Prof. Magnus on the Propagation of Heat in Gases. 3 



Despretz has since shown* that the conduction of heat in 

 water follows the laws given by Fourier for its conduction in 

 metals. A conductibility of heat in gases has never been ima- 

 gined to exist. Although it is in any case very small, it ap- 

 peared interesting to investigate what influence it might exert, 

 and whether differences existed in the conductibility of different 

 gases ; for the deportment of gases is of especial importance, not 

 only for the laws propounded by Dulong and Petit, but also for 

 any theory of the nature of heat. 



The more immediate inducement to this investigation was a 

 repetition of Grove's f interesting experiment, that a platinum 

 wire is less strongly heated by a galvanic current when surrounded 

 by hydrogen than when it is in atmospheric air or any other gas. 



On the first publication of these experiments, M.Poggendorff % 

 expressed the opinion that they depended on the laws which 

 Dulong and Petit had established for the cooling of a body heated 

 in the ordinary manner. Clausius § has since shown the con- 

 cordance between Grove's results and the numbers obtained by 

 Dulong and Petit. 



In repeating Grove's experiments, I found that hydrogen 

 exerts its preventive action even when only a very thin layer 

 surrounds the platinum wire. Two very thin, equally long pla- 

 tinum wires were enclosed in tubes of 1 millim. diameter, one 

 of the tubes being filled with atmospheric air and the other with 

 hydrogen. On passing the current through both wires in suc- 

 cession, the one surrounded by atmospheric air became strongly 

 incandescent, while the other did not even reach a red heat. It 

 is scarcely necessary to say that the result was always the same, 

 whichever of the two tubes was filled with hydrogen. Even when 

 the tube filled with hydrogen was quite horizontal, the wire did 

 not become incandescent. As the existence of currents in such 

 a narrow horizontal tube can scarcely be assumed, it appeared 

 improbable that the mobility of the particles of hydrogen was 

 the cause of the strong cooling of the wire. 



It is also impossible to conceive why currents produced by 

 differences of temperature should be stronger in hydrogen than 

 in other gases. This gas, it is well known, instead of being more, 

 is even somewhat less expansible than atmospheric air. Hence 

 the same differences in temperature produce in hydrogen less 

 change in the specific gravity than in atmospheric air. But it is 

 by these changes alone that currents are produced in gases. 

 Even if the friction of the particles exercises any influence, and 



* Ann. de Chim. S. 2. vol.lxxi. p. 206. 



f Phil. Mag. vol. xxvii. p. 445; vol. xxxv. p. 114. Pogg. Ann. 

 vol. lxxviii. p. 366. 



X Pogg. Ann. vol. lxxi. p. 197. § Ibid. vol. lxxxvii. p. 501. 



B2 



