CONTROL OF THE PUBLIC MILK SUPPLY 443 



the four laboratories have been found, within certain somewhat wide limits, 

 fairly accurate. They are at all events accurate enough to warrant three broad 

 grades, essentially three grades that have been adopted by the commission on 

 milk standards. But they are not as yet accurate enough to warrant a closer 

 grading than the commission's grades, A, B, and C, A including all below 200,000 

 (or 100,000), B from 200,000 to 1,000,000, and C including all above 1,000,000. 

 For this broad grading it is necessary to have an average of at least four or five 

 separate analyses in order to rely upon the results. Even then there will be an 

 occasional overlapping of grade B with either grade C on the one hand or grade 

 A on the other. 



"12. This series of tests has proved that if a sample of milk can be put into 

 iced water, containing floating ice, it may be kept for 20 hr. with very little change 

 in bacteria count. This makes it possible to keep samples sent to a laboratory 

 for analysis for a number of hours without any fear of change in bacterial content, 

 provided the samples are immersed in water containing floating ice. 



"13. These tests have seemed to indicate that the American peptone, made 

 by the Digestive Ferment Co., of Detroit, can be substituted for Witte peptone 

 without materially changing the results. 



" 14. Direct Microscopic Method of Bacterial Examination of Milk by the Breed 

 Method. In making a comparison of the bacteriological analysis by the plate 

 count and the microscopic count, the latter should be a count of groups rather 

 than individuals, plate colonies representing groups only. 



"15. Considerable experience by the person making the count is needed to dis- 

 tinguish between bacteria and dirt particles, particularly when the milk contains 

 minute micrococci. 



"16. When the microscopic count is made by one who has had sufficient 

 experience, the group count agrees somewhat closely with the plate count agree- 

 ing, indeed, about as closely as the plate counts of different laboratories agree 

 with each other. 



"17. Raw, fresh milk does not contain any appreciable number of dead bac- 

 teria which might disclose themselves to the microscope, but fail to grow in plates . 



"18. The direct microscopical examination of milk smears by the Breed 

 method will classify raw milk into grades A, B, and C with about the same ac- 

 curacy and much more quickly than the plate method of bacteriological analysis 

 will do. It is of no use in the study of pasteurized milk, however, since it dis- 

 closes dead as well as living bacteria, no method of distinguishing between them 

 having yet been perfected. It might be of value in telling whether such milk had 

 become old before it was pasteurized, since such would show large numbers of 

 dead bacteria by the microscopic method, though it might show small numbers 

 by the plate method. 



"19. The direct microscopical method of bacteriological analysis might be of 

 exceptional value applied at the dairy to guide the dairyman as to the general 

 grade of the milk he is marketing. It may also be of great aid to the large dealer 

 to enable him to determine promptly whether he is purchasing milk of A, B, or 

 C grade. The possibility of quick results and ease of making the smears at the 

 dairy or shipping station, subsequently sending them to the laboratory for micro- 

 scopic examination, renders the method especially applicable at the dairy end of 

 the line." 



