AUTOCATALYSIS AND INFECTION 221 



Examples of such autocatalytic reactions are seen in the ten- 

 dency of many explosive substances to spontaneous explosion. Thus 

 if gun-cotton be not most carefully washed from oxides of nitrogen, 

 these products, present in too minute quantity at first to cause 

 any change, may set up a slow and at first inappreciable reaction, 

 which, slumbering at first, gradually increases in velocity and 

 finally fires off the gun-cotton. 



A good example of autocatalysis is quoted by Ostwald in 

 the case of the action of nitric acid in dissolving metals. As has 

 been noticed by several observers, pure nitric acid free from nitrous 

 acid at first scarcely attacks many metals, such as copper, mercury, 

 and zinc, although the impure acid readily oxidises them with pro- 

 duction of nitrous acid. If a trace of a nitrite be added, however, 

 the reaction at once commences and momentarily becomes more 

 energetic, as it is continually autocatalysed with increasing energy 

 by the nitrous acid formed in the reaction. 



The writer has pointed out that such a process of autocatalysis, 

 induced by a trace of enzyme, may be the means of infection, 

 and of the reproduction of the virus in many of those infectious 

 diseases, such as the acute exanthemata, in which notwithstanding 

 much bacteriological research no causal connection of any living 

 organism has yet been demonstrated. 



The incubation period would be that required for the pro- 

 duction of the autocatalyst in sufficient quantity to cause a 

 general reaction with the tissue cells. The autocatalyst, which 

 would act as the toxin of the disease, would still go on increasing 

 in quantity until it had attained the concentration for maximum 

 effect; but at the same time the tissue cells would react to it as 

 to an ordinary toxin and produce the antitoxin, by which the 

 toxin would be neutralised and rendered inert, and so the course 

 of the disease would be limited. 



The length of time and the exactness of duration of the incuba- 

 tion period form no objection to such a view, for the time of the 

 initial period of autocatalytic reactions is often prolonged, and 

 the duration of the incubation period would be determined by 

 the reaction of the tissue cells affected, and would but little depend 

 upon the amount of the trace of autocatalyst which originally 

 carried the infection, unless this were very large. 



Further, it may be pointed out that in many cases in which 

 the toxin arises from the products of bacterial growth, the length 



