MESSRS. W. R. BOUSFIELD AND W. ERIC BOTlSFIELD 

 10. Donation of Mean Capacity of Cal^eter^e continuous _ flow 

 experimea* gave us the mean value of J for tap water between 13 C and 54 5 C. 

 J> - 4-1786 With the aid of this figure the next step was to determine the 

 mean capacity' of the calorimeter between the same limits. For this purpose 

 *lorimete7was filled with about 3 litres of tap water at a temperature of about 

 13 C which was heated for a period of 39 minutes with a current of about 

 5 ' amperes, thereby raising its temperature to about 54"5 C. The method of the 

 experiment is that described in the last preceding section, and the notes of 

 Experiment 125 in Appendix D will serve to show the course of these experiments, 

 the only difference being that there was one unbroken run of 39 minutes instead of 

 the shorter periods. The result of eight experiments is set out in Table V. The 

 corrected A0 due to electrical heating is given in the second column, the weight of 

 water employed in the experiment in the third column, and the number of joules 

 employed to produce the rise of temperature A0 in both the water and the calori- 

 meter in the fourth column. 



TABLE V. Capacity of Calorimeter. 



In the fifth column is given the number of joules required to heat the water alone 

 through A0, which is obtained by multiplying WA0 by J', where J' = 4 '17 8 6, the 

 value which we obtained as the mean value of J for tap-water over this range of 

 temperature. The difference set out in the next column is therefore the joules 

 required to heat the calorimeter alone through the interval AO. This difference 

 divided by A# is the mean capacity of the calorimeter per degree expressed in joules, 

 which works out at 595'5 joules. The mean error of the single observations works 

 out at 3'1 joules, or about per cent., but it must be remembered that it is the 

 difference of two large quantities. The mean probable error of the result works out 

 at 1 joule. With 2950 gr. water (which is the amount with which we work) an error 

 of 1 joule in the capacity of the calorimeter involves an error of less than 1 part in 

 10,000 in the value of J. 



