LOSS OF HEAT DUKING EXPERIMENT. 307 



equal to that which would have resulted for a mean excess of 

 1*5 + 2-6 90 



T = 2 ' 



The resulting correction is very insignificant ; it only amounted 

 to 0-02 of a degree in 1'5. 



We then calculate the total heat, Qi, disengaged during the 

 second stage by multiplying the sum of the masses, reduced to 

 units of water, by the apparent variation of temperature, added 

 to that corresponding to the heat lost. 



Heat lost during the first stage. The loss of heat during the 

 first stage requires a more complicated mode of calculation. 

 This first stage is divided into periods of at least five minutes, 

 according to the rapidity of the heating, until the maximum 

 temperature is attained. The mean temperature of each period 

 is written down, and also the excess of this mean temperature 

 over the initial temperature. 



The duration of the maximum temperature constitutes a 

 separate period, corresponding to the maximum excess of tem- 

 perature. 



The time following this maximum is also divided into 

 periods of five minutes, and opposite each period are written the 

 mean temperature and the excess of temperature. 



At the end of this time it was observed that the rapidity of 

 cooling was, for the same excess of temperature over the initial 

 temperature, exactly the same as in a check experiment made a 

 little later, and in which care was taken to reproduce this 

 excess under the same conditions. 



This verification proves that the reaction was quite complete, 

 and that the data of the check experiment may be applied to 

 the calculation of losses of heat during the reaction itself. 



In fact, this check experiment gives, without any hypothesis, 

 the losses of heat experienced by the calorimeter for a series of 

 excesses of temperature similar to those of the reaction, and 

 under conditions exactly parallel. 



We then write down the loss of heat experienced by the 

 system in one minute for each mean excess of temperature 

 answering to each period ; we multiply this loss by the duration 

 of the period, generally five minutes (except in the case of the 

 maximum, which is longer). We then find the sum of all these 

 losses, and add them to the variation of temperature actually 

 observed. 



To come to figures, it may be said that, the variation observed 

 being + 1/26, the correction was -f 0'234, which will not seem 

 too great in an experiment of such long duration. 



We have now only to multiply the corrected variation of 

 temperature by the sum of the heated masses (reduced to units 

 of water), in order to obtain the quantity of heat, Q 2 , disengaged. 

 It is advisable to give these details, because they afford as 



x 2 



