Third Series* — Variable Polarizability of Heat. 1 OS 



of M. Melloni's results was easy and complete. It appears 

 from the account of his experiments, that he still employs 

 piles of mica of the form I at first used, consisting of distinct 

 laminae separated by a knife, then laid together and united at 

 the edges, up to the number of 30, 60, and even more*. On 

 the other hand, the piles 1 and K, which for two years and a 

 half I have employed, are of a degree of tenuity really sur- 

 prising. The mode of their construction I mentioned briefly 

 in my last paper, art. 20, and it is so very superior to any 

 other, that it is probably from inadvertence that it has not 

 been generally employed. The piles laminated by the action 

 of violent heat, afford a multiplicity of parallel surfaces in a 

 given thickness of mica, which no mechanical method can 

 approach. The actual thickness of mica which they contain, 

 I am unable accurately to estimate. The plates marked G 

 and H are much thicker, perhaps twice as thick as those 

 marked I and K, which I commonly use ; yet the former, as 

 I roughly estimate by the tint they give in polarized light, 

 are only about one thousandth of an inch in thickness. At 

 the utmost, the plates I and K can be but one fifteen hun- 

 dredth of an inch ; and yet it appears that their polarizing 

 power (depending solely on the number of surfaces they con- 

 tain) is equal to M. Melloni's pile of ten distinct plates placed 

 at the same angle (35° to the incident rays). The mean 

 thickness of the elementary plates can, therefore, be only one 

 fifteen thousandth of an inch; and they reflect abundantly 

 the colours of Newton's rings. 



Now, I have found by the depolarization experiments, that 

 it requires a much greater thickness of mica than that tra- 

 versed by the heat in passing through the plates I and K 

 (even allowing for the obliquity) to affect materially the index 

 of polarization of heat from different sources, such as from 

 brass at 700°, and incandescent platinum. It is, therefore, a 



* Annates de Chimie, Mai 1837. At p. 17, &c. M. Melloni has given a 

 minute account of that method of constructing the piles, which, " amongst 

 several different ways, he considers the preferable one." No one could 

 doubt from his language that he is describing a new and improved form 

 of the apparatus. I regret for a moment to descend to notice an apparent 

 want of justice and courtesy towards myself; but it is im[)ossible for me to 

 not to observe, that the procedure he so exactly details, is, to almost the 

 minutest particular, identical with that which I myself used in June 1835, 

 in constructing, in M. Melloni's presence, the first pair of piles used for 

 polarizing heat which existed in France, at a time when M. Melloni ex- 

 pressed his unqualified scepticism as to the polarization of heat generally; 

 which piles I left, at his desire, where I presume they now are,— in his own 

 possession. This mode of construction I soon after abandoned, for the 

 improved one alluded to in the text. 



