MORPHOLOGIC VARIATION OF BACTERIA 11 



that they sound like established fact, particularly in the works of 

 Lohnis and Enderlein. I have assumed in such cases where illus- 

 trations or actual descriptions of the method are not furnished that 

 the statement has not been verified by actual continuous observation. 

 In most cases, however, it is quite clear that no continuous observa- 

 tion has been made, and that the author's idea of the sequence of 

 events has developed from an attempt to trace the transitions in 

 different individuals in the same preparation. This clearly involves 

 an element of error. The structures, for instance, which Mellon in- 

 terprets as zygospores, namely two cells with swollen ends in contact, 

 might just as well be interpreted as a cell with a central bulging 

 undergoing division. 



This lack of systematic observation, this attempt to patch to- 

 gether the life cycles from haphazard observation of cultures in 

 widely different media without regard to the age of the culture or 

 the phase of growth, is responsible for the vague and incoherent 

 nature of the descriptions of the life cycles presented. In practically 

 no case can one obtain from the published papers any clear notion 

 as to the author's ideas regarding the actual sequence of events in 

 the cycle; as to when gonidia are formed, and when asci and zygo- 

 spores; under what conditions simple conjunction occurs and under 

 what circumstances a symplasm results. Lohnis' complicated dia- 

 gram is absolutely unintelligible, and the one published by Mellon 

 to illustrate the ontogeny of a diphtheroid organism (1926, v), while 

 simpler, is far from being clear. 



The fact that practically all of these observed variations in mor- 

 phology have been found in old cultures, and especially on unusual or 

 unfavorable media, of course at once leads one to believe that they 

 are "involution forms" i.e., dead or degenerating cells. This argu- 

 ment is answered at once by the statement that the cultures contain- 

 ing them are perfectly viable. But the fact that growth occurs when 

 such old cultures are transferred to a new medium does not prove 

 that this growth has resulted from the cells showing the morphologic 

 variation, even when such variant cells preponderate. Only careful 

 quantitative observation of the proportion of viable cells and of mor- 

 phologically variant cells, or single cell isolations of variants to new 

 media, or continuous observation of particular variant cells trans- 



