior to elution 
is less than 5 X 10" 5 . Incubation of the long 
tracer at 50 times greater Cot indicates that less 
than 1% of the mouse satellite should be re- 
associated. The amount eluting in 5°C temper- 
ature steps is expressed in terms of percentage 
of total counts loaded on column. The total 
binding of the short tracer is 4.4% and that of 
the long tracer is 21.2%. 

(IB) tracer above 80 °C except that it 
may melt about 1°C lower. The high 
stability of IB tracer is highly suggestive 
of cross-linking which should allow the 
DNA to reform to essentially native sta- 
bility. The amount of binding is, however, 
much greater than was observed where 
cross-linking was demonstrated 32,33 for 
bacterial DNA. In those cases the 
amount of binding, when extrapolated 

linearly to the sonicated size, is about 
0.01%. Using MUP length DNA, the 
amount of IB is about 20% and more 
than half of the IB tracer has low sta- 
bility (elutes below 75°C). Low thermal 
stability appears to be quite unlikely for 
cross-linked DNA but would be expected 
to result when a single strand folds back 
on itself to form either short or imper- 
fectly base-paired regions. 

In order to determine whether the 
binding is the result of a failure to de- 
nature the DNA completely, a sample 
was boiled in 0.035 M PB instead of 0.14 
M PB, which should have lowered the 
melting temperature by 10 °C. No dif- 
ference in binding was observed. In ad- 
dition, measurements of binding follow- 
ing denaturation for 5, 10, and 15 minutes 
show a small decrease in binding which is 
compatible with estimates of the reduc- 
tion in size due to strand scission. 

The specificity of the binding to hy- 
droxyapatite was checked by measuring 
the rebinding of fractions which initially 
bound (15% total binding). The fraction 
which initially bound with high stability 
(greater than 75 °C) , rebound about 65%. 
Since less than 1% of the initially non- 
binding material will bind with high 
stability on a second pass, this fraction 
appears to be quite specific. Specificity 
of the fraction which binds with low sta- 
bility (less than 75°C) is less clear. Al- 
though it also rebinds about 65%, the 
low-stability binding on a repass of 
initially nonbinding material has been 
measured to be about 1% or about 6%, 
using different tracer preparations and 
different HAP batches. Thus, the low- 
stability fraction may contain a moder- 
ate amount of nonspecific binding. On 
the other hand, it would be difficult to 
reconcile the 65% rebinding (which is 
identical to the high-stability rebinding) 
with more than a small amount of non- 
specific binding. 

In order to determine whether cross- 
linking is present, alkaline sucrose gradi- 



ents were run, using non-IB, low-stability 
IB and high-stability IB fractions. In 
these denaturing conditions, cross-linked 
fragments should have twice the molecu- 
lar weight of single-stranded noncross- 
linked fragments. The results are not 
yet conclusive but indicate that the low- 
and high-stability IB fractions have mo- 
lecular weights about 1.7 to 2.0 times as 
large as the nonbinding fraction. This 
result suggests that cross-linking is in- 
volved, but it is possible that the IB frac- 
tions are selected on the basis of greater 

In order to determine whether the in- 
stantaneous binding sequences are rep- 
resentative of a specific fraction of the 
genome or are randomly distributed 
throughout the genome, two experiments 
were performed. In the first experiment, 
the IB fraction from DNA several times 
as large as sonicated DNA was sonicated 
along with a large excess of unlabeled 
mouse DNA. By measuring the reasso- 
ciation of tracer compared to the optical 
density of the cold DNA at various Cot 

values, one can learn whether the IB 
sequences are contiguous with sequences 
representing a kinetically defined frac- 
tion of the genome. As shown in Fig. 18, 
when one subtracts the 27% instantane- 
ous binding component, the tracer has a 
reassociation curve quite comparable to 
the whole mouse genome. There appears 
to be about 8-10% less tracer reacting in 
the intermediate Cot range of 1—100 with 
a corresponding increase in the single- 
copy region, but there are no repetition 
frequencies that are dramatically en- 
hanced or suppressed. This is similar to 
the result of Davidson et al. sl who find 
no difference in the reassociation of simi- 
larly prepared Xenopus tracer and un- 
labeled Xenopus DNA. 

The second experiment sought to de- 
termine whether there was any variation 
in IB with base composition. Fractions 
were taken across a CsCl gradient of 
mouse tracer, and the IB character of 
these fractions was determined. As 
shown in Fig. 19, there is a definite varia- 
tion in the amount of tracer IB across the 

icr 3 


10"' 0° 10 


10 2 I0 3 

10 4 

Fig. 18. Reassociation kinetics of sonicated tracer in the presence of a 10 4 -fold excess of un- 
labeled mouse DNA (except for Cot < 10 -4 points which are tracer alone). The instantaneously 
binding fraction (12%) of a longer tracer was sonicated together with the cold DNA. The frac- 
tion of tracer binding to HAP is shown as plusses, the fraction of total optical density which 
binds is shown as circles. The optical density readings of nonbinding and binding fractions were 
corrected for hypochromicities of 12.9% and 23.7%, respectively, as determined by melts of such 
material after incubation to Cot ^ 2000. 





u 1.72 


t 1.68 







O 600 


O 400 


"i — I — r 

1 — i — I — r 

i ~ t — r + l i I i i i i i i i 

10 14 18 22 26 



Fig. 19. The instantaneous binding of various 
fractions resulting from a CsCl gradient of 
mouse tracer estimated to be 1200-4000 nucleo- 
tides long. The tracer was used in amounts 
ranging from 0.025 ml to 0.100 ml in 15 ml of 
0.14 M PB. The instantaneous binding is shown 
as open circles, the number of counts per min- 
ute for 0.010 of gradient fractions is shown as 
plusses, and the density of gradient fractions is 
shown as closed circles. 

gradient. There are two sources of varia- 
tion which would occur even if the 
amount of DNA with IB properties did 
not vary with base composition. The first 
is that the specific activity of the DNA 
varies linearly across the gradient (in- 
creasing with increasing fraction num- 
ber) since the label is only in thymine. 
Thus, part of the increase in tracer bind- 
ing of fractions 12-15 could be due to 
increasing specific activity. However, 
the decrease in binding of fractions 16-21 
must be due to another effect. The sec- 
ond source of variation in tracer binding 
is the possibility that the extremes of the 
distribution of tracer in the gradient may 
be enriched in short fragments relative to 
the peak of the distribution. Shorter 
fragments would be expected to bind 
smaller amounts of tracer. However, the 

size variation would be expected to be 
symmetrical about the peak, whereas the 
amount of tracer binding peaks well 
down the heavy side of the gradient peak. 
It is therefore apparent that the amount 
of IB DNA has a dependence on base 
composition. Preliminary results indi- 
cate that the variation is more pro- 
nounced in the low-stability fraction. 

The present results suggest that fold- 
back, cross-linking, and some nonspecific 
binding all occur. The binding does not 
seem to be uniformly distributed through- 
out the genome and is specific enough 
that it may prove a useful means of frac- 
tionating the genome. It remains to be 
seen what the functions of these fractions 

Enzyme Evolution in the 

Dean B. Cowie, Paolo Truffa-Bachi* 
and Georges N. Cohen* 

The extent of homology among nucleo- 
tide sequences of the DNA of related 
organisms is a measure of their relative 
evolutionary histories. In bacteria, evo- 
lutionary divergence may arise by ran- 
dom single-base changes ; by gene fusion ; 
or by the deletion, addition, or substitu- 
tion of large DNA segments. Such 
changes may be indicated by determining 
the extent of DNA homology and the 
thermal stability of the reaction products 
formed by the hybridization of the DNAs 
of any two related species. 

Since more than one mechanism is 
involved in genetic divergence, portions 
of the genome might be expected to have 
different rates in the extent of fixation of 
base sequence change. These rates, how- 
ever, will depend upon the number of 
changes tolerable without damaging por- 
tions of the DNA which code for the 
enzymes essential for survival. 

* Service Biochimie Cellulaire. Institut Pas- 
teur, Paris, France. 



Such DNA-DNA hybridization studies 
have already indicated that certain por- 
tions of bacterial or viral genomes are 
highly conserved. Other portions show 
marked divergence and the loss or addi- 
tion of large contiguous segments of nu- 
cleotide sequences. These differences 
among the DNA sequences may result in 
differences in amino acid sequence and 
protein structure. Indeed, recent im- 
munological comparative studies of sev- 
eral enterobacterial enzymes 34,35 gave 
results that indicated the pattern of in- 
tergeneric relationships agrees with that 
derived from DNA hybridization inves- 
tigations. 36 * 37,38 

On the other hand, certain proteins — 
despite species variations — are known to 
be conserved, tolerating little change. 
Cohen et al. (1969) 39 provided strong 
biochemical evidence that throughout the 
Enterobacteriaceae distinctive features 
of the regulation of aspartokinase I- 
homoserine dehydrogenase I (AKI- 
HSDI) are conserved in spite of wide 
variations in the guanine-plus-cytosine 
(G + C) content among these bacteria. 

Last year another immunological tech- 
nique was developed (Year Book 71, p. 
253) to test whether the known genetic 
divergences among these enterobacteria 
are reflected in the structure or composi- 
tion of this essential enzyme. Sepharose 
columns were prepared containing anti- 
bodies specific for purified E. coli K12 
AKI-HSDI. When crude extracts from 
E. coli K12 were passed through the col- 
umn (in the presence of 2 X10" 3 M L- 
threonine) , AKI-HSDI was bound to the 
antibodies and removed from the crude 

A number of enterobacterial strains 
(shown in Table 7) were selected to test 
whether the homoserine dehydrogenase 
(HSD) contained in crude extracts from 
these cells would bind to the E. coli K12 
AKI-HSDI antibody column. Table 7 
also shows the G + C content of their 

TABLE 7. Bacterial Strains Employed 43 - 44 







at 60°C 

E. coli K12 Tir 8 



Serratia marcescens 



Proteus vulgaris 



Providencia sp. 7/67 



Aeromonas hydrophila 



DNAs and the degree of genetic related- 
ness as determined by DNA-DNA hy- 
bridization studies (using E. coli DNA 
as the reference material). For these 
enzyme tests, homoserine dehydrogenase 
was assayed in the direction homoserine 
— > aspartate semialdehyde by following 
NADP + reduction spectrophotometric- 
ally. 40 

The sepharose-antibody column util- 
ized in these experiments completely 
binds all the AKI-HSDI contained in 
crude extracts of E. coli K12, strain Tir 
8 * as shown in Fig. 20. The capacities 
of the immunoadsorbents were deter- 
mined by adding successive loads of crude 
extract containing known quantities of 
AKI-HSDI to the column. After each 
load of protein had passed through the 
column, it was washed with buffer A f 
and a new load added. When enzyme 
activity appears in the eluate, the col- 
umn is saturated to determine the capac- 

Usually, after each experiment the 
column was washed with 4 or 6 M urea 
(containing .001 N hydrochloric acid), 
removing all the enzyme bound to the 
antibodies. The column was then washed 
with buffer A and was ready for further 

Figure 21 shows that at least a part of 
the AKI-HSDI remains active while on 
the column; elution with 1 M urea re- 

* Strain Tir 8 is a genotypically derepressed 
mutant of E. coli K12. 

t Buffer A : 2 X 10" 2 M potassium phosphate, 
pH 7.2, containing 0.15 M KC1. 



TIR 8 

2X I0 5 










I0 4 

E. coli TIR 8 











J L 





20 40 60 


Fig. 20. Immunoadsorption of AKI-HSDI on antibody column. Open circles, appearance of 
labeled protein in eluant fractions; closed circles, appearance of AKI-HSDI activity in eluant 
fractions. Proteins labeled by growing cells in synthetic medium containing C 14 L-arginine. 

leases active enzyme and, when followed 
with 2 M urea, additional enzyme activ- 
ity is eluted. Additional protein is re- 
leased by 4 M urea, but at this concen- 
tration the enzyme is denatured. 

When crude extracts from Serratia 
marcescens or Proteus vulgaris are added 
to the antibody column, essentially all of 
the HSD enzyme activity is retained by 
the column. Saturation occurs at about 
the same level as that observed with the 
AKI-HSDI contained in the crude ex- 

tract from E. coli Tir 8 cells. The aver- 
age G + C content of Serratia DNA is 
55-58.5%, while that of Proteus ranges 
between 37 and 39%, quite different from 
that of the E. coli cells (50-52%). The 
degree of genetic interrelationship as 
measured by DNA hybridization meth- 
ods is 24%, compared with ~ 9% in E. 
coli. It is clear that, despite these dif- 
ferences, HSD contained in the crude ex- 
tracts of both Serratia and Proteus has a 
high affinity for the AKI-HSDI antibody 



column. Only a few percent of the HSD 
activity is observed in the eluate after 
passage of the crude extract through the 
column (until saturation occurs) . 

Another enterobacterium tested, Provi- 
dencia sp. 7/67, gave different results. 
No retention of the HSD activity was 
observed when crude extracts of fiese 
bacteria were loaded on the antibody 
column. Two loads of 1600 enzyme units 
were loaded on the column. The HSD 
activity contained in the eluant fractions 
was observed to be 1690 and 1600 units. 
This was the only enterobacterium tested 

in which the HSD had no affinity for the 
antibody column. 

Cohen et al., 39 assessing taxonomic in- 
terrelationship by studying regulatory 
mechanisms among the enterobacteria, 
came to the conclusion that Aeromonas 
hydrophila is closer to the coliform bac- 
teria than to the aerobic pseudomonads. 
When crude extracts of Aeromonas hy- 
drophila were passed through the anti- 
body column, 70% was retained by the 
column. Some of the HSD activity that 
passed through did not completely track 
with the bulk of the Aeromonas protein, 


2 M 


<-4M UREA- 


-8 ^-4 




•6 - 



-4 S - 






5 ~ 





2 f 





10 20 30 


Fig. 21. Release of antibody-bound protein with urea. Open circles, labeled protein released in 
eluant fractions by the addition of 1 M and 2 M urea to antibody column ; closed circles, labeled 
protein released in eluant fractions with 4 M urea; X's, appearance of AKI-HSDI activity in 
eluant fractions. Proteins labeled as in Fig. 20. 











10 20 


Fig. 22. Passage of proteins from Aeromonas hydrophila through antibody column. Open cir- 
cles, total protein as determined by optical density of eluant at 280 nm; X's, HSD activity in 
eluant fractions. 

as shown in Fig. 22. This result would 
indicate that a portion of the HSD had a 
different chromatographic characteristic 
than that retained by the column or that 
which moved with the passage of bulk 
protein. It would then appear that this 
organism has some HSD not specific to 
the AKI-HSDI antibodies. 

Exchange of antibody -bound homo- 
serine dehydrogenase. Last year exchange 
reactions were observed when crude ex- 
tracts of an E. coli mutant Gif 108 (con- 
taining aspartokinase I but devoid of 
homoserine dehydrogenase I) were loaded 
on the antibody column previously sat- 
urated with wild-type AKI-HSDI. The 
properties of eluted enzyme were unus- 
ual: When tested for threonine sensitiv- 
ity, HSD activity was enhanced rather 
than inhibited as it is for wild-type en- 

zyme. This enhancement was always 
greater than 100%. 

This result could be due to either the 
displacement of wild-type enzyme or the 
formation of hybrid molecules containing 
monomers or dimers from both mutant 
and wild-type enzymes. Whereas the 
enhancement by threonine would favor 
the hypothesis that the molecules eluted 
are different from the wild-type AKI- 
HSDI, the possibility remains that wild- 
type enzyme is eluted in a stable, less 
active conformation and that the en- 
hancement by threonine reflects a reac- 
tivation to the normal configuration. 

This year further exchange reactions 
were investigated. Experiments were 
carried out to determine whether HSD 
previously bound to the antibody col- 
umn would exchange with enzyme con- 



tained in crude extracts of other entero- 
bacteria. In one experiment the antibody 
column was saturated with radioactive 
HSD by the passage of crude extract 
from E. coli strain Tir 8 labeled with C 14 
L-phenylalanine. Two loads, one ml 
each, containing 12,000 units of HSD 
activity per ml were passed through the 
column (capacity ^20,000 units) and the 
column washed with buffer A until no 
radioactivity appeared in the eluate. The 
total radioactivity passed through the 
column was 179,000 counts per minute; 
at column saturation with HSD 2 %4ths, 
or 150,000 counts per minute, would have 
been put on the antibody column. 

The addition of nonradioactive crude 
extract from Serratia marcescens sp. 
7/72 cells caused the release of radio- 
active protein, as shown in Fig. 23. The 
first Serratia extract contained 13,500 
units of HSD activity; roughly 13,000 
units were recovered in the eluate, indi- 

cating that the column had been pre- 
viously saturated with the E. coli enzyme. 
The radioactivity eluted by the Serratia 
extract was 1000 counts per minute. 

A second load of crude extract from 
Serratia cells (2800 units of HSD) re- 
sulted in the further displacement of 
radioactive protein (750 counts per min- 
ute) and the recovery of 2600 units of 
enzyme activity in the eluate. Six molar 
urea added to the column released the 
balance of the bound radioactivity on the 
column (1800 counts per minute). No 
HSD activity appeared in this eluate 
because the enzyme is denatured by the 
high concentration of urea. 

The total amount of radioactivity 
bound (displaced by the urea or ex- 
changed with Serratia protein) was 3550 
counts per minute, or 2.4% of the total 
protein load at saturation with the E. coli 
enzyme. Approximately 50% of the 
bound protein had been displaced by the 










l 1 
1 < 

1 1 

1 I 

1 1 1 1 



: 1 






— <o 


Q _ 



1 .fc 



\ 6M 



b °^ 

I 5 



\ h- 







T <D 

9 Q 

p\ Q 

n CC 



! < 

h (? < 

[b < 

00 3 


3 [ 

! O 
1 ft> " J 

\ A o 

oil - ' 

b O 
/ \ - 1 

IV 5 




1 l\ "O 

O \ c 

D 1 "^ 

r \ "o 

/ A. c 

1 to * 



\p \ ^ 

I — 

J ^ ^ 





fl \, r 

7 \o * 



i i 

1 1 

1 1 

* w*fi h BACKGROUND 
l 1 1 

10 20 30 40 50 60 





Fig. 23. Release of labeled, antibody-bound E. coli protein with Serratia marcescens proteins 
and 6 M urea. Open circles, appearance of labeled protein in eluant fractions. Proteins labeled by 
growing cells in synthetic medium containing C 14 L-phenylalanine. 



20 30 40 50 



Fig. 24. Release of labeled, antibody-bound Serratia protein with E. coli protein and 8 M urea. 
Open circles, appearance of labeled protein in eluant fractions. Proteins labeled as in Fig. 23. 

Serratia enzymes prior to the addition of 
the urea. 

In another experiment a crude extract 
of C 14 L-phenylalanine labeled Serratia 
marcescens was loaded on the antibody 
column. Figure 24 shows that the release 
of antibody bound Serratia HSD can be 
produced by the addition of E. coli strain 
Tir 8 enzyme. 

The quantity of HSD in Serratia cells 
is less than a tenth of that found in the 
derepressed strain of E. coli Tir 8. Con- 
sequently, much more Serratia protein 
must be loaded on the column in order to 
approach saturation conditions. Further- 
more, when heavy loads of crude extracts 
are added to the column, some of the 

HSD does not bind and is eluted along 
with the bulk of the cellular proteins, 
even though the column is unsaturated. 
Only when the crude extract is diluted 
and the flow rate reduced (1 ml/10 
minutes), will complete retention of the 
HSD (99%) occur. 

In the above experiment, 5 ml of a 
crude extract of Serratia cells were 
loaded on the column. This preparation 
contained 13,000 units of HSD activity 
and 670,000 counts per minute of radio- 
activity. About 60% of the HSD activity 
was retained by the column (7480 units). 
After washing with buffer A, a non- 
radioactive preparation of E. coli Tir 8 
containing 18,400 unite of HSD was 



added to the column, which retained 
12,400 units of HSD activity. The total 
bound enzyme (12,400 E. coli plus 7500 
Serratia) was 19,900 units, a value in 
good agreement with column saturation 
results (^20,000 units) obtained when 
either E. coli or Serratia enzymes are 
used singly. A second and third load of 
crude extracts of E. coli Tir 8 (each 
18,400 units HSD) resulted in the recov- 
ery of 18,700 and 17,600 units, respec- 
tively, in the eluate fractions. 

The passage of unlabeled E. coli HSD 
released 1760 counts per minute, or '—'50% 
of Serratia protein. The total bound Ser- 
ratia protein was determined by eluting 
the balance of the radioactivity bound to 
the column with 8 M urea (3580 counts 
per minute) and adding this amount to 
the E. coli displaced or exchanged pro- 
tein. This total (5340) represents 0.8% 
of the Serratia radioactivity. 

These preliminary exchange experi- 
ments suggest that the relative antibody 
binding efficiencies of the two enterobac- 
terial enzymes are similar. More quanti- 
tative determinations will depend upon 
comparing Serratia HSD with a strain of 
E. coli having about the same ratio of 
HSD to total protein content (an E. coli 
strain that is not derepressed). These 
experiments do demonstrate that ex- 
change or displacement does occur, and 
they confirm the conclusion that the 
number of binding sites on the antibody 
column for Serratia and for E. coli HSD 
are about equal. 

Studies reported last year outlined 
some of the characteristics of the AKI- 
HSDI antibody column {Year Book 71, 
p. 253, Cowie et al* 1 1973). The studies 
emphasized the capacity of the column 
to distinguish between two conforma- 
tional states of this enzyme. The form 
predominating in the presence of threo- 
nine (T form) has antigenic determi- 
nants exposed to react with correspond- 
ing sites on the antibodies, whereas the 
enzyme is less available for binding in 
the presence of aspartate or other ligands 
that shift the equilibrium from the T to 

the R form of the enzyme. 42 In the T 
form, all of the AKI-HSDI in E. coli is 
bound by this specific antibody column. 

Studies carried out this year employed 
this column characteristic to study en- 
zyme evolution in Enterobacteriaceae. 
In the presence of L-threonine, HSD con- 
tained in crude extracts of Serratia and 
Proteus bound to the antibody to the 
same extent as that observed with the 
E. coli enzyme. This result indicates 
that among these bacteria, despite evolu- 
tionary divergences (Table 7) , HSD is 
more conserved than the DNAs. 

Another enterobacterium, Providencia, 
contained HSD that did not react at all 
with the E. coli antibody column. In ad- 
dition, HSD contained in crude extracts 
from Aeromonas hydrophila appears to 
be in two forms: — 70% binds to the 
antibody column, and the rest does not. 
This result indicates the usefulness of the 
antibody column in the detection and 
separation of isofunctional enzymes in 
crude extracts of genetically related cells. 
Thus, Aeromonas hydrophila seems to 
have an HSD with immunological fea- 
tures resembling those of enterobacteria 
and some material endowed with the 
same enzymatic activity but not able to 
bind to the antibody column. 

MENTS Suitable for Reassociation 

Bill H. Hoyer, N. R. Rice, and 
N. W. van de Velde 

Pressure-cell shearing has convention- 
ally been used to prepare DNA frag- 
ments for reassociation experiments. The 
50,000 lbs/in" 2 (50K) pressure apparatus 
is not used by many laboratories because 
of its cost and the mechanical upkeep 
required. In addition, pressure cells are 
difficult to clean and to decontaminate 
when exposed to highly radiolabeled 
preparations. Ultrasonic devices are 
relatively inexpensive, readily available, 
and very easy to clean. Because of these 
advantages we have investigated "sonica- 



tion" as a means of preparing DNA frag- 
ments and have found it to be satisfac- 
tory. However, a number of variables 
influence the size and properties of the 
DNA fragments produced by sonication, 
and we will discuss some of these. 

An obvious variable is the ultrasonic 
device and its attachments. Table 8 gives 
the specifications and calibration of the 
instrument we used. The type and diam- 
eter of the tip used, as well as the effi- 
ciency of energy transfer, are important 
variables. The properties of the instru- 
ment we used may be approximated from 
these simple measurements and compar- 
able micro tip. We have found that 
power-meter indications are not consist- 
ent from instrument to instrument and 
therefore do not include such readings in 
our data. 

Table 9 gives the results obtained 

TABLE 8. Specifications and Calibration of 
Sonifier Cell Disruptor * 

Output (gm)t 







*Heat Systems Ultrasonics, Inc., Plainview, 
Long Island, N.Y.; Model WHO with 1/8 inch 
diameter micro tip. 

fSonic output was measured on a Met tier top- 
loading balance : 50 ml of water was placed in a 
100-ml Pyrex No. 1000 beaker on a foam pad 
on the balance pan and the micro tip was 
immersed halfway (16 mm) between the meniscus 
and the bottom of the beaker. The downward 
force, in grams, was measured at the indicated 

when time of sonication and volume of 
the preparation sonicated are changed. 
Both time and volume are important 
variables. The size distribution of the 
DNA fragments becomes narrower with 
increased time of sonication, and the 

TABLE 9. Time of Sonication of Native DNA and Sedimentation of the 
Single Strands on Alkaline Sucrose Gradients * 

Volume and 







3 H-DNAf 

Width/ Width/ 

Distance! Distance § Distance! Distance § 

2-ml volume 
setting 2 

10-ml volume 
setting 4.5 































*Escherichia coli randomly labeled 14 C-DNA (50K sheared) was mixed with 
sonicated 3 H green monkey DNA and 50/ul of the mixture was layered onto 11 
ml (8 cm) linear 5-20% sucrose gradients into 10~ 2 M EDTA, 0.1 M NaCl, 
0.25 M NaOH. The gradients were centrifuged 21 hours at 39,000 rpm in a 
Te-41 horizontal rotor. Gradients were sampled from the top with a Buchler 
auto-Densiflow and a peristaltic pump. 

fFor sonication, the 3 H-DNA in 0.1 M NaCl, 0.01 M HEPES buffer, pK 
7.4, was placed in either a Cat no. 915 Schwarz/Mann 3.5 ml vial (2 ml volume) 
or a Packard glass scintillation vial (10 ml volume). The micro tip was placed 
8 mm from the bottom of the small vial and 10 ml from the bottom of the scin- 
tillation vial, Cooling was done with magnetically stirred crushed ice-water 
mixture, and temperatures within the vials did not exceed 5° C. 

{From the meniscus to the peak of the radiolabeled DNA distribution (in 

§The width of the DNA distribution at 1/2 height (in cm) was divided by 
the distance the DNA sedimented at the peak of its distribution. This number 
indicates the fragment size distribution and may be used to compare that of the 
50K marker to that of the sonicated DNA. 



average fragment size also becomes 
smaller. Less energy input is necessary 
to fragment DNA in the smaller volume. 
It is apparent that any set of conditions 
derived to produce DNA fragments of a 
consistent size must take into account 
time, volume, and geometry. The 50K 
fragments (about 450 nucleotides single 
stranded) produced by the pressure cell 
have a somewhat narrower size distribu- 
tion than those produced by sonication. 

All sonications described have been 
done continuously rather than in bursts 
because it was possible to keep the tem- 
perature at 5°C or less in vessels im- 
mersed in stirred ice- water mixtures. 

Both native and denatured DNA may 
be sonicated, and the fragment size ob- 
tained is influenced by the physical prop- 
erties of the DNAs that are processed. 
Native DNA fragments (prepared in 
neutral solutions) have the advantage 
that they may be characterized by their 
optical and chromatographic properties. 
(Evidence suggests that a small amount 
of cross-linking may occur and that son- 
icated native DNA maintains its helical 
structure.) On the other hand, introduc- 
tion of cross-links is less likely in de- 
natured (prepared in alkaline solutions) 
DNA fragments. 

Table 10 indicates that lower energies 
are required to fragment single-stranded 
alkaline DNA than are required to frag- 
ment native DNA comparably. Also, 
fragments considerably smaller than the 
50K marker may be prepared by both 
means, but even smaller fragments may 
be obtained from alkaline DNA solutions 
at easily attainable energy inputs. 

