422 Prof. H. BufF on the Thermal Conductivity/ 



quire to propagate itself to a distance of 100 millims. in the 

 short interval of five minutes ; I am rather inclined to ascribe 

 this to the very great mobility of the hydrogen molecules. 



Water-vapour in vacuum, however small the quantity, ab- 

 sorbs a considerable amount of heat, in any case more than 

 air of the same tension ; but in air of atmospheric pressure 

 this preponderance is not very striking ; in other words, rays 

 which are not absorbed by dry air pass to a great extent also 

 through moist air. 



Other experimenters have taken more or less active part in 

 the controversy on the magnitude of the absorptive power of 

 aqueous vapour for rays of low refrangibility. The researches 

 of H. Wild* are particularly deserving of attention. 



Wild studied carefully Tyndall's method, as well as that of 

 Magnus. The measurements, however, which he communi- 

 cated were obtained with an apparatus resemb-ing that of 

 Tyndallf without rock-salt plates. His somewhat altered ar- 

 rangement is well adapted for exhibiting to a large audience the 

 fact that moist air absorbs rays from a dark source better than 

 dry : but the real point at issue is not touched at all by this 

 fact ; for Magnus himself had repeatedly acknowledged the 

 higher absorptive power of moist air, denying onl}^ that this 

 was from 15 to 40 times as great as that of dry air. In this 

 respect Wild's experiments give as little information as those 

 previously made by Tyndall ; for in both cases the rays, before 

 entering the spaces saturated with moisture and reaching the 

 thermopile, had passed for a certain distance through the air 

 of the room — and consequently had lost the rays capable of 

 absorption by air, without being able to affect the thermomul- 

 tiplier. 



As already stated. Wild also submitted Magnus's process to 

 an experimental investigation ; but in consequence of his pre- 

 conceived notion that the absorptive power of air was very 

 small, he failed to draw the right conclusions from indications 

 in a contrary sense w^hich his own experiments offered J. 



One circumstance in Wild's memoir produced doubts in my 

 mind when first reading it. He assumes that the absorption 

 is proportional to the thickness of the absorbing layer, and 

 concludes from this that Magnus had obtained smaller values 

 than Tyndall, to a certain extent, because the distance of the 

 thermopile from the source of heat had been too short in his 

 experiments. 



It is quite justifiable to assume that, wdthin certain limits, 



* Phil. Mag. [IV.] vol. xxxii. p. 241. 



t Ibid. vol. xxiv. p. 270 ; also vol. xxvi. p. 21. 



t Ibid. vol. xxxii. p. 2o4. 



