BACTERIOLOGICAL STUDY OF WATER. 599 



perirnent Station has shown that gelatin of such a degree 

 of acidity as to require the further addition of from 15 to 

 20 c.c. per litre of a normal caustic alkali solution to bring 

 it to the phenolphtaleio neutral point gives, on the ^^'hole, 

 the best results. Thus, by way of illustration. Fuller 

 found that a sample of Merrimac River water gave 

 5800 colonies per c.c. on phenolphtalein neutral gel- 

 atin, 15,000 colonies on gelatin that would need 20 c.c. 

 of normal alkali solution to bring it up to the phenol- 

 phtalein neutral point — i. e., a feebly acid nutrient gel- 

 atin, and 500 colonies on a gelatin so alkaline as to 

 require 20 c.c. of a normal acid solution to bring it 

 back to the phenolphtalein neutral j^oint. 



Throughout this part of the work it is to be borne in 

 mind that when reference is made to plates it is not to 

 a set, as in isolation experiments, but to a single plate. 



Method of Counting the Colonies on Plates. 

 — For convenience in counting colonies on plates or in 

 tubes it is customary to divide the ^^hole area of the 

 gelatin occupied by colonies into smaller areas, and 

 either count all the colonies in each of these areas and 

 add the several sums together for the total, or to count 

 the number of colonies in each of several areas, ten 

 or twelve, take the mean of the results and multiply 

 this by the number of areas containing colonies. The 

 latter procedure obtains, of course, only when all the 

 areas are of the same size. By this method, however, 

 the results vary so much in different counts of the same 

 plate that they cannot be considered as more than rough 

 approximations. 



Note. — Prepare a plate ; calculate the number of 

 colonies upon it by this latter method. Now repeat 



