18 



PROCEEDINGS OP THE AMERICAN ACADEMY. 



might, before the experiment, assume the same initial temperature as the 

 acid. The final temperature was designated 6 3 , since it was attained in 

 three minutes. 



Rise of Temperature as Determined by New Experimental Method. 



The average rise of temperature as indicated by these experiments is 

 therefore 2.860 ± 0.001, while that obtained in the previous series 

 after the new correction for thermometric lag had been applied was 

 2.863 ± 0.002. Thus the two series, when all corrections were ap- 

 plied, gave results not differing by an amount greater than the sum of 

 their probable errors, and the average by the new method had only half 

 as large a probable error as that obtained with the old method. 



While it is of course true that this agreement might have been acci- 

 dental, such an interpretation to us seems improbable. 



That these two methods, each of which is apparently satisfactory, 

 should unite on the same value, gives that value a strong presumptive 

 right to be considered as the true value. If this is so, nearly all calori- 

 metric results which have ever been published, where any considerable 

 rate of cooling or warming is involved, have been incorrectly calculated. 

 In cases where the rate of cooling or warming is very slight, as in the 

 admirable work of E. Fischer and Wrede,* the error is probably so small 

 as to be negligible. 



It is a pleasure to acknowledge our indebtedness to the Romford Fund 

 of the American Academy of Arts and Sciences for the apparatus used in 

 this investigation. 



Summary. 



1. A modification of one of the devices used by Richards and Lamb, 

 permitting more careful regulation of the speed of reaction, is described. 



* Fischer and Wrede, Sitzungsber. konig. preuss. Akad. 20, 687 (1904). 



