8 



left open, thus allowing the gas to collect in the burette until the 

 velocity was again measured. In this way the velocity at any 

 moment, and the total volume of gas that had been evolved, could 

 be measured independently. However, after the first few experi- 

 ments, the burette was done away with and only the velocity 

 was measured, as it was found that in the feAV cases where a 

 knowledge of the total volume was desired it could ])e ascertained 

 with sufficient accuracy from the time-v^elocity curve. 



This simple apparatus gave complete satisfaction, cleanliness in 

 the recording tube being the only precaution necessary for its 

 success. 



THE COURSE OF THE REACTION. 



In the first experiment with the apparatus 4.4 grams of silver 

 oxide was used, the temperature was 327.5° C, and the measure- 

 ments of the reaction velocity were made for twenty-four hours at 

 intervals of one-half to one hour. For several hours the decompo- 

 sition was inappreciable; at the end of 10 hours the rate was 

 1 cubic centimeter of ox3'gen evolved per hour; after 13 hours, 

 1 cubic centimeter in 27 minutes; after 14 hours, 1 cubic centi- 

 meter in 18 minutes; after 16 hours, in 4 minutes; after 18 hours, 

 in IJ minutes, and at this point the velocity was increasing rajndly. 

 After 20 hours the rate was 1 cubic centimeter in 1 minute and 

 rapidly falling, showing that between the last observation and this 

 a point of maximum velocity had been reached. After 20^ hours 

 the rate of evolution was 1 cubic centimeter in 2 minutes; after 

 21 hours, 1 cubic centimeter in 3^ minutes; after 21|^ hours, 

 in 7 minutes, and before 24 hours had passed all action had ceased. 



This experiment gave a general idea of the course of the reaction. 

 At the beginning the velocity is very small, increasing more and 

 more rapidly to a maximum, then falling — rapidly at first, later 

 more slowly — until finally the reaction is complete. This decom- 

 position has the general character of an explosion, and in fact it is 

 nothing more or less than an explosion Avhich is slow enough so 

 that its velocity can be measured, and differing from the ordinary 

 explosion only in that in the latter case a reaction progresses with 

 accelerating velocity because of increasing temperature, due to the 

 heat given off by the reaction, while the decomposition we are 

 studying takes place at constant temperature and is in fact an 

 endothermic reaction. 



