664 BACTERIUM 



Numerous other tests have been used for the differentiation of members of 

 this group. 



The fermentation of cellobiose, a glucoside derived from cellulose (see Jones 1924, 

 Jones and Wise 1926), has been recommended for distinguishing Bad. aerogenes and 

 intermediate strains, which ferment this substance, from Bad. coli, which fails to do so 

 (see Koser 1926c, Tittsler and Sandholzer 1936). Too much rehance should not be placed 

 upon the result of this test in the identification of individual strains. Batty-Smith (1942a), 

 for example, who examined 600 strains of cohform baciUi, found that about 10 per cent, 

 of Bad. coli strains fermented ceUobiose, and that about one-third of the intermediate 

 and aerogenes strains were without action upon it. 



The production of HjS in a proteose peptone ferric citrate agar medium was found 

 by Vaughn and Levine (1934) to be characteristic of intermediate strains. Further 

 observations by the same authors (1936a, b) showed that, if cysteine was incorporated 

 in the medium, a high proportion of all coliform strains produced HgS, and that even 

 without cysteme, the result was greatly influenced by the concentration of agar. If 

 this test is to be of differential value, strict attention must be paid to the exact composition 

 of the medium. 



Leifson (1933) observed an almost perfect correlation between the V.P. reaction and 

 the fermentation of sodium malonate. 



The fermentation of polysaccharides, such as starch, by the aerogenes group was pointed 

 out by Durham (1901) and confirmed by Levine (1918). It has been made use of by 

 W. J. Wilson (1933) in the preparation of a medium for distinguishing Bad. aerogenes 

 from Bad. coli. On the other hand, a medium containing sodium sulphite and rosolic 

 acid in certain proportions, and another containing hexamine, are said to favour Bact. coli 

 at the expense of Bad. aerogenes (Wilson, W. J. 1933). 



For the differentiation of Bact. aerogenes from Bad. cloacce gelatin hquefaction and 

 the production of acid and gas from glycerol and starch are of chief value (Levine 1918). 

 Bact. aerogenes is gelatin — , glycerol +, starch +, and Bad. cloacce gelatin +, glycerol — , 

 starch — . Intermediate strains are said to be differentiated from strains of coli type, 

 not only by their production of HgS, but also by their ability to form trimethyleneglycol 

 in the anaerobic fermentation of glycerol (see Werkman and Gillen 1932). 



Besides the effect on carbohydrates, nitrogen utilization has been explored m an attempt 

 to discover reliable tests. Koser (1918), Chen and Rettger (1920), and numerous sub- 

 sequent workers (see Bardsley 1926, 1934) found that aerogenes strains were able to utilize 

 uric acid as their sole source of nitrogen, whereas strains of the coli and intermediate 

 types were not. More recently Mitchell and Levine (1938) have examined other sub- 

 stances, such as yeast nucleic acid, allantoin, hydantoin, uracil, and urea. They find 

 that aerogenes strains can make use of all these substances as their sole source of nitrogen, 

 but that intermediate strains can use only vuea, and coli strains only uracil. 



Considerable help is afforded by a test introduced as long ago as 1904 by Eijkman 

 (1904, 1914), who found that coli, but not aerogenes, strains were able to form gas in 

 a glucose broth medium incubated at 46° C. This test has had a chequered career, 

 having been reported on favourably by some workers, and utterly condemned by 

 others. Recent work (see Levine et al. 1934, Wilson et al. 1935) has rendered it 

 clear that the success of the test depends on exact standardization of the incubating 

 temperature, which must be adjusted to 43°-45° C. in the medium itself. The only 

 satisfactory means of doing this is to incubate the tubes in a constant-temperature 

 water-bath. The differential value of the test is greatly enhanced by the replace- 

 ment of glucose by lactose. Our own experience suggests that MacConkey's lactose 

 bile salt broth is the medium of choice. Using this medium in the way described, 

 it was found that of 193 M.R. -f , V.P. — , citrate — , indole -f coli strains, 180 gave 

 a positive 44° C. MacConkey reaction, whereas of 303 other strains, 40 of which 



