THE 

 LONDON, EDINBURGH and DUBLIN 



PHILOSOPHICAL MAGAZINE 



AND 



JOURNAL OF SCIENCE. 



[FOURTH SERIES.] 



JULY 1861. 



I. On the Propagation of Heat in Gases. By G. Magnus*. 



[With a Plate.] 



Conduction of Heat by Gases. 



THE cooling of a body, when it takes place in vacuo, simply 

 depends on the exchange of heat by radiation between 

 the body and the surrounding envelope. If, however, the space 

 in which the cooling takes place is filled with a gas, an ascend- 

 ing current is formed which accelerates the process. The cooling 

 is likewise promoted by the capacity of the gas to transmit heat (or 

 its diathermancy) , as well as by its conductibility, assuming that 

 gases can conduct heat. Dulong and Petit, in enunciating their 

 laws of the loss of heat in their comprehensive memoir Sur la 

 Mesure des Temperatures et sur les Lois de la Communication de 

 la Chaleur, have disregarded the latter actions, manifestly because 

 they could neglect them as being infinitely small in comparison 

 with the influence of the ascending current. Accordingly, since 

 the appearance of their memoir, it has been universally assumed 

 that the differences of cooling in various gases depend on the 

 different mobility of their particles. This was the more justi- 

 fiable, because almost simultaneously with Dulong and Petit' s in- 

 vestigation Sir H. Davy's celebrated memoirf on Flame appeared, 

 in which he says, " It appears that the property of elastic fluids 

 to withdraw caloric from the surface of solid bodies increases as 

 their density decreases, and that there is something in the con- 



* Read before the Academy of Sciences of Berlin, July 30, 1860, and 

 February 7, 1861. Translated by Dr. E. Atkinson. [A short abstract of a 

 portion of the results has already appeared in this Journal, vol. xx. p. 510.] 



t Phil. Trans. 1817, part. 1, p. 61. 



Phil. Mag, S. 4. Vol. 22. No. 144. July 1861 . B 



