9AAGL48ZMF3VN0GPWFIN5J49M

Particle Accelerators: A Brief History
M. Stanley Livingston
Associate Director, National Accelerator Laboratory
H a r v a r d University Press : C a m b r i d g e , Massachusetts : 1969

Copyright © 1969 by the President and Fellows of Harvard College
All rights reserved
Distributed in Great Britain by Oxford University Press, London
Library of Congress Catalog Card N u m b e r 69-1K03H
Printed in the United States of America
Preface
Particle accelerators are among the most useful tools for re-
search in nuclear physics and in high-energy particle physics.
T h e rapid growth of these research fields has been due, in
large measure, to the development of a sequence of electronu-
clear machines for acceleration of ions and electrons. T h e high-
intensity and well-controlled beams from these machines can
be used to disintegrate nuclei, produce new unstable isotopes,
and investigate the properties of the nuclear force. Modern
high-energy accelerators can produce excited states of the ele-
mentary particles of matter, forming new unstable particles
with mass values much higher than those of the stable parti-
cles. Fundamental questions can be asked of nature, and an-
swered by experiments with these very high-energy particles.
T h e field of high-energy particle physics is on the threshold of
a significant breakthrough in our understanding of the parti-
cles of nature and the origins of the nuclear force.
Energies achieved with accelerators have increased at an al-
most exponential rate during the past 35 years. T h e field has
been characterized by a sequence of new concepts or inven-
tions, each leading to a new machine capable of still higher
energy, and each stimulating the development and construc-
tion of a new generation of accelerators. At times, the new de-
velopments came so fast that it was difficult to determine
which laboratory or machine held the current energy record. A
chronology of the major new steps in the development is in-
cluded as an appendix, which lists the new concepts, first op-
erations of new types of accelerators, and new energy records.
Also included in the Appendix is a graph of the growth of en-
ergies achieved by accelerators during the years 1930 to 1968,
with extrapolations into the future.
iii
iv] Preface
This monograph is largely derived from manuscripts of

talks and lectures delivered at Harvard University. Four chap-
ters were presented as Morris Loeb Lectures at the Harvard
Physics Department from April 8 to 18, 1968. One chapter is a
talk given at the Physics Department Colloquium on Decem-
ber 11, 1967. And the final chapter is based on a National Ac-
celerator Laboratory Report, NAL-12-0100, dated June 18,
1968.
These chapters represent facets of accelerator history as ex-
perienced and observed by the author. T h e monograph is not
intended as a complete history of all accelerator development,
and several significant types are only briefly mentioned, al-
though they have played major roles. Rather, this is a limited
compilation of those phases of accelerator development with
which I have had close associations. It is to be hoped that ex-
perts in other accelerator categories will also tell their stories
of the origins and development of these machines.
For a more detailed description of the technical develop-
ment of accelerators the reader is referred to M. S. Livingston
and J. P. Blewett, Particle Accelerators (McGraw-Hill, New
York, 1962). A more complete report on the history of acceler-
ator development can be found in M. S. Livingston, Ed., The
Development of High-energy Accelerators (Classics of Science
Series, G. Holton, Gen. Ed.; Dover, New York, 1966).
M. Stanley Livingston
Oak Brook, Illinois
J u n e 1, 1968
Contents
1 T h e Race for High Voltage 1

2 Ernest Lawrence and the Cyclotron—An Anecdotal
Account 22
3 Synchronous Accelerators and How They Grew 39
4 T h e Story of Alternating-Gradient Focusing 60
5 Origins of the Cambridge Electron Accelerator 76
6 T h e 200-GeV Accelerator 88
Appendix: Chronology of Development of Particle
Accelerators 109
References 113
Index 119
[1]
The Race for High Voltage
W h e n J. D. Cockcroft and E. T . S. W a l t o n first disintegrated

l i t h i u m nuclei with accelerated protons of 500 kilovolts en-
ergy, at the Cavendish Laboratory of Cambridge University in
1932, they opened a new era in science. T h i s event may well be
considered the origin of m o d e r n nuclear physics. It can also be
taken as the starting point in accelerator history.
A decade earlier in the same Laboratory, Ernest R u t h e r f o r d
h a d disintegrated nitrogen nuclei for the first time, using the
naturally occurring alpha particles from radioactive elements
which had energies of about 5 million electron volts (MeV).
D u r i n g this decade much had been learned to confirm the nu-
clear character of atoms and to determine the magnitudes of
nuclear b i n d i n g energies. In a speech before the Royal Society
in 1927 R u t h e r f o r d 1 expressed his h o p e that charged particles
would some day be accelerated to energies exceeding those of
radioactive radiations, capable of disintegrating nuclei. Scien-
tists in several other laboratories held similar views of the need
for high-speed particles a n d of their value as probes for study-
ing atomic nuclei. Despite the fact that the goal of several mil-
lion volts seemed well out of reach of known techniques for
p r o d u c i n g high voltages, by 1929 work h a d started in several
laboratories in several countries to develop the electrical ma-
chines needed. T h i s competition soon took on the aspects of a
race for high voltage, and for the devices needed to accelerate
particles to energies sufficient to disintegrate nuclei. T h e ini-
tial goal was one million volts (1MV).
HIGH-VOLTAGE MACHINES
D u r i n g the next few years many of the known techniques for
p r o d u c i n g high voltage were studied to see if they could be ex-
1
2] Particle Accelerators: A Brief History
tended to even higher potentials. "Surge generators" h a d been

developed for testing electrical e q u i p m e n t at very high volt-
ages. T h e s e consisted of stacks of capacitors which were
charged in parallel f r o m a dc potential supply and then dis-
charged t h r o u g h cross-connected spark gaps to develop surges
of high voltage of a few microseconds d u r a t i o n . T h e highest-
voltage surge generator was built in the General Electric Co. 2
plant at Pittsfield, Mass., in a b o u t 1932; it was capable of pro-
ducing voltage surges of over 6 MV, and was used for break-
down tests of electrical insulators and other e q u i p m e n t .
In 1930 A. Brasch and F. Lange 3 in Germany applied 2.4-
MV voltage pulses from such a surge generator to a crude vac-
u u m c h a m b e r made of alternate rings of metal and fiber
tightly compressed between end plates. T h e peak surge cur-
rent was of the order of 1000 amperes; the discharge tube prac-
tically exploded on each surge a n d h a d to be cleaned and reas-
sembled frequently. A metal-foil window at the grounded e n d
allowed a beam of high-energy electrons and gas ions to
emerge into the air, where it produced an intense blue glow
e x t e n d i n g outward as far as 1 meter. Presumably this dis-
charge was attended by some nuclear disintegrations, b u t they
were not identified.
An even more extreme approach was an attempt to utilize
the high potentials developed in the atmosphere d u r i n g elec-
trical storms. In 1932 C. U r b a n a n d others stretched an in-
sulated cable across a valley between two peaks in the Alps.
From this cable a spherical terminal was suspended. D u r i n g
thunderstorms high potentials would develop between this
terminal a n d the valley floor; sparks several h u n d r e d feet long
were observed. Plans had been made to install a discharge tube
for the acceleration of particles b u t were a b a n d o n e d when Dr.
U r b a n was killed by lightning.
A n o t h e r engineering test installation was developed by
R. W. S o r e n s e n 4 at the California Institute of Technology in
the early 1920's, for the Southern California Edison Company,
3] The Race for High Voltage
which was using 220-kilovolt lines at 50-cycle frequency for

long-distance transmission of electric power. In this technique
three 250-kV transformers were arranged in series and
m o u n t e d on insulating platforms; an "exciter" w i n d i n g at the
high-potential end of the secondary of one transformer was
used to supply the primary winding of the next transformer.
W i t h this arrangement the rms potential of the high-voltage
terminal of the third transformer was 750 kV above ground
a n d the peak voltage exceeded 1 MV. T h e installation was
used for years for the study of high-voltage breakdown of elec-
trical devices. In about 1928 this system was taken over by C. C.
Lauritsen 5 a n d his associates of the California Institute of
Technology, to be used as a voltage source for acceleration of
particles. T h e y first developed X-ray tubes operating at poten-
tials u p to 750 kV, using a single large porcelain insulator for
the vacuum chamber. By 1934 they developed a positive-ion
accelerating tube with potentials u p to 1 MV a n d started a
program of nuclear research. Eventually, the cascade trans-
former was replaced by a belt-charged electrostatic generator,
which was found to be more suitable for nuclear experiments.
T h e Tesla coil is a resonance transformer which can gener-
ate oscillatory pulses of high potential. In its usual form the
primary circuit has large capacitance and low inductance, and
is excited by the discharge of a series spark gap; the secondary
is wound of many turns with high inductance and low distrib-
uted capacitance, with the same resonant frequency as the pri-
mary. In the early 1930's a g r o u p at the D e p a r t m e n t of T e r
restial Magnetism of the Carnegie Institution of Washington,
consisting of M. A. T u v e , G. Breit, O. Dahl, and L. R. Haf
stad, 6 attempted to develop the Tesla coil as a voltage source
for positive-ion acceleration, with the secondary coil immersed
in oil for insulation. T h e y reported peak potentials of u p to 3
MV. T h e y also developed discharge tubes in which the ap-
plied potential was divided between multiple t u b u l a r elec-
trodes; this became an accepted technique in the future. How-
ever, the oscillatory character of the potential obtained from

the Tesla coil made it unsuitable for particle acceleration. T h e
Carnegie g r o u p a b a n d o n e d it in 1932 in favor of the belt-
charged electrostatic generator developed by Van de Graaff.
Several modifications of the resonance transformer were de-
veloped by others. In 1933 D. H. Sloan 7 at the University of
California built a radiofrequency resonance device operating
at 6 megacycles per second or 6 megahertz (MHz). It was used
as an electron accelerator to generate X-rays of u p to 1.25 MV;
an installation of this type built by M. S. Livingston and M.
Chaffee gave many years of service at the University of Califor-
nia Hospital in San Francisco. Another type, resonant at 60
Hz, was developed by Ε. E. Charlton 8 and associates at the
Schenectady laboratory of the General Electric Company in
1934, which was also used as an X-ray generator at potentials
u p to 1 MV.
All the devices described above have the limitation of pro-
ducing either pulsed or alternating potentials, and this is their
basic fault as ion accelerators. Several of them have been rea-
sonably successful as electron accelerators for the production
of X-rays, where the stability requirements are not so severe
and the X-ray tube acts as a rectifier; but they have all failed as
sources of positive ions.
T h e successful techniques that have survived in the compe-
tition are those which develop steady direct voltages and
which can be regulated to m a i n t a i n constant voltage with
good precision. A system that has succeeded in this respect is
the c o m m o n direct-current power supply, consisting of an
alternating-current transformer, a rectifier, and a filter circuit
to smooth out the voltage ripple. Although the transformer-
rectifier circuit had been available for many years, the voltages
to which it had been developed (less than 100 kV) were far too
low to be useful in nuclear physics. However, a variation
known as the voltage-doubling or voltage-multiplying circuit
can produce considerably higher potentials; it had been de-
veloped 9 and used in the early 1920's for application to high-

voltage X-ray tubes.
T H E VOLTAGE M U L T I P L I E R AND
T H E FIRST DISINTEGRATION
Cockcroft and W a l t o n of the Cavendish Laboratory were in-
spired by R u t h e r f o r d to search for a device of modest size and
energy which might still be sufficient to disintegrate nuclei,
a n d chose to develop the voltage-multiplier circuit to higher
voltages; they planned their system to operate at 700 kV or
higher. T h e n in 1929 G. Gamow, 1 0 a n d also E. U. C o n d o n and
R. W. Gurney, 1 1 using the new theoretical tools of wave me-
chanics, showed that protons of relatively low energy (as low as
500 keV) should have a reasonable probability of penetrating
the potential barriers of light nuclei. T h i s more modest goal
seemed feasible and justified Cockcroft and Walton's choice.
Accordingly, they initiated experimental studies of nuclear
disintegration when they h a d achieved only 500 kV. T h e y
used the lightest practicable target (metallic lithium) a n d ini-
tially employed a scintillation-counting technique, similar to
that used by R u t h e r f o r d in his early alpha-particle experi-
ments, to observe the charged fragments of disintegration ( H e
nuclei). T h e y observed the disintegration of Li by these 500-
keV protons early in 1932. T h e i r results were reported in a
series of papers in the Proceedings of the Royal Society,12
which gave full details of the technical development, voltage
calibrations, and experimental observations. T h e s e papers are
classics in nuclear physics a n d have brought e n d u r i n g fame to
the authors, who were awarded the Nobel Prize in physics for
1951.
T h e Cockcroft-Walton voltage-multiplier circuit (Fig. 1) is

an alternating-current circuit which uses rectifiers to charge
capacitors in parallel at low potential, and discharges t h e m in
series through a load resistor to develop the high potential.
T h e arrangement at the Cavendish used vertical stacks of ca-
pacitors and rectifier tubes, with rounded corona shields at the

high-voltage terminals (Fig. 2). Cockcroft and Walton ob-
tained a fourfold voltage multiplication, with a steady dc out-
put of about 500 kV. T h e ion-accelerating tube was also a ver-
tical array of tubular electrodes within a vacuum chamber
FIG. 1. Schematic circuit diagram of the voltage multiplier, showing the

direction of charge flow d u r i n g the rectification cycle. T h e o u t p u t of a
single primary transformer (lower right) provides the charge, which is
distributed between the capacitor stacks in successive cycles.
formed of insulating glass cylinders. T h e ion source at the up-

per terminal consisted of a high-voltage discharge in hydrogen
gas in a separate chamber with a small-diameter exit hole
through which the hydrogen ions emerged into the accelerat-
ing tube. After acceleration down the column, the beam of hy-
drogen ions emerged through a channel in the grounded base-
plate, where it was directed against the target (Fig. 3).
T h e voltage multiplier has been widely copied and modi-
fied in other laboratories, but seldom for potentials higher
than 1 MV. T h e Philips Company of Eindhoven, Holland,
built a well-engineered system for the Cavendish Laboratory
which operated at 1.25 MV for many years. But the primary
use in recent years has been as a preaccelerator of particle
beams for injection into higher-energy machines.
T H E ELECTROSTATIC GENERATOR
A major contender in the race for high voltage in the early
years was the electrostatic generator. R . J . Van de Graaff
started the development of the belt-charged generator at
FIG. 2. Voltage multiplier with which Cockcroft and Walton achieved

nuclear disintegration in 1932.
Princeton in 1930, after becoming interested in the need for

high-voltage machines as a Rhodes Scholar at Oxford in 1928.
In its modern pressure-insulated form it has become so success-
ful as a voltage source that it has displaced most other types of
direct-voltage generator. In the energy range of 1 to 9 MV to
which it has been developed it can deliver a steady, parallel
beam of particles with excellent energy uniformity and stabil-
ity. It is by far the most popular type of accelerator and the
number of installations around the world is conservatively es-
timated to be over 500.
Van de G r a a f f 1 3 described his first model of a belt-charged
FIG. 3. J . D . Cockcroft in the basement room below the voltage multi-

plier where nuclear disintegrations were observed.
generator before the American Physical Society in 1931. T w o

spherical aluminum terminals, 24 in. in diameter, were
mounted on 7-ft glass rods, each with a motor-driven silk belt
to transport the charge. T h e belts were charged with ions pro-
duced by corona discharge from needle points, and discharged
with similar points in the terminals, One sphere was given a
positive charge and the other a negative charge. A potential
difference estimated to be 1.5 MV, limited only by sparking or
corona discharge, was developed between terminals. T h e sim-
plicity of the principle, and the fact that a steady direct voltage
was produced, made this technique attractive; the possibility
of extending it to higher voltages was evident. Groups in sev-
eral laboratories became interested a n d joined in the develop-

ment.
In 1932 Van de Graaff went to the Massachusetts Institute of
Technology where he started designs for a really large gener-
ator. 1 4 It consisted of two 15-ft spherical terminals m o u n t e d on
vertical insulating fiber cylinders each enclosing a 4-ft-wide
belt for charging, one carrying positive and the other negative
charges. A discharge tube for accelerating ions was to be sup-
ported between the terminals, and the laboratory for observ-
ing •experiments was to be within one of the 15-ft shells. T h e
concept of the project was exciting and the scale tremendous
for that time. W h e n operated, sparks u p to 25 ft long would
occasionally j u m p between terminals or down the columns.
W h e n the terminals were charged, the static fields would liter-
ally raise the hair of a person standing at the base. I remember
vividly the experience of standing within one sphere while
both were being charged; monstrous sparks would j u m p the
gap a n d strike the rim of the porthole just a few inches away.
However, unclean conditions in the aircraft h a n g a r near the
ocean where the generator was first assembled, a n d difficulties
in m o u n t i n g the horizontal discharge tube, were extreme, so
that the generator never performed satisfactorily as an acceler-
ator. In 1937 it was moved to M I T a n d located in an enclosed
b u i l d i n g with the two terminals in contact (Fig. 4). O n e col-
u m n was used for the charging belt and the other for a vertical
accelerating tube so experiments could be performed in a
room beneath the floor. By 1940 the modified accelerator was
in operation at 2.75 MV, 1 5 a n d it gave many years of valuable
service as a research tool. It eventually became obsolete and
was moved to the Boston Museum of Science as a p e r m a n e n t
exhibit.
Meanwhile, the g r o u p headed by M. A. T u v e 16 at the Car-
negie Institution of W a s h i n g t o n adopted the electrostatic
generator as a voltage source. W i t h the advice a n d cooperation
of Van de Graaff they built first a 1-meter sphere m o u n t e d on a
FIG. 4. C. L. Van Atta, D. L. N o r t h r u p , and L. C. Van Atta beneath the
15-ft terminals of the 2.7-MeV electrostatic generator at the Massachusetts
Institute of Technology. (From Particle Accelerators by M.S. Livingston
and J. P. Blewett. Copyright 1962 by McGraw-Hill Book Company. Used
by permission of McGraw-Hill Book Company.)
tripod of insulating struts, with a vertical accelerator tube

formed of glass tubing. By 1933 (only a year after Cockcroft
and Walton had observed disintegrations) this first practical
electrostatic generator was in operation producing hydrogen
ions of 0.6-MeV energy, and was soon put to use for nuclear
experimentation. A 2-m generator was built next, with a 1-m
inner shell as a voltage divider, which was in operation at 1.3
MeV by 1935 (Fig. 5). Major contributions of the Carnegie
group were their precise voltage calibrations and studies of the
proton range-vs.-energy relation. They used magnetic deflec-
tion of the emergent beam of protons, and current measure-
ments with a column of calibrated resistors, to determine pro-
ton energy. They measured and published values of many
sharply defined nuclear resonances, 17 which became substand-
ards for cross-calibration of the results from different labora-
tories.
Meanwhile, other groups exploited the use of high gas
pressure for insulation of the terminal, by enclosing the gener-
ator in a pressure housing. With this technique much higher
voltages were ultimately obtained. A group at the University
of Wisconsin led by R. G. Herb pioneered the development
with a series of pressure-insulated generators of steadily in-
creasing energy, which supported a continuing program of re-
search in nuclear physics. T h e Herb design used concentric
terminal shields to distribute the voltage, mounting the gen-
erator from a grounded baseplate, and with a horizontal dis-
charge tube; the pressure housing rolled back on rails to open
the chamber for servicing. T h e series started with a first model
which reached 0.75 MV in 1935,18 a second operated at 2.4 MV
by 1938,19 and a third reached 4.0 MV by 1940.20
Another development was the vertically mounted pressure-
insulated generator which was started at the Massachusetts In-
stitute of Technology by J. G. T r u m p 21 of the Electrical En-
gineering Department, Van de Graaff, and their associates.
This group made a comprehensive study of the engineering
problems of high dc potentials, and developed a sequence of
FIG. 5. View of the 2-m electrostatic generator and provisional discharge
tube at the Bureau of Terrestial Magnetism. Shown are O. Dahl (on lad-
der), a technical helper, L. R. Hafstad and M. A. Tuve.
generators for research use at M I T . T h i s development culmi-

nated in the present 9-MV machine, which has reached the
highest potential of any single-stage generator.
In 1947, T r u m p , Van de Graaff, and others formed the High
Voltage Engineering Corporation, for the commercial produc-
tion of "Van de Graaff" generators. T h i s company has pro-
duced a series of models of pressure-insulated machines of in-
creasing energy and reliability, and is the accepted leader in
the technical development of electrostatic generators. An ad-
vanced development of this company is the "tandem," a
double-ended, horizontal generator in which negative hydro-
gen ions are accelerated "up" to the central terminal where
they are stripped of electrons in a gas jet to become positive
ions and are then accelerated "down" to ground potential to
obtain particle energies of double the terminal voltage.
T H E CYCLOTRON
A close second in the race to achieve nuclear disintegration
was Ernest O. Lawrence, who with his students at the Univer-
sity of California built the first cyclotron. T h e significance of
this development was in the use of resonance acceleration, in
which high-speed particles were produced without the need
for high voltages. T h i s new principle avoided many of the
technical limitations encountered in the use of direct high
voltages due to breakdown of insulation, and opened the way
to a new category of accelerators capable of very much higher
particle energies.
Lawrence conceived the idea on reading a paper by R.
Wideröe, 2 2 who described an experiment in which charged
particles traversed tubular electrodes, in resonance with a ra-
diofrequency electric field applied to the electrodes, and
emerged with energies twice what the applied voltage would
have given them (see Chapter 2). Lawrence speculated on
variations of this resonance principle, including the use of a
magnetic field to deflect the particles in circular paths so they
would return to the first electrode, where they could utilize the
radiofrequency field in many successive traversais; the device
would also be more compact in physical size. He discovered
that the equations of motion predicted a constant period of
revolution, so the particles would be accelerated on each tra-
versal of the gap between electrodes if their motion was in res-
onance with the applied radiofrequency field. T h e concept of
magnetic resonance was first published in a brief note by Law-
rence a n d Ν. E. E d l e f s e n 2 3 in 1930, without experimental
confirmation.
I was a g r a d u a t e student at the University of California at
this time, a n d this p r o b l e m was suggested by Professor Law-
rence as the subject for a research investigation to demonstrate
the validity of the resonance principle. T h e study was success-
ful and resulted in a doctoral thesis 2 4 dated April 14, 1931,
which was not published. T h i s first model used small-sized
laboratory e q u i p m e n t including a m a g n e t with 4-in. poles. A
flat brass chamber (sealed with wax) which fitted between the
poles of the magnet h a d a single hollow electrode shaped like a
capital D to which the radiofrequency potential was applied.
Hydrogen ions were formed at the center by flooding the
chamber with gas at very low pressure a n d ionizing it with
electrons from a cathode. R e s o n a n t ions that reached the edge
of the c h a m b e r after m a k i n g scores of revolutions of increasing
radius were observed in a shielded collector cup; their energies
were confirmed by deflection in a transverse electric field. Cy-
clotron resonance was observed for hydrogen ions (H2 + ) over a
wide range of radiofrequencies a n d magnetic fields. T h e most
energetic particles obtained were hydrogen ions of 80,000 eV,
when the radiofrequency potential applied across the accel-
erating g a p was only 1000 V.
Lawrence moved rapidly to extend this first success to a ma-
chine capable of p r o d u c i n g nuclear disintegrations. In early
193J h e was awarded a grant by the National Research Coun-
cil (for $1000) to build a larger machine. A n d he arranged that
FIG. 6. T h e first operating cyclotron built by E. O. Lawrence and M.S.
Livingston at the University of California in 1931-32. It produced 1.2-
MeV protons.
16] Particle Accelerators: A Brief Histoiy
I obtain an a p p o i n t m e n t as instructor at the University to con-

tinue the development. D u r i n g the summer and fall of 1931,
with the guidance of Professor Lawrence, I designed a n d built
a magnet with poles 10 in. in diameter and a cyclotron cham-
ber to fit between its poles (Figs. 6 a n d 7). By the spring of 1932
this first practical cyclotron was completed and tuned u p to
accelerate protons to 1.2-MeV energy. T h e development was
described at scientific meetings and was published in a paper
by Lawrence and Livingston 2 5 later in 1932.
FIG. 7. Vacuum chamber for the 1.2-MeV cyclotron, built by E . O . Law-

