662 BACTERIUM 



of the observers referred to above, and it had been noted that Bact. aerogenes, as 

 opposed to Bact. coli, gave a positive reaction (Durham 1901). MacConkey (1909) 

 observed the great preponderance of Voges-Proskauer negative types among strains 

 isolated from the faeces ; positive reactions were given by 11 of 178 strains isolated 

 from human faeces, 8 of 67 strains from the faeces of the horse, and none of 87 strains 

 from faeces of the calf, goat, pig or goose. 



The fundamental importance of the Voges-Proskauer reaction, and of the 

 CO2 : H2 ratio, as compared with the presence or absence of fermentation in par- 

 ticular carbohydrates other than lactose, was not however realized by the earlier 

 investigators, so that the reaction was simply assigned a place among some selected 

 series of tests, and V.P. positive and V.P. negative strains were often allocated to 

 the same sub-group ; though it was noted by Howe (1904) during the examination 

 of strains of lactose-fermenting bacilli derived from water, that there was almost 

 perfect correlation between a positive V.P. reaction and the ability to produce 

 large amounts of gas from dextrose. 



Petruschky (1889, 1890) made the first attempt to measure, by titration, the 

 degree of acidity produced by various members of the coli-typhoid group ; while 

 Smith (1895), as noted above, called attention to the low acid production of Bact. 

 aerogenes as compared with Bact. coli. A great advance in the investigation of 

 this aspect of bacterial metabolism was marked by the introduction of indicators, 

 which rendered possible the ready determination of the hydrogen-ion 

 concentration attained during the bacterial fermentation of any test substance. 

 Clark and Lubs (1915), on this basis, devised the methyl-red test for the differen- 

 tiation of members of the coli-typhoid group. The addition of this indicator to 

 five-day cultures in dextrose phosphate peptone water distinguishes between 

 those strains which produce and maintain a high concentration of hydrogen-ions, 

 and those which produce an initial lower concentration of hydrogen-ions and then 

 cause reversion towards neutrality by the further decomposition of the organic 

 acids to carbonates, and perhaps by the formation of ammonium compounds from 

 proteins. The former type, such as Bact. coli, give a red coloration and are referred 

 to as methyl-red positive ; the latter, such as Bact. aerogenes, give a yellowish 

 colour and are referred to as methyl-red negative. It soon became clear that there 

 was a very high negative correlation between the methyl-red test and the Voges- 

 Proskauer reaction (Levine 1916a, h), and a series of intensive studies soon placed 

 on a firm foundation the conclusion, already propounded as a tentative hypothesis, 

 that the lactose-fermenting coliform bacilli could be divided into two primary 

 divisions on the basis of the COj : Hg ratio, the Voges-Proskauer reaction, and the 

 methyl-red test. The -first of these, containing strains giving a COj : Hg ratio of 

 about 2:1, V.P. positive and M.R. negative, comprised the great majority of the 

 strains isolated from plants, grain, and unpolluted soil or water. Such strains were 

 relatively infrequent in material obtained from the intestines of man or animals. 

 This group could be further subdivided, on the basis of gelatin liquefaction, into 

 the non-liquefying form Bact. aerogenes, and the much less common liquefying form 

 Bact. cloacT. The second group, containing strains giving a CO2 : Hg ratio of 

 approximately 1 : 1, V.P. negative and M.R. positive, contained the great majority 

 of those strains isolated from the intestines of man or animals, as exemplified by 

 Bact. coli commune or Bact. coli communius. This group was found to be further 

 divisible, on the basis of the ordinary fermentation tests, along lines which will 

 be considered later (Keyes 1909, Rogers et al. 1914, 1915, 1918, Johnson 1916, 



