THE SO-CALLED INORGANIC ENZYMES 123 



mixture remains quite transparent, which is due to the colloid 

 gelatine preserving the amorphous precipitate in colloidal solution. 

 No such effect is obtained with crystalline precipitates. 



The catalytic reactions of metal-sols which have been studied 

 are the conversion of hydrogen-peroxide into water and oxygen, 

 and of hydrogen and oxygen to water. These have been most 

 studied in the case of the platin-sol by Bredig and his pupils. 



The action of platin-sol upon peroxide was found to follow 

 approximately the logarithmic law for a mono -molecular reaction, 

 but in the case of hydrogen and oxygen Ernst found that the 

 velocity was proportional to the absolute amount of platinum 

 present, and in the peroxide experiments it was found that the 

 activity did not vary in simple direct proportionality to the con- 

 centration in platinum, but was expressed by the empirical law, 



k /C \& 



y 1 = ( ~ } , in which k : and Jc 2 are the velocity constants corre- 



fc 2 \C 2 / 



spending to concentrations in platinum of GI and C 2 , and 6, instead 

 of being unity, is a constant, the value of which lies between 1-6 

 and 1-3. Hence we see that there are abnormalities in the action 

 of these inorganic catalysts, similar to those already described 

 in the case of the enzymes. The abnormalities are explained upon 

 the basis of intermediate compounds between catalyst and sub- 

 stratum, or products of reaction. 



Even equal platinum concentrations do not always lead to the 

 same velocity of reaction, this depending upon variations, in fine- 

 ness of subdivision of the platinum, and upon the previous history 

 of the solution, such as its age, the temperature at which it has 

 been preserved, &c. ; to this variation van Bemmellen has in ana- 

 logy with similar phenomena in magnetisation applied the term 

 " hysteresis." 



The dilutions in which these colloidal platinum solutions exert 

 their catalytic effects are enormous. Thus a solution containing 

 only 0-000,01 grm., that i?, T ^j milligram, per c.c., acting upon a 

 solution of hydrogen peroxide containing 0-06 grm. per c.c., con- 

 verted more than the half in forty minutes. The action of TnlJZFZT 

 to 73o^ooo~ milligram of colloidal platinum upon more than a 

 million times its weight of hydrogen peroxide could still be 

 detected. 



Ernst also found that y\y milhgram of colloidal platinum 