rence and M. S. Livingston, now at the Kensington Museum of Science in
London.
A few m o n t h s before this goal h a d been reached, the publi-

cation by Cockcroft and W a l t o n describing their results be-
came available at Berkeley. Lawrence m o u n t e d a concentrated
effort to obtain and build instruments that could detect nu-
clear disintegrations. D. Cooksey and F. N . D . Kurie came
from Yale University for the summer to assist in this program.
Internal targets were installed inside the chamber of the cyclo-
tron, and a thin-foil window was mounted on the chamber

wall outside the target so the particle products of disintegra-
tion could emerge into a counter outside the window. Disinte-
grations in Li and other targets were soon observed; these re-
sults were reported in a brief note by Lawrence, Livingston,
and M. G. White 26 in September 1932.
Lawrence had already made plans for the next step to still
higher energies, and had obtained financial support from the
Research Corporation. This next size was the "27-inch." T h e
iron core for the electromagnet was taken from a dismantled
Poulsen-arc radio transmitter donated by the Federal Tele-
graph Company. By the time this machine was completed and
operating, in 1933, ions of the newly discovered heavy hydro-
gen isotope (deuterons) were available, obtained from heavy-
water samples provided by Professor G. N. Lewis of the Chem-
istry Department. This first major Berkeley cyclotron was
tuned up (primarily by "shimming" of the magnetic field) to
produce 3-MeV and then 5-MeV deuterons, and was described
in a publication by Lawrence and Livingston 2 7 in 1934. An
active research program was developed using these particles,
exploring the new fields of deuteron reactions, induced radio-
activity, and neutron production. A group of young and en-
thusiastic scientists joined in exploiting this opportunity,
including M. C. Henderson, Ε. M. McMillan, W. Coates,
J. J. Livingood, B. B. Kinsey, F. N. D. Kurie, R. L. Thorn-
ton, and, in the years to follow, many others.
In 1934 I left Berkeley to join the Physics Department at
Cornell University and to build another cyclotron, and a few
years later went to M I T to build still another. At Berkeley,
Lawrence and his growing team continued the development to
higher energies. Their next step was expansion of the magnet
poles to 37-in. diameter, which resulted in 8-MeV deuterons by
1936. T h e potential medical applications of high-energy neu-
trons, available in significant intensities for the first time,
justified and supported the next step, which was the 60-in.
"Crocker" cyclotron; 28 this was completed in 1939, operating

first at 16-MeV and later at 20-MeV deuterons or 40-MeV
doubly charged helium ions.
T h e cyclotron took the lead in particle energy in 1932, and
held first place for 10 years. Following 1934, cyclotrons were
built in many other laboratories, frequently designed by grad-
uates of the Berkeley school. Soon these other laboratories
were able to contribute to the development, including the first
high-intensity ion source and other features. More than a hun-
dred cyclotrons have been built in laboratories around the
world, mostly as sources for nuclear research studies. A few of
the larger installations have been supported by medical re-
search funds, and research programs were developed around
medical and biological problems such as the use of neutrons
for cancer therapy and of radioactive isotopes for tracer exper-
iments in living tissue.
T h e cyclotron has become a symbol of nuclear physics, and
its simple resonance principle is taught in most high school
and college physics courses. Small working models have been
built by students in high schools. Many of the larger machines
were patterned after the 60-in. Crocker cyclotron. T h e two
largest fixed-frequency cyclotrons were the 86-in. at the Oak
Ridge National Laboratory and the 225-cm at the Nobel Insti-
tute in Stockholm (Fig. 8). Both machines produced protons of
22-MeV and deuterons of 24-MeV energy. Practical limitations
restricted the m a x i m u m energy to values less than the theoret-
ical limit. T h i s limit to particle energy is due to the relativistic
increase in mass of the ions, which ultimately detunes the ions
from resonance and prevents further acceleration. However,
there is no longer any need to push the cyclotron to this energy
limit. Higher energies can be achieved more readily in syn-
chronous accelerators or by use of the special focusing tech-
niques of "isochronous" cyclotrons.
FIG. 8. T h e 225-cm cyclotron at the Nobel Institute for Physics, Stock-
holm.
ELECTRON ACCELERATORS
T h e machine that won the next heat in the race for high-
energy particles and took the energy record away from the cy-
clotron was the betatron, an electron accelerator. T h e initial
motivation for the development of electron accelerators has
been somewhat different from that for proton accelerators. In
FIG. 9. T h e first betatron built by D. W. Kerst at the University of Illi-

nois in 1940. It produced 2.3-MeV electrons.
early years the purpose was to produce high-energy X-rays for

medical therapy in hospitals. Such X-ray machines compete
with radioactive gamma-ray sources and their output is fre-
quently given in roentgen units. Another application has been
found in the metal-products industry for the survey of flaws in
castings and welds. Electrostatic generators have been widely
used in such applications, in the 1- to 4-MV range.
T h e first successful betatron was built by D. W. Kerst 29 in
1940 (Fig. 9); it produced 2.3-MeV electrons, and had an X-ray
output equivalent to that from 1 gram of radium. Next, Kerst
went to the General Electric Company, where with the help of
its experienced laboratory staff he built a 20-MeV betatron by
1942. Later he returned to the University of Illinois to build
first an 80-MeV "model" and ultimately (1950) a 300-MeV ma-
chine, 30 which represents the largest and probably the last
large betatron. Before it was completed the electron synchro-
tron came on the scene; with a lighter and more efficient ring
magnet, it displaced the betatron as an electron accelerator in

the high-energy range.
In the continuing race for high energy, synchronous accel-
erators were further developed following World War II. Also,
several new types of machine and a major increase in energy
were made possible with the introduction of alternating-
gradient focusing. For many years the energy record of 33 GeV
(109 eV) was held by the Alternating-Gradient Synchrotron
(AGS) at Brookhaven National Laboratory. This was sur-
passed by the 76-GeV machine at Serpukov, USSR, which
came into operation in time to be announced at the anniver-
sary of the October Revolution in 1967. Plans exist at various
levels of authorization for machines of still higher energy. T h e
plan with the highest credibility level at present is for a 200-
GeV alternating-gradient proton synchrotron in the United
States; a site at Weston, Illinois, has been selected by the
Atomic Energy Commission and the staff has been assembled
to formulate the final designs for the installation. Other plans
exist for 300-GeV and even for 1000-GeV machines. So the race
for high energy goes on!
[2]
Ernest Lawrence and the Cyclotron—
An Anecdotal Account
I first met Professor Lawrence as a student in his Electricity

a n d Magnetism course when I went to Berkeley as a graduate
student in 1929; h e was a young associate professor in his sec-
ond year at the University of California. I was greatly im-
pressed with his enthusiasm a n d his vivid personality. H e
seemed always to emphasize the i m p o r t a n t concepts a n d con-
clusions, but took a rather cavalier attitude toward factors of
4π or other details in theoretical developments.
I n the early summer of 1930 I asked Professor Lawrence to
propose a topic for an experimental thesis. He suggested a
study of the resonance of hydrogen ions with a radiofrequency
electric field in the presence of a magnetic field—the phe-
n o m e n o n now known as cyclotron resonance. H e claimed that
the ions should make h u n d r e d s of revolutions in the magnetic
field, gaining energy in each turn and attaining final energies
of 1 MeV or more in a magnet of suitable size. I found that
another student, Ν. E. Edlefsen, who h a d completed his thesis
the previous winter and was awaiting the J u n e degree date,
h a d made a preliminary attempt to observe this resonance
p h e n o m e n o n in the spring of 1930 using a small laboratory
magnet, a n d that Lawrence considered his results very promis-
ing. Lawrence described the concept at a meeting of the Amer-
ican Association for the Advancement of Science in Berkeley
that spring a n d submitted a brief article to Science.1
In discussions with Lawrence in later years I learned that he
h a d conceived the idea of a magnetic resonance accelerator in
the early summer of 1929, while browsing t h r o u g h the current
journals in the library at the University of California. He saw
the illustrations in a p a p e r by W i d e r ö e 2 in the Archiv für
22
23] Ernest Lawrence and the Cyclotron
Elektrotechnik for 1928, and recognized the resonance princi-

ple involved, although he could not read German readily.
Wideröe's paper described an experiment in which positive
ions of Ν a and Κ were accelerated to twice the applied voltage
while traversing two gaps at the ends of a tubular electrode to
FIG. 10. Diagram of the a p p a r a t u s used by R. Wideröe in 1928 to demon-

strate the d o u b l i n g of the energy of heavy ions in resonance with a radio-
frequency electric field. A radiofrequency generator (lower right) pro-
duces rf voltages across two gaps / a n d II at the ends of a t u b u l a r electrode
BR through which ions from the source Κ (at the left of BR) are acceler-
ated.
which a radiofrequency potential was applied (Fig. 10). This

was an elementary linear accelerator. Wideröe described the
process as "kinetic voltage transformation." He chose the type
of ions, the accelerating frequency, the applied potential, and
the gap spacing to achieve resonance. T h e doubled energy was
confirmed by electrostatic deflection measurements on the ions.
Lawrence had been searching for a method of accelerating
particles to higher energies than could be attained with dc po-
tentials, in order to study "nuclear excitations." He realized
that extension of Wideröe's technique to such high energies
would require a very long array of electrodes. He speculated
on variations of the resonance principle, including the use of
a magnetic field to deflect particles in circular paths so they
would return to the first electrode and reuse the electric field in
the gap. He found that the equations of motion predicted a
constant period of revolution in a uniform magnetic field, re-
gardless of particle energy, so the ions would remain in reso-
nance with an accelerating field of fixed frequency. Charged
particles could be m a d e to traverse the same set of electrodes
many times, gaining energy on each traversal of the gap be-
tween them; the orbit radius would increase as the velocity in-
creased. T h i s was the cyclotron resonance principle a n d the
resonance frequency is now called the cyclotron frequency.
I started experimental work that summer. I first reassem-
bled and recalibrated the 4-in. laboratory magnet used by
Edlefsen, built a replacement for the glass vacuum chamber,
a n d studied the broad "resonance" which Edlefsen h a d ob-
served when the magnetic field was varied. I soon f o u n d that
this effect was due not to hydrogen ions b u t probably to heavy
ions f r o m the residual gas, which were accelerated once in the
radiofrequency field a n d reached the unshielded detection
electrode at the edge of the chamber.
It was now my responsibility to demonstrate true cyclotron
resonance. T h e Physics D e p a r t m e n t glassblower built for m e a
sequence of flat glass chambers in which electrodes were
m o u n t e d on greased-joint seals. Glass was traditionally used
for vacuum systems in the laboratory, b u t this thin, flat glass
c h a m b e r defied o u r technical skills. I then built a c h a m b e r
formed of a brass r i n g a n d flat brass cover plates, using red
sealing wax for a vacuum seal, in which the several electrodes
could be m o u n t e d (Fig. 11). T h e radiofrequency electrode was
a single hollow D-shaped half-pillbox facing a slotted bar
placed across the diameter of the chamber called a " d u m m y
D." T h e rf potential was developed by a simple Hartley oscil-
lator; the need for a more efficient rf circuit came later with the
effort to increase energy. A 10-W vacuum tube was used as an
oscillator a n d provided u p to 1000 V o n the electrode, at a fre-
quency that could be varied by changing the n u m b e r of turns
in an external inductance coil formed of copper tubing. Hy-

drogen molecular ions (H2 + ) were produced t h r o u g h ioniza-
tion of hydrogen gas in the chamber by electrons emitted from
a tungsten-wire cathode near the center of the chamber. Ions
that reached the edge of the c h a m b e r were observed in a
shielded collector cup.
FIG. 11. Vacuum chamber of the first resonance accelerator, used by Liv-
ingston to demonstrate cyclotron resonance for a doctoral thesis at the
University of California in 1931. Now at the Lawrence Hall of Science in
Berkeley.
D u r i n g the fall I continued the technical development with

Lawrence's continued enthusiastic interest a n d supervision. It
was in November 1930 when I first observed sharp peaks in the
collector current as the magnetic field was varied over a nar-
row range. Several techniques were used to prove that the reso-
nance peaks were due to high-energy ions. A deflecting plate
a n d slit system placed in f r o n t of the collector gave a rough
check of ion energy. But the basic proof was that the magnetic
field at resonance was just that calculated from the resonance
equation using the measured value of applied radiofrequency.
O t h e r resonance peaks observed at lower magnetic fields were
explained as d u e to the 3 / 2 and 5 / 2 harmonics of the applied
frequency. Incidentally, these resonance values provided a

highly accurate check of the magnetic-field calibration. T h e
frequency (measured with a wavemeter) was varied over a wide
range a n d the corresponding values of magnetic field for the
resonance peaks were observed. W h e n a g r a p h of wavelength
vs. magnetic field was plotted, the points fell on a smooth
hyperbolic curve, as predicted by the resonance equation.
T h e small magnet used for the first studies h a d a m a x i m u m
field of 5200 gauss, for which resonance with H2 + ions occurred
at a wavelength of 76 m; the ion energy was calculated to be 13
keV (kilo electron volts) at the radius of the collector. T h i s
goal was reached on J a n u a r y 2, 1931, after working straight
t h r o u g h the Christmas and New Years holiday. A stronger
magnet was borrowed for a time, capable of p r o d u c i n g 13,000
gauss, for which resonance occurred at 30-m wavelength or 10-
MHz frequency, and for which the calculated ion energy was
80 keV. T h i s was o b t a i n e d with an applied peak rf potential of
a b o u t 1 kV, so the ions traversed a m i n i m u m of 40 turns (80
accelerations). T h i s result was reported 3 at a meeting of the
American Physical Society early in 1931.
An i m p o r t a n t consequence of our studies with this proto-
type was the experimental observation of electric a n d mag-
netic focusing. In Lawrence's original conception the electric
field inside the hollow rf electrodes should be zero and the field
in the gap should be parallel to the plane of the particle orbits.
Otherwise it was expected that small transverse fields would
produce spiraling orbits that would intersect the electrodes.
Accordingly, the electrodes initially h a d a grid of fine tungsten
wires tightly stretched across their apertures at the gap. Reso-
n a n t peaks were first observed as currents of 10~ 10 and 1 0 ~ n
amperes, r e q u i r i n g our most sensitive electrometers and gal-
vanometers. I knew that these wires were intercepting the cir-
culating beam and felt intuitively that they might not be
needed. So while I was on my own d u r i n g a trip Lawrence
m a d e to the East, I removed the grid wires a n d obtained
greatly increased currents, in the 10 _9 -A range. O n his return

Lawrence immediately recognized the focusing properties d u e
to the shape of the electric field between open electrode faces
(Fig. 12a), a n d we never again used grids. Similarly, when thin
shims of iron were inserted in the g a p between the chamber
a n d one pole of the magnet, beam intensity was increased for
(«> ( b )
F I G . 12. Sketches used by L a w r e n c e a n d L i v i n g s t o n in an early p u b l i c a -

t i o n t o i l l u s t r a t e focusing by t h e electric a n d m a g n e t i c fields ( b r o k e n
lines) in t h e c y c l o t r o n ; particles a r e deflected t o w a r d t h e m e d i a n p l a n e by
b o t h (a) t h e electric field b e t w e e n t h e D's a n d (b) t h e m a g n e t i c field at t h e
periphery.
certain sizes a n d locations of the shims. T h i s result caused us to

study the effect of the shape of the magnetic field (Fig. 12b).
T h e transverse focusing d u e to the concave-inward shape of
the fringing field at the periphery was recognized a n d checked
experimentally by observing the beam thickness at the edge;
by means of a probe o n a movable greased joint in f r o n t of the
collector cup this thickness was f o u n d to be q u i t e small, of the
order of 1 to 2 m m . From then o n the "shimming" of the mag-
netic field became an i m p o r t a n t a n d somewhat mysterious
technique for t u n i n g u p the accelerator.
Lawrence h a d moved promptly to exploit this break-
t h r o u g h . I n early 1931 he applied for a n d was awarded a grant
f r o m the National Research Council, for $1000, for construc-
tion of a m a c h i n e that could give energies useful for nuclear
research. At the e n d of March, Lawrence told me to start writ-
ing my thesis, so I could get my degree, be eligible for an In-
structorship in the D e p a r t m e n t which h e h a d obtained for me
for the following year, a n d could continue the development.
This was just two weeks before the thesis deadline date, but I
made it and presented my thesis 4 dated April 14, 1931. Inci-
dentally, I was a poorly prepared candidate. In following
Lawrence's enthusiastic lead I had been working nights, week-
ends, and holidays in the laboratory, with no time for reading
or studying. At my oral examination some members of the
F I G . 13. D i a g r a m of t h e v a c u u m c h a m b e r for t h e 1.2-MeV cyclotron b u i l t

by L a w r e n c e a n d Livingston at t h e University of C a l i f o r n i a in 1931. See
also Fig. 7.
committee were appalled to find that I had not studied

"Rutherford, Chadwick, and Ellis," the basic reference on
natural radioactivity, which they considered essential for a
person presuming to enter the field of nuclear physics. Again,
Lawrence's enthusiasm and personal recommendation pre-
vailed, and I received the degree in May.
During the summer and fall of 1931 I designed and installed
a 10-in. magnet and built the other components for a magnetic-
resonance accelerator capable of reaching 1-MeV energy, lo-

cated in Room 339 of LeConte Hall, the Berkeley physics
building. As before, the vacuum chamber was a flat brass box
and the cover plate was sealed with wax. It contained a single
hollow D-shaped electrode for the rf potential, a thermionic
cathode, and a shielded deflecting electrode and collector cup
located at a radius of 11.5 cm (Fig. 13). T h e rf oscillator used a
10-kW Federal Telegraph water-cooled power tube in a tuned-
plate tuned-grid circuit which produced peak potentials up to
50 kV across the accelerating gap at frequencies up to 20 MHz.
I was greatly aided in the development of this first high-power
rf oscillator by David Sloan, another graduate student who
had been a ham radio operator and was an ingenious student
of high-frequency radio techniques. A deflecting electrode was
used to draw the beam out of its circular orbit and into the
collector; it also provided a rough measurement of ion energy.
This first practical cyclotron produced H2+ ions of 0.5-MeV
energy by December 1931 and Hi + ions (protons) of 1.22 MeV,
with beam currents of about 10 - 9 A, in January 1932. T h e
progress was reported in several abstracts and a paper 5 was
sent to the Physical Review on February 20, 1932. This was the
first time in scientific history that artificially accelerated ions
of this energy had been produced. T h e original vacuum
chamber of this 1.2- MeV cyclotron is now on permanent ex-
hibit in the Kensington Museum of Science in London.
As a personal footnote to history, I recall the day when I had
adjusted the oscillator to a new high frequency and, with Law-
rence looking over my shoulder, tuned the magnet through
resonance. As the galvanometer spot swung across the scale,
indicating that protons of 1-MeV energy were reaching the col-
lector, Lawrence literally danced around the room with glee.
T h e news quickly spread through the Berkeley laboratory and
we were busy all that day demonstrating million-volt protons
to eager viewers.
We had barely confirmed our results and were busy with re-
visions to increase beam intensity when we received the issue of

the Proceedings of the Royal Society describing the results of
Cockcroft and Walton 6 at the Cavendish Laboratory in disin-
tegrating lithium with protons of only 500-keV energy. At that
time we did not have adequate instruments to observe disinte-
grations. Lawrence sent an emergency call to his friend and
former colleague at Yale, Donald Cooksey, who came out to
Berkeley for the summer with Franz Kurie. They helped de-
velop the necessary counters and instruments for disintegra-
tion measurements. With the help of Milton White, then a
graduate student at Berkeley, we installed a target mount in-
side the collector cup and a thin mica window on the side of
the chamber facing the target, outside which counters could be
located. Within a few months after hearing the news from the
Cavendish we were ready to try for ourselves. Targets of vari-
ous light elements were inserted, the counters clicked, and we
were observing disintegrations. These early Berkeley results 7
confirming Cockcroft and Walton and including several addi-
tional targets were published that fall.
Lawrence was planning his next step even before I had
completed the 10-in. machine as a working accelerator. His
aims were ambitious, but supporting funds were difficult to
obtain. He was forced to use many economies and substitutes
to reach his goals. In late 1931 he located a magnet core from
an obsolete Poulsen-arc magnet with a 45-in. core, and had it
machined to form flat pole faces initially tapered to a 27 '/^-in.
diameter. T h e magnet coils were formed of strip copper
wound in layers and immersed in oil tanks for cooling. (The
oil tanks leaked! We all wore paper hats when working be-
tween coils to keep the oil out of our hair.) This magnet was
installed in December 1931 in a frame warehouse near the
Physics building, later known as the "old radiation labora-
tory," which was the center of cyclotron activities for many
years (Fig. 14). Early in 1932 I turned the 10-in. machine over
to White to use for his thesis problem, and applied most of my
time to construction of the larger machine.
FIG. 14. Ε. O. Lawrence and M.S. Livingston beside the magnet frame of
the 271/2-in. cyclotron while u n d e r construction in 1932.
T h e vacuum chamber was a 27-in. brass ring fitted with iron

disks for top and bottom plates; the top "lid" was removable
and vacuum-sealed with soft wax. Initially a single rf D was
installed, supported by a Pyrex glass insulator, with a slotted
bar across the diameter for a " d u m m y D" (Fig. 15). T h i s al-
FIG. 15. Vacuum chamber of the 27>/2-in. cyclotron with lid removed,
showing the single D electrode and the collector cup set for 10-in. radius.
lowed us to locate the deflection electrode and collector at any

chosen radius. T h e accelerated beam was first observed at
small radius, and shimming and other adjustments were m a d e
to maximize intensity. T h e n the collector was moved to a
larger radius, and the t u n i n g and s h i m m i n g were repeated. If
we made too large a step a n d could not find the beam, we m a d e
a strategic retreat back to smaller radius. T h u s we learned, the
h a r d way, of the necessity for a radially decreasing field to
m a i n t a i n focusing, and produced it with thin disk shims of se-
lected radii placed in the shimming gaps. Eventually we

reached a practical maximum radius of 10 in. and installed
two symmetrical D's with which higher energies and intensi-
ties could be obtained. Technical improvements were added
day by day as we gained experience. T h e progress during this
period of development from 1-MeV protons to 5-MeV deuterons
was reported 8 in several publications from 1932 to 1934.
I am indebted to E. M. McMillan for a brief chronological
account of these early developments on the 27'/2-in. cyclotron;
it seems that earlier laboratory notebooks were lost:
J u n e 13, 1932 16-cm radius, 28-m wavelength, 1.24-

MeV H2+ ions
August 20, 1932 18-cm, 29-m, 1.58-MeV H 2 + ions
August 24, 1932 Sylphon bellows installed on filament
stem for adjustment
September 28, 1932 25.4-cm, 25.8-m, 2.6-MeV H 2 + ions
October 20, 1932 Radius fixed at 10 in., two D's installed
November 16, 1932 4.8-MeV H2+ ions; beam current 0.001
μΑ
December 2-5, 1932 Target chamber installed for studies of

disintegrations with Geiger counter;
start of long series of experiments
March 20, 1933 5-MeV H 2 + , 1.5-MeV He+, 2-MeV
(HD) + ions; deuterium ions accelerated
for first time
September 27, 1933 Neutrons observed from targets bom-
barded with D + ions
December 3, 1933 Automatic magnet-current control in-
stalled
February 24, 1934 3-MeVD + ions; beam current 0.1 μ A; ra-
dioactivity induced in C by D + bom-
bardment
March 16, 1934 1.6-MeV H + ions (protons); beam cur-
rent 0.8 μΑ
April-May, 1934 5.0-MeV D + ions; beam current 0.3 μΑ.
These were busy and exciting times. Other young scientists

joined Lawrence's group; some worked on accelerator devel-
opment and others on detection instruments. We joined in
teams for taking data and publishing results. In my own list of
publications I find 17 abstracts or articles on disintegration re-
sults, in addition to several technical papers on accelerator de-
velopment, during the years 1933-1934. David Sloan and
Wesley Coates developed a linear accelerator, using tubular
electrodes of increasing length in line, which produced 2.8-
MeV Hg4^ ions. Sloan and B. B. Kinsey built a linac for Li +
ions for energies up to 1.0 MeV. Sloan and J . J . Livingood
built a resonance transformer which produced electrons and X-
rays of 1-MeV energy. Malcolm Henderson came in 1933; he
developed counting equipment and magnet-control circuits,
and also spent long hours helping to repair vacuum leaks and
on other developments. Incidentally, Henderson invented the
name "cyclotron," which was first used only as laboratory
slang but was eventually picked up by newspaper reporters
and popularized. McMillan joined the group in 1934, and
made major contributions to the planning and interpretation
of research experiments; he overlapped my time at Berkeley by
several months, and is one of only two men still on the Law-
rence Radiation Laboratory staff who were there before I left,
the other being R. L. Thornton. And we all had a fond regard
for Commander Telesio Lucci, retired from the Italian Navy,
who was our self-appointed laboratory assistant and a friend to
all. As the research results became more interesting we de-
pended heavily on Robert Oppenheimer for discussions and
theoretical interpretations. But always Ernest Lawrence was
the leader and the central figure, enthusiastic over each new
result, intent on each new technical problem, in and out of the
laboratory at all hours up to midnight, convinced that we were
making history and full of confidence for the years ahead.
One of the exciting periods was our first use of deuterons in
the cyclotron. Professor G. N. Lewis of the Chemistry Depart-
ment had succeeded in concentrating heavy water with about

