436 



Mr. G. J. Stoney on the Penetration of 



ally to a maximum at a certain tension, after which they fall 

 away to cipher if the exhaustion is continued indefinitely. 

 This description agrees with the form determined from the 

 observations and which is plotted down in fig. 5 ; so that the 

 comparisoUj though necessarily very imperfect, lends support, 

 so far as it goes, to the conclusion that the apparently anoma- 

 lous escape of heat which De la Provostaye and Desains in- 

 vestigated was due to penetration. 



14. Hitherto I have used only the observations recorded in 

 the memoir first published in the Comptes Rendus for 1845. 

 In their second memoir, first published in the Comptes Rendus 

 for 1846, De la Provostaye and Desains record observations 

 made with the silvered thermometer within their blackened cy- 

 linder, charged successively with hydrogen, carbonic an- 

 hydride, protoxide of nitrogen, and a mixture of air and 

 hydrogen. 



In carbonic anhydride the results of experiments are repre- 

 sented by fig. 7*. In this gas the total rate of cooling in- 





creased when the tension was diminished from 12 to 4 millims. 

 This observation took De la Provostaye and Desains so much 



* De la Provostaye and Desains record the following observations, from 

 which the curve in the text has been plotted down, the ordinates "being 

 drawn proportional to the reciprocals of the observed times of cooling : — 



Tensions. . . . 



85 millims. 



12 millims. 



4 millims. 



Times 



m s 

 19 42 



m s 

 19 38 



m s 

 17 59 



The reciprocals are proportional to 282, 283, and 309. 



