President's Address. 
55 
in certain regions of the sky the influence of one or other drift will pre- 
dominate, in certain other regions the drifts will give rise to transverse 
motion in similar directions, while in yet other regions they will be 
directly opposed, renders the separation of the stars into their component 
drifts a question of very considerable delicacy. 
The difficulties in a large measure disappear if we confine our atten- 
tion to regions of the sky directly opposite to one another. In such 
regions the effects of foreshortening due to the inclination of the line of 
vision to the direction of drift velocity will be exactly similar, and any 
want of symmetry between such regions must be attributed to other 
causes. 
That there is such a want of symmetry will be clear from these 
diagrams. 
(Diagrams were here shown.)* 
The curves shown on these diagrams exhibit the proportion of stars 
contained within certain limited areas of the sky, the position angles of 
whose proper motions lie within certain assigned limits. Notice first of 
all that the majority of these curves exhibit two clearly defined humps or 
maxima, showing two prevailing directions of the proper motion for each 
region. These directions are the directions which correspond with the 
motions of Kapteyn's two drifts. The variation in form of these curves 
can in part be explained by the different aspects of view of the drifts in 
different regions, but I wish more especially to direct your attention to a 
feature which cannot be accounted for in this manner, viz., the contrast 
between the curves, which correspond with opposite regions ; this contrast 
is more clearly brought out in the third series of diagrams, the ordinates 
in which represent the differences between those of the first and second, 
and to account for it we have to suppose that, though the two drifts per- 
vade the whole sky, they are not present in equal proportions in all 
directions but that first drift stars are relatively more numerous in some 
regions and second drift stars in others. 
I say relatively more numerous, as a similar analysis of the actual 
numbers instead of their proportions seems to indicate that first drift stars 
are actually distributed fairly symmetrically throughout the sky, and that 
where they appear more numerous, the phenomenon is to be accounted 
for merely by a deficiency of stars pertaining to the secocd drift. 
Minor anomahes which it would take too long for me to discuss, al 
seem in like manner to be associated with the second drift. 
These irregularities in distribution in the stars of the second drift, 
which seem to have no counterpart in those of the first, it should be 
borne in mind, are the outstanding features over and above the fortuitous 
* A more detailed exposition of the diagrams shown has been communicated to the 
Royal Astronomical Society. 
