THE CAUSATION OF DISEASE. 3 59 



It may be asked how the " malignant " bacteria irritate 

 the tissues. I can conceive of only two ways — (i) by mere 

 mechanical irritation ; (2) through some poison which they 

 emit. The latter is, doubtless, the source of all the trouble, 

 just as it is in the case of the gall-gnat. In the latter case, 

 however, the growth never exceeds certain well-defined limits, 

 because, of course, the supply of poison is limited. 



The granulomata are admitted to be caused by parasites, 

 and why not the malignant growths also ? All these forma- 

 tions are bound together by the common bond of infectiousness, 

 and this suggests, nay, almost demands, some common cause. 

 Further, there is a marked similarity in the structure of 

 granulomata and sarcomata. The simplest way of explain- 

 ing infectiousness is to assume the existence of a parasite, for 

 this gives us an abiding mal-E, and hence a continuance of the 

 morbid process. The bacteria causing the several varieties of 

 granuloma and those causing the sarcomatous growths, are 

 equal in virulence, as is proved by the extent of dissolution 

 being equal in both. It is important to distinguish between 

 virulency and capacity to live on in the tissues. This latter 

 power is possessed in a far greater degree by the " malignant " 

 than by the granuloma parasites, for, in the first place, the 

 granulomata have far less tendency to spread at the periphery, 

 and, secondly, they are more apt to abort, as not unfrequently 

 happens with tubercle, for instance. The granulomata are 

 generally thought to be inflammatory, and some have even 

 maintained that all inflammations are due to bacteria. This 

 view is probably erroneous ; but, be that as it may, it is quite 

 certain that in those cases where inflammation is permanently 

 arrested the mal-E must cease with the cessation of the in- 

 flammation ; wherefore, if this mal-E consists of a bacterium, 

 this must lose its activity, as not unfrequently happens with 

 the granulomata. The malignant growths, on the other hand, 

 very seldom abort. 



How are we to account for the fact, for fact it unquestionably 

 is, that malignant disease so often runs in families ? Doubtless, 



B B 2 



