424 MILK 



cially since they have been able to detect large numbers of strep- 

 tococci by the direct method. 



Brew and Dotterer realize that the clumps of bacteria affect 

 the accuracy of the microscopic count, and that the counting is 

 rendered difficult when the clumps are large and dense. But 

 they add that the plate count is affected not only by the same 

 conditions, but that the clumps do not break up entirely into 

 individuals, and that thereby the plate count becomes less reliable 

 than the microscopic count. The authors emphasize the usefulness 

 of the microscopic test in rapidly grading milk and distinguishing 

 between milks of small, medium, and large germ content. 



However, modern bacterial standards, and, in fact, nearly all 

 published reports of bacterial enumerations in milk, are based 

 on the plate method. If the direct microscopic method were 

 generally introduced, ideas in regard to numbers of bacteria in 

 milk would have to be changed. However, this should not be 

 a valid objection to a new method, if such method is recognized 

 as superior. An improved method that brings us nearer the truth 

 compels us to discard old standards and establish new ones. 

 Therefore, it is desirable that comprehensive tests of the micro- 

 scopic enumeration of bacteria in milk be made and, if its superior- 

 ity be proved, it should find its proper place in bacteriologic 

 technic. 



As stated before, the discrepancies in results obtained by dif- 

 ferent workers have been instrumental in casting doubt on the 

 value of the bacterial examination of milk. It is true that the 

 precision possible in a chemical analysis cannot be obtained in a 

 bacterial examination. However, the differences are usually not 

 large enough to undermine the relative accuracy of the bacterial 

 count. It is of no consequence whether a milk is said to con- 

 tain 5000, 6000, or even 8000 bacteria; such milk is of the best 

 quality as far as the bacterial test goes. Neither is it important 

 to differentiate between a milk containing 1,000,000 and one 

 containing 1,500,000. These differences are irrelevant as long as 

 it is necessary to depend upon bacterial enumeration without 

 distinction of groups or types. When bacteriologic technic shall 

 have evolved means for determining groups and types of bac- 

 teria with reasonable certainty, then a closer relationship of 

 numbers may become of value. 



OTHER BACTERIOLOGIC TESTS 



Milk is frequently examined for streptococci with a view to 

 determining the presence of pathologic strains. Lactic acid strep- 

 tococci appear chiefly in diplococcus form and sometimes in short 



