ASTRONOMY: H. N. RUSSELL 
393 
relation between the age or evolutionary stage of a star and its position 
and motion within the galactic systems. The latter connects the prob- 
lems of stellar and galactic evolution in such a way that any notable 
advance in the solution of one is likely to be of aid in that of the other, 
while an unfounded assumption regarding either will probably confuse 
the discussion of both. 
///. Clusters and Nebulae. — So little has been known of these objects 
until very recently that the problems which they present can hardly 
yet be coordinated into a single statement. Among the most obvious 
are: 
(1) The relations of clusters and nebulae to the galactic system. It 
now appears probable that the galactic system is very much larger than 
was supposed a few years ago, and that not only the irregular clusters, 
and the gaseous nebulae, both planetary and extended, but also the 
globular clusters, and probably the Magellanic Clouds, lie within its 
confines. But the relations of the spiral nebulae are still uncertain. 
(2) Motions and dynamical relations within clusters, especially 
globular clusters, and explanation of the law of distribution of stars in 
such clusters. 
(3) Nature of the gaseous nebulae, especially of ^nebulium,^ and cause 
of their luminosity. Internal motions in gaseous nebulae. 
(4) Nature of spiral nebulae, and explanation of the rapid motions of 
their parts. 
In all these cases a persistent attempt should be made to account for 
the observed phenomena by means of the known properties of matter 
and forces of nature, and the existence of unknown forces should be 
postulated only if there is apparently no escape from the necessity of 
doing so. 
It may now be profitable to survey rapidly the different fields of as- 
tronomical investigation, and consider the bearing of various researches 
— some now under way, some practicable at present, and others desir- 
able if means for effecting them can be devised — upon these general 
problems. 
1. Spectra. — It seems to be increasingly clear that the master-key 
to these problems, so far as they have yet been formulated, lies in the 
investigation of the spectra of the stars and other bodies, and the cor- 
relation of their other characteristics with the spectra. Fortunately, 
the spectra are among the few characteristics which can be investigated 
independently of any knowledge of the distances of the various bodies, — • 
and, indeed, of the distances themselves, except for the limitation arising 
from the faintness of most of the remoter objects. 
