ASTRONOMY: C. D. PERRINE 
371 
4. A similar calculation for the case in which the axis of the magneton 
is tangential to the path indicates that the rate of radiation is greater 
than that from the electric charge alone. 
5. It should be noticed that if = z;" = g^' = 0 the expression for I 
vanishes. Hence when an electromagnetic pole and an electromagnetic 
doublet move together with constant velocity along a rectilinear path, 
it is possible for the moment of the doublet to change at a constant 
rate without there being any radiation of energy. 
^Electromagnetic Radiation, Ca.mhi. Univ. Press, 1912: London, Phil. Mag. (Ser. 6), 36, 
1918, 243. 
ON THE DISTRIBUTION OF THE APHELIA OF THE 
SECONDARY BODIES OF THE SOLAR SYSTEM 
By C. D. Perrine 
Observatorio Nacional Argentino, C6rdoba 
Communicated by E. B. Frost, July 12, 1919 
The finding of a dependence of orbital eccentricity upon the relative 
masses of the components led to an examination of the directions of the 
apheHa of the secondary bodies of the solar system, a resume of which 
is the principal object of the present paper. 
There is Httle to guide us as yet in the interpretation of such a de- 
pendence. The first explanation which suggests itself is that it is a 
residual effect of capture. In the comets we have a very exaggerated 
effect of such a dependence. Their eccentricities are essentially unity 
and whether they come from interstellar space or from the outer 
planetary regions they are, for the few journeys which they perform 
about the sun, captured bodies. This is with reference to the comets 
with very long period and those with sensibly paraboHc orbits. The 
relations of the orbits of those with comparatively short periods to the 
outer (and larger) planets establishes the fact that such comets have 
been captured for the sun by these different planets. We have, there- 
fore, in the solar system at least one class of bodies which has been 
captured. A study of the orbital characteristics of these in connection 
with the other secondary bodies (having the sun also for primary) 
should disclose any similarities which may be significant. 
One fact which stands out prominently in connection with this de- 
pendence of orbital eccentricity upon relative mass appears to be 
significant. There are two possibilities aside from that of no preference 
