Prof. Magnus on the Propagation of Heat in Gases. 5 



Davy* and others; but they., like Dulong and Petit, introduced 

 the heated thermometer into a space which was successively 

 filled with different gases, and observed the times which were 

 necessary for the same cooling in these gases. Currents were 

 thereby produced within the gas, which in my experiments 

 were very small if not entirely absent, for the tube containing 

 the gas was heated almost equally on all sides, above as well as 

 below. But as the times which the thermometer required to 

 become heated varied considerably, it appeared probable that 

 the -heating in the gases was not produced by currents alone, 

 but that there was also a propagation of heat from particle to 

 particle — in other words, a conduction. Accordingly I next made 

 a series of experiments in which the gases were only heated from 

 above, and observed the temperatures which a thermometer placed 

 in them ultimately assumed. As in this case also the tem- 

 perature was always higher in hydrogen than in other gases, 

 and was also very different in them, I was confirmed in the con- 

 clusion that gases can conduct heat. It might still be objected 

 that in the heating from above currents were formed, which 

 caused the differences in temperature. There was a ready means 

 of testing this objection. For if gases actually can conduct heat, 

 the temperature which a thermometer assumes in a space heated 

 from above, must be lower when the conducting substance is 

 absent, that is, if the space is exhausted. In order to investigate 

 whether this is the case I made use of the following apparatus. 



Experiments on Conduction. 



On a very thin glass vessel AB, fig. 1, Plate I., 56 millims. wide 

 and 160 millims. in height, a second vessel C was fixed by fusion, 

 of the same diameter, but only 100 millims. in height. A B is 

 provided with a lateral tubulure D, in which a thermometer fg 

 is hermetically fitted in such a manner that its bulb is in the axis 

 of A B, and 35 millims. under the bottom of /, while the hori- 

 zontal scale is outside AB. The lower end of A B is closed by 

 means of a cork, in which are two narrow glass tubes provided 

 with stopcocks, which serve to fill AB with different gases. 

 Boiling water was poured into C, and then, from a flask at some 

 distance in which water boiled, steam was passed into this water 

 by the glass tube pp, so as to keep it in a state of ebullition. 

 A plug of cotton wool prevented spirting. 



In order to compare the thermometer-indications, obtained in 

 using different gases, it was necessary to ensure that the space 

 surrounding the vessel AB was always at the same tempera- 

 ture. For this purpose the vessel A B, with its thermometer, 



* Philosophical Transactions for 181/, part 1. p. 60. Schweigger's 

 Journal, xx. p. 154. 



