CHAP. VI., 9.] 



HEAT. M. REGNAULT. 



150 



loni had announced did not take place) ; l next, by 

 transmission through a bundle of very thin mica 

 plates, inclined to the transmitted ray ; and after- 

 wards by reflection from the multiplied surfaces of a 

 pile of thin mica plates placed at the polarizing angle. 2 

 I next succeeded in showing that polarized heat is 

 subject to the same modifications which doubly re- 

 fracting crystallized bodies impress upon light, by 

 suffering a beam of heat (even when quite obscure), 

 after being polarized by transmission, to pass through 

 epolari- a depolarizing plate of mica, the heat traversing a 

 ID f second mica bundle before it was received on the 

 pile. As the plate of mica used for depolarization 

 was made to rotate in its own plane, the amount of 

 heat shown by the galvanometer was found to fluc- 



tuate just as the amount of light received by the eye 

 under similar circumstances would have done. This 

 experiment which, with the others just mentioned, 

 was soon repeated and confirmed by other observers, 

 still remains the only one proving the double refrac- 

 tion of heat unaccompanied by light ; and though 

 somewhat indirect, it will hardly be regarded by 

 competent judges as otherwise than conclusive. Ice- 

 land spar and other doubly-refracting substances, 

 absorb invisible heat too rapidly to be used for effect- 

 ing directly the separation of the rays, which requires 

 a very considerable thickness of the crystal. I also 

 succeeded in repeating Fresnel's experiment of pro- Circular, 

 ducing circular polarization by two internal reflec- polariza- 

 tions. The substance used was of course rock-salt. 8 tlon ' 



9. M. REGNAULT. Numerical Laws of Expansion by Heat ; Rudberg. Vaporization; 



Dulong. Latent Heat ; Ilygrometry . 



THE limits of this Essay will not permit me to do 

 more than allude in very general terms to the merito- 

 rious services of M. HENRI- VICTOR REGNAULT in the 

 science of heat. In the seventh section of this chapter 

 I have mentioned his name in connection with that 

 of Dulong, whose researches he has prosecuted, and 

 whose position in the College de France he now fills. 



The attention of M. Regnault has been devoted 

 chiefly to heat in its combinations with matter to 

 dilatation and vaporization. I have already said, in 

 speaking of Dulong, that, in point of numerical pre- 

 cision in the results of experimental physics, the 

 French are unrivalled. The talent which they have 

 shown in the construction of apparatus, skill in its use, 

 and patience in deducing results with due attention to 

 every numerical correction, have not been equalled 

 either in England or Germany, much less elsewhere. 



We must, however, note that doubts were first 



thrown upon the accuracy of Gay-Lussac's coefficient Coefficient 

 of the expansion of the gases (0-375 of the vo- [ * x P f an " 

 lume at 32 for the expansion between 32 and 212 ases and 

 Fahrenheit) by Rudberg a Swedish philosopher, who mercury ; 

 determined a new coefficient (0-3645). M. Reg- Rudberg 

 iiault finds for air of the ordinary density a co-~~ 

 efficient nearly the same as that of Rudberg, but 

 differing slightly for the same fluid under differing 

 pressures, and also for the various gases. 4 The ex- 

 pansibility of all of these fluids appears to tend to 

 the same limiting value when they are sufficiently 

 attenuated. As a preliminary to these experiments, 

 the expansion of mercury was ascertained by its hy- 

 drostatic equilibrium at different temperatures, as 

 had already been done by Dulong, and with almost 

 coincident results. The dilatation of mercury was 

 used to ascertain that of the glass vessels employed. 



The irregularity of the dilatation of glass is one of (722. 



1 Annales de Chimie, torn. Iv. (1833). 



2 I was led to polarize heat by transmission through mica films from having observed the extraordinary permeability of those 

 films to radiant heat, and from the facility of adapting them to tubes applied to the pile. The idea of using bundles of mica for 

 reflecting heat did not occur to me until some time after. But I cannot here omit mentioning a circumstance of which I only 

 became aware some years after the publication of my researches. In arranging my correspondence, -I found some letters from 

 Sir David Brewster, with whom I had communicated as to the best means of polarizing heat, during my earliest and unsuccessful 

 attempts with common thermometers. In one of these letters he recommends, among other methods, the reflection of radiant 

 heat from mica bundles. This suggestion was not put in practice ; for, owing to change of residence and other circumstances, 

 my attention was diverted to other subjects, and only recalled, after a lapse of some years (as stated in the text), to the polari- 

 zation of heat, by the invention of the Thermomultiplier. Nor was Sir D. Brewste.r's suggestion recollected by me until I ac- 

 cidentally met with it (after another long interval), in the manner which I have just stated. I am glad to have an opportunity 

 of acknowledging the friendly assistance and encouragement in all matters of science which at an early age I received from him 

 when I was an obscure, though ardent student, and when he was my only scientific adviser. 



3 I have not thought it proper to go into farther details concerning my own experiments on radiant heat. Those who desire 

 more information will find it in Professor Powell's Second Report on Radiant Heat, in the Brit. Assoc. Reports for 1840. But I 

 mayhere state, that M. Melloni's first experiments on polarization were made with mica piles, furnished to him by myself in 1835. 



4 M. Regnault's experiments were published in 1841. Professor Magnus of Berlin was at the same time engaged on similar 

 experiments, and with nearly coincident results. The following table contains the summary of all these experiments ; 



Dalton's coefficient 0-391 



Gay-Lussac's coefficient .-. 0-375 



Rudberg's coefficient 03645 



M. Regnault'e coefficient (from the expansion observed under a constant pressure) 0'367 



M. Regnault's coefficient (from the elasticity observed under a constant volume) 0'3665 



M. Magnus' coefficient (from the elasticity observed under a constant volume) 0'3665 



Dalton's experiments were made between 55 and 212, and after allowing for the expansion of glass, he obtains for the relative 

 volumes of air at those temperatures 1000 and 1325, giving T 3 of the volume at 55 for 1 Fahr., or jfo of the volume at 32% 

 which agrees with the co-efficient given above. See Manchester Memoirs, vol. v., p. 598-9. 



