The 1920 Shapley-Curtis Discussion: Background, Issues, and Outcome

     ___________________________________________________________________

 

                           Virginia Trimble

 

              Physics Department and Astronomy Department

          University of California     University of Maryland

          Irvine CA 92717              College Park MD 20742

                                       vtrimble@astro.umd.edu



                      [received: December 20, 1995]

             (first printed - Vol. 2, No. 6 - January 1996)



Prepared for the 1995 75th Anniversary Astronomical Debate 

and for publication in Publications of the Astronomical Society of the Pacific

 

Reproduced by the kind permission of the Author and Editor, from:

Virginia Trimble, Publications of the Astronomical Society of the Pacific 

(PASP) vol. 107, p. 113 (1995). 

Copyright 1995, Astronomical Society of the Pacific; 

reproduced with permission.

 

 

ABSTRACT

________

 

	No one now living attended the original lectures by Curtis and 

Shapley, and the scientific and other worlds in which they moved are connected 

to ours only by the written record and second-hand stories. Depending on which 

corners you choose to peer into, those worlds can seem remarkably modern or 

remarkably ancient. As is often the case for classic dichotomies, the wisdom 

of hindsight reveals that each of the speakers was right about some things and 

wrong about others, both in choosing which data to take most seriously and in 

drawing conclusions from those data. Modern (mostly casual) discussions of the 

1920 event leave the impression that Shapley was, on the whole, the winner. 

But the two men's reactions to Hubble's discovery of Cepheids in the Andromeda 

galaxy makes clear that both felt that the issue of existence of external 

galaxies (on which Curtis had been more nearly correct) was one of greater 

long-term importance than the size of the Milky Way (on which Shapley had been 

more nearly correct). Shapley is much the better known today and is generally 

credited in text books with the Copernican task of getting us out of the 

center of the galaxy. Under modern conditions, he would probably also have 

gotten most of the press notices. Curtis's repeated theme, "More data are 

needed," is remarkably difficult, then as now, to turn into a headline.

 

1. INTRODUCTION

_______________

 

	The suggestion came originally from George Ellery Hale, whose father 

had endowed a lecture series for the National Academy of Sciences. After some 

initial hesitation, the NAS Home Secretary, C.G. Abbot, agreed that the 1920 

William Ellery Hale lectures would be a discussion on "The Distance Scale of 

the Universe," with Harlow Shapley of Mt. Wilson Solar Observatory and Heber 

Doust Curtis of Lick Observatory as the discussants. Both the published 

versions of their presentations (Curtis 1921, Shapley 1921) and the notes from 

which they spoke (Hoskin 1976) are now available, as is a good deal of 

information on the lead-up to what much later came to be called "the great 

debate" and on its scientific aftermath.

 

	We first examine the cultural and scientific environments in which the 

1920 event occurred, then the event and its participants, ending with an 

examination of the scientific issues as then perceived and as now understood. 

It is not clear whether any very useful lessons for the case of gamma ray 

bursters can be drawn. As is frequently (but not always!) the case in 

scientific disputes, Shapley and Curtis each had hold of portions of the 

correct elephant.

 

2. THE WORLD IN 1920

____________________

 

	At the time of their Academy encounter, Heber Doust Curtis and Harlow 

Shapley were employed respectively at Lick and Mt. Wilson Observatories. A 

born Californian, I thought first of probing their world by comparing the road 

maps they would have used with the ones that now guide us to the observatory 

sites. At first glance, the differences seem small. The main north-south route 

into the Oakland - San Francisco area, then as now, split to go both ways 

around the Bay. And a motorist striving to get over the mountains surrounding 

Los Angeles had a choice of two routes, one now called the Hollywood Freeway 

and one the Golden State Freeway, which follow the routes then called Cahuenga 

Pass and San Fernando Road, the latter nearly the old Spanish El Camino Real 

from Mission San Gabriel to Mission San Fernando.

 

	The speed limit on the open road was, however, 35 mph (30 mph in 

Oregon), and the driving instructions rejoiced in stretches that were "paved 

all the way" and presented "no grades steeper than 12%." Alum Rock Road, where 

one begins the modern climb out of San Jose to Lick was on the maps, but 

petered out within a few miles into randomly-places images of hillocks and 

mountains that might almost have been labeled "here be dragons." The Mt. 

Wilson road was both better marked and more often traveled by casual visitors, 

but the site of Palomar Observatory was simply a random part of northern San 

Diego County, between the Pala Indian Reservation and "Nellie Warner's Hot 

Springs." The modern access road, from the south, was built by San Diego 

County much later. According to a contemporary hand-drawn map, the site could, 

however, be reached from the west, via a route later called Harrison Grade 

(and then carrying a name so politically incorrect that I dare not mention it) 

past landmarks like "Doane's old cabin" and "Elbow Creek Telephone Line."

 

	The auditorium in which we meet had existed for about seven years and 

contained seats made of materials suitable for the pre-microphonic age. Curtis 

and Shapley necessarily filled the room with their own voices.

 

2.1 Politics, History, and Demographics

_______________________________________

 

	The 9:30 PM Conversazione following the 1920 William Ellery Hale 

lectures took place without the customary glasses of wine, for the 19th 

amendment to the US constitution took effect on January 16th, ushering in "the 

great experiment" of prohibition (which, though it had the desired effect of 

considerably decreasing ethanol consumption, is nevertheless generally held to 

have failed).

 

	That year also American women went to the poles nationwide for the 

first time, increasing voter turn out nearly 25% over the previous two 

elections and helping to elect Warren Gamliel Harding and Calvin Coolidge over 

James M. Cox and Franklyn Delano Roosevelt by 16.1 to 9.1 million votes (by 

modern standards an overwhelming majority). Eugene V. Debs, running for the 

Socialists, also lost, for the fifth and last time. Only Norman Thomas, his 

successor, with six defeats, ever equaled or beat his record. Levi P. Morton 

(vice president under Benjamin Harrison) died at the age of 96, and I mention 

it because he was born in 1824 and so overlapped by two years the lives of 

Thomas Jefferson and John Adams. We are a young country! (My grandmother, 

dying at 98 in 1984, had lived through more than half the life of our 

Constitution.)

 

	Outside the US, the League of Nations was established (fatally, 

without the US); Austria held her first elections; and the Communist party 

completed taking control over the newly-named USSR. Two Georges ruled England 

(David Lloyd as prime minister and "V" as king); Poland retook Wilno/Vilna 

from Lithuania (with long-term implications for the universities and 

demographics of the region); and Benedict XI was pope, in succession to Pius 

X, the last occupant of the chair of St. Peter so far elevated to sainthood 

(we hope in spite of, not because of, his abolition of solos in liturgical 

music).

 

	World population was roughly 2 Gigapersons, with 108 million of them 

resident in the US. Within the US, only 4.7% of the population was over 65, 

and the male:female ratio was 1.04 (and greater than unity even for the 

over-65's, the last census for which this was true). The foreign-born fraction 

was about 13%, higher than at any time since. Women made up 22% of the labor 

force, and unemployment was 5.2%, quite close to the current level.

 

	Our national debt, left from the first world war (and the first one 

never significantly repaid) stood at $24.3 million, or $228.32 per person. 

This was something like 10 weeks salary for a semi-skilled craftsman and so 

not so very different either from the current level.

