386 British Association for the Advancement of Science. 



and d in different elastic media, we shall have data for ascertaining 

 their relative capacities for caloric. Such a method, however, though 

 theoretically exact, is beset with difficulties so great that it may be 

 considered as practically impossible. The artificial gases, as usually 

 collected, are saturated with moisture, a state in which they are 

 quite unsuited for the necessary experiments ; and even though this 

 difficulty were overcome, it would probably be impossible to deter- 

 mine their dew-points by direct experiment. 



If, however, we suppose that/'' =0, or that the gas is perfectly 



dry, the above value of a will become £ — -x — , an expression in- 

 volving no unknown quantities but/' and d, and which will there- 

 fore enable us to calculate the specific heat of a gas when we have 

 observed the stationary temperature t\ to which, when in a state of 

 perfect desiccation, it brings the wet-bulb thermometer. In order 

 to the determination of t', and of t—t'z=d, the following method of 

 experimenting was, after a trial of several others, finally adopted. 



Into a bent tube, a b c d e, about 50 inches long, and -rVhs of an 

 inch in diameter, oil of vitriol was poured to the height marked by 

 the horizontal line x y, and to one extremity of this siphon a pair 

 of bladders furnished with stopcocks were attached, through the 

 intervention of a three-armed copper pipe, while to the other extre- 

 mity of the apparatus there was connected by a caoutchouc collar a 

 glass tube carrying the dry thermometer D, and wet one W. Mat- 

 ters being thus arranged, an assistant pressed, by means of a deal 

 board, first upon the bladder A, containing atmospherical air, and, 

 when it was exhausted, upon the bladder G, containing the gas 

 which was the immediate subject of experiment. The air, in pass- 

 ing through the oil of vitriol, was deprived of its vapour, and in 

 subsequently traversing the tube containing the thermometers, pro- 

 duced in the wet one such a reduction of temperature, that, upon 

 continuing the experiment as rapidly as possible with the gas, the 

 wet thermometer soon acquired a stationary temperature,— which, 

 when attained, was, as well as the indication of the dry instrument, 

 carefully noted. The residual gas was now-passed into a glass jar 

 on the mercurial trough, with a view to a subsequent analysis; and 

 both bladders being refilled with atmospherical air alone, a second 

 experiment was performed precisely as just described. 



From the values of t t' obtained in the first experiment, we ob- 



