GROWTH OF BACTERIA IN NORMAL MILK. I9 



amount is preferable. We find it advantageous to buy the dry 

 litmus in rather large quantities, and then after a single ex- 

 periment has shown us the amount of litmus of the particular 

 sample that is needed to produce the desired color, we continue 

 to use the same percentage until the whole quantity of litmus 

 has been used. A new sample of litmus will require a new 

 percentage. 



This culture medium is used exactly as ordinary gelatin. 

 The milk to be studied is diluted in sterilized water to the de- 

 sired amount, and a cubic centimeter of the final dilution is 

 added to a test tube containing a definite quantity of the steril- 

 ized litmus gelatin. The material is poured into petri dishes, 

 hardened, and set aside at ordinary temperature for develop- 

 ment. In actual use we endeavor so to dilute our milk that 

 the plates which we obtain will have from a hundred to a 

 thousand colonies each. If more than a thousand are present 

 on a plate it is almost impossible satisfactorily to differentiate 

 the colonies from each other and to determine their species. If 

 less than a hundred are present, while the differentiation of 

 species is very satisfactory, we do not think that the sample 

 gives as close average results as when the number is larger. 

 A plate containing two or three hundred is the most satis- 

 factory. It is of course not possible in all cases to regulate the 

 dilutions so that the plates shall have the proper number, of 

 colonies. Only long experience with the milk under conditions 

 of experiment enable us to determine this accurately, and as 

 will be seen in the following experiments, even long experi- 

 ence is sometimes at fault. It frequently happens that two 

 samples of milk under identical conditions show very great 

 differences in the number of bacteria, and no rule as to the 

 amount of dilution can be applied to the two samples and give 

 the same results with certainty. The determination of the 

 proper dilution of the milk is one of the most difficult problems 

 in the experiments. 



In order to differentiate the colonies and determine from their 

 study the species of bacteria present, it is necessary that the 

 plates should be allowed to grow several days before examina- 

 tion. The study of the plates when they are two days old will 

 enable one readily enough to count the number of bacteria, but 

 not properly to differentiate the species. The same plate when 



