582 Prof. J. Larmor on the Intensity of the 
emitting body being opaque, the source of the radiation is 
vibratory disturbance of electrons in its surface-layer ; these 
constitute a self-damped current-sheet which is pushed back 
by the magnetic field it produces, precisely as happens for the 
corresponding current-sheet at a conducting surface on which 
radiation is incident as above *. : 
We now proceed to our problem of the radiation from a 
moving body. Consider an enclosure, with ideal perfectly 
reflecting walls, at a uniform temperature throughout, and 
thus pervaded by the steady natural radiation corresponding 
to that temperature. The principle of Carnot requires that 
we cannot by cycles of slow movement of the bodies in the 
enclosure transform any of this energy at uniform temperature 
into mechanical effect through the agency of the pressure of 
radiation. There must therefore be a unique state of density 
of the total enclosed radiation, independent of the nature of 
the surfaces of the bodies in movement; for otherwise direct 
movement with one kind of surface combined with the reverse 
movement with another kind would constitute a working 
cycle. The steady aggregate density of radiant energy in the 
enclosure is therefore not affected by the motion of the 
bodies; indeed, if this were not so, by opening and closing a 
window in the enclosure while it is moving at different speeds, 
cycles could be established which would violate Carnot’s 
principle. Now compare a moving perfectly reflecting sur- 
face, which reflects back all the incident radiant energy, with 
the same moving surface rendered perfectly absorbing : this 
is allowable, the analogous change from conducting to non- 
conducting being contemplated in elementary thermal 
reasoning about Carnot’s principle. It follows from the 
theory of exchanges, that in the state of equilibrium the 
radiation that is returned must be the same as regards consti- 
tution and intensity in both cases. Now the solution of the 
electrodynamic problem of reflexion from a moving perfect 
reflector is known f : therefore the law of the radiation from 
a perfect radiator in motion is determined in complete detail. 
When the reflector is advancing in a stationary enclosure, 
the energy-density of the reflected radiation is greater than 
that of the incident, and the excess is a fraction of the latter 
equal to four times the ratio of the velocity of the reflector 
* [April 22.—There are, however, intricacies here, owing to the 
internal radiation in the conductor, which are evaded for the present 
purpose by considering the aggreeate force on a moving slab due to 
nee radiation emitted from both faces, as given by the formula at the 
eoinning’. 
+ Cf. Tmeley British Association Report, 1900; Encyclopedia 
Britannica, article ‘ Radiation,” xxxii. 1903. 
