204 PROCEEDINGS OF THE AMERICAN ACADEMY 



Calculations. — The unit of heat employed, the calorie, is the quan- 

 tity required to raise one kilogramme of water one degree at the 

 temjierature of the experiment. The mean temperature after mixing 

 is calculated by the following formula : 



_ 04+^0 Tg + BCn-t) . 

 in which 



T„^ = mean temperature. 



A = water equivalent in grammes of solution A in calorimeter. 



d = water equivalent in grammes of calorimeter, stirrer, and ther- 

 mometer. 



B = water equivalent in grammes of solution B. 



Ta = observed initial temperature of solution A in calorimeter. 



71 = observed initial temperature of solution B in flask. 



t = correction in degrees to make the reading of the two thermom- 

 eters correspond. 



The temperatures 2'^ and 7\, recorded in the tables, are those 

 actually observed. In the calculation of T„^ the correction d is used, 

 but no correction is made for difference between the observed zero 

 and the actual zero, as it would not affect the final result. The tem- 

 perature recorded as T^ is the difference between the calculated mean 

 temperature and the resultant temperature ; in other words, the change 

 of temperature due to the reaction, and is positive unless otherwise 

 specified. 



The column flu the tables gives the factor by which T^ is multi- 

 plied to reduce the heat disturbance of the reaction in calories to 

 molecular jiroportions, and H is this final result. The factor y is 

 obtained as follows : 



f=(A-{-B-\-d) X "'°^' 7?" (of alum). 

 •^ ^ ' 1 / '^ grms. taken ^ ^ 



In all the experiments of this paper the water equivalent of the 

 calorimeter and its belongings is 8.5 grm., while A and B are the 

 volumes in culdc centimeters of the constituent solutions reckoned as 

 grammes of water (Berthelot's method of calculation*), each numeri- 

 cally equal to 500 c.c. unless otherwise specified, t = .008°. 



Limit of Error. — The reading of temperature is by far the most 

 considerable element limiting the certainty of experimental observa- 

 tion. The quantity of alum used, 20.667 grm., is weighed to one 



* Essai de Me'canique Cliimique, i 190. 



