ASTRONOMY: C. D. PERRINE 
291 
of velocity in all three classes near the smallest proper motions. An 
examination of the results of Kapteyn and Adams indicates somewhat 
similar rather sudden increases in velocity in the classes F, G, and K, 
but in proper motions somewhat larger than in the B and A stars. Such 
an effect is not noticeable in the M stars. The application of a mag- 
nitude velocity correction would tend to reduce the peculiarity slightly. 
As these velocities (Vi) for the B stars had been derived by includ- 
ing the constant error K, it was decided to see what the effect might 
be of neglecting it and using Vi = — 19.5 km., the same solar motions 
as were used in obtaining results for the A stars. The fifth column (V) 
contains these results. The general decrease of velocity with increasing 
proper motion is still indicated rather uncertainly but the sudden in- 
crease of velocity alluded to has been reversed in the B-B5 stars. This 
reversal seems to result chiefly from the application of K to the small 
negative velocities in the two groups. 
It seems rather curious that this peculiar maximum in the velocities 
shifts about with the term K — where a considerable value has been 
found for it, the maximum and peculiar sudden increase of velocity are 
shown only if it is used. On the contrary, where no K term resulted, 
the maximum and sudden increase show with its omission. If con- 
firmed, this indicates a relation of some sort. 
The parallactic motion of the Sun is very noticeable in the proper 
motions of both the B and A classes and it seems not improbable that 
some systematic effect may have crept into the velocities when classi- 
fied according to proper motions which have not been freed from the solar 
motion — that the peculiarities observed may be due to a preference of 
the faint stars and those of larger velocity for certain regions of sky. 
The dependence of velocity upon magnitude is clearly shown in both 
classes B-B5 and A in the above classifications particularly in the 
residual radial velocities of table II. The effect of the few apparently 
abnormal velocities of bright stars in class A is here very marked also. 
An examination of table II shows that in B-B5 stars with the possible 
exception of the smallest radial velocities (very brightest stars) there is 
little or no connection of the proper motion with the radial velocities. 
There is, however, a peculiarity in the case of the class A stars — the 
largest proper motions being associated with velocities of medium 
value. The largest velocities of all show the smallest proper motions — 
even after making liberal allowance for a magnitude-velocity effect. 
Until systematic effects such as magnitude-velocity equation, constant 
error, etc., are better understood, it seems useless to attempt a dis- 
cussion of such small peculiarities. It seems probable, however, that 
