TERMINOLOGY ' 289 



and, so long as we regard bacteria as asexual organisms multiplying by simple 

 binary fission, we must avoid the tendency to misapply concepts which have their 

 essential basis in the segregation, and conjugation, of a special system of repro- 

 ductive cells. Some of the concepts are applicable ; as we noted in Chapter 2, 

 the persistence of specific characters through a large number of generations of 

 a bacterium implies an hereditary mechanism of some kind, and one that behaves 

 as a unit within the bacterial cell. For the purposes of bacterial genetics, it is 

 irrelevant whether a nuclear apparatus has been demonstrated morphologically 

 or not. A logical deduction from the mathematical data of genetical experiments 

 with other species leads to a chromosome and gene mechanism. We have as yet 

 insufiicient data on bacterial heredity to justify the assumption that here also 

 there is a similar mechanism, but in the meantime, it is convenient, and indeed 

 sensible, to use those concepts of general genetics which are applicable to bacteria. 



We may postulate a haploid or a polyploid nucleus ; scanty morphological 

 studies suggest that haploid and perhaps diploid nuclei are likely to be commonest 

 in bacteria. The mutations observed in bacteria are characterized by absence 

 of the intermediate forms, and a moderate readiness to reversion. If these are 

 due to genie changes in the nucleus, the most likely of all known types of genie 

 change is transgenation (Lindegren 1935). It should be noted, however, that 

 single-cell cultures, which are essential for the proper study of bacterial genetics, 

 do not necessarily represent a single cell in the genetical sense, for many single 

 cells appear to contain more than one nuclear unit. In this respect, at least, a 

 technique for the morphological demonstration of a nucleus is a necessary founda- 

 tion for the formulation of a "mathematical" nucleus referred to above. 



In one sense, bacteria offer a fruitful field for the study of mutations. Observ- 

 able mutations occur with great rarity among more complex multicellular organisms. 

 If bacterial mutations are of the same order of rarity, the bacteriologist will have 

 ample opportunities for observing them, since the colony that grows from a single 

 cell after a day's incubation on an artificial medium may contain lO'-lO^ individual 

 viable cells, representing the end result of an even greater number of divisions 

 during which mutation could have occurred. 



The term " bacterial dissociation " is frequently employed to denote a par- 

 ticular type of bacterial variation (see Hadley 1927, 1937, and Morton 1940, for 

 detailed reviews). This term, in its generally accepted sense, denotes the appear- 

 ance, in a bacterial culture, of forms which differ sharply, in one or more char- 

 acters, from the " normal " forms of the parent strain ; that is, the strain may 

 be said to have undergone dissociation into two types, differentiated from each 

 other in colony-form, in antigenic structure, or in some other way. The varia- 

 tion must be discontinuous in type, even though there is some overlapping ; and 

 the dissociated, or variant form, must be sufficiently stable to maintain its new 

 characters over several generations, whether or not it eventually reverts, wholly 

 or in part, to the normal form from which it was derived. 



Those who uphold the view that bacteria pass through a complex life-cycle have 

 naturally sought to relate the phenomenon of dissociation to the phases of this 

 cyclical development (Hadley 1927) ; but the term may be employed in a purely 

 descriptive sense, without any reference to the possible existence of modes of 

 reproduction other than simple binary fission. It is, however, doubtful whether 

 " dissociation " has any advantage over " variation " as a descriptive term ; and, 

 since it is desirable, at the present stage, to avoid any implication in regard to 



P.B. L 



