THE COLON GROUP OF BACILLI 133 



present, red colonies upon a blue field will be visible. 

 The Htmus-lactose-agar plate may become blue again 

 after 48 hours, owing, presumably, to the formation of 

 amines and ammonia by the action of the bacteria 

 upon the nitrogenous matter present. If the dilution 

 is too low, the resulting colonies will be small and 

 imperfectly developed, making it difficult to be sure 

 of pure cultures for the subsequent tests. A great 

 number of colonies will also prevent the change of 

 reaction from acid back to alkaline. 



In the selection of those red colonies which are to be 

 fished from the litmus-lactose-agar plate the appearance 

 of the growths must be closely noted. A colony of 

 irregular contour, surrounded by a very faint area of 

 reddening, will probably belong to some member of the 

 B. mycoides group (Winslow and Nibecker, 1903); 

 small, compact, bright-red colonies are characteristic 

 of the streptococci, and Gage and Phelps (Gage and 

 Phelps, 1903) have pointed out that of these there are 

 two types, one of a brick-red color, and of such con- 

 sistency as to be readily picked up by the needle-point, 

 and the other smaller and of an intense vermilion 

 color. The colonies of the colon bacillus are usually 

 well formed, pulvinate on the surface and fusiform 

 when growing deeper down. 



If no red colonies appear on the litmus-lactose-agar 

 plate after a positive result in dextrose broth one of 

 four things has occurred: There may be an organism 

 present which forms gas in dextrose but no acid in 

 lactose; there may be present forms which individually 

 fail to attack lactose but growing together, symbiotically, 



