On Kirkwood’s Analogy. : 27 
"ich ihed tends ao strikingly to confirm it—so much, indeed, 
that if his latter be true, I do not know how any one can resist 
the argument which it furnishes in favor of the former, in so far 
as it applies to our solar system. It is then no longer a hypothe- 
sis, but becomes a probable theo 
I will give‘a very short actly of the quantities I have used in 
ae “repeating Mr. Walker’s computation 
Dr. Gould then gave upon the blackboard the masses of the 
planets, and the periods of rotation which he had used, differing 
from those used by Mr. Walker. ] 
"These are the masses which I have used; and these are the 
_ times of rotation as given in the books. I do not know how ac- 
curate ~ —_ may be considered ; perhaps to minutes, perhaps 
even less I believe they have all been determined by the ob- 
servation of spots. If so, what proof have we that the spots do 
not move, no matter what the number of rotations used in deter- 
mining the period? I have marked the period of Uranus as 
doubtful, because I do not know upon what authority it rests, 
having only found it in a table of a popular work by Sir J. Her- 
schel, with a mark of doubt prefixed. . does not agree at all 
with this theory. 
n considering a question of this kind, we must remember the 
nature of our investigations. The subject i is to a certain extent, 
necessarily general, and the appearance of precise harmony could 
not be expected even were our data exact, which they are not. 
The nature of the problem requires a general, not a special agree- 
ment between observation and theory. When we are consider- 
ing the evolution of order from chaos, we cannot pretend to a 
knowledge of all the physical forces which exerted an influence. 
€ go back to a supposed time when the planetary spaces were 
filled with nebular matter; we assume the existence of certain 
nuclei or centres of attraction ; and, from our knowledge of the 
Solar system, as it now is, infer the relative foree which these 
Sea centres of attraction must have exerted, and assign to 
its proportionate realm. If now we find that the spheres of 
Seine belonging to the several nuclei are harmoniously con- 
hected, by a simple formula, with the periods of rotation as ob- 
Served to-day, —an element before omitted in our investigations— 
We discover a remarkable corroboration of the probability of our 
hypothesis. This is what Kirkwood’s formula professes to be— 
asimple relation between the time of rotation and the diameter 
the sphere of attraction 
ranted in demanding that exactness of numerical agreement 
quisite- for the verification of theories of a more special. natu 
» 88 circumstances appear to indicate, more careful investigation 
should lead to the general adoption of the theory of Kirkw wood, 
The subject being then a strictly general one, we are not 2 
