New York Agricultural Experiment Station. 103 



during the first five days were separated from those during the 

 remainder of the first month. These first samples were within 

 the colostrum period of flow and a majority of them w r ere obtained 

 at the first milking. The relatively high average of these samples 

 is characteristic of practically all of them and it seems probable 

 that under average conditions the germ content of the colostrum 

 is higher than that of the milk at any later period of lactation. 



The high average for the twelfth month, on the other hand, is 

 due to high counts in two of the samples coming from a single 

 cow so that the mathematical average of all of the counts is de- 

 ceptive. The small numbers found in six of the eight samples 

 during the twelfth month, taken in connection with the results 

 from the two preceding months, suggest that toward the close of 

 the lactation period there is a slight tendency for the germ con- 

 tent of the udder to fall. 



It should be observed that these deductions have been ob- 

 tained by massing together observations without regard to any 

 peculiarities of the individuals from which the samples had been 

 obtained. Deductions from such massed data can well be tested 

 by the results obtained from individual cows. As has been al- 

 ready explained, successive samples were taken over a consider- 

 able period from seven cows and a smaller number of samples 

 from 11 others. The results from these individual sets of samples 

 are in almost complete accord with the massed data. The colos- 

 trum contains rather high numbers of bacteria, the numbers vary- 

 ing greatly with the different animals. With the secretion of 

 normal milk the numbers fall noticeably and remain reasonably 

 constant during the remainder of the lactation period, usually de- 

 creasing slightly toward its close. It should be remembered, how- 

 ever, that even under the most favorable conditions only a single 

 set of three samples was taken once per month and more frequent 

 sampling might disclose minor variations. 



It should be further remembered that these statements are in- 

 tended to apply only to the germ content under normal conditions. 



