GROWTH OF BACTERIA IN NORMAL MILK. 23 



In addition to these common forms there are some twenty or 

 more other species of bacteria which are found very frequently 

 in milk in this vicinity. Most of these types are easily sepa- 

 rated from each other by the study of their colonies in gelatin. 

 Some of them produce pigments, some of them produce colo- 

 nies of peculiar shapes, and in regard to others there are char- 

 acters which enable us to recognize the species by the study of 

 the colony. These different species are included in the follow- 

 ing tables in appropriate columns, and on later pages will be 

 given further descriptions as to the method by which the more 

 important colonies are distinguished in gelatin. 



We recognize that the differentiation thus obtained is never 

 exact and that there is always a considerable possibility of 

 error. It is not possible with certainty' to separate B. aero- 

 genes and B. coli; it is quite certain that the Streptococcus group 

 consists of more than one kind of bacteria, and the liquefiers 

 are certainly several species. It is probable in regard to the 

 other types also that errors occur and that sometimes different 

 species of bacteria produce colonies so much alike that they are 

 not differentiated in our tables. Moreover, in cases where 

 liquefiers are abundant it is sometimes necessary to study the 

 plates before the colonies are grown sufficiently for proper 

 differentiation. At best, therefore, the differentiation which 

 is given in the tables must be recognized as only approximate. 

 We are convinced, however, that the error is not very great 

 and that the numbers which are given in the tables herewith 

 may be taken as expressing an approximation to the actual 

 number of the species present in the milk. At all events, 

 while recognizing the possible errors of this method it seems 

 to us that the results obtained are considerably in advance of 

 those hitherto obtained in the study of bacteria in milk. They 

 give much new information and are therefore likely to be use- 

 ful in the solution of various problems even though not strictly 

 accurate. 



CALCULATION OF RESULTS. 



A little explanation is necessary as to the method of calcu- 

 lating results. The plates have been allowed to grow three or 

 four days, and then studied as follows: The total number of 

 colonies is counted. The number of colonies of each clearly 



