MR HAYCRAFT on the Specific Heat of the Gases. 197 



This last method was superior to that of Dr CRAWFORD, inas- 

 much as greater quantities of Gas could be employed. In other 

 respects it was far inferior, because the experiments were not, 

 strictly speaking, comparative. Atmospheric air, whose capacity 

 was their standard of comparison, was subjected to trial, and the 

 results were remarked. The other Gases, at different periods, 

 with the surrounding media of different temperatures, and 

 under different barometrical pressures, were examined ; this plan 

 involved endless and very difficult calculations, in order to adjust 

 those differences. But the greatest imperfection in those expe- 

 riments, was the neglect of depriving the Gases of their watery 

 vapours previous to their examination. The apparatus itself 

 would not admit of this, because the water employed in the pro- 

 cess would necessarily keep the gas and the whole apparatus in 

 a state of moisture. Besides, this very great source of error was 

 materially increased by the high temperature to which the 

 Gases were exposed, being a condition in which they are disposed 

 to unite with a greater quantity of watery vapour than at ordi- 

 nary temperatures. Considering the subject in this point of 

 view, therefore, the experiments of Messrs DE LA ROCHE and 

 BERARD may be supposed to determine the capacities of the dif- 

 ferent Gases united with watery vapour at the boiling point, but 

 by no means of those Gases in their dry state, and at ordinary 

 temperatures. 



The apparatus now to be described, will perhaps be found to 

 unite the advantages, and avoid the defects of both methods. 



It consisted of two hollow brass-cylinders, (see Plate VIII.), 

 in each of which was a piston attached to a spindle by means of 

 two levers of equal length ; to the spindle was attached another 

 lever, terminating in a handle, to be moved by an assistant. Each 

 cylinder was closed at each end, excepting where the tubes were 

 attached, which served to conduct the Gases. By means of four 

 valves to each cylinder, fixed in such a way as, though difficult to 



