588 
Proceedings of the Royal Society 
2 1 
P = 
3 y-1 ' 
if y be the ratio of the specific beats, and /3 the ratio of the whole 
energy to the translational part of it. For air, the value of y found 
by experiment, is 1*408, which makes /3 = 1*634. For steam, 
Maxwell says, on the authority of Eankine, /3 “may be as much 
as 2*19, but this is very uncertain.” If the molecules of gases are 
admitted to be elastic corpuscules, the validity of Clausius’ prin- 
ciple is undeniable ; and it is obvious that the value of the ratio /3 
must depend upon the shape of each molecule, and on the distribu- 
tion of elastic rigidity through it, if its substance is not homo- 
geneous. Farther, it is clear that the value of /3 for a set of equal 
and similar corpuscules will not be the same after collision with 
molecules different from them in form or in elastic rigidity, as 
after collision with molecules only of their own kind. All that is 
necessary to complete Le Sage’s theory of gravity in accordance 
with modern science, is to assume that the ratio of the whole 
energy of the corpuscules to the translational part of their energy 
is greater, on the average, after collisions with mundane matter 
than after inter-collisions of only ultramundane corpuscules. This 
supposition is neither more nor less questionable than that of 
Clausius for gases which is now admitted as one of the generally 
recognised truths of science. The corpuscular theory of gravity is 
no more difficult in allowance of its fundamental assumptions than 
the kinetic theory of gases as at present received ; and it is more 
complete, inasmuch as, from fundamental assumptions of an ex- 
tremely simple character, it explains all the known phenomena of 
its subject, which cannot be said of the kinetic theory of gases so 
far as it has hitherto advanced. 
Postscript , April 1872. 
In the preceding statement I inadvertently omitted to remark 
that if the constituent atoms are aeolotropic in respect to perme- 
ability, crystals would generally have different permeabilities in 
different directions, and would therefore have different weights 
according to the direction of their axes relatively to the direction 
of gravity. No such difference has been discovered, and it is 
