MORPHOLOGIC VARIATION OF BACTERIA 15 



by artificial selection quite different races may be obtained from 

 cultures of protozoa in which no sexual reproduction has occurred. 

 No clear evidence of the occurrence of hybridization in bacteria has 

 as yet been furnished, Almquist's experiments with so-called typhoid- 

 dysentery crosses being too incomplete. It is not enough to show that 

 variation has occurred; it must also be shown that the variation fol- 

 low the Mendelian law. This has recently been claimed for the muta- 

 bile variant of the colon bacillus by Stewart, but his article is not con- 

 vincing since it was possible to obtain from the "hybrid" only one 

 of the assumed parent factors. His assumption that the other char- 

 acter was associated with a lethal factor that suppressed growth is 

 hardly warranted, since the same conditions hold with practically 

 all bacterial mutants, and it is hardly likely that in every hybrid- 

 ization one of the characters is associated with a lethal factor. Topley 

 and Ayrton have pointed out that these variations can be as well 

 explained by a non-equational cell division; that we may have segre- 

 gation without syngamy. With most bacterial variations it is 

 rather difficult to distinguish between temporary modifications and 

 true mutations, and the possibility of inheritance of acquired modi- 

 fications in microorganisms has not been disproved. In short, our 

 know^ledge of the mechanism of variation in unicellular forms is quite 

 too incomplete to permit us to use such variations as a premise from 

 which to argue the occurrence of sexual reproduction. 



With these tw^o unwarranted assumptions, that similarity of form 

 indicates indentity of function, and that variation is necessarily an 

 evidence of conjugation, we are led into a sort of circle of false logic 

 which runs something like this: Certain cell forms of bacteria look 

 like conjugating cells or products of conjugation. Let us call them 

 such; they are ascospores (or what you like). Since they are sexual 

 spores they should at times give rise to hybrids. Variations are 

 found; therefore sexual reproduction occurs; therefore our assumption 

 of the sexual nature of the observed structures was valid. But there 

 is no proved fact anywhere in this argument, save the existence of 

 both morphologic and cultural variation. 



In fact, neither the data nor the logic of these new pleomorphists 

 are adequate to convince us that bacteria possess complex fungoid 

 life cycles. We may sum up the present status of the problem by 



