and Diatherinancy of Air and Hydrogen. 413 



The average of five experiments gives for air of atmospheric 



density the number — - — =20*7, whilst for a vacmim of only 



1*5 millim. tension the average value was — - — =43*4. Thus 



half of the rays of low refrangibility entering dry air of atmo- 

 spheric density are absorbed by those layers which are nearest 

 the sources of heat. 



With decreasing density the diathermancy increases. But 

 the experiments 9 to 15 of the Table show that the increase is 

 not inversely proportional to the tension of the air, but takes 

 place much more rapidly^ so that for a tension of 100 millims. 

 the diathermancy approaches very near that of a vacuum. 



The effect upon the junction of the thermopile was next de- 

 termined when the source of heat was at a greater distance 

 from it. This was done in order to determine whether the 

 absorption took place immediately upon the entrance of the 

 rays, or whether the lower layer of air likewise absorbed part 

 of the rays. 



The connecting wires of the thermopile were therefore 

 shortened so that the soldered junction was 100 millims. below 

 the brass plate. At this depth there was no rise of tempera- 

 ture perceptible on the outer surface of the glass cylinder. 



The results of the experiments carried out in this manner 

 did not answer my expectations. The deflections of the needle 

 diminished to such a degree that, even under the most favour- 

 able circumstances, they scarcely ever exceeded 4° ; the ab- 

 sorptive powder of air, instead of proportionately increasing, as 

 I had supposed, seemed to decrease from the 50 per cent, 

 previously observed to 20 and even 15 per cent. 



To obtain greater deflections of the needle, the lower side 

 of the bottom of the brass vessel, which had hitherto remained 

 metallic, was also covered with lampblack, but without favour- 

 able result. The deflections increased but feebly, though the 

 radiating-power of the source of heat must have been six or 

 seven times greater than previously ; on removing the cool- 

 ing-water, however, the deflections increased to such a degree 

 as to become useless. 



The soldered junction of the thermopile, when placed 100 

 millims. below the bottom of the brass cylinder, was thus 

 nearly entirely withdrawn from the influence of the source 

 of heat, and was dependent only upon the temperature of the 

 cooling water. This ought not, of course, to have affected the 

 relation of the forces which deflect the needle, according as the 

 space was exhausted or filled with air, if in both cases the influ- 

 ence of the cooling-water had remained proportional to the 