20 percent of deuterium from battery-acid residues; we elec-
trolyzed it to obtain gas for our ion source. Soon after we had
tuned in the first beam of deuterium ions we observed alpha
particles from a Li target with longer range and higher energy
than any observed in natural radioactivities, coming from the
d(Li 6 , α ) reaction. This led to an intensive program of re-
search on deuteron reactions. We installed a wheel of targets
on a greased joint, with targets of many light elements which
could be turned into the beam inside the collector cup. A 2-in.
re-entrant tube on the side of the chamber, with an end face
made of a thin mica foil supported by a grid, allowed us to put
our counters close to the target and to measure the range of the
charged-particle products.
Chadwick had reported the discovery of the neutron in 1932,
and we were aware of its basic properties. As soon as we had
developed linear amplifiers capable of observing single parti-
cles in thin ionization chambers, we inserted a paraffin layer in
front of the ionization chamber and were able to observe the
recoil protons from neutrons. When deuterons became avail-
able for bombardment, we observed neutrons in large intensi-
ties from essentially every target used. This first observation of
neutrons 9 was in September 1933.
We had frustrations—repairing vacuum leaks in the wax
seals of the chamber was a continuing problem. As a leak de-
tector we sprayed ether on the seals and observed a sharp in-
crease in current in the ionization gauge; the laboratory often
smelled like a hospital or chemical lab. T h e ion-source cath-
ode was another weak point and required frequent replace-
ment. And sometimes Lawrence could be very enthusiastic! I
recall working till midnight one night to replace an ion-source
filament and reseal the chamber. T h e next morning I cau-
tiously warmed up and tuned the cyclotron to a new record
beam intensity. Lawrence was so pleased and excited when he
came into the laboratory that morning that he jubilantly ran
the filament current higher and higher, exclaiming each time

at the new high beam intensity, until he pushed too high and
b u r n e d out the filament!
W e m a d e mistakes too, owing to inexperience in research
and the general feeling of urgency in the laboratory. W e h a d
observed neutrons from every target b o m b a r d e d by deuterons
(originally called "deutons"), although we could not measure
their energy with any accuracy. W e also observed protons from
every target, including o n e g r o u p having the same range from
every target. T h i s led to the now forgotten mistake in which
the n e u t r o n mass was calculated on the assumption that the
deuteron was breaking u p into a proton and a n e u t r o n in the
nuclear field. T h e n e u t r o n mass was computed from the en-
ergy of the common proton group, and came o u t lower than
the value determined by Chadwick. Later, T u v e , Hafstad, and
Dahl in Washington, D.C., using the first electrostatic gener-
ator to be completed and employed for research, showed that
these protons and neutrons came from different reactions, the
D(D,p) and D(D,n) processes, in which the target was
deuterium-gas contamination deposited in all targets by the
beam. W e were chagrined and vowed to be more careful in the
future.
We also h a d many successful and exciting moments. I recall
the day early in 1934 (February 24) when Lawrence came rac-
ing into the lab waving a copy of the Comptes Rendus a n d ex-
citedly told us of the discovery of induced radioactivity by
Curie and Joliot in Paris, using natural alpha particles on
boron and other light elements. T h e y predicted, that the same
activities could be produced by deuterons on other targets,
such as carbon. Now we h a d a deuteron beam in use, a carbon
target on the target wheel in the chamber, a n d a Geiger point
counter a n d counting circuits in service at that time. W e h a d
been m a k i n g 1-min runs on another target, measuring the
range of the long-range alpha particles, with the oscillator o n
one pole of a double-pole knife switch a n d the counter on the
second pole. W e quickly disconnected this counter switch,
turned the target wheel to carbon, adjusted the point-counter

circuits to count electrons, and then bombarded the target for
5 min. When the oscillator switch was opened, the counter was
turned on, and click-click-click—click—-click. We were observ-
ing induced radioactivity within a half hour after hearing of
the Curie-Joliot results. This result was reported by Hender-
son, Livingston, and Lawrence 10 in March 1934.
I left the laboratory in July 1934, to go to Cornell (and later
MIT) as the first missionary from the Lawrence cyclotron
group. Cooksey returned to stay permanently at Berkeley, and
joined in Lawrence's next stage of development, which was to
expand the pole faces to 37-in. diameter and to build a larger
chamber which soon reached 8-MeV deuterons; this greatly
improved version was reported 11 in 1936. Many other young
scientists joined the group; also the first professionally trained
engineers arrived, notably W. W. Brobeck and Winfield Salis-
bury, and brought high-quality engineering into the labora-
tory. T h e first biological experiments were started in 1935,
when Dr. John Lawrence, Ernest's brother, arrived. When a
mouse was bombarded with neutrons in the re-entrant counter
tube on the side of the 271/2-in. machine, it came out dead.
This created a strong impression in the laboratory, and was
the start of the radiation-control program, though it was soon
found that the mouse had really died of suffocation. Neutron
shielding was installed around the cyclotron, formed of stacks
of 5-gallon cans filled with water. Lawrence and his brother
developed the biological program rapidly, with the strong
hope that neutrons could be used for cancer therapy. Law-
rence obtained support for the 60-in. Crocker cyclotron, 12
which was completed in 1939 (Fig. 16). This machine was a
beautifully engineered and reliable instrument, and became
the prototype of scores of cyclotrons around the world. T h e
year 1939 was also notable as the year in which Ernest Law-
rence received the Nobel Prize.
And perhaps this is the place to leave my story of Ernest
Lawrence and the cyclotron. T h e prestige of the Nobel Prize
FIG. 16. Ernest Lawrence and his laboratory staff in 1938 within the
magnet frame of the 60-in. cyclotron: (first row) J. H. Lawrence, R. Ser-
ber, F. N . D . Kurie, R. T . Birge, Ε. O. Lawrence, D. Cooksey, A. H.
Snell, L. W. Alvarez, P. H. Abelson; (second row) J G. Backus, Marin, P.
Ebersold, E. E. McMillan, E. M. Lyman, M. D. Kamen, D . C . Kalbfell,
W . W . Salisbury; (third row) A. S. Langsdorf, Jr., Simmons, J . G . Hamil-
ton, D . H . Sloan, J . R . Oppenheimer, W. M. Brobeck, R . A . Cornog,
R . R. Wilson, Viez, J. J. Livingood.
gave him the opportunity to implement his dreams of even

larger cyclotrons and made him a scientific figure of interna-
tional importance. T h e Lawrence Radiation Laboratory, with
its 184-in. synchrocyclotron and the 6-GeV bevatron, is a con-
tinuing memorial to this great innovator of science.
[3]
Synchronous Accelerators and How They Grew
T h e principle of synchronous or phase-stable acceleration

which was introduced in 1945 changed the course of acceler-
ator history and opened a new field of scientific research, that
of particle physics. T h e principle of phase stability was dis-
covered independently and nearly simultaneously in two
countries. In early 1945 V. Veksler in the U S S R submitted a
p a p e r 1 presenting the principle. And later in 1945, before
Veksler's paper reached the United States, the same principle
was announced by Ε. M. McMillan 2 at the University of Cali-
fornia. Both papers presented methods by which particles in
resonance-type accelerators could be kept in resonance with
the radiofrequency fields indefinitely, and could be accelerated
to much higher energies.
T h e early accelerators used in nuclear research before
World War II are all limited in maximum energy. Electro-
static generators and Cockcroft-Walton machines are re-
stricted by physical limitations such as the breakdown of insu-
lation in high electric fields. In the cyclotron the relativistic
increase in mass of protons and deuterons destroys the validity
of the resonance principle for energies above about 25 MeV.
T h e betatron, although capable of reaching the 300-MeV
range, is limited by the rapid increase in radiation loss by the
orbiting electrons at higher energies. None of these machines
can maintain acceleration indefinitely.
In resonance acceleration, particles traverse an alternating
(rf) electric field applied across a series of gaps, and acquire an
increment of energy on each successive traversal of this field.
Synchronous accelerators use a type of phase stability possessed
by certain particle orbits, so that particles having the incorrect
phase or the wrong energy on arrival at the gap will experi-
39
enee an automatic correction toward the proper phase and en-

ergy at succeeding gaps. Particles with small phase or energy
errors will continue to be accelerated, with minor oscillations
in phase and energy around the correct values, and these oscil-
lations are damped to smaller amplitude as acceleration con-
tinues. T h e significant feature of synchronous accelerators is
that, in principle, acceleration can be continued indefinitely.
At present, the upper energy limits are set only by economic
considerations.
TABLE 1. Characteristics of synchronous particle accelerators.
Frequency
Magnetic of electric Radius of
Type field field orbit
Electron synchrotron Increasing Constant Constant
Synchrocyclotron Constant Decreasing Increasing
Proton synchrotron Increasing Increasing Constant
Linear Zero Constant Infinite
M icrotron Constant Constant Increasing
T h e stability principle has had important applications in

five types of accelerator: electron synchrotrons, synchrocyclo-
trons, proton synchrotrons, microtrons, and linear acceler-
ators. These several types differ in the combinations and geo-
metric arrangements of magnetic and electric fields; they also
differ in the shape of particle orbits, type of magnetic field and
frequency of the accelerating rf field. T h e i r special features are
summarized in Table 1.
ELECTRON SYNCHROTRON
We shall illustrate phase stability with the electron synchro-
tron, one of the simpler types of synchronous accelerator. In
the synchrotron the electrons are retained in circular orbits of
approximately constant radius in a ring-shaped transverse
magnetic field. Acceleration takes place at one or more gaps
around the orbit, across which a radiofrequency electric field is
41] Synchronous Accelerators
applied which has a frequency equal to the orbital frequency

of the electrons (or an integral harmonic of it). Orbital fre-
quency is essentially constant for electrons having a velocity
approaching that of light, that is, for energies above a few mil-
lion electron volts. As electron energy and m o m e n t u m in-
crease, the magnetic field must increase to maintain constant
orbital radius; this is accomplished by pulsing the magnet cy-
clically from low to high field intensity. T h e electrons are ac-
celerated in a succession of bursts at the cycling frequency.
In synchronous acceleration the electrons must acquire just
the right energy per turn (on the average) to maintain con-
stant orbital radius in the rising magnetic field; this can be
called the "equilibrium" energy increase per turn. Consider
an electron crossing the gap at a phase of the rf field when the
potential has just the equilibrium value. T h e r e are two phases
on the rf wave at which this is true, one when the voltage is
rising and one when it is decreasing. T h e stable phase in the
synchrotron is that for decreasing voltage. If the applied fre-
quency is equal to the orbital frequency and the electron has
the equilibrium energy, it will arrive at the gap in successive
turns at the same phase and will stay in resonance.
Now consider a particle that arrives early (at too large a
phase angle) and receives more than the equilibrium incre-
ment in energy; it will traverse a circle of larger radius than
the equilibrium orbit and take a longer time than the equilib-
rium particle. This particle will therefore arrive at the next
gap when the phase of the applied radiofrequency has shifted
to a smaller value, will receive a smaller energy increment, and
will traverse an orbit of smaller radius. T h i s process continues
until the energy of the particle is reduced below the equilib-
rium value for resonance, when the opposite response occurs.
So the particle oscillates in phase of crossing the gap, its energy
oscillates around the equilibrium value, and the individual
orbital radii oscillate around the radius of the equilibrium or-
bit (Fig. 17). A similar argument shows that particles arriving
(a)
(b)
F I G . 17. (a) D i a g r a m of t h e r a d i o f r e q u e n c y cycle of a s y n c h r o t r o n , in

w h i c h t h e p h a s e oscillates a r o u n d a n e q u i l i b r i u m p h a s e a n g l e Φ,, d u r i n g ac-
c e l e r a t i o n ; (b) a n a l o g u e i l l u s t r a t i n g t h e m o t i o n of a d i s p l a c e d p e n d u l u m
o s c i l l a t i n g a b o u t its e q u i l i b r i u m a n g l e of d i s p l a c e m e n t Φ,.
at the gap with the wrong energy will also oscillate in phase
and in energy around the equilibrium values. Furthermore,
the equations of motion show that these oscillations are
damped to smaller amplitude as particle energy increases. Ac-
celeration continues until the magnetic field reaches its maxi-
mum, which determines maximum energy. Note that the other
phase of the rf, when voltage is rising, is unstable. An electron
crossing the gap slightly off this phase will have its phase mis-
match amplified and it will be lost from resonance. So the par-
ticles become bunched tightly around the equilibrium phase
and, as the motion is damped, occupy a smaller fraction of the
phase angle and have smaller azimuthal and radial spreads.
This oscillation in phase and energy about the equilibrium
values is analogous to the hunting in phase of the armature of
a synchronous electric motor, which led McMillan to the
choice of the name "synchrotron."
A virtue of the electron synchrotron is that the energy radi-
ated by the electrons owing to their transverse acceleration in
the magnetic field is automatically compensated by phase
shifts to larger phase angles and higher values of volts-per-
turn. 3 (The radiation loss reaches 4.5 MeV per turn in the
Cambridge Electron Accelerator synchrotron, for example,
with no reduction in the phase stability.)
Seldom has a new scientific principle been exploited with
such promptness. T h e reason was that it was announced just at
the end of World War II when scientists returned to their lab-
oratories from their wartime assignments, eager to resume re-
search activities and equipped with new skills and experience.
Many new technical devices and materials were available,
which had been developed during the war in fields such as
microwave radar, electronics, and nuclear physics. Experience
in wartime crash programs was carried over to speed up accel-
erator developments. But most significant was the increased
national prestige of scientists, which brought prompt and
generous financial support from their governments.
Unlike earlier accelerators, most of which required slow de-

velopment, starting with small sizes at low energies, the syn-
chronous machines were conceived in their full stature as high-
energy devices capable of producing mesons. T h e first electron
synchrotron to be started, by McMillan at the University of
California, was designed for 320 MeV; it came into operation
in J a n u a r y 1949. Meanwhile, in 1946, before the California
machine could be completed, Goward and B a r n e s 4 in En-
gland demonstrated the validity of the synchronous principle
with a much smaller magnet using an old betatron, in which
the magnet excitation was made to over-run the betatron
phase a n d acceleration was continued to higher energy (8
MeV), a radiofrequency circuit being used to provide the ac-
celerating field. T h e next to succeed, in 1947, was a group at
the General Electric Co. Research Laboratory 5 with a 70-MeV
electron synchrotron, based on their previous experience with
betatrons. Synchrotron radiation was first observed with this
machine, as a forward-directed beam of white light.
T h e first wave of construction in this country was in the 300-
to 350-MeV energy range. In addition to McMillan's machine
at the University of California, synchrotrons were started in
the years before 1949 at Massachusetts Institute of Technology
(Fig. 18), the University of Michigan, P u r d u e University, a n d
Cornell University; most of these were supported by the
United States Office of Naval Research. At the General Electric
Research Laboratory a 300-MeV synchrotron was built using a
magnetic field produced by electrical coils w i t h o u t an iron
core. Also, d u r i n g this phase, a 140-MeV synchrotron was built
at Oxford, England. Between 1946 and 1949 several theoreti-
cal studies of orbit stability and phase stability were pub-
lished, 6 and the dynamics of particle motion were well under-
stood. T h i s period was climaxed by the completion in 1954 of
a 350-MeV synchrotron at the University of Glasgow, which
incorporated all the best features of the earlier machines.
More recently the urge for still higher energies started an-
FIG. 18. Ring-shaped electron synchrotron magnet built at the Massachu-
setts Institute of Technology in 1948. (From Particle Accelerators by M.S.
Livingston and J. P. Blewett. Copyright 1962 by McGraw-Hill Book
Company. Used by permission of McGraw-Hill Book Company.)
other wave of construction in the 1000-MeV (1-GeV) energy

range. At California Institute of Technology the magnet for
the quarter-scale model of the bevatron (see below) was rebuilt
by 1958 and adapted for acceleration of electrons to 1.1 GeV.
At Frascati, Italy, a 1.2-GeV machine was completed in 1959.
In other laboratories the principle of alternating-gradient
focusing was adopted, and machines were built at Cornell for
1.2 GeV, at Bonn University in Germany for 0.5 GeV, at the
Royal Institute in Stockholm for 1.2 GeV, and at the University
of Tokyo for l.S GeV. Still higher energies were to come, start-
ing with the Cambridge Electron Accelerator, which more
fully exploited the AG focusing principle; these will be de-
scribed in the following chapter.
SYNCHROCYCLOTRON
Synchrocyclotron development started even before that of
the electron synchrotron, in late 1945. In this case the problem
was the relativistic limitation of fixed-frequency cyclotrons,
due to the increasing mass and decreasing orbital frequency of
the ions. T h e 184-in. magnet at the University of California
had originally been conceived as a giant standard cyclotron by
Professor Lawrence and his coworkers. A paper by Bethe and
Rose 7 in 1938 had disclosed the relativistic limitation, but
Lawrence and his group were confident that with sufficient rf
voltage they could push the energy up to 100 MeV. T h e mag-
net was assembled and used for other experimental purposes
during the war but was not completed as a cyclotron. At the
end of the war, when McMillan proposed the use of frequency
modulation, it became obvious that this method would result
in higher energies, and plans were made to convert the 184-in.
magnet into a synchrocyclotron.
McMillan (and Veksler) proposed that the applied fre-
quency be modulated (reduced in frequency) cyclically, to
match the decreasing orbital frequency of the ions in the uni-
form magnetic field as mass increased. T h e equilibrium phase
would be shifted ahead during acceleration just enough to

provide the necessary accelerating voltage to maintain an in-
creasing equilibrium energy. In this situation a stability argu-
ment similar to that for the synchrotron applies. An ion arriv-
ing at the gap too early receives too large an energy increment,
traverses a larger orbit and takes a longer time, thus arriving at
the next gap with a phase closer to the equilibrium phase.
Again, this leads to stable oscillations about the equilibrium
phase and energy, and to radial oscillations about the radius of
the equilibrium orbit. Particles are bunched about the equi-
librium phase, which means in their azimuthal location in the
orbit. Each sausage-shaped bunch moves radially outward in
the magnetic field as the applied frequency is decreased, arriv-
ing at the periphery with maximum energy at the end of the
modulation cycle. T h e rf modulation is cyclic with a frequency
of 100 Hz or more, so high-energy particles arrive at the target
in pulses at this frequency.
Another technique for maintaining resonance at constant
frequency in the cyclotron would be to shape the magnet poles
so as to increase the field in the outer regions to compensate for
the increase in mass of the particle. However, it was known
that a radially decreasing field is required for focusing in the
transverse plane. If the field is azimuthally uniform and in-
creases radially, the beam will blow up axially. In 1938
Thomas 8 proposed the use of radial sectors with alternately
strong and weak magnetic field, but with the average field
having the desired increase with radius. He showed that addi-
tional axial restoring forces exist at the edges of the magnet
sectors which compensate for the defocusing of the average
field. At this time the accelerator art was in its infancy, the the-
ory of particle orbits was not well understood by accelerator
builders, and many practical problems of engineering devel-
opment took precedence. As a result, the idea lay fallow for
many years. It was reactivated during the war at the University
of California Radiation Laboratory, under the stimulus of po-
tential military applications of very high beam intensity, but

the work was classified secret and not published until the field
was declassified in 1956. By this time it was evident that
Thomas's proposal was a special case of alternating-gradient
focusing as applied to constant magnetic fields, and it has had
its fruition in the "sector-focused" or "isochronous" cyclotrons
of the present day. This is discussed in more detail in Chapter
4.
So we return to the story of the synchronous accelerators.
T h e first test of the stability principle at Berkeley was made on
the older 37-in. cyclotron magnet in late 1945, by an ingenious
method of simulating the expected relativistic mass change in
the 184-in. machine with an exaggerated radial decrease in the
magnetic field of the 37-in.9 T h e orbital frequency decreases at
high energy owing both to the relativistic mass increase and to
the slight radial decrease needed for axial focusing. It can be
made to decrease further in a small cyclotron if the magnetic
field at large radii is reduced still further below the central
field. For this test the 37-in. pole faces were tapered radially to
produce a 13-percent decrease in field and so in orbital fre-
quency; thus the techniques of frequency modulation of the
radiofrequency could be studied and the principle of phase-
stable synchronous acceleration could be tested. A rotating ca-
pacitor in the rf circuit was used to modulate the frequency.
T h e results were completely successful. Very low D voltages
were required as compared with conventional cyclotron opera-
tion, and the deuterons remained in resonance for many thou-
sands of revolutions.
This success justified conversion of the 184-in. machine to
utilize the principle. A crash program of development and
construction resulted in initial operation in November 1946
(Fig. 19). It was successful in producing good intensities of
deuterons of 190 MeV and He + + ions of 380 MeV. T h e eight
authors listed in the first r e p o r t 1 0 were leading members of the
Berkeley laboratory staff; many others were also involved.
F I G . 19. T h e 184-in. s y n c h r o c y c l o t r o n at t h e University of C a l i f o r n i a soon
a f t e r c o m p l e t i o n in 1946 a n d b e f o r e s h i e l d i n g was a d d e d . ( F r o m Particle
Accelerators by M . S . L i v i n g s t o n a n d J . P. Blewett. C o p y r i g h t 1962 by
M c G r a w - H i l l Book C o m p a n y . Used by p e r m i s s i o n of M c G r a w - H i l l Book
Company.)
T h i s was an example of the "big-team" accelerator develop-

ments that became standard in the years to follow. T h e im-
pressive array of scientific and engineering talent was one rea-
son that such a tremendous j o b could be completed in less than
a year. But another reason was the basic soundness of the prin-
ciple of phase stability and the relative simplicity of the tech-
niques required. Equivalent success a n d speed in t u n i n g u p a
synchrocyclotron have been achieved subsequently in other

laboratories.
T h e immediate success at the Berkeley laboratory led others
to build similar accelerators. In the United States, most of
them were supported by the Office of Naval Research. Within
a few years five synchrocyclotrons were completed and in oper-
ation, at Rochester, Harvard, Columbia, Chicago, and Carne-
gie Institute of Technology. Other installations were built
abroad, at Amsterdam, Harwell, Montreal (McGill), Uppsala,
Liverpool, Geneva, and in the USSR. T h e 184-in. machine was
rebuilt in 1957, achieving an energy of 720 MeV, which is the
record energy for cyclotrons; other large synchrocyclotrons are
at CERN, Geneva (600 MeV), Dubna, USSR (680 MeV), and
the NASA laboratory near Norfolk, Virginia (600 MeV).
PROTON SYNCHROTRON
T h e proton synchrotron is the culmination of phase-stable
accelerators and has produced the highest energies. T h e
synchrocyclotron requires a solid-core magnet; at relativistic
energies the weight and cost of the magnet increase roughly as
the square or cube of a magnet dimension such as pole-
face diameter; power cost also increases about as the square of
the diameter. For energies in the GeV range the weight and
cost of the magnet would become exorbitant. T h e obvious
method of reducing magnet cost is to use a ring magnet cover-
ing only a narrow annular band, and to use a pulsed magnetic
field. In both McMillan's and Veksler's papers the possibility
of acceleration of protons in a ring-shaped magnet was im-
plicit, although neither paper was primarily concerned with
this more complicated application of the principle.
Actually, the first proposal of a proton accelerator using a
ring magnet was made in 1943 by Prof. M. L. Oliphant of the
University of Birmingham, to the British Directorate of
Atomic Energy. Because of wartime activities and security re-
strictions the proposal was not published or acted upon at that
time. T h i s antedated the discovery of phase stability by two

