as deduced by himself and Dr. Suerman. 271 



first, with atmospherical air deprived by the oil of vitriol of 

 it moisture, pressure was made upon it by an assistant, so as 

 to force its contents in a rapid current into the second bell, 

 D, through the tubes containing the wet and dry thermome- 

 ters. During this operation the observer kept his eye, armed 

 with a lens, steadily fixed on the thermometers, and registered 

 the indications of both as soon as the wet one became and 

 continued stationary for a kw seconds. The height of the 

 barometer being now taken, the necessary data were obtained 

 for calculating from formula (A), or 



fit - f _ ^1^ V P 



the elastic force of the vapour still existing in the air of the 

 gasometer. This air being now replaced by one of the gases 

 which were to be the subject of experiment, and this being 

 left during the same time with the air in contact with the oil 

 of vitriol, the very manipulations and observations just detailed 

 were repeated. This same experiment was again and again 

 performed, and it having been ascertained, after a consider- 

 able number of repetitions, that the results were uniform and 

 consistent, and that they might therefore be relied upon, the 

 mean of all the observations was taken, and from this the spe- 

 cific heat of the gas was deduced by means of the formula 



that value being assigned to /" which resulted from formula 

 (A) applied to the preliminary experiments on atmospherical 

 air. The analysis of the gas was next very carefully per- 

 formed, and it having been ascertained that n volumes, e. g. 

 of atmospherical air per cent, were present, the proper cor- 

 rection was applied by the formula 



, ia — e)n 



a — a +-^ '— 



100— w 



already explained, in which e = '267 is the specific heat of 

 air, a' the true specific heat of the gas, and a the specific 

 heat of mixture of air and gas /as previously determined. 

 Such was the course pursued in the case of each of the gases 

 submitted to experiment. 



The particulars of the entire series of experiments are com- 

 prehended in tables (1.) and (2.) Table (3.) contains the 

 final results, alongside which are placed the numbers of De 

 la Roche and Berard, and those of Dulong, for the purpose 

 of comparison. 