 

	Among the 300,000 people who graduated from high school, women 

outnumbered men by 50%, but men outnumbered women nearly 2:1 among the 48,000 

college graduates. Another legacy of the "great war," Spanish influenza, wound 

down after killing roughly 20 million people in three years, compared to about 

8.5 million in WWI itself (insert your own best estimate for AIDS fatalities 

to date).

 

2.2 Sports and Culture

______________________

 

	Somehow these items seem to present the most striking contrast of 

ancient and modern. The Cleveland Indians (their name not yet threatened by 

the forces of political correctness) defeated the Brooklyn Dodgers (long gone) 

in the 1920 World Series. Harvard edged out Oregon 7-6 in the Rose Bowl, in 

striking contrast to the "Fight Fiercely, Harvard" image we inherit from Tom 

Lehrer. Jack Dempsey was 7th world heavyweight champion, while Emanuel Lasker 

of Germany, the first man ever declared world chess champion, still held the 

title. At the 7th Olympiad, Pavlo Nurmi won his first gold medals (one of a 

large number of Finnish track and field winners). The American men raking in 

gold for swimming events carried Hawaiian surnames, a reminder of the time 

when swimming was a survival skill rather than a recreation.

 

	The winner of the Kentucky Derby (Paul Jones - horse, not rider) had a 

winning time only marginally longer than current records, though the purse at 

$30,375 sounds small till you inflate it. But the winner of the Indianapolis 

500 (Gaston Chevrolet, driving a Monroe) had an average speed of 88.62 mph, 

slower than many of us have driven our production models. Bill Tilden (from 

the US) won Wimbledon, and the Ottawa Senators carried away the Stanley cup.

 

	The Academy Awards (Oscars) had yet to be invented, but Eugene O'Neill 

won the 1920 Pulitzer for "Beyond the Horizon". It was not a great year for 

the Nobel Prizes, several of the winners inviting a "hoo hee" response from 

non-experts. Physics went to Charles Guillaume, Peace to Leon Bourgeois, 

Literature to Knut Hamsun, Physiology or Medicine to August Krogh, and 

Chemistry to Walter Nernst (who illustrates the advantages of having a theorem 

named after you).

 

	Enrico Caruso sang his last performance (La Juive) - which feels 

infinitely long ago, and Agatha Christie published her first murder mystery 

"The Mysterious Affair at Styles" - which was obviously only yesterday, since 

she brought out new volumes long enough to see many of us through high school 

and beyond. Sinclair Lewis published "Main Street", which remains a classic 

(something everybody wants to have read, but nobody wants to read).

 

	The first regular transcontinental airmail opened between Boston and 

San Francisco. Deaths during the year included artist Modigliani and explorer 

Admiral Robert E. Peary -- both controversial figures down to the present. An 

incomplete list of those born in 1920 includes Ravi Shankar, Isaac Stern, Nat 

"King" Cole, Alex Hailey, Isaac Asimov and Ray Bradbury, Mickey Rooney, 

Federico Fellini, Yul Brynner, Eileen Farrell, Lana Turner, Tony Randall, 

David Brinkley, Dave Brubeck, Jack Webb, Stuart Udall, Walter Mathau, and 

Patti Andrews. Christmas came on a Saturday, and the 26 April debate on a 

Monday.

 

	Finally, what would prove to be the trial of the year or even the 

decade began with the arrests of Nicola Sacco and Bartolomeo Vanzetti, the 

good shoemaker and the poor fish peddler, for a murder most now think they 

never committed (though they died for it in 1927), but really for the crime of 

not being upper middle class WASPs. The drawing of modern analogies is left to 

the reader.

 

2.3 Astronomy in 1920

_____________________

 

	The Astrophysical Journal was already a quarter of a century old and 

under the joint editorship of Hale, Frost, and Gale. They had just added 

abstracts to the standard paper format and admitted that page charges were 

here to stay, owing to the numbers of overseas subscribers not having 

recovered after the War, at least for authors or institutions who incurred 

more than $200 of production expenses in any one volume (of which there were 

two per year, with fewer than 30 papers each).

 

	Very few of our current "best buy" theories were yet in place 

(Russell, Dugan, and Stewart 1926; Eddington 1926). The Chamberlin-Moulton 

(dynamic encounter) hypothesis for the origin of the solar system was in 

favor, largely because the sun seemed to have too little angular momentum to 

have come from a "nebular hypothesis." The solar wind eventually resolved that 

issue. Elements common in the earth (silicon, iron, oxygen) were supposed also 

to dominate the stars, giving them (with ionization) a mean molecular weight 

close to 2.1. It took Cecilia Payne's 1925 Harvard thesis on K giants and H.N. 

Russell's later work on the sun to sort this one out.

 

	Not surprisingly, the source of stellar energy was unknown. The 2 Gyr 

age of some earth rocks (found by Rutherford and his colleagues) and the 

stability of Cepheid pulsation periods had already demonstrated that neither 

gravitational potential energy nor radioactivity was sufficient. New ideas in 

the air were "subatomic energy" that might power the sun for 10^10 years 

without much changing its mass (advocated by Eddington) and some form of total 

annihilation of electrons and protons (the only known particles) that would 

suffice for 10^12 years (favored by James Jeans because he thought that much 

time was needed to allow star clusters to relax). The only picture of stellar 

evolution sufficiently developed for comparison with observations was 

Russell's giant and dwarf theory, whose imprint lingers today in the use of 

"early" and "late" for spectral types. The idea was that stars begin bright 

and red, contracting toward the main sequence until they had used up all their 

"giant stuff", whatever it was, and then move diagonally down the main 

sequence, living on their "dwarf stuff" for a much longer time, fading out as 

red or white dwarfs. The debaters were both more or less subscribers to this 

point of view, and Shapley invokes it as part of the theoretical argument for 

his point of view.

 

	Events of 1920 within the astronomical community included the deaths 

of Lockyer (discoverer of helium and founder of Nature), Brashear (of the 

process), and Hermann Struve. The Royal Astronomical Society marked its 

centenary, with Frank Dyson (whose successor is our moderator) as astronomer 

royal. Warner & Swazey Observatory was dedicated, and installments of the 

Henry Draper Catalogue (spectral types) and Wolf Catalog of proper motions 

were published. The International Astronomical Union, the first of the 

international scientific unions established under the Treaty of Versailles, 

which specifically abolished all international organizations of the pre-war 

period, came into being. The losers in the recently-ended conflict were 

specifically barred from membership, and Germany did not adhere to the Union 

until another war had come and gone.

 

	Publications during the year relevant to "the scale of the universe" 

included Shapley on globular clusters, Haber claiming that Cepheids were 

eclipsing binaries (a well known crank in his day, now nearly forgotten), 

Kapteyn and van Rhijn arguing for a small, nearly sun-centered galaxy on the 

basis of star counts, and H.N. Russell demonstrating that the large positive 

velocities of the spiral nebulae could not be caused by radiation pressure 

from the Milky Way. Shapley apparently thought this a possible mechanism while 

he was preparing his manuscript. That anyone could have entertained the idea 

for more than five minutes suggests a painful shortage of envelope backs. The 

Thompson cross section and the momentum carried by light were already old 

ideas.

 

	Some of the less relevant papers were remarkably prescient. Albert 

Michelson was advocating use of the 60" and 100" telescopes (the latter only 3 

years old) at Mt. Wilson for interferometry. Eleanor Seiler suggested the use 

of photoelectric cells as photon detectors for astronomy. And Walter S. Adams 

and Coral Burall pointed out that novae must really be ejecting material. 

