THE RATIO OF THE CALORIE AT 73° TO THAT AT 20°. 



By Arnold Romberg. 



Received April 21, 1922. 



Presented by P. W. Bridgman. 



The present paper, publication of which has been deferred for 

 various reasons, describes an experiment designed as a crucial test of 

 the correctness of the temperature variation curve of the specific heat 

 of water. The desirability of such a test is apparent from figure 1, 



which gives the results of some of the best researches on the subject, 

 and which shows the great discrepancies existing at the higher temper- 

 atures.^ The work was suggested by H. N. Davis, and carried out in 



1 The task of finding from the published papers the reasons for the differ- 

 ences in these curves must be regarded as hopeless. In connection with the 

 excellent work of Barnes a possible sovu-ce of error may however be pointed out. 

 The heating current in Barnes's calorimeter is carried by a platinum wire 

 between copper terminals at which a certain amount of heat is absorbed and 

 liberated, respectively, due to the Peltier effect. The thermometers are 

 situated within these hollow terminals, and it is conceivable that an appre- 

 ciable error might result. 