Small DNA fragments, when reassoci- 
ated, have lower thermal stabilities than 
larger fragments. 45 In agreement with 
their sedimentation properties, the Te 50 
values 46 of reassociated DNA fragments 
in Table 10 indicate that the DNA frag- 
ments prepared in alkaline solutions are 
smaller at a given energy input than 
those prepared from native DNA sonic- 
ated in neutral solutions. 

The fragments produced by sonication 
at setting levels 2 and 4 (Table 10) in 
0.25 M NaOH were reassociated to about 
Coti/2, adsorbed to hydroxyapatite in 
0.05 M phosphate buffer (PB) at 50°C 
and eluted by a linearly ascending con- 
centration of PB. The single-stranded 
Escherichia coli DNA fragments both 
eluted with 0.14 M PB; the reassociated 
level-four fragments eluted with 0.24 M 
PB ; and the reassociated level-two frag- 
ments with 0.28 M PB. Thus, the size of 

TABLE 10. Sedimentation and Thermal Elution Properties of Escherichia coli DNA 
Sonicated in Neutral and Alkaline Solutions * 


Tesot of 
DNA (°C) 


14 C-DNA 


32 P-DNA 



Distance § 


Distance § 















*Marker and sonicated DNAs were mixed and layered onto alkaline sucrose gradients as described 
in Table 9. The neutral solution contained native DNA 0.1 M NaCl, 0.01 M HEPES (pH 7.4), and 
10- 3 M EDTA. The alkaline solution was 0.2 M NaOH and 10~ 3 M EDTA. All sonications were 
done in 2 ml with conditions described in Table 9; the time of sonication was 10 minutes. 

fThe Te b0 (midpoint of the thermal elution curve) was determined by eluting adsorbed, reassociated, 
sonicated DNA from hydroxyapatite. A linear thermal gradient of 0.14 M PB was used as the eluant. 

^Distance was determined as described in Table 9. 

§Width/distance ratio was determined as described in Table 9. 



the DNA fragments influences their 
phosphate elution properties and may 
therefore influence fractionations based 
upon PB concentrations. 

It has been determined that native 
DNA in a 10-ml volume under the con- 
ditions described in Table 9 is frag- 
mented to about 50K size in 5 min at a 
setting of about 4.5. However, it is es- 
sential that one calibrate his own ultra- 
sonic device to produce the required 
DNA fragment size. Volume, geometry, 
energy input, state of the DNA, and time 
of sonication appear to be the important 
variables. The concentration of DNA 
and the salt concentration of neutral 
solutions appear to have little influence 
within reasonable limits. 

Expression and Divergence of 
Repeated DNA Sequences 

Michael J. Byers 

Repeated sequences are found in the 
DNA of eukaryotic cells, but little is 
known of their function. Previous ex- 
periments have investigated the expres- 
sion and divergence of single-copy DNA 
(Year Book 70, p. 376). Experiments 
now in progress are directed at the fol- 
lowing questions: Are repeated nucleo- 
tide sequences transcribed into messenger 
RNA? If so, are these transcribed se- 
quences more highly conserved than re- 
peated sequences as a whole or than those 
transcribed from single-copy DNA? 

It has been demonstrated in several 
laboratories that cytoplasmic mRNAs 
have a sequence of polyadenylic acid 
covalently attached at the 3' end of the 
messenger molecule. The use of affinity 
chromatography on oligo-dT cellulose 
allows separation of these poly A at- 
tached molecules from the bulk of the 
cellular RNA. DNA sequences comple- 
mentary to this class of RNA molecules 
can be isolated by hybridization. The 
degree of conservation of these tran- 
scribed sequences can be determined by 
thermal chromatography of the double- 

stranded molecules formed in reactions 
with DNA sequences from homologous 
or heterologous species. 

Efforts to date have concentrated on 
developing a method of isolation that 
minimizes the probability of enzymatic 
degradation and physical loss of RNA 
during extraction procedures. One 
method that has shown promise is the use 
of dithiothreitol (DTT) in the lysing 
buffer. As a reducing agent, DTT re- 
duces the disulfide linkages of pancreatic 
RNase and destroys its activity. DTT, 
in conjunction with 8 M urea, solubilizes 
most of the protein in a tissue sample, 
and CHC1 3 extraction does not precipi- 
tate the solubilized protein. Loss of 
RNA at the interface is thus eliminated, 
but the solubilized proteins are difficult 
to remove. As a compromise, DTT has 
been included in a lysing medium without 
urea, and resulting precipitates at or- 
ganic-aqueous interfaces are reextracted. 
This modification shows promise for the 
extraction of undegraded RNA. 

Labeling of DNA with 125 I 
M.B. Davis 

Comparison of DNAs by reassociation 
techniques has been hindered by the 
difficulty of obtaining high specific ac- 
tivities using in vivo isotopic labeling. 
Commerford 47 has recently described a 
method for labeling polynucleotides in 
vitro with 125 I. In order for 12r T-labeled 
DNA to be useful in hybridization ex- 
periments, it is necessary to determine 
the effects of the iodination reaction on 
piece size, thermal stability, and ability 
to form hybrids. 

A series of iodination reactions was 
carried out in 0.5 ml of buffer A* (pH 5) , 
with 2.5 X 10" 5 M KI (carrier) and 
1.5 X 10" 4 M TICI3 (oxidizing agent). 
In one case the reaction was carried out 
in buffer A adjusted to pH 6. Each re- 
action mixture contained 100 /xg of 
double-stranded E. coli DNA (sonicated 

♦Buffer A: 0.1 M sodium acetate, 0.04 M 
acetic acid (pH 5). 


to an average length of approximately tities of non-TCA precipitable counts. 

760 nucleotides) and 0.05 mCi of 125 I as To test for the presence of precipitable 

Nal. The reaction mixtures were heated labeled products other than DNA (or 

for various periods of time at 40 °C or RNA), a phenol-chloroform extraction 

60 °C, cooled to 0°C, combined with 0.025 was done on one of the samples. Less 

ml of 0.1 M Na 2 S0 3 (to reduce excess than 5% of the counts remained in the 

T1C1 3 ) and 0.1 ml of buffer B f (to raise phenol phase. For the further testing 

the pH to 9), and heated for 20 minutes described below, DNA from batch 1 

at 60 °C (to remove any noncovalently (optimal labeling conditions) was used 

bound iodine). The fraction of 125 I unless otherwise stated. In all cases, the 

which was stably bound to the DNA was unlabeled reference was E. coli DNA 

determined by measuring the ratio of that had been sonicated to an average 

TCA precipitable counts to total counts length of 760 nucleotides, 

for 10A samples of each mixture. The The stability of the final product was 

results are shown in Table 11. The per- tested by measuring the TCA precipi- 

centage of cytosine that was iodinated table counts before and after heating to 

was calculated, using the assumption 100 °C for 5 minutes. There was no sig- 

that all precipitable counts were due to nificant difference in the number of pre- 

the binding of iodine to cytosine. The cipitable counts before and after the 

specific activity of batch 1, measured treatment. 

under optimal conditions in a Packard The effect of the iodination reaction on 
well-type scintillation counter (approxi- DNA piece size was determined by com- 
mately 25% efficient), was 1.5 X 10 4 paring the sedimentation velocities (at 
cpm per ^g of DNA. The specific activ- p H 13) of E. coli before and after iodina- 
ity could be greatly increased by increas- t j on> j n one set f measurements, the 
ing the ratio of 125 I to carrier I; during sedimentation velocities of both treated 
these reactions the ratio of 125 I to carrier anc j untreated DNA were determined 
I was approximately 1:500. Commer- optically, using a Spinco Model E ultra- 
ford's results also indicate that the spe- centrifuge with ultraviolet optics. The 
cific activity could be further increased average length of the iodinated DNA was 
(perhaps tenfold) by denaturing the approximately 85% of the length of the 
DNA before iodinating. untreated DNA. Another set of measure- 
The DNA from each reaction mixture men t s was made in an alkaline sucrose 
was precipitated with ethanol and dis- gra dient on a Spinco Model L ultra- 
solved in 0.2 ml of water. The final cent rifuge. The relative distances trav- 
solutions contained no significant quan- ded by the ^ DNA and by the labeled 

t Buffer B: 1 M ammonium acetate, 0.5 M P^ces of DNA were determined by UV 

ammonium hydroxide (pH 9). absorbance (OD) and radioactivity 

TABLE 11. Extent of Iodination of Sonicated E. coli DNA under 
Various Labeling Conditions 

Percentage of 

I Which is 

Percentage of 



Cytosine Which 




Is Iodinated 


60°C 15 min pH 5 




60°C 5 min pH 5 




60°C 15 min pB. 6 




40°C 15 min pH 5 








1 1 1 1 



A 1 1 


r a 


/o\ • 




> 4 1 



/ ' 1° 




/ / lo 





/ ° \ 



/ ' ,° 


J 1 1 \ 




/ / \ \ 



/ o I . 


/ / \ ft 


/ \ o 



o- - 

^" 1 1 1 





20 30 

40 50 



- 10000 




6000 S 





Fig. 25. Distribution of 125 I-labeled (open circles) and unlabeled (closed circles) E. coli DNA 
in an alkaline sucrose gradient. A mixture of iodinated and untreated DNA, in 0.2 ml of 0.1 A 7 
NaOH, was layered on top of a 5%-20% alkaline sucrose gradient in a 5-ml centrifuge tube, and 
the tube was spun in a SW39 head on a Spinco Model L ultracentrifuge at 30,000 rpm for 19 hours 
at 5°C. The bottom of the tube was pierced, and 67 fractions of 6 drops (0.07 ml) were collected. 
After adding 0.6 ml of water to each fraction, the UV adsorbance and counting rate were meas- 
ured. Fraction 67, at the top of the tube, contained the lightest molecules. 

measurements, respectively (see Fig. 25). 
The labeled DNA was found to be ap- 
proximately 80% as long as the un- 
treated DNA, in general agreement with 
the figure given above. There was no 
evidence for the existence of large num- 
bers of small labeled fragments. 

The thermal stabilities of the labeled 
and the untreated DNA were compared 
by applying a mixture of labeled and un- 
labeled E. coli DNA to a hydroxyapatite 
column at 60°C in 0.12 M PB and meas- 
uring the amounts of radioactivity and 
OD which were eluted as the temperature 
was raised stepwise. The thermal elution 
curve is shown in Fig. 26. It was found 
that 59% of the counts and 7% of the 
OD passed through the column at 60 °C 
without binding. Approximately 15% of 
the bound counts were associated with 
DNA which was eluted at low tempera- 
ture (less than 80°C). The difference in 
Te 50 (the temperature at which half the 
material has been eluted) between the 
labeled and unlabeled samples was ap- 
proximately 1°C. 

The ability of the iodinated DNA to 
form stable hybrids with normal DNA 
was investigated. A mixture of iodinated 
and untreated E. coli DNA (in the ratio 
2:100) was heated at 100°C for 5 min- 
utes and was incubated in 0.12 M PB 
at 60°C to C t 300. The DNA was then 
put on a HAP column in 0.12 M PB at 
60 °C, and the bound DNA was eluted 
with increasing temperature as before. 
Approximately 80% of the counts and 
95% of the OD bound to the column at 
60 °C. The thermal elution curve is 
shown in Fig. 27. Nearly all the labeled 
DNA was able to form stable hybrids, 
including the labeled DNA which had 
previously failed to bind to the column. 
The difference in Te 50 between the 
labeled and unlabeled products was 
approximately 1°C. 

A further study was made to determine 
why more than half of the counts from 
the iodinated DNA failed to bind to the 
HAP column in 0.12 M PB at 60°C, yet 
most were later able to form stable hy- 
brids with noniodinated DNA. A sample 














70 80 90 



Fig. 26. Thermal stability of iodinated E. 
coli DNA. A mixture of native (closed circles) 
and iodinated (open cirles) E. coli DNA was 
applied to a HAP column in 0.12 M PB at 
60 °C, and the bound DNA was eluted with in- 
creasing temperature. Approximately 59% of 
the counts and 7% of the OD did not bind. 
Te 50 was 93.5°C for the native DNA and 92.5°C 
for the iodinated DNA. 

of labeled E. coli DNA (iodinated under 
the same conditions as batch 1) was 
applied to a HAP column at 60°C in 0.12 
M PB, and the amounts of radioactivity 


70 80 90 



Fig. 27. Thermal stability of hybrids. A mix- 
ture of iodinated and normal E. coli DNA 
(ratio 2:100) was heat-denatured and incubated 
at 60° C in 0.12 M PB to Cot 300. The sample 
was then applied to a HAP column at 60° C, 
and the bound DNA was eluted with increas- 
ing temperature. Approximately 12% of the 
counts and 5% of the OD did not bind. Te 50 
was 93.5°C for the homologous product and 
92.5°C for the heterologous product. 

and UV absorbance eluting in 0.12 M PB 
were measured at 60°, 85°, and 100 °C. 
The results are shown in Table 12. Note 

TABLE 12. Differences in the Extent of 125 I-Labeling for Fractions of 
Labeled E. coli Eluting from HAP at Different Temperatures 

in 0.12 M PB 







of DNA, 


260 nm 

% of total 


% of total 


60°C (nonbind- 

ing fraction 


























that the nonbinding fraction (that elut- 
ing at 60°C) contained 43% of the radio- 
activity but only 6% of the UV- 
absorbing material. This fraction ap- 
peared to be labeled more than 10 times 
as extensively as the rest of the DNA. 

To test for the possibility that the non- 
binding counts were associated with 
RNA contamination, aliquots of the 
60°C and the 100°C fractions were in- 
cubated at 60° for three hours at pH 
12.5. The decrease in TCA precipitable 
counts was the same for both fractions, 
approximately 10% ; RNA associated 
counts were therefore negligible. It was 
therefore concluded that the nonbinding 
fraction was single-stranded DNA which 
had been labeled approximately 12 times 
as extensively as the double-stranded 
DNA. This conclusion is consistent with 
Commerford's finding that denatured 
DNA is labeled more than 10 times as 
effectively as native DNA. 

The source of the single-stranded 
DNA is uncertain. It is unlikely that the 
denaturation occurred during the iodina- 
tion reaction, since T m for E. coli DNA 
in buffer A was measured to be 81.7 °C, 
considerably higher than the temperature 
of the iodination reaction mixtures. The 
single-stranded DNA was probably pro- 
duced during sonication, since, as men- 
tioned above, approximately 6% of the 
unlabeled sonicated DNA did not bind to 
HAP at 60°C in 0.12 M PB. This per- 
centage of denatured DNA would prob- 
ably not be detected during an optical 

Preliminary studies of the extent of the 
single-stranded sections have been made, 
using the single-strand-specific nuclease 
Si, prepared by the method of Sutton. 48 
A quantity of 125 I-labeled E. coli DNA 
(from batch 2) was applied to a HAP 
column at 60°C in 0.12 M PB. The 
bound fraction (containing 88% of the 
OD and 31% of the radioactivity) was 
eluted with 0.4 M PB, dialized against 
water, lyophilized, and dissolved in 1 ml 
of 0.05 M sodium acetate buffer (pH 
4.5). To this solution, 10A of 10" 2 M 

ZnCl 2 and 5A of Si nuclease preparation 
were added. The solution was then in- 
cubated at 50 °C for one hour and 
chromatographed on a Sephadex G-100 
column with 0.12 M PB. The DNA 
which was digested (18% of the bound 
fraction, based on OD measurements) 
contained 54% of the 125 I; the digested 
DNA was therefore labeled approxi- 
mately five times as extensively as the 
intact DNA. Although further studies 
of the properties of the Si preparation are 
necessary before quantitative results are 
possible, it is evident that single-stranded 
sections are present in sonicated DNA 
and that these sections are labeled con- 
siderably more effectively than the 
double-stranded sections. 

In conclusion, Commerford's iodina- 
tion procedure is a convenient method 
for obtaining labeled DNA of high spe- 
cific activity and of near native size, 
thermal stability, and ability to hybrid- 
ize. To obtain uniformly labeled DNA 
with a well-defined specific activity, how- 
ever, precautions are necessary. The 
DNA sample being labeled must not 
contain both single- and double-stranded 
sections, since the single-stranded sec- 
tions will be preferentially labeled. 

Cooperative Research at 
California Institute of Technology 

A research program on the mechanism 
of regulation of genetic activity is being 
carried out at the California Institute of 
Technology under the joint direction of 
Eric Davidson, Associate Professor of 
Developmental Biology at Cal Tech, and 
Roy Britten, member of the Biophysics 
Section, Department of Terrestrial Mag- 
netism.* The work is being done both at 

* Other members of the group are : Senior 
Research Fellow Barbara R. Hough; Research 
Fellows Robert B. Goldberg. Dale E. Graham. 
William H. Klein, Berney R. Neufeld, Michael 
J. Smith; graduate students Christopher S. 
Amenson and Glenn A. Galau ; and research 
staff, Margaret E. Chamberlin. Dennis B. Coch- 
ran, Charles F. Collett, Michael Henerey, Marta 
Rico. Jane Rigg, Vasile Sapatino, Michelle T. 
West, and Paul H. Yancey. 



the main campus in Pasadena, California, 
and at the Kerckhoff Marine Laboratory 
at Corona del Mar, California. This 
program is in its second year and the first 
report of this program appeared in Year 
Book 71, p. 270. 

The focus of interest in these labora- 
tories continues to be on the molecular 
basis of gene regulation in higher orga- 
nisms and the operation of the genomic 
apparatus in oogenesis and early devel- 
opment. Much of our effort this year 
was devoted to analysis of DNA se- 
quence organization, since understanding 
sequence order is clearly prerequisite to 
understanding gene regulation. 

maining single stranded are single-copy 
sequences. These data indicate that 
there are two classes of interspersed 
repetitive sequences: a closely spaced set 
averaging about 1000 nucleotides' spac- 
ing and a more widely spaced set with a 
4000-nucleotide or greater spacing. The 
repetitive sequence elements themselves 
appear to average about 300 nucleotides 
in length. These observations are almost 
identical to those described last year for 
Xenopus DNA, even as far as the spac- 
ings and the quantities of DNA involved. 
It is astonishing and probably significant 
to find such a detailed and quantitative 
similarity between the genomes of an 
echinoderm and an amphibian. 

Interspersion of Sea Urchin Repetitive 
and Nonrepetitive DNA Sequences 

Dale E. Graham, Berney R. Neufeld, and 
Roy J. Britten 

Measurements of sequence intersper- 
sion have been made for the DNA of the 
sea urchin, Strongylocentrotus purnura- 
tus, by methods essentially the same as 
those used previously for Xenopus ( Year 
Book 71, p. 273). Labeled fragments of 
DNA of many different lengths were re- 
associated with an excess of short (450 
long) fragments to a Cot at which most 
of the repetitive DNA but no nonrepeti- 
tive DNA had paired. The fraction of 
the labeled fragments with paired re- 
gions, i.e., repetitive sequence elements, 
was measured (by binding to hydroxy- 
apatite) as a function of fragment length. 
About 25% of the DNA is actually made 
up of repetitive sequences so that for 
very short fragments about 25% of the 
DNA binds. Fractions containing repeti- 
tive sequences increase rapidly with 
length— up to about 55% for 1000- 
nucleotide-long fragments. Above that 
length the amount increases more slowly 
until 80% of 4000-nucleotide-long frag- 
ments contain repetitive sequences. In 
each case the parts of the fragments re- 

size Distribution of Interspersed Repeti- 
tive Sequences in Sea Urchin DNA by 
Reassociation and Single- Strand- 
Specific Nuclease Digestion 

Dale E. Graham, Michael Henerey, 
Roy J. Britten, and Eric Davidson 

In these experiments sea urchin DNA 
was sheared to about 2000 nucleotides in 
length, denatured, and reassociated to a 
C t which would pair most of the repeti- 
tive DNA elements but not the nonrepet- 
itive DNA. The DNA preparation was 
then digested with S-l nuclease (Ando, 
T., Biochim. Biophys. Acta, 114, 158, 
1966) under conditions such that duplex 
structures that would not be melted at 
60 °C, 0.12 M phosphate buffer were pre- 
served and less stable structures were 
digested by the enzyme. At C t 20, about 
25% of the DNA was enzyme resistant. 
One quarter of this DNA (6% of the 
total) was in fragments greater than 
2000 nucleotides long. A smaller amount 
of DNA was distributed in a range of 
sizes between 500 and 200 nucleotide 
pairs, with the majority of the resistant 
DNA fragments present in a peak of 
about 300 nucleotide pairs. The single 
strands that make up the 300-long du- 
plexes were also about 300 nucleotides 



long. Thus, relatively few of the individ- 
ual strands had been attacked at internal 
regions of imperfect base pairing. The 
long fragments consisted of single- 
stranded DNA with the same size dis- 
tribution as the original sheared DNA 

These size distributions show that 
there are extensive regions of repetitive 
sequences which we have termed "clus- 
tered repetitive regions." This DNA 
gives rise to the long, enzyme-resistant 
DNA fragments. The short (300 nucleo- 
tide, average) enzyme-resistant fragments 
represent the repetitive sequence ele- 
ments that are interspersed throughout 
the genome. Their size, when isolated by 
S-l nuclease treatment, is about the same 
as that deduced for the short-period 
interspersed repetitive sequences men- 
tioned above. 

Similar treatment with S-l nuclease of 
Xenopus DNA reassociated to a repeti- 
tive C t confirmed that, in this organism 
also, the repetitive sequence element is 
about 300 nucleotides long, measured 
either as duplex or after denaturation. 

Samples of each of the size classes of 
enzyme-resistant, reassociated, double- 
stranded sea urchin DNA were melted in 
the spectrophotometer. As expected, the 
hyperchromicity was approximately that 
of native DNA. The thermal stability 
of the enzyme-resistant regions was cor- 
related with their length. The fraction of 
greatest length (excluded peak) melted 
within one or two degrees of the tempera- 
ture at which long native DNA melts. 
Each succeeding smaller size fraction 
melted at a lower temperature, down to 
the 300-base-pair, long-peak regions, 
which melted as much as 15° below na- 
tive long DNA. The reduction of melting 
temperature observed is too great to be 
due simply to the length of the duplex 
region and must result from increasing 
sequence divergence. There is a clear 
correlation between the amount of se- 
quence divergence among the repetitive 
sequences and their length. 

Hyperpolymers and Another Approach to 
Repetitive Sequence Organization 

Dale E. Graham and Roy J . Britten 

It has been known for many years that 
the reassociation of repetitive DNA leads 
to the formation of extended structures 
termed "hyperpolymers." This happens 
because the DNA fragments are ran- 
domly broken and two complementary 
strands that pair seldom terminate at the 
same place. Thus, after the first pairing 
there remain single-stranded ends which 
can pair again with other fragments. The 
resulting structures may be very large 
after a reassociation reaction is near 

Recent observations indicate that in- 
terspersed repetitive sequences do not 
form extended hyperpolymers. In these 
tests 300-nucleotide-long fragments were 
dissociated and incubated to a C () t 50 or 
100 times that required for half reasso- 
ciation. The size of the reassociated 
DNA was assayed on an Agarose A50 
column. Treated in this way, T4 viral 
DNA is mostly excluded, indicating that 
most of the DNA is incorporated in 
structures containing at least 20 frag- 
ments. The same experiment was done 
with total sea urchin DNA and with iso- 
lated highly repetitive and middle repet- 
itive DNA fractions. In each case very 
little of the DNA was excluded from the 
A50. The excluded DNA quantity cor- 
responded to that known to be in clusters 
from the measurements described above ; 
clustered repeated DNA is expected to 
form hyperpolymers. The remainder of 
the duplex structures consists of only 
pairs or tetramers of the original frag- 
ments. This result confirms the fact that 
most of the repetitive sequences are pres- 
ent as interspersed short-sequence ele- 
ments whose length is about that of our 
fragments, i.e., about 300 nucleotides. 
Furthermore, it shows that these ele- 
ments are internally nonrepetitive with 
homologous elements beginning and end- 
ing at homologous positions. 



General Assessment of Advances in the 

Knowledge of DNA Sequence 


Dale E. Graham, Berney R. Neujeld, and 
Roy J. Britten 

During the year, a number of related 
observations have been made of sequence 
arrangement in the DNA of several 
species, including the marine worm 
Urechis, a sea urchin, an amphibian and 
a mammal. Although there are differ- 
ences in genome size and repetition fre- 
quency among these animals, neverthe- 
less there are common features and a 
pattern emerges from the comparisons: 
in all of these species (1) the majority of 
the repetitive sequences are interspersed 
with nonrepetitive sequences; (2) a 
minor component of clustered repetitive 

sequences exists; (3) the clustered re- 
gions extend for at least several thousand 
nucleotides, while (at least for the three 
deuterostomes) the length distribution of 
the interspersed repetitive sequences has 
a peak at about 300 nucleotides; (4) the 
clustered sequences display relatively 
precise sequence repetition, and the inter- 
spersed sequences show divergence rang- 
ing from 5% to 20% or more. 

This evidence shows that there are 
general patterns of organization of the 
repeated and single-copy sequences in 
living systems and helps to establish the 
foundation for our current investigations 
of nuclear RNA and polysomal messen- 
ger RNA. These patterns appear almost 
certain to lead us to a deeper understand- 
ing of the functional organization of the 


L. T. Aldrich, L. Beach, K. D. Burrhus, R. C. Fletcher, A. Cuyubamba, S. E. Forbush, 

B. Grauert, S. R. Hart, A. Hofmann, D. E. James, A. T. Linde, H. Okada, 

I. S. Sacks, N. Shimizu, D. Simoni, T. J . Smith, and A. Snoke 


C. A . Andersen, J. Bannister, F. R. Boyd, C. Brooks, I, Casaverde, D. Comajord, G. L. Davis, 

M. L. Crawford, S. del Pozo, D. W. Evertson, A. Flores, A. A. Giesecke, B. J. Giletti, 

L. Hall, J. R. Hinthorne, T. E. Krogh, I. Kushiro, Y. Motoya, J. Mattinson, 

L. Ocola, G. Olajsson, R. Quiroga, J. E. Ramirez, A. Rodriguez B., 

R. Salgueiro, J. G. Schilling, M. G. Seitz, H. Sigtryggsson, 

S. Suyehiro, L. Tamayo, E. Triep, M. A. Tuve, 

V . Volponi, M. E. Wagner, and Y. Yamagishi 

L. T. Aldrich 

For several years there have been 
comments in this introduction extolling 
the opportunities at Carnegie for express- 
ing one's creative concerns in earth sci- 
ence on a global scale. We have implicitly 
and, on occasion, explicitly expressed 
both our appreciation for these opportu- 
nities and our hope that we are using 
them fruitfully. This year I would like 
to outline briefly some of the collabora- 
tive efforts making up the fabric of the 
work in earth science at the Department. 
The most extensive program is that of 

Selwyn Sacks, whose broad-frequency- 
band, high-dynamic-range seismometers 
occupy a seven-station net, ranging from 
New Guinea to Iceland with two stations 
each in Peru and Japan. He even has one 
on the DTM campus! Without the help 
of devoted scientists and technicians in 
these remote locations, these observations 
would be impossible. Among these col- 
laborators, Dr. S. Suyehiro in Japan 
should be specially mentioned for his 
continued interest. The contributions of 
another of our collaborators, Ing. D. 
Simoni of Peru, have been recognized by 
his appointment as a Staff Member of 
the Department. 



Dr. Suyehiro and Mr. Dale Evertson, 
of the University of Texas, Austin, are 
full partners with Sacks in the installa- 
tion and testing of a tripartite net of 
Sacks' volume strainmeters in Japan, the 
fruits of which are yet to be appreciated 
fully. David James includes among his 
collaborators not only those listed in his 
reported geochemical studies below but 
also numerous geologists in the geological 
surveys of Peru and Bolivia. The inter- 
est of local men in these studies provided 
physical and logistic support during the 
collecting activities this year. My elec- 
trical conductivity activities are based 
on the commitments of Mateo Casaverde, 
Lupe Tamayo, John Bannister, and Rey- 
naldo Salgueiro — all of whom are install- 
ing and operating magnetic variographs 
in their home countries as this is written. 

Stanley Hart has for several years 
been interested in samples of ridge ba- 
salts as indicators of process in the mod- 
ern theory of global tectonics. Such 
samples are collected only by large-scale 
oceanographic expeditions. A measure of 
the value he places on these samples is 
expressed by the fact that he will act as 
co-chief scientist on such an expedition 
in the coming report year. 

