BACTERIA DEVELOPING AT DIFFERENT TEMPERATURES 235 



tion. From these figures we see that, with one exception, a much 

 greater percentage of the total bacteria are determined at 30 for 

 polluted waters than for the pure waters; that is to say, the dis- 

 tinction between pure and polluted waters is emphasized by the 

 30 counts. On the other hand, the counts at 40 materially decrease 

 such a distinction, as is shown by the fact that the ratios are greater 

 for the good waters than for the polluted waters. If we assume, 

 however, that the presence of bacteria capable of rapid growth at 

 40 is an indication that the water contains disease-producing bac- 

 teria, the fact that the proportion of such bacteria is greater in the 

 purer waters than in the waters known to be seriously polluted 

 would signify that our filtered waters were not of such excellent qual- 

 ity as their low bacterial content would indicate, which supposition 

 is well worth further study. 



No such lesson is apparent in the ratios between the total bacteria 

 at 20 and the acid-producing bacteria at 30 and 40, the values 

 for the raw waters and filtered waters being much the same. A sharp 

 distinction is noted between the ponds and the other samples in the 

 30 values, although this distinction does not appear in the 40 

 values. The ratios between the bacteria growing at each tempera- 

 ture and the number of acid-formers at that temperature appear 

 to be of little value in distinguishing between the different waters. 

 As will be shown later, the chief use of these ratios seems to be in 

 locating errors and abnormal values in the other ratios and in the 

 counts from which they are computed. 



TABLE 3. 

 BACTERIAL RATIOS FOR DIFFERENT CLASSES OF WATERS, 20, 30, AND 40 C. SERIES. 



