192 



PROCEEDINGS OF THE AMERICAN ACADEMY. 



Incidentally it may be noted tliat Greek letters are used to desig- 

 nate the several two-liter flasks containing different portions of the 

 same acid. 



TABLE III. 



Check Determinations with Water. 



The agreement of these results is highly satisfactory, two of them 

 giving essentially the calculated values and two others deviating by an 

 amount corresponding to only about 0.001°, the probable error of 

 experimentation. The last experiment was made with the last 

 portion of the original 12 liters of acid, nearly four months haA'ing 

 intervened since the first experiment. 



The results recorded in these tables warrant the assinnption that 

 the acid furnishes a definite quantitative source of heat, which may 

 be depended upon; and also that the method of evoh'ing and measur- 

 ing this heat has now been brought to within a range of accuracy as 

 great as is needful under the circumstances. 



Having thus prepared the way, we may now proceed to describe 

 the work with the \'arious solutions under consideration. 



From the above determinations, it was possible to compute the 

 exact amount of heat which would be liberated by the neutralization 

 of a given portion of acid within the temperature limits of the stan- 

 dardization. By replacing the water of these earlier experiments by a 

 definite amount of some otlier liquid, a direct determination of the 

 specific heat of the latter could be made. As has already been stated, 

 the experiments were so conducted — by regulating the amounts of 

 liquid used — that the temperature rise in all the determinations was 

 substantially the same. The case of HCl 100 HoO may be taken as 



