New Yoke Agricultural Experiment Station. 



59 



It is evident from these comparisons that there is no reasonably 

 constant relation between the germ contents of samples collected 

 in this way; the number of germs found in any particular sample is 

 only a general indication of the germ content actually present in the 

 pail of milk. 



A further study of the distribution of the bacteria in a pail of milk 

 drawn freshly by the milking machine was made as follows: With 

 a sterile long-handled spoon a sample of foam was collected, a second 

 sample of the milk was drawn from approximately the middle of the 

 pail with a sterile pipette and a third sample was taken after 

 thoroughly stirring the entire pail of milk with a long-handled sterile 

 spoon. The germ content of these samples was determined as in 

 the preceding study except that only three plates were prepared from 

 each sample and the dilutions used were 1 to 10, 1 to 20 and 

 1 to 50. The numerical results from 36 sets of samples are given in 

 Table III. 



The average numerical results from these three sets of samples are 

 in no better accord than those recorded in Table I. These averages 

 bear the same relation to each other as 75, 82 and 100, which 

 are practically the same ratios as found in the preceding sets of 

 samples. 



The rankings of the various samples in the above table are sum- 

 marized in Table IV. 



While there is no very striking agreement in the rankings of the 

 samples as shown in this table it will be noted that 50 per ct. of the 

 highest samples were among those taken after a thorough stirring of 

 the pail of milk and 39 per ct. of the medium samples were also of this 

 class. These last samples have the added advantage that they were 

 taken under conditions which would ordinarily be considered as being 