Finally, James, Sacks, Linde, Simoni, 
and I were a part of a five-nation group 
that organized and executed a major geo- 
physical study of the region between 
2° S and 4° N latitude, and 74° and 82° 
W longitude this year. We and our col- 
leagues in North and South America were 
joined by a German team of specialists 
in explosion seismology to complete 
seismic studies all over the area and 
magnetic and gravity observations at 
sea. Of special importance to the suc- 
cess of our plans were the efforts of Dr. 
J. E. Ramirez, S. J., in Colombia. 

It is anticipated that we will all be 
interested in continuing our far-flung 
activities in earth science. We again ex- 
press our appreciation for the support of 
these studies from both the Carnegie 
Institution and the National Science 

Cosmic-Ray Research 
S. E. Forbush and L. Beach 

The variation, with -period twice that 
of the sunspot cycle, in the cosmic-ray 
diurnal anisotropy. As described in pre- 
vious Annual Reports, the annual means 
of the total cosmic-ray diurnal aniso- 
tropy vector from Carnegie Institution 
of Washington ionization chambers are 
well represented as the sum of two com- 
ponents, V and W, where V and W are in 
the asymptotic directions 90° and 128° 
east of the sun, respectively. The com- 
ponent V varies with the period of the 
sunspot cycle and is well correlated with 
magnetic activity. When data were avail- 
able only through 1965, annual means of 
W were found to be well fitted by a sine 
wave with a period of two solar cycles, 
which amounted to twenty years during 
the interval covered by the data. This 
wave passed through zero in 1958, which 
is near the time that Babcock found, 
from regular measurements begun in 1953 
at Mount Wilson Observatory, that the 
north polar magnetic field of the sun 
changed from positive to negative. 

The interval between successive sun- 
spot maxima (or minima) varies from 
about 9 to 13 years. Thus, within the 
interval from 1937 to about 1965, it is 
fortuitous that this sunspot period re- 
mained close to 10 years and that a wave 
with a period of 20 years fitted W. 

Thus, not unexpectedly, the observed 
points (squares) in Fig. 28 after about 
1966, definitely depart from the indicated 
sine wave, W s , that fitted quite well the 
observed points (circles) until about 
1965. Figure 28 shows a "guesstimated" 
curve, Wg, fitted to observed values after 
1965. W G shows that the component W 
changed sign shortly after the beginning 
of 1971. 

From continuing measurements at 
Mount Wilson Observatory, Howard 
found that the north polar field of the 
sun, having been negative since 1958, be- 
came significantly positive about the 



6 2> 

~ E 

5 - 


O / r 

\ O 

," ' \ 

' \ 

/ \o 




/ o 

/ ° 


V) o 


I ° 
/ o 



o, o .Annual means of W * 

Wg : Sine wave with period 

10 years \ 

Wq :"Guesstimated" 
variation in W 
after 1966 




Fig. 28. Annual means, 1937-1972, of the component W of the diurnal anisotropy in the asymp- 
totic direction 128° east of the sun. 

middle of 1971. Thus the two observed 
reversals in the sun's polar magnetic field 
since observations began in 1953 occurred 
near the time of the last two reversals in 
the sign of the component W of the 
cosmic-ray diurnal anisotropy. 

Although the mechanism responsible is 
still not understood, this variation of W 
with a period twice that of the sunspots 
is probably the first terrestrial effect ob- 
served to be related to the variation in 
the direction of the sun's polar magnetic 
field. The components V and W combine 
to give an excellent fit to the observed 
annual mean diurnal anisotropy. It is 
evident that this satisfying aspect could 
only have been uncovered from a long 
series of records. 

Observations and reduction of data. 
Cosmic-ray ionization chambers were 
operated throughout the report year at 
Huancayo, Peru; at Fredericksburg, Vir- 
ginia; and at Christchurch, New Zealand. 
Scalings and reduction of records have 
been maintained on a current basis for 
these three stations. The reductions have 
been greatly facilitated by the use of the 
IBM 1130 computer. 

Cooperation in operation of cosmic-ray 
meters. Grateful appreciation is ex- 
pressed to the National Oceanic and 
Atmospheric Administration and the 
staff of its magnetic observatory at 
Fredericksburg for efficient operation of 
the cosmic-ray meter there during the 
past report year, and to the Government 
of Peru and the Director and staff of the 
Instituto Geofisico del Peru for making 
cosmic-ray records from Huancayo avail- 
able. Grateful appreciation is also ex- 
pressed to the Director and staff of the 
Geophysical Observatory at Christ- 
church, New Zealand, for excellent main- 
tenance of the equipment there and for 
the fine records obtained. 

A Comparison of the Anelasticity 

Structure between Western 

South America and Japan 

/. S. Sacks and H. Okada 

Although the driving forces for sea- 
floor spreading and continental drift are 
largely unknown, these motions are cer- 
tainly influenced by the viscosity struc- 
ture of the crust and upper mantle and 
perhaps also by that of the lower mantle. 



Viscosity in the earth is difficult to de- 
termine directly but may be estimated 
from the anelasticity (Q _1 ) even though 
at this time the formal relationship be- 
tween these quantities has not been 
established. Anderson (1966) 49 and 
Sacks (1972) 50 have proposed empirical 

South America and Japan, the two re- 
gions studied, have broad similarities 
(ocean trenches, dipping seismic planes, 
volcanoes) and important differences 
(South America is a continent, whereas 
the Japan region is an island-arc type 
structure). In addition to these differ- 
ences, the seismicity is different in the 
two regions. In Japan the seismicity is 
restricted to a dipping seismic plane and 
shallow activity west of the ocean trench. 
In South America this dipping seismic 
plane is more (northern Chile) or less 
(central Peru) obvious, but the wedge 
above the dipping plane contains earth- 
quakes, relatively few in northern Chile 
but many in central and northern Peru. 
In Japan the seismic activity is con- 
tinuous down the dipping plane, though 
the most energy is released by the shal- 
low quakes and the least energy by 
events of intermediate depth. 

Above the deep earthquakes in South 
America, however, there is an aseismic 
zone which extends at least from Ecuador 
to central Argentina (some 30 degrees of 
latitude) and covers the depth range 350 
to 530 km. The energy released by deep- 
focus events is comparable to that from 
shallow events. Thus it is likely that a 
comparison of their anelasticity struc- 
tures should yield information pertinent 
to large-scale earth movements. 

In order to make valid comparisons of 
Q structures, the fewest possible assump- 
tions were made and identical techniques 
were used in the two regions. Since the 
radiated spectrum of an earthquake de- 
pends on its magnitude and location, this 
variable has been eliminated by use of 
spectral ratio techniques for reliable 
quantitative Q determinations (Sacks, 

1968). 51 Nevertheless, spurious effects 
(e.g., different crustal transfer functions 
at different stations) necessitate calcula- 
tion of the spectral ratios over a wide 
frequency range. The results presented 
here derive from recordings by specially 
developed wide frequency range seismo- 
graphs (Sacks, 1966) r>2 which have been 
installed in the two regions. 

Summary of results. Q results from 
the South American region, areally 
(Sacks, 1969) 53 and as a function of 
depth (Sacks, 1971) , n4 have been re- 
ported previously. In essence, the South 
American Q structure consists of a 
high-Q continental block about 350 km 
thick overlying a variable low-Q layer 
of uncertain dip. The Q s of the thick 
block increases with depth from a value 
of 500 for the upper 100 km to 1500 at 
350 km. There are, however, important 
variations in this block. Some values in 
the northern Chile— northwestern Argen- 
tina region are less than one-half of those 
farther north. The presence of a down- 
going lithospheric slab has not yet been 
detected from the Q study or from time 
residual studies. This does not mean that 
such a slab does not exist in the South 
American region but that its Q structure 
is similar to that beneath the continent 
since possibly neither region contains sig- 
nificant partial melt. We believe that 
deep continental roots are not confined to 
South America though anelasticity evi- 
dence is confined to this continent at 
present. Studies of surface wave disper- 
sion across shield areas in Africa and 
South America by Alexander (1972)"' 
show an absence of velocity reversal in 
the uppermost mantle (circa 200 km) in 
both cases. A velocity reversal is found 
commonly in oceanic areas and occurs at 
depths of approximately 70 km. 

The situation beneath Japan is some- 
what different. Figure 29 shows two 
seismograms from earthquakes at depths 
of about 285 km. The shear arrival at 
MAT in Japan shows little high-fre- 
quency energy, i.e., the path from the 



MAT, 29 AUG. 1970, M*5.2, DEPTH = 284KM, A=7.I0 

0.1 sec 

CUZ, 16 DEC 1968, M«4.l, DEPTH* 283, A=6.4 

Fig. 29. Analogue filtered seismograms indicated the effects of anelasticity in the upper 280 km. 
Both seismographs, MAT in Japan and CUZ in Andean South America, are situated well above 
the dipping seismic plane (160 to 200 km) in roughly equivalent positions relative to the tecton- 
ics. The earthquakes are at the same depth, circa 285 km. The MAT seismogram (top) shows 
that the shear arrival, while of similar amplitude to the compressional (P) at lower frequencies 
(upper traces), has little energy at the higher frequencies. In contrast, the record from CUZ 
(bottom), shows considerable energy at the high frequencies for the shear arrival. This is because 
the region above the dipping seismic plane in South America is generally of high Q. 



earthquake to the seismograph must be 
low-Q. The shear arrival recorded at 
CUS in South America, which is simi- 
larly situated relative to the dipping 
seismic plane and the volcanic ranges, 
shows high frequencies and therefore a 
high-Q path from the same depth. The 
gross features of the anelasticity struc- 
ture can be seen by comparing the areal 
distribution of average Q from the deep 
earthquakes to the surface. The results 
can be seen in Fig. 30. For Japan, mean 
intensity contours from deep earth- 
quakes, determined by Utsu (1967) , 56 
have been plotted on a map showing the 
volcanic front. There is substantial simi- 
larity between the shape of the intensity 
contours and the form of the volcanic 
front. The correlation between the inten- 
sities and average Q values has been con- 
firmed by a study similar to the one done 
for the South American region. No ex- 
ceptions were found to the observation 
that there are high-Q paths to the seis- 
mographs on the east coast and low-Q 

paths to the stations on the west coast, 
i.e., to those 150 to 200 km above the 
seismic plane. 

In South America, on the other hand, 
high-Q and low-Q paths from deep earth- 
quakes are not related in any simple way 
to the tectonics as expressed by surface 
features. This is because the average Q 
of these paths is dominated by variations 
in a low-Q region between depths of 
350 and 530 km. The structure between 
the deep earthquakes and the surface is 
therefore more regular beneath the Japan 
region than beneath South America. The 
Q structure of the dipping seismic zone 
(descending slab) was determined mainly 
from the broad-band seismograph at 
Kamikineusu (KMU) on the Pacific coast 
of the north island, Hokkaido. All paths 
from earthquakes in the dipping plane 
north of central Honshu to this sta- 
tion have high Q. By studying earth- 
quakes at various depths in the seismic 
plane, it was found that the average Q 
to the seismograph increases with depth 


ntensrty Anomalies 

Caused by 
Deep Earthquakes 

Volcanoes •»• 

Trenches Jzw 

Fig. 30. Areal distribution of the average Q structure between the deep earthquakes and the 
surface beneath Japan and South America. Intensity anomaly contours (after Utsu, 1967) are 
drawn. Intensity anomalies reported from deep earthquakes (after Utsu, 1967) have been con- 
toured on a map of Japan (left). The higher numbers indicate greater vibration and therefore 
higher average Q. The correlation of the intensity contours with the form of the volcanic front 
is apparent. Beneath South America no such simple pattern exists (Sacks, 1969). The variations 
in average Q are caused by variations in a region of generally low Q in the depth range 350 to 
530 km. 






Fig. 31. Sketch showing ray paths and seismic geometry beneath Japan. The two ray paths 
shown were found to have identical average Q, even though the path to MAT is much longer in 
the aseismic wedge than that to KMU. Therefore, the absorption in the section "A" in the aseis- 
mic wedge must be exactly compensated by that beneath the seismic plane "B," i.e., the average 
Q structure on either side of the high-Q seismic plane must be the same. 

of the earthquake. Inversion to actual 
Q p shows that it increases from a value 
of approximately 1000 near the surface 
(100 km), to 2000 at about 200 km, to 
more than 3000 below 400 km. We do 
not yet have a good constraint on the 
thickness of this dipping high-Q zone, 
but, from a simple ray path considera- 
tion, it must be either fairly thick (more 
than 100 km) or have special interfaces 
at upper and lower boundaries to effi- 
ciently reflect the seismic energy at near 
grazing incidence. 

The lithosphere beneath Japan has not 
been studied yet, but the appearance of 
seismograms of elastic waves traveling in 
it suggests that it is of high Q. 

The Q values of the aseismic wedge 
between the surface and the dipping seis- 
mic plane were determined mainly from 
the broad-band seismograph in Matsu- 
shiro, central Honshu (MAT) , again by 
studying earthquakes situated at various 
depths in the seismic plane. The average 
Q p of this region was found to be 400. 
We have not yet determined how much 
structure there is in the anelasticity of 
this aseismic wedge. It is of interest to 
note that no exceptionally low-Q regions 
were found in the vicinity of the vol- 
canoes. Asamayama, one of the nearest 

(30 km) to the seismograph (MAT), is 
presently active. This is consistent with 
no observations of low Q in the vicinity 
of volcano El Misti (near Arequipa) in 
southern Peru, South America. Farberov 
et al. (1971 ) 57 actually measured the ab- 
sorption of elastic waves beneath vol- 
canoes in the Kamchatka region and 
found a rather narrow high-absorption 
zone. This suggests that there is no 
great magma pool beneath the volcanoes. 
Instead, there must be a narrow pipe 
transporting the magma from depths. 

Comparison of the Q values on either 
side of the dipping seismic plane yields 
the interesting result that they are iden- 
tical. Figure 31 shows a schematic of the 
method used for this comparison. Spec- 
tra from deep events below Fiji were 
compared at the two broad-band instal- 
lations in Japan. The observed spectra 
whose ratio is shown in Fig. 32 were 
identical over a considerable frequency 
range (0.1 to 5.0 cps), indicating that 
absorption in the longer path to MAT in 
the aseismic wedge is exactly compen- 
sated by that in the path to URA below 
the seismic plane. This is confirmed by 
comparison of spectra of near vertical 
shear waves ScS at MAT and URA. The 
spectra are essentially identical in slope. 














9 Oct. 1967 Fiji 


P(MAT) / P(KMU) 

M(ISC)= 6.2 H = 654 KM 






~ o 

CD CcF° O 

o o 

° °o o 

° o o ° ° 


D o & 
°° o ° 


-° ocP ^o 


O o 


- °o° 
_ o 


•/' ■ W- :& 



/ ■./fj^zmimb 

\' "' 


X mA.t^^ 'K 

1 1 












Fig. 32. The spectral ratio as a function of frequency of the compressional wave (P) arrival at 
MAT to KMU from a deep earthquake in the Fiji region. The ray path geometry is shown in Fig. 
31. A gross slope of the spectral ratio would indicate a difference in the Q values of the two paths 
compared. Undulations in the curve are caused by dissimilarities in the crustal structure of the 
two seismograph stations. 

Thus, at least for the upper 250 km or so, 
materials which may be released from the 
downgoing slab are not of a type or 
quantity to affect the gross anelasticity 
of the wedge above the slab. 

At greater depths, however, much 
lower Q values are found. Spectral ratios 
of P P vs. P from the Fiji deep earth- 
quakes recorded at MAT and URA gave 
Q p values of 100. A deep earthquake in 
the Okhotsk Sea region recorded at the 
broad-band seismograph at Port Mores- 
by, New Guinea (PMG) also gave a Q p 
of 100 from a P P vs. P spectral ratio cal- 
culation. The spectral ratio of sScS vs. 
ScS at MAT from the same earthquake 
yielded a Q s of 50, which is in good agree- 
ment with the Q p values. Above the 
South American deep earthquakes, Q p is 
350 from P P/P and Q s is 160-200 from 

sScS/ScS and also from ScS/SKP. These 
generally higher Q values are due to the 
very thick lithospheric section (or ab- 
sence of the low-Q wedge) which has a 
high Q for the 350 km below the surface. 
Very low (70-200) Q p values were found 
in the 350 to 530 km aseismic zone be- 
neath South America, but these appear 
to be of lesser extent. 

Discussion. The anelasticity struc- 
tures of the two regions are compared in 
Fig. 33. The similarities allow us to 
make certain conjectures. We have found 
that Q p values in regions which have 
earthquakes, even if the seismicity is low, 
are above 1000. These regions must be 
able to store strain energy, i.e.. have very 
high viscosity and probably very little, if 
any, partial melt. The Q of such a mate- 
rial increases with pressure (depth) and 




Trench ^ ^ ^ 

200 400V^600 800 - 1000 1200 1400 KM 

Fig. 33. The Q P structure beneath Japan (left) and South America (right). Dots indicate areas 
of high Q (1000-3000); slant lines, intermediate-Q values (300-500); and cross-hatching, low-Q 
values (50-100). The small numbers in the dotted region indicate earthquake hypocenters. Seis- 
mograms to the stations indicated, MAT and CUS, are shown in Fig. 29. The region above the 
dipping seismic plane is of intermediate Q in Japan, whereas in South America it has high Q. The 
high-Q slab in Japan is continuous down to the deepest earthquakes; in South America there is 
a substantial decrease in Q below 350 km even though there are earthquakes 600 km deep. 

is probably not much affected by rock 
type. The increase of Q with depth in the 
dipping slab beneath Japan was similar 
to that in the continental block beneath 
Cuzco in Peru, which is well east of any 
possible slab. Absence of earthquake 
activity does not necessarily imply low Q 
since sufficient tectonic stress must also 
be present. 

Q p values for the aseismic wedge be- 
neath Japan are as high as 400, and we 
assume that the viscosity is too low to 
store strain energy, presumably due to 
some partial melt. Far lower Q values 
are possible, of course, and have been 
found above the deep earthquakes in Fiji 
and Japan and in pockets in South Amer- 
ica. The tectonic significance of the very 
low Q's above the deep earthquakes is 
not clear to us at this stage, but its com- 
mon existence suggests that it is in some 
way fundamental. 

Some tectonic implications suggested 
to us by these results are illustrated in 
Fig. 34. For island arcs, such as Japan, 
the normal plate tectonic model of a 
relatively thin (—70 km) slab dipping 
into the upper mantle from the vicinity 

of an oceanic trench in an "oceanic plate" 
(or possibly "oceanic plate— marginal sea 
plate") interaction is adequate and is 
shown on the right-hand side of the 

Fig. 34. Tectonic implications of the Q struc- 
ture. Oceanic lithosphere is 70-100 km thick, 
subducting at trenches as confirmed by the Q 
structure beneath Japan. Continental litho- 
sphere, however, is much thicker (350 km be- 
neath South America) and penetrates below 
what is traditionally considered asthenosphere 
(the weak layer). It is probable, therefore, that 
continental roots are more strongly coupled to 
the higher rigidity (mesophere) region of the 
mantle than are the oceanic plates. Whole 
mantle convection would therefore drive con- 
tinents strongly (e.g., South America), just as 
lack of whole mantle convection beneath a 
continent (possibly Africa) would tend to 
anchor it. 



figure. However, for "oceanic plate- 
continent" interactions, the difference is 
that the continent has a very thick (350 
km), high rigidity "root." The shape of 
the subduction zone is forced by the 
shape of the continent before subduction 
starts and will in general not be the ac- 
curate shape found commonly in ocean- 
ocean interactions. 

Empirical Models for Anomalous 

High-Frequency Arrivals from 

Deep-Focus Earthquakes in 

South America 

A. Snoke, I. S. Sacks, and H. Okada 

Anomalous high-frequency arrivals 
(AHFA) from deep-focus earthquakes 
have been observed for epicentral dis- 
tances of 13-21° in Japan, 58 Tonga, 59 - 60 
and South America. 61,62 Their arrival 
time is up to a minute later than the 
direct S; their path has a higher Q than 
that of direct S; their duration is longer 
than that of direct P or direct S; and 
they have an apparent phase velocity of 
4.7 km/sec. If Jeffreys' velocity-depth 
model for the mantle is used, a phase 
velocity of 4.7 km/sec corresponds to an 
S phase which has a maximum possible 
depth of 270 km. The fact that the rele- 
vant foci are as much as 300 km below 
this depth and that the path has anomal- 
ously high Q implies a radically different 
path for an AHFA from those for normal 
arrivals. A feature common to all the 
geographical regions mentioned above is 
that lithosphere descends towards the 
focal region. On the basis of seismic ac- 
tivity and Q studies, the lithosphere is 
believed to be continuous down to the 
relevant focal depths in all regions except 
in South America where there is no evi- 
dence of lithosphere between a depth of 
about 350 km and the focal depths of 
540-600 km. 

Isacks and Barazangi 60,62 have in- 
troduced a model which satisfies the 
travel time and Q data for AHFA ob- 
served at NNA. In this model the de- 

scending lithosphere must be taken as 
continuous down to the focal depths. Ac- 
cording to their model, an AHFA results 
from a phase which propagates from focus 
to station by traveling inside the litho- 
sphere as a guided wave and then con- 
verting to an Sn phase near the earth's 
surface. The wave guide is obtained by 
postulating that the upper surface of the 
lithosphere is a perfect reflector for *S 
phases and by assuming a velocity gradi- 
ent increasing from this surface inwards. 
Assuming the uniqueness of their model 
in its ability to fit the data, Isacks and 
Barazangi have used their model as 
evidence for the existence of structure 
and phenomena heretofore not observed 
in South America, such as the continuous 
lithosphere, a totally reflecting upper 
surface of the lithosphere for S phases 
and the corresponding wave guide. 

We have also been studying South 
American AHFA and are seeking a model 
which can provide a satisfactory expla- 
nation for them. Our data come from 
the DTM broad-band stations CUS and 
TRU and from the WWSSN stations 
NNA and ARE (see Table 13). The 
NNA data are the same as those used by 
Isacks and Barazangi, and the ARE 
data are cited by them but not used. The 
locations of these stations and the loca- 
tions of events that gave rise to clear 
AHFAs observed at at least one station 
are shown in Fig. 35. A graph of travel 
time versus epicentral distance for ob- 
served AHFA is given in Fig. 36. For 
five of these twelve events, AHFAs are 
observed at two stations. For some 
events, including three of these twelve, 
there are potentially usable data from 
stations for A less than 13° or greater 
than 21°, but in these cases we observe 
no AHFA near the anticipated travel 
time. It is difficult to distinguish between 
the direct S and an AHFA at the shorter 
distances because the time difference be- 
tween S and the AHFA is so short, while 
for the longer distances there is simply no 
observable high-frequency arrival within 



TABLE 13. AHFA-Producing Events as Plotted on Figure 36 


Magnitude Depth, km Station 

Travel Time, 
A, degrees sec 

Dec. 9, 1964 






Mar. 5, 1965 






May 13, 1965 






Dec. 11, 1965 






Dec. 20, 1966 









Jan. 17, 1967 








483 ±2 

Sept. 9, 1967 






Jan. 31, 1968 









May 11, 1968 






Aug. 23, 1968 








491 ±5 






July 25, 1969 









30 S 

80 W 

Fig. 35. Map showing the locations of stations and AHFA-producing events in South America. 
Also included are projections of the conversion region for Models A and B (see text) and projec- 
tions of seismic activity for depths of 100-250 km (numbers) and for depths of greater than 500 
km (open triangles). The AHFA-producing events are solid triangles. 










1 1 1 


i 1 

' ' 'l / 

• NNA 


■ TRU 


( DEPTH = 580 KM) 

X* / 

r i i 

i i 

1 1 1 

14 16 18 



Fig. 36. Total travel time versus A for AHFA 
observed at NNA, ARE, CUS, and TRU. The 
straight line drawn through the data points has 
a slope corresponding to a phase velocity of 4.7 

a couple of minutes of the arrival of 
direct S. 

We have attempted to formulate 
models that require no geological struc- 
ture not observed independently. Our 

two most successful models to date have 
the following common features: a phase 
(phase 1) travels from the focus to some 
point where it is converted to another 
phase (phase 2) which travels to the ob- 
serving station. The paths of these 
phases are restricted by the condition 
that the resultant arrival should have a 
path of higher Q than that of the direct 
S. We have been guided by Sacks' find- 
ings 63 of a low-Q region in South Amer- 
ica between depths of 350 km and 500 
km. Accordingly, phase l's trajectory 
through this low-Q region should be 
steeper than that for direct S, and phase 
2 should have a maximum depth of less 
than 350 km. A long duration for the 
AHFA can be explained by the possibil- 
ity of nonplanar paths and irregularities 
in the interface producing the conversion. 
This interface is discussed below. Our 
two models differ in the proposed paths 
for phase 1 : In Model A, phase 1 travels 
directly from the focus to the conversion 
point; while in Model B, phase 1 is re- 
flected from the earth's surface before 
arriving at the conversion point (see Fig. 
37) . In both models we have left the 




)/ 15 

* 300 





: '*fe%i/T\_~l 


-*j?w ; > 

[iA. ' . 



30° S 

l^fe™ to 


Fig. 37. The paths according to Models A and B for an AHFA observed at NNA from the 
event on May 13, 1965, superimposed on a seismicity plot for the region between the event and 




conversion process unspecified at present. 
We chose as the first test for the ac- 
ceptability of a model its ability to pro- 
duce total travel times which fall along a 
straight line whose slope is the observed 
phase velocity for an epicentral range of 
13-21° (see Fig. 36). In Table 14 the 
results for our most successful models are 
presented for a Model A case of a P 
phase converting to an S phase and for a 
Model B case of a P phase that remains 
a P phase after reflection from the earth's 
surface but is then converted to an S 
phase at the conversion point. The final 
column in Table 14 gives the time differ- 
ence between the total travel time cal- 
culated according to the model (using 
Jeffreys' velocity-depth model for S 
phases and Herrin's model for P phases) 
and the time calculated from the straight 
line through the data points in Fig. 35. 

We note that for both models this time 
difference is only a few seconds (out of 
about 400) throughout the range of in- 
terest. This result is fairly insensitive to 
the choice of the particular velocity 
models chosen or to the choice of the 
depths for the focus and the conversion 
point; the depths used in Table 14 are 
589 km and 150 km, respectively. 

If we assume that the boundary of the 
lithosphere is near the boundary of the 
seismically active region, a test of our 
models derives from the correlation be- 
tween the location of conversion points 
predicted by the models and the bound- 
aries of the seismically active region. 
Projections of possible conversion regions 
for Models A and B are shown in Fig. 
35 along with the seismicity for depths of 
100-250 km between the AHFA-produc- 
ing events and the stations, and Fig. 37 

TABLE 14. Results from Models A and B for a Focus at 589 km and a Conversion at 150 km 






Angle at 

Angle from 

Depth of 





Phase 2, 






A 2 

A total 


Model A: P to S 









































































- 9.2 

Model B 

: P to P to S 





































































- 9.2 





Angles in degrees are measured from the horizontal and are positive for upward-directed rays. 



includes a vertical seismicity plot for the 
region between a particular event and 
NNA. These figures show that the con- 
version points are indeed in the vicinity 
of the eastern boundary of the inter- 
mediate-depth seismic region. 

A further test of our models derives 
from an examination of the Q along the 
proposed path for each event producing 
an AHFA. Figure 38 contains a photo- 
graph of a seismogram showing the 
AHFA at NNA from the event of Fig. 
36 along with a photograph of a seismo- 
gram showing the direct S arrival at 
NNA for an event, rj, which lies along the 

path of phase 2 for either model (as in- 
dicated by the arrow in Fig. 36). The 
path length for the direct S from the 
event -q includes about 40% of the total 
path length, resulting in an AHFA for 
either model. It is essential to our models 
that this be a high-Q path; the high- 
frequency content of the S arrival from rj, 
compared to that of the AHFA, is con- 
sistent with this requirement. 

