16 



ROYAL SOCIETY OF CANADA 



to include the furnace and an ammeter of the main circuit as indicated 

 in Fig. 1. The direction of the current from these cells was opposed to 

 that of the main circuit and by adjusting the voltage of the cells so that 

 it was nearly or quite equal to the fall in voltage of that circuit between 

 the points a and h, Fig. 1, i. e., between the terminals of the furnace, any 

 considerable increase in the voltage of the main circuit traversed the cells 

 as a charging current, while any considerable decrease was corrected by 



Hawp RKeostcit 



the discharge of the cells. By this means the fall of voltage between the 

 terminals of the furnace could be kept at any desired value, and constant 

 temperatures secured. 



The temperatures were measured by means of a LeChatelier pyro- 

 meter, the rhodium platinum juncture of which was bound to the out- 

 side of the bulb containing the chlorate, by means of a thread of asbestos. 

 The accuracy of the readings of the pyrometer were tested by means of 

 steam, the vapor of sulphur, and the melting point of potassium chlorate, 

 and were found to be satisfactory. 



In carrying out the experiments, the bulb with thermocouple 

 attached was quickly inserted into the furnace which had been previously 

 heated so that the temperature at which the decomposition was to be 



