396 PEOF. H. L. CALLENDAE, PEEFACE TO 



flow, which is clearly shown in the table of observations. The application of this 

 correction has accordingly the effect of reducing the mean value of the specific heat 

 from the value S = 2'074, originally obtained, to 2'060 joules per gr. per 1 C., or from 

 0'496 calories to 0'493, over the range 103 C. to 113 C. 



The value so obtained has still to be corrected for the water present in the steam, 

 which was probably of the order of 2 parts in 10,000 at the initial temperature of 

 103 C. The correction can be obtained from the reduction formula, regarded simply 

 as expressing the results of experiment, without requiring an exact knowledge of the 

 nature of the impurity present. Assuming that the formula applies up to a 

 temperature of 113 C., which was within the range of the experiments, and that all 

 the water would probably be evaporated by the time the steam had reached this 

 temperature, the value of the specific heat at 113 C. given by the formula of 

 reduction, should be approximately that of dry steam at this temperature. If the 

 correction to the mean specific heat over the range 9' to 0" is given by the formula 

 173/0' 0", the correction to the specific heat at any temperature 6 is given by 173/0 2 . 

 The value of the specific heat at 113 C. will be less than the mean over the range 

 103 C. to 113 C. by 173/3 x 13-173/13 2 , which is OU342 joules per 1 C. If the 

 value from 103 0. to 113 C. is 2'060, the value at 113 C. will be 2'026 joules, which 

 agrees very well with the value 2'030 at 105 (J. obtained by my collaborator, when 

 special precautions were taken to secure dry steam free from impurities. Briefly 

 enumerated these precautions include (l) a higher initial temperature ; (2) a much 

 longer steam jacket ; (3) pure water for the boiler ; (4) tubes of pure tin in place of 

 glass or rubber, for conveying the steam to the calorimeter; (5) special attention to 

 the arrangement of the separator and drain tubes for removing water of condensation. 



Indirect, Verification. 



The results obtained for steam with the apparatus above described were indirectly 

 verified in various ways, so that there could be little doubt that they represented cor- 

 rectly the specific heat of the steam actually employed, although there might be a slight 

 uncertainty as to the quality of the steam. The first method of verification applied 

 was to use the same apparatus for finding the specific heat of air. There was in this 

 case no difficulty with regard to wetness or insulation, and it was possible to employ 

 thermometers of bare wire in addition to the tube thermometers. The results found 

 for the specific heat of air at atmospheric pressure lay between 0'240 and 0'241 gr. 

 calories per gr. per 1 C. and were about 2 per cent, larger than those given by 

 KEGNAULT, but agreed very well with the value deduced from JOLY'S experiments at 

 constant volume, and with the known value of the adiabatic index for air. The chief 

 difficulty encountered was the regulation and measurement of the air current. This 

 was overcome by constructing a delicately balanced and compensated gasometer for 

 .maintaining a steady pressure, and passing the air current through a resistance 



