face of the enclosure, the heat which enters the surface from 
the moving body will not be the same as that which the 
surface gives out. 
Suppose for instance that the walls of the enclosure are 
made of glass, and that the temperature of the whole 
enclosure including that of the moving body is 0°C., then, 
were the whole at rest, the heat which strikes the glass 
surface will all be absorbed at a very short distance below 
the surface, and in like manner the heat radiated by the 
glass will all emanate from a short distance below the sur- 
face. But let us now suppose, to take an extreme case, that 
the moving body is approaching one of the glass surfaces so 
rapidly that the heat which it emits has been so much 
increased in refrangibility as to enter the boundary of the 
visible spectrum. 
Then while the heat radiated by the glass will still con- 
tinue to proceed from a very short distance beneath the 
surface, the heat absorbed by the glass from the moving body 
will be able to penetrate to a very considerable depth beneath 
the surface of the glass. 
The outer layer of glass will thus lose, while the inner 
layer will gain, heat. 
Now it is possible to conceive an enclosure with a fixed 
diaphragm, and containing a revolving body, so arranged 
that the heat which leaves it in the direction of a certain 
part of the enclosing surface, shall always be given out by 
that part of the revolving body, which is moving towards 
the surface ; while on the other hand, the heat given out 
by the revolving body to another surface, shall be given out 
when the revolving body is moving from that surface. 
There will thus be a want of temperature equilibrium 
among the various layers, those near the surface being some- 
what different in temperature from those beneath. But 
when we have a temperature difference of this kind have we 
not acquired the power of converting heat into work ? It 
