2 Harold N. Allen, 



portance is to be attached than as showing the short time 

 needed for heating the body experimented on to ioo° C, 

 and the very slow rate at which the weight of the sub- 

 stance and condensed water changes after that time. Ex- 

 periments on the latent heat of evaporation of alcohol 

 remained without definite result owing to water contained 

 in it, but a certain amount of success was obtained with 

 bisulphide of carbon, though here too the same difficulty 

 was encountered. 



I have to thank Professor F. Kohlrausch, Director of the 

 Physical Institute in Strassburg, for the great help given in 

 the course of these experiments, and the kindness he has 

 uniformly shown me. 



THEORY OF THE METHOD. 



Suppose that a number of substances, with weights W-^, 

 W2, Wz, etc., and specific heats Si, So, S^, etc., and with the 

 common temperature f, are plunged into the saturated vapor 

 of a liquid, the latent heat of vaporization of which is A,, and 

 the temperature T. Then a certain weight tv of the liquid 

 will be condensed such that the heat given up in condensa- 

 tion is equal to that required to raise the temperature of all 

 the bodies from f to T°. 



W\ = ( WiSi + W,S^ + W,S, + etc.) (T- 1). 



This affords a means of determining the specific heat of one 

 of the bodies if the specific heats of the others and the latent 

 heat \ are known, or of finding \ if the specific heats are 

 known. The great difficulty is in the determination of zu, — 

 the weight of the condensed liquid, — as a removal of the sub- 

 stances from the vapor causes in general instant evaporation. 

 Both Bunsen and Joly adopt the expedient of weighing 

 the whole suspended in steam, the difference between the 

 methods being that while Bunsen plunges substance and 

 carrier into a vessel of steam, Joly surrounds them with 

 steam as quickly as possible by suddenly passing it into the 

 vessel in. which they hang. 



196 



