6 SCHMITZ, On a Method of Ice Calorimetry. 



of an alteration in the time of transference. They were 

 given by a flat-ended copper cyh"nder, weigifhing a little 

 over loo grammes, suspended with its axis vertical, and 

 carried steadily from cooler to calorimeter. It may be 

 remarked that the time required for complete evaporation 

 of the liquid air from the surface of the cylinder was 1 5 

 to 20 seconds. 



Time of transference 



Apparent 



in seconds. 



specific heat. 



5 



•0796 



10 



■0792 



15 



•0800 



20 



•0772 



25 



•0774 



30 



■0773 



60 - 



•0743 



Subsequent experience showed that the shape of the 

 object here used was not well adapted for obtaining 

 consistent results by the present method ; but the 

 numbers above given are sufficient to show that any 

 heating which may occur in five seconds is not of much 

 importance.* 



* Five experiments with the same cylinder, in which the transference was 

 made as rapidly as possible, gave respectively '0793, '0799) 'o787j '0805, 

 •0806, mean "0798. A small copper cover was made, closely fitting over the 

 curved surface and base of the cylinder, and provided with a hook below so 

 that it could be rapidly withdrawn by means of an attached thread. In three 