years. But at that time there was no theoretical assurance that
the proposal was sound or technically practical. In 1947, fol-
lowing the McMillan and Veksler papers, a theoretical analy-
sis of orbit stability based on their work and a more detailed
design study were published 11 by the Birmingham group. An
accelerator following these designs was built and was in opera-
tion at 1-GeV energy from 1953 (a year later than the cosmo-
tron) to 1967 when it was closed down. T h e Brimingham ma-
chine lacked some of the advantageous features developed at
Brookhaven and Berkeley and operated at rather low beam
intensity.
A fixed orbital radius requires synchronous acceleration.
But unlike electrons, which approach the velocity of light at
relatively low energies and so have essentially constant orbital
frequency with increasing energy, protons do not approach
this velocity until they reach energies of many giga electron
volts. So the velocity and orbital frequency of protons increase
d u r i n g the entire acceleration interval. T h i s requires modula-
tion of the frequency of the applied rf field, which must syn-
chronizewith theincreasingorbital frequency. T h i s modulation
must extend over a wide range of frequency, from a low value
at injection of the protons to a high value at m a x i m u m energy.
T h e rate of change of frequency must also match the rate of
rise of the magnetic field, which depends on the characteristics
of the power supply and the properties of the magnet iron.
T h i s introduces new and complicated technical problems in
the design of the rf accelerating units and of the high-
frequency oscillator, which are unique to the proton synchro-
tron.
T h i s extension of the principle of synchronous acceleration
was obvious to many designers, once the principle had been
announced and success had been achieved with the synchro-
tron and the synchrocyclotron. Design studies were started
early in 1947 in two laboratories supported by the United
FIG. 20. T h e 3-GeV "cosmotron" at Brookhaven National Laboratory,
before shielding was added.
States Atomic Energy Commission, at the University of Cali-

fornia and at Brookhaven. At Berkeley, W. M. Brobeck 12
made a preliminary study of a possible design for 10-GeV pro-
tons, which was primarily a comparison of alternative types of
power supply for such a large ring magnet. Designs for a simi-
lar accelerator were begun at Brookhaven 13 u n d e r my direc-
tion, stimulated by Professor I. I. Rabi of Columbia Univer-
sity; the staff members involved here J. P. Blewett, G. K.
Green, L. J . Haworth, and myself (on leave from M I T ) .
Both of these design studies were supported and encouraged
by the Research Division of the AEC. W h e n preliminary de-
signs and cost estimates became available in 1948, a decision
was made by representatives of the AEC and of the two labora-

tories for the construction of two machines, a 2.5- to 3.0-GeV
"cosmotron" at Brookhaven and a 5- to 6-GeV "bevatron" at
Berkeley. In both laboratories teams of scientists and engi-
neers were rapidly assembled to complete designs and proceed
with construction. T h e Brookhaven cosmotron 14 was the first
to be completed, in 1952, at 2.3-GeV energy; it was brought to
its maximum design energy of 3.0 GeV in 1954 (Fig. 20). T h e
Berkeley designers chose to build first a quarter-scale model, 15
and then continued with the full-scale machine, which was
brought into operation at 5 GeV in 1954 and reached 6.2 GeV
in 1955. These two machines have contributed greatly to the
field of particle physics.
Three other proton synchrotrons have been based on cosmo-
tron design and experience: the 3-GeV "Saturne" at the Saclay
laboratory of the French C.E.A., a rapid- cycling "PPA" ma-
chine for 3 GeV at Princeton, and a 7-GeV "Nimrod" machine
at the Rutherford Laboratory in England. At Dubna, USSR, a
10-GeV "synchrophasotron" based on the bevatron design and
experience was completed in 1957. And at the Argonne Na-
tional Laboratory a 12.5-GeV machine called the zero-gradient
synchrotron (ZGS) was finished in 1962. In the ZGS the bend-
ing magnets have uniform fields, and axial focusing is ob-
tained by shaping the end faces of the eight magnetic sectors.
LINEAR A C C E L E R A T O R S
T h e earliest proposal for an electron linear accelerator was
made by G. Ising 16 in Sweden in 1925. He suggested resonance
acceleration of electrons down a linear array of tubular elec-
trodes of increasing length, using a spark-gap oscillator and
transmission lines to supply the radiofrequency fields of the
proper phase to the electrodes. But Ising did not build a work-
ing model. This speculation prompted Wideröe to perform
the first experiment in which radiofrequency resonance was
observed, in an elementary linear accelerator, as described in
previous chapters. And this idea of Wideröe's, in turn, led

Ernest Lawrence to develop his concept of the magnetic reso-
nance accelerator, or cyclotron.
During the early 1930's several resonance linear accelerators
were built and tested. J. W. Beams 17 at the University of Vir-
ginia, in 1934, experimented with a traveling-wave accelerator
for electrons using transmission lines of different lengths at-

tached to a linear array of tubular electrodes and fed with po-
tential surges generated by a capacitor-spark-gap circuit,
similar to the system proposed by Ising. Bursts of electrons
were occasionally accelerated to 1.3 MeV, but the device was
not sufficiently steady or reliable for nuclear experiments.
While I was working under Lawrence at the University of
California developing the first cyclotron, another graduate
student, David Sloan, was building a linear accelerator (linac)
with ten or more tubular electrodes in line connected alter-
nately to a source of radiofrequency potential (Fig. 21). Ade-
quate radio power tubes for high frequencies were not avail-
able at that time, so rather low frequencies were used (30 MHz)
for which resonance could be obtained only with heavy ions.
By 1931 Sloan and Lawrence 18 had accelerated mercury ions

to 1.25 MeV. Working with Sloan, W. M. Coates 19 built a
linac for Hg1" ions of about 2.8-MeV energy and B. B. Kinsey
produced 1.0-MeV Li + ions. These heavy-ion linacs were not
capable of producing significant nuclear disintegrations, and
the effort was not continued.
5
FIG. 22. Basic structure of positive-ion linear accelerator. As the drift-
tube length increases, the outer diameter decreases to maintain resonance.
(From Particle Accelerators by M.S. Livingston and J· P. Blewett. Copy-
right 1962 by McGraw-Hill Book Company. Used by permission of
McGraw-Hill Book Company.)
Proton Linacs. T h e development of the proton linac had to

wait for 12 years, until the improved technology in radio and
radar developed during World War II made it possible. It was
initiated by Luis W. Alvarez at Berkeley in 1946. While Law-
rence and his associates were building the 184-in. synchro-
cyclotron and McMillan was engaged on his electron synchro-
tron, Alvarez, with the able assistance of W. K. H. Panofsky,
built the first proton linac to attain an energy sufficient for nu-
clear experiments (32 MeV). T h e availability of surplus radar
equipment, including rf power tubes for 200 MHz, speeded
the Berkeley development. T h e basic structure is a long, cylin-
drical cavity, with drift tubes of increasing length mounted
along its axis (Fig. 22). It is supplied by protons of 4-MeV en-
ergy from a Van de Graaff generator. T h e cavity operates in a

standing-wave mode in which axial rf fields between the drift-
tube electrodes provides acceleration. In this structure an ion
with the correct phase sees only accelerating fields when it
crosses the gaps; while the fields are in the reverse direction,
the ions are shielded by the tubes. T h e drift tubes are basically
cylinders, with rounded ends, increasing in length and de-
creasing in outer diameter, supported on slender stems along
the axis of the cavity. T h e precise shapes and spacings are de-
termined by calculations and model studies so that each "unit
cell" has the same frequency. Power is fed into the cavity at
several points from power tubes which are excited in parallel.
T h e Alvarez design has been the basis for essentially all sub-
sequent proton linacs, and the same rf frequency of about 200
MHz has been retained. Linacs of this type are used as injec-
tors for higher-energy machines; 50-MeV linacs provide pro-
tons for the ZGS at Argonne (Fig. 23), the AGS at Brookhaven,
and the CPS at CERN. T h e highest-energy machine at present
is the 100-MeV linac used as an injector for the proton syn-
chrotron at Serpukhov, USSR. Construction is now in process in
several laboratories for linacs to produce 200-MeV protons,
and at Los Alamos construction has started on a "meson-
factory" linac for 800 MeV.
These proton linacs utilize a simple type of phase stability.
No magnetic fields are used and particles are accelerated in a
linear path by an rf electric field of constant frequency. T h e
separation between accelerating gaps increases along the path
in such a way that the time required for a particle to traverse
each space is the same, usually a full period of the rf field. T h e
equilibrium phase is on the rising side of the voltage wave ap-
plied across the gaps. If a particle arrives too soon it will re-
ceive less than the equilibrium increment in energy, will have
too small a velocity, and will arrive at the next gap at a later
phase which is closer to the equilibrium phase. T h e reverse is
also true. So the phase of crossing the gaps oscillates around
FIG. 23. T h e 50-MeV proton linac used as an injector for the
ZGS machine at the Argonne National Laboratory.
the equilibrium phase, and particle energy oscillates around

the steadily rising equilibrium energy.
Electron Linacs. Modern linear electron accelerators are
based on the work of W. W. Hansen and his associates at Stan-
ford, starting in the 1930's. One of the early concepts was a
high-Q cavity resonator for single-stage acceleration of elec-
trons. T o excite this cavity, rf sources of unusually high power
and high frequency were needed. This requirement led to the
invention and initial development of the klystron by Hansen
(in collaboration with the Varian brothers), which has since
played an important role in the rf power field. At the end of
the war, Hansen, E. L. Ginzton, and others started a double-
headed development program, to develop waveguides for ac-
celeration of electrons at microwave frequencies, and to pro-
vide klystrons of adequate power at these frequencies. 20 T h i s
program, ably continued after Hansen's death in 1949 by
Ginzton and later by Panofsky, has resulted in a sequence of
electron linear accelerators of increasing length, power, and
output energy.
T h e accelerating structures for electron linacs are quite dif-
ferent from those for protons. On account of the higher elec-
tron velocities, beam apertures can be smaller and resonant
frequencies higher than for protons. Largely owing to the
availability of radar components in the 10-cm range, the fre-
quency of 2855 MHz has been chosen for most electron linacs.
T h e structure used is an iris-loaded waveguide, of about 4-in.
diameter, in which a traveling wave of the TMoi mode pro-
vides an axial electric field with a wave velocity equal to that of
light. T h e iris diaphragms have 2-cm apertures for the beam
and are spaced along the cylindrical waveguide at intervals of
about 3 cm to provide the correct wave velocity.
Electrons approach the velocity of light at an energy of a few
million electron volts, so the mechanism of phase focusing is
applicable only during the early part of acceleration. How-
ever, although the phase is no longer oscillating around the
equilibrium value at higher energies, it now approaches it

asymptotically; the electron continues to be "locked" to the ac-
celerating wave. T h e electrons can be considered to ride on the
advancing electric wavefront much as a surfboard rides on the
front of a water wave.
A large fraction of the development of the electron linac has
taken place at Stanford, and most of it has been reported in the
literature, in particular the "Linear Accelerator Issue" 2 1 of
the Review of Scientific Instruments (February 1955). This no-
table development at Stanford has culminated in the present
"SLAC" 2-mile linac, which was completed in 1966 and oper-
ates at 20 GeV.
Microtron. T h e principle of the microtron is similar to that
of the cyclotron, but it is designed to accelerate electrons rather
than positive ions. It was proposed originally by Veksler. 22
T h e accelerating cavity is small and is near the outer bound-
ary of the magnetic field. T h e successive circular orbits of the
electrons are all tangent to one another within the cavity. Be-
cause the electrons very quickly attain nearly the speed of
light, they gain energy and momentum but practically no
speed in passing through the accelerating cavity. Each succes-
sive orbit is longer than the preceding one by one wavelength
of the accelerating electric field, and the sequence of orbits
forms a series of tangent circles of uniformly increasing diame-
ter. An advantage of the microtron is the ease with which the
beam can be ejected from the largest, final orbit.
A type of phase stability exists in the microtron, based on
stable oscillations in the phase of crossing the accelerating cav-
ity. A particle which arrives behind the equilibrium phase
gains less than the synchronous energy, traverses a smaller cir-
cle and takes less time, and thus arrives at the next gap closer
to the equilibrium phase. Although the number of accelera-
tions is small (typically 10 to 20), some phase focusing results.
[4]
The Story of Alternating-Gradient Focusing
Progress in nuclear physics and particle physics has come in

large measure from the use of a sequence of particle accelera-
tors of higher and higher energy. T h e beams of charged parti-
cles which they produce, and their secondary radiations, can
disintegrate nuclei and create many new types of particles. T h e
energies achieved with accelerators have increased at an al-
most exponential rate during the past 35 years, averaging a
tenfold increase every 6 years (see Fig. 34). T h e energy record
has been held in turn by voltage multipliers, cyclotrons, beta-
trons, synchrotrons, synchrocyclotrons, proton synchrotrons
and—most recently—alternating-gradient synchrotrons.
T h e latest in the sequence of improved concepts in the de-
sign of accelerators has been the use of alternating gradients in
the guide-field magnets of synchrotrons, to provide focusing
for the accelerated particles. This development has led to sev-
eral new types of synchronous and resonance accelerators, ca-
pable of either much higher energies or much higher intensities
than earlier machines. This principle of "strong" focusing was
first publicly announced at the Brookhaven National Labora-
tory in the summer of 1952.
A "gradient" magnetic field is stronger on one side than on
the other; it can be formed by tilting the magnet pole faces. A
uniform gradient is one in which the rate of increase (or de-
crease) of the field is constant across the magnetic aperture;
this can be achieved if the pole cross section has the shape of a
rectangular hyperbola. In such a field particles entering on
parallel paths will either converge or diverge; they will be ei-
ther focused or defocused, depending on the sense of the gradi-
ent. Alternating-gradient (AG) focusing utilizes a sequence of
focusing (F) and defocusing (D) sectors, spaced by field-free
(O) regions.
60
61] Alternating-Gradient Focusing
A basic property of such gradient fields is that a cegion

which is convergent (F) in one transverse plane (say the hori-
zontal) is divergent (D) in the other transverse plane (the
vertical), and vice versa. So a repetitive sequence of F, O, and D
sectors will result in net focusing in both transverse planes,
over a rather wide range of gradients, sector lengths, and field
strengths. If the magnet sectors form a closed circular orbit, the
particles will oscillate about a central orbit and be retained
within the aperture of the uniform gradient fields. T h e fre-
quency of these "betatron" oscillations is higher and the wave-
length shorter than in the more weakly focusing magnets of
earlier cyclotrons or betatrons. Amplitudes of particle oscilla-
tions about the central orbit are smaller. T h e magnets and the
enclosing vacuum chambers of a synchrotron can be made
physically smaller.
At the time when the strong-focusing principle was con-
ceived, plans were being made for proton synchrotrons in the
energy range up to 10 GeV. T h e reduction in magnet cross sec-
tion offered by the AG principle would reduce costs for the
magnets and their power supplies by a significant factor. It be-
came economically practical to design accelerators for much
larger orbits and so for higher energies.
B R O O K H A V E N , 1952
I was fortunate to have had a direct association with the
conception and development of the alternating-gradient prin-
ciple at Brookhaven, starting in 1952. It now appears useful to
record my memory of the history of this development, and to
identify the sequence of events and the major contributors.
As general background, I should note that I served as chair-
man of the Brookhaven Accelerator Department, on leave from
Massachusetts Institute of Technology, from the date Brook-
haven was organized in 1946 through the period when the de-
sign of the cosmotron was essentially complete in 1948, when I
returned to MIT. But I retained a strong professional interest
in the successful completion of the cosmotron. Early in 1952

the cosmotron was b r o u g h t into operation for the first time. I
made plans to return to Brookhaven for the summer, bringing
a graduate student and some meson-detection instruments, to
participate in the initial research p r o g r a m at the cosmotron.
However, the new machine could not be completed for re-
search operations d u r i n g the summer, b u t required engineer-
ing consolidation of the control system and modifications of
some of the components.
Before I arrived at Brookhaven I h a d learned of the im-
p e n d i n g visit of a delegation of scientists representing the
newly organized E u r o p e a n laboratory at Geneva now known
as C E R N , to assess the 3-GeV cosmotron as a model for a 10-
GeV accelerator. T h e y p l a n n e d also to visit the University of
California to study the bevatron, which was u n d e r construc-
tion.
In anticipation of this visit by the C E R N delegation, I felt it
would be useful to review the design features of the cosmotron,
to see if it could be extended to 10 GeV. As a start, I set myself
the task of improving the efficiency of the magnet. I h a d been
largely responsible for choosing the C-shaped cross section for
the yoke of the cosmotron magnet, a n d was sensitive to criti-
cisms based on its limitations. It was k n o w n that magnetic
saturation effects reduced the useful radial a p e r t u r e seriously
at high fields, to about one-third of the width at injection
fields.
In the standard synchrotron, o r b i t stability is achieved by
use of a magnetic field with a small radial decrease, commonly
expressed in terms of the "η-value," where η is the exponent of
the variation in the radially decreasing field: Bz = B0(r0/r)n.
For stability of the orbit in both radial and vertical transverse
coordinates in a synchrotron, the rc-values must lie between
zero a n d unity. T h e value chosen for the cosmotron was n =
0.6; this required a field that decreased by a b o u t 6 percent
across the 30-in. width of the radial aperture. T h e poles were

flat and nearly parallel.
It seemed to me that the asymmetric saturation of the C-
shaped cores could be compensated, and yet the advantages of
the C-shape retained, by alternating the locations of the back
yokes of the magnets from inside to outside the orbit circle.
However, I found that, if the pole faces were shaped to give
η = 0.6 for both types at injection fields, the asymmetric satura-
tion at high fields would result in positive gradients in much
of the gap between pole faces for one yoke location and nega-
tive gradients for the other. My first concern was whether this
alternation in gradients would destroy orbit stability.
I discussed this question of orbit stability with my theoreti-
cal colleague Ernest Courant, and he took the problem home
with him that evening. T h e next morning he reported, with
some surprise, that preliminary calculations showed the orbits
to be stable and to have even smaller transverse amplitudes
than in the weak gradient of a standard synchrotron. As I re-
call, the set of gradients used by Courant in this first calcula-
tion was f o r n i = + 1.0 and «2= —0.2, which was assumed togive
an average value of Τΐ= +0.6 as used in the cosmotron. T h e
significance of this result was discussed with others in the labo-
ratory, notably Hartland Snyder, and no fault could be found
with Courant's analysis.
This was the start of an exciting period at Brookhaven. It
seemed that if a little alternating gradient was good, more
should be better. Courant's next calculations were for n-values
of about ± 10, which showed even stronger focusing and
smaller amplitudes of oscillation. It became clear that the
average value of ñ = + 0 . 6 was unimportant, but that a new
type of stability was associated with the alternation in gradi-
ents.
Larger and larger gradients were assumed in further stabil-
ity calculations, with η values of 10, 100, 1000, and even more.
As they increased, I developed sketch designs of magnetic cir-

cuits to provide the high-gradient fields (Fig. 24). T h e magnet
poles became more sharply tilted and narrower. We realized
that the pole faces should be shaped to a rectangular hyper-
bola to provide a uniform gradient across the aperture. As the
FIG. 24. Early sketch of a synchrotron magnet with very high gradients.
See reference 4.2. T h e back-legs of the gradient magnets would be located
alternately on the inside and outside of the orbit.
experimentalist on the team, I kept busy designing the

strangely shaped magnetic circuits with hyperbolic pole faces
and small cross sections. Some of these early speculations led to
such large gradients and small apertures that construction was
obviously impractical. I recall the day when the speculations
reached η-values of 10,000, resulting in pole faces so narrow
and steep and a magnetic aperture so small that the largest
vacuum chamber which could be installed was less than 1 in.
in diameter; at this point I objected that we had passed the
bounds of practicability.
Courant also studied the synchronous oscillations in gradi-
ent fields and found them to be stable as in the normal syn-

chrotron. Furthermore, the orbits of particles having a consid-
erable spread in momentum were found to be compacted into
a narrow radial band whose width varied inversely as the ri-
valile. So as η-values increased and pole faces became nar-
rower, the acceptable spread in particle momentum remained
large. Stability limits were identified, and illustrated with the
"necktie" diagram which became a familiar symbol of A G fo-
cusing. Suitable configurations of F and D magnets and O
straight sections were devised for the arrangement of magnet
units in circular orbits, such as F O D O , FOFDOD, and
FOFODODO.
Snyder recognized and developed the generality of the sta-
bility principle. He noted that the alternating magnetic forces
on charged particles resulted in a type of dynamic stability
that has many analogues in mechanical, optical, and electrical
systems. For example, an inverted pendulum is unstable under
static forces, and will fall to one side with any small displace-
ment from the vertical. However, if the base is oscillated
rapidly up and down through a short stroke, the pendulum is
stable in the inverted position over a wide range of oscillation
frequencies.
T h e use of gradient fields as lenses for charged particles in
linear beams was studied. T h e magnet proposed for such ap-
plications had four poles of alternating polarity, so field inten-
sity was zero on the axis (Fig. 25). A doublet formed of two
such "quadrupoles," in which the gradients in the second unit
are oriented at 90° to the first, forms a lens that focuses diver-
gent charged particles in both transverse directions (Fig. 26).
T h i s was the first development of the quadrupole lens systems
which are now commonly used in accelerator laboratories for
control of linear particle beams. T h e strength of the focusing
possible with such a quadrupole lens greatly exceeds that of a
solenoidal magnetic field, and the power requirements for fo-
cusing high-energy beams are much less. W e noted that similar
FIG. 25. Cross section of a quadrupole magnet for the focusing of linear
particle beams.
focusing could be achieved with quadrupoles formed of per-

manent magnets. In a frivolous moment we speculated on the
use of small-diameter evacuated beam pipes surrounded by
permanent-magnet quadrupoles in cables many miles long,
through which high-energy particle beams from the Brook-
haven accelerators could be piped to the several Associated
Universities that operate the Brookhaven National Labora-
tory.
FIG. 26. Schematic diagram of the focusing of particle beams traversing

two quadrupole magnets in which the directions of gradients alternate.
Meanwhile John Blewett showed that alternating electric-

field gradients had the same focusing properties as magnetic-
field gradients. Transverse electric-field gradients, alternating
in sign, between sets of electrodes of hyperbolic cross section in
a quadrupole array, will also focus particle beams along the
axis. Strong-focusing electric lenses of this type appeared to be
practicable for other types of accelerators. This feature was first
utilized in the "electron analogue" of an AG accelerator built
at Brookhaven to test the AG principle.
EARLY P L A N N I N G FOR AG A C C E L E R A T O R S
When the CERN delegation, consisting of Odd Dahl, Frank
Goward and Rolph Wideröe, arrived in Brookhaven, the AG
concept had been developed sufficiently to be presented to
them as a significant improvement over the cosmotron design.
They were impressed with the potentialities of alternating
gradients, and on their return they stimulated studies of AG
orbit stability in British and European laboratories. By the
time they were ready to proceed to the University of Califor-
nia, we realized that the development at Brookhaven had oc-
curred so rapidly that other United States laboratories had not
yet been informed. It seemed particularly important that the
Berkeley group should not have to learn of this basic develop-
ment through their CERN visitors. So Leland Haworth made
a long-distance telephone call to Berkeley to inform them of
the AG concept before the CERN delegation arrived.
Later we learned that the Berkeley staff were in the embar-
rassing position of being unable, owing to security classifica-
tion, to describe their own developments of focusing in the
cyclotron by azimuthally varying fields, as originally proposed
by L. H. Thomas 1 in 1938. T h e use of such sector focusing in
cyclotrons has since led to the development of a category of
high-intensity "isochronous" cyclotrons in the energy range
from 50 to 100 MeV. By the time these two lines of develop-
ment merged a few years later, it became evident that Thom-
as's proposal of sector focusing was a special case of the general

