230 MR. W. F. G. SWANN ON THE SPECIFIC HEATS OF AIR AND 



(p. 221). Taking the pressure coefficient over this range to be 0'003674, the 

 correction can be shown to be additive and to amount to 1*9 parts in 10,000. 



(3) A correction to represent the deviation of the ice-point and fundamental 

 interval of the mercury thermometer to which the temperatures of the reservoir were 

 referred from their proper values. The reading of the thermometer at C. was 

 0'00 C., and the correction to be applied to the fundamental interval was - 03 C. 

 The correction to be applied at 17 C. is conseqiiently quite negligible. 



(4) A correction for the reduction of the readings of the above mercury ther- 

 mometer to the absolute scale. The correction is subtractive, and amounts to 2'2 parts 

 in 10,000 on the specific heat. 



(5) A correction for the fact that the volume of the reservoir was altered by the 

 external pressure of the water in the bath which surrounded it. The corrections for 

 internal pressure were taken into account when working out the results for the 

 calibration of the tubes, but the above correction was left to be applied to the 

 calculated vahie of the specific heat. It is dealt with fully in the records preserved 

 in the archives, and is shown to be additive and to amount to 2 parts in 10,000. 



Hence the several corrections amount to total additive corrections of 1'5 parts per 

 1000 on the value of the specific heat at 20 C., and 2'9 parts per 1000 on the value 

 at 100 C. The corrected values of the specific heat are 



S = 0-24173 cal. per gr. degree at 20 C., S = 0-24301 cal. per gr. degree at 100 C., 



the results being expressed in terms of the calorie at 20 (J. 



(22) Comparison of the Results with Theory. Prof. CALLENDAR has shown* that 

 the characteristic equation of a gas may be simply and accurately expressed in the 

 form v b = IW/pc, where ?> is a constant, II is the gas constant, and c is a function 

 of the temperature of the form c = c (0 n /6)". The behaviour of air is well represented 

 if we take c (l = 1"48 and n = 1'5. It is convenient to measure p with a unit of 

 pressure of 10 6 dynes (75 cm. Hg), in which case if S represents the value of the 

 specific heat at a pressure p and a temperature 0, and S the value at zero pressure 

 and at the same temperature 



S = S + n(n+l)cpf0, 



where S and S are both measured in terms of a unit of heat of 10 8 ergs or 1/41 '80 

 calorie, the calorie referred to being the calorie at 20 C., which is taken equal to 

 4-180 xlO 7 ergs. Substituting the value of c at 50 C. we find S-S = G'00032 

 calorie. Assuming a linear variation of the specific heat with temperature over the 

 small range of variation of 0'5 per cent, between 20 C. and 100 C., my own 

 observations give '24221 cal. per gr. degree for the value of S at 50 C. Hence 

 S = '241 89 at 50 C. It can be shown that if s is the specific heat at constant 



* 'Phil. Mag.,' S. 6, vol. 5, p. 91 (1903). 



