384 PROF. H. L. CALLENDAR, PREFACE TO 



After circulating round the thermometer pocket, the throttled steam, before leaving 

 the apparatus, was made to circulate twice round the calorimeter to minimise external 

 loss. The whole apparatus, including the steam-pipe and the throttle tube, were 

 also well lagged in the usual way. The pressure after passing the throttle was 

 generally atmospheric, but an exit throttle and gauge were provided for raising the 

 pressure and observing its value when required. The exit could also be connected, 

 if desired, to a condenser vacuum. 



The wetness of the steam (l q) was determined in the first instance by means of 

 the usual formula with REGNAULT'S coefficients, namely, 



(l-g)L' = 0-305 (*'-100)-0'475 ("-100), (l) 



where L' is the latent heat of dry saturated steam at the initial temperature, t' and 

 t" is the temperature observed after throttling to atmospheric pressure. In spite of 

 all precautions suggested by previous experience in calorimetry, the wetness of the 

 steam persisted in coming out negative (which was obviously impossible) to an extent 

 beyond the probable limits of error of observation. Since many at that time suspected 

 REG:STAULT'S value 0'475 of the specific heat, and had proposed other values, ranging 

 from 0''380 (GRAY) to 0'5G8 (ZEITNEK), it was decided to make a direct determination 

 of the specific heat in the neighbourhood of 100 (J. by the continuous electric method, 

 which had been devised some years previously, for the determination of the mechanical 

 equivalent of heat. 



The first experiments on the specific heat by this method were made by passing a 

 current of slightly superheated steam at atmospheric pressure over an electric heating 

 coil, and observing the rise of temperature with a pile of several thermojunctions in 

 series. The electric energy supplied was measured with standard Weston instruments, 

 and the steam current by the usual method of condensation and weighing. This was 

 a comparatively rough method, because the steam current could not be very accurately 

 controlled, or the external loss exactly determined. Some difficulties were also 

 experienced with the insulation of the thermojunctions. The results obtained showed 

 discrepancies of 1 or 2 per cent., but indicated a value of the specific heat falling from 

 0'50 to 0'49 over the range of temperature from 110C. to 160 C. covered by the 

 experiments. This was regarded as a decisive confirmation of REGNAULT'S value 

 0'475 over the range 125 C. to 225 C., since a slight diminution of the specific heat 

 with rise of temperature was to be expected owing to the existence of the Joule- 

 Thomson cooling effect. 



Having thus obtained by direct experiment the values of the mean specific heat at 

 atmospheric pressure over the required range of temperature, it was possible to deduce 

 the values of the total heat of saturated steam at temperatures above 100 C., from 

 the observations already taken with the throttling calorimeter, by means of the 

 equation, 



(2) 



