Detennining the Specific Heat of a Liquid. 109 



while later on the difference of temperature is larger but the 

 jackets have had time to become wai^med up. In order to 

 avoid tliis uncertainty in tlie radiation correction, the second 

 air-space was converted into a water-space containing about 

 1250 grams of water whose temperature could be read by two 

 thermometers. This made the heat-capacity of the jacket so 

 large that the small amount of heat radiated coukl not cause 

 any perceptible increase in the teinperature of the jacket and 

 consequently in the rate of cooling, and the second curve shows 

 that the rate of cooling, though now much faster than before, is 

 the same whether the contents of the calorimeter have just 

 been heated or whether cooling has been going on for some 

 time. 



Throughout these experiments the author has had the great 

 advantage of the personal advice and assistance of Professor 

 T. R. Lyle, in whose laboratory at the Melbourne University 

 the research was carried out ; for valuable assistance in the 

 construction of apparatus the author is also indebted to Mr. 

 L. Meyei', the mechanical assistant at the laboratory. 



The appended tables of figures from actual experiments will 

 best show the accuracy actually attained, but in addition it 

 may be well to summarise the salient features of the method : — ■ 



1. The total absence of evajooration. 



2. A calorimeter very efficiently protected and with a very 



small and very constant radiation loss. 



3. Efficient contact between the hot liquid run into the 



spiral and the cold liquid in the calorimeter. 



4. Efficient mechanical stirring and free circulation in the 



calorimeter. 



5. Ease of repeating and checking experimental observa- 



tions. 



6. A degi'ee of accuracy of at least 1 in 500. 



