Latent Heat of Evaporation of Benzene. 7 



of tank- water was forced by means of a pump driven by a 

 water-motor : thus the stem-temperature was always that (or 

 nearly that) of the tank, and the readings were not affected 

 by changes of temperature in the room. The actual value of 

 6 was of small consequence, an accuracy of y 1 ^ C. being 

 sufficient ; but it was necessary to read changes of temperature 

 during an experiment to nearly o, 001 C, and this we believe 

 we were able to do. A full description of the various pre- 

 cautions taken will be found in Paper A, pp. 55 and 56, and 

 Paper TV, pp. 275, 276. 



The observations were taken by means of a reading-tele- 

 scope fitted with a micrometer-eyepiece, which directly divided 

 1 millim. on the thermometer-stem into 10 parts : thus 0*1 of 

 a millimetre could be directly read, and *01 could be estimated. 

 There is no doubt that the actual readings could be taken to 

 •025 millim., that is about 0°'001 C. 



Thermometer A had a range of 16° to 26° C, and about 

 27 millim. were equivalent to 1° C. 



Thermometer II. had a range of 28° to 53° C, and about 

 20 millim. were equivalent to 1° 0. 



All temperatures, both when obtained from these and from 

 the platinum-thermometers, are throughout this Paper ex- 

 pressed in terms of the nitrogen-scale. 



5. The Calorimeter and its Connexions. 



A full description will be found on pp. 276-281 of Paper W, 

 and drawings and sections on plates 5, 6 of the same paper. 

 The brief account given in the description of the method 

 supra is sufficient to indicate the nature of the arrangements 

 to those who are not familiar with that paper. 



The differences of temperature between the calorimeter and 

 surrounding walls were determined by means of differential 

 platinum-thermometers. A description of the method of 

 standardization and observation of those thermometers is 

 given on pp. 52-56 of Paper A, and some further details on 

 pp. 285-290 of Paper W. Experimental evidence is there 

 adduced in support of the following statement : — " It follows 

 that differences of temperature could be determined to '0004°, 

 and differences of '0001° could be detected." 



The following Table shows the difference in temperature 

 corresponding to a difference of 1 mean millim. of the bridge- 

 wire whose opposite ends were connected with the thermo- 

 meters :— 



