534 



PROCEEDINGS OF THE AMERICAN ACADEMY. 



It will be noticed that the figure in the last column falls from 537.4 

 calories per gram, the result of the fastest determination, to 531.8 

 calories, in the two slowest, — a change of about 1 per cent. 



This definite march of the results seems to be referable only to the 

 loss of heat by premature condensation in the part of the tube between 

 the trap and the condenser, — a defect in the method which is inevi- 

 table, for the vaporizer could hardly be brought closer to the calorim- 

 eter than it was in these experiments. 



When the results are plotted, the values for the heat of vaporization 

 being laid out in the direction of ordinates, and the time required for 

 vajjorization of 1 gram laid out in the direction of abscissai, the linear 



X 



Figure 6. The Heat of Evaporation of Water; First Series. 



Time in fractions of a minute is plotted in the direction of abscissa;, and 

 heat of evaporation (in 21° calories) in the direction of ordinates. The dotted 

 line is an extrapolation, giving the value for a hypothetical instantaneous 

 experiment. 



tendency of the results is manifest. The greatest departure of any 

 single result from the straight line representing their average tendency 

 is only 0.6 calorie, or 0.1 per cent of the total thermal quantity being 

 measured. This corresponds to an error of thermometric reading of 

 0.004°. The agreement therefore is as close as could be expected. 



By extrapolation to zero time the value 539.6 (cal. 21°) is obtained 

 for the heat of vaporization of a gram of water weighed in air — a 

 value from which the effect of premature condensation must have been 

 eliminated, because there is every reason to believe that this disturbing 

 phenomenon is directly dependent upon the time consumed in the 

 experiment, and that if the experiment could be performed instantane- 

 ously the error would wholly disappear. This value becomes 538.8 in 

 terms of the calorie at 15°. 



Inspection of the curve shows that the loss of heat in 1 minute 

 must have been 7.0 small calories from this apparatus under these cir- 

 cumstances, and each of the results is evidently to be reduced to a 

 common basis by adding to it this value multiplied by the fraction of 

 a minute required for the vaporization of 1 gram. The results, then, 

 become respectively 539.9, 539.4, 539.7, 540.2, 539.1 and 539.C in the 



