M. MELLONl ON THE POLARIZATION OF HEAT. 343 



source of heat, or emitted by different sources, are very unequally 

 affected by the cause through which the phnenomena of polarization in 

 the tourmalines are rendered sensible. Some of them appear to un- 

 dergo no action of this kind, others produce indices of polarization 

 more or less marked, and others are, like rays of light, completely po- 

 larized. Tourmalines in general, and the green tourmalines in particu- 

 lar, absorb the most polarizable rays, and transmit tliose species which 

 seem totally or partially to escape polarizing action. The consequence 

 is, that their apparent index of polarization is generally very feeble and 

 sometimes even inappreciable. But it rises even to xW» and perhaps 

 higher, in those systems of plates which are permeable to a greater pro- 

 portion of heat susceptible of a high degree of polarization, as we see 

 in the plates of yellow, brown, or violet tourmalines. The index of 

 apparent polarization in a given system varies considerably in passing 

 from one source to the other, because a change takes place in the qua- 

 lity and the grouping of the rays constituting the calorific stream issuing 

 from the focus of heat. In fine, this index varies, and in certain cases 

 attains its two extreme limits, and 100, when we introduce between 

 the same source and the same system of tourmalines diathermanous 

 plates of a different kind ; because the particular absorption of these 

 screens affects the relations of quantity existing between the several 

 groups of rays composing the calorific stream naturally transmitted by 

 the polarizing system." 



In all these statements we have taken care to apply the qualifying 

 term apparent to the signs of feeble polarization exhibited by the tour- 

 malines ; and, in fact, all the rays of heat, whether direct or transmitted 

 through a screen, might be completely jDolarized, as light is, in the in- 

 terior of these crystallized bodies, and yet the polarization not be ren- 

 dered perceptible by any diminution in the quantity of heat transmitted 

 by the plates when the parallel is exchanged for the perpendicular po- 

 sition of the axes. 



In order to understand this proposition it is necessary to recollect the 

 phaenomena which take place in the polarization of light by tourmalines. 



When a ray of natural light falls perpendicularly on a plate of tour- 

 maline cut parallel to the axis of the needles, the double refraction first 

 divides the ray into pencils possessing sensibly equal intensities and 

 polarized at right angles ; but, in proportion as these pencils penetrate 

 into the substance of the tourmaline, they suffer very different degrees 

 of absorption, that of the pencil which has undergone the ordinary 

 refraction being much the greater; so that, beyond a depth often veiy 

 inconsiderable, one of the pencils is entirely absorbed, while the other 

 pursues its path, emerges from the plate, and shows itself in its proper 

 direction of polarization. This inequality of absorption is proved by 

 the following experiment, for which we are indebted to M. Biot. We 



