and Diathermancy of Air and Hydrogen. 



411 



che more rapid cooling of a body in the latter medium, even if 

 the conductivity of both gases were the same. 



According to the results of the fourth series of experiments, 

 the air impedes the passage of heat-rays in a striking manner. 

 The air employed had been, as stated, carefully dried ; never- 

 theless about half of the incident rays were absorbed by the 

 layer of air between the bottom of the heating- vessel and the 

 soldered junction of the thermopile. As this result was in con- 

 tradiction to the experience hitherto collected, further obser- 

 vations were required in order to confirm it. 



To increase the deflection of the needle, the upper surface 

 of the soldered junction was covered with lampblack ; by this 

 means the desired result was attained. 



In the following Tables, T denotes the temperature of the 

 water in the brass cylinder at the beginning of an experi- 

 ment, T' the temperature of the water when the needle had 

 reached its maximum deflection t. The diff'erences of tempe- 

 rature at which r was obtained are given under t, and T — Hs 

 therefore the temperature in the brass cylinder before the in- 

 troduction of the hot water. Finally, the figures under / 

 give the duration of each experiment from the time when the 

 hot water was introduced, and column /> represents the interior 

 tensions. 



Table V. 



Diathermancy of Dry Air at different Densities. 



T. 



T'. 



t. 



T. 



z\ 



millims. 



100 r 



t ' 



41-8 



36-3 



291 



120 



6 



1-5 



41-2 



510 



430 



32-5 



13-5 



7 



1-5 



41-5 



540 



46-5 



38-2 



17-6 



4 



1-5 



460 



630 



550 



47-4 



20-2 



5 



1-5 



42-6 



840 



70-3 



63-2 



27-1 



(29-0) 



5 



15 



45-9 



48-0 



41-3 



311 



60 



7 



748-0 



19-3 



520 



45-3 



34-3 



7-0 



7 



7440 



21-8 



85-0 



70-1 



611 



130 



5 



748-0 



212 



81-0 



73-2 



65-7 



13-5 



6 



7504 



20-5 



67-3 



58-8 



50-8 



12-7 



7 



520-6 



250 



82-4 



67-5 



60-3 



16-4 



7 



414-5 



27-2 



64-0 



54-5 



46-5 



13-5 



7 



254-7 



290 



620 



510 



43-2 



160 



7 



108-0 



37-0 



64-0 



53-4 



46-9 



18-8 



5 



121 



401 



660 



54-5 



48-0 



100 



7-5 



747-3 



20-8 



The first five experiments in the above Table, in which the 

 thermal motion could have been caused by radiation alone, 

 furnish a very convenient test of the correctness of the method 



