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



THERMOMETER K. 



Experiment 1. Fundamental Intervals .... 5*1454 ohms. 



2. .... 5-1445 



,,3. .... 5*1446 



Mean ,, .... 5'1450 



Eesistance at C 13'3039 



The relationship between the temperatures t and t pt on the absolute scale and on the 

 platinum scale, respectively, can be expressed by the formula t t pt = d . t (t 100) . 10~ 4 , 

 where d is a constant which may be taken equal to 1'5 for pure platinum. This 

 expression, combined with the above observations, enables the values of a. and ft in 

 the expression R, = K ( 1 + at + fit 2 ) to be calculated. The values for the thermometer K 

 are a = '0039253, /3 = -0 '000, 000580 ; and for the thermometer H, a = 0'0039394, 

 = -0-000,000582. 



The difference in the values of the coefficients is probably due to a difference in 

 the treatment to which the thermometer wires were subjected in winding. It is 

 satisfactory to observe that the higher coefficient is found in the thermometer H, in 

 which we should expect the effect of thermal conduction through the leads to be the 

 greater ; this circumstance points to a true difference in the values of the coefficients. 



Two pieces of solid caustic soda were placed in the bottoms of the tubes which held 

 the thermometers during the determinations of the lower fixed points. The insulation 

 of the thermometers was tested from time to time, and found to be almost perfect. 



(7) The Cadmium Cells. The cadmium cells were of the well-known H pattern. 

 The amalgam contained 12 - 5 per cent, of cadmium, and was prepared electrolytically 

 according to the specification given by F. E. SMITH (' B.A. Report,' South Africa, 

 1905). The mercurous sulphate and cadmium sulphate used were supplied by 

 KAHLBAUM, the cadmium sulphate being recrystallised and tested in the manner 

 described by SMITH. Four cells were made up, but only two of these cells, (2) and (3), 

 were used in the work which follows. A comparison of these cells with a cell (7) made 

 up two years previously showed that (2) and (3) were lower in E.M.F. than (7) by 

 0'00014 volt and 0'00012 volt respectively. At a later date these cells were sent to 

 the National Physical Laboratory, where they were tested over a period of seven 

 months. It was found that while cell (7) had remained practically constant, a slight 

 fall was observed in the case of cells (2) and (3) ; this was immaterial, however, since 

 the cells were compared with cell (7) during the actual research. The full data 

 referring to the tests of the cells are preserved in the archives, the mean value for the 

 cells (2) and (3) at the time when they were actually used in the calorimetric 

 experiments being 1 '01 83 5 volts at 17 C. in terms of the international ohm and the 

 ampere (10" 1 C.G.S. unit). This value corrected to the temperature of the experi- 

 ments has been taken for the calculation of the specific heats. That value, 4 '180 