Because the data for AHFA, such as 
travel time and Q, can be fitted by widely 
differing models, further tests are re- 
quired to find the operative model for a 
given geographical region. (It is not in- 

NNA 13 MAY 1965 
h = 589 KM A =14.52° 

NNA 26 MAY 1970 
h= 159 KM A = 7.2° 



Fig. 38. Photographs of seismograms from NNA for an AHFA and for an event, v. occurring 
along the proposed path of the AHFA-producing phase (see Fig. 37). 



conceivable that no single model will 
prove universal.) An obvious further 
test for our models involves calculating 
expected amplitudes resulting from re- 
fraction or reflection from interfaces sug- 
gested by the seismicity. Although 
Isacks' and Barazangi's model also in- 
volves a conversion, more restrictive tests 
for that model should result from impli- 
cations of their heretofore unverified 

We have found that a number of dif- 
ferent models fit travel time and Q data 
for AHFA. Therefore, in the absence of 
other restrictions, no conclusions based 
on uniqueness of any one model can 
validly be drawn. 

Converted P Phase from the ScS Phase 
at the Inclined Deep Seismic Zone 

Hiromu Okada 

A longitudinal forerunner to the ScS 
phase has been observed on short-period 

seismograms for nearby deep earth- 
quakes at Kamikineusu (KMU) and 
Urakawa (URA), both stations located 
on the Pacific coast of Hokkaido, north- 
ern island of Japan. At these stations a 
time difference of eight seconds between 
the forerunner and the ScS phase is es- 
sentially constant. The particle motion 
for the ScS phase showed horizontal 
polarization (SH). The longitudinal 
forerunner for the clear arrival, however, 
showed a different incident azimuth 
which corresponds to the dip direction of 
the deep seismic zone. The proposed 
interpretation for these forerunners was 
an ScSp phase, the P phase converting 
from the ScS phase at the upper bound- 
ary of the dipping seismic zone. 64 Figure 
39 shows the schematic ray paths and 
conversion interfaces discussed. 

Additional data were found at Matsu- 
shiro (MAT) in central Honshu, Japan. 
Figure 40 shows a traced record example 
(left below). The amplitude of the ScSp 

S °UT H 

4M£ R 

w Station 

• Hypocenter 
++ Earthquakes 

i a J ?r schematlc model showing the paths of the ScSp phase (solid line) and the ScS phase 
(dotted line). The ScSp phase, a P phase converted from the ScS phase at the inclined seismic 
zone, exists at certain stations in the Japanese and South American regions. 

200 400 







1 43 r 

4351 1 1 1 






1000 KM 

y>t , 



s s 

„- 1 



15 / ./// 

20 25 


35 sec 

1970 Sep. 5 OKHOTSK^ s 
m = 5.7 H = 580 KM 
A =18.2° AzB= 26.6° 

Fig. 40. Traced seismograms (left below) of the ScSp and ScS phases at MAT. The vertical 
cross section of the earthquake activity is shown by the numbers of the events located inside the 
oblong on the map (USCGS and NOAA, 1964-1971). Assuming the normal incidence for the 
ScS phase, the ScS-ScSp time referred to J-B travel times is indicated by the dotted lines. The 
time difference contours at 21 seconds are hatched and just cover the upper boundary of the dip- 
ping seismic zone. 

phase is about half that of the ScS phase. 
The time difference, however, is 21 sec- 
onds. There are no significant differences 
for this lead time. Sometimes the ScSp 
phase is hard to identify because of poor 
signal-to-noise ratio. 

The seismic activity has been projected 
onto a vertical cross section in Fig. 40. 
Assuming normal incidence for the ScS 
phase, the time-difference (ScS-ScSp) 
contour is calculated using J-B travel 
times and is shown by the dotted lines. 

The general pattern of the time differ- 
ence curves suggests a dip angle of about 
32°. The hatched zone — 21 seconds for 
MAT — covers the upper boundary of the 

dipping seismic zone. In this case the 
interpretation referred to above is ade- 
quate. Evidence suggests that for MAT 
the interface at the deeper part of the 
dipping zone might be sharp enough to 
produce a clear conversion. 

In South America it was found that 
the ScSp phase exists at some stations, 
but the relationship of this phase to the 
seismicity appears more complex. Exam- 
ination of the ScSp phase on the short 
period WWSSN seismograms for Are- 
quipa (ARE), southern Peru, La Paz 
(LPB), Bolivia, Antofagasta (ANT), 
northern Chile, and Nana (NNA) , cen- 
tral Peru, revealed the following: 









400 ARE 
_i I_ 



1000 KM 



11 1 
1 11 11% v 

V 1 

1 1 

V. N N 1 

1 V* 





1 12 J 2 T 1 1 1 1\ 1\ x. 

^ N 

^ 1 N -%c N s - 

25 20 15 10 5 sec 

1965 Nov. 3 
m = 6.2 H = 593 KM 
A = 7.3° AzB = 0.7° 







20 sec 


Fig. 41. Vertical cross section of earthquake activity and the traced seismograms (right below) 
of the ScSp a ad ScS phases at ARE. Time difference contours as for Fig. 39. 

Fairly clear longitudinal ScSp phases 
were observed at Arequipa about 11 sec- 
onds before the ScS phase (Fig. 41). In 
southern Peru the deep seismic zone is 
clearer and steeper than in central Peru. 
The interpretation previously referred to 
is also adequate in this case. No clear 
phase is recognized at La Paz and Anto- 
fagasta. The La Paz station is east of 
the dipping seismic activity, and there- 
fore no dipping interface is expected. For 
the coastal station, Antofagasta, since the 
precise form of the deep seismic zone is 
convex, the converted wave from the 
deeper part could arrive above the inter- 
mediate activity but not near the coastal 

There is a 10-second precursor at 
Nana. The seismicity is more compli- 
cated (Fig. 42). The ScSp phase looks 

similar to the ScS phase. A seismicity 
cross section is shown in Fig. 42. Three 
regions show relatively dense activity: 
inland shallow activity, ocean-side shal- 
low activity, and intermediate activity. 
Isacks and Molnar (1971) 65 suggest that 
the slab-like seismic zone has a dip of 
only 10° to 15° (inset in Fig. 42) (Fu- 
kao, 1972, 66 Sykes, 1972 67 ). The hatched 
zone around 10 seconds, indicating the 
probable interface at which the ScS to 
ScSp conversion takes place, is below the 
slab. If this slab is a good representation 
of the deep seismic zone, the ScSp phase 
is not converted at the upper boundary 
and the proposed hypothesis apparently 
fails. The conversion might occur at the 
bottom of the slab. Some other explana- 
tion might be possible, considering that 
the hatched part can be a good indication 









200 NNA 
_J I L 

1 1 1 11 
2 1112122 1 

2436 14433984625343 21 

12132235C3331 v 
2 112 2444513 21 ^ 

1 1444622111 1 





1000 KM 

I 1311 1 
1 21 111 

II 141 
1 1 


14A41 1 

1131 11 1 

1 1111 1222141444 1 1 

21 12 2 1221 4 412247AC71321111 1 

K 111 2 221 1421 1 1 

V 1 \ 1 11 12 11 1 

1 1 11 1 

1 11 1 

v < 1 1 1 11 

i A v 3 \ 1 1 1 11311 1 1 

TInA 11K3 \1 11 1 11 1 1 1 1 

11<SL 1 \ \ 1 1 113 2 11 1 

TV \ \1 11 1 1111 21 1 

1 V 1 \ 1 V 1 11 222 21 1 

1 %. ~\ \1 11 1 12 12311 

31434 31 
% 111 

%. ^ 1 31 

1 1 


%.B V 1\1 

\ 1 





1 \ 
\ % 

20sec15 10 










1000 KM 

1965 Mar. 5 SANTIAGO DEL 

m = 5.5 H = 573 KM ESTERO 

A = 19.6° AzB = 141.5° 





20 sec 

Fig. 42. Vertical cross section of the earthquake activity and the traced seismograms (right be- 
low) of the ScSp and ScS phases at NNA. Isacks and Molnar's gently dipping slab model (inset, 
left below) also indicates only large events for the same region, with magnitude greater than 
about five (USCGS, 1961-1967). The upper boundary of the slab is far above the 10-second time 
difference zone (hatched zone), and the hypothesis of wave type conversion at the top of the 
inclined seismic zone apparently fails. 

of the seismic zone. There is clear evi- 
dence that the ocean-side shallow activ- 
ity shows significant bending to a steeper 
angle (Fig. 42) . The extrapolated zone 
of that activity is in contact with the 
hatched zone (Fig. 42). 

Zoeppritz's equations describe the am- 
plitude relations in terms of elastic pa- 
rameters and geometry. 68 ' 69 For ver- 
tically polarized shear waves (SV) inci- 
dent from the high velocity side, the am- 
plitude of the refracted P wave gradually 
increases with incident angle and forms a 
small minimum at the first critical angle 
(reflected P critical). After the first cri- 
tical angle, considerably larger amplitude 

is expected up to the second critical angle 
(refracted P critical). The larger ampli- 
tude corresponds to greater velocity con- 
trast. Figure 43 shows these amplitude 
ratios ( refracted P/refracted S). When 
the velocity of the incident side is 10% 
higher, the amplitude ratio can reach 
nearly one-fifth. 

Assuming normal incidence for the 
ScS phase, and using the Nafe-Drake 
velocity-density relation (Talwani et al, 
1959) 70 and a Poisson's ratio of 0.25, the 
velocity ratio R (transmitted/incident) 
and the amplitude ratio are calculated 
for the incident angle i 8 and the take-off 
angle 6 P at the conversion point (Fig. 




20° 30° 40 


Fig. 43. The transmitted amplitude ratios (refracted P/refracted S) for the incident angle is. 
The velocity contrast R (transmitted /incident) = 0.7, 0.8, 0.9, 1.1, 1.2, 1.3, and the Poisson's ratio = 
0.25. In the high velocity side, the P wave velocity is constant at 8 km /sec. 

44) . The P wave velocity in the high 
velocity side is constant at 8.0 km/sec. 
For the normal incidence, i s is equal to 
the dip angle 8 of the dipping seismic 
zone. The converted P to transmitted S 
amplitude ratio is indicated by the 
broken lines in percent. The velocity 
ratio R (transmitted/incident) is shown 
by the solid lines. The zones that satisfy 
the time differences in Figs. 40 to 42 cor- 
respond to the dotted columns and those 
marked A, B, C. For NNA, two cases are 
examined: NNA-1 for the time difference 
zone and NNA-2 for the bottom conver- 
sion of the gently dipping slab (Fig. 44) . 
It is not easy to determine dip angle, espe- 
cially in South America. The incident 
angle (equal to the dip angle) for KMU, 
MAT, ARE, and NNA-1 is about 30 to 

35°, and for these cases i s is slightly less 
than the first critical angle. From studies 
of travel-time anomalies in typical island 
arcs, it is believed that the velocity in the 
dipping seismic zone is 5 to 10% higher 
than the surrounding material. Al- 
though this velocity anomaly is a path- 
integrated averaged value, it is appro- 
priate to constrain the velocity ratio R 
at about that value. For shallower con- 
version, A (in Figs. 40 to 42) lies near the 
R = 0.9 line, and the amplitude ratio is 
about 5 to 15% (Fig. 44). Since the 
velocity of the dipping seismic zone is 
higher than the underlying part, the inci- 
dent angle i s might be several degrees 
larger than the normal incident case. 
After correction, the incident angle i s 
may become larger than the first critical 






Fig. 44. The geometry, the velocity contrast, and the amplitude ratio for S to P conversion. 
Under certain assumptions, the velocity ratio R (transmitted/incident) is calculated (solid lines), 
and the amplitude ratio (converted P/transmitted S) is shown in percent by the dotted curves, 
from the Zoeppritz equations. The hatched zones in Figs. 39, 40, and 41 correspond to the dotted 
columns. The marks A, B, and C indicate the arbitrary position corresponding to the same marks 
in Figs. 40, 41, and 42. See text for further discussion. 

angle, and more efficient conversion will 
occur in a slightly deeper part between A 
and B (Figs. 40 to 42). For the con- 
version at the bottom of the slab (inset, 
Fig. 42), R = 1.3 and the amplitude 
ratio is less than 10% (Fig. 44, NNA-2). 
This velocity ratio (R = 1.3) is unrea- 
sonably large and, for more probable 
velocity ratios, the conversion will be too 
inefficient to cause the large observed 
arrivals. Conversion on the upper bound- 
ary could explain not only the time dif- 
ference but also the geometry and the 
amplitude tendency. 

Some spectra from the Carnegie broad- 

band seismograms have been analyzed 
(Fig. 45) . At KMU, spectra of the ScS and 
the ScSp phases show the same form and 
slope in the frequency range 0.2 to 0.9 
Hz. The spectral ratio ScSp/ ScS is about 
unity. Because of the low signal-to-noise 
ratio at MAT (for the example shown 
in Fig. 45) and the considerable differ- 
ence in two horizontal spectra, it is not 
easy to compare them. In other cases at 
MAT and other stations the apparent 
wave characteristics show no significant 
differences between the ScSp and ScS 
phases. Regional differences in Q may 
affect spectral characteristics, but it ap- 






Fig. 45. The spectral characteristics of the ScSp and ScS phases. Traced seismograms (top), the 
spectra (middle), and the spectral ratio (bottom). The 10-sec boxcar window was applied to 
KMU (right) and a 16-sec window to MAT (left). The digitizing rate was 40 points per second. 
The spectral ratios are calculated after 1/3 octave averaging. For KMU, the three spectra show 
the same slope and the ratio is about unity. 

pears more likely that conversion is fre- 
quency dependent because of velocity 
structure at the interface. 


1. The ScSp phase exists in a certain 
area in Japan and in South America 
where there is a clear seismic zone. The 

time difference, the geometry, and the 
amplitude tendency agree with the con- 
version at the upper boundary of that 
zone (Fig. 39). 

2. NNA's case is especially interesting 
because when the gently dipping slab 
model 65 is taken, the proposed hypothe- 
sis apparently fails. This suggests that 



at intermediate depths the slab itself 
is aseismic and may be continued from 
the ocean-side shallow active seismic 

3. The amplitude ratio ScSp/ScS can 
be quite large, and for these cases the 
conversion must be at nearly critical 
angles. The velocity contrast might also 
be quite large. Averaged velocity anom- 
alies from the travel-time studies may 
not be enough to produce a clear conver- 
sion. Considering the amplitude, the 
geometry, and the spectral characteris- 
tics, it is likely that the boundary is a 
thin transition zone instead of a simple 
boundary. Application of a Thomson- 
Haskel matrix method to this problem is 
under study. 

4. The ScSp phase can be a new tool to 
investigate a boundary in the upper 
mantle and provides a new approach to 
understanding the tectonic processes at 
island arcs or island arc systems. 

Source-Spectral Relations for 

A. Snoke, A. T. Linde, and I. S. Sacks 

Archambeau's elastodynamic source 
theory 71,72 postulates an infinite medium 
subject to a uniform shear field from 
which strain energy is released as a con- 
sequence of the creation of an expand- 
ing rupture zone in which the rigidity 
vanishes. The only geometry published 
to date is a spherical zone which expands 
to some maximum radius R () with a con- 
stant radial velocity V R (although Ar- 
chambeau and Minster have also found 
solutions for various propagating-sphere 
geometries 73 ). In some calculations it 
appeared convenient to truncate integrals 
over the prestressed medium to the vol- 
ume of a sphere containing the source 
with finite radius R s . This seemed justi- 
fied by Archambeau's finding 72 that 
more than 95% of the total radiated 
energy comes from a region surrounding 
the rupture zone of radius equal to five 
i^o and by Linde's finding 74 that for 

Ro, V R , and the signal velocity T T a held 
constant, the far-field displacement spec- 
trum was independent of R s for R s large. 
Using an equivalent but computation- 
ally simpler formulation of the same 
theory, 72 we have redone Linde's calcula- 
tions and have found some of his results 
to be in error. We now find that the 
shape and the position of the peak of the 
far-field displacement spectrum have a 
strong dependence on R H . Typical results 
are shown in Fig. 46, where we have 
plotted the far-field displacement spec- 
trum for Rs/Ro equal to 5, 10, 20, and 40 
while Ro, V R , and V a are held constant. 
(For comparison we have included the 
spectrum as calculated previously by 
Linde). As Rs increases, the frequency 
at the peak, v peak , decreases so that Rs X 
vpeak remains constant and the curve gets 
broader. As pointed out by Randall, 75 
in the limit that Rs becomes infinite, the 


Fig. 46. Far-field displacement spectra for 
Ro = 10 km, V R = 4.5 km /sec. V a = 10 km/sec 
and RJRo = 5. 10. 20 and 40. Also included 
(dashed line) is the spectrum as calculated by 
Linde for Ro, Vf, and T" a the same as above 
and RsIR, > 8. 



spectrum approaches the flat-to-d.c. 
shape predicted by dislocation mod- 
els. 76, 77 - 78 Thus, Archambeau's theory 
produces peaks only for noninfinite val- 
ues of Rs- Since the model assumes an 
infinite uniformly prestressed medium, 
it is inadmissible to choose a finite value 
for R s . 

These findings have some bearing on 
the previous study by Linde and Sacks 79 
in which dimensions of South American 
deep earthquakes were obtained by inter- 
preting observed spectra in terms of the 
theoretical spectrum labeled "Linde" in 
Fig. 46. Although this interpretation is 
not correct, it might be noted that these 
dimensions were also consistent with 
measurements of radiation time, and so 
we do not expect serious errors in these 
estimates. It also bears mention that 
energy calculations in that study are 
independent of the source theory, as is 
the argument which leads to the conclu- 
sion that radiation efficiency increases 
with magnitude. 

In order to derive any relations be- 
tween source dimensions and displace- 
ment spectra, we require a theory that 
reproduces observed spectra. When con- 
sidering possible modifications of existing 
source theories, we examine assumptions 
implicit in the current formulations of 
those theories which, if altered, would 
make the theory more physically realistic 
but would not make it mathematically 
intractable, and we then ask if altering 
these assumptions could alter the low- 
frequency behavior. In Archambeau's 
theory we are examining assumptions 
concerning the geometry of the rupture 
zone and the size of the prestressed me- 
dium. A spherical geometry for the rup- 
ture zone is a very unrealistic model for 
an earthquake, but it is believed that 
going from a spherical to a more planar 
geometry will have little effect on the 
low-frequency part of the spectrum. A 
radial propagation may not be bad for 
describing bilateral earthquakes, but it 
is a very poor approximation of a uni- 
lateral event. Archambeau and Minster 

report 73 that linear propagation brings 
in higher-order multipole contributions 
which can affect the spectrum near the 
corner frequency. In the earth, techn- 
ically stressed regions are definitely 
finite, and for the South American events 
we have been studying, the observer is 
far outside any prestressed region. The 
spectra in Fig. 46, in which R s is treated 
as finite, are probably meaningful to the 
extent that they indicate that increasing 
the size of the prestressed region tends to 
increase the amplitude of the low- 
frequency part of the far-field displace- 
ment spectrum. 

Borehole Strainmeter Installation 
in Japan 

/. S. Sacks and S. Suyehiro 

The major emphasis of the initial phase 
of the operation of the DTM borehole 
strainmeters in an active earthquake re- 
gion in Japan was an assessment of the 
capabilities of this type of instrumenta- 
tion. Initial tests using explosive-gen- 
erated vibrations showed that the instru- 
ments and their immediate environment 
could tolerate the severe accelerations 
caused by near earthquakes without 
spurious behavior (Sacks et al., 1971 ). 80 
What remained to be confirmed was 
whether strain measurements made very 
near the surface were reliable, since rock 
at shallow depths is often fractured and 
jointed, particularly in the Matsushiro 

We have already recorded the strain- 
field changes of a number of earthquakes 
which occurred within the net, and we 
find that consistent strain steps are re- 
corded at the different sites. For example, 
an earthquake of magnitude 3.9 on 26 
October 1972, which was located about 
midway between the Nagano and Matsu- 
shiro installations (10 km), gave strain 
steps of 1.4 X 10" 8 at Nagano and 
2.1 X 10" 8 at Matsushiro. Body waves 
and surface waves from distant earth- 
quakes write almost identical records at 
the three sites. 



We now know that strain steps, body 
waves, and surface waves can be re- 
corded with fidelity by borehole strain- 
meters in shallow holes (50 km), but 
in the period range of 10 minutes to 
several hours, the local noise is domi- 
nated by pressure changes in the aquifer. 
This is a very local effect; noise in this 
period range differs by more than a fac- 
tor of 100 in two instruments just 300 
meters apart. Because of this low co- 
herence, we expect that installation of a 
strainmeter well below the level where 
the aquifer saturation changes due to 
such actions as pumping would reduce 
this 10-minute to 2-hour noise. 

At very long periods (of about a year) , 
the results from the borehole instrument 
agree well with the 100-meter quartz ex- 
tensometer at Matsushiro. Further anal- 
ysis of the data will be done during a 
forthcoming visit by Sacks to Japan. 

Project Narino 

L. T. Aldrich, D. E. James, A. T. Linde, L. Ocola, 
G. R. Poe, I. S. Sacks, and D. Simoni 

Collaborating Institutions 

Bolivia: Comision de Geofisica; Instituto Pan 
Americano de Geografia e Historia. Colombia: 
Instituto Geojisico de Los Andes, Universidad 
Javeriana; Instituto Geogrdfico "Agustin" Co- 
dozzi; La Armada Colombiana; Texas Petro- 
leum Corp.; International Petroleum Corp.; 
INGEOMINAS. Ecuador: Instituto Geogrdfico 
Militar Ecuatoriano; Direccion National Geo- 
logico; Escuela Politecnica; Instituto Hidro- 
grdfico de La Armade Ecuatoriana. Germany: 
German Seismic Group. Panama: Inter-Amer- 
ican Geodetic Survey. Peru: Instituto Geojisico 
del Peru; Instituto Geojisico, Universidad Na- 
tional de San Agustin; Centro Regional de 
Sismologia. United States: University oj Texas; 
University oj Wisconsin; University oj Hawaii; 
University oj Washington. 

The Andean cordillera, an active tec- 
tonic belt that extends thousands of kilo- 
meters along the west coast of South and 
Central America, has been a major focus 
of geophysical research by the Depart- 
ment of Terrestrial Magnetism (DTM) 
for more than a decade. Most of these 
studies have been confined to the region 

of the central Andes — southern Peru, 
Bolivia, and northern Chile. The central 
Andes, dominated by two mountain 
chains — the eastern and western ranges 
— and an intervening high plateau, lies at 
the heart of the Andean system and stud- 
ies of the region have never failed to 
reveal unique and extraordinarily com- 
plex structures. Nonetheless, there are 
other sectors of the Andean cordillera 
that present equally fascinating geolog- 
ical and geophysical anomalies and com- 
plexities. Such a region is southwestern 
Colombia. Here the Andes, which to the 
south form a compact chain of mountain 
belts trending parallel to the coast, di- 
verge to form three distinct north-north- 
easterly trending ranges separated by 
broad low-lying valleys (see Fig. 47). 
The point of confluence of these three 
ranges near the city of Pasto marks the 
approximate latitude north of which the 
Peru-Chile trench shallows and nearly 

Gravity investigations by Case and 
others (1971, 81 1973 82 ) demonstrate that 
this region is not only physiographically 
unusual but tectonically complex as well. 
The Bouguer anomalies indicate that 
parts of the Andean cordillera within this 
zone are either out of isostatic equilib- 
rium or are compensated by unusual 
crustal and upper mantle structures. For 
example, the Cordillera Occidental of 
western Colombia — some peaks of which 
attain elevations of 10,000 feet or more — 
is characterized by position Bouguer 
anomalies. Case and his co-workers have 
suggested that the Cordillera Occidental 
has been built atop oceanic crust, possi- 
bly as an island arc that has subse- 
quently collided with and been incor- 
porated into the South American conti- 
nent, and that the crust is therefore dense 
and comparatively thin. Other parts of 
this region of the Andean cordillera are 
similarly complex, and the magnitude of 
the gravity anomalies suggests that con- 
trasts in crustal structure may be large. 

To examine the crustal structure of 
southwestern Colombia and northwestern 

















Fig. 47. Location map showing position of shot lines and observing profiles (asterisks) for Proj- 
ect Narifio. Land shot point is situated in Laguna de la Cocha, east of Pasto. 



Ecuador, the DTM participated in the 
organization of a multinational study 
carried out early in 1973. The program 
was coordinated by Fr. J. E. Ramirez, 
S.J., of the Instituto Geofisico de los 
Andes Colombianos. 

Shot points, shot lines, and observing 
profiles are shown in the map given in 
Fig. 47. The only genuinely fixed shot 
point was that in Laguna de la Cocha 
near Pasto where some 20 explosions of 
one-half to two tons were fired by DTM. 
A series of shots was fired from the ship 
Gorgona near Tumaco and Buena- 
ventura by the University of Texas at 
Dallas. Two-ship seismic profiling be- 
tween Tumaco and Buenaventura was 
carried out using both the Colombian 
vessel Gorgona and the oceanographic 
ship Kani-Keoki of the University of 
Hawaii. Finally, a series of shots about 
50 km apart was fired by the Kani-Keoki 
along two profiles, Malpelo Island to 
Tumaco and Buenaventura to Malpelo 
Island. Seismic observations along short 
profiles near Malpelo Island were made 
by the University of Washington oceano- 
graphic ship, the T. G. Thompson, which 
further carried out heat flow, bathymetry, 
and magnetic measurements. 

Recording sites on land were organized 
along four principal profiles, as shown 
in Fig. 47. Profile I was occupied by ob- 
serving groups from the University of 
Wisconsin and the German Seismic 
Group along that part of the profile 
north of Pasto; south of Pasto and in 
Ecuador the observations were primarily 
the responsibility of DTM. Four DTM 
and IGP recording groups observed along 
profile II which trended along the valley 
of Huila. Observations along profile III 
were carried out by the University of 
Texas at Dallas and the University of 
Hawaii. All groups observed along pro- 
file IV. 

Preliminary results are still fragmen- 
tary and contradictory. There is some 
indication that no apparent velocities 

greater than 7.5 km/sec are observed 
along the north-south profiles even at 
distances well in excess of 300 km. If 
this conclusion holds, it will mean that 
we have yet to observe definite "normal" 
mantle velocities anywhere along the 
Andean system. This work was sup- 
ported in part by the National Science 
Foundation and the Harry Oscar Wood 
Fund of the Carnegie Institution. 

Electrical Conductivity Studies 

L. T. Aldrich, M. Casaverde, J. Bannister, 

L. Beach, T. J . Smith, L. Tamayo, E. Triep, 

R. Quiroga, R. Salgueiro, and S. del Pozo 

Observations to study the electrical 
conductivity structure across the Andean 
cordillera in central Chile and Argentina 
had just been completed at the end of 
the previous report year. These observa- 
tions were obtained using a 15-station 
net of magnetic variographs operated at 
stations located as shown in Fig. 48. Five 
nighttime events have been digitized to 
provide good values for the AZ/AH and 
AD/ AH relationships among the stations. 
Two or three daytime events are needed 
to complete the data required to give 
AZ/AD and AH /AD as a function of fre- 
quency. This work has been and is being 
carried out both at the University of 
Chile in Santiago and here at the De- 
partment. Preliminary results from the 
nighttime events substantiate those found 
earlier with two stations. Any abnormal 
conductivity with an E-W component, if 
it exists, lies west of the net (under the 
ocean). The daytime storms with varia- 
tion in D are needed to examine struc- 
tures with an alignment parallel to the 
Andean cordillera (N-S). 

Of major importance to future analy- 
sis are two developments this year. An 
interpolating algorithm developed by 
Professor Shelton Alexander of the Penn- 
sylvania State University has diminished 
the time required for digitizing our rec- 
ords by at least a factor of 2. This 



70° W I 














El Quisco 




Mario Pinto 



30° S 

""^Paso de las Aguas Negras 

/+ + 

+ Jachal 

Las Flores 





Cerro Negro 

32° S 


+ Uspallata 


34° S 

70° W 

Fig. 48. Map showing locations of magnetic variograph stations in Chile and Argentina for elec- 
trical conductivity studies in 1972. 

algorithm requires digitization only at 
points where the first and second time 
derivatives of the record are zero. Com- 
parison with records digitized at one- 
minute intervals showed that the two 
digitizing systems produced identical 
records. Initial steps have been taken to 
adapt new techniques of frequency anal- 
ysis to our problem. These techniques 
look most promising and would not only 
further reduce our digitizing require- 
ments but also increase the precision of 
our analysis. 

