CONTRIBUTIONS FROM THE CHEMICAL LABORATORY 

 OF HARVARD COLLEGE. 



AUTOCATALYTIC DECOMPOSITION OF SILVER OXIDE. 



By Gilbert N. Lewis. 



Presented by T. W. Richards, March 8, 1905. Received March 8, 1905. 



In the course of an investigation to determine the equilibrium pressure 

 at a given temperature in the system 2Ag20 = 4 Ag + O2, a singular phe- 

 nomenon appeared. The experiment consisted simply in heating silver 

 oxide at a constant temperature of 335" Centigrade, the stout glass tube 

 which contained the oxide being connected with a manometer. After 

 heating several hours, the manometer gave no evidence that any decom- 

 position had occurred. This occasioned some surprise, since 250'' C. is 

 frecpiently given as the temperature at which silver oxide begins to decora- 

 pose. As I was about to stop the experiment in order to test the tight- 

 ness of the joints, 1 detected a slight deflection of the manometer. This 

 increasedconsiderably during the next hour, and I noticed that the greater 

 the pressure became, the greater became the rate of its increase, until 

 finally the pressure had reached twenty atmospheres, and was increasing at 

 the astonishing rate of between three and four atmospheres per hour. At 

 this point the cement joining the tube and the manometer began to give 

 way, and the experiment had to be discontinued. It had progressed far 

 enough, however, to stimulate considerable curiosity as to the explanation 

 of so unusual a beha\aour. 



The gradual acceleration in the decomposition which was observed 

 could only be attributed to one of the following causes : 1 . The increas- 

 ing pressure. 2. The increasing concentration of oxygen. 3. The ac- 

 cumulation of metallic silver. 4. Some change in the silver oxide, such 

 as the slow agglomeration of its grains. Of these four, the first two were 

 eliminated at once by a rough experiment carried on at atmospheric 

 pressure, which showed the same phenomenon of accelerating decom- 

 position. It seemed worth while to study more carefully this decom- 

 jiosition under atmospheric pressure, and for this purpose the following 

 apparatus and methods were devised. 



