THIRD SERIES.— VARIABLE POLARIZABILITY OF HEAT. I77 



of this course has been the gradual abandonment of such doubts, and the entire 

 adoption of my conclusions. 



6. I believe that only a single exception remains to this statement. I expressed 

 my belief in uryjii^st paper that heat was differently polarizable, according to the 

 source whence it was derived. M. Melloni* failed to verify this result, and the 

 opposite conclusion, namely, that all kinds of heat are equally polarized by a given 

 pile of mica, was prominently put forth by himself and M. Biot as an important 

 discovery, f Without any undue confidence in my first, confessedly imperfect, 

 researches, I proceeded in my second paper (art. 22, et seq.) to give what I con- 

 sidered ample proofs of the correctness of the statement, though the great dissi- 

 milarity of the numbers arrived at from those of my first paper, shewed that the 

 latter were worthy of very little confidence on the ground of numerical exactness, 

 which, indeed, 1 never claimed for them. The later experiments, however, were 

 made with a view to accurate results, and I stated certain forms of the experiment 

 which I had devised on purpose to meet the objections of M. Melloni, although I 

 avoided mentioning his name. 



7. It seems, however, that M. Melloni, returning to the subject with his ac- 

 customed diligence, after receiving my second paper, still confirmed his former 

 results, and he has attempted to shew, in a very long paper, published in the 

 Annates de Chimie for May 1887 (which only appeared in October), that his results 

 must be exact, and the probable source of my errors. I contented myself with 

 giving a very brief answer to this paper in the Philosophical Magazine for Decem- 

 ber 1837, admitting the improbability that so experienced an operator as M. Mel- 

 loni should be wrong in his numerical results, but stating convincing grounds for 

 believing that his explanation of my conclusion, founded on experimental errors, 

 was inapplicable. The inquiry which I have since been led to make, and the en- 

 tirely satisfactory explanation at which T have arrived of a difference so puzzling, 

 terminating in a confirmation of my original statement, I now proceed to lay be- 

 fore the Society. 



8. I have not the remotest intention of examining and criticising M. Mel- 

 loni's paper in the Annales de Chimie for May 1837, as respects trifling or per- 

 sonal matters, which I readily confide to the impartiality of those best qualified to 

 judge : but it is quite necessary to state the facts which I had observed, and M. 

 Melloni' s mode of accounting for them. 



9. With two polarizing mica bundles of great tenuity, prepared in the method 

 described, (II. ^ art. 20) marked I and K, I found that, with heat from an Ar- 

 gand-lamp, 72 to 74 per cent, of the incident rays were polarized, that is, — of 100 



* Comptes Rendus de TAcademie des Sciences, ii. 140. f Ibid. p. 194. 



X To avoid circumlocution, I shall denote by I. II., &c. the First, Second, &c. Series of Researches, 

 and by the succeeding Arabic numeral the Article referred to. 



VOL. xlv. part I. Z 



m 



