the Elements Carbon, Boron, and Silicon. 169 



heat lost by the substance must be inappreciable, considering 

 that it is surrounded with such bad heat-conductors (glass, air, 

 cotton-wool, and glass). The temperature of the substance at 

 the moment of its deposition in the calorimeter was then taken as 

 equal to the reading of the thermometer in the bath during the 

 five or ten minutes immediately preceding the removal of the 

 heating-vessel. If during this time the temperature varied so 

 much as a tenth of a degree, the mean temperature during these 

 five or ten minutes was taken as the final temperature of the 

 substance as it fell into the calorimeter. 



All temperature-readings were reduced to the readings of the 

 air-thermometer. The thermic expansion of the glass vessel of 

 this thermometer, within the range 0° to 200°, was found to be 



V T = V (1 + 0-0000235 T + 00000000095 T 2 ). 

 The constant temperatures of the solid carbon dioxide and of the 

 snow and salt freezing-mixture were determined directly by the 

 air-thermometer ; the mean of a number of readings gave for 

 these temperatures the values — 79°*5 C. and — 21°'3 C. 



For temperatures between 0° and 100° two standard thermo- 

 meters graduated to tenths of a degree were employed. The 

 correction to be made in reducing the readings of these thermo- 

 meters to air-thermometer readings was found to be 



For the temperature T, o^on 



For temperatures between 100° and 300° a thermometer gra- 

 duated to whole degrees, but easily capable of reading to a tenth, 

 was used. To reduce the readings of this thermometer to those 



of the air-thermometer the correction o*nn was ap- 

 plied. Every necessary correction in reading the thermometer 

 was allowed for. In all experiments which did not involve the 

 raising of the substances to a red heat, Bunsen's ice-calorimeter* 

 was used for the determination of the amounts of heat W lf 

 W 2 , . . . W n . Had it not been for the calorimeter of Bunsen, 

 the following work could not have been accomplished. By means 

 of this instrument I was able to measure with great accuracy 

 very small amounts of heat even when using less than 1 grm. 

 of the substance. For a description of the instrument reference 

 must be made to Bun sen's original paper (loc. cit.). For cal- 

 culating the amount of heat from the readings of the calorimeter 

 I adopted the following plan. 



In order to graduate the scale-tube employed, I dill not use the 

 method of employing a certain quantity of water at a known tem- 

 * Pogg. Ann. vol. cxli. p. 1. [Phil. Mag. S. 4. vol. xli. p. 161.] 