theory of A G focusing as applied to constant magnetic fields,
a n d that the Berkeley g r o u p were also working on a type of
alternating-gradient focusing at that time.
Most of these developments at Brookhaven occurred within
a few weeks' time and involved primarily four staff members.
T h e first report was sent to the Physical Review on August 21
a n d was published 2 in the December 1, 1952 issue. T h e report
discussed the principle of alternating-gradient focusing with
magnetic fields, a n d described design concepts for a 30-GeV
accelerator. T h i s design was a radical d e p a r t u r e from past
magnetic accelerators, using an extremely high gradient (n =
3600), a vacuum c h a m b e r with an a p e r t u r e of only 1 X 2 in., and
FIG. 27. E. D. Courant, M.S. Livingston, H. S. Snyder, and J. P. Blewett

demonstrating the relative cross sections of the cosmotron magnet and a
speculative alternating-gradient magnet of very large gradient.
an orbit circle of 300-ft radius (Fig. 27). T h e report also de-

scribed a quadrupole magnetic lens system for focusing high-
energy beams in linear systems. A companion paper by
Blewett 3 presented the parallel case of focusing by electric-
field gradients, and described its application to linear accel-
erators.
By this time it had become clear that the AG principle had
important potentialities and deserved thorough study. A
major step was reported by Adams, Hine, and Lawson 4 of the
Harwell Atomic Energy Establishment early the next year.
They identified and studied the problem of orbital resonances
which might threaten orbit stability. If the frequency of the
transverse "betatron" oscillations were an integral multiple of
the orbital frequency, the effect of even a very small orbit
perturbation would be cumulative and could build up disas-
trous amplitudes in a few revolutions, resulting in loss of the
beam. T o a lesser extent the same is true of half-integral and
other subintegral resonances. For a time this seemed to set a
serious limitation to the use of alternating gradients. But fur-
ther work showed that objectionable resonances could be
avoided by care in design and by use of suitable control systems
to maintain constant nonintegral frequencies during the accel-
eration cycle, even with η-values as large as 100.
As so frequently happens in a technical field, this concept
was developed independently elsewhere. N. C. Christofilos, an
electrical engineer of American birth, educated and working
in Athens, had been studying accelerators as a hobby for some
years. He suggested several new and unusual ideas on acceler-
ator design, in the form of private reports and patent applica-
tions. An unpublished r e p o r t 5 dated 1950 presented the con-
cept of AG focusing and the conceptual design of an acceler-
ator using this principle. He also applied for United States
and European patents. A copy of his report was privately
transmitted to the University of California Radiation Labora-
tory at that time, but was not given serious consideration.
After the Brookhaven publication in 1952, Christofilos came

to the United States and demonstrated his priority. This was
recognized in a brief note published by Courant, Livingston,
Snyder, and Blewett 6 in 1953. Christofilos joined the staff of
the Brookhaven Laboratory for a time, where he continued his
speculative designing of accelerators and other devices and
also contributed to the laboratory program leading toward a
large AG accelerator and to its linear-accelerator injector.
DESIGN STUDIES
In the United States, I started a design study late in 1952 at
Massachusetts Institute of Technology, with the assistance of
members of the M I T and Harvard Physics Department staffs.
It resulted in a laboratory report 7 dated June 1953. This was
the earliest relatively complete design study for an AG acceler-
ator, but it was dropped when financial support for the AGS at
Brookhaven was authorized by the U. S. Atomic Energy Com-
mission in 1954. However, the Cambridge group promptly
transferred their interest to electrons and designed the first
multi-GeV electron synchrotron using AG magnetic focusing,
which was later constructed and is now in operation as the 6-
GeV Cambridge Electron Accelerator at Harvard.
Both the Brookhaven and the CERN groups initiated de-
sign studies in 1953 for AG accelerators in the energy range of
25 to 30 GeV. This energy was ten times that of the cosmotron,
the highest-energy accelerator then in operation, and five
times that of the bevatron, which was under construction. T h e
collaboration between the two groups, which started in 1952,
continued with exchange of staff and design information. As a
result the two machines have striking similarities. Theoretical
groups at both laboratories were concerned with the effects of
magnet misalignments and with beam behavior at phase tran-
sition, and both groups wished to check the theory with a
working model before committing the final design. A compro-
mise acceptable to both was for Brookhaven to build an elee-
FIG. ¡28. T h e 28-GeV alternating-gradient proton synchrotron (CPS) at
the CERN Laboratory in Geneva.
tron model using electrostatic-gradient fields, while CERN

proceeded with detailed design. This "electron analogue" was
first operated in 1955 at an energy of 10 MeV, and supported
the most optimistic theoretical predictions. Construction
could proceed with confidence.
T h e CERN group was initially under the direction of O.
Dahl of Norway and F. Goward of England. Later, Dahl re-
turned to Norway and, following Goward's untimely death,
direction of the design study passed to J. B. Adams of England
and C. Schmelzer of Germany. T h e CERN CPS was brought
into operation at 26 GeV in 1959 (Fig. 28). T h e Brookhaven
design group was headed by L . J . Haworth, G. K. Green, and

J. P. Blewett. T h e AGS was completed in July 1960 and soon
exceeded its design energy to operate at 33 GeV (Fig. 29). Both
machines have been highly successful and each has supported
an impressive program of research experiments. Beam intensi-
ties have far surpassed the designer's estimates, and the use of
multiple targets and emergent beams has broadened their
capabilities and effectiveness for research support.
T h e possibility of extension of the AG principle to fixed-
field magnets (FFAG) was soon recognized, by scientists in
many places, including Brookhaven, Japan, and the USSR.
Another United States group formed a design laboratory
called Midwestern Universities Research Association
(MURA), whose primary function was to design a high-energy
accelerator for the region; they emphasized the FFAG princi-
pie in their design studies. T h i s is n o t the place to go into the

history of M U R A , b u t the g r o u p has m a d e m a j o r contribu-
tions to accelerator theory a n d t e c h n i q u e . T h e most i m p o r t a n t
a p p l i c a t i o n of fixed-field a l t e r n a t i n g - g r a d i e n t focusing u p to
t h e present is t h a t used in "sector-focused" or "isochronous"
cyclotrons, m e n t i o n e d earlier as based o n t h e 1938 p r o p o s a l by
T h o m a s a n d first developed at the University of California.
T h e first w o r k i n g accelerator to use t h e A G focusing princi-
p l e was a 1-GeV electron synchrotron being constructed at
Cornell University at t h a t t i m e by R . R . Wilson a n d his asso-
ciates. T h e m a g n e t h a d replaceable pole-face u n i t s i n t e n d e d
originally to be of the u n i f o r m - g r a d i e n t type. However, w h e n
t h e C o r n e l l g r o u p h e a r d of the B r o o k h a v e n developments,
they rapidly designed a n d o r d e r e d pole-face u n i t s w i t h alter-
n a t i n g gradients. By late 1953 the Cornell m a c h i n e was oper-
a t i n g at 1-GeV energy. A l t h o u g h the A G pole tips were n o t
essential to the o p e r a t i o n of this s y n c h r o t r o n , they p r o v i d e d
t h e e q u i v a l e n t of a larger a p e r t u r e a n d s o m e w h a t h i g h e r in-
tensities. T h e Cornell g r o u p has recently (1967) c o m p l e t e d a
10-GeV electron accelerator using a l t e r n a t i n g - g r a d i e n t fo-
cusing.
T h e most i m p o r t a n t a p p l i c a t i o n of A G focusing has been to
p r o t o n synchrotrons. It has led to t h e C E R N CPS w h i c h oper-
ates at 28 GeV, a n d the B r o o k h a v e n A G S w h i c h for m a n y years
h e l d the world's record w i t h an energy of 33 GeV. A larger ma-
c h i n e has recently been c o m p l e t e d in S e r p u k h o v in the USSR,
o p e r a t i n g at 70 to 80 GeV. T h e A G p r i n c i p l e has completely
c h a n g e d t h e basic designs for the m a g n e t s used for synchro-
trons. Particle orbits can be r e t a i n e d between m u c h smaller
m a g n e t poles a n d w i t h i n smaller v a c u u m chambers. T h e basic
p r i n c i p l e is still t h a t of s y n c h r o n o u s acceleration, b u t m a g n e t
a p e r t u r e s a n d cross sections d o n o t need to increase as o r b i t
r a d i u s is m a d e larger. It has b e c o m e economically practical to
design s y n c h r o t r o n s for m u c h h i g h e r energies.
PLANS FOR T H E F U T U R E
Research scientists in the field of particle physics are clamor-
ing for accelerators of still higher energies, and designs are in
process in several countries. In all of these design studies the
decision has been simply to enlarge the orbital dimension of
the AG proton synchrotron. In general, the basic design con-
cepts and technical features of the present 30-GeV machines
are believed to be capable of extension to energies of several
hundred GeV. T h e transverse dimensions of magnets and vac-
uum chambers do not increase with increasing energy. All the
scientific and engineering problems of machines of 200- to 300-
GeV energy appear to be solvable. T h e major problem is that
of cost, which is nearly proportional to energy and is estimated
at present to be about $1,000,000 per GeV for the machine and
basic laboratory facilities for research utilization.
Plans are under way in the United States for an accelerator
of 200-GeV energy or higher. A design study at the Lawrence
Radiation Laboratory of the University of California at Berke-
ley for a 200-GeV machine provided the first basis for plans in
the United States. T h e National Accelerator Laboratory at
Oak Brook, Illinois, under the direction of R. R. Wilson, has
prepared a revised design study and a proposal for construc-
tion for the machine, which has been submitted to the Atomic
Energy Commission and the U. S. Congress. A site has been
chosen at Weston, Illinois, and the staff of this National Ac-
celerator Laboratory has been assembled. In Europe a group
at CERN has submitted a preliminary design for a 300-GeV
machine to their governments, and approval is anticipated.
More speculative designs exist for energies up to 1000 GeV.
T h e only known limit is cost. Scientists in the field are con-
vinced that these new machines are necessary to further
progress in particle physics and the study of nuclear forces.
It is clear that the principle of alternating-gradient focusing
has played an essential part in the development of multi-GeV
accelerators and of the research field of particle physics. T h e

reduction in unit costs that it has made possible has allowed
the energy range of accelerators to be extended into a region of
extraordinary scientific interest. T h e even higher-energy ma-
chines of the future will involve costs that will become signifi-
cant items in national budgets. T h e i r most significant product
will be new knowledge about the fundamental particles of na-
ture and the properties of nuclear forces. Justification for sup-
port of this research field must be found in the value of this
new knowledge to our society. Many scientists believe that we
are at the threshold of a significant upward step in human un-
derstanding. Ultimately, the impact of this new knowledge
will be felt on a very broad intellectual and cultural level.
High-energy physicists and accelerator enthusiasts are hopeful
that the giant accelerators now being planned will be built
and will contribute significantly to this growing understand-
ing of nature.
[5]
Origins of the Cambridge Electron Accelerator
T h e Cambridge Electron Accelerator can be described in

vector notation as the sum of three forces: the Harvard Physics
Department's desire to have an accelerator capable of produc-
ing mesons, the continuing M I T interest in high-energy elec-
tron and photon physics, and my own experience in the design
of accelerators. It was indeed fortunate that these vectors
pointed in the same direction in the planning years of 1952—
1956, so their sum was the unique and successful 6-GeV syn-
chrotron at the CEA. T h e joint administration was made pos-
sible by the growing interest at Harvard and M I T in provid-
ing closer cooperation between the two institutions. Nation-
ally, it came at just the right time to take advantage of policy
decisions within the Atomic Energy Commission to support
accelerator facilities in the larger universities in addition to
the National Laboratories.
T h e CEA was the first multi-GeV electron accelerator using
alternating-gradient focusing, and held the energy record for
electrons from its completion in 1962 until shortly before the
SLAC 2-mile linac at Stanford came into operation in 1966. It
was the first synchrotron to take full advantage of the small
apertures possible with AG focusing, so its magnets were
small, compact, and relatively inexpensive. T h e magnet cy-
cling rate of 60 Hz was achieved with a unique resonant pow-
ering circuit of modest power level. It utilized an unusually
high radiofrequency for acceleration and phase focusing, and
high rf power for compensation of radiation losses. T h e accel-
erated beams were reduced by radiation damping to such
small dimensions and energy spread that they could be ejected
and focused onto external targets on a spot a few millimeters
across, a tremendous experimental advantage.
Long before the CEA was completed, its unique design fea-
76
77] The Cambridge Electron Accelerator
tures and the growing scientific interest in electron and pho-

ton physics made it the model for developments in other labo-
ratories. T h e Deutsches Elektronen Synchrotron Laboratory
in Hamburg, Germany, sent a sequence of staff members to
Cambridge to learn-by-doing at the CEA; this "DESY" ma-
chine was completed in 1965, and has been run at energies up
to 7.0 GeV. T h e "NINA" machine at the Daresbury Labora-
tory (near Liverpool, England) was completed and dedicated
in 1967; it operates at 4.5 GeV. At Cornell University a large-
orbit, somewhat simplified AG synchrotron rated for 10 GeV
was brought into preliminary operation in 1967. And at the
Physical Institute of the Armenian Academy of Sciences in
Erivan, a 6.5-GeV machine based on the CEA called "ARUS"
was completed in 1967. So the CEA has had an impact far out-
side Cambridge, its staff members have earned the respect of
accelerator scientists throughout the world, and it has "put
through school" a long list of younger "sisters."
T h e Harvard interest which culminated in the CEA can be
traced back to about 1948, even before the synchrocyclotron
which is just terminating its career was completed, when the
production of π-mesons by the 184-in. cyclotron at Berkeley
showed that the 150-MeV energy of the Harvard machine was
just below the meson threshold. T h e cyclotron was built by
K. T . Bainbridge, R. R. Wilson, W. Davenport, and others,
and a research program was initiated, but the Harvard group
anticipated the need for, and kept alert to the possibilities of, a
higher-energy machine. T h e 1952 Report of the Visiting Com-
mittee to the Physics Department (J. Robert Oppenheimer,
chairman) said: "We believe it not unlikely that within the
next five years Harvard will desire to build at least one new
major piece of equipment to fill this gap." So in 1952 the Har-
vard physics group were actively searching for an opportunity
to supplement the cyclotron, when the first discussions started
with M I T on the feasibility of a jointly sponsored accelerator.
At MIT, the postwar choice of a high-energy accelerator was
an electron synchrotron, with an energy of 300 MeV. T h e

principle of synchronous, or phase-stable, acceleration of elec-
trons in a r i n g magnet using a pulsed magnetic field was an-
nounced in 1945, as we have seen, independently by Ε. M. Mc-
Millan at Berkeley a n d V. Veksler in the USSR. By 1946 the
technique had been demonstrated experimentally and theo-
retical analysis of particle motion was in an advanced stage.
An experimental g r o u p headed by Ivan Getting built the
M I T electron synchrotron in 1946-47 and started studies on
the p h o t o p r o d u c t i o n of mesons. T h i s was one of the first elec-
tron machines to use a C-shaped core for the ring magnets, a n d
the M I T g r o u p contributed several other i m p o r t a n t design
innovations. By 1952 this g r o u p was deeply involved in π-
meson physics, and were p l a n n i n g for a larger synchrotron
with an energy tentatively chosen as 1.5 GeV.
Meanwhile, I r e t u r n e d to M I T in 1948 from my position as
chairman of the Accelerator D e p a r t m e n t a l Brookhaven, where I
h a d helped initiate the accelerator program a n d the prelimi-
nary designs for the cosmotron. By early 1952 the cosmotron
was a p p r o a c h i n g completion a n d I h a d t u r n e d my interests to
the local needs in Cambridge. I joined in discussions with Jer-
rold Zacharias, then chairman of the M I T Nuclear Science
Laboratory, a n d N o r m a n Ramsey, representing the Harvard
physics group, on the feasibility of an accelerator in Cam-
bridge. W e also knew that certain E u r o p e a n scientists were
p l a n n i n g a proton synchrotron of about 10-GeV energy, to be
based on the cosmotron or the Berkeley bevatron designs. O n
May 14, 1952, I sent a m e m o r a n d u m to President Stratton of
M I T proposing a summer design-study g r o u p to anticipate
these needs, a n d approached Leland Haworth, director at
Brookhaven, to release Ernest Courant, H a r t l a n d Snyder, a n d
others for such a study.
H a w o r t h felt that he could not release any Brookhaven staff
for that critical time a n d m a d e a counterproposal that I come
to Brookhaven for the summer (1952), in which case their key
people could devote part of their time to a design study. I ac-

cepted this offer and went to Brookhaven. One of the first re-
sults of this study, as we have seen, was the concept of
alternating-gradient focusing, which was immediately recog-
nized as of potential importance in the design of higher-energy
machines. Before the end of the summer, I joined with Cou-
rant and Snyder in publishing the first report of the AG prin-
ciple, including a preliminary design for a 30-GeV machine.
Meanwhile, Ramsey came to Brookhaven in August and be-
came enthusiastic about the prospects for AG machines. We
immediately started preparing a draft of a proposal to the
Atomic Energy Commission for a jointly sponsored accelerator
of this new type to be located in Cambridge, and discussed the
possibility with Thomas Johnson, then director of the AEC
Research Division, when he visited Brookhaven later that
summer. On our return to Cambridge we organized a self-
constituted "Joint Accelerator Committee," consisting of
Ramsey and J. Curry Street from Harvard and Zacharias and
myself from MIT, to carry on the planning. Although this
initial committee was an ad hoc group, other department
members and both university administrations were kept in-
formed and gave us full support. On October 10, 1952, we sub-
mitted a preliminary draft, and on October 31 a more definite
request to the AEC, for support of a joint design study of a 10—
20-GeV AG proton synchrotron to be carried on in Cambridge
under my direction. This request was approved, and arrange-
ments were made for funds to be supplied through a subcon-
tract from Brookhaven to M I T for the period from November
1, 1952, to J u n e 30, 1953.
I quickly organized a design study and arranged for labora-
tory space at MIT. Bulletin No. 1, dated November 7, 1952,
scheduled a series of six planning conferences within the next
week on the systems design and the several major components.
This study involved at the start at least 20 members of the
M I T and Harvard departments. A small group of full-time
people were enlisted, including R. Q. Twiss from England, to

extend the theoretical analysis, a n d J . A. H o f m a n n , A. Vash,
a n d J . F. Frazer, who built dc and ac models of high-gradient
magnets a n d instruments to measure gradient fields. T h e re-
sult of this joint effort was a 200-page report: "Design Study for
a 15-BeV Accelerator," published as MIT Laboratory for Nu-
clear Science, Report No. 60, dated J u n e 30, 1953. T h i s was the
first relatively complete design study of a multi-GeV AG accel-
erator; it included theoretical studies of resonances, synchro-
nous stability, phase transition, d a m p i n g , and the like. It also
included a table of parameters (for example, orbit radius 124
ft) a n d a cost estimate (about $8,000,000). O n August 11, 1953,
the J o i n t Accelerator Committee, acting with the approval of
the university administrations, forwarded to the AEC a pro-
posal, backed u p by the design study, for the construction of
this accelerator in Cambridge.
Meanwhile, the Brookhaven staff were engaged in their own
design study of a larger A G accelerator, initially conceived to
be for 50 GeV a n d later reduced to 25-30 GeV, to be located
at Brookhaven. Paralleling this effort, some of the E u r o p e a n
scientists w h o were then organizing the C E R N laboratory co-
o r d i n a t e d their design studies in several countries; this effort
culminated by late 1953 in a design for an A G p r o t o n synchro-
tron of a b o u t 25-GeV energy.
I n September 1953, T . J o h n s o n of the AEC informed us that
his office would r e c o m m e n d s u p p o r t of the Brookhaven pro-
posal for a 25—30-GeV machine b u t could not s u p p o r t a competi-
tive C a m b r i d g e proposal. A l t h o u g h we b r o u g h t the heaviest
guns from both university administrations to bear, this recom-
m e n d a t i o n was confirmed, and we were informed in December
of the Commission's decision to refuse the C a m b r i d g e pro-
posal. T h i s led to renewed political a n d technical activities in
Cambridge, including discussions with the National Science
F o u n d a t i o n as a possible alternative source of s u p p o r t i n g
funds. T h e s e discussions led the N S F to schedule a meeting of
an Advisory Panel on U l t r a h i g h Energy Nuclear Accelerators

for early spring 1954, at which Ramsey testified for the Cam-
bridge group, a n d which proposed a consistent U n i t e d States
accelerator program.
ELECTRON SYNCHROTRON
O n the technical side, the AEC decision led to f u r t h e r dis-
cussions between Harvard and M I T scientists, in which an
electron accelerator was seriously considered as an alternate
possibility. Although protons were clearly superior for studies
of strong nuclear interactions, the prospects of a large proton
accelerator at Brookhaven reduced the attractiveness of a
lower-energy proton m a c h i n e in Cambridge. Also, the M I T
g r o u p working with their 300-MeV synchrotron wished to con-
tinue their electro- a n d p h o t o p r o d u c t i o n studies of mesons.
For a time a hybrid machine for 6-GeV protons and 2.5-GeV
electrons was considered. But soon it became clear that o u r
best h o p e for a u n i q u e facility in C a m b r i d g e would be a high-
energy electron machine.
I was personally convinced that A G magnetic focusing
could be applied to an electron synchrotron, despite the very
fast cycling rate required. T h e AG magnet lattice with field-
free straight sections between magnet sectors was ideal for in-
stallation of an rf system having many high-Q resonant cavi-
ties to develop the large value of the volts-per-turn ratio
needed to compensate for radiation losses. T h e fast cycling a n d
the high source intensity possible with electrons should in-
crease average beam current and largely compensate for the
lower production cross section for nuclear processes t h r o u g h
the electromagnetic interaction. And we hoped that theoreti-
cal interpretations might be simpler when the known electro-
magnetic interaction was involved. D u r i n g the winter and
early spring of 1954,1 drew sketches a n d calculated parameters
for a sequence of AG electron synchrotrons with increasing
energies, which were discussed by the H a r v a r d - M I T group.
As our confidence grew we realized that the choice was one of

balancing the obvious advantages of very high-energy elec-
trons against the increasing complexity and cost of such a high-
energy machine. O u r desire was to build a facility for univer-
sity research, convenient to the scientists and students and op-
erated as nearly as possible like a university research labora-
tory.
A site location at Harvard was agreed on at an early date.
T h e filled land near M I T was clearly unsuitable for the
underground-tunnel building needed for shielding. Sites out-
side Cambridge would be much less convenient for the univer-
sity users. T h e Harvard administration gave tentative ap-
proval to our plans to use the area around and beyond the Cy-
clotron Laboratory, which was undeveloped at that time, was
close to the existing Harvard physics buildings and laborato-
ries, and was suitable for excavation. Both M I T and Harvard
representatives approved of this location.
T h e National Science Foundation Advisory Panel on Ul-
trahigh Energy Nuclear Accelerators, referred to earlier, met
on March 23, 1954 in Washington, and Ramsey presented
arguments for accelerators in universities. T h e resulting Panel
Report recommended federal support for some university ma-
chines in addition to those in the existing National Laborato-
ries, and was important in establishing sound principles for
support of the newly developing field of high-energy physics.
It also opened the possibility that the NSF might support uni-
versity accelerators if the AEC did not, which may have had
some influence on AEC policy.
T h e group of interested Harvard and M I T scientists con-
tinued their discussion meetings during the spring of 1954,
and the design of an electron accelerator took form. I summar-
ized the status in a report, "Design Study for an AG Electron
Synchrotron," April 26, 1954, which was later given the report
number CAP-1 when the "Cambridge Accelerator Project" or-
ganization was established. This report gave design parame-
ters for 6-GeV protons and 2.5- and 5.0-GeV electrons, with
rough cost estimates. T h e Cambridge Joint Accelerator Com-
mittee used the 5-GeV electron design data from this report
and recommended it to the Harvard-MIT group in an un-
numbered report in May 1954. This became the clear choice of
the Cambridge group. It was presented to the Harvard Nu-
clear Physics Committee and approved unanimously at a
meeting on June 7, 1954. During the summer the design stud-
ies continued, and the energy being considered rose to 6.0
GeV. Representatives of the Joint Accelerator Committee
(Livingston and Ramsey) prepared a "Proposal for a 5- to 6-
BeV Electron Accelerator," dated August 17, 1954 (later given
the number CAP-2), which included parameters and cost esti-
mates, and submitted it to the AEC.
On August 18, I sent a request to Johnson at the AEC for
funds to continue the design study during the interval before
the proposal could be acted upon. This request was deferred
from month to month, and the local design group was reduced
to a few part-time people.
During this interval our understanding with the AEC Re-
search Division was that they would recommend construction
of two university accelerators, one at Cambridge and another
at Princeton (later Princeton-Pennsylvania) and that we
could expect Congressional action in the spring of 1955.
During the Physical Society meetings in New York in Janu-
ary 1955,1 heard some disturbing rumors of a change in policy
and planning within the AEC. When traced down they proved
to be valid. Following a recommendation by the General Ad-
visory Committee, the AEC had removed the construction
items for the two accelerators from their budget, and was re-
questing only design funds for the coming fiscal year. Further-
more, Johnson had decided that the AEC would make a new
survey of all accelerator requests throughout the country, and
would support only the most promising with design funds.
First priority would go to the National Laboratories, second to
groups of universities, and third to individual universities.

