6o 
Philosophical Magazine for September, 1876 (p. 173), “ The 
tetlier, being assumed to be susceptible of variation of density, 
must be conceived to be atomically constituted, because we 
have no experience of variation of density and pressure which 
is not the result of atomic constitution. But for the purposes 
of physical research, it suffices to regard the asther as a con- 
tinuous substance, and apply calculation to it as such, just as 
the air is treated mathematically in hydrodynamics, although 
it is known to be composed of discrete atoms.'” Thus, Pro- 
fessor Challis, while treating the aether as continuous for con- 
venience of calculation, declares that it, in fact, consists of 
discrete atoms, like the air. There is, therefore, nothing in 
the view adopted by him which at all militates against 
Mr. Herbert Spencer's remark, that by supposing the inter- 
vention of an aether we are brought no nearer to the con- 
ception of action at a distance than we were without that 
supposition, because the atoms of the aether itself are at dis- 
tances from each other which are very great when compared 
with their magnitude. Since, then, we cannot conceive any 
mode by which gravitation produces its effects, surely it would 
be taking a great deal upon us to accept it as an axiom that 
its amount can never vary. Unless we knew its mode of 
action, we could not possibly assert this even as a fact, much 
less as an axiom. If we did not know it to be a fact that a 
heated mass gradually loses its heat, it seems to me that there 
would be quite as much reason in pronouncing upon the 
invariability of its heat as upon that of its attracting power. 
If one power of matter can be subject to variation, why not 
another; especially when both are believed to act through the 
same medium, viz., the aether ? 
29. It might perhaps be said in reply, that even the sup- 
position of a gradual diminution of the attracting power of 
the sun or the earth would not be inconsistent with the per- 
sistence of force, because that power might be dissipated, as 
heat is believed to be dissipated, but never actually lost. But 
the question which is now being dealt with is the persistence, 
or rather the invariability, of the attractions of the sun and of 
the earth upon bodies to which their attractive force can reach ; 
for it is by examples drawn from these that Mr. Spencer 
illustrates his principle. To admit that these forces may be 
dissipated would be to admit that the conservation of energy 
is not an established principle ; for then the kinetic and 
potential energies of a planet or of a pendulum would not be 
complementary, the unit of force which is assumed in dynamical 
calculations to be constant being no longer so. Thus, Mr. 
Spencer's principle of the persistence of force would not only 
VOL. xv. f 