The magnetic variographs are cur- 
rently deployed in Peru, Bolivia, and 

northern Chile to study in more detail 
the anomalies in southern Peru and cen- 
tral Bolivia. Revisions of the DTM 
variographs to make them even more 
easily used have been completed. Figure 
49 shows a record obtained at Las Flores, 
Argentina, of an event recorded on May 
16, 1972. The characteristics of the rec- 
ord which make it easy to analyze are: 
(1) one-minute digitization rate, (2) 
independent time marks at one hour and 
thirty minutes, and (3) twelve-hour 
marks on the Z (vertical) trace to help in 
time keeping and trace identification. 
The unit operates at 7.5 volts with aver- 






1 K 
i % 

II / 





*4 K 

<r i 



I I I 

I i 

! I 



>^ 1 \ 


J * 

% 1 

* f 
m f 

I 1 

I m 

il I 







#• ' 



UJ o 



*i — I— I 


~ 3 ^ 
3 C •/. 

S45 O 

~C - ~ 

:- x — 

O ^ ■- 

•^ '— ~ 

'- T - ~ 

'"_ 5 o 

■ a § I 

X - - 

~ 2 

bp a 

E-i -■ 

W — ' 

o — 

S 5 

— -5 

3 w 


i-i c 

5 § 

g s 

X! BD .2 b 

O 3 O 
« 2 f 3 

;- — . 

X X 

•- 3 

<M O - O 

Oi ^ g '- 

CD r3 — S 

3 O C 

r 3 o 

x - "~ 

£ ™ O 

3 © 




«~ 5 

O x « ' 

— X 

2 ~C *-■ ~ 

2 g ~ 

H -H 

J3 ^ 

■f r- !> _z: 

■SP § £ 2 

o S 



age current of 60 ma, which is easily 
provided by aircells, which in turn re- 
quire no other source of power during 
their lifetime. 

Geochemistry and Geochronology 
Geochemical Trends in Andean Rocks 
D. E. James, C '. Brooks, and A. Cuyubamba 

We have reported in previous Year 
Books on our efforts to understand the 
evolution of the Andean orogen. Over 
the years of studying the central Andes 
Mountains, the direction and emphasis of 
our research have been altered to pursue 
promising new avenues to fresh informa- 
tion. As the implications of the role of 
plate tectonics in continental evolution 
have become better recognized and un- 
derstood, the Andean system has emerged 
as a classic example of an evolving 
orogen produced by ongoing subduction 
along a continental margin. 

Our previous studies of the Andes were 
confined largely to geophysics, with an 
emphasis upon crustal and upper mantle 
structure. Those early studies form the 
necessary underpinnings of an under- 
standing of evolution as read from the 
geologic record. Plate tectonic theory, 
however, has led to a confluence of for- 
merly distinct disciplines within earth 
science, and there now exists clear com- 
plementarity between geophysics, geo- 
chemistry, and geology in the study of 
orogenic processes. Thus, while geo- 
physics tells us much about the existing 
tectonic regime of the central Andes, it 
can only hint at the tectonic evolution 
which has led to the present-day moun- 
tain system. To read the record of past 
events it has been necessary to turn to 
geology and geochemistry. 

The kinds of geological or geochemical 
studies which can be fruitfully pursued 
by one or two individual research men 
must be carefully framed. It is clear, for 
example, that geological field studies are 
pursued best within the context of large, 

organized programs such as those car- 
ried out over decades by national geo- 
logical surveys. Thus, we at DTM have 
concentrated our efforts on geochemical 
studies which heretofore have constituted 
a virtually unexplored domain in the 
central Andes. By focusing our attention 
upon those geochemical parameters that 
have been shown to be sensitive indica- 
tors of plate tectonic processes, we hope 
to shed light on the early tectonic history 
of the Andes. 

A model for Andean evolution from 
Paleozoic times to the present was de- 
scribed in Year Book 70, pp. 344-349. 
That model was based largely upon geo- 
physical and geological data but may be 
at least partially tested by studying cer- 
tain aspects of the geochemistry of the 
igneous rocks that form the Andean 
cordilleran arc. In Year Book 71 we out- 
lined our program of geochemical study 
and our specific objectives. The results 
reported below represent preliminary 
data now emerging from these studies. 

Sr 87 /Sr 86 ratios for igneous rocks of the 
central Andes. One hypothesis related to 
plate tectonic theory holds that the ande- 
sitic magmas extruded from the vol- 
canoes of island and cordilleran arcs 
originate by partial melting of the upper 
parts of the descending plate or of the 
mantle immediately above the plate. 
Partial support for this hypothesis comes 
from the measurement of Sr 87 /Sr 86 ratios 
in igneous rocks of the island arcs. In 
these rocks the ratios are uniformly 
around 0.7038 and it is generally thought 
that ratios this low imply magma origins 
from within the mantle or oceanic crust. 
As regards island arcs, this hypothesis 
has met little opposition, for there ap- 
pears to be no other probable source of 
magma generation. 

For cordilleran arcs such as the Andes, 
this conclusion, though following logic- 
ally by analogy with island arcs, is not 
so simple or so readily verified. It is 
possible, for example, that cordilleran 
arcs sitting astride sialic crust are fed 



by anatectic melting of preexisting crus- 
tal rocks. Indeed, until the advent of 
plate theory, such an origin was generally 
assumed; Pichler and Zeil (1967, 83 
1969, 84 1971 85 ) still maintain that the 
late Cenozoic volcanism of northern 
Chile is caused by anatectic melting of 
lower crustal rocks. Arguments like 
these are kept alive by the fact that 
Sr 87 /Sr 86 ratios in rocks of the cordilleran 
arcs such as the Sierra Nevada are con- 
sistently higher than those of the island 

We have previously concluded, on the 
basis of geophysical evidence indicating 
that the crust of volcanic arc has nearly 
doubled in thickness over the past 50 to 
100 m.y., that the magma source of the 
late Cenozoic volcanism must lie beneath 
the crust. We sought to further discrimi- 
nate between various possible magma 
sources by measured Sr isotopic ratios in 
the igneous rocks of the arc. Clearly, 
nothing in the Sr ratios can bear either 
for or against a lower crustal origin, for 

we have no knowledge of the isotopic 
composition of those rocks. Nonetheless, 
isotopic ratios near those of island arcs 
should lend credence to the hypothesis 
that the magma originates subcrustally. 
If isotopic values are very high, in excess 
of 0.710, for example, they could suggest 
anatectic origin by melting of preexisting 
crustal rocks. 

Strontium 87/86 present-day ratios 
uncorrected for age are given in Table 15. 
Radiometric ages have yet to be deter- 
mined and hence a correction for radio- 
genic Sr 87 has not been made. The rock 
ages given in the table are on the basis 
of field relations and provide only a gen- 
eral guide to the relative chronology. For 
late Cenozoic volcanic rocks, the present- 
day ratios are equivalent to initial ratios. 
For certain late Cretaceous-early Terti- 
ary rocks, the Rb/Sr ratio is sufficiently 
high that the correction for radiogenic Sr 
may be quite large. For example, sample 
86 has a Rb/Sr ratio of greater than 1.0; 
thus if the rock is 100 m.y. old, the iso- 

TABLE 15. Sr 8 7Sr 86 Present-Day Ratios for Rocks of the Central Andes of Southern Peru 


Sr 87 /Sr ! 


Depth to 


Zone, km 


PE 26 0.7071 ±0.0001 0.0547 155 Quaternary andesite 

PE 27 0.7070 ±0.0003 0.0547 155 Quaternary andesite 

PE 30 0.7072 ±0.0002 0.0548 155 Quaternary andesite 

PE 46 0.7076±0.0001 0.0883 160 Quaternary andesite 

PE 71 0.7052±0.0002 0.1570 200 Quaternary andesite 

PE 1 0.7069 ±0.0002 1.095 130 Cretaceous-early Tertiary granite 

PE 13 0.7070 ±0.0003 0.3569 78 Cretaceous-early Tertrary diorite/gabbro 

PE 17 0.7057 ±0.0002 0.1341 145 Cretaceous-early Tertiary tonalite 

PE 58 0.7057 ±0.0002 0.2456 1155 Cretaceous-early Tertiary tonalite 

PE 85 0.7052 ±0.0002 0.0455 140 Cretaceous-early Tertiary gabbro 

PE 86 0.7088±0.0002 1.234 140 Cretaceous-early Tertiary granodiorite 

PE 87 0.7082 ±0.0003 0.2283 140 Cretaceous-early Tertiary diorite 

PE 90 0.7060±0.0001 0.1854 140 Cretaceous-early Tertiary granodiorite 

PE 92 0.7056±0.0003 0.0811 140 Cretaceous-early Tertiary diorite 

PE 12 0.7046±0.0001 0.0168 105 Jurassic basalt (altered) 

PE 114 0.7052±0.0002 n.d. 120 Jurassic basalt (altered) 

PE 115 0.7058±0.0001 n.d. 120 Jurassic basalt (altered) 

PE 117 0.7049±0.0002 n.d. 120 Jurassic basalt (altered) 

PE 118 0.7040 ±0.00007 n.d. 120 Jurassic diabase 

PE 18 0.7749 ±0.0001 0.971 110 Precambrian banded gneiss 

PE 19 0.7724 ±0.0002 0.765 110 Precambrian banded gneiss 

PE 37 0.7324±0.0001 0.264 145 Precambrian banded gneiss 

PE 111 0.7379±0.0002 0.962 70 Precambrian banded gneiss 

* Ages not radiometric. 



topic correction will be of the order of 

Histograms of 87/86 ratios for samples 
within age categories are presented in 
Fig. 50. This series of histograms illus- 
trates that the 87/86 ratios tend to be 
higher for younger rocks than for older 
ones. The oldest rocks of the arc, Juras- 
sic extrusives, exhibit the lowest isotopic 
ratios and the youngest extrusives, late 
Cenozoic in age, exhibit the highest 

The change in isotopic ratio correlates 
not only with time but with depth to the 
present-day Benioff zone and crustal 
thickness. In detail, however, these cor- 
relations fail to survive close scrutiny. 
Fig. 51 shows, for example, 87/86 ratios 
of late Cretaceous-early Tertiary and 
late Cenozoic samples plotted against 
depth to the present-day Benioff zone. 
Within any given age group, no correla- 
tion exists between isotopic ratio and 
depth to the Benioff zone. The same lack 
of correlation holds for crustal thickness. 

All of the samples, with the possible 
exception of some Jurassic extrusives, ex- 
hibit higher 87/86 ratios than the aver- 
age for island arcs. This result is in ac- 
cord with other data that show that rocks 
of cordilleran arcs consistently exhibit 
higher Sr 87 /Sr 86 ratios than rocks of is- 
land arcs. Table 16 gives some repre- 
sentative averages for island arc and 
cordilleran arcs. The Andean rocks have 
87/86 ratios that lie comfortably within 
the range of values for cordilleran arcs. 

If we grant that island arcs represent a 
comparatively simple system of ande- 
sitic magma genesis, then it is reasonable 
to suppose that the isotopic ratios meas- 
ured are of uncontaminated rocks. The 
consistency between rocks of various arcs 
supports this notion. Thus, if we are to 
assume that the andesites of cordilleran 
arcs are of similar origin to those of 
island arcs, it is necessary to postulate 
that the strontium isotope ratios have 
been altered by contamination in the 
passage of magma through sialic crust. 





703 704 765 706 707 708 709 JlO 



v 5 


m 4 

o J 

« 2 





.703 .704 .705 .706 .707 .708 .709 .710 
Sr 87 /Sr 86 

Fig. 50. Histogram of Sr 87 /Sr 86 ratios for 
igneous rocks of various ages of coastal and 
western cordilleran regions of southern Peru. 
Sample ages are on basis of field determinations 
and are not precise. All ratios are uncorrected 
for age. 

To explore at least one facet of the con- 
tamination problem, and at the same 
time to examine the question of possible 
crustal origins for these rocks, we meas- 
ured isotopic ratios of the Mollendo 
gneiss of southern Peru — the only wide- 





N V -707h 



70 80 


120 140 

160 180 200 

Fig. 51. Sr 8 7Sr 86 ratios plotted against depth 
to the present-day Benioff zone for samples of 
late Cretaceous— early Tertiary age and late 
Cenozoic age. 



TABLE 16. Sr 8 7Sr 86 Ratios for Andesitic and Granite Rock Suites of the 

Circum-Pacific Region * 


Rock Type 


Sr 87 / 86 

Reference Source 




Hedge (1966); Pushkar (1968) 

New Britain 



Peterman et al. (1970) 

Cascade Range 

Quaternary basalt and 
andesite, N. Cal., 
Oregon, and Wash. 


Hedge et al. (1970) 
Peterman et al. (1970a) 

Central America 

Quaternary andesite 


Pushkar (1968) 


Cretaceous granitic average 
of Sierra Nevada 


Hurley et al. (1965) 

Northern Chile 



Hedge (in press) cited in 
Pichler and Zeil (1972) 

Southern Peru 

Quaternary andesite 
Cretaceous/E. Tert. 
granitic rocks 
Jurassic basaltic rocks 





present study 
present study 

present study 

15 Adapted from Dickinson, 1970, p. 825, Table 2. 

spread pre-Mesozoic rocks exposed in the 
cordillera. The isotopic ratios are plotted 
against Rb/Sr ratios in Fig. 52. Shown 
in parentheses beside each point is the 
model age (in b.y.) for an initial ratio 
of 0.710. The Sr 87 /Sr 86 ratios of the 
gneiss establish two important points: 

1. The gneiss is almost certainly Pre- 
cambrian in age and, thus, predates any 
of the Paleozoic marine sedimentary 
rocks of the eastern ranges. 

2. The Sr 87 /Sr 86 ratios are so high that 
they preclude the possibility that the 

Fig. 52. Sr 8 7Sr 86 ratios vs. Rb/Sr ratio of 
samples of Precambrian (?) Mollendo gneiss. 
Model ages (in b.y.) assuming initial ratios of 
0.710 are shown beside each point. 

gneiss has participated in any significant 
way in the genesis of the igneous rocks of 
the post-Paleozoic arc. The gneiss, or 
rocks of similar isotopic composition, 
does pose imporant possibilities for con- 
tamination, for even minor assimilation 
of such rocks by a rising andesitic magma 
could increase isotopic ratios signifi- 

The rocks of Jurassic age exhibit pres- 
ent-day ratios much closer to those of 
island arc extrusives. Most of the Juras- 
sic extrusives exhibit alteration which 
may have increased the 87/86 ratio. 
Moreover, no correction has been made 
for radiogenic Sr 87 , although, in general, 
Rb/Sr ratios appear to be low. The least 
altered rocks typically give the lowest 
87/86 ratios and analysis of compara- 
tively fresh plagioclase separates from 
the altered samples is now being under- 
taken to ascertain whether isotopic com- 
position varies. In general, however, the 
Sr 87 /Sr 86 ratios of the Jurassic rocks, 
even without consideration of the effects 
of alteration and radiogenic strontium, 
are not substantially higher than those 
ratios measured for rocks of modern 



island arcs. This conclusion is consistent 
with what we believe to be the origin of 
the Jurassic arc — namely, as an island 
arc formed partly under water off the 
coast of Mesozoic South America. The 
arc may have been underlain by a thin 
sialic crust or it may have formed en- 
tirely upon simatic crust. Further meas- 
urement of Sr isotopic compositions may 
give us confidence to answer that ques- 
tion more fully, for the key is whether or 
not the ratios are within the range of 
island arc values or whether, as now, 
they lie between island arc and cordil- 
leran arc averages. 

The strontium isotope ratios that have 
been presented provide for a few tenta- 
tive conclusions: 

1. The Jurassic arc may have formed 
on simatic or thin sialic crust, and the 
Sr ratios of the Jurassic extrusives are 
close to those measured for rocks of 
island arcs. 

2. Rocks of post-Jurassic age exhibit 
Sr isotope ratios significantly above those 
for island arcs but very close to the 
mean value of rocks of other cordilleran 
arcs, such as the Sierra Nevada. 

3. The very high (0.732 to 0.775) pres- 
ent-day 87/86 ratios of Precambrian (?) 
gneiss of the western cordillera exclude 
the possibility that these rocks or any 
older sialic rocks of similar isotopic com- 
position have directly participated in the 
genesis of the igneous rocks of the arc. 

Potash variation within the western 
cordillera. Dickinson and Hatherton in 
a series of papers 86 ' 87,88 have shown 
that potash concentration at given Si0 2 
levels in modern volcanic rocks of island 
arcs increases roughly linearly with 
depth to the Benioff zone. This empirical 
correlation has served as one of the bases 
for the hypothesis that island arc mag- 
mas originate in or near the Benioff zone. 
It is further asserted by some, ourselves 
included (Year Book 70, pp. 344-349), 
that the magma has formed by partial 
melting of the oceanic crust of the de- 
scending lithospheric slab. The defense 
of these speculations is not our objective 

here, but they show the direction of our 
thinking as regards potash variation, for 
if potash varies with depth to the 
Benioff zone, then it may prove possible 
to reconstruct positions of past Benioff 
zones from the measurement of potash 
in igneous rocks. Before proceeding on 
such an optimistic undertaking, how- 
ever, it is instructive to consider the pot- 
ash content of Cenozoic volcanic rocks of 
the western cordillera with a view toward 
determining how much significance can 
be attached to K 2 0-/i (depth to Benioff 
zone) relationships within a given arc. 
Following Dickinson, we first obtain 
variation diagrams of K 2 vs. Si0 2 for 
various suites, then plot K 2 at fixed 
Si0 2 values for the appropriate depths 
to the Benioff zone. All analyses re- 
ported here are XRF and were com- 
pleted at the University of Montreal 
with the assistance of B. Gunn. 

The late Cenozoic suites have been 
selected to represent sampling sites that 
lie at different elevations above the 
Benioff zone. It should be noted, how- 
ever, that all samples included here were 
collected at or west of the crest of the 
western cordillera, a distance spanning 
considerably less than half of the total 
east-west extent of Andean volcanic 

Variation diagrams of K 2 vs. Si0 2 
are presented in Fig. 53 for analyses of 
late Cenozoic rocks of southern Peru, in 
Fig. 54 for analyses of late Cenozoic 
rocks of northern Chile, 83, 84 ' 90 and in 
Fig. 55 for analyses of late Cretaceous- 
early Tertiary intrusives of the southern 
Peruvian part of the Andean batholith. 
Jurassic volcanic rocks are not included 
here because of paucity of analyses and 
general uncertainty as to the extent of 
change in initial K 2 concentrations due 
to later alteration. 

The variation diagrams are self- 
explanatory, but it is appropriate to 
mention a few anomalies. The andesitic 
series of Fig. 53(a) was collected from 
the flanks of volcano Misti, a large ande- 
sitic cone near Arequipa. The slope of 



§ 5 
O 4 

* 3h 


1 - 












60" 65" 

Si0 2 (o/ ) 





* h=165 km 
■ h=190km 






% Si0 2 

Fig. 53. (a) Variation diagram K 2 vs. Si0 2 for late Cenozoic volcanic rocks of southern Peru. 
Andesitic volcanic rocks were all taken from the flanks of Misti volcano near Arequipa and dis- 
play a slope greater than that observed over the larger range of compositions. Ignimbrites form 
the more silicic end members of this series, (b) Variation diagram K 2 vs. Si0 2 for late Cenozoic 
volcanic rocks situated along a zone east of sample localities of Fig. 53a and overlying a deeper 
part of the Benioff zone. 

increasing potash with silica is much 
higher than for any other suite collected 
in southern Peru and is much greater 
than would be expected on the basis of 
typical increases even within cordilleran 
arcs. A second anomaly may be analyt- 
ical, for Guest's 90 analyses show consist- 
ently higher potash than those of Pichler 
and Zeil for northern Chile (see Fig. 54). 
The potash measurements of Guest are 
consistent with our results for southern 

A plot of K 2 vs. depth to the Benioff 
zone is shown in Fig. 56 for each of these 
suites at Si0 2 of 55%, 60%, 65%, and 

70%. Where rock compositions within a 
given suite did not extend to a given Si0 2 
concentration, those points were not in- 
cluded. Shown for comparison are the 
approximate best (eyeball) fit K 2 0-h 
lines from the accumulated data pre- 
sented by Dickinson (1970). 91 It is im- 
mediately clear that no simple relation- 
ship exists for these various suites 
between potash and depth to the Benioff 
zone. Indeed, within the limits of error 
of the variation diagrams, there is no 
significant difference in potash concen- 
tration between the included suites. 
There is some indication that the late 




O 3 


h = 175 

▲ GUESTU969) 

• Z. & P. (1967) 
P. & Z. C1969) 



60 65 

% Si0 



Fig. 54. Variation diagram K 2 vs. Si0 2 for 
late Cenozoic volcanic rocks of northern Chile 
(Zeil and Pichler. 1967; 83 Pichler and Zeil, 
1969; 84 Guest, 1969 90 ). Guest's analyses tend to 
show consistently higher K L >0 values than those 
of Pichler and Zeil and could indicate analyt- 
ical differences. Guest's analyses are chiefly of 
ignimbrites and tend to lie along the best fit 
line of Fig. 53a. Trend lines of Fig. 53a are 
plotted for comparison. In general, potash 
values, as measured by Pichler and Zeil, tend to 
be lower than those observed for late Cenozoic 
volcanic rocks of southern Peru. 

Cenozoic suite at 190 km above the 
Benioff zone exhibits slightly higher pot- 
ash, but the increase is slight. In addi- 
tion, potash concentrations for Si0 2 > 
65% diverge markedly from the trend 
lines of Dickinson's island arc volcanic 
rocks. Interestingly, the late Cretaceous- 
early Tertiary intrusive rocks exhibit 
potash concentrations nearly identical to 
those measured for late Cenozoic vol- 
canic suites. This fact may indicate that 

O 3 

2- !*#*""• • •• 





6 65 

Si0 2 (%) 


Fig. 55. Variation diagram K 2 vs. Si0 2 for 
intrusive granitic rocks of late Cretaceous- 
early Tertiary (?) age for region west of Are- 




1 2 3 

© e 

' - " _ - 65 - 


_-©- -<5 



__ „ »- 
_ --o- 

Si0 2 
O 55 
■ 60 
© 65 
* 70 





Fig. 56. Plot of K 2 vs. h (depth to Benioff 
zone) for various suites samples in southern 
Peru and northern Chile. Values for h are 
taken from James (1971). 92 

the Benioff zone has not altered its posi- 
tion dramatically since Cretaceous time. 

The fact that the potash content of 
Cenozoic volcanic rocks 150 km above 
the Benioff zone does not differ signifi- 
cantly from that measured for rocks 200 
km above the Benioff zone suggests that 
the extrusives are all derived from 
roughly the same source region. The dif- 
ferent extrusive sites may be a function 
more of available fracture systems along 
which magma can be transported than of 
varying source region. The Cenozoic 
volcanic rocks of northern Chile exhibit 
the same phenomenon. Although not 
shown here, we sought initially to split 
the Chilean analyses into groups for 
rocks 150 ± 25 km above the Benioff 
zone and for those 200 ± 25 km above 
the zone. The best fit lines for the two 
variation diagrams proved identical. 

We have not yet examined trends for 
other trace elements across the arc. One 
feature of special importance does arise 
from the major element analyses: The 
intrusive rocks of the Andean batholith 
exhibit average compositions less silicic 
than the average rock of the Cenozoic 
volcanic arc. Mean compositions of the 
intrusive rocks collected are presented in 
Table 17, together with some mean com- 
positions for representative andesitic and 
intrusive suites. It should be noted that 
the Andean volcanic suites do not include 
any rhyolitic ignimbrites, and their mean 
silica content is thus lower than would 



TABLE 17. Mean Chemical Compositions of Andesitic and Granitic Suites 










Si0 2 










A1 2 3 










Ti0 2 










Fe 2 3 +FeO 




































Na 2 










K 2 














1. Cenozoic volcanic rock of Little Sitkin Island, Aleutian Islands (Snyder, G. L., 1959, Geology 
of Little Sitkin Island, Alaska, U. S. Geol. Surv. Bull. 1028-H, pp. 169-210; cited in Dickinson. 1970). 

2. Average analyses of Cenozoic rocks from Kurile Islands (Markinin, E. K., 1968, Volcanism as 
an agent of formation of the earth's crust, in The Crust and U^pper Mantle of the Pacific Area, Mono- 
graph 12, 413-423, AGU, Washington, D. C; cited in Dickinson, 1970). 

3. Average analyses of andesites from Kurile Islands (Gorshkov, G. S., 1970, Volcanism and the 
Upper Mantle, Plenum Press, New York, 385 pp.). 

4. Average composition of Cenozoic volcanic rocks (exclusive of ignimbrites) of Arequipa area of 
southern Peru (28 analyses). 

5. Average andesite (13 analyses) from Misti volcano near Arequipa, southern Peru. 

6. Average dacite (15 analyses) of western flanks of the western ranges near Arequipa, southern 

7. Weighted average of Mesozoic plutonic rocks of southern California batholith (Larsen, E. S., 
1948, Batholith and associated rocks of Corona, Elsinore, and San Luis Rey quadrangles, southern 
California, 182 pp., Geol. Soc. Amer. Mem. 29; cited by Dickinson, 1970). 

8. Weighted average of Mesozoic plutonic rock of the Sierra Nevada batholith (F. W. Dodge, 
unpublished data, cited by Dickinson, 1970). 

9. Average of 34 analyses of plutonic rocks of the Arequipa area of southern Peru. 

be obtained by averaging volcanics of the 
late Cenozoic series. From Table 17 it is 
apparent that the mean composition of 
the intrusive rocks collected (some 34 
specimens not representing a weighted 
sample) is closer to a typical andesitic 
series than a granitic series. As some 
effort was made to sample rocks in ap- 
proximate proportion to their occurrence, 
we suggest that the Andean intrusives 
are compositionally more akin to ande- 
sitic suites than to batholithic granitic 
suites. In particular, the composition of 
Andean intrusives is significantly more 
mafic than those of the Sierra Nevada 
batholith. The similarity of composition 
between the intrusive rocks of the 
Andean batholith and the extrusive rocks 
of the volcanic chain lends support to the 
concept that granitic plutons are the 
deep-seated comagmatic equivalents of 
the extrusive rocks of the arc. Moreover, 
it appears possible that the late Cenozoic 

volcanic rocks of the central Andes will 
prove to have average compositions more 
silicic than the andesitic suites of island 
arcs, possibly quite comparable to the 
batholithic means for other cordilleran 

The Geochemistry of Basalts from Ice- 
land and the Reykjanes Ridge 

Stanley R. Hart and Jean-Guy Schilling* 

Basalts from spreading ocean ridges 
differ in a number of ways from basalts 
erupted on oceanic islands and island 
arcs. While all three tectonic environ- 
ments may produce low-K tholeiites, the 
spreading ridge tholeiites are generally 
characterized by unusually low concen- 
trations of elements such as Rb, Cs, Ba, 

* Narragansett Marine Laboratory. Univer- 
sity of Rhode Island, Kingston, Rhode Island. 



and Sr, and by relatively low Sr 87 /Sr 86 

The distinctive character of spreading 
ridge basalts is generally ascribed to deri- 
vation from mantle which is depleted in 
these elements relative to the mantle which 
acts as a source for oceanic island and 
island arc basalts. In Year Book 70, pp. 
353-355, we commented on the problem 
of how two distinct mantle sources can 
maintain their identities when both are 
in convective motion and are in close 
proximity. For example, many oceanic 
islands occur near the crests of spreading 
ridges and yet the basalts from both re- 
gions are quite distinct. 

