ON THK SPKCIFIC 1IKAT OF WATER. 225 



t.iined by a separate experiment (No. 171). This experiment was conducted in the 

 same way as the other continuous-flow experiments, except that there was no 

 electrical heating, no heating of the obturator, and no stirring. Under these circum- 

 stances water flowing in at 13'00 C. flowed out at 0'02 C. to 0'025 C. higher. In 

 the actual experiments, owing to the heating of this tul>e by the contents of the 

 calorimeter, the viscosity of the entering water and its consequent rise of temperature 

 would be diminished. We have taken the lower figure 0'02 C. as representing this 

 frictional heating, and this is deducted from A0 in order to get the rise of temperature 

 due to electrical heating only. 



The data furnished by these three series of experiments are summarised in 

 Appendix C. 



The first series embraces eight experiments, the second six, and the third five. 

 The headings of the columns are in most cases self-explanatory. The temperatures 

 are given corrected. For the first two series the correction for heating by obturator, 

 stirrer, &c., is taken as 0'43 watt for all the experiments of both series and the total 

 heating is given in the column headed " C'lt + O^S." Also M 3 in the first two series 

 includes the resistance of the primary leads of the watt balance as well as that of the 

 mercury leads of M a . The leads are deducted to obtain the figure entered in the 

 column headed " Resistance of heater." Ci (see fig. 4) is the total current through 

 MI and C = CiR /(Ro + M 2 ) is the current through the heater. For the third series 

 the correction for heating by obturator, stirring, &c., is given for each experiment in 

 the column " Obturator, &c." 



Looking at the columns headed " Difference from mean," it will be seen that there 

 is an ascending degree of accuracy as between individual experiments in the three 

 series. This is due in the second series to the fact that the runs were longer than in 

 the first, and in the third to the additional fact that the correction for heating by 

 obturator, stirring, &c., had been reduced to a much smaller amount. When 

 corrected to correspond with the exact period from 13 C. to 55 C. the result of the 

 first two series is to give the mean value of J for tap-water from 13 C. to 55 C. as 

 4'177 and that of the third series 4' 179. Owing to the largeness of the correction for 

 IM -a ting by obturator, &c., in the first two series we do not bring the results into the 

 average, but take the exact corrected figure yielded by the third series as the basis 

 of our calculations, that is to say, for " tap-water " the mean value of J for the 

 interval from 13 C. to 54'25 C. is 4'1785, or corrected to the interval from 13 C. to 



55 C. we have 



J 13 M = 4-1788. 



This figure is also entitled to special weight owing to the fact that the thermometer 

 and the standard resistance and battery of nine standard cells had all been sent 

 to the National Physical Laboratory to be re-calibrated, and had been returned 

 a t'r\v days before this third series of continuous-flow experiments was made, 

 and that the bridge and mercury thermometer-resistances were also re-calibrated from 



VOL. CCXI. A. 2 G 



