﻿and Expansions of some Liquids, 141 



Now the time occupied by the experiments remains the same 

 whatever the elevation of temperature obtained, since the maxi- 

 mum is attained, towards the end, so much the more quickly as 

 the excess of temperature, being greater, determines a stronger 

 radiation. Hence the first of the causes of loss above mentioned 

 is nearly proportional to the elevation of temperature attained, 

 so that the ratio of the variations of temperature of the balloon 

 and of the calorimeter is always reduced by it in the same pro- 

 portion ; consequently the ratio remains constant for all the ex- 

 periments of one and the same series. 



As to the loss of heat produced by the second cause, it is pro- 

 portional both to the amount by which the interior temperature 

 exceeds the exterior (and this will be proportional to the heigh 

 of the temperature of the calorimeter), and to the weight of 

 water contained in the balloon. 



It follows that only those experiments can be compared in 

 which the liquid contained in the balloon represents the same 

 value in water, but it is not necessary that the initial temperature 

 shall be always the same. 



Consequently I united in one series all the experiments, made 

 at different times on pure sulphide of carbon, to deduce from 

 them the value of the balloon and the thermometer, only taking 

 care to use always nearly the same weight of that liquid (about 

 43 grms.), and afterwards to take of the various solutions weights 

 equivalent to that quantity as regards their specific heat. 



In stating the results of my observations, I will commence 

 with the aqueous solutions. 



For the formulae of compound bodies I have adopted the atomic 

 notation, hydrogen being taken as unity; they are therefore 

 identical with those of M. Thomsen. 



In the Tables of results I designate by 



n the number of molecules of water to one molecule of the 

 substance dissolved ; 



c the specific heat of the solution per unit of weight ; 



p the molecular weight of the solution ; 



C =pc the molecular heat, both that given by experiment, and 

 that resulting from an empiric formula indicated further on 

 for each case. 



A final column gives the value of C — iSn ; that is, the dif- 

 ference between the molecular heat of the solution and that of 

 the water contained in it. To obtain these differences I have 

 used the values of C calculated by means of the interpolation- 

 formula, in order that the law of their progression might not 

 be masked by the uncertainty of the values of C given by expe- 

 riment in the case of very dilute solutions. 



Although I give in the Tables the value of c to four decimals, 