Other 1920 authors who are part of our folklore include Joel Stebbins (who, 

with a 64 year history of publications in ApJ, 1901-64, may be the 

longest-productive astronomer ever, Abt 1995), E.O. Hulbert, Francis G. Pease, 

Karl T. Compton, F.H. Seares, Edwin Hubble, Robert Millikan, Leigh Page, R.S. 

Dugan, Gustave Strömberg, and Seth Nicholson. Among those who lived long 

enough that I (and undoubtedly many of you) had a chance to meet them were 

Alfred H. Joy, Ira S. Bowen, Harold D. Babcock, Bancroft W. Sitterly, and Paul 

Merrill. The proportion of women authors was not so very different from the 

current mix. In addition to Burall and Seiler just mentioned, I spotted Mary 

Fowler (on eclipsing binaries), Mary Ritchie and Helen David (both Shapley 

co-authors in the Harvard tradition of measuring project lengths in 

woman-years).

 

	Shapley and Curtis were not the only well known scientists to speak at 

the 1920 Academy meeting, though the usual difficulties of travel were 

compounded by, as secretary Abbot described it, "Washington [being] still 

somewhat congested" in the aftermath of the war. What would he think of the 

place now, when a change in power structure means that the Republicans arrive 

but the Democrats don't leave (or conversely)? There were no parallel 

sessions, but a good many speakers were allotted only 5, 10, or 15 minutes.

 

	In any case, Frank Boas spoke on "growth and development as determined 

by environmental issues." He meant of people, and the issue is still (or 

again) a burning one. Robert Yerkes presented the results of a psychological 

study of Army doctors. Robert H. Goddard proposed "possibilities of the rocket 

in weather forecasting." Hale described recent results from the 100" telescope 

(as old then as Keck is now), Edward Kasner discussed "geodesics and 

relativity," Millikan "reflection of molecules from surfaces," Michael Pupin 

"wave balance," whatever that is, and Arthur Noyes (brother of the poet 

Alfred) the direct combustion of nitrogen and chlorine. Some of these would be 

perfectly possible titles or subjects for this year's academy meeting. Some 

definitely would not.

 

	Topics whose presenters are less familiar to our selective memories 

were a similar mix of ancient and modern -- "conservation of nature 

resources," "rate of growth of the population," "Indian tribes of the Klamath 

River region," "common foods as sources of vitamines" (but note the spelling; 

they were all still thought to be true amines), "specific heat of powder 

gases," "alternating current for submarine transmission," "improvements in 

telegraphy," and two presentations on the properties of Springfield rifles! 

Yes, American militia units really carried the black powder, smoky "trapdoor" 

right up to, and occasionally beyond, the moment we went "over there." 

(Sweeney 1995; Pinckney 1995)

 

3. HISTORY OF THE DEBATE AND DRAMATIS PERSONAE

______________________________________________

 

	The background and circumstances of the 1920 lectures have been 

described by Struve and Zebergs (1962), Whitney (1971), Jaki (1972), and 

Berendzen et al. (1976), among others, on the assumption that the printed 

versions of the talks (Curtis 192l, Shapley 1921) were a close approximation 

to the material presented orally. Hoskin (1976) has shown that this is not the 

case, and his discussion therefore takes precedence.*

 

	William Ellery Hale I, having presciently moved his family out of the 

center of Chicago shortly before the 1871 fire, made his fortune by 

constructing elevators for the buildings that grew up afterwards as well as 

for the Eiffel tower and other structures (Wright 1966, Osterbrock 1993). Some 

of the profits of these ventures bought his elder son, George Ellery, his 

first microscopes and telescopes, and, eventually, much of the Mt. Wilson 60". 

In addition, he endowed a fund for the National Academy of Sciences to be 

used, among other purposes, for invited lectures at annual meetings. Shapley 

and Curtis each received $150 honorarium (plus, presumably, travel expenses).

 

	Not surprisingly, G.E. Hale (elected to the Academy in 1902) had some 

considerable say in how these funds were expended. In late 1919, he spoke to 

Charles G. Abbot, Home Secretary of the NAS, proposing that there be a William 

Ellery Hale Lecture at the 1920 April meeting in the form of a debate or 

discussion on either general relativity or the distance scale of the universe. 

Abbot's reaction was that it might be difficult to stir up interest in so 

specialized a topic as the existence of island universes, and that everyone 

would be heartily sick of relativity by then. He counter proposed causes of 

the ice ages or some topic in zoology or biology. Hale had originally 

suggested that the discussants on island universes and the distance scale 

should be W.W. Campbell (director of the Lick Observatory), presenting the 

conventional view, and Harlow Shapley (Hale's junior associate at Mt. Wilson), 

putting forward his new, larger distance scale, based on variable stars in 

globular clusters and other considerations.

 

	When the dust settled, they had agreed on two talks, by Harlow Shapley 

and Heber D. Curtis (of Lick) on "the distance scale of the universe," and 

Hale sent out telegrams of invitation on 18 February. Shapley's invitation 

still exists, in the possession of Vera C. Rubin, who found it in a book she 

bought from Shapley's collection. After some discussion, the lecturers agreed 

to exchange their ideas in advance and each to give a single talk, with 

Shapley going first, and to include responses to each other's viewpoints 

therein, rather than to adopt a debate format, with rebuttals. The 

participants in the 1995 commemorative even similarly considered several 

possible formats, but made a different choice, opting for a formal debate 

structure.

 

	Table 1 presents some aspects of the lives and works of the four 

people most closely associated with the 1920 debate: Hale who suggested it, 

Shapley and Curtis who carried it out, and Edwin Hubble who, a few years 

later, collected the data that settled the issue of island universes. All were 

born in the midwest, within 21 years of each other, and all had doctorates of 

some sort, though Hale's were all honorary. I mention their activities during 

WWI because at least part of the source of the life-long coolness between 

Hubble and Shapley was that Shapley, remaining at Mt. Wilson, carried out some 

project that Hubble had intended to pursue as soon as he could take up his 

profered position there after returning from active duty overseas (Hoffleit 

1995). Hubble had volunteered immediately after defending his thesis and 

apologizing to Hale for not being able to accept the Mt. Wilson position 

immediately. We was wounded in France and rose to the rank of major. 

Correspondingly, during the second world war, while Shapley remained at 

Harvard (helping resettle refugees), Hubble moved to Aberdeen Proving Ground 

to direct its ballistics lab.

 

                             TABLE 1

                             _______

 

SOME HIGHLIGHTS OF THE LIVES OF HALE, CURTIS, SHAPLEY AND HUBBLE

_________________________________________________________________

 

  HALE                  CURTIS             SHAPLEY              HUBBLE

  George Ellery         Heber Doust        Harlow               Edwin Powell

  _____________         ___________        ___________          ____________

born:

  1868 Chicago          1872 Muskegon      1885 rural           1889 rural

                        Mich.              Missouri             Missouri

1st degree:

  1890, BS              1893, classics     1911, astronomy      1910, BS 

  MIT                   U. Michigan        U. Missouri          Chicago

                                                                & Oxford

PhD :

  12 honorary           1902, U. Virginia  1913, Princeton      1917, Chicago

WWI:

  National Research     Taught             --------             Active Duty,

  Council               navigation,                             France 

                        Berkeley & San                          1917-19,

                        Diego; NBS                              infantry

                        optical section                         battallion CO

Career:

  Harvard & private     Taught Latin &     Mt. Wilson 1914-20   Mt. Wilson

  observatory 1886-96   Greek, later       Director HCO         1919-53

  Yerkes/Chicago        math, 1893-1900    1921-52 Emeritus to

  1887-1904 Mt. Wilson  Lick 1902-20       1972

  1904-23 honorary      Director,

  director & private    Allegheny 1920-30

  observatory 1923-38   Director, U.

