ADVIvSORY COMMITTEE ON ASTRONOMY 1 33 



luminous objects to be studied witli certainty served as a foundation 

 for the recent remarkable development of this subject. Secchi, 

 shortly after the chemical constitution of the Sun had been discov- 

 ered, placed a prism over the objective of his telescope in the manner 

 instituted by Fraunhofer, and proceeded to survey the spectra of the 

 stars. In 1863 he published his first classification, v/hich showed 

 that white, yellow, and red stars have characteristic spectra, so re- 

 lated as to indicate a general line pf development. A short time 

 later Riitherfurd, who had initiated such work in the United States, 

 published a similar classification. In 1864 Huggins entered upon 

 his career as a pioneer in celestial spectroscopy. The record of his 

 discoveries is too extensive for more than the barest reference here. 

 He first directed bis attention to stellar spectra, and after making 

 visual observ^atious which established the presence of iron and other 

 elements in the stars, he attempted to record their spectra by pho- 

 tography. His first plates, though they contained impressions of 

 the spectra, did not reveal the lines. These lines were first success- 

 fully photographed in 1872 by Dr. Henry Draper, who actively pur- 

 sued this work until his death in r 882 . In 1876, with more powerful 

 instrumental means, Huggins returned to the task with great suc- 

 cess. Thanks to the use of suitable refractive and dispersive media^ 

 his photographs reach far into the ultra-violet. They revealed at 

 once the characteristic series of lines in the spectrum of hydrogen > 

 previously unknown even in laboratory researches. The publication 

 by Balmer of an empirical formula which accurately represents the 

 lines of this series stimulated the studies of spectral series by Kayser 

 and Runge, Rydberg, and otjiers, which have so greatly increased 

 tlie range and significance of spectroscopic research. 



Two fields of investigation, both successfully cultivated in recent 

 years, were opened up by Huggins. His identifications of spectral 

 lines were signiScunt, not merely in disclosing the chemical compo- 

 sition of stellar atmospheres, but also in suggesting relationships of 

 star to star which could be possible only on the assumption of de- 

 scent from a similar ancestor. The investigation of stellar evolu- 

 tion, foreshadowed in the classifications of Secchi and Rutherfurd, 

 was thus placed upon a sure footing. Through the researches of 

 Huggins, Yogel, Duner, Pickering, L-ockyer, and others, the line 

 of development frotn gaseous nebulae through Sirian and solar stars 

 to the cooling red stars has been traced out with few elements of 

 uncertainty. Where uncertainty still exists, it almost invariably 

 arises from the fact that many of the most interesting and impor- 



