12 



PROCEEDINGS OF THE AMERICAN ACADEMY. 



With this apparatus experiments were first made according to each of 

 the usual methods of calorimetric work. Two distinct procedures have 

 been used by various experimenters ; in one case the experiment is started 

 at about the temperature of the room, and a large final cooling correction 

 is added, while in the other case the experiment is started below the 

 temperature of the room, and a small initial correction for warming is 

 subtracted. 



In previous experiments in this laboratory it has been found that the 

 latter of these methods gives the more constant results ; and the same 

 conclusion must have been reached by Stohmann and by Atwater, since 

 they have usually used it. 



Because of the lesser accuracy of the former of these methods, more 

 experiments were made by it than by the latter, in order that the 

 averages might have about equal weight. 



Below are given the results of successive experiments by these two 

 methods. The first column of the table designates the method ; the sec- 

 ond, the number of the experiments; the third, the rise in temperature as 

 calculated by the Regnault-Pfaundler method, but not corrected for ther- 

 mometric lag; the fourth, the correction for thermometric lag, as found 

 in the way already described; and the fifth, the final value with all 

 corrections. 



Preliminary Experiments. 



