New York Agricultural Experiment Station. 99 



Thirty-six of the sixty-five samples which showed a plate count of 

 less than 10,000 per cubic centimeter were morning milk and pre- 

 sumably many of the bacteria present were derived from the udder. 

 The temperature requirements of these organisms tend to prevent 

 their growth on agar when incubated at 21° C. Consequently many 

 of these udder organisms would fail to reveal themselves on the 

 plates, while they would be seen and counted with the microscope. 

 The remaining twenty-four samples where the plate counts were 

 less than 10,000 per cubic centimeter were night milk, but had 

 been kept at temperatures below 10° C. (50° F.), thereby retard- 

 ing all bacterial development. Therefore, the greater part of their 

 bacterial flora was probably of udder origin and the same explana- 

 tion of the discrepancy in the count would apply. 



The closer agreement between results secured by the two methods 

 when the number of bacteria is high is probably explained in a 

 similar way. Freshly-drawn milk contains bacteria adapted to 

 udder conditions. But as the temperature is lowered and the 

 milk grows older and other organisms gain access, the bacteria 

 from the udder gradually die out or are overgrown by bacteria 

 which thrive better at lower temperatures. Thus in such cases if 

 the agar plates are incubated at a temperature fairly comparable 

 to that under which market milk is usually kept, it is reasonable 

 to expect that practically all of the bacteria will grow, and it is 

 not surprising to find that this actually happens. It has also been 

 observed that as the milk approaches the souring point there are 

 proportionately more isolated bacteria and smaller clumps. This 

 being true, then, since a colony on a plate develops from a single 

 source, there would naturally be a closer agreement between the two 

 counts. Even though some of the udder organisms do not grow 

 on agar plates at 21° C. they are ordinarily so few in number as to 

 have no appreciable effect upon the counts, except when the counts 

 are low. 



As stated, there is a general relationship between the counts made 

 by the two methods, yet occasional very wide variations from the 

 normal differences between the two counts are found. For example, 

 a sample taken on February 19 (Table II) showed a plate count 

 of 120,000 colonies per cubic centimeter while the microscopic count 

 showed the total number of individual bacteria to be 10,160,000 

 per cubic centimeter. On the following day a sample was taken 



