Prof. Forbes on the Polarization of Heat, 



351 



stead of making the plates parallel, as in fig. 1., I make them 

 incline opposite ways, as in fig. 3., and still get the same re- 

 sults, whilst in the intermediate position of fig. 2. forty per 

 cent, of the heat which reached the pile in the other cases is 

 stopped, and that instantly ^ without any appearance of that 

 gradual process which conduction implies? I imagine that 

 those most urgent for mathematical precision will perceive that 

 the effects of conduction above alluded to must be quantities 

 far below the powers of our most delicate instruments to mea- 

 sure, and at all events of an order quite inferior to the great 

 and striking effects I have described, and which can only be 

 due to a peculiar influence of the plates upon the freely trans- 

 mitted ray. 



Fig. l. 



Fig. 2. 



Fig. 3. 



I speak with the more confidence on this point because 

 the objection in question is not a new one. It was urged to 

 me several months ago in Paris by a French philosopher of 

 great eminence, who, however, had not had recourse to the 

 only fair test, namely, an experimental one. Unprovided with 

 any of the apparatus I had employed, I hastily and rudely pre- 

 pared bundles of mica for the purpose of demonstrating the 

 independence of the result upon any nice adjustment of the 

 distances or relations of the plates, and I showed these expe- 

 riments to MM. Melloni and Libri, whose names are im- 



