820 EXPERIMENTS ON SPECIFIC HEAT. 



the heat of the hot water is applied to heating the glass 

 vessel and the surrounding air. But if we neglect this 

 loss we may say that 500 grammes, or 0'5 kilogramme 

 of hot water, in cooling from 95 to 55, parts with 

 0-5 x 40 = 20 thermal units or kilogramme-degrees, 

 and these 20 kilogramme-degrees are just sufficient 



20 

 to raise the temperature of O kgr< 5 through = 40, 



U't) 



that is, from 15 to 55. 



But if equal weights of different substances, each at 

 a different temperature, be mixed, the temperature of 

 the mixture will no longer be the mean of the two tem- 

 peratures. Into a somewhat large flask l kgr of water is 

 poured, and l k f of mercury is heated in a small flask 

 until the temperature of the mercury is as nearly as 

 possible 74 higher than that of the water ; for example, 

 if the water is at 15, let the mercury be heated to 89. 

 As soon as the mercury has the desired temperature, it 

 is poured in a thin stream into the water, briskly stirring 

 the mixture at the same time. The water and the 

 mercury will soon have the same temperature, which 

 will be found to be only 2 higher than that of the water 

 before adding the mercury. If the water was at 15, and 

 the mercury at 89, the temperature of the mixture will 

 be 17. One kilogramme of mercury, therefore, in 

 cooling from 89 to 17, that is, through 72. has only 

 parted with as much heat as is required to raise the 

 temperature of 1 kilogramme of water through 2, or 

 2 thermal units. Now, if one kilogramme of mercury, 

 in cooling through 72, parts with 2 thermal units, it 



2 1 



will part with only -- = th of a thermal unit in 



i ~~ ob 



