17 



reaction of which we are nicasuriiii'- the velocity? — tliat is. Which 

 is the slower reaction ? 



We have many reasons to helieve that the reaction 0., = 20 is an 

 extraordinarily slow one. For example may be cited the inac- 

 tivity of oxygen and the preponderating tendency in slow oxidi- 

 zation at ordinary temperatures for the oxygen to enter into the 

 resulting compounds as the radical (O^), forming the unstable 

 peroxides. We may assume that to an extremely small extent 

 oxygen gas is dissociated, that in ordinary 0^ there is a small 

 concentration of in equili])riuni with it. If when the equilib- 

 rium is destroyed l)y removing some monatomic oxygen it is 

 restored with great slowness, then it will be true also that if the 

 equilil)rium is destroyed in the opposite direction the recovery, 

 according to the reaction 20 = 0,, will likewise be slow. 



It is not unlikely that this latter reaction is the one whose 

 velocity we have been studying and that it is the one which 

 is catalyzed by the presence of the silver. In fact, silver is known 

 to catalyze many gaseous reactions. 



The above line of reasoning would a])ply cciually to other 

 reactions in which oxygen is evolved. Let us consider two cases 

 as different from this one as possible, namely, the decomposition 

 of hydrogen peroxide and of potassium chlorate. Both of these 

 reactions may be accelerated by catalyzers; the catalysis of the 

 latter by manganese dioxide and of the former l)y manganese 

 dioxide, platinum, and .other substances is familiar to all 

 chemists. If in these cases also we assume that the union of two 

 atoms of oxygen is the reaction the slowness of which retards the 

 decomposition, then this reaction is the one which is catalyzed Ijy 

 ])latinum and by manganese dioxide. Therefore, these sul)stances, 

 if our assumptions are correct, should aid the decomposition of 

 silver oxide as silver does. We have thus the means of testing our 

 suppositions by direct experiment. 



Two tubes, each containing 5 grams of silver oxide, to one of 

 which 0.25 gram of pure platinum black was added, were put 

 into the thermostat simultaneously. The one containing ])latimim 

 l>egan to give off oxygen at once. At the end of two hours it was 

 evolving 1 cubic centimeter in twelve seconds, while the other 

 was evolving 1 cubic centimeter in ten minutes. 



A similar experiment was made with another i)reparation of 

 silver oxide, in which case the silver oxide containing platinum 

 30644 2 