In a study of one such transition 
zone between an oceanic island and 
a spreading ridge environment, Schil- 
ling 93 studied a suite of basalts collected 
from Iceland and dredged at intervals 
along the Reykjanes Ridge south of Ice- 
land. The basalts showed only minor 
variations in .major element chemistry, 
with most of the ridge basalts being 
quartz tholeiites and the Iceland samples 
being dominantly olivine tholeiites. The 
basalts were quite different in minor and 
trace element chemistry, however, with 
K, P, Ti, and the rare-earth elements 
(REE) decreasing fairly regularly from 
Iceland to a point some 550 km south on 
the Reykjanes Ridge. This variation was 
accompanied by a progressive depletion 
of the light REE relative to the heavy 
REE. We report here an extension of 
this study, to include the elements Rb, 
Cs, Ba, and Sr and the isotopic composi- 
tion of Sr. There is currently much inter- 
est in the volcanology of Iceland because 
of plume convection or mantle hot spot 
models 94,95 in which Iceland is consid- 
ered the surface volcanic expression of 
one such mantle plume. The geochemical 
results reported here obviously have an 
important bearing on the concept of 
mantle plumes. 

Figure 57 shows the location of sam- 
ples from Iceland and the Reykjanes 
Ridge which were analyzed for this 
study. Sr 87 /Sr 86 data (Fig. 58) show two 



Fig. 57. Map of Iceland showing location of 
the Reykjanes ridge spreading axis and loca- 
tions of analyzed samples. 

general groupings of values, with a rela- 
tively sharp transition between them. 
Samples from Iceland and for 200 km 
south along the ridge show remarkable 
uniformity, with an average value of 
0.70304 (the standard deviation of these 
13 samples is ±0.003%). This is the 
most precise demonstration of regional 
uniformity of Sr isotope ratios in vol- 
canic rocks yet achieved. In the zone 
from 200 to 250 km south on the ridge, 
some fluctuation in Sr 87 /Sr 86 values is 
noted, whereas beyond 250 km south on 
the ridge the values are consistently 
lower by 0.05%, averaging 0.70270. Thus 
the Sr isotope ratios from this area con- 
form to the pattern observed throughout 
the ocean basins: island volcanics in- 

<? .7028 


in .7027- 


1 1 

• • 























• • 



1 1 








IOO O 100 


Fig. 58. Plot of Sr 87 /Sr 86 ratio as a function of 
distance from southern tip of Reykjanes Penin- 
sula on Iceland. Values are relative to 0.70800 
for E and A Sr standard. 



variably having Sr 87 /Sr 86 greater than 
0.7030 and spreading ridge basalts show- 
ing, with a few exceptions, Sr 87 /Sr 86 less 
than 0.7030 (the average Sr 87 /Sr 86 for 18 
spreading ridge basalts analyzed in this 
laboratory is 0.70265). The isotopic data 
demonstrate that there are two discrete 
mantle sources being tapped for magmas, 
one feeding Iceland and the northern 200 
km of the Reykjanes Ridge, and the 
other feeding the ridge from 250 km on 
south, with perhaps some mixing or over- 
lap in the region between 200 and 250 km. 
The trends of selected trace elements 
and trace element ratios are shown in 
Figs. 59 and 60 as a function of distance 
from the southern end of the Reykjanes 
Peninsula in Iceland. The samples far- 
thest south on the ridge (60° N) show 
the depleted trace element characteris- 
tics (for example, K and Sr, Fig. 59) 
which are typical of spreading ridge ba- 
salts. In fact, these samples have lower 
concentrations of these elements than 
any other spreading ridge basalts yet 
analyzed. Spreading ridge basalts aver- 
age about 1000 ppm K and 130 ppm Sr, 
compared to 330 ppm K and 70 ppm Sr 
for the southern Reykjanes Ridge ba- 
salts. The elements Rb, Cs, and Ba and 
the REE are also unusually low. Various 
trace element ratios, on the other hand, 

Fig. 59. K and Sr concentration as a function 
of distance from tip of Reykjanes Peninsula. 

Fig. 60. K/Sr and K/Ba ratio variation as a 
function of distance from tip of Reykjanes 

are not as highly fractionated as average 
spreading ridge basalt. For example, 
spreading ridge basalts average — 1000 for 
K/Rb, 75,000 for K/Cs, and 120 for 
K/Ba, whereas the basalts from 60-61 ° N 
on the Reykjanes Ridge average ^770, 
59,000, and 60 for these ratios. If the low 
trace element concentrations of typical 
ridge basalts are ascribed to previous 
depletion processes operating on the 
mantle source, it seems that more ex- 
treme operation of this depletion process 
is much less effective in producing ele- 
ment fractionation. It is also interesting 
to note that mantle which has under- 
gone this severe depletion in the LIL 
(large-ion-lithophile) trace elements is 
still capable of generating a tholeiitic 
basalt that is normal as far as major ele- 
ment chemistry is concerned. 

The trace elements show a fairly regu- 
lar increase in concentration ( and change 
in ratios such as K/Sr and K/Ba, Fig. 
60) going north toward Iceland. At 200 
km, where the step in Sr 87 /Sr 86 occurs, a 
small but noticeable step also occurs in 
the trace element trends. On Iceland 
most of the samples closest to the ridge 
have higher concentrations of the LIL 
trace elements than any of the ridge 



samples. However, concentrations ap- 
pear to decrease on going north across 
Iceland, and the final range in concentra- 
tions on Iceland is about the same as on 
the ridge, though the average is some- 
what higher on Iceland. Though some 
of the Iceland tholeiites have element 
concentrations as low as spreading ridge 
basalt (for example, K = 700 ppm, Sr = 
90 ppm) , the element ratios are quite un- 
like those for basalts derived from de- 
pleted mantle (for example, the K/Rb, 
K/Cs, and K/Ba ratios of the Iceland 
basalts are about 500, 45,000, and 32, 
respectively, and are relatively uniform 
and independent of actual element con- 
centrations) . Thus, the delineation of 
two discrete mantle source regions in the 
Iceland area on the basis of Sr isotope 
data is corroborated by trace element 
ratios such as K/Rb and K/Ba (and 
La/Sm 1 ). 

While the Sr isotope data would seem 
to suggest only limited mixing of these 
two mantle sources ( or the derived mag- 
mas), the trace element concentrations 
do show a progressive change with dis- 
tance from Iceland which is suggestive 
of a mixing process. This possibility can 
be evaluated with two-element mixing 
plots (Fig. 61 is one example) . This plot 
and others like it suggest that the surface 
and dredge samples are aligned on sepa- 
rate trends and that mixing of a high- 
concentration Iceland basalt with a low- 
concentration ridge basalt is too simple a 
model to explain all the variations ob- 



U 0.02 


20 30 40 

Ba , ppm 



Fig. 61. Cs content vs. Ba content of Iceland 
basalts (open circles) and ridge basalts (solid 
circles). Note approximate fit to trends with 
different Ba/Cs ratio. 

served along the ridge. Other, more com- 
plicated mixing models have not been 
evaluated but may be allowed by the 
data. Note that two-element plots (such 
as Fig. 61) suggest a division of the sam- 
ples at the shoreline, not at the 200 km 
break shown by the Sr isotope data. 

A model which we are in the process of 
evaluating quantitatively includes the 
following features: 

1. Existence of both depleted and un- 
depleted mantle, with a fairly sharp 
boundary about 200-250 km south of 
Iceland. This boundary coincides with a 
concave break in slope of the ridge crest. 

2. Variable degree of partial melting 
of these mantle sources, with the greatest 
degree of melting occurring under both 
the north edge of Iceland and the south- 
ern end of the ridge. 

3. Change in mantle mineralogy, with 
the mantle underlying the submarine 
ridge containing a phase such as plagio- 
clase which is absent under Iceland itself. 
This feature is required to explain the 
two different trends shown by the trace 
elements (such as those shown in Fig. 61) 
and to explain the large scatter of Sr con- 
centrations for Iceland basalts as com- 
pared to the ridge basalts. 

4. Near-surface crystal differentiation 
will produce minor variations in trace- 
element concentrations. 

Geochemical data such as these ob- 
viously cannot prove the existence of a 
deep-mantle plume under Iceland; how- 
ever, the clear delineation of two mantle 
source regions is consistent with a plume 
model for Iceland. These data show that 
such a plume does not feed significant 
quantities of material into the spreading 
ridge beyond about 200 km of Iceland. 

Geochemistry of Ultramafic Inclusions 
from Salt Lake Crater, Hawaii 

N. Shimizu 

Trace elements in basalts have been 
studied by many investigators because 
they offer important information on the 
chemistry of the upper mantle. Another 



possible link to upper-mantle chemistry 
is the ultramafic inclusions in basalts 
and kimberlites. Judging from the major- 
element compositions, some of these in- 
clusions appear to be potential sources of 
basaltic magma; others appear to be 
residua left after partial melting or crys- 
tal accumulates from basaltic magma. 

To better understand the trace element 
chemistry of these rocks and especially 
to understand the possible correlation be- 
tween major element and trace element 
chemistry, the samples to be analyzed 
should be petrologically well documented. 
Preferably, a suite of samples collected 
from a single locality should be studied. 

Ultramafic inclusions found in an 
olivine nephelinite tuff of Salt Lake 
Crater, Hawaii, are among the best sam- 
ples in these respects. Seventeen rock 
samples have been made available for 
the study, all of which were described by 
Kuno (1969) 9G and Kuno and Aoki 
(1970). 97 On petrological grounds they 
are classified into three groups: lherzolite 
(olivine + orthopyroxene + clinopyrox- 
ene + spinel), olivine eclogite (olivine 
-j- clinopyroxene + garnet ± spinel ± 
orthopyroxene) and olivine-free eclogite 
(clinopyroxene + garnet ± orthopyrox- 
ene ± spinel). Olivine-free eclogite is 
better called garnet pyroxenite, because 
clinopyroxene in this rock type is not 
quite omphacitic (White. 1966; 98 Jack- 
son and Wright, 1970 ") . 

K, Rb, Cs, Sr, and Ba abundances and 
Sr 87 /Sr 86 were measured in separated 
clinopyroxenes because in these mineral 
assemblages the above trace elements are 
largely concentrated in clinopyroxene. 
The mineral separation was made only 
by handpicking. Grains surrounded only 
by fresh surfaces and free from any 
visible inclusions or alteration product 
were carefully selected under a binocular 
microscope, and this process was repeated 
several times with successive grinding 
into finer grain size. The separates were 
then washed with acetone, w r arm 2.5 N 
HC1 with ultrasonic agitation, and dis- 
tilled water prior to processing. The sam- 

ples were processed with low-level blank 
chemistry procedures. The total blanks 
are: 7 ng K; 0.01 ng Rb; 0.3 pg Cs; 0.2 
ng Sr; 0.8 ng Ba. These low-level blanks 
made it possible to handle a small 
amount of sample, for example, 10-20 
mg, which, in turn, made it easy to obtain 
extremely pure mineral separates. The 
chemistry blanks are negligible, except 
for Ba in a few cases. 

In general, concentrations of these ele- 
ments in clinopyroxenes are low com- 
pared with those in the same mineral 
from the same locality reported in the 
literature (Griffin and Murthy, 1969; 10 ° 
Philpotts et al, 1972 101 ) . Figures 62 and 
63 show Rb-Sr and K-Rb relationships. 
Arrows in these figures indicate the 
trends of increasing whole-rock MgO/ 
SFeO ratios. From these figures, the 
trace element variations appear to be 
interrelated, and particularly in Fig. 62, 
three groups of rocks fall in separate re- 
gions of different Rb-Sr characteristics. 






-[ 1 1 1 — i — r-r 

1 1 r 



i i i i i 

50 100 200 
Sr, ppm 

Fig. 62. Rb vs. Sr plot for clinopyroxenes 
from ultramafic inclusions of Salt Lake Crater. 
Open circles, lherzolites; solid circles, olivine 
eclogites; crosses, olivine-free eclogites. Ar- 
rows indicate the trend of increasing whole-rock 




50 - 



-i 1 1 1— -i — i i i 


-i 1 1 1 — i — i — r~r 






j i i i i i i 

J I L 

I I 

.01 .02 .05 .1 .2 Rb, ppm 

Fig. 63. K vs. Rb plot. Symbols are same as those in Fig. 62. 

The relationship between major and 
trace elements is better illustrated in Fig. 
64, in which trace element abundances in 
clinopyroxenes are plotted against whole- 
rock MgO/^FeO ratios. The correlation 
is particularly good in olivine eclogites. 
It is rather surprising that even at sub- 
ppb concentration levels (Cs in olivine 
eclogites) , a trace element appears to 
behave quite systematically relative to 
major elements. This may indicate that 
contamination from the enclosing magma, 
if any, has been very minor. Correlation 
also exists between the trace elements 
and other major elements because the 
oxide components in these rocks vary in 
a systematic manner with respect to 
MgO/SFeO ratio (Kuno, 1969). The 
whole-rock MgO/3FeO ratio was chosen 
here because it should be a good indicator 
of magmatic processes ( such as partial 
melting or fractional crystallization). 

According to Kuno (1969) 96 and Kuno 
and Aoki (1970), 97 the lherzolites would 
represent residua left after partial melt- 
ing, and the eclogites (both olivine- 
bearing and olivine-free) would represent 
crystal accumulates from the basaltic 
magma. Following the idea that these 

rocks represent the solid phases which 
were once equilibrated with liquid, the 
trace element variations illustrated in 
Fig. 64 may be explained in terms of 
partition relations of these elements be- 
tween solid and liquid. An important 
feature in this context is the increase of 
Sr with increasing MgO/^FeO observed 
in the lherzolites and olivine eclogites. 
If MgO/^FeO represents the degree of 
partial melting (the higher the ratio, the 
greater the degree) or the degree (or 
stage) of crystallization (the higher the 
ratio, the earlier the stage), the increase 
of Sr can be interpreted in two ways: 
that Sr was more enriched in residual 
clinopyroxene left after the larger degree 
of partial melting or that Sr was more 
enriched in cumulative clinopyroxene 
crystallized in the earlier stage of frac- 
tionation. These interpretations implic- 
itly assume that the present mineral 
assemblage represents the primary as- 
semblage at the time of solid-liquid 
equilibrium. Both interpretations re- 
quire Sr to be partitioned more into 
clinopyroxene than liquid. However, this 
is quite unlikely in the light of the avail- 
able data on Sr partitioning (Philpotts 



| |_|C-R7p)| ITC 

— - /'"M C~/'~" 1 

K, Sr 

LnL i\l W L_ 1 

2 45 6 

8910 12 

UL t.(wL. 

17 21 

27 30 


( III 

I I \( 

| | 

1 ^ 

4 00 



.00 3'"-* 1 

Cs / 



0021 Y 


^ Sr 


I 1 




' 1 



* -— 


. / 



x ' 


- / 



r •". 0004 

.' / 


1 _ 

/ f 



-4 :- 


i > 





i :~ 
i ■ 


K" v — 

'••x' "'■•. 


i i 


-4 --— 



,' -x 

x x 
x Rb 

i i 



- .2 

. 1 


I .02 

4.5 4 3.5 3 

MgO / FeO" 



Fig. 64. Variations of trace element contents of clinopyroxenes with respect to whole-rock 
MgO/2FeO. Sample numbers for lherzolites and olivine eclogites are given at the top of the 
figure. Cs in no. 30 is 0.0039 ppm, in no. 12 is 0.0004 ppm. 

and Schnetzler, 1970; 102 Shimizu and 
Akimoto, 1971 103 ). In addition, Sr con- 
tents (as high as 140 ppm) and low K/Sr 
ratios (about 0.3) are difficult to recon- 
cile with clinopyroxene-liquid relation- 

Therefore, it is indicated that the pres- 
ent mineralogy is a recrystallized assem- 
blage and that there must have been a 
phase at the time of the primary solid- 
liquid equilibrium into which Sr was pre- 
ferentially partitioned. This phase must 
have been stable only at higher tempera- 

ture and must have disappeared during 
recrystallization at subsolidus tempera- 
tures. One possibility for this phase from 
the standpoint of trace element geochem- 
istry is plagioclase, because Sr is strongly 
partitioned into plagioclase relative to 
liquid (Berlin and Henderson, 1968; 104 
Philpotts and Schnetzler, 1970; 102 Drake 
and Weill, 1972 105 ). A possible positive 
Eu anomaly due to plagioclase eval- 
uated by a model calculation based 
on the normative plagioclase contents 
and the partition coefficients was 



found to be about 5%, being practically 
invisible in the chondrite-normalized 
REE diagram. The problem, then, is 
whether this is plausible in the light of 
phase petrology. 

The stability of plagioclase at high 
pressures has been studied in both syn- 
thetic and natural systems of basaltic 
composition in connection with the 
basalt-eclogite transition (Kushiro and 
Yoder, 1966; 106 Green and Ringwood, 
1967; 107 Ito and Kennedy, 1971 108 ) and 
in systems of lherzolitic composition in 
connection with defining the mantle min- 
eral facies (Green and Hibberson, 
1970; 109 Herzberg, 1972 110 ). 

There is no way to estimate the P and 
T of the primary equilibrium, but, as- 
suming that the subsequent recrystalli- 
zation took place by isobaric cooling, the 
estimated P and T for these rocks should 
be reasonably close to the stability field 
of plagioclase-bearing assemblages re- 
ported by those investigators. Pressure 
and temperature conditions for reequilib- 
rium were therefore estimated for lherz- 
olites by the major element compositions 
of pyroxenes. The major element analy- 
sis was made by a Material Analysis 
Corporation electron probe at the Geo- 
physical Laboratory. 

The method used here is the same in 
principle as that discussed by Boyd 
(1973). 111 Temperature was estimated 
from the diopside composition on the 
diopside-enstatite solvus determined for 
the pure binary system by Davis and 
Boyd (1966) 112 at 30 kb. As shown in 
Fig. 65, the effect of iron was considered 
by extrapolating the diopside composi- 
tions in the quadrilateral to the binary 
join Di-En. The effect of A1 2 3 was, 
however, ignored. Although the effect of 
A1 2 3 on the temperature estimate seems 
to be negligible at 30 kb as inferred from 
the data by Boyd (1970), 113 it is totally 
unknown at pressures in the stability 
field of spinel instead of garnet. Pres- 
sure was estimated from the A1 2 3 con- 
tent of enstatite, using the data by Mac- 

Fig. 65. Compositional relations of olivines, 
orthopyroxenes and clinopyroxenes in lherzo- 
lites. Temperatures ( 1000-1300° C) are from 
Davis and Boyd (1966). 112 

Gregor (1973) 114 on the system MgO- 
Al 2 3 -Si0 2 . According to Boyd (1970), 
at 30 kb and 1200°C there is about 1% 
difference in A1 2 3 between two enstat- 
ites, one at the binary join En-Al 2 3 and 
the other at the apex of the three-phase 
triangle En-Di-Gar. At lower pressures 
this difference is presumably greater. 
Therefore, 1.5% A1 2 3 was added rather 
arbitrarily to the analyzed values, and 
pressure was then estimated from the 
A1 2 3 isopleths. Figure 66 illustrates the 
estimated P and T for lherzolites to- 
gether with some relevant phase bound- 
aries. As shown in the figure, the esti- 
mated conditions are reasonably close to 
the stability limit of plagioclase in lherz- 
olitic composition. It thus seems possible 
that plagioclase participated in the pri- 
mary equilibrium for these lherzolites. 

Another problem is whether or not this 
also holds true for the olivine eclogites. 
Since the major element compositions of 
the olivine eclogites are neither lherz- 
olitic nor basaltic, the phase boundaries 
shown in Fig. 66 cannot be applied with- 
out modification. As discussed by Green 
and Ringwood (1967), important factors 
controlling the positions of boundaries 
would be a complicated combination of 
the MgO/MgO + FeO of the rock, the 




i 1 1 






'lherz. dry 


'< \i 






1 1 /^^^LHERZ. Px 

■ >'' 


A B / 


C ( FO g2+ AN 59 



i i i 





P, kb 

Fig. 66. Estimated P-T conditions of reequi- 
librium for lherzolites. The wet and dry solidi 
of a lherzolite are from Kushiro et at. (1968). 115 
The phase boundaries A and B are from Green 
and Ringwood (1967) 107 and C from Green and 
Hibberson (1970). 109 

degree of silica saturation, the normative 
plagioclase composition, and the molecu- 
lar ratio of normative olivine and hypers- 
thene to plagioclase. It is impossible to 
discuss this possibility in great detail 
until the phase relations are studied ex- 
perimentally on the specific composition 
of olivine eclogite. However, it seems 
important to note the relative position of 
the boundaries B and C in Fig. 66. These 
are remarkably close to each other in 
spite of the difference in chemical com- 
positions. The figure also indicates that 
at 1000°C, for example, the eclogitic 
assemblage is stable even at lower pres- 
sures in basaltic composition than the 
spinel lherzolite assemblage in lherzolitic 
composition. Since the major element 
compositions of the olivine eclogites are 
of intermediate nature between lherzolite 
and olivine tholeiite in many respects, it 
may not be unreasonable to assume that 
the relative position of boundaries B and 
C still holds true for the olivine eclogite 

and the lherzolite now in question. The 
molecular ratio of normative olivine to 
plagioclase is about 1 in olivine tholeiite 
B (Green and Ringwood, 1967), while 
those in olivine eclogites are definitely 
higher. This would be a good indication 
that these rocks contain excess olivine, 
assuming that both breakdown of plagio- 
clase and formation of garnet are essen- 
tially reactions between equal numbers of 
molecules of plagioclase and olivine. The 
presence of excess olivine components in 
olivine eclogites suggests that olivine is 
stable at any pressures of present 

From these arguments, it seems possi- 
ble that the hypothesis based on the trace 
element variation also holds true for the 
olivine eclogites. Judging from the pres- 
sure, the liquid in equilibrium with those 
rocks may well have been tholeiitic 
rather than alkalic. 

In contrast to this, the olivine-free 
eclogites (or garnet pyroxenites) may 
be of a different origin. Green (1966) 116 
noted the similarity in chemical composi- 
tions between the liquidus clinopyroxene 
crystallized from an alkali basalt at 13- 
18 kb and a whole-rock eclogite (No. 18 
of Kuno, 1969). The texture also sug- 
gests that garnet and orthopyroxene are 
exsolved from clinopyroxene. Green 
(1966) then concluded that the eclogite 
of this type was originally monomineralic 
clinopyroxene crystallized from alkali 
basalt at high pressures which later ex- 
solved orthopyroxene and garnet during 
cooling. The sample used in his discus- 
sion was also studied here (No. 18 in Fig. 
62). The concentrations of trace ele- 
ments in a hypothetical liquid calculated 
from the measured trace-element abund- 
ances in clinopyroxene and partition 
coefficients experimentally obtained at 
high pressures are 7000 ppm K, 10 ppm 
Rb, and 800 ppm Sr, which are reason- 
able as alkalic basalt. This appears to 
support Green's argument on the origin 
of this type of rock. 



Differential Dissolution Technique 

(DDT): Chemical Separation of 

Crystals from Glass 

N. Shimizu and S. R. Hart 

One of the most serious problems in the 
experimental studies on partitioning of 
trace elements between silicate crystal 
and liquid is the separation of phases 
from the small amount (less than 20 mg) 
of high-pressure run product. Even if the 
grain size is large enough for separation 
by the conventional heavy liquid tech- 
niques, it is impossible to eliminate crys- 
tals with a very thin lining of glass on the 
surface or those with tiny glass inclu- 
sions. Even 0.5% impurity would change 
the results significantly, because the 
trace elements now in question (K, Rb, 
Cs, Sr, and Ba) are strongly partitioned 
into liquid relative to crystal. In order to 
remove this difficulty, a new technique 
has been developed and is presented here. 

This technique is based on the fact 
that clinopyroxene and glass (quenched 
liquid) are dissolved by dilute HF at 
significantly different rates. Several tests 
using natural clinopyroxenes and glass 
showed that glass is dissolved much 
faster than clinopyroxene. Therefore, in 
principle, starting with a mixture of 
clinopyroxene and glass, we can separate 
them by putting the mixture in dilute HF 
for an appropriate period of time to dis- 
solve only the glass. By repeating the 
treatment several times, only pure clino- 
pyroxene is left behind. If the amounts 
of trace elements in the solution are 
measured, and if the weight loss of the 
sample during the treatment is measured, 
we can calculate the concentrations of 
elements in the dissolved sample (glass). 
Ideally, in successive treatments the 
amounts of trace elements in the solution 
should come down to the chemical blank 
levels before any clinopyroxene is dis- 
solved. If, however, clinopyroxene is dis- 
solved simultaneously, or if any selective 
leaching of trace elements from clino- 
pyroxene occurs, the amounts of trace 

elements in the solution would never 
come down to the blank level. 

In order to test this technique, we pre- 
pared a mixture of clinopyroxene and 
glass (clinopyroxene from a spinel lherz- 
olite from Salt Lake Crater, Hawaii; 
tholeiitic basalt glass from the Juan de 
Fuca Ridge) in a proportion so that the 
simultaneous dissolution or the selective 
leaching of trace elements from clinopy- 
roxene could be detected. In this mix- 
ture, the amount of Sr in the clinopyrox- 
ene is more than twice that in the glass, 
and Ba in the clinopyroxene is about 
equal to that in the glass. A detailed 
experimental setting is described below: 

A pulverized mixture of clinopyroxene 
(6.5 mg) and glass (2.8 mg) was put in a 
Teflon centrifuge tube and 0.5 ml dilute 
HF (4.9%) and a few drops of 2.5 N HC1 
were added. After leaving the mixture 
for 20 minutes at room temperature, the 
solid was centrifuged down and the solu- 
tion was withdrawn by a Teflon pipette 
and transferred to a Teflon beaker. The 
residue was rinsed three times with 2.5 N 
HC1 and then three times with H 2 and 
the rinses were added to the Teflon 
beaker. Spike solution was weighed into 
the beaker and the solution was evapo- 
rated to dryness. It was then dissolved 
with 2.5 N HC1 and the elements were 
separated by a cation-exchange resin col- 
umn and analyzed by a mass spectrom- 
eter. The residual solid was dried in the 
centrifuge tube and weighed to determine 
the weight loss of the sample during this 
treatment. The treatment was repeated 
three times and the amounts of the trace 
elements in the successive solutions came 
down to levels close to the chemical 

The results are summarized in Table 
18 and illustrated in Fig. 67. As shown 
in the figure, the amounts of trace ele- 
ments in the solution are close to the 
blanks at the third step. The largest dif- 
ference was observed for Sr and may be 
explained by the persistence of about 5 
fig of glass (0.2% of that originally pres- 
ent) . The concentrations of these ele- 



TABLE 18. Experimental Results 




1st, 2nd, and 4th 















































*Basalt glass D10 from Juan de Fuca Ridge, pulverized in reagent grade acetone. 
fSame glass not pulverized with acetone. 

ments were calculated and given in the 
table. It is interesting to note that glass 

1 is enriched in these elements relative to 
glass 2. The glass was initially pulver- 
ized in reagent grade acetone, so it might 
have been contaminated by the acetone. 
Glass 1 is in excellent agreement with 
D10F (pulverized in acetone) and glass 

2 with that not pulverized in acetone. It 
is likely that the contaminant went into 
the first solution and the pure glass (free 
from contamination) was dissolved in 

the second step. In the fourth step, the 
remaining sample was dissolved with the 
normal HF and HC10 4 . The concentra- 
tions are in good agreement with the 
clinopyroxene of the starting material. 

These results indicate that clinopyrox- 
ene is clearly separated from glass in this 
technique and that selective leaching of 
trace elements from the pyroxene, if any, 
is insignificant. 

In the course of the experiments, we 
observed armoring of undissolved grains 

10 r 


1.0 - 


1.0 r 



.012 5 











Fig. 67. Variations of trace elements in solutions during successive HF treatments. Dashed 
lines with B designate blank levels of these elements. 



with insoluble precipitate (presumably 
CaF 2 ) and adsorption of dissolved ions 
on the surface of remaining grains, but 
these were removed by adding a few 
drops of 2.5 N HC1 to the HF and by 
rinsing the sample with 2.5 N HC1 after 
the HF was withdrawn. 

