16 MASS. EXPERIMENT STATION BULLETIN 414 



B. G. B. is so selective in its action that it prevents growth, or gas formation at 

 least, by some strains of coliform bacteria. 



Of the solid media employed for confirmation, both Endo's agar and E. M. B. 

 agar gave false confirmations with the same combinations of bacteria: A, C, and 

 D. MacConkey's agar gave false confirmations with combinations A and D. 

 The Tonney and Noble medium did not give reactions with any of the combina- 

 tions that could have been confused with coliform reactions, and thus it would 

 seem to be an ideal medium for use in water analysis. The medium, however, 

 has limitations, not the least of which is the difficulty of preparing and using it. 



Table 3. — Combinations of Non-coliform Bacteria Giving 

 Positive Tests. 



Presumptive tests 



Confirination tests 



MacConkey's Tonney and 



Combination Endo's agar E. M. B. agar agar Noble agar 



24 hr. 48 hr. 24 hr. 48 hr. 24 hr. 48 hr. 24 hr. 48 hr. 



+ + 



+ + 



Note: + on confirmatory media means that readings could be interpreted as coliform reactions. 



E. Dissolved Metallic Iron 



For many years bacteriologists have observed that the colon bacillus produces 

 a greater degree of acid from carbohydrates than the aerogenes bacillus or many 

 of the intermediates. This fact is the basis of the methyl-red test designed by 

 Clark and Lubs (12). In the course of certain experiments in this laboratory to 

 determine the possible influence of metals on bacterial growth, it was observed 

 that when certain members of the coliform group were cultivated in unbuffered 

 dextrose broth containing metallic iron, sufficient acid was produced to dissolve 

 the iron. Application of the potassium ferrocyanide test detected the dissolved 

 iron. Experiments were done (13) to investigate the correlation of dissolved 

 iron with the methyl-red test. 



In the first experiment a buffer (K2HPO4) was added to Clark and Lubs' 

 glucose broth containing iron filings. The concentrations of buffer ranged from 

 0.2 to 0.5 percent, at intervals of 0.1 percent. Both E. coli and A. aerogenes 

 cultures were inoculated separately into this medium. Tests for dissolved iron, 

 made after suitable incubation periods, showed that 0.3 percent buffer in the 

 medium gave the best differentiation between E. coli and A. aerogenes. Tubes 



