ADVISORY COMMITTEE ON ASTRONOMY 137 



of this instrument differ in no important respect from those of the 

 Mills spectrograph. The complete realization of laboratory condi- 

 tions in stellar spectroscopy is the essential purpose of the experi- 

 ments now being pursued at the Yerkes Observatory. A large grat- 

 ing spectrograph, rigidly mounted in a constant- temperature labora- 

 tor>', will receive the star's light from a 30-inch ccelostat telescope. 

 With this apparatus it should be possible to photograph the spectra 

 of the brightest stars with a dispersion equal to that now employed 

 for the Sun. 



Spectra of Planets. — Excepting the objective prism, which can be 

 used advantageously only with objects of small angular magnitude, 

 the instruments employed in stellar spectroscopy can be equally well 

 used in the study of nebulae, planets, comets, and other heavenly 

 bodies. The planets already afford a fruitful field of investigation, 

 which promises to become more important in the future. As the 

 light from the Sun reflected to us by the planets must pass twice 

 through their atmosphere, the absorption thus produced can be 

 detected in their spectra. In this way it has been found that 

 Saturn possesses a dense atmosphere, which is absent or extremely 

 'rare on the rings. Perhaps the most interesting application of the 

 spectroscope to the study of planets is the determination of their 

 period of axial rotation. B}- proving spectroscopically that the 

 matter composing the inner edge of Saturn's rings completes a rota- 

 tion about the planet in a shorter time than the matter at the outer 

 ^<i^<c, Keeler showed observationally, as Maxwell long before had 

 done theoretically, that the rings must consist of discrete particles. 

 As soon as more powerful instruments become available, the dis- 

 puted question of the rotation period of Venus will doubtless be 

 settled by spectroscopic means. 



Spectra of NcbidcB. — Comparatively few of the nebulas liave been 

 investigated spectroscopically, and a most important ^€iC\. of research 

 lies open here. The discovery of the terrestrial counterpart ot the 

 chief nebular litie, the chemical and physical differences which char- 

 acterize various parts of the same nebula, and the corresponding 

 differences of radial velocity, are among the problems that most 

 readily suggest themselves. The certain nature of the spectrum of 

 5uch an object as the Great Nebula in Andromeda should be ascer- 

 tained without further delay. 



The Sun. 



A complete knowledge of the phenomena and physical constitu- 

 tion of the Sun is of fundamental importance in connection with the 



