BACTERIA AND RURAL WATER SUPPLIES 17 



inoculated with E. coli gave positive tests for dissolved iron and those inoculated 

 with A. aerogenes gave negative tests. A few drops of a 2 percent solution of 

 K3Fe(CN6) added to a tube gave a blue color if iron had been dissolved. The 

 color was deeper in cultures with a greater degree of acidity because more iron 

 was dissolved. 



To determine the pH value at which iron goes into solution, tubes of the sterile 

 medium containing iron filings were prepared. Sterile acetic acid was added in 

 quantities to give values ranging from pH 4.5 to pH 6. These tubes were un- 

 inoculated and were incubated for the same length of time as were the cultures. 

 Above pH 5.14, no dissolved iron was detected; at pH 5.05, the results were 

 plus-minus; and at pH 4.97 to 4.99, they were plus or double plus. 



A series of 236 coliform cultures, including E. coli, A. aerogenes, and inter- 

 mediates, was differentiated on the basis of the Voges-Proskauer, methyl-red, 

 sodium citrate, and uric acid tests. Then these cultures were inoculated sep- 

 arately' into buffered glucose broth containing iron filings. After 48 hours' in- 

 cubation, tests for dissolved iron were made. Results may be seen in table 4. 



Table 4. — Comparison of Methyl- red Test With Tests 

 FOR Dissolved Metallic Iron. 



From the table it may be observed that all but 11 of the M. R. positive cultures 

 gave plus or plus-minus reactions for dissolved iron. The failure of these eleven 

 cultures and of most of the methyl-red plus-minus cultures to give positive tests 

 may be explained possibly by the observed fact that the pH value did not go as 

 low in cultures containing metallic iron as in those with the same buffer con- 

 centration but no iron. 



F. Erythrosin Precipitation 



A similar study was made (14) of the precipitation of erythrosin by acid- 

 producing coliform bacteria as an indicator of the amount of acid produced. It 

 had been observed, in some previous experimental work, that acid-forming bac- 

 teria precipitated the d^-e in a medium containing dextrose. 



The medium employed was the same as that employed for the dissolved-iron 

 study (Clark and Lubs' medium with 0.3 percent K2HPO4), except that 0.1 

 percent of erythrosin was used instead of metallic iron. The cultures, for the 



