RESISTANCE OF BACTERIA 231 



veyed into any region of the body whatsoever. It may be found in 

 plants, for in the nodules of leguminous plants the nitrifying bacteria 

 are not in pure culture but in a culture mixed with the bacteriophage. 

 None of the bacteria are necessarily protected from the bacteriophage. 

 Wherever the bacterium may be found the bacteriophage intervenes 

 and a struggle takes place. But the bacteria resist and symbiotic 

 cultures or mixed cultures result. Without treating at length all of 

 these facts, which will not be emphasized here, it may be well to remark 

 that the fact that mixed cultures are found among laboratory stock 

 cultures is not particularly incomprehensible. 



We know that unless the bacteriophage possesses a maximum viru- 

 lence the formation of secondary cultures after bacteriophagy is the 

 usual thing. It would be indeed strange if this were not also true 

 under natural conditions. To assert the contrary would be to affirm 

 that bacteriophagy is solely a laboratory phenomenon. If we trans- 

 plant on to agar an artificial mixed culture we know that certain colonies 

 are ultrapure, while others, more or less numerous, are contaminated by 

 a bacteriophage. If the secondary culture forms in nature, by starting 

 with any natural product whatever and culturing this on agar we 

 should obtain both ultrapure colonies and mixed colonies. Trans- 

 planting one of these ultrapure colonies the strain isolated will remain 

 ultrapure indefinitely, but if by chance we select for transplanting 

 a mixed colony the strain may perpetuate itself in the form of a mixed 

 culture contaminated by bacteriophage corpuscles. 



I have said "may perpetuate" for experiment shows that a mixed 

 culture prepared artificially frequently undergoes an auto-purification 

 through the elimination of the bacteriophage corpuscles in the course of 

 successive transfers. The same thing certainly takes place with 

 naturally mixed strains and I have noted upon two different occasions a 

 purification of this type during the series of transplants. Both of these 

 instances occurred with two strains of B. coli isolated from cases of 

 cystitis. It would seem that if this ultrapurification by elimination 

 did not take place the number of mixed strains to be found in labora- 

 tories would be far greater than it actually is. 



Let us bear in mind that it may be very difficult to prove whether a 

 bacterial strain is ultrapure or contaminated. Experiments on artificial 

 mixed cultures show that after a certain number of transfers there is 

 not only no further elimination of corpuscles, but a perfect equilibrium 

 is established between the capacities of the two antagonists to attack and 

 to react. No plaques are to be found on agar and the appearance of 