This technique was originally designed 
to study the partitioning of trace ele- 
ments between clinopyroxene and liquid 
and is found to be applicable to other 
problems, such as those described by 
Shimizu and Kushiro (this Report, be- 
low). An advantage of this technique is 
that one does not have to worry about 
the grain size of the samples. During 
high-pressure synthesis of clinopyroxene, 
it is generally difficult to obtain a grain 
size large enough to do fission track map- 
ping or even microprobe analysis very 
easily. A disadvantage is that it is not 
applicable to systems containing more 
than two phases. It is not certain at this 
stage whether the technique is directly 
applicable to a system such as amphibole- 
liquid, or whether some other acid or 
more diluted HF would be needed. 

Trace Element Content of Liquid Formed 

by Partial Melting of a Garnet 

Lherzolite at High Pressures: 

A Preliminary Report 

N. Shimizu and I. Kushiro 

Many experimental works have been 
carried out on the genesis of basalt mag- 
mas from major element chemistry, but 
very little has been done on the trace 
element behavior during melting proc- 
esses at high pressures. This was due to 
the difficulty of separating phases by 
conventional magnetic and heavy liquid 
techniques from fine-grained run prod- 
ucts. An attempt has been made to over- 
come this difficulty by developing a new 
technique of separation. A detailed ac- 
count of the technique (DDT, differen- 
tial dissolution technique) is given else- 
where in this annual report (Shimizu and 
Hart, pp. 268-270). Applying this tech- 

nique, we have analyzed trace element 
concentrations in liquid formed by par- 
tial melting of a garnet lherzolite at high 
pressures. Although the present paper 
reports only some preliminary results, 
more detailed study along this line would 
provide possible explanations of trace 
element variations in basaltic magmas 
and a quantitative treatment of basalt 
genesis from the standpoint of trace ele- 
ment geochemistry. 

The material used in this study is a 
sheared garnet lherzolite in kimberlite 
from Thaba Putsoa Pipe, Lesotho. The 
sample was supplied by Dr. F. R. Boyd 
of the Geophysical Laboratory. Major 
element chemistry of whole-rock and 
constituent minerals was described by 
Nixon and Boyd (in preparation) . High- 
pressure melting experiments were car- 
ried out with a solid-medium piston cylin- 
der apparatus using a graphite crucible 
as a sample container. Phase relations 
and major element compositions of co- 
existing phases were given by Kushiro 
(1973) , 117 

K, Rb, Cs, Sr, and Ba were analyzed 
by isotope dilution on the glass (quenched 
liquid) of two samples, one run at 15 
kb, 1450 °C, and the other at 20 kb, 
1500 °C (both anhydrous) by sepa- 
rating the glass from crystals with the 
differential dissolution technique. Al- 
though the chemical blanks are negligi- 
ble, a large error is involved in the 
weighing of samples of 1 mg or less. The 
overall uncertainty in concentrations is 
estimated to be 10-20%. 

Table 19 summarizes the trace element 
contents of whole rock, clinopyroxene 
separated from the same rock, and two 
glasses formed at the conditions given 
above. The rock appears very fresh as 
an inclusion in kimberlite; nevertheless, 
the whole-rock trace element system has 
been affected by alteration judging from 
the unusually high Rb/Sr ratio for a 
mantle material. Cs could not be meas- 
ured because of unfavorable spiking ratio 
(133/135 > 100). From K/Cs ratio in 
one of the glasses (liquid A) and from K 



TABLE 19. Trace Element Concentrations 

Liquid A 

Liquid B 

PHN 1611 

15 kb, 

20 kb, 










K, ppm 





3476 § 


Rb, ppm 








Cs, ppm 



Sr, ppm 








Ba, ppm 




































Sample wt. 

(mg) anal. 





*Enrichment factor, glass/starting material. 
jConcentration ratios of liquid A/liquid B. 
tK 2 = 0.68%; microprobe value = 0.70%. 
§K 2 = 0.42%; microprobe value = 0.47%. 

in the whole rock, Cs in the whole rock 
would be as high as 0.5 ppm. 

Berg (1968) 118 noted that high Cs and 
resultant low K/Cs in kimberlite inclu- 
sions are associated with secondary zeo- 
lite (probably analcime) . The presence 
of zeolite also increases Rb and Rb/Sr. 
The situation may be similar for the sam- 
ple used in the present experiments, al- 
though zeolites have not been observed. 
In Table 19, K 2 values determined in- 
dependently by microprobe and DDT 
are compared. They are in excellent 
agreement considering the possible 10— 
20% weighing error of the present experi- 

If a liquid is in equilibrium with a 
solid phase in which a particular trace 
element is preferentially partitioned, or 
if two trace elements are effectively frac- 
tionated by the solid phase, trace element 
abundances (absolute as well as relative) 
in the liquid are to some extent con- 
trolled by this phase. In the present case, 
however, the major solid phases coexist- 
ing are olivine, orthopyroxene, and 
clinopyroxene for liquid A; and olivine 
and orthopyroxene for liquid B. Con- 
ceivably these trace elements may not be 
fractionated by these phases. It is ex- 

pected, therefore, that the concentrations 
of trace elements in the liquid are de- 
pendent only on the degree of partial 
melting. In other words, the enrichment 
factors for all elements are equal. This 
appears to be approximately correct, al- 
though there are some variations of en- 
richment factors (Table 19) which re- 
main unexplained at this stage. For 
example, in both liquids A and B, Rb 
and Ba seem to be more enriched than K 
and Sr. Judging from the concentration 
ratios of the two liquids (A/B in Table 
19), enrichment of K, Rb, and Sr is 
rather straightforward, whereas Ba does 
not behave as systematically (which 
might partly be due to inhomogeneity of 
the starting material with respect to Ba) . 
As shown by Kushiro (1973), the 
liquids are tholeiitic in major element 
composition (quartz tholeiite for A and 
olivine tholeiite for B), while trace ele- 
ments in these liquids are more compli- 
cated. For example, although the concen- 
tration levels of K, Sr, and Ba are in the 
range shown by many tholeiites, the 
ratios of the elements are anomalous (i.e., 
low K/Rb and Sr/Ba, high K/Sr) . These 
peculiarities are undoubtedly due to the 
anomalous trace element pattern of the 
starting material. 



It may be pointed out, however, that 
it is possible to obtain concentration 
levels of trace elements in natural basalts 
by direct partial melting of a lherzolite 
with concentrations of these elements 
similar to those in the garnet lherzolite 
used in the present experiments. If this 
is true, about 1% phlogopite may be 
present as a primary host phase for the 
alkali trace elements and Ba. If phlogo- 
pite does not persist during partial melt- 
ing of 20-40% at 15-20 kb pressures, the 
liquid should contain about the same 
amounts of trace elements as those meas- 
ured in the liquids of the present experi- 
ments. In this case there is no need to 
consider wall-rock reaction (Green and 
Ringwood, 1967) 119 or eclogite fractiona- 
tion (O'Hara, 1968) . 12 ° If, however, the 
trace element content in the upper man- 
tle is much lower, then less partial melt- 
ing is required, or it is necessary to con- 
sider that some process such as wall-rock 
reaction or eclogite fractionation is oper- 

Trace Element Contents of Clinopyrox- 

enes from Garnet Lherzolites in 


N. Shimizu and F. R. Boyd 

The garnet lherzolite is widely as- 
sumed to be a major constituent of the 
upper mantle. It is actually abundant 
among ultramafic inclusions found in 
kimberlites. The recent works by Boyd 
and Nixon (1972) 121 and Boyd (1973) 122 
revealed that there are two groups of 
garnet lherzolites, the sheared nodules 
and the granular ones. The sheared nod- 
ules appear to be more primitive upper 
mantle material, judging from the major 
element compositions, and can be a 
potential source of basalt magmas (Ku- 
shiro, 1973), 123 while the granular nod- 
ules are depleted in the basaltic compo- 
nents (CaO, A1 2 3 , and FeO relative to 

The trace element characteristics of 
garnet lherzolites are therefore impor- 

tant for a better understanding of the 
upper mantle chemistry and processes 
operating in the upper mantle. There 
have been only a few trace element data 
on these nodules available (Griffin and 
Murthy ,1969; 124 Allsop et al, 1969; 125 
Hutchison and Dawson, 1970 ; 126 Phil- 
potts et al., 1972 127 ). The sheared nod- 
ules, in particular, have not been studied 
because they are rare compared with the 
granular ones and their chemical charac- 
teristics have only recently been recog- 
nized. Two technical difficulties that 
were observed by these authors are the 
low-level concentrations of trace ele- 
ments and the difficulty of obtaining pure 
mineral separates. 

In this article, we report K, Rb, Cs, 
Sr, and Ba abundances and Sr 87 /Sr 86 
ratios of clinopyroxenes separated from 
garnet lherzolites. Our samples are pre- 
dominantly from the Thaba Putsoa kim- 
berlite pipe, Lesotho, with one (JJG 352) 
from the Bultfontein kimberlite pipe 
(supplied by Dr. I. Kushiro of the Geo- 
physical Laboratory). Two of them, 
PHN 1611 and 1596, are sheared nodules, 
and PHN 1569 and JJG 352 are typical 
granular nodules. PHN 1600E4 is a 
discrete nodule, a diopside single crystal 
about 2 cm in its smallest dimension, and 
we analyzed the central part and the edge 
(C and A, respectively) to find if there 
are any significant variations within a 
grain. A garnet separated from PHN 
1596 was also analyzed for comparison. 
Clinopyroxenes were selected for inten- 
sive study because trace elements are en- 
riched in this mineral relative to other 
primary phases except phlogopite. The 
mineral separation was made only by 
handpicking under a binocular micro- 
scope. The separates were washed with 
warm 2.5 TV HC1 and water prior to the 
chemical processing. The chemical blanks 
are : 7 ng K ; 0.01 ng Rb ; 0.3 pg Cs ; 0.2 ng 
Sr; and 0.8 ng Ba (negligible except in 
garnet) (see Table 20). These low-level 
blanks made it possible to handle small 
amounts (less than 20 mg) of extremely 
pure samples. Sr 87 /Sr 86 ratios were 







GO t> 










































oo co 

CM rH 

id oo 



rH CO 

o o 

o o 

© d 

rH CD 


CM i—l i-H 


N CO iO 

d co h 

CO CO ^ 
TjH ^H t^H 


-H -H 



















lO OS 




r-3 d 

1— I 





i— i 




i— i 

1— 1 
o oo 

O rH 

1— 1 











l— H 




d d 








io io 










i— 1 















id cm 








CO iO 


r— 1 






00 iO 






rH OS 





i— i 

iO i— i 








1— 1 







1— 1 



1— 1 


T— 1 1— 1 

1— 1 






rH CM 

1— 1 








CO i— i 














1— 1 



i-h CM 









iO OS 














1— 1 



CM 00 















d d 


rH 00 

i-H O 

o o 

o o 

o o 

d d 


iO co 

rH O 

O O 

d d 

CM t> 


iO o 


rH O 


rH 00 

o o 
o o 
o o 

d d 

OS i— i 


o o 

d d 

co oq 
d rH 
oo oo 















PhPh W 



OS . O 

iO , +T CO 
XPh - 



















HH — 

.5 CM 


& 08 





CD hh 

§ a 


2 "° 

r3 "S 






.a -a 

W PQ <? 












measured with a precision of ±0.00014 
or better at 95% confidence level. The 
results are summarized in Table 20 and 
illustrated in Figs. 68 and 69. 

The whole-rock trace element concen- 
trations (PHN 1611) are not accounted 
for by the clinopyroxene data, indicating 
that greater quantities of trace elements 
should be along the grain boundaries 
and/or in the secondary minerals, al- 
though the rock appears fresh as an in- 
clusion in the kimberlite. Potassium K a 
radiation image of this rock showed that 
potassium is concentrated in the grain 
boundaries. The intergranular phase in 
which the potassium is concentrated can- 
not be resolved under the microprobe but 
is intergrown with thin veinlets of ser- 

The results for diopsides show that K 
ranges from 110 to 430 ppm, and Rb 
ranges from 0.015 to 0.69 ppm, being var- 
iable by more than a factor of 40. The 
Sr contents of the clinopyroxenes show 

a bimodal distribution, one concentration 
at about 90 ppm (the sheared nodules) 
and the other at about 500 ppm (the 
granular nodules). K/Rb ratios in the 
clinopyroxenes from the sheared nodules 
and a discrete diopside crystal are un- 
usually high, especially the value ob- 
served for PHN 1596 clinopyroxene 
(22301), which is higher than any other 
values ever observed. There seems to be 
no coherence between K and Rb in the 
clinopyroxenes, which is quite anomalous. 
However, the coherence between Rb and 
Cs is relatively good judging from the 
narrow-range variation of Rb/Cs ratios 
(59-139). As shown in Fig. 69, the 
clinopyroxenes from the sheared nodules 
appear to be distinguished from those in 
the granular nodules primarily by the 
difference in their Sr contents. The Sr 
contents of the clinopyroxenes from the 
sheared nodules are very similar to those 
observed in clinopyroxenes from the 
Hawaiian nodules. 




T 1 » I I I I | 







' ' i i ' 


t — r 


T 1 I I I 





i i il 


Rb ppm 

Fig. 68. K vs. Rb plot for clinopyroxenes from garnet lherzolite inclusions in kimberlite. Open 
circles, data found in the literature (Griffin and Murthy, 1969; Allsop et al, 1969; Philpotts et 
al, 1972). Solid circles, present study. Dots, clinopyroxenes from Hawaiian ultramafic inclusions 
(Shimizu, this Report). 



i i i i i i 

I " 1 


i i i ! t j r j ■ 














/ # 1569 





.00 3 


A • 





l i i i i l 

i . i 

• i i i i i i l 



Sr , ppm 


Fig. 69. Rb vs. Sr plot, Symbols are same as those in Fig. 68. 

Comparing the two groups, the clino- 
pyroxenes from the sheared nodules con- 
tain higher K but lower Rb, Cs, Sr, and 
Ba concentrations than those from the 
granular nodules. These concentrations 
are not consistent with the hypothesis 
that the granular nodules are residua 
produced by partial melting and extrac- 
tion of liquid from more primitive rocks 
like the sheared nodules because, if this 
were true, all these trace elements would 
be depleted in the granular nodules be- 
cause they are strongly partitioned into 
silicate melt relative to clinopyroxene. 

The high Sr contents of the granular 
nodules cannot be explained by partition 
of Sr between clinopyroxene and silicate 
melt because the partition coefficient of 

about 0.1 for wt Sr in cpx/wt Sr in liquid 
(Shimizu and Akimoto, 1971) 128 requires 
that the clinopyroxene should have been 
equilibrated with a silicate melt con- 
taining more than 5000 ppm Sr, and that 
is unrealistic. 

The peculiar trace element character- 
istics of these clinopyroxenes cannot be 
primary features; most represent altera- 
tion by some unknown secondary proc- 
esses operating in the mantle. The low 
Sr 87 /Sr 86 ratios of these clinopyroxenes 
(Table 20) may rule out the possibility 
of crustal contamination. An indication 
of a secondary process may be the pres- 
ence of phlogopite in some of the granu- 
lar nodules. The phlogopite is sometimes 
present in two generations with different 



major element compositions. These 
phlogopites are unusually high in Ti0 2 
(about 2-4 wt %) , and an enrichment of 
Ti0 2 in clinopyroxene has been observed 
in the immediate vicinity of second- 
generation phlogopite which is especially 
rich in Ti0 2 . These facts suggest that the 
trace element variations may also be due 
to partial equilibration with phlogopite 
or with the fluid phase from which the 
phlogopite was formed. 

The high Sr content of clinopyroxenes 
from granular nodules may be another 
indication of the effects of secondary 
processes. Bell and Powell (1969) 129 
have measured concentrations of 5000- 
11,000 ppm of Sr in carbonatites and 
hence it appears possible that the high 
content of Sr in the diopsides from kim- 
berlite nodules is due to secondary reac- 
tions between the nodules and carbonate- 
rich kimberlite magma. 

The unusually high K/Rb and K/Cs 
ratios in the clinopyroxenes from the 
sheared nodules remain unexplained. If 
these high ratios also resulted from sec- 
ondary reaction with some fluid or 
magma, such a magma or fluid should 
have anomalously high K/Rb and K/Cs 
but normal Rb/Cs, or some phase should 
be present by which K, Rb, and Cs abun- 
dances in clinopyroxene are controlled. 

From the results presented it is obvious 
that until a plausible mechanism is found 
to explain their characteristics and varia- 
tions, caution must be exercised in apply- 
ing trace element data from kimberlite 
inclusions to any geochemical model of 
the upper mantle. 

Metasomatic Theory; Some Examples of 
Diffusion and Infiltration 

R. C. Fletcher and A. W. Hojmann 

The importance of metasomatism (i.e., 
the change in bulk composition of a 
rock) in metamorphic processes has been 
debated by petrologists for many decades. 
Dehydration and decarbonation reac- 
tions are familiar and well-established 
examples; alkali metasomatism and 

granitization are perhaps among the 
more controversial subjects of petrology. 

In recent years, new methods have been 
employed to determine the extent to which 
mass transport is involved in metamorphic 
processes. Examples are the use of isotopic 
tracers (e.g., Shieh and Taylor, 1969) 13 ° 
and the experimental and field observa- 
tions of Vidale (1969). 131 The electron 
microprobe makes possible the detailed 
analysis of chemical gradients in meta- 
morphic-metasomatic rocks; from these 
gradients and a knowledge of the mineral 
equilibria involved, it is in principle 
possible to deduce the transport mecha- 
nism responsible for these gradients. 

In the last several decades, D. S. 
Korzhinskii has established a physico- 
chemical basis for the thermodynamics 
and the modes of transport involved (for 
a recent summary, see Korzhinskii,. 
1970) . 132 In order to derive simple trans- 
port equations, he used the concepts of 
diffusion and infiltration metasomatism, 
where both mechanisms involve chemical 
transport through a pore fluid that is in 
local equilibrium with the solid phase. 
In the case of infiltration metasomatism, 
the fluid percolates through and reacts 
continuously with the rock. The result- 
ing chromatographic effects have been 
reviewed by Hofmann (1972) 133 (see also 
Korzhinskii, 1970, 132 1973 134 and Hof- 
mann, 1973 135 ). In diffusion metasoma- 
tism, the mobile components diffuse 
through a standing pore fluid. Korzhin- 
skii (1970) has given some general trans- 
port equations for this case. 132 Our pur- 
pose here is to show solutions for a few 
specific cases of pure diffusion and for 
one case of combined diffusion and in- 
filtration. These solutions were derived 
by analytical and numerical methods 
given by Carslaw and Jaeger (1959) 136 
and by Crank (1956). 137 These examples 
illustrate important qualitative features, 
namely, the effects of different liquid- 
solid equilibria (referred to as "iso- 
therms") on the concentration profiles in 
metasomatic replacement fronts. The 
isotherm shapes are qualitatively the 



Fig. 70. Diffusion profiles for three continuous isotherms A, B, and C. The fluid concentrations 
Ca and Cc are shown as dashed lines; solid concentrations Sa and Sc are shown as solid lines. A 
single solid line represents the concentration profile for Sb and Cb. The distance units on the z 
axis are arbitrary. The isotherms C — f(S) are shown in the inset. 

same as those discussed by Hofmann 
(1972) , 133 

The simple graphical representation of 
an isotherm (as used in Figs. 70 through 
73) implies an isothermal, quasi-isobaric 
equilibrium relationship between a solid 
and a fluid where only one component is 
independently variable. This includes 
idealized cases of two-component ion ex- 
change where the total ion concentration 
in the fluid cannot be fixed. Because the 
variability in the total ion concentration 
is generally small, our examples are qual- 

itatively, and probably semiquantita- 
tively, applicable to two-component ion 
exchange, which is geologically more in- 
teresting than the one-component system. 
The general mass balance for the sys- 
tem containing porous rock and fluid is 

(If), +(£),=» •" 

if the concentration C i<r of the ith com- 
ponent in the rock (solid plus fluid) and 
the flux Ji are continuous and differenti- 
ate functions of time t and distance z 

30 40 


Fig. 71. Diffusion profiles for a continuous, S-shaped isotherm. See legend for Fig. 70. 









Fig. 72. Diffusion profiles for a discontinuous isotherm. The solubility gap is represented by a 
dotted line. See legend for Fig. 70. 

and if the transport is one dimensional 
in the z direction. 

When the flux Ji is specified by the 
diffusion and infiltration processes, the 
mass balance becomes 

= —fiUi 


On the right-hand side of equation 2, the 
first term accounts for infiltration, the 
second for diffusion; Di is the diffusion 
coefficient of the ith component in the 
fluid ; Cij is its concentration in the fluid ; 
Ui is its velocity due to infiltration; and 
(3 is the porosity of the rock. The formu- 
lation of equation 2 requires that all 
transport occur in the fluid phase, that is, 
longitudinal diffusion in the solid is 

Fig. 73. Combined infiltration and diffusion profiles (labeled a) and pure diffusion profiles 
(labeled b) for an isotherm with a "pseudosolubility gap." The a profiles are distinguished by 
nearly horizontal segments (concentration plateaus) at the origin (C = 1, S = 1) and at the con- 
centrations C = 0.5 and S = 0.15. For further explanations, see text. 



We now impose the condition of local 
equilibrium, represented by the simple 
isotherm C«, r = / (C*,/) , and obtain 


V dt ) z " dc^/dd, 

fiDt ( d 2 C u \ 

,t V ^ 2 ) 

( dC u \ 


Cl^i,r/ d^i,) 

Equation 3 illustrates the dependence of 
both infiltration and diffusion on the 
slope of the isotherm dC i>r /dCij. If the 
isotherm contains a solubility gap, there 
is a concentration discontinuity and 
equation 3 is not valid at this boundary. 
This means that the solid acts as an in- 
stantaneous sink or source of the com- 
ponent in the fluid. 

The examples discussed below and il- 
lustrated in Figs. 70 through 73 are com- 
puted for diffusion (and infiltration) into 
a semi-infinite medium in which the con- 
centrations are initially equal to zero. A 
constant finite concentration is main- 
tained at the origin z = 0. For conveni- 
ence, we have assumed that the porosity 
(3 is sufficiently small so that the concen- 
tration of i in the total rock, C ir , is ap- 
proximately equal to the concentration in 
the solid phase alone, C is . For the pur- 
pose of representing several isotherms 
and diffusion profiles in one diagram, we 
have relabeled the fluid concentration C 
and the solid concentration S. 

An important feature of all solutions 
for diffusion with these boundary condi- 
tions is that the distance from the origin 
of any given concentration is propor- 
tional to the square root of time. This 
means that the shapes of the profiles 
shown in Figs. 70 through 72 (which have 
arbitrary distance units) are independent 
of time. An equivalent number of time 
steps has been used for these profiles in 
order to permit comparison of the rela- 
tive rates of movement. 

Diffusion profiles for curved isotherms 
are most easily obtained by finite- 
difference methods. We have used a 
modified Schmidt method (see Crank, 
1956, chapter X) , 137 Figure 70 shows the 

concentration profiles for three continu- 
ous isotherms, A, B, and C. The straight- 
line isotherm (line B in Fig. 70) yields a 
profile, C B and S B , that can be tested by 
an analytic solution (in the form of an 
error function) of the diffusion equation. 
When the isotherm is convex upward 
(line A), the fluid concentration C A 
moves faster than C B but the solid con- 
centration S A is retarded compared to S B 
over most of the concentration range. 
The most interesting profile is obtained 
for the concave-upward isotherm C. The 
fluid concentrations C c form a nearly 
linear profile with appreciable curvature 
only in the lower ten percent of the con- 
centration range. The solid concentra- 
tion S c forms an S-shaped profile. This 
shape is quite distinctive and may be 
used to distinguish between different dif- 
fusion mechanisms. If, for example, the 
operating mechanism were simple diffu- 
sion through the solid phase with a con- 
stant diffusion coefficient and no trans- 
port through a pore fluid, there could be 
no reversal in curvature, at least not for 
the kind of boundary conditions we have 

The isotherm shapes shown in Fig. 70 
are typical of solid solutions such as the 
Na-Ca exchange equilibrium of plagio- 
clases with chloride solutions (Orville, 
1972). 138 The shapes shown in Figs. 71 
and 72 resemble the alkali exchange 
equilibria for alkali feldspars at different 
temperatures as determined by Orville 
(1963) , 139 The fluid-concentration pro- 
files for the latter two isotherms are simi- 
lar, but the solid-concentration profiles 
differ in that there is a discontinuity at 
the solubility gap. 

The diffusion profiles for a solubility 
gap (Fig. 72) were computed by Neu- 
mann's method (see Carslaw and Jaeger, 
1959, chapter XI) , 136 In this method, the 
concentration profiles on either side of 
the concentration discontinuity are ob- 
tained by two separate analytic solutions, 
and the velocity of the moving boundary 
(concentration discontinuity) is obtained 
numerically. The solubility gap acts as a 



sink, and the concentration profile be- 
tween the origin and this sink closely 
resembles that of a steady-state gradient, 
i.e., in this case a straight line. 

For the purpose of showing the effect 
of combined diffusion and infiltration, we 
have used an isotherm (Fig. 73) which 
exhibits a "pseudo-solubility gap." Anal- 
ysis of a true concentration discontinuity 
is not possible with our present finite- 
difference procedure. Comparison of the 
diffusion profiles in Figs. 72 and 73 shows 
a close similarity : The concentration dis- 
continuity of Fig. 72 is replaced by a 
steep concentration gradient which rep- 
resents the "pseudo-solubility gap" in 
Fig. 73. 

When infiltration is superimposed on 
diffusion, both solid and fluid concentra- 
tion profiles become distinctly different. 
The pronounced concentration plateaus 
are characteristic of the infiltration proc- 
ess. The sharp replacement front that 
would be produced by infiltration alone 
(Hofmann, 1972) 133 is here damped by 

These examples show that the shape of 
a metasomatic concentration profile can 
be diagnostic of both the shape of the 
isotherm and the transport mechanism. 
In general, it will be necessary to know 
the approximate shape of the isotherm in 
order to determine the metasomatic 
mechanism from a given concentration 

Ion Microprobe Analysis of an Isotonic 
Diffusion Experiment on Biotite 

A. W. Hofmann, B. J. Giletti, J. R. Hinthorne, 
D. Comaford, and C. A. Andersen 

The ion microprobe offers several po- 
tential advantages in the analysis of dif- 
fusion experiments on minerals. Among 
these are in situ spot analysis of isotopic 
composition to obtain diffusion profiles 
on individual mineral grains and the 
ability to study isotopic variations in 
depth (with a resolution on the order of 
tens of angstroms) as successive layers of 

the mineral are eroded by ion sputtering. 
The resulting high resolution makes the 
technique particularly suitable for the 
analysis of the very slow diffusion rates 
common in geological processes. Experi- 
ence with metamorphic rocks shows that 
characteristic transport distances due to 
solid-state diffusion are commonly on the 
order of 0.1 to 10 mm. As this distance 
is usually (depending on boundary con- 
ditions) proportional to the square root 
of time, it is reduced to the order of 0.01 
to 1 /xm during an experiment lasting 0.1 
year if the natural process required 10 
million years. The ability to analyze 
diffusion gradients with a resolution of 
10 to 100 A would therefore enable us to 
perform the diffusion experiments at geo- 
logically realistic temperatures with no 
need for extrapolation. 

