ON THE SPECIFIC HEAT OF WATER. _.;; 



bridged by tliirty secondary cells in series, which were not themselves capable of 

 giving out a continuous current of more than 4 amperes, but served to take off the 

 small variations of the dynamo current and rendered the regulation of the current 

 more easy. 



(h) For continuous-flow experiments water from the water-mains was used (which 

 we may shortly designate as " tap-water "). The volume of water required in these 

 experiments, which involved a flow of many litres of water per hour for several 

 hours for one experiment, rendered the use of distilled water inconvenient. The 

 water was passed through a large thermostat, and the regulation was such that it 

 was easy to keep the temperature constant within 0'01 0. The relative value of 

 the specific heats of distilled water and tap-water was subsequently determined. 



The fundamental experiment was the determination of the mean value of J by the 

 method of steady flow and steady heating, and its accuracy depended on maintaining 

 all the conditions steady for a sufficient time. The temperatures at the points of 

 inflow and outflow could be read by means of mercury thermometers to 0"005 C. 

 This involved a possible error on the temperature interval of 0'01 C. To obtain the 

 order of accuracy at which we were aiming, it was therefore necessary to have a 

 temperature interval of about 40 C., so that an error of 0'01 C. would not vitiate 

 our results by more than t part in 4000. For convenience the temperature of inflow 

 was fixed at 13 C., as during the portion of the year in which the continuous-flow 

 experiments were made, water could always be obtained from the mains l)elow that 

 temperature. The temperature of outflow was in the neighlxmrhood of 54 '5 C. 

 This temperature range had many conveniences. The liberation of air was negligible 

 up to alwut 55 C., and with a heating effect of about 240 watts the weight of water 

 and the time of flow were such as to be easily ascertainable with an accuracy equal 

 to that required. Moreover, the necessary corrections could be more accurately 

 determined below this temperature than for higher temperatures. 



The result of these experiments was to give the mean value of J for tap-water over 



the interval from 13 C. to 54'5 C. as 



4-179. 



second set of experiments was to determine with the aid of the figure 4'17'J 

 the mean capacity of the calorimeter. For this purpose a weighed quantity of tap- 

 water at the temperature of 13 C. was placed in the calorimeter and heated through 

 the same interval. The result was to give the mean capacity of the calorimeter from 

 13 C. to 54'5 C. as 595'5 joules per degree. 



The third set of experiments was similar to the last, except that distilled water 

 was used. The result of these experiments, in conjunction with the capacity figure 

 previously obtained, was to show that the mean value of J for distilled water over 

 the interval of 13 C. to 54"5 C. was slightly higher, that is to say, 



4-182. 

 2 D 2 



