224 



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



between the values for the arrangement Qq and Pp is due to the fact that the 

 platinum loops (see p. 203) associated with Qq and Pp were not exactly equal in 

 resistance. 



TABLES of Mean Values of (Bt/t) 100. 

 Air Temperature. 



Steam Temperature. 



The correction is thus a small one, only amounting to about 2 per cent, of the total 

 heating effect. It will be observed that the heating effect of the leads was slightly 

 greater for the slow than for the quick flow at air temperature, which is what we 

 should expect if, in those experiments, there was a term proportional to the square 

 of the electric current, which represented heat conduction along the leads. The 

 variation of the above numbers under different conditions is only what might be 

 expected in view of the complicated conditions governing the heat loss in the leads, 

 which would vary with the temperature of the calorimeter and the rate of flow of the 

 gas through it. It is only by a method such as the above, in which the corrections 

 are determined under the conditions which hold in the main experiments, that such 

 small variations as are shown in the above tables could be detected and successfully 

 measured. 



(20) Calculation, oftfie Results. In the main runs the current passed either through 

 the leads Qp or Pq. If the ratio of the rise in temperature due to the leads to the 

 total rise in temperature be denoted by a for the first arrangement, and b for the 

 second, the values of (a + b) for different conditions are equal to the one-hundredth 

 parts of the corresponding quantities in the third line of the above table. Now, in 

 taking into account the heating effect of the leads, the simple equations on pp. 200 and 

 201 must be modified by replacing 80 by 80(1 a), and 86 (1 b) in the case of the 

 arrangements Qp and Pq respectively. In those experiments where the run with 



