36 



DISCOVERY 



obscrs'atioiial treasures of the past two centuries, 

 which Dr. Abbot has called " that precious store of 

 exact knowledge whose value time cannot impair, 

 but can only enhance." The observations of Bradley, 

 for instance, become more precious with the passage of 

 the years ; so do the tabulated surveys of the Herschels, 

 of Bessel, and of Argclander and Schonfeld, the 

 photographic surveys of Gill, and the pioneer spectro- 

 scopic work of Secchi, Vogcl, and Pickering. The 

 nineteenth-century astronomers were largely occupied 

 in amassing observations ; in the twentieth, we are, 

 in some measure at least, reaping the fruits of their 

 labours. A second factor in the remarkable progress 

 of this branch of astronomy is the rapid development 

 of the powers of the telescope — and consequently of 

 the spectroscope — during the last thirty years. The 

 great refractors of the Lick, Lowell, and Yerkes Ob- 

 servatories in the United States have been employed — 

 the two former in superb climates — with the highest 

 advantage to stellar astronomy. More recently the 

 great reflectors of Mount Wilson have carried us still 

 further into the depths of space. 



As remarked in the previous article, astronomers in 

 the nineteenth century were obliged to concentrate 

 on the study of stellar distribution ; comparatively 

 few stellar distances and motions had been determined 

 with accuracy. By the opening years of the century, 

 however, it was evident that the investigation of 

 stellar motion was to prove a valuable line of attack 

 on the ultimate problem. In 1904 Professor Kapteyn, 

 of Groningcn, read at the Astronomical Congress at 

 St. Louis, U.S.A., a paper of far-reaching importance, 

 in which he announced the discovery of star-streaming. 

 .\s is well known, there are two components in the 

 motion of each star : (i) the parallactic motion due 

 to the movement of the sun, carrying w^ith it the earth 

 and the other planets, and (2) the individual proper 

 motion. Formerly it was believed that after the 

 parallactic motions were eliminated the individual 

 motions were more or less at random. This assump- 

 tion Kapteyn showed to be erroneous. After elimin- 

 ating the parallactic motion from the motions of the 

 bright stars in the catalogue of Bradley, revised by 

 .\uwcrs, Kapteyn found that the motions of the stars 

 fell into two opposite directions in the galactic plane. 

 This result he interpreted as indicating that the 

 brighter and nearer stars belong to one or other of 

 two vast interpenetrating star-streams. 



Shortly afterwards these results were abundantly con- 

 firmed by the independent work of Professor A. S. 

 I'lddington, then chief assistant at Greenwich, now 

 director of the Observatory at Cambridge. In his first 

 paper on " The Systematic Motion of the Stars," read 

 before the Royal Astronomical Society in November 

 1906, Eddington made what he called " the attempt 



to subject Kapteyn's theory to a quantitative test 

 by examining the Greenwich-Groombridge propcT 

 motions. The follosving reasons, among others, seemed 

 to make this desirable. The Bradley stars are all 

 bright stars ; of those of the Groombridge catalogue 

 a large number fall between the seventh and ninth 

 magnitudes. The 2,500 stars of Bradley's catalogue 

 cover most of the sky ; the 4,500 stars of Groom- 

 bridge are all within 52^ of the north pole." The 

 result of the investigation was a complete confirmation 

 of Kapteyn's theory. Eddington ascertained that 

 the velocity of the first drift of stars relative to the 

 sun was very much greater than that of the second, 

 and that there was no appreciable difference in the 

 main distances of stars of the two drifts. 



A further stage was reached by the paper which 

 Sir F. W. Dyson — then Astronomer-Royal for Scot- 

 land — read before the Roj-al Society of Edinburgh in 

 1908. His investigation covered 157 stars within 52' 

 of the north pole, taken from Groombridge's cata- 

 logue ; 165 stars within 50° of the South Pole from 

 the Cape Astrographic Catalogue, from Porter's Cin- 

 cinnati Catalogue, and the Cape General Catalogue ; 

 and 778 stars within a belt extending 30° north and 

 south of the equator, from the Cincinnati Catalogue. 

 The results of this study of stars, scattered all over 

 the sky, were strongly in support of the two-drift 

 hypothesis. " The quick-moving stars here con- 

 sidered show the streams in a very pronounced light." 

 Confirmation of the two-stream theory was also forth- 

 coming from Professor Campbell's investigation of 

 radial motion. Dr. Halm, of the Cape Observatory', 

 not onlyconfirmed the hypothesis by his own researches, 

 but came to the conclusion that there were not only 

 two drifts, but three. 



Various hypotheses were put forward to explain the 

 existence of the two star-streams. Local streams had, 

 of course, been known since Proctor pointed out the 

 community of proper motion of five of the stars of 

 the Plough ; but the work of Kapteyn, Eddington, 

 and Dyson indicated that the great mass of the 

 brighter stars belonged to two or three streams, and 

 this called for a theoretical explanation. The problem 

 before the theoretical astronomer was to preserve the 

 unity of the universe. In 1908 Schwarzschild, of 

 Potsdam, put forward his ellipsoidal hypothesis, and 

 in 1912 Turner, of Oxford, outlined a new theory re- 

 conciling the observed duality of the universe with 

 the unity which theory seems to demand. In 191 1 

 Eddington remarked, however, that " the two star- 

 streams probably involve at least half a million of 

 the stars around us ; but there has never been any 

 evidence that they prevail in the extremely remote 

 parts." This has been confirmed in recent years, and 

 the two-drift phenomenon finds its explanation in the 