This seemed to us to be a reversal of AEC policy regarding
university accelerators, and it certainly would cause a delay of
a full year in our plans for Cambridge.
This delay was very disappointing to the Cambridge group.
We felt it was essential to enlist an effective design staff imme-
diately and to proceed with the detailed design. T o meet the
increased competition it would be necessary to submit a much
more detailed proposal than our first one, but we had no funds
to carry on the design study. In this emergency the Joint Ac-
celerator Committee approached the Harvard and M I T ad-
ministrations for special funds. I prepared a minimum budget
of $33,000 for six months, half to be provided by each of the
institutions. On May 17, Ε. M. Purcell and I, representing the
Joint Accelerator Committee, met with administration offi-
cials to work out details, and on June 23, 1955 we received au-
thorization for the amount needed from each institution, so we
could proceed with the design study. T h e study was to be
known as the Cambridge Accelerator Project (CAP) and
would be administered by the Harvard Office for Research
Contracts.
During the spring of 1955 Congress somewhat unexpectedly
reinserted $10,000,000 in the AEC appropriations bill for two
university accelerators. However, the AEC continued with its
plans for what we called the "box-top" contest. On July 6,
Thomas Johnson sent letters to the presidents of Harvard,
MIT, and a large number of other universities requesting ac-
celerator proposals, with a deadline for submission of October
1. At the request of other applicants this date was extended to
January 1, 1956. M. G. White at Princeton saw the virtues of
a joint proposal such as ours and revised his plans to include
the University of Pennsylvania. A few other proposals were
submitted.
Meanwhile, the CAP took form in Cambridge and our de-
sign activities intensified. During the summer Giovanni
Lanza, Kenneth Robinson, and Ralph Waniek were employed

full time, and 20 Harvard and M I T staff members contributed
part-time help. T h e final result of this effort was a 64-page
"Proposal for a 6-BeV Electron Accelerator" (CAP-15), dated
December 15, 1955, which was submitted to Johnson of the
AEC by Presidents Killian and Pusey of M I T and Harvard.
T h e CAP group continued technical studies during the fol-
lowing spring, producing a total of 25 reports and continu-
ously modifying the design parameters. I note, for example,
that the final (enlarged) orbit radius of 118 ft was first reported
in CAP-18, January 6, 1956, and that revised parameters for
the radiofrequency system were set in CAP-25, dated May 1,
1956. We found that the cost estimate in the proposal would
not be sufficient to build the enlarged accelerator and also an
experimental laboratory sufficient to utilize its capabilities.
This resulted in a supplementary proposal for an additional
$2,000,000, dated June 1, 1956.
A "letter of intent" contract between the AEC and Harvard
University for funds to construct the accelerator was signed on
April 2, 1956. This became the official starting date of the
CEA. A letter of agreement had been exchanged earlier be-
tween the presidents of the two institutions, dated March 7,
1956, which gave it the name Cambridge Electron Accelerator,
described the intent of sharing the research opportunities
equally, established the administrative system of an Executive
Committee with equal representation from Harvard and M I T
and, since the accelerator was to be located on Harvard prop-
erty, named Harvard University to be the contracting agency
and to provide contractual and budgetary supervision. T h e
total funds authorized in the original definitive contract were
$6,500,000, for construction of the laboratory and the acceler-
ator. Later proposals covered the anticipated annual cost for
research operation following completion. T h e eventual cost of
construction of the machine and the full-size laboratory was
$11,600,000; this is just under $0.002/volt.
86]
F I G . 30. Artist's conception of the Cambridge Electron Accelerator.
T h e first addition to the staff following the official start of

the CEA was our late dearly beloved business manager, Robert
E. Cummings. Lanza, Waniek, and Robinson continued from
the CAP roster. Others joining us during the next year were
Paul Cooper, John Rees, Lee Young, T o m Collins, Janez De-
kleva, Maurice Vallauri, Manfred Wächter, and several part-
time staff such as Andrew Koehler and Frank Robie. After oc-
cupying various temporary quarters, we settled into the pres-
ent CEA building when it was completed in 1958 (Fig. 30).
T h e accelerator was completed and research activities were
started in late 1962. It has performed beyond design specifica-
tions in almost all respects. One of its most valuable features is
the high quality and the flexibility in control of the emergent
photon and electron beams as developed by the CEA staff. T h e
number and variety of emergent beams have provided oppor-
tunities for an unusually large number of experimental
groups. T h e number of research scientists and students in-
FIG. 31. T h e magnets, radiofrequency system, and other components of
the Cambridge Electron Accelerator in 1962.
volved in experiments at the CEA is over 150. T h e quality of

the research has been consistently high and the significance of
the scientific results can hardly be overestimated. It has be-
come a successful and valuable university research laboratory
(Fig. 31).
[6]
The 200-GeV Accelerator*
T h e 200-GeV accelerator is the highest-energy machine in the

sequence described in this monograph, and is the culmination
of over 30 years of scientific and technical development. It is
basically an alternating-gradient proton synchrotron of very
large orbital radius. T h e conceptual design represents the
cumulative efforts of hundreds of professional accelerator sci-
entists and engineers, including many from foreign laborato-
ries. T h e origin of individual technical concepts is becoming
increasingly difficult to identify, and the assignment of credit
must be widely shared between many individuals. A large
group of research scientists who have supported the proposal
and hope to use the accelerator for experiments are broadly
spread across the country. Also, much of the planning effort
which has made the 200-GeV machine a reality has come from
senior scientists who are not themselves experts in this field of
science, and from government administrators and political
leaders responsible for maintaining national scientific prog-
ress. T h e design staff that has been assembled has come from
many accelerator laboratories and universities. T h e site has
been chosen to make the laboratory a facility for scientists
throughout the nation. It is expected to become also an impor-
tant international center for research in high-energy particle
physics. T h e project was authorized for construction by the
U. S. Congress on April 19, 1968, and is scheduled for initial
operation in 1972.
ORIGINS
T h e conceptual planning for an accelerator in the hundred-
GeV energy range became possible with the discovery of the
principle of alternating-gradient focusing in 1952 (Chapter 4)
which led to the design and construction of the 28-GeV CPS at
88
89] The 200-GeV Accelerator
CERN and the 33-GeV AGS at Brookhaven National Labora-

tory. Although these two accelerators represented the practical
first steps toward exploiting the new principle, speculative
thinking which was aimed at much higher energies started as
early as 1952 in Brookhaven and at several other laboratories
in the years to follow.
At the University of California Radiation Laboratory the
design staff had been thinking for years about the next step
beyond the 6-GeV bevatron. W. M. Brobeck made an early
feasibility study of design features and methods of powering
magnets for synchrotrons in the 100- to 150-GeV range in 1955.
In 1956 a local Accelera tor-Building Committee was formed,
consisting of E. M. McMillan, E. J. Lofgren, R. L. Thornton,
W. M. Brobeck, L. Smith, and D. L. Judd, which coordinated
planning during the next 5 years during which nearly 50 in-
ternal notes and memorandums were prepared. T h i s program
was summarized by J u d d in I960.1
T h i s early planning was brought to a focus at a summer
study in 1959 of the Midwest Universities Research Associa-
tion (MURA) at Madison, Wisconsin, where questions were
first raised and discussed publicly concerning the scientific
need and feasibility of an accelerator for energies of several
hundred GeV. A memorandum 2 circulated that fall by Mat-
thew Sands, then at California Institute of Technology, de-
scribed a possible 300-GeV "cascade synchrotron" using a very
large ring of alternating-gradient magnets and a smaller syn-
chrotron as an injector. This discussion came at a time when
the CERN CPS had just been brought into operation at 26
GeV and the Brookhaven AGS was approaching comple-
tion.
Even at that time it was recognized that costs would be
nearly proportional to energy and that the unit cost would be
similar to that at CERN and Brookhaven, about $1 million
per GeV. Many realized that such a large and costly accelerator
might properly be supported by an international laboratory
similar to CERN. Discussions were held at the Rochester High

Energy Physics Conference 3 in August 1960 with many for-
eign delegates attending. In September, at a meeting held at
the American Institute of Physics in New York between
United States and Soviet physicists, it was agreed that both
countries should explore further the feasibility and desirabil-
ity of accelerators for energies above 200 GeV. Plans were
made to report progress at the next International Accelerator
Conference, scheduled to be held at Brookhaven in September
1961.
Three United States laboratories carried on design studies
during the early 1960's, at Brookhaven, California Institute of
Technology, and the University of California Radiation Lab-
oratory at Berkeley. Abroad, a design group consisting of ex-
perienced members of the CPS staff at CERN initiated studies,
and later events showed that preliminary studies were also un-
der way in the Soviet Union.
At Brookhaven, a group led by J. P. Blewett started think-
ing seriously about higher-energy machines in 1960. They
chose energies of 400, 700, and 1000 GeV for their initial study.
A preliminary design report was issued in May 1961, and con-
clusions were revised during an extensive study sponsored by
Brookhaven during August 1961. At one time, about 25 accel-
erator experts were assembled from nine centers in the United
States, with others from CERN and the Rutherford Labora-
tory in England. T h e R e p o r t 4 of this study group was pub-
lished in late 1961; it included both a preliminary design and
an analysis of experimental program requirements.
These early concepts for superenergy machines were
strongly influenced by the design of the Brookhaven AGS, but
even more by the belief that large gradients could be achieved,
with very small magnets. T h e magnets designed were compact
and simple. But the orbital radii were so very large that the
engineering planning for other components and for powering
the magnets was quite nebulous. A sequence of improved and

revised tables of parameters for the three energy ranges was
developed during the next few years.
T h e Caltech group organized by Sands made a series of stud-
ies during 1960 and 1961 and issued several internal reports. 5
They studied a variety of special problems, such as the preac-
celeration of the ions in a smaller "booster" synchrotron and
injection into the large orbit. When it became clear that the
University of California Radiation Laboratory and Brook-
haven were becoming involved in similar design thinking, a
meeting was held on neutral ground at UCLA in December
1961 between members of the three groups. An agreement was
reached under which U C R L would carry on the West Coast
design efforts in the 100- to 300-GeV range, and Brookhaven
would continue to explore the higher energy range. T h e
Caltech effort was phased out in the following year.
Design planning was followed up much more seriously at
the University of California Radiation Laboratory. T h e status
of U C R L studies was reported by J u d d and Smith 6 in late
1961. In February 1962 the laboratory submitted a formal re-
quest to the United States Atomic Energy Commission for
support of a design study in the 100-300-GeV range and re-
newed the request in December 1962. Early concepts at U C R L
were similar to those at Caltech and Brookhaven in that they
involved a large ring of alternating-gradient magnets for the
main synchrotron. But initially they explored the use of a 1- to
2-GeV proton linac as an injector into the main ring. T h e
name of the laboratory was changed to the Lawrence Radia-
tion Laboratory (LRL) in 1959, in honor of Ernest O. Law-
rence, who died in 1958; Ε. M. McMillan became the new
director.
At an International Conference on High-Energy Particle
Accelerators held at Brookhaven in September 1961, an exten-
sive program was arranged for the exchange of information on
designs of multihundred-GeV accelerators. Scientists from

CERN and the European countries attended and described
their existing designs; the three U. S. laboratories also pre-
sented detailed status reports. Unfortunately, however, the
USSR delegation did not arrive, so the exchange was incom-
plete.
T h e Atomic Energy Commission had been for years the pri-
mary source of supporting funds for construction of acceler-
ator facilities and high-energy physics research in the United
States. T h e Commission was kept informed of the planning for
higher energies in the several laboratories and encouraged the
exchange of ideas with scientists abroad. In November 1962 a
special panel was appointed jointly by the General Advisory
Committee of the AEC and the President's Scientific Advisory
Committee, to study the status of high-energy physics and to
recommend a program for the future. T h e Report 7 of this
panel (called the "Ramsey panel" after its chairman, Professor
N. F. Ramsey of Harvard University) was released on May 10,
1963. In addition to making general and extensive recommen-
dations for support of the existing high-energy physics pro-
gram, the Report made several specific suggestions regarding
the extension of facilities into the very-high-energy range. T h e
panel proposed a two-step approach, starting with early au-
thorization of construction for the 200-GeV proton accelerator
then being considered at the LRL, and continued support of
design studies at Brookhaven of a national accelerator in the
600- to 1000-GeV range to be authorized at a later date. It also
suggested development of plans for a proton-proton storage
ring at Brookhaven as an intermediate step toward the study
of higher-energy interactions. T h e recommendations of this
panel justified the AEC in implementing plans at L R L and
Brookhaven. Both laboratories were authorized to proceed
with their design studies starting in April 1963.
93] The 200-Ge V Accelerator
T H E L R L DESIGN STUDY
At the Lawrence Radiation Laboratory, authorization of
the 200-GeV design study in 1963 resulted in a major effort ex-
tending over four years with an average professional staff of 35
persons. E . J . Lofgren headed the study and Lloyd Smith led
the theoretical section. T h e culmination was an Interim Re-
port presented to the AEC in December 1964, followed by a
"Design Study for a 200-GeV Accelerator" 8 in J u n e 1965. T h e
design study covered the scientific, technical, and engineering
features of the accelerator and the associated facilities, and in-
cluded a preliminary engineering cost estimate and time
schedule for completion. T h e cost of the facility plus basic ex-
perimental equipment was estimated to be $350 million, with
a continuing annual operations cost of $50 to $60 million.
T h e L R L "Design Study" was a description of a single in-
tegral design that was feasible, with realistic cost estimates.
During the study many alternative concepts were considered
and evaluated. Although a fixed set of parameters was chosen,
it was recognized that changes and improvements could be ex-
pected with further developments. T o make the cost estimates
meaningful, a single site was selected, in the Sierra Nevada
foothills above Sacramento. It was conceived that the acceler-
ator would be designed and its construction supervised by an
expanded staff coming primarily from the LRL, but after
completion it would become a National Laboratory available
to all qualified scientists.
Following publication of the L R L "Design Study" and dur-
ing the following year before the AEC announced the selection
of another site for such an installation, work continued on pos-
sible improvements, optimization of parameters and refine-
ment of cost estimates. A summer study held in 1966 explored
further the instruments and facilities needed for experimental
use. Work went on at Berkeley for two more years, until the
National Accelerator Laboratory was organized and took over

the task.
Meanwhile, designs were in progress in other laboratories.
At Brookhaven the design study for an accelerator in the 600-
to 1000-GeV range continued, but on a considerably lower
scale of effort than at LRL. Emphasis was on analysis of feasi-
bility and general parameters for these very high energies,
without much engineering detail or cost estimating. A sum-
mer study held at Brookhaven in 1963 was attended by a large
number of physicists and accelerator experts from this country
and abroad. This study showed that there was considerably
more interest in increasing intensity of the existing AGS than
in a more detailed design study for superhigh energies or for a
set of proton storage rings to utilize the 30-GeV protons for
beam-beam interactions. A program for conversion of the AGS
to produce ten times higher intensities was initiated, and soon
authorized by the AEC.
At the CERN Laboratory in Geneva, a group of design ex-
perts from the CPS division started a design study as early as
1961, with alternative initial goals of 150- and 300-GeV accel-
erators, and continued for the next 3 to 4 years. By the time
United States planning began to focus around 200-GeV en-
ergy, the CERN planning concentrated on the 300-GeV ma-
chine. A design study was published 9 in 1964, and a proposal
was submitted to the CERN Council for further planning and
negotiation between the member States. A Committee was ap-
pointed by the Council to initiate studies of possible sites for a
300-GeV accelerator in western Europe.
During these years exchange of ideas continued between the
design groups at LRL, Brookhaven, CERN (and interested
individuals elsewhere). A series of meetings took place at ap-
proximately 6-month intervals between representatives of
these groups. A few exchange visits of extended duration by
group members maintained continuity in this collaborative ef-
fort. And exchange of ideas continued with Soviet scientists at
international meetings, with the possible goal of an interna-

tional accelerator for the superhigh-energy range.
T h e general conclusion from these design studies was that
the basic principle of the AG proton synchrotron could be ex-
tended with certainty to the 200- or 300-GeV range or even
considerably higher. Peak energy would be set by orbit radius
for conventional iron-cored magnets, and would probably be
determined by economic rather than by technical considera-
tions. Another consequence of the use of very large orbits was
the availability of very high beam intensities per pulse, owing
to the large orbit circumference. Intensities of the order of 1013
protons per second could be expected, with beam power ap-
proaching 1 megawatt. T h e higher-intensity advantage of
other techniques, such as use of fixed-field alternating gradi-
ents (FFAG) or high-duty-cycle linear accelerators, was to a
great extent compensated by the large orbits, in the hundred-
GeV range. Design cost estimates varied among the several
laboratories, based on different engineering styles, but unit
costs were in the range of $100 million to $150 million per 100
GeV energy.
S C I E N T I F I C J U S T I F I C A T I O N AND G O V E R N M E N T POLICY
T h e studies at LRL, Brookhaven, and other laboratories
also included surveys of the purposes of high-energy physics
and theoretical justifications for higher energies. Considerable
effort went into preliminary planning and feasibility studies
of beam-separation and beam detection equipment in the
hundred-GeV range. In December 1964, a Brookhaven Re-
port 10 edited by L. C. L. Yuan presented statements by about
25 leading theoretical scientists in the field. These statements
were unanimously favorable, and even urgent, in their ad-
vocacy of the need for new accelerators in the higher energy
range.
In the spring of 1964, the National Academy of Sciences-
National Research Council established a physics survey com-
mittee, under the chairmanship of George E. Pake, to study

future requirements in relation to national needs in physics
(and other fields of science). A subpanel on elementary particle
physics, Robert Walker, chairman, brought in a report which
was basically in accord with the conclusions of the Ramsey
panel, but recommended that future high-energy accelerators
be considered national rather than regional facilities. T h i s re-
port of the Pake committee was instrumental in developing
national policy in the field of high-energy physics. It made
basic recommendations which were repeated in a policy paper
prepared within the AEC. T h e most significant result was the
identification of future high-energy accelerators as national
rather than regional facilities, with the implication that
choice of site was an important aspect of this function. In sub-
sequent actions the A E C turned down the Midwest Universi-
ties Research Association ( M U R A ) proposal for a regional
F F A G accelerator in the midwest designed to produce very
high beam intensities at less than 100-GeV energy. T h e L R L
proposal for a 200-GeV machine to be located near and oper-
ated by the University of California was reconsidered in the
light of its significance as a national facility.
A study paper on "Policy for National Action in the Field of
High Energy Physics," referred to above, was prepared by
members of the A E C Research Division, and became the basis
for a U.S. J o i n t Committee P r i n t 1 1 published in February
1965. T h i s policy statement reviewed progress in high-energy
physics and high-energy accelerators, summarized the needs
for higher energy and higher intensity, and made specific rec-
ommendations for government action. In particular, it noted
the need for new high-energy facilities for the large number of
user-scientists within universities. It recommended a two-step
approach, starting with a 200-GeV machine similar to that de-
signed at L R L to be located at a site convenient to a nation-
wide distribution of users, to be followed in the 1970's by a ma-
chine of about 1000-GeV energy.
Hearings on the high-energy research program were held

before the Subcommittee on Research, Development and Ra-
diation of the Joint Committee on Atomic Energy of the
United States Congress. Many scientists from government lab-
oratories and from universities were called as witnesses to
present their views. T h e L R L and Brookhaven design studies
were reported, and summaries entered in the testimony. High-
energy physicists described the scientific justification and
Yuan's collection 10 of essays by theorists was presented. AEC
and other government officials presented program plans and
budget estimates for the future. A full report of the hearings
was published. 12
A general conclusion from these hearings was that the scien-
tists had made a strong case. Government officials seemed per-
suaded that a national facility for research in the multi-
hundred-GeV range was well justified.
UNIVERSITIES RESEARCH ASSOCIATION
A group of leading scientists from the National Academy of

Sciences began in early 1965 to consider the problems of man-
agement of the large scientific facility required to support re-
search operations of a large multi-GeV accelerator. T h e pur-
pose was to provide university backing and support with a
broad national basis to the planning and management of the
proposed 200-GeV laboratory and the subsequent research
program. Frederick Seitz, President of the National Academy
of Sciences, acted as coordinator of scientific opinion in calling
a meeting of university presidents, first in January and again
in J u n e 1965. T h e outcome of these meetings was the organiza-
tion of the Universities Research Association (URA), consist-
ing originally of 34 universities distributed across the United
States which carry on significant research programs in the
physical sciences.
T h e URA is incorporated in the District of Columbia and
maintains its principal office at 2100 Pennsylvania Avenue,
N.W., Washington, D.C. T h e Council of Presidents of the

member universities meets about once a year. T h e president of
URA is Professor Norman Ramsey of Harvard University. Ac-
tive management is placed in a Board of Trustees elected by
the Council of Presidents. T h e Trustees include a representa-
tive from each of 15 regional groups of universities and 6 at
large to represent the public interest. Each member university
agreed to contribute up to $100,000 to URA if and when called
upon by the Trustees, primarily for the expenses of organiza-
tion and operation of URA. By January 1968 the number of
member universities had grown to 48.
T h e URA offered its services to the AEC as a management
organization to contract with the AEC and to operate the 200-
GeV facility. T h e organization was not involved in selection of
the site for the laboratory, but indicated willingness to under-
take construction and operation at any site selected by the fed-
eral government.
SELECTION OF T H E SITE
T h e choice of Weston, Illinois (35 miles west of Chicago) as
the site for the 200-GeV laboratory was made by the Atomic
Energy Commission after extensive site-selection studies. 13
T h e search for the site started in April 1965, when the AEC
issued a press release inviting statements of interest in propos-
ing sites for this new scientific facility. A total of 125 proposals
were ultimately received, relating to more than 200 different
locations, one or more from each of 48 states. By September
1965 the Commission had reduced the list to 85 proposals re-
lating to 148 sites. T h e AEC then requested the National
Academy of Sciences to enlist a site evaluation committee com-
posed of eminent scientists to review and evaluate the site pro-
posals and make recommendations to the AEC. T h e chairman
of this committee was E. R. Piore. T h e committee visited and
studied the sites which met the basic selection criteria, and in
March 1966 reported to the Commission recommending six
sites as of equal merit. As part of the final evaluation effort,

three commissioners visited each of the six recommended sites.
T h e Commission announced its unanimous decision for the
site at Weston, Illinois, on December 16, 1966.
As was to be expected, the selection of a site was not popular
with the proponents of many other sites. T h e choice caused
considerable discussion in the Congress and in Joint Commit-
tee Hearings. 13 Much of the discussion centered on the avail-
ability of open housing in the Chicago suburban area, a prob-
lem raised by the National Committee against Discrimination
in Housing and other civil rights groups. In the testimony, the
state of Illinois was called upon to enact fair-housing legisla-
tion and suburban communities were importuned to the same
effect. In partial answer, the Commission made clear its basic
adherence to a policy of nondiscrimination.
In the hearings, the AEC proposed a reduced-scope acceler-
ator dictated by the Bureau of the Budget for budgetary rea-
sons. In this reduced scope the energy was to be retained at 200
GeV but the intensity could be decreased initially to about one-
tenth that of the L R L design and the number of experimental
target stations reduced. An initial cost figure of about $240
million was suggested, with the expectation that additional
funds would be provided later to recover the original scope.
T h e general result of these hearings was congressional au-
thorization to the AEC to proceed with the planning for a 200-
GeV accelerator, but at the lower initial cost. T h e first step was
the execution of a contract with the URA to initiate a design
study to accomplish this purpose, which was signed on Jan-
uary 23, 1967.
ESTABLISHMENT OF T H E NATIONAL
ACCELERATOR LABORATORY
Following selection of the Weston site, the URA moved
promptly to select a director and initiate activities. T h e first
scientist asked to direct the design study declined, and Profes-
sor Robert R a t h b u n Wilson of Cornell University was offered

the position of director of the National Accelerator Labora-
tory (NAL). He announced his intention to prepare a design
for high intensity with a higher-energy option within one year
at a cost not to exceed $240,000; but he required significant
concessions of authority from the AEC and URA and a fast
time schedule (5 years) for construction. These conditions were
acceptable. Wilson accepted the position as director on March
7, 1967; he took up the position full time on J u n e 15, when he
moved to Chicago.
President Ramsey called a meeting of potential scientific
users of the 200-GeV accelerator for April 7-8, 1967 at the
Argonne National Laboratory, at which time announcements
were made of the organization of NAL and the appointment
of Wilson as director. Wilson used this opportunity to call a
special meeting of accelerator scientists attending the meeting
and announced a summer design program to start on June 15
in the Chicago area.
In April, a preliminary contract for architectural-
engineering services was signed with the firm of "DUSAF," *
and their engineers started meeting with the director and his
associates. They were asked to perform site surveys and make
preliminary site plans (Fig. 32). Also in April, arrangements
were made for temporary office space in Oak Brook, Illinois.
By May a number of scientific and administrative staff
members had been recruited and accepted appointments, in
most cases to start at some later date. These included E. L.
Goldwasser (University of Illinois) as Deputy Director, F. T .
Cole (LRL), A. L. Read (Cornell), J. DeWire (Cornell), M. S.
Livingston (CEA), D. E. Young (University of Wisconsin),
C. D. Curtis (University of Wisconsin), and on the administra-
tive side Donald Getz and Donald Poillon.
* DUSAF is a consortium of four firms: Daniel, Mann, J o h n s o n , and

