RECENT ADVANCES IN SCIENCE 355 



infinitesimal geometry of space, see R. Occhipinti (ibid. 335-40, 

 373-80; 1917, 55, 31-4), P. Calapso (Ann. di Mat. 1917, 26, 

 151-90), L. Bianchi (ibid. 199-223), S. Finsterwalder (Abh. 

 Milnchen, 28), and E. Cartan (Bull. Soc. Math. France, 191 6, 

 44, 65-99). 



ASTRONOMY. By H. Spencer Jones, M.A., B.Sc, Royal Observatory, 

 Greenwich. 



The Classification of Stellar Spectra. — The first four hours of 

 Right Ascension of the Henry Draper Catalogue have been 

 published as vol. 91 of the Annals of the Harvard College 

 Observatory. This is the first instalment of the results ob- 

 tained from an extensive and laborious piece of work which 

 was planned several years ago. The importance of a knowledge 

 of the spectra of the stars was beginning to be realised, for it 

 was found that many properties of the stars depended upon the 

 type of their spectra. The Director of the Harvard Observatory 

 had the foresight to see that a classification according to the 

 class of their spectra of as many stars as possible would be of 

 great value ; moreover, the only material which was available 

 for such an investigation which should cover in a uniform 

 manner the whole sky was contained in the Harvard collection 

 of photographs. 



The task would have been a heavy one, even if the classifi- 

 cation had been confined to the comparatively bright stars, 

 but it was wisely decided to extend it to stars as faint as possible 

 so as to make its early repetition unnecessary. The photographs 

 available had been taken for the northern stars with the 8-inch 

 Draper telescope of the Observatory, and for the southern stars 

 with the Bache telescope, mounted in Arequipa, Peru, a prism 

 being placed over the objective of the telescope in each case. 

 In this way, spectra of all the stars of sufficient brightness in 

 the field of the instrument were obtained on one plate : the 

 photographic image of each star was dispersed by the prism 

 into a line, and these lines were broadened into bands by slightly 

 altering the rate of the telescope clock. Thus, spectra of all 

 stars of sufficient brightness, without exception, were available. 



The classification was entrusted to Miss Annie J. Cannon, 

 and was commenced in 191 1 and practically finished by the 

 end of 191 5. The classification adopted was that previously 



