18 



was entirely decomposed in twenty-four hours and the other not 

 until after three days. In a similar experiment manganese dioxide, 

 which had been previously heated to redness, Avas used in place of 

 the platinum and it also showed marked catalyzing power. 



These results speak strongly in favor of the truth of the assump- 

 tions which we have made. It is true that catalysis of silver oxide 

 and hydrogen peroxide 1iy platinum may be specifically different, 

 but let us consider a few of the reactions, where oxygen is a factor 

 or a product, which are catalyzed by platinum; the combustion of. 

 gases such as hydrogen, methane, ammonia, vapor of methyl 

 alcohol; the decomposition of oxidizing agents such as hydrogen 

 peroxide; the electrochemical solution of oxygen at the cathode 

 and its separation at the anode. Certainly there seems to he 

 ground for believing that in all these cases, while several reactions 

 occur together, an important, and in many cases the only, factor in 

 determining the speed of the whole reaction, is the speed in one 

 direction or the other of the intermediate reaction, 20 % 0^, and 

 that this reaction is catalyzed by platinum and by a number of other 

 substances, of course as well in one direction as in the other. 



At least we may conclude in regard to the decomposition of 

 silver oxide that since there are substances besides silver which 

 act as catalyzers the autocatalysis is not of the same nature as 

 that of precipitation from a supersaturated solution, and further, 

 that there is some ground for believing that the velocity which we 

 measured is either that of the reaction 20 = 0.^ or that of the diffu- 

 sion of the factors or products of this reaction, to or from the 

 catalyzing substance. 



We should expect the catalytic power of the silver to be propor- 

 tional to the extent of its surface, while we have found it to be 

 proportional to the amount of silver. But probably these two 

 quantities are very nearly proportional since the silver is produced 

 as a slightly cohesive mass which readily falls into a fine white 

 powder. If the grains are of fairly uniform size the surface 

 exposed is proportional to their number, or, in other words, to 

 their mass. 



SUMMARY. 



A satisfactory thermostat for high temperatures is described, 

 together with a method of measuring the velocity of decomposition 

 of silver oxide. 



