﻿140 M. C. Marignac on the Specific Heats, Densities, 



only reached at the end of three or four minutes, so that the 

 slightest difference in the action of exterior causes, or in the 

 agitation of the balloon, may exercise a perceptible influence on 

 the result. Very concordant results are nevertheless obtained 

 when the external temperature remains nearly constant, provided 

 great care is taken to keep the temperature of the calorimeter, 

 for more than five minutes, absolutely invariable. 



By reason of the small specific heat of sulphide of carbon, I 

 should have been obliged, if I had wished to make use of water 

 in the experiments for comparison, to put only 10 grammes of it 

 into the balloon, which would not have been sufficient to com- 

 pletely bathe the thermometer ; and it is to be feared that, under 

 these conditions, the experiments would have been hardly com- 

 parable. I therefore preferred to take for comparison sulphide 

 of carbon itself, a liquid the specific heat of which is well known. 

 M. Regnault has determined very carefully, and has given the 

 following formula for the calculation of the quantity of heat ne- 

 cessary to raise its temperature from to t° : — 



Q=0-2352314* -f 0-00008143131 * 2 . 



By means of this formula it is calculated that the mean spe- 

 cific heat, between 17° and 44°, the limits of my experiments, 

 is 0*2374. The experiments of M. Him* indicate a rather 

 higher number ; for he finds 0*23878 for the specific heat at 30°. 

 I have adopted the mean number 0*238. 



Before commencing the determinations which relate to these 

 solutions, I made a great number of preliminary determinations 

 on water, in order to know the conditions necessary to be fulfilled 

 to make them comparable. 



I ascertained that as long as the quantity of water in the bal- 

 loon remains the same, we may vary its temperature at the 

 moment of immersion, even to a considerable degree, without 



T — P 

 producing any sensible change in the value of the ratio r = — f — ■ ; 



so that all the experiments made in these conditions lead to the 

 same number for the equivalent in water of the balloon and the 

 thermometer contained in it. 



But if we change the weight of water in the balloon, we find 

 a different number to express the equivalent in question, and 

 the new series of experiments is not comparable with the former. 



It is easy to account for these results, if we remark that there 

 are two principal causes of loss of heat — the exterior radiation 

 of the calorimeter, and the fact that, at the moment the maxi- 

 mum is indicated, the liquid in the balloon has still a tempera- 

 ture higher than that attributed to it in the calculation when it 

 is supposed to be brought to the temperature of the calorimeter. 

 * Ann. de Chim. et de Phys. 4th Series, vol. x. p. 84. 



