690 



PROCEEDINGS OF THE AMERICAN ACADEMY. 



the calonmetnc system increase with rise in temperature, the water- 

 equivalent of the calorimetric system varies but slightly with change 

 in temperature. In fact, the roughly calculated water-equivalent of 

 our calorimetric system at 20° varied from its water-equivalent at 25° 

 by less than 0.03 per cent. The estimated water-equivalents used in 

 this comparison were, of course, not accurate, but the ratio between 

 them must be very close to the truth. On the assumption that the 

 relation between the weights of the water and the metallic parts in the 

 calorimetric system of other investigators did not differ much from 

 ours, it is evident that but little error will arise in comparing our re- 

 sults obtained at an average temperature of 21.4° with the results of 

 Stohmann and of Fischer at average temperatures of 17° and 15° 

 respectively. 



The results of previous investigators, expressed both in gram-calories 

 and kilojoules, are tabulated below. 



Results of Other Experimenters. 



Thus our ratio for the heats of combustion of benzol and sugar, 

 2.534, is about 0.3 per cent higher than 2.527, Stohmann's correspond- 

 ing ratio. It is interesting to note that had we used the value for 

 benzol obtained when the bomb was filled with air at the beginning, 



our ratio would have been 

 same as Stohmann's. 



2.655 

 1.0504 



= 2.528, which is practically the 



" Stohmann and Langbein, Jour. prak. Chem., 45, 313 (1892) ; 40, 77, 81 

 (1889) ; Stohmann, Rodatz, and Ilerzherp, ibid., 33, 103 (1886). 



" Sitzb. Berl. Akad. d. Wisson, 19, 20, 21, 087 (1904). Value corrected for 

 error in calibration of bomb, vide, Zeit. phys. Chem., 53, 164 (1905). 



" Berthelot and Vielle, Ann. Oliim. T'hys. (6), 10, 442 and 458 (1887). 

 Berthelot, Ann. Chim. Rhys. (5), 23 ,193 (1881). 



