CLASSIFICATION OF B. COLI TYPE 145 



described above. There is also a danger of mistaking pin point 

 red colonies produced by acid-forming streptococci for those 

 produced by attenuated B. coH and it will be found advisable 

 to ignore all such colonies when examining the plates. By 

 this procedui-e, only organisms in a fairly vigorous state are 

 counted, and, although it is somewhat empirical in character, 

 it produces results that are of greater sanitary significance. 

 Of 271 red colonies fished from rebipelagar, the author found 

 that 236 (87 per cent) were of the B. coH group so that even if 

 all the red colonies are counted no serious errors will be intro- 

 duced. 



One difficulty in connection with the use of rebipelagar is the 

 quaUty of the bile salt. Many brands of this salt are pur- 

 chasable but very few are satisfactory. Sodium taurocholate, 

 sodium glycocholate, and many brands of the commercial bile 

 salt are too restrictive in their action on B. coli and if the 

 amount is reduced to avoid this, the selective action is de- 

 stroyed. With bile salt of satisfactory quality, vigorous B. 

 coH will produce colonies 3 to 5 mm. in diameter in twenty- 

 four hours at 37° C. and all brands that fail to do this should 

 be rejected. 



Classification of B. Coli Type. It has been indicated earlier 

 in this chapter (page 136) that an attempt to regard one par- 

 ticular type of B. coU as having more sanitary significance 

 than others has been a comparative failure. The present 

 problem is not the definition of the properties of a distinct 

 biotype such as B. coli communis or B. acidi lactici but the 

 correlation of properties with the immediate previous environ- 

 ment. The fsecal types of B. coli can apparently be distin- 

 guished from those occurring on grain ^^ by the hydrogen ion 

 concentration produced in dextrose broth containing 0.5 per 

 cent of dextrose, 1.0 per cent of peptone, and 0.2 per cent of acid 

 potassium phosphate. This can best be determined by the 

 methyl red reaction of Clark and Lubs ^^ which Levine ^^ has 

 shown to be correlated with the Voges and Proskauer reaction. 

 The precise sanitary significance of these so-called grain types 



