514 Messrs. W. E. Bullock and W. Cramer. On a New 



some of the bacteria are taken up by phagocytes and slowly undergo 

 disintegration within the phagocytes, but that these two processes, by 

 means of which the animal defends itself against the infecting bacteria, do 

 not take place when the bacteria are injected together with toxin. It is 

 clear, therefore, that the toxin paralyses the defensive mechanism of the 

 animal. But it also follows that this defensive mechanism is normally so 

 efficient that the formation of toxin in a concentration sufficient to produce 

 this paralysing action is prevented. 



We were thus led up to the problem : Why does gas gangrene ever occur 

 in man ? Since the defences of the body against the bacteria of gas 

 gangrene are so efficient, the infection of wounds with these bacteria is not 

 in itself an adequate explanation of the occurrence of gas gangrene. And 

 clinical experience leads up to the same problem. Very many wounds, 

 especially those received on the Western Front, were infected with the 

 bacillus of Welch, but only a very small percentage of those infected wounds 

 developed gas gangrene. 



It is clear that a new factor, extrinsic to the infection, enters here, and it 

 was in the search for this factor that the phenomenon recorded in this paper 

 was observed and found to have a more general significance. 



The small incidence of gas gangrene in wounds infected with the bacteria 

 of gas gangrene is, of course, well known. The explanation generally given 

 is that, owing to the special conditions obtaining in war wounds, the latter 

 offer an exceptionally favourable nidus for the growth of these bacteria. 

 Thus, interference with the circulation, the presence of large masses of dead 

 and lacerated muscle, the presence of cloth, dirt or foreign bodies generally, 

 have all been suggested as factors capable of eliciting gas gangrene. But 

 clinical experience shows that many wounds, in which these factors are 

 undoubtedly present, do not develop gas gangrene, and that there are no 

 clinical features of a wound pathognomonic of gas gangrene. And 

 experiment completely fails to give any support to the nidus theory. 

 Splinters of wood, pieces of muscle, of paper, wool, cotton-wool, and of 

 khaki cloth, have been soaked in a suspension of B. Welchii, and then 

 introduced underneath the skin of a mouse without ever producing gas 

 sansrene. Even the minced muscle from a meat broth culture of B. Welchii 

 does not produce gas gangrene if introduced subcutaneously after having 

 been freed from toxin by washing with saline. Nor was it found possible to 

 produce gas gangrene by cutting the femoral artery and injecting a sus- 

 pension of B. Welchii into the leg or into the heematoma, or by injecting 

 suspensions of B. Welchii, together with staphylococci, streptococci, B. coli, 

 B. proteus, or B. sporogenes. The injection of lactic acid in varying con- 



