512 PROCEEDINGS OF THE AMERICAN ACADEMY. 



result. Most of the methods used, particularly the earlier ones, be- 

 long to the latter class; because, although those belonging to the first 

 class have the advantage of being independent of the values for the 

 specific heats of the substances used, they are more complicated in 

 manipulation and calculation, and none has proved to be entirely 

 satisfactory. 



Joseph Black 1 was the first to make the interesting observation, 

 that in the formation of the vapor phase from the liquid phase heat is 

 absorbed. His first experiments, though extremely crude, were sufii- 

 cient to show that in the case of water the amount of heat absorbed by 

 this transformation is considerable. Somewhat later Black and Irvine 

 made further experiments in this direction, and found the value 520 

 calories per gram, which is surprisingly close to the now generally 

 accepted value, considering the crude method with which they worked. 

 In 1781 Watt ^ for a short time attacked the subject at the suggestion 

 of Black, and about fifteen years later returned to it and made a num 

 ber of measurements, the details and results of which he published 

 He pointed out that heat is lost through radiation, but found no 

 method of measuring this loss. He showed also that the condenser 

 gains heat by conduction when connected directly to the boiler by 

 means of a metallic tube, and sought to overcome this source of error 

 through making the connection by means of a cork, so that metallic 

 contact was avoided. The average of eleven separate determinations 

 gave him the value 525.2 calories, or 625.2 from zero,*^ the values vary- 

 ing from 612.9 to 637.1, a difference of about four per cent; but he 

 expressed his opinion that the true value is not far from 633 calories. 



Somewhat later Count Rumford * made three experiments concern- 

 ing the same constant, finding the mean value 667 as the total heat 

 of evaporation from zero, — a result much higher than Watt's. Rum- 

 ford also determined the heat of evaporation of two or three organic 

 liquids, but these results have no value, since, as he admits, the sub- 

 stances were not pure. In this work he sought to overcome the error 

 due to radiation by starting a determination with the calorimeter 

 water at a temperature as much below the surrounding temperature 

 as it would be above it at the completion of the determination, on the 

 assumption that the gain in heat during the first half of the total time 



^ See Regnault's Experiences, Part I, p. 635 (1847). 



2 Rogmuilt, M6n.de rinst.de France, 21, 035(1847); Robinson's Mechan- 

 ical Philosophy, 2, 5 (1822). 



^ This latter number includes the heat necessary to raise the gram of 

 water from 0°C to 100°C. 



* See Biot's Traite de Physique, Vol. 4, p. 710. 



