240 Dr. W. Marcet. 



Although there was in one case as much as 8*04 per cent, differ- 

 ence between the calories found and the calories lost from the jar, 

 still the mean of the eleven experiments differs only by 1*41 per 

 cent., which is a near result considering the difficulty of the experi- 

 ment. 



The determination of the heat produced by the combustion of 

 hydrogen was certainly a more satisfactory method than the former 

 for testing the calorimeter ; hydrogen was prepared for the purpose 

 in the usual way by the action of sulphuric acid on zinc, the gas 

 being purified through solutions of potassium hydrate and cupric 

 sulphate, and collected over water in a bell-jar carefully graduated. 

 The receiver was supplied with a gauge, showing the pressure to 

 which the gas was subjected, and a thermometer ; from 20 to 29 litres 

 of gas were used in each experiment. 



After making the required preliminary essays with the agitators, 

 the experiment was proceeded with as follows : 



First of all it was necessary to find out and to adjust carefully the 

 speed of the gas delivery, and with that object a weight was placed 

 on the bell-jar, while the rate of issue of the gas was regulated at 

 will by means of a screw clamp on the track of the gas tube.* In 

 that way the speed of the gas delivered was adjusted so as to produce 

 on burning about the same heat as a person would emit in the calori- 

 meter in a given time. 



The delivery tube led from the bell- jar into the calorimeter 

 through a fixed metal tube carried across the walls of the two 

 chambers of the calorimeter, its end being connected with a suitable 

 burner ; when lighted, the gas burnt with but a very small flame. 



Before commencing the experiment the tube was rinsed out with 

 hydrogen and the thermometers were read, together with the pointer 

 on the scale of the bell-jar. Then the gas was turned on and 

 lighted, the vessel containing the ice hung in position, the stop- 

 watch started, the calorimeter closed, and the agitators put in 

 motion. Of course every care was taken to keep the temperature of 

 the calorimeter constant, which was done without any difficulty, the 

 temperature of the copper varying seldom by more than 0'1 or 0*2 C. 



When forty minutes or an hour had elapsed (mostly forty minutes), 

 the temperatures were read, the gas turned off, and the agitators 

 stopped. Next the calorimeter was rapidly opened, and the flow of 

 water from the ice to the flask arrested ; the temperature of the ice 

 water was then read off, these various operations being carried out 

 as rapidly as possible. It was necessary to determine the heat 

 absorbed by the burner, which was done by plunging the burner, 

 immediately after turning off the gas, into 200 c.c. of water at a 



* The gas was carried as much as possible through glass tubing, in order to 

 avoid the loss by diffusion through india-rubber. 