                        Michigan

                        Observatories

                        1931-42

WWII:

  d.                    d.                 refugee              Aberdeen

                                           resettlement         ballistic 

                                                                missle

                                                                laboratory

died:

  1938, Pasadena        1942, Ann Arbor    1972, Boulder CO     1953, Pasadena

                            MI

obituaries:

  NAS (W.S. Adams)      PASP 54 (McMath)   QJRAS (1972, B.      PASP 66 (H.P.

  and at least                             Bok) Nature (1972,   Robertson)

  8 others                                 Z. Kopal)

 

 

 

	Of our four protagonists, Hale was far the most wide-ranging in his 

activities (Wright 1966; Osterbrock 1993). Astronomers know him as the founder 

and initial fund raiser for Yerkes, Mt. Wilson, and Palomar Observatories. A 

strong believer in international cooperation, he was among the prime movers in 

establishing the International Union for Cooperation in Solar Research in the 

years before WWI. Not easily discouraged in those days, he reacted to its 

abolishment by the Treaty of Versailles (which dissolved all pre-war 

scientific and cultural international organizations) by starting over with a 

still larger vision and persuading into existence the entity now called the 

International Council of Scientific Unions, as well as the International 

Astronomical Union under it.

 

	During the war years, Hale was the first pure scientist to try 

seriously to persuade President Woodrow Wilson (awkwardly stuck with the 

slogan "He kept us out of war") that the services of his colleagues would be 

needed to win the war and the peace that followed. The organization he founded 

with that goal in mind is now the National Research Council. The Yerkes 

Primate Lab at Chicago is another of his inspirations. Curiously, the Robert 

Yerkes for whom it is named was not a close relative of the industrial magnate 

whose name the observatory bears. Hale early encouraged psychologist Yerkes to 

turn his attentions from people to other primates. Under the circumstances, 

one can only be astounded that Hale also made fundamental contributions to our 

understanding of the solar spectrum, magnetic field, and activity cycle, 

though he failed in a life-long ambition to photograph the solar corona 

outside of eclipse.

 

	Curtis, too, was interested in the sun and participated in 11 eclipse 

expeditions between 1900 and 1932 (McMath 1942). His years at Lick were, 

however, devoted primarily to photographing spiral nebulae with the Crossley 

telescope, the work that resulted in his being asked to face off with Shapley 

in 1920. Curtis moved later the same year to the directorship of Allegheny 

Observatory (having already served as president of the Astronomical Society of 

the Pacific in 1912) and later to the corresponding position at the University 

of Michigan. He was an important force in the transformation of Hulbert-McMath 

Observatory from a private endeavor to a serious research facility. His own 

research days essentially ended when he left Lick, though his name continued 

to grace the astronomical journals with papers on subjects as unlikely as "A 

Voyage to the Moon." Curtis guided to their PhDs Helen Dodson Prince (1934), 

Ralph B. Baldwin (1937), and K.O. Wright (1940) among others. The University 

of Michigan had, incidentally, been admitting women students to its astronomy 

graduate program since before 1920, when Julia May Hawkes received her PhD for 

work on the positions of stars and nebulous knots in the Great Nebula of 

Andromeda (Sears 1955). Curtis died in the observatory director's house in Ann 

Arbor, with his directing, if not his observing, boots on.

 

	Ralph Baldwin (1955, whose thesis was on the spectrum of Nova Cygni 

1920 and its relationship to that of Nova Herculis 1934) remembers Curtis as 

"a small, quiet man with a remarkable sneeze." Curtis was "not full of wild 

enthusiasm for Einstein's theory," to which he had a long list of objections, 

and once ended a graduate course by throwing out the final exams of the five 

of so students on the grounds that if he hadn't given them enough work over 

the whole semester to get to know them, "the three hours isn't going to tell 

me anything new." The grades were all A's. He described the 37 inch telescope 

at Michigan as "focusing like a dish pan," and had great expectations for the 

98 inch mirror he had cast at Corning in 1936 (while Corning was in the 

process of learning to produce the 200" blank for Hale and Palomar). 

Unfortunately, the money to turn it into a telescope never materialized, and 

the 98 inch sat next to the observatory parking area for many years until it 

became the primary of the Isaac Newton Telescope, and so sat next to 

Herstmonceux Castle for an additional number of years (contributing at least 

slightly more to astronomy in the latter location).

 

	Curtis, like Hubble, was a confirmed pipe smoker, who sporadically set 

his wastebasket on fire. It was a search for the correct pronunciation of his 

middle name that eventually led to my making contact with Baldwin. The correct 

answer is "to rhyme with soused." And if you think you have heard of Baldwin 

in some other context, you probably have. He was one of the very first and 

most vocal proponents of impact cratering as the explanation for "The Face of 

the Moon" (Baldwin 1949).

 

	Shapley, too, spent more of his life as an observatory director than 

as a research astronomer, taking up the reins of Harvard College Observatory 

shortly after the 1920 debate as successor to Pickering (and handing over to 

Donald H. Menzel more than 30 years later). He brought Harvard firmly into the 

20th century, though he retained always a preference for relatively small 

telescopes with wide fields of view (Kopal 1972). In the post-war years he 

served as president of the American Astronomical Society, the American 

Association for the Advancement of Science, and the honorary scientific 

fraternity Sigma Xi, and was a firm opponent both of the communist witch hunts 

in the US and of the nonsense propounded by Velikowski.

 

	Cecilia Payne Gaposchkin (at Harvard from 1923 to her death in 1979) 

described Shapley's style of leadership as "divide and rule" (Haramundanis 

1984, p. 224). His decision that she must switch from spectroscopy after her 

thesis work (which was the first clear demonstration that stars consist mostly 

of hydrogen) to variable stars, leaving the spectroscopy to Menzel, hurt her 

deeply without in the least making Menzel dislike Shapley less (Hoffleit 

1995). On the other hand, it was Shapley who persuaded Hoffleit to go on for 

her PhD (on spectroscopic parallaxes), though it would take her away from the 

work she was doing for him, and he welcomed her back at Harvard from war work 

at Aberdeen, though it had been done under the supervision of Hubble. 

Shapley's commitment to international cooperation rivaled that of Hale, and he 

is generally credited as the man who put the S in UNESCO.

 

	Hubble, in contrast, was primarily a research astronomer all his life. 

He never directed an observatory or held an AAS office, though he served two 

3-year terms as President of the IAU Commission now called Galaxies. While we 

remember him here for the discovery of Cepheids in NGC 6822, M33 and M31, 

which settled the issue of the existence of external galaxies, he is at least 

as well known for helping to draw the distinction between emission and 

reflection nebulae, discovering the linear redshift-distance relation that 

bears his name, classifying galaxies into their "Hubble types," and 

demonstrating that virtually all spiral galaxies rotate in the same direction, 

with their arms trailing. That Hubble was not personally known more to us is a 

consequence of his having been the shortest-lived of our protagonists. I have 

not attempted to assemble any personal impressions of him, but suggest that 

readers should take the one presented by Florence (1994) with some 

reservations, based on his treatment of Hale (Osterbrock 1993, 1995).

