RICHARDS AND MATHEWS. — HEAT OF EVAPORATION. 517 



Brown, working with a single apparatus, passed a current of meas- 

 ured value for a measured time through a wire of known resistance, 

 immersed under the litjuid. The portion evaporated was weighed ; 

 thus all the data for calculating the heat of evaporation were obtained. 

 As in Ramsay's and Marshall's method the liquid must be at the boil- 

 ing point before the measured current is passed through, but no 

 knowledge of the specific heat of the liquid is necessary. The results 

 obtained by Brown agree well among themselves, but are even higher 

 in value than those obtained by Marshall and Ramsay. 



F. Henning ^4 used a method similar in principle to that used by 

 Brown, but superior in detail, because greater precautions were taken 

 to provide against possible errors. The original paper must be con- 

 sulted for particulars, most of which do not immediately concern the 

 present investigation. The greatest difficulty in this interesting and 

 painstaking work seems to have been that considerable heat was 

 necessarily carried away by the conducting wires, and there seems to 

 have been no entirely satisfactory method for measuring or calculating 

 this uncertain quantity. As the mean of his determinations made at 

 an average temperature of 100.59°, Henning found the heat of vapor- 

 ization of a gram (weighed in vacuum) of water to be 538.25 calories 

 at 15°. A number of determinations were made also at reduced pres- 

 sures, but they do not concern us here. 



A modification of this electrical method has been proposed by A. 

 Cameron Smith, 25 who suggested that the electrically heated vapor- 

 izer be suspended from one arm of an analytical balance. This ap- 

 paratus is suitable only for lecture demonstrations; many possible 

 sources of error tend to diminish its accuracy. 



Among these many methods, Kahlenberg's modification of Berthe- 

 lot's seemed to be preferable, partly because of its simplicity, and 

 partly because it retains the advantage of the electrical method of 

 heating while avoiding the disadvantages. Hence we used this 

 method as the basis of our own, seeking to discover and correct any 

 sources of error which might still remain in it. 



The Evolution of the Present Apparatus. 



One of the most serious causes of error in all calorimetric work is the 

 more or less uncertain correction for cooling. Hence one of the first 

 steps of the present research was the application of the new method of 

 adiabatic calorimetry to the problem.26 This method, first put into 



2* Loc. cit. 26 pi-oc j^Qy goc., Edinburgh, 24, 450 (1903). 



2^ Richards, Forbes, and Henderson, Proc. Am. Acad., 41, 1 (1905). See 

 also Richards and Jesse, as well as Richards and Burgess, Jour. Am. Chem. 

 Soc, 32, 268 (1910) and 32, 431 (1910). 



