THE GOLI-AEROOENES GROUP 661 



type, named it Bad. coli communius. The application of a more extended series 

 of tests resulted in further subdivision of this group, and elaborate classifications 

 were suggested by various observers on the basis of the results obtained (Bergey 

 and Deehan 1908, MacConkey 1905, 1909, Jackson 1911). It may be noted that 

 one important correlation between biochemical activity and natural habitat had 

 already been detected. The Bad. aerogenes type was found to be a relatively in- 

 frequent inhabitant of the intestine, but was frequently isolated^from certain grasses 

 and from the soil, while Bad. coli commune and Bad. coli communius were noted 

 to be typically intestinal parasites (Winslow and Walker 1907). This correlation 

 was of practical as well as of theoretical importance. The presence or absence of 

 " B. coli " in water supplies, and the relative number of this organism if present, 

 soon came to be recognized as a very valuable indication of the presence and 

 degree of faecal pollution (see Chapter 92), and it became very desirable, apart 

 from any question of systematic classification, to difierentiate between those types 

 which were of intestinal origin, and those which might occur in unpolluted waters. 

 The investigations of those who have been primarily concerned with the practical 

 aspects of the bacteriological analysis of water supplies have added materially 

 to our knowledge of this group. 



Apart from the merely positive or negative results, as regards acid or gas 

 production in the various sugars, certain observations made in the earlier days had 

 indicated a difference in kind between the fermentation of one and the same carbo- 

 hydrate by different strains of bacilli of the colon type. Thus Smith (1895), using 

 the method of the fermentation tube, noted that gas was produced more rapidly 

 and in greater amount by Bad. aerogenes than by Bad. coli ; and a rough estimation 

 of the ratio of CO2 to Hj in the gas evolved showed that this was higher with the 

 former organism than with the latter. He noted also that the degree of final 

 acidity was lower with Bact. aerogenes than with Bad. coli. In both respects Bad. 

 cloacce, a coliform organism isolated from sewage by Jordan (1890) and differen- 

 tiated from all other types of coliform bacilli by its power of liquefying gelatin, 

 corresponded with Bact. aerogenes. Kussell and Bassett (1899) confirmed the 

 differential value of a high or low COj : Hj ratio, and noted that the high-ratio 

 strains appeared to be normal soil forms, rather than intestinal parasites. This 

 question was placed on an entirely new footing by the careful quantitative studies 

 of Harden and his colleagues (Harden 1901, 1905, Harden and Walpole 1906), who 

 showed that strains of coliform bacilli were divisible into two well-defined classes. 

 In one, typified by Bad. coli, the CO2 : Hj ratio of the gas evolved gave a value 

 closely approximating unity. In the other, typified by Bact. aerogenes, the CO2 : Ha 

 ratio gave a value of 2 : 1, or thereabouts. These observations have been amply 

 confirmed by later observers. 



Voges and Proskauer (1898) had described a colour reaction given by certain 

 bacteria, but not by Bact. coli. It is obtained by adding a few drops of a 

 strong solution of potassium hydrate to a culture grown in a dextrose medium. In a 

 positive reaction a red, fluorescent coloration appears, which may develop relatively 

 slowly. The nature of this reaction was elucidated by Harden and his colleagues 

 (Harden 1906, Harden and Norris 1911), who showed that it depends on the pro- 

 duction of acetylmethylcarbinol (CHs-CHOH-CO-CHa) ; this, in the presence 

 of alkali and of atmospheric oxygen, is oxidized to diacetyl (CHa-CO-CO-CHj). 

 which reacts with the peptone of the broth to give the red colour. 



This reaction had been applied to the examination of the colon group by some 