 

	Of course, a very large number of other astronomers contributed 

relevant data and ideas before, during, and after the epoch of the "great 

debate." Vesto Melvin Slipher (1875-1969) measured the first wavelength shifts 

of spiral nebulae. Johannes C. Kapteyn (1851-1922) was a life-long proponent 

of a small Milky Way, centered nearly on the sun, and his "Kapteyn universe" 

continued to bedevil attempts to picture the large scale distribution of stars 

for decades after his death (both Trumpler and Shapley trying to picture 

Kapteyn's star cloud as part of the disk of some larger structure traced out 

by the globular clusters).

 

	Adriaan van Maanen (1884-1946) was responsible for most of the 

measurements of apparent rotation of spiral galaxies that prevented Shapley 

from considering the possibility of their being at large distances until very 

late. Van Maanen's plates and equipment were not at fault. Although the 

instrument at Mt. Wilson bore the legend "Do not use the stereocomparator 

without consulting A. van Maanen", Knut Lundmark (1889-1958), visiting from 

Sweden, actually used it a few years after the debate to remeasure van 

Maanen's plates. He found no rotation, and, while the non-existence of the 

rotation is no longer in question, nobody has ever been quite sure what van 

Maanen did wrong. Lundmark was also the first to write, in 1920, that some 

novae might be so bright as to be detectable even at millions of light years 

from us. He advocated a quadratic relationship between redshift and distance 

(as expected in a de Sitter universe) before Hubble promulgated his law. 

Though van Maanen's sign remained on the blink comparator through my own 

graduate days (1964-68) and down to the time Berendzen photographed it (1972), 

I and others did eventually use it without consulting him. A minor point of 

possible confusion: "Mt. Wilson" was long used to mean, indifferently, the 

Mountain site and the administrative offices on Santa Barbara Street in 

Pasadena (the blink comparator was in the basement at Santa Barbara Street). 

Both places still exist, though the latter has undergone name changes to "Mt. 

Wilson and Palomar Observatories," "Hale Observatories," "Mt. Wilson and Las 

Campanas Observatories," and "Las Campanas Observatory," and "Observatories of 

the Carnegie Institution of Washington." And I may have forgotten one or two.

 

4. IMAGES OF THE MILKY WAY AND OUR DISTANCE TO THE GALACTIC CENTER

__________________________________________________________________

 

	For more than a century after Herschel (1785), astronomers lived 

essentially at the center of a galaxy not much more than 6000 LY across 

(illustration 18 in Jaki 1972). Herschel arrived at his result by counting 

stars as a function of apparent magnitude in various directions ("star 

gauging") and, according to Kopal (1971) increased the diameter to 20,000 LY 

in 1806. The issue of whether the spiral nebulae might constitute other island 

universes was discussed sporadically through the 19th century, but was not the 

focus of anyone's research. Simon Newcomb (1882; illustration 19 in Jaki 

1972), for instance, put "the region of the nebulae" immediately above and 

below a Herschel-like disk. It is widely believed that Newcomb was Walt 

Whitman's "Learned Astronomer," but this should probably not be held against 

either of them.

 

	Cornelius Easton's (1900; Berendzen et al. 1976 p.56) galaxy was also 

small and sun-centered, but he was the first to give the Milky Way spiral 

arms. An honest examination of the sky forced him to displace the center of 

the spiral pattern away from us by more than half the galactic radius in the 

direction of Cygnus, and his drawing gives the impression of a man struggling 

with the truth and losing. Parsecs gradually replace Light Years as the unit 

of choice between 1900 and 1920. Karl Schwarzschild's (1910) galaxy was 10 kpc 

across, 2 kpc thick, and sun centered, while Arthur S. Eddington (1912, 

picture p. 196 in Whitney 1971) put us 60 LY above the center of the galactic 

plane. Hugo von Seeliger, the most thorough counter of stars since Herschel, 

and many others, concurred (Seeliger 1911).

 

	Shapley (1918, 1919 and earlier references therein) shows a certain 

youthful exuberance in his distances - 67 kpc for NGC 7006 and 13.9 kpc even 

for M3. The centroid of his distribution slid from 13 to 25 kpc, with the 1919 

paper settling on 20 kpc and a total diameter at least three times that. 

Shapley's universe had precious little room for anything outside this enormous 

galaxy, and he attempted at one point (Shapley 1930) to describe the Milky Way 

as more like the Coma-Virgo cloud of galaxies than like a single spiral or 

disk system. This is also the purport of his remark, quoted in Russell, Dugan, 

and Stewart (1926) that, if the spiral galaxies are islands, the Galaxy is a 

continent. Anton Pannekoek (1919) agreed with Shapley in placing the sun far 

off center but in a smaller galaxy (Ro = 40-60,000 LY, d = 80-120,000 LY).

 

	At the time of the debate, Curtis's Milky Way was only 10 kpc across, 

with the sun at Ro = 3 kpc. Meanwhile, Kapteyn and van Rhijn (1920, Kapteyn 

1922) were counting stars more precisely than they had ever been counted 

before, but with no allowance for absorption by dust. Their first result was 

Ro = 0 and d = 24 kpc; the second Ro = 3 kpc, d = 17 kpc (shown on p. 24 of 

Berendzen et al. 1976). But Shapley's numbers dominated people's thinking very 

quickly. Sir Harold Spencer Jones (1923, General Astronomy), Sir James Jeans 

(1927, Astronomy and Cosmology), as well as Russell et al. 1927, vol. 2) place 

the galactic center 20 kpc away. Jeans describes the Milky Way and other 

spirals as having the relationship of a cake to a bunch of bisquits. All 

attempt to fit Kapteyn's "universe" in somewhere as a local stellar subsystem.

 

	Trumpler (1930, reproduced p. 93 of Berendzen et al. 1976) made a 

valiant attempt to declare all parties correct. His drawing shows a coordinate 

system centered at the sun in the middle of a slightly-tilted 10 kpc Kapteyn 

universe, but globular clusters scattered over an 80 kpc spheroid, centered 

about 18 kpc away from us.

 

	Jan Oort's discovery of galactic rotation quickly led to a new 

calibration of distance scales. His first version (shown in Oort 1927) 

reported Ro = 6300 +/- 2000 pc, soon revised upward to 10 kpc (shown in Oort 

1932). This value was widely used over the next 20 years and incorporated in 

many images (see, e.g., Bok 1937).

 

	Walter Baade (1953), however, looked again at the globular clusters 

and their RR Lyrae variables and settled on Ro = 8.16 kpc. This value, rounded 

off to 8.2 kpc, was generally accepted as the standard for reducing galactic 

rotation curves over the next decade (as shown by Westerhout 1956 and Kerr 

1962). Nancy Grace Roman (pr. comm. 1992), who attended the symposium where 

Baade shrank the galaxy, described herself as having gone to college at 10 kpc 

and to graduate school at 8.2 kpc.

