224 MESSRS. W. R. BOUSFIELD AND W. ERIC BOUSFIELD 



The flasks for noeiving the outflow water were introduced one after another at 

 intervals of half an hour. The weighings are given in the last three columns of the 

 table, and are set opposite the middle of each period of 30 minutes. For the purpose 

 of the calculation the last four results are taken, the mean of which is a i 

 86 '04 gr. of water per minute. 



The current from the dynamo was maintained throughout so as to balance 

 E.M F. of the nine cadmium cells on the resistance M, as indicated in fig. 4. 



Readings were recorded every quarter of an hour, the time being shown in the first 

 column of the table. Thermometer readings are recorded uncorrectkl, except that 

 the hath tlu-rmometer is corrected so as to read with the outflow thermometer, in 

 ,,nler to show the difference of temperature between the bath and the interior of the 

 Dewar vessel. It will be observed that the slight variations of the temperature of 

 the supply cistern (which are shown in the second column) do not affect the 

 temperature of the inflow water (shown in the third column). To secure this result 

 a bottle containing about 8 litres was sunk in the cistern and the inflow water was 

 taken from near the bottom of this bottle, the temperature of which was practically 



constant. 



At 3 o'clock the conditions had been steady for over an hour and the main run 

 then ceased, and the subordinate observations were made to check the resistances and 

 determine the heating effect of the obturator, &c. The flow of electricity and of 

 water through the interior of the calorimeter was stopped, everything else being 

 continued as before, including the flow of water through the pipe b (see figs. 1 and 3) 

 into the water jacket of the inflow and out by the pipe e. A very slow rise of 

 temperature took place under these circumstances, the record of which for an hour is 

 shown in the notes (Appendix B). The total rise of temperature during the hour 

 was 0'04 C. The water contents of the calorimeter together with its water 

 equivalent amounted to 3170 gr., giving a total heating effect of 015 watt, or 

 9 joules per minute. 



The measurements of the resistances MI and M 2 are also recorded in the notes 

 (Appendix B). The formulae for those resistances are those given before in Section 5 

 except for the constant terms. These are obtained as shown in bridge ohms and 

 converted into international ohms by the appropriate factors. For the third series of 

 experiments the resistance of the mercury leads of the heater was re-measured and 

 found then to be 0'0055. 



With the exception of the complication introduced by the watt balance the notes 

 of Experiment 66 recorded in Appendix A are of the same character and seem to 

 require no further explanation. 



In order that the inflow water might be brought to a temperature near to that of 

 the contents of the calorimeter before mixing therewith, it was carried through a 

 zig-zag tube of a length of about 4 feet immersed in the calorimeter. The resistance 

 thus offered to the passage of the water produced a heating effect which was ascer- 



