BACTERIA DEVELOPING AT DIFFERENT TEMPERATURES 229 



water, we may compute the ratios between each pair of results on the 

 different samples, and by averaging the ratios so obtained for different 

 classes of samples we may ascertain what value each of the different 

 ratios may have in indicating the quality of a water. As a con- 

 sideration of all of the experimental data, and of all the possible 

 ratios, would occupy more space than is available, many of the data 

 have been excluded, and, in addition to a consideration of the num- 

 bers of bacteria determined on agar after four days' incubation 

 at 20 C., and on lactose agar after 24 hours ' incubation at 30, 40, 

 and 50 C. respectively, and the numbers of bacteria producing red 

 colonies on lactose agar at the same temperatures, only the follow- 

 ing ratios will be presented, these ratios being expressed in every 

 case as the per cent which the smaller value is of the greater. 



1. The ratios between the total number of bacteria determined 

 on agar after four days' incubation at 20 C., and the numbers of 

 bacteria determined on lactose agar after 24 hours' incubation at 

 30, 40, and 50 C. respectively. 



2. The ratios between the total number of bacteria determined 

 on agar after four days' incubation at 20 C. and the numbers of 

 bacteria producing acid fermentation on lactose agar in 24 hours 

 at 30, 40, and 50 C. respectively. 



3. The ratios between the numbers of bacteria and the number 

 of acid-producing bacteria determined at each temperature as above. 



Sources oj the data included. The data used in the discussion 

 and tables have been obtained in part by recomputation of certain 

 of the routine results obtained during the past eight years at the 

 Lawrence Experiment Station, and in part from a special study 

 of the relative counts of bacteria obtained on plates incubated at 

 different temperatures. In addition to the incubators regularly 

 operated at 20 C. and 40 C., we have one spare incubator, which 

 has been operated part of the time at 30 C. and part of the time at 

 50 C. In considering the significance of the numbers of bacteria 

 and the numbers of acid-producing bacteria determined at the 

 different temperatures, it is therefore necessary to divide the results 

 into two series. The two experiments covered much the same 

 classes of waters, and had determinations of bacteria and acid- pro- 

 ducers at 20 C. and at 40 C. common to both experiments. No 



