vi. J PROFESSORIAL L1F1-. 



it between polarizing and analysing plates of 

 mica, as described in my last. When the plane of total 

 reflection coincided with that of primitive polarization, 

 or rather was perpendicular to it, the heat was as much 

 polarized as before the rhomb was interposed ; when it 

 was inclined 45 it was wholly un polarized, apparently, 

 or even the longer axis of the ellipse turned a little the 

 other way, corresponding to p for heat. This I made 



11 with a very imperfect rhomb, and with heat 

 wholly unaccompanied ly light. ... I congratulate 

 Scotland upon {Smith's distinction. . . .' 



1 EDINBURGH, March I0th, 1836. 



1 . . . Pray be the depository of these facts in case 



Biot and Melloni, with the odds of two men at leisure 



Insl one man with his hands full, take them from me. 



.... Since writing the above I have solved, partly, 



'iibt which has much puzzled me. The action of 



mrtal gives a maximum polarizing angle for heat greater 



i for light, whence I concluded that the index of 



iction must be greater for the former than, the latter, 



contrary to my general views. I have just found in 



Brewster's paper that a precisely similar fact occurs in 



tin- n"iion of metals on light ; the red ray is polarized at 



i- incideiiee tlian the blue. It is most satisfactory 



thus to find the truth of experiments confirmed by 



1 anomaly, as has several times occurred to 



in*' lately.' 



The following letter to Mr. Leslie Ellis, before he 

 <] on his ( amhridov eanvr, will interest many who 

 l'\rd and admired that remarkable man: 



INIJUROH, February Utk, 1836. 



... I assure you also that I had been long looking 



to tin- fulfilment of your promise to write to me, 



began to f el some anxiety as to your state of health. 



Let no stimulus of fame or advantage induce you to 



.e a sacrifice of the first of earthly blessings. I hope 



will not go to Cambridge, unless you are equal to 



