114 Report of the Department of Bacteriology of the 



a clean, high-grade milk to secure the highest price, and, second, to 

 enable milk dealers, butter-makers and cheese-makers to select their 

 high-grade milk and separate it from that of inferior quality. 



Apparently the method has a greater usefulness in this way than 

 it has where samples are to be examined the history of which may 

 be unknown. Until the value of this method for the examination 

 of pasteurized milk is determined it will be uncertain as to how 

 much use can be made of it in the examination of samples of this 

 sort where some are pasteurized and others not. 



The improvement of market milk supplies is primarily an econ- 

 omic question which involves the grading of milk and the paying 

 of a better price for that of high grade. 15 The microscopic method 

 of milk examination will aid in bringing this about because it per- 

 mits the establishment of grades involving a bacterial standard more 

 readily than the plate method. There are two grades into which 

 milk can be divided naturally by this method, each being separated 

 from the other by a fairly distinct border line. One includes those 

 samples of milk in which bacteria cannot be seen readily after 

 searching through a few fields of the microscope and which usually 

 give a plate count below 100,000 per cubic centimeter. The other 

 includes those samples in which bacteria can be seen readily in a few 

 fields of the microscope and which ordinarily give a plate count above 

 100,000 per cubic centimeter. There are no other natural points 

 by which more grades can be established and any such grades must 

 be arbitrarily fixed by more extensive investigation and by practical 

 experience. The basis for these statements is found in the results 

 discussed on pages 97 to 104. 



CONCLUSIONS. 



There appears to be a thoroughly well defined relationship existing 

 between the direct microscopic count and the plate count. This 

 is more apparent in long series of examinations than in short series 

 because of wide variations between results secured by the two 

 methods on single samples. The relation between the two counts 

 is so variable in individual samples that it is impossible to establish 



16 Harding, H. A. Publicity and payment based on quality as factors in improv- 

 ing a city milk supply. N. Y. Agr. Exp. Sta., Bui. 337, 1911: Harding, 

 H. A., and Brew, J. D. The financial stimulus in city milk production. N. Y. 

 Agr. Exp. Sta., Bui. 363, 1913. 



