interior of one continuous medium. This transmission we have examined, and, as we have just seen, it presents nothing contrary to its analogy with the transmission of light through coloured media. There is however a particular case in which two homogeneous screens act in so singular a manner in respect to light that it must be interesting to know whether something analogous does not take place in respect to caloric.
The optical phænomena presented by most of the slices of tourmaline cut parallel to the axis of crystallization are universally known. If these slices are placed one over the other and their axes laid in the same direction, they transmit light in considerable quantities. But if they be laid at right angles to one another, the light is totally intercepted. Do these phænomena, arising, as is well known, from the polarization of the light in the interior of the slices, take place in respect to calorific rays also; or, in other words, is radiant heat capable of being polarized in its passage through tourmaline?
In order to ascertain this I have taken two square plates of the same dimensions. I have made an aperture in the centre of each. This aperture was likewise a square having its sides parallel to those of the plate and each equal to the least breadth of the two polarizing slices. I then took some soft wax and attached a tourmaline to each aperture, holding the axis of the former parallel to one of the sides of the latter. These two plates being laid one over the other, it evidently depended on one of the sides of the one plate being placed parallel or perpendicular to a side of the other whether the light was to be transmitted or intercepted. Yet this pair of plates being placed vertically on the stand of my thermoelectric apparatus and exposed to the radiation of a lamp or incandescent platina, uniformly produced the same calorific transmission, whatever might be the relative direction of the sides of each plate.
That this fact might be put beyond the reach of doubt the galvanometric index was carried to the 18th or 20th degree, and the calorific communication now established was suffered to remain while we placed one of the plates on each of its sides in succession. The flame or the incandescent platina was then observed to appear and disappear alternately while the magnetic needle continued invariably at the same point of deviation.
This experiment was repeated many times with several tourmalines, and the angle formed by tlie intersection of their axes varied. The result was in all cases the same. The quantity of calorific rays transmitted through the two polarizing slices is then independent of the respective directions given to their axes of crystallization; that is to say, the heat radiating from terrestrial sources is not polarized in its passage through tourmalines[1].
- ↑ This result seems opposed to the experiments of M. Bérard on the polarization of reflected heat; but, ignorant as we are of the nature of the relations
