THE COLI-AEROOENES GROUP 



663 



Hulton 1916, Leviiie 1916c, d, 1917, Burton and Rettger 1917, Chen and Rettger 

 1920, Winslow et al. 1919, Bardsley 1926, 1934, Pawan 1931). 



Besides the division rendered possible by the tests we have just outlined into 

 a coli group on the one hand and an aerogenes-cloocce group on the other, further 

 work revealed the occurrence of a third group of strains possessing properties 

 intermediate between those of the two main groups. This group is, as yet, not 

 completely defined, and is therefore most conveniently referred to as the " inter- 

 mediate " group. Brown (1921) drew attention to the usefulness of a medium con- 

 taining citrate for the differentiation of Bact. coli from Bad. aerogenes. Koser (1923, 

 1924, 1926a, b) devised a synthetic medium in which citrate was provided as the 

 sole source of carbon. He found it possible to differentiate coliform bacilli into a 

 M.R. +, V.P. — , citrate — coli type,. a M.R. — , V.P. +, citrate + aerogenes type, 

 and a M.R. +, V.P.—, citrate + intermediate type. Examination of 104 soil 

 strains from fields subjected to only chance pollution showed that 23-1 per cent, were 

 of the coli, 67-3 per cent, of the aerogenes, and 7-7 per cent, of the intermediate type. 

 Further work in numerous countries soon revealed the value of this test in differenti- 

 ating the intermediate from the coli group (see Bardsley 1926, 1934, Pawan 1931). 



As pointed out by Vaughn, Mitchell and Levine (1939), and Levine (1941), it is advisable 

 to carry out the methyl-red test on a culture incubated for 5 days at 30° C, and the 

 Voges-Proskauer test on a culture incubated for not more than 2 days at 30° C. At 37° C, 

 cultures of some strains of the aerogenes-doacoe group may fail to revert to alkaUne ; and 

 acetylmethylcarbinol is either not formed or is destroyed (see Tittsler 1938). Some 

 of the discrepancies in the literature are doubtless ascribable to carrying out these tests 

 luider unfavourable conditions. Special precautions have also to be taken with the 

 citrate test. It may be noted that, according to a number of authors (Kline 1935, Stuart, 

 Griffin and Baker 1938, Parr 1939, Griffin and Stuart 1940), some strains of Bact. coli 

 may acquire the power of utiUzing citrate. 



In Table 40 we have summarized the results recorded by various observers 

 (see Bardsley) with regard to the percentage of strains isolated from different 

 sources, which give the particular reactions to which we have referred. We have 

 included the indole reaction, because recent work suggests that it is particularly 

 significant in relation to habitat. It will be noted that figures are available for all 

 tests only in the case of the strains derived from animal fseces ; but it may safely 

 be assumed that the strains with a high gas ratio would have given positive V.P. 

 and negative M.R. reactions, and vice versa. 



TABLE 40 



Showing the Percentage of Lactose-fermenting, Coliform Bacilli, prom various 

 Sources, which give the Reactions indicated (various Authors). 



