BACTERIAL INHIBITION 537 



toward their own udder organisms, while in table 1 (columns 1, 

 2, and 3) their milk shows distinct bactericidal properties for 

 the mixed flora of several cows. This also emphasizes the speci- 

 ficity of the germicidal property and suggests an immunity of 

 the organisms constantly present in the cow's own udder. 



Observations were made on agglutination in connection with 

 the numerical determination of the bacteria by plate count. 

 The microscopical examinations for agglutination were made 

 according to Breed's method for direct count from the milk 

 cultures just before inoculation, and from the raw and heated 

 milk just after sampling for plating. There was in no case 

 agglutination in the heated milk or in the milk cultures used for 

 the inoculation, which were made from sterilized milk. In the 

 raw milk there was a very well-defined agglutination of Culture 

 R, which appeared in large clumps of from twenty to several 

 hundred, as opposed to an even distribution of individual organ- 

 isms throughout the preparations made from the comparable 

 heated milk. Bad. lactis-acidi showed a similar but weaker ac- 

 tion in the raw milk, forming clumps of from four to twenty 

 organisms. The preparations with Bad. coll in the raw milk 

 were not so satisfactory, for where the counts were low the scar- 

 city of organisms made microscopical enumeration difficult. 

 A few small clumps were noted in the raw milk smears made at 

 240 minutes on January 26 and February 9 (table 2, columns 1 

 and 2). 



Several theories have been advanced as explanations of germi- 

 cidal action. Its specificity and reaction to heat suggest a sero- 

 logic origin. The presence of agglutinins, antitoxins, hemolysins, 

 opsonins, and other antibodies in raw milk has been frequently 

 reported since Ehrlich's work in 1893. The decrease in bacteria 

 in raw milk has been explained by some investigators as an ap- 

 parent one, due to agglutination, rather than as a true bacteri- 

 cidal action, but the observations reported above do not show a 

 correlation of agglutination and germicidal action. Summariz- 

 ing the experimental data, Culture R gives the best agglutination 

 but no germicidal action; Bad. ladis-acidi gives a good aggluti- 

 nation and an immediate increase in growth; and Bad. coli shows 



