108 Proceedings of the Rouol Societif of Victoria. 



calorimeter began to follow the diurnal variations with a certain 

 amount of diminution of amplitude, but its mean tenperature 

 was not perceptibly higher than that of the room in which it 

 stood ; nor was any perceptible difference to be noticed when 

 the observations were continued after the motor had been 

 stopped. No correction for the heat generated by the stirrer 

 has therefore been applied. 



The question of applying a correction for the loss of heat 

 by radiation during an experiment was much more important. 

 Originally the usual plan was followed of noting the amount of 

 cooling that occurred in a time, directly after the experiment, 

 of equal length to that required for the experiment itself, and 

 it was found that this amount was, in the summer months, 

 quite imperceptible, the rate of cooling being extremely slow. 

 Recently, however, with a view to increased accuracy that may 

 be expected from the use of an electrical-resistance thermometer, 

 the radiation of this calorimeter was thoroughly tested ; the 

 results, embodied in the curves appended to this paper, led to 

 a modification of the apparatus consisting in the inti'oduction 

 of a water-jacket in place of the second air-space. 



The " curve of cooling " obtained from the calorimeter in its 

 original state shows that whenever the contents have been 

 freshly heated the cooling is much more rapid at first than it 

 would Ije under the same external and internal temperatures, 

 it' these had been reached by gradual cooling from a higher 

 temperature. The explanation lies in the fact that, after 

 gradual cooling has been going on for some time, the various 

 air-spaces and shields have acquired the temperatures correspon- 

 ding to steady flow, and it is this steady flow that is observed ; 

 on the other hand, when the calorimeter has just been heated 

 all the air-spaces and shields are colder than they would be in 

 the other case, and thus heat leaves the calorimeter to warm 

 up the air and metal, and thus the rate of cooling is very 

 much faster at first than it is when the steady state has been 

 reached. With this arrangement it is therefore a matter of 

 doubt how exactly to determine the radiation correction ; during 

 the first part of the experiment the temperature of the 

 calorimeter is rising and the rate of cooling is probably rapid, 

 but on the other hand the difference of temperature is smaller, 



