10 MASS. EXPERIMENT STATION BULLETIN 414 



being old, were curbed with loosely laid field stones and were covered with flat 

 stones, both of which would unavoidably admit some surface water. 



On the basis of sanitary surveys, these wells would have been approved as 

 sanitary, yet bacteriological tests strictly interpreted would have condemned 

 the water of all of them as unfit for household use. A majority of positive cul- 

 tures from these wells were aerogenes bacilli or coliform intermediates. Thus, 

 the study of the wells confirmed the information obtained from the study of 

 water samples sent in to the laboratory to be tested. 



The results from the water samples and from abandoned wells indicated the 

 need for a simplification of the procedure for detecting sewage contamination in 

 water. It would be desirable if one medium could be found to take the place of 

 the several required by the "Standard Methods." Two new media favorably 

 reported by investigators were tried (C). They are designated in the abstracts 

 as B. G. B. and B. C. P. broths. The first medium contains bile and a dye in- 

 tended to promote the growth of sewage types of coliform bacteria and suppress 

 non-sewage bacteria; the second contains dyes for the same purpose and an 

 indicator for the detection of acid which coliform bacteria produce in media con- 

 taining sugars. 



These media were inoculated with water from 76 routine samples, and the 

 standard lactose broth was inoculated for comparison. The results did not 

 encourage the substitution of either of the media for the "Standard Methods" 

 procedure, even though the latter requires two media and several days to get 

 results. 



In the next study (D) several media were investigated to determine their 

 ability to detect false presumptive tests which have already been explained. 

 Several combinations of species capable of symbiotically producing gas from 

 lactose have been described, and the abstract of this study lists some of these. 

 It is quite common for at least one member of a symbiotic association of bacteria 

 to so resemble the coliform bacteria as to make its detection difficult. A number 

 of media commonly employed in water analyses were tried with several sym- 

 biotic combinations of bacteria to see if one or more media could be found that 

 would fail to support growth of the symbiotic bacteria. Included were the "Stand- 

 ard Methods" media, the B. G. B. and B. C. P. broths employed in the previous 

 study, and a special agar medium (Noble's) for which much was claimed, at the 

 time, as a differential medium. B. G. B. gave better results that did lactose broth 

 or B. C. P., but it has been claimed that the B. G. B. medium will prevent growth 

 of gas production by some colon bacilli (E. coli). Noble's agar medium gave 

 results that were encouraging, but it has disadvantages that make its use im- 

 practicable for most technicians; so again the experiments failed to find any short 

 cut to dependable results. 



The colon bacillus produces acid very actively when cultivated in broth media 

 containing simple sugars, particularly if the sugar happens to be dextrose (glucose) 

 or lactose. The aerogenes bacillus and many of the intermediates produce acid 

 weakly or not at all. These facts are the basis for the methyl-red test wideh- 

 used to differentiate the colon bacillus from other members of the group. The 

 next studies (E, F) were based on observations previously made that when metallic 

 iron or erythrosin (a dye) were incorporated in a broth containing dextrose and a 

 buffer to partially counteract the acid, and the mixture was inoculated with the 

 colon bacillus, the iron was dissolved or the erythrosin was precipitated. 



In this study, cultures of colon and aerogenes bacilli were inoculated separately 

 into both iron-dextrose broth (E) and erythrosin-dextrose broth (F). The results 