In order to test the general feasibility 
of this approach we analyzed a biotite 
flake that had previously been subj ected to 
a tracer diffusion experiment. The biotite 
was sealed in a gold tube with an aqueous 
2.0 normal alkali-chloride solution that 
contained sodium, potassium enriched 
with K 41 tracer, and rubidium. The 
K 39 /K 41 ratio of the solution was 0.95 
(which contrasts with a natural potas- 
sium value of about 13.5), the Na/K 
mole ratio was 5.73, and the Rb/K ratio 
was 480. The charge was subjected to a 
temperature of 650 °C and a pressure of 
2.0 kb for 41 days. The diffusion experi- 
ment is described in greater detail by 
Hofmann and Giletti (1970) . 14 ° 

The isotopic analyses reported here 
were made on an ion microprobe mass 
analyzer (Andersen and Hinthorne, 
1972) 141 at Applied Research Labora- 
tories at Sunland, California. Figure 74 
shows a sketch of the biotite flake and 
the spots analyzed. Each spot was bom- 
barded by a beam of 0.4 nanoamps of 
negatively charged oxygen-16 ions which 
had been accelerated to 20 kV in the pri- 
mary mass spectrometer. The K 39 /K 41 
ratios were recorded at successively 
deeper levels as continued ion sputtering 
removed material from the surface. The 



Fig. 74. Schematic sketch of biotite flake. The small rectangles show the sites of the spot analy 
ses. The large rectangle represents the rastered area. 

results of the spot measurements are 
shown in Fig. 75. The contours were 
drawn by hand to show the overall con- 
sistency of the data. The K 39 /K 41 ratios 
increase from the edge toward the center 
of the flake. From these contours a rough 
estimate can be made of the relative 
transport rates in the different crystallo- 

graphic directions. Because each spot 
has a diameter of about 6 fim, much of 
the steep lateral concentration gradient 
near the edge was not resolved. If we 
take the measured isotopic ratio to rep- 
resent the value at the center of a spot, a 
ratio K 39 /K 41 = 11.5 is found at a depth 
of >2000 A when measured 5 pm from 

10 20 30 40 


% 600 \- 




Fig. 75. Analytical results of the spot analyses. Each black dot represents the site of an analy- 
sis, and the associated number is the K 39 /K 41 ratio measured at that site. The approximate depth 
is measured from the top surface of the flake and is estimated from the elapsed sputtering time. 
The black bars show the diameter of each spot. The contours were drawn by hand. They repre- 
sent the ratios K 39 / 41 = 9, 11, 12, 13, and 13.5. 



the left edge of the flake (spot 1). This 
isotope ratio is reached at a depth of 400 
A (measured in the c direction) in the 
center of the flake (spot 4). Thus, the 
transport distance appears to be greater 
in the a,b direction than in the c direc- 
tion by a factor of 125. The same factor 
calculated using spot 5 instead of spot 1 
is only about 40. Because of the steep 
gradient near the edge, a small error in 
the estimate of the location of the spot 
or a small amount of beam wobble will 
cause considerable errors in the measured 
isotopic ratio of this spot. At present, we 
can only conclude that the transport rate 
of potassium parallel to the mica cleav- 
age is considerably faster than in the 
direction parallel to the c axis. In addi- 
tion to the spot analyses, we rastered an 
area 25 X 30 /xm (see Fig. 74) by rapid 
scanning of the primary ion beam. This 
procedure erodes the analyzed region 
slowly and yields more precise, though 

horizontally averaged, analyses at suc- 
cessively deeper levels from the surface 
of the grain. The dashed line in Fig. 76 
gives the combined results of the rastered 
analyses and the analyses from spot 4 
(open circles). 

Because of the preliminary nature of 
the experiment, a number of unverified 
assumptions are made in order to derive 
a diffusion coefficient from these results: 
(1) The depth has been estimated cor- 
rectly from the sputtering rate. (2) The 
first 80 A may be ignored because rem- 
nants of the carbon film deposited on the 
surface of the grain were sputtered to- 
gether with the first few unit cells of 
mica. (3) The beam drills a flat-bot- 
tomed hole and does not sputter from the 
hole walls. (4) The degree of vertical 
isotopic mixing in the crystal due to ion 
bombardment is negligible. (5) The 
measured peak ratio of masses 39 and 41 
equals the K 39 /K 41 ratio (a check on the 










4 - 

2 - 

I ' I 

I ■ I ' I ' I ' I ' I i I i I 'J 


J i i i i L 

J L_L 


J i L 





Fig. 76. K 39 /K 41 ratio as a function of depth from the surface of the flake. The dashed line rep- 
resents the combination of the analytical data of the rastered area for depths less than 350 A, and 
the results from spot 4 for depths greater than 400 A. The solid line represents an ideal diffusion 
profile (for the specific assumptions used to calculate this profile, see text). 



validity of this assumption is given by 
the fact that the measured ratio K 39 /K 41 
= 13.5 in the interior of the grain is 
within the range of currently accepted 
values for the isotopic abundances of 
natural potassium). (6) The total potas- 
sium concentration is constant over the 
range of the measured concentration pro- 
file (in fact, some potassium was lost 
from the biotite because of K-Na ex- 
change with the hydrothermal solution). 
(7) The boundary conditions can be ap- 
proximated by constant concentrations of 
K 39 and K 41 at the edge of the grain. 
The last assumption may be justified 
from our experience with diffusion ex- 
periments on biotite (Hofmann and Gi- 
letti, 1970) . 14 ° We have found that the 
isotopic composition of the fluid changes 
very rapidly at first, probably because of 
surface exchange and a small amount of 
sample dissolution. After this initial ex- 
change, the fluid composition and there- 
fore the boundary conditions for the 
solids change much more slowly as a re- 
sult of volume diffusion. The situation is 
further complicated by the apparent re- 
versal in the profile at an approximate 
depth of 40 to 80 A. If this reversal was 
caused by a dislocation or an incipient 
splitting along a cleavage plane, the 
effective boundary for diffusion at greater 
depths may have been at 80 A. 

Because of these uncertainties we used 
a constant boundary value of K 39 /K 41 = 
9.0 to obtain a theoretical diffusion profile 
for a semi-infinite solid. Given these 
boundary conditions and the above as- 
sumptions, the concentration C of a 
tracer isotope, say K 41 , is given by the 
equation (see Crank, 1956, chapter 3) 

L — C i 

= erf 



where C x is the concentration at the sur- 
face, C is the concentration in the inte- 
rior at time t = 0, x is the distance from 
the surface, and D is the diffusion co- 
efficient. The theoretical curve can be 
fitted to the measured values by adjust- 

ing the free parameter D. This procedure 
yields a diffusion coefficient for transport 
parallel to the c axis, D = 2.4 X 10" 18 
cm 2 sec" 1 . 

Clearly, these results are not definitive 
measures of potassium diffusion in bio- 
tite. They do show that extremely slow 
diffusion can be measured with the sput- 
tering technique and that diffusion in 
anisotropic crystals can be evaluated by 
three-dimensional mapping of isotopic 
abundances in a single grain. 

Direct Determination of Rock Rheology 
from Wavelength Selection in Folding 

Raymond C. Fletcher 

An assumption about the rheology of 
rock in slow tectonic flow (strain rate 
— 10~ 14 /sec) in the crust and mantle is a 
necessary part of any tectonic model. 
Data from steady-state creep tests at 
strain rates >10" 9 /sec can be adequately 
described (Goetze and Brace, 1972) 142 
by a power law (strain rate proportional 
to the deviatoric stress raised to a power 
n) . These results show significant non- 
linearity, n =3. The question remains as 
to whether this nonlinearity persists 
through a reduction in strain rate of five 
orders of magnitude or whether rocks at 
tectonic strain rates may be adequately 
treated as linearly viscous fluids (n = 1) . 
Most modelers of tectonic deformation 
assume a linear viscous rheology. 

In principle, a direct determination of 
rock rheology at tectonic strain rates is 
possible by interpreting rheology-sensi- 
tive features of natural structure on the 
basis of theoretical models which employ 
a range of rheologies. An analysis of 
buckling of a layer of power-law fluid 
embedded in a power-law medium has 
been carried out and used to fit data on 
wavelength selection in a natural single- 
layer fold. 

An appropriate tensor generalization 
of a single-component power law (Nye, 
1953; 143 Calladine and Drucker, 



1962 144 ), written for plane flow in the 
x, z plane, is 

&xx — "ft € XX Pi 

<*zz = 2ft e zz — p, 

&xz — "ft *-xz 


where p is the pressure, 2/x = \BJ 2 in 2) ] _1 , 
Jz — (Vi) (vxx — <?zz) 2 + vxz 2 , and B is a 
constant. If x and z are the principal 
axes of a stress vif, a small stress incre- 
ment atj and the associated strain-rate 
increment e# will approximately satisfy 
the linearized constitutive relations 

S% x = 2fi° (l/n)££, — p 7 


2/x° (1/n) 



— 2a° 

ft e xz 


where, from (1), o-^/ — a zz ° = 4/x° c^ . 
These relations are equivalent to those 
for an anisotropic viscous fluid whose 
viscosity in shortening or extension par- 
allel to x and z is less than its viscosity 
in shear by the factor (1/n). 

In the initial stages of buckling of a 
single embedded layer (Sherwin and 
Chappie, 1968), 145 wavelength compo- 
nents in the initial irregularity of the 
layer surfaces are selectively amplified, 
and the form of the layer tends, generally 
with much imperfection, towards a regu- 
lar train of folds with a preferred arc 
wavelength to thickness ratio, L p /h. 
During this phase the layer accommo- 
dates itself to bulk shortening by thick- 
ening approximately uniformly. When 
the limb dip of the folds reaches 10-15°, 
wavelength selection ceases, a sharp 
transition to buckle-shortening of the 
layer takes place, and thereafter the arc 
wavelengths of existing folds and the 
layer thickness are preserved. 

The analysis of this process shows that 
the power-law exponent of the layer n L 
and the parameter Q = (n L /n M ) 1/2 
(ftu°/ftL°) , where the subscripts L and M 
refer to quantities of the layer and em- 
bedding medium, respectively, may be 
determined from (1) L p /h; (2) A/A , the 
amplification of this component at the 

time wavelength selection ceases; and 
(3) the amount of layer thickening, rep- 
resented, for example, by T s = (h/h ) 2 , 
where h and h are the final and initial 
thicknesses. L p /h is readily observed, 
and from consideration of the probable 
initial amplitude of irregularities, A'/A 
can be estimated. T s may be estimated in 
the fortunate circumstance that some in- 
dicator in the layer records its finite 
strain or is bounded by the amount of 
shortening of the embedding medium. 
Only a single case is presently known for 
which T s can be reliably estimated (Gro- 
shong, 1971 ; 146 Fletcher and Groshong, 
1973 147 ). 

For a limestone layer embedded in 
shale in the unmetamorphosed Rochester- 
McKenzie Formation (Silurian) of the 
Valley and Ridge Province, central Penn- 
sylvania, L p /h = 9.4, A/A is estimated 
to lie in the range 20-100, and 1.07 < T s 
< 1.27. The rheological interpretation of 
these data is illustrated in Fig. 77. The 
upper bound on the layer thickening, T s 
= 1.27, which is determined from the de- 
formation of brachiopods in the shale far 
from the layer, implies that the limestone 
exhibits significant rheological nonline- 
arity, n L > 3. This might be expected for 
a rock deformed at tectonic strain rates 
and low temperature. The value T s = 
1.07 is derived from a calcite-twin strain- 


L d /h = 9.4 

Fig. 77. Evaluation of rheological parameters 
from data for a natural fold. 



gauge procedure (Groshong, 1971) 146 and 
probably underestimates the amount of 
thickening. It implies a value n L « 12. 
Corresponding values of Q range from 
about 1/40 for n L — 3 to 1/60 for n L = 
12. The results are not sensitive to the 
amount of amplification assumed. 

This is the only example of a direct 
determination of rock rheology at tec- 
tonic strain rates known to the author. 
A variety of finite strain gauges may be 
found in deformed rocks: deformed peb- 
bles, fossils, oolites, and twinned crystals, 
for example. The theoretical results are 
therefore applicable to a wide variety of 
rocks, provided the appropriate detailed 
observations are made. 

U-Pb Isotonic Studies of Zircons from 

the Baltimore Gneiss of the Towson 

Dome, Maryland 

B. Grauert 

The Towson dome is the easternmost 
protrusion at the end of a chain of man- 
tled gneiss domes along the crest of the 
Washington-Baltimore anticlinorium in 
the Maryland Piedmont (see Hopson, 
1964, Fig. 12) , 148 The complex formation 
of crystalline rocks exposed in the domes 
represents a Precambrian basement and 
is generally referred to as Baltimore 
Gneiss. The mantle of the domes is 
formed by the younger metasediments 
and metavolcanics of the Glenarm Series 
(see Higgins, 1972). 149 The lowest unit 
of the Glenarm Series, the Setters Forma- 
tion, consists chiefly of quartz-rich 
schists and mica quartzites (see Knopf 
and Jonas, 1929; 15 ° Hopson, 1964 148 ). 
Both the geologic evidence and previous 
isotopic age determinations (Tilton et al., 
1958, 151 1970 ; 152 Wetherill et al, 1966, 153 
1968 154 ) indicate that the Baltimore 
Gneiss had a polymetamorphic history: 
(1) A strong orogeny at about 1000-1100 
m.y. transformed earlier rocks to gneisses 
and migmatites (Hopson, 1964). (2) 
During an early Paleozoic orogeny, prior 
to 425 ± 20 m.y., the basement was re- 

activated and rose as domes beneath its 
mantle. The Gunpowder granite, a gra- 
nitic "intrusion" into the gneisses of the 
Towson dome and its southern mantle 
(see also p. 288, this Report), is inter- 
preted as rheomorphic Baltimore Gneiss 
(Hopson, 1964, p. 45; 148 Steiger and 
Hopson, 1964 155 ). (3) Temperatures re- 
mained sufficiently high for argon diffu- 
sive loss and strontium isotopic reequili- 
bration in minerals until about 290-280 
m.y. The northeastern part of the Tow- 
son dome, where the zircon samples were 
collected, consists of coarse augen gneiss 
grading into or intercalated with biotite 
and hornblende gneisses. The augen gneiss, 
which locally contains large crystals of 
microcline, is also known as Hartley 
augen gneiss (Knopf and Jonas, 1929) . 150 
Knopf and Jonas thought the porphyritic 
gneiss ". . . represents an igneous magma 
that has in other places formed a banded 
ribbon gneiss by its injection into the 
Baltimore Gneiss sediments" and was 
later crushed and recrystallized into 
augen gneiss. Hopson (1964, p. 43), how- 
ever, believes ". . . the augen gneiss is a 
local porphyroblastic phase of the Balti- 
more Gneiss, formed chiefly by potassium 
metasomatism along movement zones be- 
neath the Setters quartzite, in part dur- 
ing post-Glenarm doming and meta- 

To test these interpretations, I have 
separated and analyzed zircons from 
three gneiss samples, all collected within 
30 feet of a fresh, blasted outcrop at 
Gunpowder Falls (Towson quadrangle, 
Maryland, 3 69 420 / 43 64 470 , 1000-meter 
grid). At the sample locality the pre- 
dominant rock type is a coarse augen to 
flaser gneiss (sample BAL-11) with few 
migmatized intercalations of biotite 
gneiss (BAL-12). Veins of a medium- 
grained granite (BAL-13), showing more 
or less sharp contacts, have been de- 
formed along with the augen gneiss. All 
rock samples are relatively rich in zir- 
cons (about 2-5 g zircon in 10 kg of 



A microscopic study reveals great simi- 
larities in the morphological and optical 
properties of the zircons, reflecting three 
stages of zircon growth: (1) Translucent 
cores, sometimes clear but more often 
dark, can be observed in many of the 
larger crystals. They most likely repre- 
sent the oldest presumably inherited zir- 
con component. (2) The second stage is 
characterized by euhedral zircon growth, 
evident from well-displayed repeated 
zoning (up to ten zones) in the majority 
of the crystals. The zoned part of the 
zircons, which generally makes up more 
than 80% of the volume, is clear in the 
uranium-poor zircons and brownish 
translucent in those rich in uranium. 
From the microscopic picture it appears 
that the zircons went through a period 
during which the development of per- 
fectly euhedral and repeatedly zoned 
crystals was possible, thus suggesting a 
magmatic or migmatic environment of 
crystallization. (3) During a third stage, 

clear uranium-poor overgrowths modi- 
fied the shape of many crystals, forming 
all sorts of grains from nearly euhedral 
individuals to very irregularly shaped 
clods. The assumption that the clear 
overgrowths are poor in uranium is based 
on analyses of 17 zircon fractions from 
various types of Baltimore Gneiss which 
yielded the following results: clear frac- 
tions, 400—600 ppm U; brownish to pink- 
ish translucent fractions, 500-2500 ppm 
U; metamict fractions, more than 2500 
ppm U. The isotopic data of six zircon 
fractions are plotted on the concordia 
diagram in Fig. 78. The points fall 
within analytical uncertainty on a chord 
which intersects the concordia curve at 
about 450 m.y. and at 1180 m.y. Like the 
zircons from the Baltimore Gneiss of the 
Phoenix dome (see Year Book 71, pp. 
301—305) the pattern on the diagram does 
not fit any simple continuous diffusion 
model (Tilton, I960, 156 Wasserburg, 
1963 157 ) although the degree of discord- 

pb 20 6/u 238 



i 1 1 r 



BAL-II 663ppm 
BAL-II 660 ppm - 

BAL-II 789 ppm U 

BAL- 12 1393 ppm U 

BAL-13 1672 ppm U 
BAL- 12 2032 ppm U 



I I I I L 


Pb 207 /U 235 

Fig. 78. Concordia diagram showing the isotopic data of zircons from the Baltimore Gneiss of 
the Towson dome. 



ance could be interpreted as a function 
of the uranium concentration (Silver, 
1963) 158 or radiation damage. 

A straightforward explanation of the 
linear pattern in Fig. 78 would be the 
simple episodic loss of lead from 1180- 
m.y.-old zircons. Considering, however, 
the morphological properties mentioned 
above, the question arises whether the 
zircons represent mixtures of old (zoned) 
cores with 450-m.y.-old overgrowths, or 
whether they are entirely old, having 
been subject to radiogenic lead loss or 
uranium gain during an Early Paleozoic 
metamorphism. To test these possibil- 
ities, the uranium concentrations have 

been plotted versus the apparent Pb 206 / 
TJ238 ages m ;pjg 79 (i) if we assume 

a 1180-m.y.-old zircon population with 
about 500 ppm uranium as the starting 
material, a model of episodic uranium 
gain at 450 m.y. without new zircon 
growth involved (curve A) would ap- 
proximately fit the measured pattern; 
however, it is indistinguishable from a 
model of episodic lead loss from zircons 
with primary differences in uranium con- 
centration. (2) A model of uranium-rich 
overgrowth, e.g., with zircon substance 
containing 3000 ppm uranium, is shown 
by curve B. This model does not fit all 
the points and requires additional pri- 














ppm U 

Fig. 79. Diagram to explain various mixing models of 1180-m.y.-old zircons with zircon over- 
growths or uranium gain at 450 m.y. 



mary differences with respect to the 
uranium concentration. However, since 
the overgrown part is low in uranium, 
this possibility can be excluded. (3) In 
accordance with the relatively uranium- 
poor overgrowth observed, mixing curves 
have been drawn for sample BAL-12 
(2023 ppm U) assuming 500 ppm and 
1000 ppm uranium in the overgrown part 
(curves C and D, respectively). Ade- 
quate curves for the other zircon frac- 
tions analyzed would be similar but 
steeper. This model requires large differ- 
ences in the uranium concentrations of 
the older zircon components. In addition, 
the concentrations in the cores of the 
uranium-rich crystals must be extremely 
high — more than 8000 ppm. Such values, 
however, occur only in completely met- 
amict zircons. 

From these considerations the follow- 
ing conclusions are drawn: (1) The iso- 
topic pattern shown in Fig. 79 is pri- 
marily the result of episodic lead loss 
about 450 m.y. ago. Minor episodic ura- 
nium gain may be possible. (2) The 
clear overgrowth may be old or young. 
If it is young it would have a minor mix- 
ing effect in addition to the episodic dis- 
turbance of the older U-Pb systems. 

U-Pb Isotopic Studies of Zircons from the 

Gunpowder Granite, Baltimore County, 


B. Grauert 

Along Gunpowder Falls in Baltimore 
County, the Baltimore Gneiss of the east- 
ern Towson dome and its metasedimen- 
tary mantle have been intruded by a 
fine- to medium-grained gneissic quartz 
monzonite, the Gunpowder granite 
(Knopf and Jonas, 1929) . 150 According 
to Hopson (1964, p. 47) 148 the Gun- 
powder granite is a rheomorphic offshoot 
of the Baltimore Gneiss and was em- 
placed during deformation and regional 
metamorphism. Where the granite breaks 
through the mantle rocks the material is 
relatively uniform and shows intrusive 

contacts, but traced back to its roots in 
the core of the dome, the granite becomes 
more migmatic and the contacts against 
the Baltimore Gneiss are often grada- 

From the field observations and from 
similarities in mineralogy and bulk com- 
position with the Baltimore Gneiss, Hop- 
son (1964) 148 concludes that the Gun- 
powder granite was formed by anatectic 
conversion of Baltimore Gneiss. This 
interpretation seems to be supported by 
the findings of Steiger and Hopson 
(1964) , 155 who analyzed two zircon-size 
fractions from the granite (their data are 
included in Fig. 80) . In the coarse frac- 
tion (point C) they observed many short, 
rounded grains, often cloudy and pinkish, 
similar to zircons found in the Baltimore 
Gneiss. The fine fraction (point F) , 
however, was composed of about 70% 
clear, needlelike euhedral zircons. Steiger 
and Hopson therefore interpreted the 
coarser zircons as inherited from the 
Baltimore Gneiss, whereas the zircons of 
the fine fraction should have been crys- 
tallized during migmatization ^conver- 
sion) of the Baltimore Gneiss. To estab- 
lish a minimum age for the migmatization 
and intrusion, Steiger and Hopson use the 
apparent Pb 207 /Pb 206 ages. They add, 
however, that by assumption of episodic 
lead loss during Paleozoic metamorphism 
or of continuous diffusive lead loss, the 
true age of the zircons would be higher 
(up to 700 m.y.). 

The major regional deformation and 
metamorphism was terminated 425 ± 
m.y. ago. This age is given by a Rb-Sr 
isochron on minerals from various coarse 
pegmatites which occur in the surround- 
ings of the gneiss domes and which cross- 
cut the foliation of the metasediments 
(Wetherill et al, 1966). 153 It is, how- 
ever, to be noted that the mineral iso- 
chrons in some pegmatites have been 
reset during later thermal (and tectonic?) 
events: Two pegmatites yielded ages of 
about 390 m.y. and three others gave 



Pb 206 /U 238 

0.08 - 

0.06 - 

0.05 - 





^ 4.0 

500 ^/ 
<■/ □ C 600 ppm U 



125 - 150 Mm 364 ppm U 
75 -I25^m 484 ppm U 



Aj < AOfim 

810 ppm U 



F - 325 mesh 

(<44/*m) 900 ppm U 

300 > 


1 1 






Pb 20 7u 235 

Fig. 80. Concordia diagram showing the isotopic data of zircons from the Gunpowder granite. 
Data for points C and F after Steiger and Hopson (1964) 155 and Davis et al. (1965). 159 The grain 
size of the zircon fractions and the uranium concentrations are indicated on the diagram. 

ages of 345-347 m.y. (A Rb 87 = 1.39 X 

lo-^r 1 ). 

As the age of the Gunpowder granite 
and the problem of its origin are of par- 
ticular interest, I have collected rock 
samples from two different localities 
along Gunpowder Falls in order to carry 
out isotopic analyses of small, well- 
defined zircon fractions. Because episodic 
lead loss appears to be the most likely 
interpretation for the discordant zircons 
of the Baltimore Gneiss (see this Report, 
p. 288), detailed analyses were initiated 
to find out whether this interpretation is 
also applicable to the zircons of the 
Gunpowder granite. 

So far, I have separated zircons from 
two blocks of slightly foliated, homo- 
geneous-looking granite (GUN-1, GUN- 
10) . Compared with the zircon yield from 
various Baltimore Gneiss samples (0.2-0.5 
g zircon/kg of rock) , the crop from the 
Gunpowder granite was about a factor of 

10 smaller (ca. 0.03 g/kg of rock). The 
zircons are clear, mostly euhedral, and 
have high elongations. They are appar- 
ently identical to those described by 
Steiger and Hopson (1964). 155 How- 
ever, zircons that may be interpreted as 
inherited from the Baltimore Gneiss were 
found only in sample GUN-10 and 
amount to only about 3% of the total 
zircon crop. The two types of zircons are 
clearly distinguishable, and no zircons 
have been found that could be regarded 
as partly resorbed or as inherited from 
the Baltimore Gneiss with later euhedral 
overgrowth in the Gunpowder granite. 

The gross differences in zircon yield be- 
tween the Baltimore Gneiss and the Gun- 
powder granite and the occurrence of 
only few and distinctly different zircons 
of Baltimore Gneiss type do not support 
Hopson's (1964) 148 conclusion that the 
Gunpowder granite was formed in situ by 
anatectic conversion of Baltimore Gneiss, 



at least not at the present-day exposed 
level. I would rather interpret the rock 
as a homogeneous, although palingenic, 
intrusive which locally migmatized and 
assimilated parts of the country rock. 

So far, three sizes of fractions of zircon 
sample GUN-1 have been analyzed, and 
the results are plotted on the concordia 
diagram in Fig. 80. The points fall well 
on a straight line that intersects the con- 
cordia curve at about 330 m.y. Although 
the zircon fractions contain no grains 
that could be clearly related to the Balti- 
more Gneiss, they obviously contain a 
component of older radiogenic lead. Such 
observations are quite common for mag- 
matic zircons of presumably palingenic 
intrusives. This problem is the major 
subject of another part of this year's Re- 
port (see Grauert and Hofmann) . 

The new data, particularly the point 
of the smallest size fraction, do not agree 
with the results of Steiger and Hopson 
(1964) , 155 If Steiger and Hopson's 
method of interpretation were applied to 
the new data, the minimum age of crys- 
tallization for the euhedral zircons would 
be considerably lower, about 430 m.y. 
instead of 500 m.y. It is, however, strik- 
ing that the age of 330 m.y., given by 
the lower intersection of the best-fit line 
with the concordia, nearly coincides with 
the ages of the Rb-Sr mineral isochrons 
(biotite, muscovite, and feldspars) of 
three pegmatites determined by Wetherill 
et al. (1966) , 153 It is therefore possible 
that a distinct thermal (and tectonic?) 
event took place 320-350 m.y. ago. In 
that case, the data pattern in Fig. 80 
would be the result of mixing of old and 
newly formed zircon components com- 
bined with later episodic disturbances of 
the U-Pb systems at about 300 m.y. This 
model, however, gives no satisfactory ex- 
planation for the highly discordant ages 
of Steiger and Hopson's point F, which 
suggests instead a continuous diffusive 
lead loss or episodic disturbance some 
time after 300 m.y. 

U-Pb Isotopic Analyses of Zircons from 

Granulite and Amphibolite Fades Rocks 

of the West Chester Prong and the 

Avondale Anticline, Southeastern 


B. Grauert, M. L. Crawford* and 
M.E. Wagner] 

The Precambrian basement of south- 
eastern Pennsylvania, collectively termed 
Baltimore Gneiss, outcrops along an axis 
of intense Paleozoic deformation and 
metamorphism. As in the Piedmont of 
the Baltimore-Washington area, the 
basement has been uplifted and thrust 
over metamorphosed sediments of the 
Glenarm Series (a detailed discussion of 
the stratigraphic age of the Glenarm 
Series is given in Higgins, 1972). 149 The 
parts of the basement known as the West 
Chester Prong and the Avondale Anti- 
cline, where the zircons were collected, 
are largely bounded by faults and mylo- 
nite zones which separate the Baltimore 
Gneiss from mica schists and serpenti- 
nites in the north and from serpentinites, 
schists, and gneisses, reaching the silli- 
manite-orthoclase subfacies of the am- 
phibolite facies in the south. 

The Baltimore Gneiss in the middle 
part of the West Chester Prong has been 
recently studied by Wagner (1972) 160 
and Wagner and Crawford (1973) . 161 
From the mineral parageneses observed, 
they distinguish three areas of different 
metamorphic facies (see Fig. 81). (1) 
In the western granulite facies area, 
the PT conditions were the highest and 
reached approximately 10-11 kb and 
900-950 °C. The main lithologic types 
are pyroxene granulites (presumably the 
metamorphic derivatives of olivine gab- 
bros), quartzofeldspathic granulites, 
and quartzose rocks. (2) In the eastern 
granulite facies area, th