THIRD SE«IES.— DEPOLARIZATION OF HEAT.; 187 



of a pair of callipers constructed for such purposes by Troughton. These results 

 are the mean of ten measures each, which were rendered difficult by the elastic 

 and fissile nature of the substance. 



ThiL-kuess in parti 

 of an inch. 



No. 1 0026 



No. 2 • 0044 



No. 3 0074 



No. 4 0060 



No. 5 0167 



35. With these mica plates in succession, employed for depolarizing, I pro- 



ceeded to determine the ratio pj (art. 32) for the most part exactly in the way 



described and illustrated by an example in art. 71, First Series, which I found 

 preferable to any other. This laborious investigation I performed for heat from 

 three sources ; (1), an Argand-lamp with glass chimney ; (2), incandescent plati- 

 num ; and, (3), brass heated (not to visible redness) by an alcohol flame. The 

 thickness of the plates No. 3. and No. 4. being very nearly the same (and giving, 

 as they ought to do, almost exactly the same measure of depolarization), I pre- 

 fen-ed using the united thickness of Nos. 2. and 3. as an interpolation between 

 Nos. 3. and 5. The swings of the needle, or dynamical effects, (II. 8) were always 

 observed, and are alone given. The polarizing and analyzing plates were the 

 same, marked I and K, before fully described (II. 20), and a plate is said to be at 

 0° or at 90° as its plane of refraction is vertical or horizontal. With these expla- 

 nations, and a reference to art. 71, First Series, the following specimens of obser- 

 vations will, it is hoped, be intelligible. 



Argand-Lamp : 16 inches from centre of Pile, depolarizing Mica Plate No. 3. 



