THE COLON GROUP OF BACILLI 103 



in milk, nitrate solution, pep ton solution and gelatin; 

 28 per cent failed to produce indol and 5 per cent did 

 not reduce nitrates. Numerous other results indicat- 

 ing similar variations are cited in Chapters VII and 

 VIII. 



Where shall the line be drawn? The English bac- 

 teriologists usually require in addition to the morpholog- 

 ical characters mentioned above, motility, non-liquefac- 

 tion of gelatin, fermentation of dextrose and lactose 

 media, coagulation of milk, production of indol, and 

 reduction of neutral red. The usual American pro- 

 cedure has included reactions in dextrose broth, milk, 

 peptone solution (for indol), gelatin (absence of liquefac- 

 tion), and reduction of nitrates. Of late years, how- 

 ever, there has been a growing feeling that such 

 arbitrary definitions went either too far or not far 

 enough. The whole group of lactose-fermenting bacilli 

 is characteristically of intestinal origin. That we be- 

 lieve to have been clearly established by results to 

 be cited later in this chapter and in the succeeding one. 

 A differentiation between various sub-tj-pes of this 

 group can only be properly justified by the fact that 

 some of them are less resistant in water than others, 

 and hence are indicative of fresh and recent pollution. 

 There is some evidence that such is the case which will 

 be discussed in Chapter VIII, but whatever may be 

 concluded from the somewhat conflicting opinions 

 on this point it appears certain that, in temperate 

 climates at least, the whole class of lactose fermenters 

 should be absent from safe water supplies. As the 

 Conmiittee on Standard Methods of Water Analysis 



