298 BACTERIAL VARIATION 



rafl&nose. From Miiller's account it would appear that the rhamnose-fermenting 

 variants of Salm. typhi, and the raffinose-fermenting variants of Salm. paratyphi B, 

 are non-reverting modifications of the parent strain. Apparently, also, subcultures 

 from the non-fermenting portions of the colonies, in these two organisms, showed 

 the same tendency to throw off fermenting variants in the form of papillae, as was 

 observed by Massini in the case of Bact. coli mutahile. Penfold's observations do 

 not, however, confirm the absence of a tendency towards reversion. In connection 

 with the appearance of dulcitol-fermenting forms of Salm. typhi, in particular, he 

 finds that subcultures from the dulcitol-fermenting papillae, or from fermenting 

 cultures in dulcitol peptone water, show a marked tendency to revert to the non- 

 fermenting parent form during the earlier generations. If the selective process is 

 continued through a long series of successive generations the tendency to reversion 

 becomes less and less ; though it appears doubtful whether absolute permanency is 

 ever attained. Penfold concludes from his results that the more rapidly a par- 

 ticular species acquires the ability to ferment a particular substrate, the less 

 tendency is there for subsequent reversion ; while the longer and more rigorous 

 is the training required to bring about the appearance of fermenting variants, the 

 longer must that training be afterwards continued to make a lasting impression 

 on that particular strain. 



Certain experiments carried out by Penfold (19116, c) and by Kevis (1911, 

 1912) have brought to light a different type of impressed variation in bacteria. 

 Penfold found that, by growing certain strains of Bact. coli on an agar medium 

 containing sodium monochloroacetate, he was able to isolate strains which retained 

 the power of producing acid in all the usual carbohydrate media, but which had 

 lost the power of producing gas in many of them. Moreover, he found that while 

 the power to produce gas from sugars, or from substances giving rise to sugars 

 on hydrolysis, was usually suppressed, the power of forming gas from alcohols, 

 such as mannitol or dulcitol, was usually unaffected, or but slightly diminished. 

 The only exception noted was in the case of rhamnose, a methyl pentose, which 

 was fermented with gas formation by the variant strains. 



Goodman (1908) obtained variants showing differences in fermentative ability 

 by an essentially different technique. Starting with a particular strain of G. 

 diphthericB, which produced a certain degree of acidity in dextrose broth, he in- 

 oculated 15 tubes of this medium from a single colony, and determined the degree 

 of acidity attained after a few days. From the tube showing the highest acidity 

 he inoculated 15 tubes of the same medium and a similar number from the tube 

 showing the lowest acidity. This process he repeated through 36 successive 

 subcultures. At the end of this series the high-acid strain produced a titratable 

 acidity more than twice as great as the parent strain, while the low-acid strain 

 produced no acidity at all. As would be expected this strain had also lost its 

 power to produce acid from maltose, but it is of interest to note that its power 

 to produce acid from dextrin was almost unaffected. These observations would 

 appear to afford an example of the separation of a bacterial strain into a fermenting 

 and a non-fermenting variant, by a simple process of selection without any modifica- 

 tion of the environmental conditions ; for though it is true that the high-acid strain 

 was in fact subjected to a high concentration of hydrogen-ions during its successive 

 subcultures, while the low-acid strain was not, it is difficult to understand how 

 the latter condition could lead to the production of a variant which had lost its 

 power to ferment dextrose and maltose, while retaining its power to ferment dextrin. 



