38 



DISCOVERY 



magnitudi'S and thus to compute the parallaxes of 

 nearly a hundred of these. Fourthly, in the large 

 class of variables known as Cepheids, the length of 

 the period of variation has been ascertained to depend 

 on the absolute brightness. The mean parallaxes of 

 the nearer of these variables being known, it is possible 

 to determine the absolute magnitude, and thus the 

 distance of cverj' Ccpheid variable in the heavens. 

 Finally, Russell and others have indicated theoreti- 

 cally a method of ascertaining the minimum distance 

 of the bright helium stars of type B. Only stars of 

 very large mass arc able to attain to the unusually high 

 temperature of B-typc stars ; hence a B-star, however 

 faint it may appear, must have a certain minimum 

 absolute magnitude. All these new methods of deter- 

 mining and estimating distance have contributed to a 

 great extension of our knowledge of the extent and 

 structure of the universe. Further, in recent years 

 it has become possible to ascertain the hypothetical 

 spectral type of the ver>' faint stars. As is well known, 

 the photographic plate is more sensitive than the 

 eye to certain wave-lengths of light ; the difference 

 between the photographic and the visual magnitude 

 of a star is therefore due to the colour of the star, 

 and is called the colour-index. Thus it is possible to 

 determine the colour, and hence the approximate 

 spectral type, of verj- faint stars whose spectra cannot 

 be directly observed. 



As a result of these new methods of investigation 

 several researches of the utmost importance have been 

 undertaken within the last few years. In 1916 Pro- 

 fessor Charlicr, of Lund, published a paper on " The 

 Galaxy of the B-type Stars," in which he started 

 from the assumption that the B-stars do not differ 

 much in absolute magnitude. Having made this 

 assumption, he was enabled to determine the distances 

 of stars of this type brighter than the fifth magnitude. 

 He found that these stars form a well-defined flattened 

 cluster — whose ,plane is very similar to that of the 

 Galaxy — with its centre of gravity in the constellation 

 Carina. The farthest limits of the cluster Charlier 

 found to be at a distance of 750 parsecs • — roughly 

 corresponding to about 2,000 light-years. " We are," 

 said Charlier, " in a position to get, with the help of 

 the B-stars, what might appropriately be called a 

 skeleton image of the Milky Way." In other words, 

 Charlier assumed that the system of B-stars was 

 co-extensive with the stellar universe. 



In 1913 Professor Hertzsprung, of Potsdam, from a 

 study of the Cepheid variable" stars in the Lesser 

 Magellanic Cloud, estimated the absolute magnitudes 

 of these stars, and thus their distances and the distance 



' The parsec is the distance which corresponds to a parallax 

 of one second of arc. It is equal to nineteen billions of 

 miles. 



of the cluster. The distance at which he arrived was 

 10,000 parsecs — corresponding to about 30,000 light- 

 years. In the following year Dr. Harlow Shapley, 

 the brilliant American astronomer, commenced, with 

 the aid of the great 60-inch reflector of the Mount 

 Wilson Observatory, his studies on the colours and 

 magnitudes in stellar clusters. The clusters investi- 

 gated were of two classes, the compact globular clusters 

 and the open galactic clusters which are undoubtedly 

 condensations of distant Milky Way stars. The 

 results reached by Dr. Shapley during the first two 

 years of his investigation were sufficiently startling. 

 In the galactic star-clouds in the vicinity of the 

 open cluster Messier 11, he discovered faint blue 

 stars obviously of type B. " The cluster-stars," 

 wrote he in April 1917, " are probably giants in 

 luminosity, and accordingly the distance of the group 

 must be of the order of 15,000 light-years. The wide 

 dispersion in magnitude of both blue and red stars 

 indicates a similarly great distance for the neighbour- 

 ing galactic clouds. It suggests that the extent of 

 the stellar clouds in the line of sight is relatively very 

 great — in fact, the depth may be as great as, or greater 

 than, the distance to the nearer boundary." 



No less remarkable were Shapley's preliminary 

 conclusions regarding the brighter globular clusters. 

 The first of these exhaustively studied was the great 

 cluster in Hercules (Messier 13). In his paper on 

 this cluster, dated August 1915, Shapley estimates its 

 distance from " considerations of its variable-stars, 

 its fragmentary luminosity-curves, and the average 

 apparent brightness of certain colour groups," as 

 100,000 light j-ears, a distance much greater than had 

 ever been imagined for the cluster.- 



Dr. Shapley's subsequent researches have been given 

 to the world in the series of " Contributions from the 

 Mount Wilson Observatory," and have been admirably 

 summarised in his recent pamphlet on " Star-Clusters 

 and the Structure of the Universe." By means of 

 the various methods of determining absolute magni- 

 tude and distance, Shapley has fixed the positions 

 in space of eighty-six globular clusters. He finds 

 these to be " cosmic units," sub-systems dependent 

 on the greater galactic system, at enormous distances 

 from our earth. The nearest. Omega Centauri, is 

 20,000 light years away, and the most distant — known 

 as N.G.C. 7006 — is so distant that light requires 

 220,000 years to travel from the cluster to the earth. 

 Shapley finds the clusters arranged in a certain 

 symmetrical way, which indicates that they are 

 dependents of the sidereal system — as Professor 

 Perrine suggested in 1917 — and that the sun is not 

 near the centre of the universe, as was generally 



• A later measure by Dr. Shapley gives a distance of 

 36,030 light-years. 