 

	The present author did precisely the opposite; for in 1963 Oort (1964, 

cf. Schmidt 1965) moved us back out to 10 kpc. And there the official IAU set 

of galactic rotation constants kept us until the 1985 General Assembly in 

Bangalore, where the Commission on Galactic Structure (cf. Kerr and 

Lynden-Bell 1986) voted to reduce Ro to 8.5 kpc. This number is the average of 

along table that includes numbers between 6 and 11 kpc. Subsequent trends have 

perhaps been toward the small end of the range. Thus our present distance from 

the galactic center is quite close to the geometric mean of the numbers 

advocated by Shapley and Curtis in 1920.

 

5. SCIENTIFIC ISSUES IN 1920 AND THEIR RESOLUTION

_________________________________________________

 

	Shapley and Curtis disagreed to some extent on at least 14 

astronomical issues. These are presented in the following paragraphs in 

roughly the order in which they occur in the printed texts (Shapley 1921, 

Curtis 1921), which is neither in order of importance nor according to any 

other pattern a modern reviewer would be likely to choose. According to the 

actual texts reproduced by Hoskin (1976), no other additional scientific 

points were made during the main talks, though some may have arisen during 

Russell's rebuttal or other parts of the discussion, no record of which has 

been preserved. Each paragraph indicated an issue, what each disputant thought 

(or anyhow wrote or said), what we think now and sometimes why, and who should 

be counted the winner on each issue.

 

1. Resolved F, G, and K stars in globular clusters. Shapley believed they were 

giants like local F-K giants, with absolute magnitudes near -3, placing 

average globular clusters 10-30 kpc from us. Curtis said they were like the 

commonest sorts of stars around us, F-K dwarfs, with average visual magnitudes 

of about +7, putting the clusters at a kpc or two. As became unambiguously 

clear when the first 200" color-magnitude diagrams of globulars reached the 

main sequence turn-off (e.g. Sandage 1953), Shapley was essentially right on 

this one.

 

2. B stars in globular clusters. Shapley said they should have absolute 

magnitudes near 0, like nearby main sequence late B and early A stars. Curtis 

responded that something very strange must be going on, since the brightest 

blue stars in the solar neighborhood are brighter than the brightest red 

stars, while the opposite is true in the clusters. It took the insight of 

Walter Baade and his data gathered during the black outs of WWII to sort this 

one out, with the concept of two stellar populations. Each of the speakers was 

right about the particular point he emphasized.

 

3. Cepheids as distance indicators. Shapley used the relative 

period-luminosity relation found in the Large Magellanic Cloud with its zero 

point calibrated on a handful of Milky Way disk examples using statistical 

parallax. He noted that the nearby Cepheids of the cluster type (that is, RR 

Lyrae stars) are high velocity objects and must not be used for the 

calibration. Curtis responded that there was no evidence for a 

period-luminosity relation in the Milky Way, and that a larger sample, 

including some stars with geometric parallax measurements, even ruled it out. 

This was the point on which he said most firmly "more data are needed." When 

they came, Milky Way Cepheids did display a P-L relation, based both on 

secular parallaxes (or statistical) and on open cluster members. But the zero 

point was offset from the globular cluster one by more than a magnitude. This 

also was the work of Baade, who knew something was wrong the day (or rather 

night) he turned the 200" toward Andromeda and saw no RR Lyrae stars. Curtis 

was right about "more data" but wrong about what they would show -- he had 

placed too much faith in tiny geometric parallaxes, though he had more sense 

(paragraph 14) than to be misled by tiny proper motions. Shapley was right 

that Cepheids are generally good distance indicators.

 

4. Spectroscopic parallaxes in general. Shapley believed these could be 

trusted as long as you could see any of the line ratios indicative of giant 

surface gravities in nearby stars. Curtis believed they should be trusted only 

in the region of less than 100 pc where they had been calibrated. Errors and 

omissions expected (like some high latitude B stars), Shapley was right on 

this, though one shudders to think of the faith of eye required to see 

luminosity indicators like the ratio of 4215 (Sr II) to 4454 (Ca I) in spectra 

of individual globular cluster giants taken before 1920.

 

5. Interpretation of star counts. Curtis said, correctly, that star counts, 

straightforwardly interpreted, require a small Milky Way. His idea that spiral 

nebula dust existed as a ring around the stellar disk prevented him from 

suggesting absorption as relevant to the problem. Shapley did not address the 

issue, presumably because his use of globular clusters had already committed 

him to the "negligible absorption" camp, and he could, therefore, say nothing 

to rebut the point. Robert Trumpler (1930), by correlating apparent diameters 

of open star clusters with their apparent brightnesses revealed the importance 

of interstellar absorption (though Jesse Greenstein and others had come very 

close to discovering it earlier).

 

6. Stellar evolution theory. Shapley claimed that if and only if the globular 

clusters were put at large distances would their stars fit the Russell giant 

and dwarf theory and Eddington's models of gaseous giants. Curtis opined that 

spiral nebulae as a phase of stellar evolution didn't fit anywhere in any 

reasonable theory (remember protostellar nebulae were Out for solar system 

formation and encounters were In that year, and Jeans' idea that they were 

places where new stuff was pouring into the galaxy from Elsewhere had yet to 

be espoused and modified by Victor Ambarsumyan and others). While both points 

were true enough, we have to count Curtis the winner on this one, since we no 

longer adhere to the giant and dwarf theory!

 

7. Distribution of spiral nebulae on the sky. Shapley doesn't really mention 

this, but for a "single system" man, it was no more unreasonable for spirals 

to avoid the galactic plane than for OB stars to favor it. Curtis was forced 

to deal with the problem and concluded that it was "neither impossible nor 

implausible" for the Milky Way to have an occulting ring around it, as many 

edge-on spirals seem to, so that we would not be able to see nebulae in the 

plane. Curtis was closer to the truth than Shapley, but missed the critical 

point that stars and absorbing material are mixed together.

 

8. Nova brightness at maximum light. Both speakers agreed that "new stars" had 

been seen in the Milky Way and in several spiral nebulae. Shapley felt 

strongly that the implied real brightnesses would be totally ridiculous if the 

spirals were separate galaxies. Curtis said that, for four events with 

estimated distances in the Milky Way and a handful of novae in spirals, peak 

luminosity would be the same, provided the Milky Way had his preferred small 

size and the spirals were separate systems of similar physical diameter. He 

agreed that S Andromeda in 1885 was much brighter than this general run of 

events, said that Tycho's nova probably had been too, and concluded "a 

division into two classes is not impossible." One of the participants in our 

modern debate presumably feels the same way about the gamma ray bursters. 

Notice that Curtis was willing to trust a calibration based on four examples 

when he liked the answer, but not for the Cepheids, where he didn't. Two 

classes was, of course, the solution. Lundmark (1920) hinted at it, and Baade 

and Zwicky (1934) said it firmly from December 1933 onward, dubbing the 

brighter class super-novae (the hyphen disappeared the year Hale died; not 

causal). Curtis gets the points for this topic.

 

9. Nova mechanisms. Shapley suggested, seemingly with a straight face, that 

both the star and the nebulosity had existed to begin with, and that nebulae 

(with their large velocities) overtook and enveloped stars, producing nova 

events. He claimed to get the right rate of a few per year in the Milky Way 

from the numbers of stars and nebulae in his model universe. Curtis countered 

that the proposed mechanism would yield a rate of 1 per 500 years in 

Andromeda, where several had already been caught in the last 20 years. Once 

again, Curtis 1, Shapley 0.

