METHODS 99 



living being under observation relatively to the substance 

 on which the diastases act. Now they are mixed in the 

 culture bouillon with excrementitial substances (substances 

 R) ; and so, when we study the effect of a filtered culture x 

 of a microbe or yeast on an animal or any reagent what- 

 ever, we cannot distinguish what is due to the diastase and 

 what to the excrementitial substance. This is a hindrance 

 in the study of the diastases. 



As we have said, the diastases are dead colloids. Probably 

 their activity is due to their peculiar colloid state ; and this 

 in fact, manifests itself for the most part by variations in- 

 troduced into the colloid state of other substances. We 

 know that, secondarily to these colloid variations, chemical 

 reactions resulting from them must often be produced ; 

 but that does not prevent our believing the action of the 

 diastase to be of colloid order, even though we can verify only 

 the secondary result, which is chemical. 



Sucrase is an example ; it is a diastase produced by 

 yeasts and other moulds ; and it has the property of in- 

 verting cane sugar. This simply proves that the solution 

 of cane sugar, as a chemically defined compound or saccha- 

 rose, is stable only in a colloid state differing from that 

 of sucrase. There is a relation of equilibrium between 

 the chemical and physical states of this particular sub- 

 stance. 



This idea of the purely colloid action of diastases suggests 

 a general interpretation of their origin. A body living in a 

 must, bouillon, or any other colloid, multiplies by assimila- 

 ting the medium, that is, by fabricating at the medium's 



1 Filtered or centrifugated ; for certain diastases do not pass 

 through the niters, and so it is preferable to separate the bodies 

 of the microbes by means of an apparatus acting centrifugally 

 like that used in removing the cream from milk. 



