40 Mr. P. J. Hartog and Dr. J. A. Harker on 

 We may write that 



or 



L = J^>_(T-«,) 



If we take account of the changes in specific heat of the 

 bodies concerned, between the limits of temperature which 

 enter into the experiment, and if C = the sp. heat of water, 

 and ^ = the calorific equivalent of the calorimeter and 

 contents, both being expressed as functions of the tem- 

 perature, we have 



'■ j. 



w I cdt 



' Cdt 



W 



The variation of specific heat with temperature will 

 affect mainly the second term, as the interval (t f — 1 ) is 

 always small. The differences between the results of 

 different observers do not allow of this correction being 

 accurately taken into account at present, and it is not 

 likely to affect the value of L to more than the extent of 

 m part. 



The first difficulty which presents itself with the 

 apparatus of M. Berthelot is the determination of the 

 initial temperature t . During the preliminary period, the 

 thermometer in the calorimeter is rising, owing to the 

 radiation from above. At a certain moment the rate of 

 the rise will shew a sudden increase, namely, when 

 the first portion of vapour is condensed. But this 

 will certainly occur before the liquid begins to boil. 

 M. Berthelot, however, takes as his initial temperature the 



