President's Address. 
57 
be derived from a study of the radial velocities in regard to the sys- 
tematic structure of the universe. 
A year or two must elapse before the present observing programme is 
completed, but a preliminary discussion of the observations already secured 
in combination with the published results derived from similar observa- 
tions in the Northern Hemisphere has revealed the existence of anomalies 
similar to those found from the study of the transverse motions — anomalies 
which can only be reconciled with the two-drift hypothesis put forward 
by Kapteyn, by the further hypothesis that though both drifts pervade the 
whole sky, they are not similarly distributed throughout it. 
At present, through scantiness of material, from a study of the radial 
velocities, we have been able to do little more than discriminate between 
the two halves of the sky, which contain respectively the greatest and the 
least proportion of second drift stars. It is, however, a fact of some 
significance that the former corresponds very closely with that hemisphere 
which contains the Milky Way, suggesting the phenomenon that Kapteyn's 
second drift might be identified with the galaxy. It was with a view to 
examining this suggestion, in the light of the evidence which could 
be secured from the transverse motions of the Bradley stars, that the 
discussions I have sketched to you to-night were undertaken by Dr. 
Halm. 
While they have established almost beyond question the rough features 
of distribution demanded to reconcile the radial velocity determinations, 
they further point to an even more detailed correspondence between the 
distribution of galactic stars and the distribution of stars of the second 
drift, leaving but little doubt as to the identity of this second drift with 
the galaxy. 
I have pointed out how it is this second drift which exhibits evidence 
of structural unity. As regards the Milky Way, the mere appearance on 
any fine night affords evidence of a similar character, and it is on this 
account that we have been able to identify the Milky Way with the 
second drift rather than the first. 
The significance and origin of this structure is as yet obscure, but the 
more its details are elucidated and the essential features established 
the nearer are we to an answer to the question — 
What is the Milky Way? 
To revert to my original text, I have endeavoured to point out to you 
the methods of research by which an answer is sought to this and similar 
questions, and to explain to you the reasons why the highest precision 
attainable is a sine qua non in the conduct of such research. Thus it is 
that the study of the large-scale phenomena of the universe resolves itself 
