538 PROCEEDINGS OF THE AMERICAN ACADEMY. 



work of others, speaks strongly in its favor. The value given by Reg- 

 nault, 536.7, is unquestionably too low. Henning's value, 538.25 

 cal.i5 was found at 100.6°. Corrected to 100°, this would be 538.7 

 cal.150, a value only a trifle above ours. Joly,^^ Harker,33 and Smith, 3* 

 likewise obtained values near 540. 



In conclusion we are glad to express our gratitude to the Carnegie 

 Institution of Washington, for generous pecuniary assistance. 



Summary. 



1. The method of Berthelot and Kahlenberg for determining heats of 

 vaporization has been modified in such a way as to diminish greatly 

 the errors inherent in the experimentation. A Dewar vessel was used 

 as a vaporizer and the calorimetric work was strictly adiabatic. 



2. A serious persistent disturbing effect, amounting to about 0.1 

 per cent of the total per minute, was eliminated by conducting experi- 

 ments at different speeds and extrapolating the results to a hypothet- 

 ical instantaneous experiment from which the disturbing effect may be 

 supposed to be eliminated, because this was found to depend essentially 

 upon time. The disturbing effect was probably premature condensa- 

 tion in the very narrow zone between the vaporizer and the condenser. 



3. The heat of vaporization of a true gram of water was found by this 

 method to be 538.9 cal.2i<' or 2.251 kilojoules per gram. A gram mole- 

 cule therefore requires 9.707 Cal.2i° or 40.54 kilojoules, when the 

 vaporization is conducted at 100° (0 = 16.000, 1 Cal.2i° = 4.177 

 kilojoules). 



4. Comparison of this figure with the results of others shows that 

 the method is trustworthy and suitable for general use. 



5. Numerous other liquids also have been used in the apparatus, 

 and consistent results with them have been obtained. These will be 

 communicated in a future paper, when the specific heats of the liquids 

 have been determined. 



32 Phil. Trans., 186, 322 (1895). Dependent on Joly's value for the mean 

 calorie between 12° and 100° as determined by the steam calorimeter. 



33 Loc. cit. 



34 Loc. cit. 