 

10. The large, positive average velocities of the spiral nebulae. Shapley 

suggested the cause might be repulsion by radiation pressure from the Milky 

Way (a mechanism Russell showed to fail by many orders of magnitude the same 

year). Curtis simply proposed that large (mostly) positive wavelength shifts 

might somehow be intrinsic to the nebulae, and a large velocity also 

characteristic of the Milky Way. There are cases where "I haven't a clue" is 

the correct answer. It took the combined force of observations by Hubble, 

Milton Humason, and others and theoretical advances by Einstein, Alexander 

Friedmann, and others to come up with expansion of the universe as the 

explanation. Curtis over Shapley again, though perhaps not full marks. 

Incidentally, in case I forgot to mention it elsewhere, Einstein did not 

attend the 1920 debate, pace Florence (1994) and could not have, being still 

in Europe.

 

11. Properties of Galaxies, I. Shapley pointed out that the observed central 

surface brightnesses of spiral nebulae are much larger than anything seen in 

the Milky Way and the radial distributions of colors and surface brightnesses 

are different. Curtis remained silent on the issue. The answer, of course, is 

absorption and reddening, so Shapley was right about the data, but wrong about 

the interpretation. Love-love.

 

12. Properties of Galaxies II. According to Curtis, spiral nebulae have colors 

and line spectra a lot like those of star clusters, implying that the nebulae 

are mostly large assemblages of stars. Shapley did not mention this, and 

Curtis was right.

 

13. Central location of the sun. Shapley claimed this was an illusion, caused 

by the local star cloud now called Gould's belt. Curtis said it was God's own 

truth, and that our location kept us from readily seeing our own spiral arms. 

Once again, dust is an important part of the picture, but Shapley was nearly 

right.

 

14. Rotational proper motions of spirals as measured by van Maanen. Shapley 

said these were "fatal to the comparable galaxy theory." Curtis fully agreed, 

but said that you should never trust a proper motion of less than 0.1/yr for 

fuzzy things measured from a base line of 25 years or less. A round of 

applause for Curtis and sympathy for Shapley, who said later that van Maanen 

was his friend, so of course he believed him.

 

6. AFTERMATH OF THE DEBATE

__________________________

 

	Immediate reaction to the two lecturers was undoubtedly driven by the 

two men's styles of public speaking. Comments have come down to us indicating 

that Curtis was by far the more experienced lecturer and expounder to the 

public. He had, at any rate, taught at Detroit High School, Napa College 

(California), and College of the Pacific (first Latin and Greek, later 

mathematics and astronomy) for about five years before seeking his PhD (McMath 

1942, Stebbins 1950). Russell's private reaction (Hoskin 1976) was that 

Shapley ought to be persuaded to offer a lecture course to hone his skills in 

this direction. From our modern vantage point, it is hard to see things this 

way. Shapley springs to mind as the man for whom, rightly, the AAS Shapley 

Lectures are named, while Curtis is the man with the rimless glasses (and 

without the hair) who was prone to describe astronomical hypotheses as "not 

impossible" and "neither implausible nor impossible," while intoning a refrain 

of "more data are needed." Shapley, however, appears literally to have read 

his paper (from a typewritten text with long- and short-hand corrections), 

while Curtis had his lecture notes on slides. He might have even used overhead 

plastics, like the 1995 debaters, if they had existed in his time.

 

	Although the participants continued to speak and write among 

themselves about the "famous debate," "memorable set-to," and "memorable 

discussion" for several years after 1920 (Hoskin 1976), the event seems to 

have attracted very little attention in the popular or scientific press. 

Berendzen et al. (1976) were able to locate only one contemporary report 

(given by a historian of science, Peter Doig, at the December 1921 meeting of 

the British Astronomical Association). Several contemporary reviews of 

distance scales refer to the work of one or both debaters, but not to the 

debate, and conclusions drawn are essentially those held by writers before 

April 1920. A splashy headline on a May, 1921 issue of the Boston Sunday 

Advertizer (p. 72 of Berendzen et al. 1976) refers only to Shapley's work on 

the distance scale and seems to have been featured primarily because he was, 

by then, a "Harvard astronomer." As noted in Sect. 4, Shapley's structure for 

the Milky Way was rapidly adopted by writers of astronomical textbooks, but 

without any reference to the 1920 event or 1921 publications.

 

	At the time of Curtis's death, the discussion at the NAS was 

sufficiently forgotten that McMath's (1942) obituary makes no mention of it. 

Shapley, in his 1969 autobiography, similarly averred that it had for long 

escaped his memory. The first commemorative account I have seen is Stebbins' 

(1951) talk at the dedication of the Curtis Memorial Telescope in 1950. Otto 

Struve (1960), writing on the 40th anniversary, spoke of "a historic debate," 

and described (William) Albert Whitford (PhD U. Michigan 1942) and his 

students at Wisconsin as having restaged the event several times around 1950. 

Following Struve's account, others (based largely on the published texts) 

appear in many books and articles. Struve could not have been at the original 

debate. Stebbins (who was there to welcome Curtis to Lick in 1902) could have 

been, but apparently was not.

 

	Each debater at some point expressed the opinion that he had won, 

perhaps not surprising, given the near-equality of their scores (Sect. 5). The 

question of correct distance scales within (and without) the Milky Way has 

been iterated on many times since 1920. According to recent rounds, Shapley's 

galaxy was too big and Curtis's too small, but, more seriously, centered far 

too close to the sun. The sort of sketch map most of us would draw has not 

changed much since that of Plaskett (1939, shown as fig. 202 in Berendzen et 

al.)

 

	The question of the existence of separate, external galaxies, island 

universes, or whatever you want to call them, was resolved much more cleanly. 

A contribution by Edwin Hubble to the December/January 1924-25 meeting of the 

AAAS (read by H.N. Russell) announced the presence of Cepheids in the nebulae 

at apparent brightnesses that put them firmly outside even Shapley's bloated 

Milky Way. The result had actually been published in the 23 November 1924 New 

York Times, without attracting much attention. But Stebbins, Russell, and 

others who were at the AAAS meeting felt that the issue had been fully 

resolved.

 

	The debaters apparently agreed. Curtis (quoted in Berendzen et al. 

1976, p.138) wrote in April, 1925 "I have always held this view [that spirals 

are separate galaxies], and the recent results by Hubble on variables in 

spirals seems to make the theory doubly certain." This sounds like a calm, 

reasoned reaction, appropriate to a scientist who had distrusted earlier 

conclusions based on Cepheids (and I cannot say whether the disagreement 

between subject and verb was Curtis's or accidentally introduced by Berendzen 

et al.).

 

	Shapley's predictably much more flamboyant reaction was recalled long 

after by Cecilia H. Payne-Gaposchkin, who had come as Harvard's first PhD 

student in astronomy in 1923 (Haramundanis 1986, p. 209). She was in Shapley's 

office when a letter arrived from Hubble, describing the period-luminosity 

relation for Cepheids in M31. "Here is the letter that destroyed my universe," 

said Shapley, holding it out. She also recalled him saying, "I believed in van 

Maanen's results . . . after all, he was my friend." And she herself resolved 

(Haramundanis 1986, p. 227) that she would "not accept the conclusions of 

another astronomer simply because I am fond of him, or reject them because I 

dislike him (though I admit there is a temptation here)."

