VARIATIONS IN MORPHOLOGY 291 



Barber (1907), starting with a single strain of Bact. coli, selected and subcultured, 

 by means of a micromanipulator, individual bacterial cells that had grown to an 

 unusual length. In this way he was able to isolate three strains that grew in 

 the form of long rods throughout many successive generations, showing no tendency 

 to revert to the modal short bacillary form of the original parent strain. This 

 would appear to afford an instance of the perpetuation of a natural heritable 

 variation by simple selection. 



We have already, in Chapter 8, noted the occurrence of the H — >■ variation 

 — the loss by a flagellated organism of the capacity to produce flagella, associated 

 of course with a loss of motility. When this variation occurs naturally, the variant 

 form usually shows no tendency towards reversion. The importance of this 

 change from the point of view of antigenic analysis led to a search for methods 

 by which it could be induced at will. It has been found that growth on agar 

 containing 0-1 per cent, phenol largely, or completely, suppresses the formation of 

 flagella (Braun 1918) ; but the non-flagellated cells so obtained give rise to the 

 normal flagellated form when subcultured on ordinary media ; so that we are here 

 dealing not with an impressed variation but with a temporary adaptation to 

 environment. 



Another striking morphological variation is the occurrence of asporogenous 

 variants of such spore-bearing bacilli as B. anthracis (Preisz 1904, Eisenberg 1912). 

 Many of these naturally-occurring asporogenous variants have shown no tendency 

 to revert to the normal spore-bearing form. Pasteur (1881a, b) found that 

 the growth of B. anthracis at 42-5° C. for about a month resulted in a great decrease 

 in the frequency of spore formation, as well as in a decrease in virulence ; and 

 Roux (1890) obtained asporogenous strains of this organism by growing it in the 

 presence of low concentrations of antiseptics. Whether the asporogenous strains 

 obtained by these methods were examples of an impressed variation or of a 

 temporary adaptation to environment is difficult to determine. Later experi- 

 ments by Bordet and Renaux (1930), however, strongly suggest the occurrence of 

 an impressed variation of the genetic type. They found that certain strains of B. 

 anthracis, yielding the normal, flat, filamentous colony when grown on solid media, 

 gave rise on prolonged incubation to a characteristic type of papillary daughter 

 colony consisting of asporogenous bacilli. By repeated subculture from these 

 daughter colonies a completely asporogenous strain of B. anthracis could be isolated. 

 When the medium on which these strains of B. anthracis were grown was deprived 

 of its calcium by treatment with oxalate, spore formation was stimulated, and the 

 papillary, asporogenous daughter colonies did not appear. When the calcium con- 

 tent of the medium was increased, by the addition of a little calcium chloride, the 

 frequency of spore formation decreased and the papillary, asporogenous daughter 

 colonies became very numerous. Repeated subcultures from these daughter 

 colonies again produced a completely asporogenous strain (see also Bordet, P. 

 1930). 



There would seem to be an interesting difference between the asporogenous 

 strains of B. anthracis obtained by the method of Pasteur, and those obtained by 

 the method of Bordet and Renaux. Although growth at 42-5° C. for several 

 days leads to a decrease in the average virulence of a culture and to the appearance 

 of various abnormal bacillary forms, many of which are non-spore-bearing, Preisz 

 (1911) was unable to find any correlation between the capacity to form spores and 

 virulence as tested by animal inoculation. Spore-bearing strains might be virulent 