Mendenhall; Max O. U r b a h n ; Seelye, Stevenson, Value, and Knecht; and
George A. Fuller.
FIG. 32. Site plan for the 200-400 GeV*accelerator to be built by the
National Accelerator Laboratory at Weston, Illinois.
Meetings were held at an increasing rate during the spring

of 1967: of the URA Trustees, of a Scientific Advisory Commit-
tee appointed by the director, with representatives of DUSAF,
and with accelerator design experts from many laboratories.
T h e AEC took a leading role in planning and in solving ad-
ministrative problems, and their representatives attended all
planning meetings and discussions. All obstacles were cleared
away, offices for the summer program were prepared, contracts

for financial support by URA and the AEC were formally
signed, and commitments were obtained from accelerator sci-
entists from many sources to attend and contribute to the sum-
mer program.
S U M M E R DESIGN P R O G R A M OF 1967
T h e purpose of the 1967 summer program was to develop
concepts for a new design for a 200-GeV accelerator, signifi-
cantly simpler and of lower cost than that proposed in the
L R L design report. T h e leader in this conceptual study was
the new director, R. R. Wilson, who brought to bear his recent
experience in building the Cornell 10-GeV AG electron syn-
chrotron at a lower cost per GeV than previous electron
synchrotrons of this type. Some of these simpler and lower-cost
concepts became part of the new design from the start, such as
a compact magnet structure which embodied a girder-type
support in its construction, and a minimal-size magnet tunnel
enclosure without overhead crane or magnet foundation piers.
Other simplifying concepts were provided by visiting acceler-
ator experts and scientists. Several significant alternative solu-
tions were developed and presented by visitors from LRL,
which had several months' lead in searching for cost reduc-
tions following the AEC requirement of a reduced-scope accel-
erator discussed in the hearings in February. T h e theme of the
study became a search for new and different solutions to design
problems, with the emphasis on reduction of cost without ex-
cessive loss of quality or reliability.
T h e program started on J u n e 15, 1967, with 20 accelerator
scientists attending the first week, including Wilson and
Goldwasser. Additional members arrived later and others
came for a few weeks, with an average attendance of about 25.
Credit for this development should be distributed among the
large number of scientists and engineers who contributed. A
total of 63 persons were involved in the study during the sum-
mer and fall, of whom 30 eventually joined the NAL staff.
T h e more significant features considered and ultimately ac-

cepted in the new design deserve to be identified individually:
(a) Option for Higher Energy. Well before the congres-
sional hearings in January-February 1967, it was becoming
evident that the two-step program (for a 200-GeV machine to
be followed by a 600- to 1000-GeV machine) might postpone
too long the attainment of truly high energies. Many scientists
began to regret their commitment to 200 GeV and wished for
higher energies. Also, the reduced scope specified by the AEC
required rethinking about the energy limitations of a ring of
magnets limited by saturation of the iron. One concept for re-
ducing initial cost (originated by Garren, Lambertson, Lof-
gren, and Smith at LRL) consisted of filling half the ring with
magnets, with an initial operating energy that could be in-
creased later by the addition of other magnets; it was called the
"expanditron" in laboratory slang. Prior to the summer pro-
gram, this concept was modified to consider use of an orbit
filled with magnets but initially powered at half field to reduce
cost, with additional power supplies to be added later; this
procedure would minimize the downtime required for a future
conversion. Wilson adopted this concept and extended it to
imply an orbit size capable of ultimate powering to reach 400
or even 500 GeV but initially powered for 200-GeV operation.
He hoped that this option for higher energy could be included
within the $240 million budget. T h e concept was discussed
with enthusiasm as early as January 1967, in local groups and
at U R A meetings. It typified the basic theme of the new ap-
proach to design and had great appeal at AEC and congres-
sional levels.
(b) Separated-Function Magnets. Alternating-gradient ac-
celerator designers had long known of the option of separating
the bending and focusing functions of the AG magnets in the
main ring. For the large orbit (radius 1 km) required for 400
GeV the focusing properties could be provided by relatively
short and widely spaced quadrupole magnets, allowing most
of the orbit to be filled with bending magnets which would
have u n i f o r m and very high fields, flat poles, a n d simplified

construction. T h i s separated-function design significantly re-
duced estimated costs for the magnet system.
(c) Minimum Magnet Enclosure. Experience at electron
synchrotron laboratories such as CEA a n d Cornell which used
small AG magnets has increased confidence that installation
a n d m a i n t e n a n c e can be accomplished in a small tunnel with-
out an overhead crane. I n typical maintenance procedures,
complete magnet units (20 ft long) will be replaced by means
of special h a n d l i n g vehicles. T h e desire to minimize mainte-
nance time a n d reduce radiation exposure results in removing
m u c h of the auxiliary e q u i p m e n t from the tunnel a n d into
separate e q u i p m e n t galleries. T h e s e plans justified a simply
constructed and minimum-size tunnel, formed of precast con-
crete sections of 10-ft width (Fig. 33). T h e cost savings in de-
sign estimates were considerable.
(d) Magnet Foundations. W i t h the development toward
A G magnets of small cross section and light weight, confidence
has increased a m o n g accelerator designers that concrete piers
for the magnet are not necessary. Experience shows that mag-
nets can be aligned with the aid of position monitors for the
beams. As a consequence, magnet piers were eliminated from
the design a n d the magnets are to be m o u n t e d directly on the
slab floor of the tunnel. Although there were many critics f r o m
outside the N A L group, this concept prevailed.
(e) Fast-Cycling Booster. All early plans for a m u l t i h u n d r e d -
GeV machine divided the acceleration into several steps, with
one or more preaccelerator stages feeding into the main ring.
A multi-GeV linear accelerator has the obvious advantage of a
linear beam for injection into the ring, b u t has an undesirably
high cost per GeV. T h e p r o b l e m with a smaller injector syn-
chrotron was the uncertain efficiency of ejection of the preac-
celerated beam and the difficulty of injecting it into the m a i n
ring. Success in the ejection of high-quality emergent beams
from the AGS and PS synchrotrons d u r i n g the m i d d l e 1960's
FIG. 33. Full-scale mock-up of the tunnel enclosure and a magnet sector
for the 2 0 0 - 4 0 0 GeV accelerator.
solved this problem and made it possible to plan for the use of
a 10-GeV "booster" synchrotron as an injector into the 200-
GeV accelerator. T h e booster is operated at a fast cycling rate
so that successive booster output pulses can be used to fill the
main ring completely. This concept has been adopted at the
NAL, with the booster operating at 15 pulses per second and a
sequence of 13 pulses (0.8 sec) used to fill the main ring. T h e
result is a high-intensity circulating beam in the main ring, at
the cost of a slightly lower cyclic repetition rate (15 to 20 per
minute) than might otherwise be achieved.
(J) Long Straight Sections. T h e need for relatively long

straight sections for ejection of an emergent 200- to 400-GeV
beam was obvious at the start of the design study. T h e ac-
cepted technique is to use m a t c h i n g q u a d r u p o l e pairs at the
ends of each straight section. Six long straights each with 160 ft
of free space are planned, to be used for injection, ejection, a n d
f u t u r e options.
(g) Vertical Injection and Ejection. T h e compact magnet
design, with small vertical height above a n d below the b e a m
has offered an o p p o r t u n i t y of injecting the beam from the
linac into the booster in the vertical rather than the horizontal
plane. T h e same principle is used for ejection f r o m the booster
a n d injection into the m a i n ring. O r b i t analysis shows that
such vertical injection gives somewhat larger beam acceptance
in the m a i n ring. It also removes injection e q u i p m e n t from the
radial plane where it m i g h t be damaged by radiation or in-
duced radioactivity p r o d u c e d by the m a i n beam.
(h) Single Emergent Proton Beam. A n u m b e r of members
of the s u m m e r p r o g r a m were experimental high-energy physi-
cists who devoted their efforts to the p l a n n i n g of experiments
a n d the s u p p o r t i n g facilities. Originally, several emergent-
beam facilities h a d been p l a n n e d , f a n n i n g out f r o m successive
long straight sections. T h e concept of using a single emergent
beam, with switching magnets to utilize three or more target
stations, was originally introduced as a means of reducing ini-
tial costs. However, as the study progressed, excellent reasons
were developed to justify such a single emergent beam r u n as
the total external beam facility. W i t h one extended beam
path, communications and transportation of e q u i p m e n t be-
come simpler, and ejection efficiency can be m a d e higher, t h a n
with several shorter paths. T h e location of the main r i n g a n d
emergent-beam r u n on the site were arranged to maximize ac-
cessibility to the experimental facilities.
(i) Options for the Future. T h e m a j o r f u t u r e option is that
for increasing power to operate at 400- or even 500-GeV energy.
It was also considered essential to include the option of adding

a colliding-beam facility in the future. T h e main-ring location
and the single emergent-beam run allowed plans for a future
storage-ring facility to be fed from other, unused long straight
sections. Options include the use of bypass sectors external to
the main ring and storage rings of either small or large diame-
ter.
By September 15 most of the basic concepts of the new de-
sign had been crystallized, at least in principle. Preliminary
descriptions and cost estimates were prepared and circulated
to critics in other laboratories; several discussion sessions were
held at which criticisms were considered and changes made
when they proved valid. A number of critics remained unsatis-
fied, but no alternative concepts were proposed that would not
significantly increase the estimated costs. As director, Wilson
made a sequence of basic decisions freezing one by one the
major concepts and parameters. Visitors from other laborato-
ries returned to their home bases; NAL staff members were en-
listed and joined the continuing staff. T h e total staff at design
headquarters during the fall did not drop below the 25 aver-
aged during the summer program.
Activities initiated during the summer continued through
the fall, with increasing emphasis on more detailed designs,
improved parameters, and improved cost estimates. T h e labo-
ratory organization took form and business and clerical staff
were added. A machine shop was initiated and installed in
quarters in Downers Grove, 5 miles from Oak Brook, and ad-
ditional office space was leased. T h e professional staff grew to
30 by December 30, with 8 appointees to start after the first of
the year. Total laboratory staff, including offers outstanding,
was 90 on December 30.
One initial deadline for the laboratory was preparation of
the AEC Construction Data Sheet (Schedule 44), to be submit-
ted by October 15, 1967. This represented the final and com-
plete cost estimate for construction. It was prepared and
submitted on time, with a total construction-cost estimate of

$242 million. As a result of schedule revisions required by
budgetary limitations, the AEC changed and rounded off the
estimated cost to $250 million.
T h e final product of the design program was a design report
to the AEC presenting the results of the study, detailed param-
eters and justifications, and a cost estimate. T h e original date
requested by the AEC Research Division for this report was
December 15. On that date a final copy of the "200 BeV Accel-
erator, Parameters and Specifications" 14 was completed; addi-
tional copies were available for general distribution before
January 15, 1968. This design report was submitted to the
Joint Committee on Atomic Energy and was incorporated in
the AEC Authorizing Legislation for fiscal 1969, published in
a U.S. Government P r i n t 1 5 in February 1968. T h e bill autho-
rizing start of construction of the 200-GeV accelerator was
passed by the Congress and was signed into law by the Presi-
dent on April 19, 1968.
[Appendix]
Chronology of Development of Particle Accelerators
T h i s listing gives n a m e s of investigators o r g r o u p s , d a t e s of n e w

c o n c e p t s a n d of first o p e r a t i o n s of n e w types, a n d n e w energy rec-
ords.
PREHISTORY
1919—1921 S c h e n k e l , G r e i n a c h e r ; v o l t a g e - m u l t i p l i e r circuit.
1924 Ising; p r o p o s a l of electron linac w i t h d r i f t tubes, spark-gap
excitation.
1925 Sorensen; cascade t r a n s f o r m e r at 750 kV for ac electrical test-
ing.
1928 L a u r i t s e n ; cascade t r a n s f o r m e r for X-rays at 750 kV.
W i d e r ö e ; r a d i o f r e q u e n c y r e s o n a n c e w i t h heavy ions in two-
s t e p linac.
1930 Brasch a n d L a n g e ; s u r g e g e n e r a t o r , 2.4 MV, d i s c h a r g e in
chamber.
T u v e et al.; Tesla-coil r e s o n a n c e t r a n s f o r m e r , 1—2 MV.
V a n d e G r a a f f ; electrostatic g e n e r a t o r , 1.5 MV.
1931 L a w r e n c e a n d L i v i n g s t o n ; d e m o n s t r a t i o n of cyclotron reso-
nance.
Sloan a n d L a w r e n c e ; l i n e a r accelerator, 1.2-MeV H g ions.
1932 Bellaschi; s u r g e g e n e r a t o r , 6 MV, for ac testing.
U r b a n et al.; t h u n d e r s t o r m p o t e n t i a l s in Alps.
EARLY HISTORY
1932 C o c k c r o f t a n d W a l t o n ; voltage m u l t i p l i e r , 0.5-MeV p r o t o n s ;
disintegration.
L a w r e n c e a n d L i v i n g s t o n ; cyclotron, 1.2-MeV p r o t o n s ; disin-
tegration.
1933 T u v e , H a f s t a d , a n d D a h l ; electrostatic g e n e r a t o r , 0.6-MeV
protons.
1934 L a w r e n c e a n d L i v i n g s t o n ; cyclotron, 5.0-MeV d e u t e r o n s .
T u v e , H a f s t a d , a n d D a h l ; electrostatic g e n e r a t o r , 1.2-MeV
protons and deuterons.
1935 H e r b et al.; p r e s s u r e electrostatic g e n e r a t o r , 0.7-MeV p r o t o n s .
1936 L a w r e n c e a n d Cooksey; cyclotron, 8.0-MeV d e u t e r o n s .
109
110] Appendix: Chronology
1938 Herb et al.; pressure electrostatic generator, 2.7-MeV protons

and deuterons.
Bethe and Rose; relativistic limitation of cyclotron.
Thomas; proposal of sector focusing for cyclotron.
1939 Lawrence et al.; 60-in. cyclotron, 16-MeV deuterons, 32-MeV
He ions (later 20-MeV deuterons, 40-MeV He ions).
1940 Kerst; betatron, 2.3-MeV electrons.
1942 Kerst et al.; betatron, 20-MeV electrons.
R E C E N T HISTORY (HIGH-ENERGY ACCELERATORS)

1945 Veksler; synchronous-accelerator proposal.
McMillan; synchronous-accelerator proposal.
1946 Goward and Barnes; electron synchrotron, 8-MeV electrons.
Lawrence et al.; 184-in. synchrocyclotron, 200-MeV deuterons,
400-MeV He ions.
1947 Lawson et al.; electron synchrotron, 70-MeV electrons.
Hansen et al.; Stanford Mark I linac, 6-MeV electrons.
1949 McMillan; electron synchrotron, 320-MeV electrons.
1950 Stanford Mark II linac, 35-MeV electrons.
Kerst; betatron, 300-MeV electrons.
Alvarez et al.; linear accelerator, 32-MeV protons.
1952 Brookhaven National Laboratory; proton synchrotron (cos-
motron), 2.2-GeV (later 3.0-GeV) protons.
Courant, Livingston, and Snyder; proposal of alternating-
gradient focusing.
1953 Wilson et al.; Cornell AG electron synchrotron, 1.2-GeV elec-
trons.
1954 Lawrence Radiation Laboratory; proton synchrotron (beva-
tron), 6.2-GeV protons.
1957 Lawrence Radiation Laboratory; synchrocyclotron (rebuilt),
720-MeV protons.
Dubna, USSR; synchrophasotron (proton synchrotron), 10-
GeV protons.
1959 CERN; AG proton synchrotron, CPS, 28-GeV protons.
1960 Brookhaven National Laboratory; AG proton synchrotron
AGS, 33-GeV protons.
Stanford Mark IV linac, 1.0-GeV electrons.
1961 High Voltage Engineering Co.; tandem Van de Graaff gen-
erator, 12 MeV.
1962 Argonne National Laboratory; ZGS proton synchrotron, 12.5
GeV.
Ill] Appendix: Chronology
Livingston et al.; Cambridge A G electron synchrotron, 6-GeV

electrons.
1966 Stanford 2-mile electron linear accelerator, S L A C , 20 GeV.
1967 Cornell; A G electron synchrotron, 10-GeV electrons.
Serpukov, U S S R : A G proton synchrotron, 76-GeV protons.
National Accelerator Laboratory; start o f 2 0 0 - 4 0 0 - G e V ac-
celerator design.
1930 1940 1950 I960 1970
FIG. 34. Energies achieved by accelerators from 1932 to 1968. The linear
envelope of the individual curves shows an average tenfold increase in
energy every 6 years.
References
C H A P T E R 1. T H E RACE FOR H I G H VOLTAGE

1. E. R u t h e r f o r d , Proc. Roy. Soc. (London) A117, 300 (1927).
2. P. L. Bellaschi, Trans. Am. Inst. Elee. Engrs. 51, 936 (1932).
3. A. Brasch a n d F. L a n g e , Naturwiss. 18, 769 (1930); Z. Physik
70, 10 (1931).
4. R . W . S o r e n s e n J . Am. Inst. Elee. Engrs. 44, 373 (1925).
5. R . C r a n e a n d C. C. L a u r i t s e n , Rev. Sci. Instr. 4, 118 (1933).
6. M. A. T u v e , L. R. H a f s t a d , a n d O . D a h l , Phys. Rev. 35, 1406;
36, 1261 (1930).
7. D. H . S l o a n , Phys. Rev. 47, 62 (1935).
8. E. E. C h a r l t o n , W . F. W e s t e n d o r p , L. E. D e m p s t e r , a n d G.
H o t a l i n g , J. Appi. Phys. 10, 374 (1934).
9. H . G r e i n a c h e r , Z. Physik 4, 195 (1921).
10. G. G a m o w , Ζ. Physik 52, 510 (1929).
11. E. U. C o n d o n a n d R . W . G u r n e y , Phys. Rev. 33, 127 (1929).
12. J . D . C o c k c r o f t a n d E. T . S. W a l t o n , Proc. Roy. Soc. (Lon-
don) A136, 619 (1932); A137, 2 2 9 ( 1 9 3 2 ) ; A144, 704 (1934).
13. R . J . V a n d e G r a a f f , Phys. Rev. 38, 1919A (1931).
14. R . J . Van d e G r a a f f , Κ. T . C o m p t o n , a n d L. C. Van A t t a ,
Phys. Rev. 43, 149 (1933).
15. L. C. V a n A t t a , D. L. N o r t h r u p , R . J . V a n d e G r a a f f , a n d
C. M. V a n A t t a , Rev. Sci. Instr. 12, 534 (1941).
16. M. A. T u v e , L. R. H a f s t a d a n d O . D a h l , Phys. Rev. 48, 315
(1935).
17. L. R. H a f s t a d a n d M. A. T u v e , Phys. Rev. 47, 506 (1935); 48,
306 (1935).
18. R . G. H e r b , D. B. P a r k i n s o n , a n d D . W . Kerst, Rev. Sci.
Instr. 6, 261 (1935).
19. D. B. P a r k i n s o n , R. G. H e r b , E . J . B e r n e t , a n d J . L. McKib-
b e n , Phys. Rev. 53, 642 (1938).
20. R . G. H e r b , C. M. T u r n e r , C. M. H u d s o n , a n d R . E. W a r r e n ,
Phys. Rev. 58, 579 (1940).
21. J . G . T r u m p a n d R . J . V a n d e G r a a f f , Phys. Rev. 55, 1160
(1939).
22. R . W i d e r ö e , Arch. Elektrotech. 21, 387 (1928).
113
114] References
23. E. O. Lawrence and Ν. E. Edlefsen, Science 72, 376 (1930).

