17 
in both cases the magnetic action varies very nearly inversely 
as the cube of the distance between the centres of the magnets. 
I have accounted for both these laws by means of a hydrody- 
namical theory of magnetism founded on the same a priori 
hypotheses. (See the Philosophical Magazine for July, 1869, 
p. 42.) 
(3.) The Astronomer Royal has given in vol. clxii. of the 
Philosophical Transactions the results of an experiment for 
determining the intensity and direction of the action of a 
galvanic coil on a small magnet placed in various positions 
round the coil, and restricted to oscillate, with its middle 
point fixed, in a plane passing through the axis of the coil. 
On the principles of hydrodynamical theories of galvanism and 
magnetism resting on the same basis, I have been able to 
account for the laws of this action, and to make a successful 
numerical comparison of theoretical values, giving the direc- 
tion and intensity of the galvanic force, with values obtained 
directly from the experiment. (This problem is discussed at 
length in the Numbers of the Philosophical Magazine for 
September, November, and December, 1874.) 
(4.) Much interest has recently been excited by the pheno- 
mena of Mr. Crookes’ Radiometer, which have been supposed 
to give indication of the existence of a new physical force. It 
might, therefore, reasonably be asked whether the proposed 
general hydrodynamical theory of the physical forces was 
competent to explain these phenomena. On making applica- 
tion of it, I was led to conclude, without adding to, or deviating 
from, in any particular principles previously admitted, that 
the action of light on the surfaces of the vanes induces, in 
combination with the law of heat-exchanges always in operation 
between neighbouring bodies, an abnormal disposition of the 
superficial atoms not unlike that produced in electrical experi- 
ments by friction, and that in consequence of the inequality 
of this action on the opposite black and bright surfaces of the 
vanes, steady ethereal currents are generated (just as in the 
hydrodynamical theory of frictional electricity), the pressure 
of which on the individual atoms causes the movement of the 
vanes. An experiment by Mr. Crookes which showed that a 
pith ball, suspended near the revolving vanes in a cup inclos- 
ing the Radiometer and very nearly exhausted of air, w r as 
made to oscillate if the rotation was not very rapid, seems to 
justify the supposition of an electric action. (I have treated 
of the theory of the Radiometer in the numbers of the Philo- 
sophical Magazine for May and November, 1876, and April, 
1877.) 
It may here be mentioned, as peculiarly confirmatory of the 
VOL. XII. 
c 
