440 



G. S. WILSON 



TABLE 22 



Experiment 16 



DISCUSSION 



How, it may be asked, can the discrepancy between the total 

 and viable counts during the logaritlimic phase of growth be 

 explained? It is only necessary to consider three hypotheses. 

 (1) It might be referred to technical errors; that this is not so 

 has been shown already. Moreover the uniformity of the indi- 

 vidual counts suffices to alleviate one's few remaining suspicions 

 on this point. (2) It might possibly be concerned mth a method 

 of reproduction other than that by binary fission. If, for in- 

 stance, there were during certain phases of the culture a multi- 

 plication of the organisms by a process of budding or even by a 

 sexual mode of some kind, it is conceivable that a certain number 

 of parent cells might die or pass into a resting stage, and thus 

 lead to the presence of a proportion of dead or inactive bacteria 

 in the medium, even during the height of growth. Fortius, 

 however, there is no direct evidence. Division of the particular 

 strain of Bact. suipestifer used has been watched for several 

 generations in gelatin film preparations on a large number of 

 occasions, and never has any other method of reproduction than 

 that by binary fission been encountered. (3) The most likely 

 explanation would appear to be that of every generation produced 

 a certain number of bacteria die. If, for example, 80 per cent 

 of the organisms produced during a given generation continued 