24. M. S. Livingston, " T h e production of high-velocity hydrogen
ions without the use of high voltages," Ph.D. thesis, University of
California, April 14, 1931.
25. E. O. Lawrence and M.S. Livingston, Phys. Rev. 40, 19
(1932).
26. E. O. Lawrence, M. S. Livingston, and M. G. White, Phys.
Rev.42, 1950(1932).
27. E. O. Lawrence and M. S. Lvingston, Phys. Rev. 45, 608
(1934).
28. E. O. Lawrence, L. W. Alvarez, W. M. Brobeck, D. Cooksey,
D. R. Corson, E. M. McMillan, W. W. Salisbury, and R. L. Thorn-
ton, Phys. Rev. 56, 124 (1939).
29. D. W. Kerst, Phys. Rev. 58, 841 (1940).
30. D. W. Kerst, G. D. Adams, H. W. Koch, and C. S. Robinson,
Rev. Sci. Instr. 21, 462 (1950).
C H A P T E R 2. E R N E S T L A W R E N C E A N D T H E CYCLOTRON
1. E. O. Lawrence and N. E. Edlefsen, "On the production of
high speed protons," Science 72, 376 (1930).
2. R. Wideröe, Arch. Elektrotech. 21, 387 (1928).
3. E . O . Lawrence and M.S. Livingston, Phys. Rev. (A) 37, 1707
(1931).
4. M. S. Livingston, "The production of high-velocity hydrogen
ions without the use of high voltages," Ph.D. thesis, University of
California, April 14, 1931.
5. E . O . Lawrence and M.S. Livingston, Phys. Rev. (L) 38, 834
(1931); E. O. Lawrence and M. S. Livingston, Phys. Rev. (A) 38, 862
(1931); E . O . Lawrence and M.S. Livingston, Phys. Rev. 40, 19
(1932).
6. J. D. Cockcroft and E. T . S. Walton, Proc. Roy. Soc. (London)
A137, 229 (1932).
7. E. O. Lawrence, M. S. Livingston and M. G. White, Phys. Rev.
(L) 42, 150 (1932).
8. M.S. Livingston, Phys. Rev. (L) 42, 441 (1932); M.S. Living-
ston and E . O . Lawrence, Phys. Rev. (A) 43, 212 (1933); M.S. Liv-
ingston, Phys. Rev. (A) 43, 214 (1933); E . O . Lawrence and M.S.
Livingston, Phys. Rev. 45, 608 (1934).
9. M.S. Livingston, M. C. Henderson and E . O . Lawrence,
"Neutrons from deutons and the mass of the neutron," Phys. Rev.
(L) 44, 782 (1933).
115] References
10. M. C. H e n d e r s o n , M. S. L i v i n g s t o n a n d Ε. O. L a w r e n c e , "Ar-
tificial radioactivity p r o d u c e d by d e u t o n b o m b a r d m e n t , " Phys.
Rev. (L) 4 5 , 4 9 7 (1934).
11. E. O. L a w r e n c e a n d D. Cooksey, Phys. Rev. 50, 1131 (1936).
12. E. O. L a w r e n c e , L. W . Alvarez, W . M. Brobeck, D. Cooksey,
D. R . Corson, E. M. M c M i l l a n , W . W . Salisbury a n d R . L. T h o r n -
ton, Phys. Rev. 56, 124 (1939).
CHAPTER 3. SYNCHRONOUS ACCELERATORS AND HOW

THEY GREW
1. V. V e k s l e r , J . Phys. (U.S.S.R.) 9, 153 (1945).
2. E. M. M c M i l l a n , Phys. Rev. 68, 143 (1945).
3. E. M. M c M i l l a n , Phys. Rev. 68, 144 (1945).
4. F. K. G o w a r d a n d D . E . Barnes, Nature (London) 158, 413
(1946).
5. F. R. E l d e r , A. M. G u r e w i t s c h , R . V. L a n g m u i r , a n d H . C.
P o l l o c k , / . Appi. Phys. 18, 810 (1947).
6. D. B ö h m a n d L. Foldy, Phys. Rev. 70, 249 (1946).
7. H . A. B e t h e a n d M. E. Rose, Phys. Rev. 52, 1254 (1937).
8. L. H . T h o m a s , Phys. Rev. 54, 580, 588 (1938).
9. J . R. R i c h a r d s o n , K. R . M a c k e n z i e , E . J . L o f g r e n , a n d B. T .
W r i g h t , Phys. Rev. 69, 699 (1946).
10. W . M. Brobeck, E. O. L a w r e n c e , K. R . M a c k e n z i e , E. M. Mc-
M i l l a n , R . S e r b e r , D . C . Sewell, Κ. M. S i m p s o n , a n d R . L. T h o r n -
ton, Phys. Rev. 71,449 (1947).
11. M. L. O l i p h a n t , J . S. G o o d e n a n d G. S. H y d e , Proc. Roy. Soc.
(London) 59, 666 (1947); J . S. G o o d e n , H . H . J e n s e n , a n d J . L. Sy-
m o n d s , Proc. Roy. Soc. (London) 59, 677 (1947).
12. W . M. Brobeck, Rev. Sci. Instr. 19, 545 (1948).
13. M . S . Livingston, Phys. Rev. 73, 1258 (1948).
14. M . S . L i v i n g s t o n , J . P. Blewett, G. K. G r e e n , a n d L . J . Ha-
w o r t h , Rev. Sci. Instr. 21, 7 (1950).
15. E . J . L o f g r e n , Science 111, 295 ( 1950).
16. G. Ising, Arkiv Mat. Astron. Fysik. 18, 45 (1925).
17. J . W . B e a m s a n d L. B. Snoddy, Phys. Rev. 44, 784 (1933),
J . W . Beams a n d H . T r o t t e r , J r . , Phys. Rev. 45, 849 (1934).
18. D. H . Sloan a n d E. O . L a w r e n c e , Phys. Rev. 38, 2021 (1931).
19. D. H . Sloan a n d W . M. Coates, Phys. Rev. 46, 539 (1934).
20. E. L. G i n z t o n , W . W . H a n s e n a n d W . R . K e n n e d y , Rev. Sci.
Instr. 19, 8 9 ( 1 9 4 8 ) .
21. L i n e a r A c c e l e r a t o r Issue, Rev. Sci. Instr. 26 ( F e b r u a r y 1955).
116] References
22. V. Veksler, Compi. Rend. Acad. Sci. U.S.S.R. 43, 444 (1944);
44, 392 (1944); J. Phys. (U.S.S.R.), 9, 153 (1945).
CHAPTER 4. THE STORY OF ALTERNATING GRADIENT

FOCUSING
1. L. H. Thomas, Phys. Rev. 54, 580, 588 (1938).
2. E. D. Courant, M. S. Livingston, and H. S. Snyder, Phys. Rev.
88, 1190(1952).
3. J. P. Blewett, Phys. Rev. 88, 1197 (1952).
4. J. B. Adams, M. G. N. Hine, and J. D. Lawson, Nature (Lon-
don) 171,926 (1953).
5. N . C . Christofilos, "Focusing systems for ions and electrons
and applications in magnetic resonance particle accelerators"
(1950, unpublished).
6. E. D. Courant, M. S. Livingston, H. S. Snyder, and J. P.
Blewett, Phys. Rev. 91, 202 (1953).
7. "Design study for a 15-BeV accelerator," M.I.T. Laboratory
for Nuclear Science, Report No. 60 (June 30, 1953).
CHAPTER 6. THE 200-GeV ACCELERATOR

* Taken largely from M.S. Livingston, "Early History of the
200-GeV Accelerator," National Accelerator Laboratory Report,
NAL-12-0100 (June 18, 1968).
1. David L. Judd, "Development from 1952-60 at LRL of plan-
ning for a larger accelerator," UCRL Memo (September 29, 1960).
2. M. Sands, "A proton synchrotron for 300 GeV," MURA Re-
port 465 (August 1959).
3. R. R. Wilson, Science 133, 1602 (1961).
4. Brookhaven National Laboratory Report, BNL 772 (T-290)
(August 28, 1961).
5. California Institute of Technology—Design Study Reports:
CTSL-10, Matthew Sands, "A proton synchrotron for 300 GeV"
(September 1960); CTSL-11, Robert I. Hulsizer, "Magnet position-
ing problems" (October 20, 1960); CTSL-12, M. H. Blewett, "Pre-
liminary cost estimate" (October 20, 1960); CTSL-14, Matthew
Sands, "Injection criteria" (January 5, 1961); CTSL-15, Kenneth
W. Robinson, "A radiofrequency system" (January 19, 1961); CTSL-
16, R. L. Walker, "Beam transfer in the cascade synchrotron" (Jan-
uary 1961); CTSL-21, Α. V. Tollestrup, "RF synchronization dur-
ing beam transfer" (March 21, 1961); CTSL-24, Vincent Z. Peterson,
117] References
" V a c u u m r e q u i r e m e n t s " (April 11, 1961); CTSL-25, J o n Mathews,

"Effects of m a g n e t non-linearities on b e t a t r o n frequencies" (April
3, 1961).
6. D. L. J u d d a n d Lloyd Smith, "Summary of accelerator studies
since J u n e 1960 at L R L , " U C R L Memo (December 26, 1961).
7. Reference 12, A p p e n d i x 1.
8. "Design study for a 200-BeV accelerator," vols. I and II, U C R L
16,000 ( J u n e 1965).
9. "Design study of a 300-GeV p r o t o n synchrotron," vols. I and
II, C E R N A R / I n t . SG-15 ( N o v e m b e r 19, 1964).
10. " N a t u r e of m a t t e r — P u r p o s e s of high energy physics," ed.
L. C. L. Yuan, Brookhaven National Laboratory R e p o r t , B N L 888
(T-360) (January 1965).
11. J o i n t C o m m i t t e e on Atomic Energy, Print 42-613 (Govern-
m e n t P r i n t i n g Office, W a s h i n g t o n , D.C., February 1965).
12. H e a r i n g s — J o i n t C o m m i t t e e on Atomic Energy, Print 46-982
( G o v e r n m e n t P r i n t i n g Office, W a s h i n g t o n , D.C., March 2 - 5 ,
1965).
13. H e a r i n g s — J o i n t C o m m i t t e e on Atomic Energy, Print 76-329
( G o v e r n m e n t P r i n t i n g Office, W a s h i n g t o n , D.C., J a n u a r y - F e b r u a r y
1967).
14. "200-BeV accelerator, parameters a n d specifications," Na-
tional Accelerator Laboratory Design R e p o r t (January 1968).
15. H e a r i n g s — J o i n t C o m m i t t e e on Atomic Energy, Part 2,
P r i n t 90-281-0 ( G o v e r n m e n t P r i n t i n g Office, Washington, D.C.,
February 7 and 21, 1968).
Index
accelerator: electron linac, 53; lin- Bureau of the Budget, 99

ear, 54; magnetic resonance, 22,
29; synchronous, 39, 48; travel- California Institute of Technology
ing-wave, 54; 200-GeV, 74, 88, (Caltech), 2, 3, 46, 89, 90, 91
93, 96, 98, Design Study, 93, California, University of, 4, 13, 22,
108; 300-GeV, 74, 94; 400-GeV, 39, 44, 46, 52, 54, 67, 69, 73, 74,
103; 1000-GeV, 96 89, 90, 91; Hospital, 4
Adams, J. B., 69, 71 Cambridge Accelerator Project
Alvarez, Luis W., 55, 56 (CAP), 82, 83, 84
alternating-gradient focusing, 46 Cambridge Electron Accelerator,
alternating-gradient principle, 79 43, 46, 70, 76, 77, 85, 87, 104
alternating-gradient synchrotron Carnegie Institute of Technology,
(AGS), 21,72, 73, 89; fixed- 50
field (FFAG), 72, 95, 96 Carnegie Institution, 3, 4, 9
American Institute of Physics, 90 cascade transformer, 3
Amsterdam, 50 Cavendish Laboratory, 1, 5, 30
analogue, electron, 71 CERN, Geneva, 56, 62, 67, 70, 71,
Argonne National Laboratory, 53, 73, 74, 80, 89, 90, 94
56, 100 Chadwick, J., 35, 36
Armenian Academy of Sciences Chaffee, M., 4
(ARUS), 77 chamber, vacuum, 29, 32
Atomic Energy Commission Charlton, E. E„ 4
(AEC) , 21, 52, 74, 76, 79, 80, Chicago, University of, 50
83, 85, 92, 93, 96, 98, 100, 101, Christofilos, N. C„ 69, 70
103, 108; Scientific Advisory Coates, W„ 17, 34, 55
Committee, 92 Cockcroft, J. D„ 1, 5, 6, 11, 16, 30
Cockcroft and Walton, voltage
Bainbridge, K. T., 77 multiplier, 5
Barnes, D. E„ 44 Cole, F. T., 100
beam, emergent, 106 colliding-beam, facility for 200-
Beams, J. W„ 54 GeV accelerator, 107
betatron, 20, 38, 53, 62, 89 Collins, T., 86
Bethe, Η. Α., 46 Columbia University, 50
Birmingham, University of, 50 Condon, E. U., 5
Blewett, J. P., 52, 67, 70, 72, 90 Cornell University, 17, 37, 44, 46,
Bonn, University of, 46 73, 77, 104
Boston Museum of Science, 9 cosmotron, 53, 61, 62
Brasch, Α., 2 CPS, CERN, 71, 73, 88, 89
Breit, G., 3 Cooksey, D„ 16, 30, 37
Brobeck, W. W„ 37, 52, 89 Cooper, P., 86
Brookhaven National Laboratory, Courant, E. D„ 63, 64, 70, 78
(BNL), 21, 52, 53, 56, 60-62, 66, Cummings, R. E., 86
67, 70-73, 78-80, 89-95, 97 Curie, I., 36
119
120] Index
Curtis, C. D„ 100 pressure-insulated, 11; tandem,

cyclotron, 13, 16, 34, 54; Crocker, 13
18, 37; electric-focusing, 26; fre- Geneva, 50
quency, 24; isochronous, 18, 48, Getting, I., 78
67, 73; magnetic-focusing, 26, Getz, D„ 100
32; resonance, 22, 24, 39; sector- Ginzton, E. L., 58
focused, 48; 1.2-MeV, 16,29; Glasgow, University of, 44
27i^-in., 7, 33; 37-in., 7, 37, 48; Goldwasser, E. L„ 100, 102
184-in., 38, 48 Goward, F. K„ 44, 67, 71
gradient magnetic field, 60
Dahl, O., 3, 36, 67, 71 Green, G. K., 52, 72
Davenport, W., 77 Gurney, R. W., 5
Dekleva, J., 86
deuterone, 17, 35 Hafstad, L. R„ 3, 36
Deutsches Elektronen Synchrotron Hansen, W.W., 58
(DESY), 77 Harvard University, 50, 70, 76,
DeWire, J., 100 77, 81, 82, 84, 85
disintegration, nuclear, 30 Harwell Atomic Energy Establish-
drift tubes, 55, 56 ment, 50, 69
Dubna, USSR, 50, 53 Haworth, L. J., 52, 67, 72, 78
DUSAF (architectural-engineering Henderson, M.C., 17, 34, 37
firm) , 100, 101 Herb, R. G„ 11
High Voltage Engineering Corp.,
Edlefsen, N. E., 14, 22, 24 13
ejection, vertical, in 200-GeV ac- Hiñe, M.G.N., 69
celerator, 106 Hofmann, J. Α., 80
electron synchrotron, 20, 40, 43
Illinois, University of, 20
electrostatic generator. See gen-
injection, vertical, in 200-GeV ac-
erator, electrostatic
celerator, 106
equilibrium orbit, 41, 47
Ising, G., 53
expanditron, 103
Johnson, T., 79, 80, 83, 84
Federal Telegraph Company, 17 Joint Committee on Atomic En-
fixed-field alternating gradient ergy, 97, 108
(FFAG), 72, 95, 96 Joliot, F., 36
focusing: alternating-gradient, 60, Judd, D.L., 89, 91
68, 69, 79, 88; electric, 26; mag-
netic, 26, 32; sector, 67 Kensington Museum of Science,
Frascati, Italy, 46 29
Frazer, J. F., 80 Kerst, D.W., 20
frequency modulation, 47, 51, 69 Killian, J. R„ 85
Kinsey, B.B., 17, 34, 55
Gamow, G., 5 klystron, 58
Garren, Α. Α., 103 Koehler, Α., 86
General Electric Company, 2, 4, Kurie, F.N.D., 16, 17, 30
20, 44
generator, electrostatic, 3, 4, 7, 9, Lambertson, G. R., 103
11, 36; van de Graaff, 13, 56; Lange, F., 2
121] Index
Lanza, G., 85, 86 nuclear resonances, 11, 16

Lauritsen, C. C., 3
Lawrence, Ε. O., 13, 14, 16, 17, 22, Oak Ridge National Laboratory,
23, 25-30, 34-37, 46, 54, 55, 91 18
Lawrence, John, 37 Office of Naval Research, 44, 50
Lawrence Radiation Laboratory Oliphant, M. L., 50
(LRL), 34, 38, 74, 91-94, 96, 97, Oppenheimer, J. R., 34, 77
102 open housing, 99
Lawson, J. D., 69 orbit stability, 63
Lewis, G.N., 17, 34 oscillations: betatron, 61; syn-
linac: electron, 58; heavy-ion, 55; chronous, 64
proton, 55, 56, 91 Oxford University, 7, 44
Liverpool, University of, 50
Livingood, J. J., 17, 34 Pake, G. E., 96
Livingston, M.S., 4, 14-17, 24, 28, Panofsky, W. K. H., 55, 58
37, 52, 61, 63, 70, 79, 83, 84, 100 Pennsylvania, University of, 84
Lofgren, E. J., 89, 93, 103 phase stability, 39, 40, 49, 56, 59
Los Alamos Scientific Laboratory, Philips Company, Eindhoven,
56 Holland, 6
Lucci, T., 34 Piore, E. R., 98
Poillon, D., 100
Massachusetts Institute of Tech- Poulsen-arc magnet, 30
nology, 9, 11, 13, 17, 44, 61, 70, Princeton-Pennsylvania Acceler-
76-79, 81, 82, 84 ator (PPA) , 53, 83
McGill University, 50 Princeton University, 7, 53, 84
MacMillan, E. M„ 17, 33, 34, 39, Purcell, E. M., 84
44,46, 50, 51,55,78, 89,91 Purdue University, 44
mesons, 44 Pusey, N. M„ 85
microtron, 59
Michigan, University of, 44 quadrupole, 65, 66
Midwestern Universities Research
Association (MURA), 72, 73, Rabi, 1.1., 52
89, 96 radioactivity, induced, 17, 36
Momentum compaction of orbits radiofrequency resonance, 53
in AGS, 65 Ramsey, N. R„ 78, 79, 81-83, 92,
98, 100; panel, 92, 96
National Academy of Sciences, 95, Read, A. L., 100
97, 98 Rees, J., 86
National Accelerator Laboratory, Research Corporation, 17
74, 94, 99, 102, 107 resonance, orbital, 69
National Research Council, 14, 27 resonance acceleration, 13, 39
National Science Foundation, 82 resonance transformer, 4
neutron, 35, 36; production, 17 Robie, F., 86
Nimrod (accelerator), 53 Robinson, K., 85, 86
Nobel Institute, 18 Rochester, University of, 50;
Nobel Prize, 5, 37 High Energy Physics Confer-
Northern Institutions National ence, 90
Accelerator (NINA) , 77 Rose, M. E., 46
nuclear disintegration, 5 Royal Institute, Stockholm, 46
122] Index
Royal Society, 1 Tuve, Μ. Α., 3, 9, 36

R u t h e r f o r d , E., 1, 5 Twiss, R . Q., 80
R u t h e r f o r d Laboratory, 53, 90
University of California Radia-
Saclay, France, laboratory, 53 tion Laboratory ( U C R L ) , 69,
Salisbury, W., 37 89-91; see also Lawrence Ra-
Sands, M„ 89, 91 diation Laboratory
Saturne (accelerator), 53 Uppsala, University of, 50
Schmelzer, C„ 71 Universities Research Associa-
scientific justification of large tion ( U R A ) , 97-101
accelerators, 75, 95 Urban, C„ 2
sections, long straight, in 200-
GeV accelerator, 106 Vallauri, M„ 86
Seitz, F., 97 Van d e Graaff, R . J., 4, 7, 9, 11,
separated-function magnets, 108 13; generator, 13, 56
Serpukhov, USSR, 21, 56, 73 Varian, R . H., 58
shimming, 27 Vash, Α., 80
site selection, 98 Veksler, V., 39, 46, 50, 51, 59, 78
SLAC (accelerator), Stanford Virginia, University of, 54
University, 58, 59, 76 voltage multiplier, 5, 6
Sloan, D. H „ 4, 29, 34, 54, 55
Smith, L., 89, 91, 93, 103 Wächter, M., 86
Snyder, H., 63, 65, 70, 78 Walker, R „ 96
Sorensen, R . W „ 2 Walton, E. T . S., 1, 5, 6, 11, 16, 30
stability, dynamic, 65 Waniek, R „ 85, 86
storage ring, 107 waveguides, 58
Stratton, J. Α., 78 Weston, Illinois, 21, 74, 98, 99;
Street, J . C„ 79 see also National Accelerator
surge generators, 2 Laboratory
synchrocyclotron, 38, 46, 50 White, M. G„ 17, 30, 84
synchronous acceleration, 41, 51 Wideröe, R., 13, 23, 53, 54, 67
synchrophasotron, 53 Wilson, R . R . , 73, 74, 77, 100,
synchrotron, 43; alternating- 102, 103
gradient, 60, 70, 73, 74, 81; Wisconsin, University of, 11
booster, 91, 104, 105; cascade,
89; electron, 20, 40, 43; pro- X-rays, 20
ton, 50, 73; radiation, 44
Yale University, 16, 30
Tesla coil, 3, 4 Young, D. E., 86, 100
Thomas, L. H „ 47, 67, 73 Yuan, L. C. L„ 95, 97
T h o r n t o n , R. L„ 17, 34, 89
Tokyo, University of, 46 Zacharias, J. R „ 78, 79
transformer, resonance, 4 zero-gradient synchrotron (ZGS),
T r u m p , J . G., 11, 13 53

9AAGL48ZMF3VN0GPWFIN5J49M

Uploaded by

Copyright:

Available Formats

9AAGL48ZMF3VN0GPWFIN5J49M

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

9AAGL48ZMF3VN0GPWFIN5J49M

Uploaded by

Copyright:

Available Formats

Particle Accelerators: A Brief History

H a r v a r d University Press : C a m b r i d g e , Massachusetts : 1969

This monograph is largely derived from manuscripts of

1 T h e Race for High Voltage 1

The Race for High Voltage

W h e n J. D. Cockcroft and E. T . S. W a l t o n first disintegrated

tended to even higher potentials. "Surge generators" h a d been

which was using 220-kilovolt lines at 50-cycle frequency for

ever, the oscillatory character of the potential obtained from

veloped 9 and used in the early 1920's for application to high-

T h e Cockcroft-Walton voltage-multiplier circuit (Fig. 1) is

pacitors and rectifier tubes, with rounded corona shields at the

FIG. 1. Schematic circuit diagram of the voltage multiplier, showing the

formed of insulating glass cylinders. T h e ion source at the up-

FIG. 2. Voltage multiplier with which Cockcroft and Walton achieved

Princeton in 1930, after becoming interested in the need for

FIG. 3. J . D . Cockcroft in the basement room below the voltage multi-

generator before the American Physical Society in 1931. T w o

eral laboratories became interested a n d joined in the develop-

tripod of insulating struts, with a vertical accelerator tube

generators for research use at M I T . T h i s development culmi-

I obtain an a p p o i n t m e n t as instructor at the University to con-

FIG. 7. Vacuum chamber for the 1.2-MeV cyclotron, built by E . O . Law-

A few m o n t h s before this goal h a d been reached, the publi-

tron, and a thin-foil window was mounted on the chamber

"Crocker" cyclotron; 28 this was completed in 1939, operating

FIG. 9. T h e first betatron built by D. W. Kerst at the University of Illi-

early years the purpose was to produce high-energy X-rays for

magnet, it displaced the betatron as an electron accelerator in

I first met Professor Lawrence as a student in his Electricity

Elektrotechnik for 1928, and recognized the resonance princi-

FIG. 10. Diagram of the a p p a r a t u s used by R. Wideröe in 1928 to demon-

which a radiofrequency potential was applied (Fig. 10). This

in an external inductance coil formed of copper tubing. Hy-

D u r i n g the fall I continued the technical development with

frequency. Incidentally, these resonance values provided a

greatly increased currents, in the 10 _9 -A range. O n his return

F I G . 12. Sketches used by L a w r e n c e a n d L i v i n g s t o n in an early p u b l i c a -

certain sizes a n d locations of the shims. T h i s result caused us to

F I G . 13. D i a g r a m of t h e v a c u u m c h a m b e r for t h e 1.2-MeV cyclotron b u i l t

committee were appalled to find that I had not studied

resonance accelerator capable of reaching 1-MeV energy, lo-

visions to increase beam intensity when we received the issue of

T h e vacuum chamber was a 27-in. brass ring fitted with iron

lowed us to locate the deflection electrode and collector at any

lected radii placed in the shimming gaps. Eventually we

J u n e 13, 1932 16-cm radius, 28-m wavelength, 1.24-

December 2-5, 1932 Target chamber installed for studies of

These were busy and exciting times. Other young scientists

ment had succeeded in concentrating heavy water with about

the filament current higher and higher, exclaiming each time

turned the target wheel to carbon, adjusted the point-counter

gave him the opportunity to implement his dreams of even

Synchronous Accelerators and How They Grew

T h e principle of synchronous or phase-stable acceleration

enee an automatic correction toward the proper phase and en-

TABLE 1. Characteristics of synchronous particle accelerators.

T h e stability principle has had important applications in

applied which has a frequency equal to the orbital frequency

F I G . 17. (a) D i a g r a m of t h e r a d i o f r e q u e n c y cycle of a s y n c h r o t r o n , in