14 SCIIMITZ, On a Method of Ice Calorimetry. 



The value used for the latent heat of water is 8o"0. 

 This is (very nearly) the number obtained by Bunsen in 

 1870. Bunsen's unit of heat is the one one-hundredth part 

 of the heat required to raise unit mass of water from 0*^0. 

 to lOO^C. ; this is shown by recent researches to be very 

 nearly equal to the heat required to raise unit mass of 

 water from \']° to 18*'.* 



The Time of Immersion in the Calorimeter. 



In the preceding tables for iron and aluminium a 

 variation in the time of immersion in the calorimeter 

 produces no well-marked difference in the result. But 

 in several series of experiments in which the time of 

 immersion was varied I found a considerable variation in 

 the result. For instance, a copper cylinder weighing 

 approximately lOO grammes gave the numbers shown in 

 the table on p. 15. 



Here an increase in the time of immersion gives 

 apparently an increased mass of ice, and from the bad 

 conductivity of ice and the possible existence of a film of 

 air between the metal and its ice-jacket, it might be 

 suspected that it would take a long time for the tempera- 

 ture of the metal to reach that of the surrounding water. 

 But the true explanation is different. This is at once 

 evident when the ice is weighed during immersion. 

 Details of a single experiment of this kind will suffice. 

 The object experimented on was a copper sphere 

 weighing between 62 and 63 grammes. It has already 

 been stated that the ice-coating does not completely 

 enclose the piece of metal. The imprisoned air always 

 keeps open one or more passages from the interior 

 to the exterior of the ice-coating. In the case of 



* The values given in this paper for results obtained by the method of 

 mixtures are expressed in terms of the latter unit. 



