Heats of Evaporation of different Liquids. 39 



and intend to apply the method to a number of other liquids 

 that we have not yet had time to investigate. 



In such an arrangement as this there is no direct heat 

 measurement, and therefore a fruitful source of error is elimi- 

 nated. As the current passes, heat is developed in each 

 carbon filament ; and this heat is entirely expended in con- 

 verting a portion of the liquid into vapour. The thermal 

 capacity of the liquid no longer needs to be known, since the 

 experiment is conducted entirely at the temperature of its 

 boiling-point. 



Each vessel is weighed at the temperature of the atmo- 

 sphere at the beginning and end of the experiment ; and the 

 loss of weight of the vessel during the experiment gives the 

 amount of liquid vaporized. If the same current be passed 

 for the same time through two filaments of equal resistance, 

 equal amounts of heat will be developed : hence the ratio of 

 the loss of weight of the first vessel to the loss of weight 

 of the second is the inverse ratio of the heats of vaporization 

 of the liquids employed. 



If the filaments have not equal resistances, the amounts of 

 heat developed will be directly as the resistances. If the heat 

 of evaporation of any one liquid be known, the absolute value 

 of the heat of evaporation of any other liquid can be calcu- 

 lated from the ratio directly found. 



3. [Note, July 1895, by D. Marshall. — At the time when 

 this work w r as begun, the only liquid for which L was at all 

 certainly known was water ; but it was not considered ad- 

 visable to compare directly with water in all the experiments : 

 (i.) because the heat of evaporation of water is so very much 

 greater than that of any other liquid ; (ii.) because it is so 

 exceedingly difficult to get it sufficiently pure ; and (iii.) 

 because the continued use of water seemed to make the 

 apparatus more liable to break. 



Ethyl alcohol was in every way more suitable for a standard, 

 and I therefore tried to make a particularly careful comparison 

 between alcohol and water. But both in those first attempts, 

 and again in several other attempts that I made later on, I 

 had the greatest difficulty in getting even moderately con- 

 cordant results ; and I have never, when working with water, 

 been able to approach the degree of accuracy that I know 

 from my other results I am justified in expecting. 



Thus I was finally obliged to conclude that water is not a 

 liquid to use in these experiments at all, much less is it fitted 

 to form our ultimate standard. It became necessary there- 

 fore to find some other liquid for which L was accurately 

 known ; and as we did not feel sufficient confidence in any 



