as the distance from the surface of entrance increases; but the diminution becomes less and less perceptible, so that it must become invariable when the rays have penetrated to a certain depth. This is precisely what happens to a pencil of ordinaiy light when it enters a coloured medium; for, those rays that are of a colour different from that of the medium being extinguished in the first layers, the losses of intensity sustained by the luminous pencil are at first very great, but they afterwards become gradually less and are at last very small, but constant when the only rays remaining are those of the same colour as the medium.
In fine the successive transmissions through heterogeneous screens furnish a third proof of the analogy which the action of diathermanous bodies on radiant heat bears to that of coloured media on light. The luminous rays issuing from a coloured plate either pass in abundance through a second coloured plate or undergo in it a powerful absorption according to the greater or less analogy of the colour of the second to that of the first plate. Now we observe facts perfectly similar to this in the successive transmission of radiant heat through screens of different kinds. And in this case too the rock salt acts in respect to the other bodies as it does in the case of rays emanating from sources of different temperatures. A given plate, if it be of rock salt, being successively exposed to calorific radiations of the same force emerging from different screens, transmits a constant quantity of heat; if the plate be of any other diathermanous substance the quantity transmitted will be variable.
There is therefore but one colourless and diaphanous body that really acts in the same manner on luminous and calorific rays. All other diaphanous bodies besides this indiscriminately suffer all kinds of light to pass through them, but of the rays of heat they allow some to pass while they absorb others: thus we discover in this one substance a real calorific coloration, to which, as it is invisible, and therefore totally distinct from coloration properly so called, we have given the name of diathermancy.
The colours introduced into a diaphanous medium always diminish its diathermancy in a greater or a less degree, without communicating to it any tendency to arrest certain calorific rays rather than others: they affect the transmission of radiant heat as dusky bodies affect the transmission of light. There is, it is true, an exception to be made in respect to green and opake black, at least in certain kinds of coloured glass. But these two colouring matters appear, in this case, to do no more than modify the quality to which we give the name of diathermancy, and which, as we have already seen, is totally independent of coloration.
The quantity of radiant heat which passes through polarizing plates of tourmaline is not affected by any change made in the angle at which their axes of crystallization are made to cross one another. Rays of heat are therefore not polarized in this mode of transmission and are in
