VARIATION AND MUTATION IN ENZYME SYSTEMS 297 



evidence that it depends on a major variation in cell organization, though it must 

 be noted that Penfold (1910a) found a lactose-fermenting strain of this organism 

 to be very unstable, with a marked tendency to revert to the non-lactose-fermenting 

 form. 



In this connection we may refer to an interesting type of variability in en- 

 zymic activity which was first demonstrated by Massini in 1907. This observer 

 described a coliform bacillus which, on first isolation, failed to ferment lactose, 

 and hence gave rise to colourless colonies on an agar medium, containing lac- 

 tose and an indicator that gave a red colour in the presence of acid. From the 

 third day of incubation onwards, small papillae began to appear on these colonies, 

 and took on a red tint, indicating that the bacilli composing them were break- 

 ing down the lactose with the formation of acid. Subcultures from these papillae 

 gave non-papillated red colonies, showing that the power to ferment lactose had 

 been transmitted to the descendants of the bacilli which had originally formed 

 the red papillae ; and repeated subcultures showed that this power was not sub- 

 sequently lost. Subcultures from the colourless parts of the original colonies, 

 however, gave rise to colourless colonies on which red papillae appeared after about 

 3 days, just as in the case of the original culture ; and repeated subcultures from 

 the colourless portions of the colonies of successive generations gave similar results. 

 Thus, the non-lactose-fermenting form of this organism showed a constant tendency, 

 when grown on a lactose-containing medium, to give off lactose-fermenting variants 

 in which the new character appeared to be permanent. To this organism Massini 

 gave the name of Bacterium coli mutabile. Such a bacterial strain may, as Dobell 

 has pointed out, be likened to the ever-sporting races of plants which have frequently 

 been described. Massini's observations have been confirmed, in all essentials, by 

 many subsequent workers (Burk 1908, Benecke 1909, Burri 1910, Kowalenko 1910, 

 Baerthlein 1912«, b) ; while Benecke, and Kowalenko, added greatly to the signifi- 

 cance of their results by starting with a culture obtained from a single bacterial cell, 

 thus eliminating the possibility that the phenomena resulted from an original 

 admixture of strains. In a careful quantitative study, Lewis (1934) showed that 

 the mutation rate in this organism was such that when it was grown in a lactose- 

 free culture medium, one in every 100,000 viable cells produced a lactose-fermenting 

 colony. Lewis also observed that the proportion of variants was relatively constant. 

 A relatively constant proportion of variant and normal cells has been recorded 

 by other observers as characterizing certain mutating strains (see, for example, 

 Solotorovsky and Buchbinder 1941). 



Experiments along similar lines, employing other species belonging to the same 

 bacterial group, have been carried out by Penfold (1910a and b, 1911a, b and c, 

 1912) and Miiller (1908, 1911). The observations recorded by these observers 

 have made it clear that the behaviour of Massini's Bad. coli mutabile is by no 

 means a bacteriological curiosity, but that, given a substrate appropriate to the 

 particular species under investigation, many members of the coli-typhoid-dysentery 

 group will adapt themselves to a particular nutrient material which they do not 

 immediately attack by giving rise to variants endowed with the power of breaking 

 down this particular substrate. 



Thus, Salm. typhi usually fails to ferment dulcitol ; but when grown on a 

 medium containing that alcohol, it gives colonies which develop dulcitol-ferment- 

 ing papillae (Penfold 1910a, b, 1911) ; the same organism behaves similarly towards 

 rhamnose (Miiller 1911) ; while Salm. paratyphi B behaves similarly towards 



