THIRD SERIES.— DEPOLARIZATION OF HEAT. 185 



§ 2. On the Depolarization of Heat. 



29. In the first series of these researches, ^ 4, I entered pretty fuEy into the 

 subject of depolarization. The establishment of the fact was of the highest im- 

 portance, since there is httle probability of proving in any more direct manner 

 the doubly refi-active energy of crystals with respect to heat. But, besides the 

 demonstration of the fact, I pointed out in that paper the important numerical 

 determinations to which it might lead ; determinations of the first consequence to 

 the theory of heat, and the discrimination of heat from light. The measure of 

 depolarization in the case of light, or the quantity of light which has become po- 

 larized in a new plane by passing through a doubly refracting plate, such as mi- 

 ca, depends, 1. upon the length of a wave of light ; and, 2. upon the retardation 

 which one of the doubly refracted pencils of light suffers, upon the other, in pass- 

 ing through the mica, which retardation differs with the material of the plate, 

 varies directly as its thickness, and mai/ also vary with the quality of the inci- 

 dent ray. 



30. Hence, as a little reflection clearl}' shews, if the quantity of light (or, by 

 analogy, of heat) depolarized by a plate of given thickness be numerically esti- 

 mated, we may, if the length of the wave be given, determine the retardation, or 

 energy of double refraction ; or, if the latter be assumed or known, we may find 

 the length of a wave. Considering the latter element as the more important, and 

 not being then in possession of any more direct mode of determining it numeri- 

 cally, I proposed to assume the retardation due to double refraction as the same 

 for heat as in the case of light, (considering heat as but less refi'angible light), 

 and to determine the length of a wave in the way which I fully explained in the 

 First Series, art. 68-75. 



31. Two circumstances require notice by way of precaution. The first is, 

 that, for the very reason that we have periodical colours in the case of light, there 

 are (liferent thicknesses of mica and different measures of retardation, which, for 

 the same length of a wave, will give the same measure of depolarization ; these 

 dubious cases (which the formula of depolarization completely expresses) must 

 be distinguished. The second remark is, that all our sources of heat furnishing 

 heterogeneous rays, each has its own period of maximum and minimum inten- 

 sity, just as in the case of solar light, and since our means of numerical estima- 

 tion embraces the sum of all the effects of heterogeneous rays, we cannot ex- 

 pect results which shall rigorously satisfy a formula, in which homogeneity (or 

 constancy of x, the length of a wave), is assumed, but consider the approxi- 

 mate result as representing the mean or predominating character of the heat 

 employed. 



VOL. XIV. PART I. A a 



