412 
ASTRONOMY: H. N. RUSSELL 
(h) As van Maanen^^ and others^*^ have shown, the proper motions of 
some spiral nebulae — both of the mass as a whole and of the condensa- 
tions in the arms relatively to the centre — are apparently large enough 
to be determined by the careful comparison of plates taken only a few 
years apart. This opens up another wide field of study, and will make 
it possible before long to determine the mean parallax of many such 
nebulae by comparison of the proper motions and radial velocities of 
their nuclei. There is also reason to hope that the distances of some 
individual nebulae, which are seen at a suitable angle, can be deter- 
mined by comparing the radial and transverse components of motion 
along the arms. Enough is already known to convince us that the 
distances of these nebulae must be measured in thousands of parsecs, 
and their diameters in parsecs, and that direct measures for parallax 
are utterly hopeless. 
{c) Photometric measures, both of the total light of the spirals and the 
relative brightness of their parts, would be of value, especially if accom- 
panied by determinations of color. Seares^^ has recently shown that the 
outer convolutions are far bluer than the centre — which is the part that 
shows the spectrum of solar type. Spectroscopic observations of these 
outer regions, if possible, would <be of great interest. Another matter 
calHng for further study is the nature of the dark bands which cross 
many nebulae which appear to be spirals seen edgewise, and look as if 
they were due to the interposition of opaque material in the outer 
regions of the nebula. 
{d) The distribution of spiral nebulae in the heavens — so utterly differ- 
ent from that of any other objects — ^may be explainable when their real 
distribution in space is even partially known. It is hardly time as 
yet to consider the greater question of their real nature, except to note, 
with van Maanen,^^ that, unless they are in process of very rapid 
dissipation into space, their masses must be exceedingly great. 
15. Finally, it must not be forgotten how important a place theoretical 
investigations will occupy in the solution of the larger problems of sidereal 
astronomy. The increasing observational data are already furnishing 
just those guides which point the skilled mathematician in the right 
direction, and these indications have been very successfully followed, 
especially by certain members of that 'Cambridge school' which com- 
bines keen mathematical analysis with a thorough knowledge of modern 
physics. Results of remarkable generality have already been obtained. 
In the field of stellar evolution, Eddington^^ j^^g worked out in de- 
tail the importance of radiation pressure in determining the conditions 
of internal equilibrium of the stars, and the approximate equality in 
