the Specific Heat of Air under constant Volume. 435 



of which was a small and extremely good English one; and within 



the limits of accuracy attainable by a single observer I have 



always obtained the same value. I see no reason why the result 



should not deserve at any rate the same confidence as the older 



experiments with the water manometer. 



Yet the value found above ( = 1*302) would be in disaccord 



with the observed velocity of sound and with the number assumed 



for the mechanical equivalent of heat ; for it would lead to 



319'4 metres for the velocity of sound, taking 0*0012934 as the 



density of dry air at 0° and 756 millims. pressure. Taking, 



with Regnault, the specific heat of air under constant pressure 



as 0*2377, the mechanical equivalent of heat would be 532, 



taking Delaroche and Berard's number (0\2669) it would be 



equal to 473 kilogrammetres. The most recent experiments of 



llegnault have given 330 "3 metres for the velocity of sound, 



c 

 from which — = 1*392, and the mechanical equivalent of heat 



c \ 

 (putting c = 0-2377) would be equal to 437 kilogrammetres. In 

 our experiments y would have to be = 14*5 millims., instead 

 of 10*61, to agree with this result. 



It would be difficult to discover a source of error to this 

 amount in the above measurements. It is, however, advisable 

 to repeat the experiments with improved instrumental means. 

 In the latter we should include first of all a method of pro- 

 ducing the change of density in a time much shorter, but 

 capable of accurate measurement. Both the motion of the 

 cocks and the observation itself would be best effected by mecha- 

 nism. Moreover a metallic manometer of as small moment of 

 inertia as possible should be constructed. Doubtless, too, by 

 using a larger receiver with badly conducting sides, the equali- 

 zation of temperature might be considerably retarded. . 



I doubt not that if these conditions be fulfilled a trust- 

 worthy direct determination may be made by the above method 



c 

 of the ratio — (and thus an important gap in physics be filled), 



° l 

 not merely for atmospheric air, but also (with no greater difficulty) 



for other gases, which is of especial interest. Apart from this, 

 the indication of this simple method of quantitatively determin- 

 ing with approximate accuracy the heating produced by com- 

 pression in gas may be welcome to many a lecturer. 



Gottin^en, January 1869. 