 

7. LESSONS FOR TODAY AND TOMORROW

_________________________________

	(AND POSSIBLY UP TO NEXT TUESDAY OR THEREABOUTS)

	________________________________________________

 

	It is possible to discern, or perhaps imagine, several patterns in the 

1920 debate and long-term repercussions. First, Curtis and Shapley each seem 

to have got things more or less right when they relied on data they had 

collected for themselves (Shapley's photometry of stars in globular clusters; 

Curtis's images of spiral nebulae), and to have gone astray when they 

attempted to make use of data assembled by others. This is not a happy omen 

for the 1995 debate.

 

	Second, conclusions that they drew by attempting to rely on 

astrophysical theory did not have a very high batting average. You can argue 

about just what belongs in this paragraph, but Shapley invoking the 

giant-dwarf theory of stellar evolution, radiation pressure for the large 

redshifts of spiral nebulae, and an encounter hypothesis for nova events, and 

Curtis attempting a sort of generalized Copernican approach to stellar 

populations strikes me as good examples. Most of us would, of course, say that 

correct theories do not get you into this kind of trouble, but rather add 

strength to observational conclusions by enabling you to understand them. This 

is what Eddington had in mind when he said he refused to believe an 

observation until it was confirmed by theory. But then, how much confidence 

should we have that any of the astrophysical theory brought to bear on the 

bursters so far is of this type?

 

	Third, each of the 1920 protagonists had hold of part of the truth and 

so could claim partial victory. Other famous scientific disputes have ended 

this way, for instance that between the 18th century neptunists (who believed 

only in sedimentary rocks, laid down under Neptune's oceans) and the 

plutonists (who believed only in igneous rocks, rising up from Pluto's 

underworld). Both, of course, exist. I am inclined to suspect that the claims 

and counterclaims of star bursts vs. monster central engines to modelactive 

galaxies will prove to be like this.

 

	Not all scientific disputes can end in such mergers or compromises. 

There is nomiddle ground between a planetary system co-forming with the sun 

and one draggedout of an already-established star by an intruder. Nor have 

various attempts to combine the virtues of standard hot big bang cosmology with 

those of steady state succeeded.

 

	What about the gamma ray bursters? Could both galactic and very 

distant (and perhaps even "other") sources lurk among the classical events, 

with perhaps very different physical mechanisms at work in each? Since I have 

always felt that the popular image of the Curtis-Shapley debate gave the elder 

astronomer rather shortshrift, I would like him to have the last word, taken 

from his comments on novae, "A division into two magnitude classes is not 

impossible."

 

ACKNOWLEDGMENTS

_______________

 

	First and foremost, we all thank Robert Nemiroff for the enormous 

amount of work he has put in to organize this event, comprising among the 

wishes of people at least as discordant as their scenarios. I am personally 

deeply grateful to Ralph Belknap Baldwin, Dorrit Hoffleit, Vera Cooper Rubin, 

and Richard Sears for sharing their memories and records that still connect 

us, tenuously, to the era of the original Curtis-Shapley debate, and to 

Katherine Gaposchkin Haramundanis for an invitation to write the introduction 

to the second edition of her mother's biographical volume, which led to my 

rereading it at just the right time.

 

                     _________________________________

 

[* The mistake of placing the debate in 1921 is curiously common. Bok (1972) 

does it in his obituary of Shapley, as do several of the secondary accounts of 

the debate. And Florence (1994) manages to make several chapters out of the 

events of "April 1921." The cause is, presumably, the date of the publication 

and the fond belief that refereeing didn't take so long in those days! ]

 

REFERENCES

__________

 

Abt, H.A. 1995. Preprint

Baade, W. 1953, in Symposium on Astrophysics (Ann Arbor: U. Michigan) p. 25

Baade, W. and F. Zwicky 1934. Phys. Rev. 45, 138; Proc. NAS 20, 254 & 259

Baldwin, R.B. 1949. The Face of the Moon (U. Chicago Press)

. . . . 1995. Private communication

Berendzen, R. et al. 1976. Man Discovers the Galaxies

	(NY: Science History Publications)

Bok, B.J. 1937. The Distribution of Stars in Space (U. Chicago Press)

. . . . 1972. QJRAS

Curtis, H.D. 1921. Bull. NRC. 2, 194

Easton, C. 1900. ApJ 12, 136

Eddington, A.S. 1914. Stellar Movements (London: Macmillan) p.31

. . . . 1926. The Internal Constitution of the Stars

	(NY: Dover Reprints, 1959)

Florence, R. 1994. The Perfect Machine (NY: Harper Collins)

Haramundanis, K. (ed.) 1984. Cecilia Payne-Gaposchkin

	(Cambridge University Press)

Herschel, W. 1785. Phil. Trans. Roy. Soc. 75, part 1

Hoffleit, D. 1995. Private communication

Hoskin, M.A. 1976. J. Hist. Astron. 17, 169

Jaki, S. 1972. The Milky Way (NY: Science History Publications)

Kapteyn, J.A. 1922. ApJ 55, 65

Kapteyn, J.A. and P.J. van Rhijn 1920, ApJ 52, 23

Kerr, F.J. 1962. MNRAS 123, 327

Kerr, F.J. and D. Lynden-Bell 1986. MNRAS 222, 1023

Kopal, Z. 1971. Widening Horizons (NY: Taplinger) p. 21

. . . . 1972. Nature 240, 429

Lundmark, K. 1920. Sven. Vetkapsakad. Handingar 60, No. 8

McMath, R.R. 1942. PASP 54, 69

Newcomb, S. 1882. Popular Astronomy - 4th Ed. (NY: Harper & Bros.) p. 493

Oort, J.A. 1927. BAN 4, 69

. . . . 1932. BAN 6, 279

. . . . 1964. IAU Inf. Bull. 11

Osterbrock, D.E. 1993. Pauper and Prince (Tucson: U. Arizona)

. . . . 1995. Private communication

Pannenkoek, A. 1919. MNRAS 79, 500

Pinckney, R. 1995. Am. Her. Sci. & Tech. 10, No. 4, 29

Plaskett, J.S. 1939. Pop. Astron. 47, 255

Russell, H.N., R.S. Dugan, and J. Q. Stewart 1926. Astronomy

	(Boston: Ginn & Co.)

Sandage, A.R. 1953. AJ 58, 61

Schmidt, M. 1965. in Galactic Structure (U. Chicago Press) p. 513

Schwarzschild, K. 1910. AN 185, 81

Sears, R. 1995. Pri. communication

Seeliger, H. von 1911. Sitzungsberichte der Akad. Wiss. zu Muenchen, p. 413-61

Shapley, H. 1918. ApJ 48, 176

. . . . 1921. Bull. NRC 2, 171

. . . . 1930. Star Clusters (Harvard Obs. Monograph No. 2, McGraw Hill) p. 209

. . . . 1969. Through Rugged Ways to the Stars (NY: Scribner) p. 79

Shapley, H. and M.B. Shapley 1919, ApJ 50, 116

Stebbins, J. 1951. Pub. Obs. U. Mich. 10, 1

Struve, O. 1960. Sky and Telescope 19, 398

Struve, O. and V. Zebergs 1962. Astronomy of the 20th Century (NY: Macmillan)

Sweeney, D. 1995. Am. Her. Sci. & Tech. 10, No. 4, 20

Trumpler, R. 1930. LOB 14, 154 (no. 420)

Westerhout, G. 1956. BAN 13, 201

Whitney, C.A. 1971. The Discovery of Our Galaxy (NY: Knopf)

Wright, H. 1966. Explorer of the Uiverse (NY: EP Dutton)