376 
ASTRONOMY: C. D. PERRINE 
of the positive and neutral bundles showed the hydrogen series lines in 
both. 
The second point is that there is a sharp separation between the two 
bundles at the proper pressure. This shows that Hght is not emitted 
only during the return of a corpuscle to a positive centre, as very often 
assumed, for the neutrals cause light in a region where all free corpuscles 
have been swept out by the field. Also the sharp separation of the 
bundles shows that at these pressures changes take place slowly, so 
that if a positive were formed in the neutral bundle it would be deflected 
out before becoming neutrahzed again. We are thus led to conclude 
that light excitation may occur directly because of the collision of a neutral 
particle with a neutral molecule of the gas. We may regard the light 
emission as taking place during the rearrangement of the electrons in 
the atom after one has been detached by the collision (Stark's theory), 
or we may retain the picture underlying Bohr's theory, if we regard the 
displaced electron as not leaving entirely the centre to which it is 
attached. 
IE. Gehrcke and R. SeeHger, Verh. D. Physik. Ges., 15, 897 (1913); H. Rau, Ber. 
Phys.-med. Ges., Wiirzburg, Feb. 1914; J. Frank and G. Herz, Verh. D. Physik. Ges., 16, 
512 (1914); J. C. MacLennan and J. P. Henderson, London, Proc. R. Sac, A., 91, 485 (1915). 
2J. Stark, Leipzig. Ann. Physik., 13, 390 (1904). 
3A. Wehnelt, Ibid., 14, 464 (1904). 
AN APPARENT DEPENDENCE OF THE APEX AND VELOCITY 
OF SOLAR MOTION, AS DETERMINED FROM RADIAL 
VELOCITIES, UPON PROPER MOTION 
By C. D. Perrine 
OBSERVATORIO NACIONAL. ARGENTINO, CORDOBA 
Received by the Academy, June 3, 1916 
The position of the solar apex which had been derived from the 
stars of class B both from proper motions and radial velocities differed 
so much from the apex derived from radial velocities of A stars, that an 
investigation was undertaken of the radial velocities of all of the spectra 
classes. The discordances were not only between the spectral classes 
as a whole but appeared to be between the results from northern and 
southern stars as well. 
In the course of the work it was observed that frequently there ap- 
peared to be marked discordances between the radial velocities of stars 
whose proper motions also differed in size and sign. This, and the 
desire to test the effect of differences of distance as evidenced by 
