1895.] The Latent Heat of Evaporation of Water. 



215 



There can be no question as to the accuracy with which Q 6 e was 

 measured. The term Q s 5 is probably the one least accurately deter- 

 mined, but the probable error is not 1 in 100, and the quantity 2gr 

 was found to a far greater degree of accuracy than the above table 

 shows to be necessary, as it was certainly known to 1 in 1000. 



I will not here attempt any proof of the above statements, full 

 particulars being given elsewhere, nor do I propose to describe the 

 mode of carrying out an experiment, as I find it impossible to suffi- 

 ciently compress the account of the operations. 



I performed a large number of experiments when the evaporation 

 of the water was promoted by the passage of a gas. Extreme pre- 

 cautions were taken to dry the gas and to prevent any carrying of 

 "heat by it either to or from the calorimeter, but the results of these 

 experiments cannot be called satisfactory, as the differences between 

 individual experiments at the same temperature in extreme cases 

 amounted to as much as 1 per cent. The 'following table gives the 

 results : 



Table II. 



The rate of evaporation was greatly varied during these experi- 

 ments, for the potential difference was changed from that of 1 to 3 

 Clark cells, hence the rate of evaporation was in some cases nine 

 times as great as in others. 



There is one curious coincidence, which, however, may be merely 

 fortuitous. In two of these experiments nitrogen was passed through 

 the water instead of air, and these two experiments give almost 

 exactly the same value as my final results, viz. : 



Temp. 

 24-96 

 39-98 



L (Nitrogen expts.). L. Final values 



581-68 5817 



572-72 572-7 



The method finally adopted was that of allowing the water to fall 

 drop by drop on to the interior surface of the silver flask, and causing 



identical, and it will be seen that any errors of this kind would be eliminated during 

 the calculations. This remark, however, does not apply to Professor Schuster's 

 correction for the specific heat of the air displaced, for we are not dealing here 

 with differences in rate of rise caused by the displacement of air by water. His 

 correction (1 in 4000) is therefore included in the above value. 



