14 N. H. Agr. Experiment Station [Bulletin 291 



insulated cooling tanks used about 40 pounds of ice for every 100 pounds 

 of milk cooled. Many times where non-insulated tanks were found, slow 

 and unsatisfactory cooling took place even though twice as much ice was 

 used per 100 pounds of milk. 



Gargety milk from the cows contributed to 21 per cent of the producers' 

 troubles. The cause of this was definitely traced to misuse of milking ma- 

 chines. It was most pronounced in those herds where a twin milking 

 machine was in use. Also several of the men did not strip the cows out clean 

 after using the machine. A number encountered high counts when drying 

 off the cows and milking only once a day. Others milked too close to fresh- 

 ening or saved the milk too soon after freshening. Not washing the udders 

 clean before milking was the reason given for a series of high counts by two 

 dairymen, who blamed the situation on the hired men. Not only have many 

 of these men corrected the above causes but they have learned the value of 

 using a strip cup and checking the first streams drawn from each cow for 

 signs of garget. In one instance, however, only the largest streams drawn 

 showed such signs. 



Although most of the high counts were due to previously named factors, 

 some of the trouble could be attributed to the lack of human interest. When 

 a new hired man started in, high counts were almost sure to follow. On the 

 other hand, there were dairies where the proprietor was not a Grade A man. 

 He knew how the work should be done, but his nature would not allow him 

 to follow through consistently so as to produce quality milk regularly. 



Only a few traced their high counts to stable conditions. In one instance 

 the stable was so small and cramped that it was practically impossible to 

 keep the cows clean and the milk from being contaminated. 



In previously discussing the principal factors causing high counts and 

 the distribution of the producers according to these factors, a producer was 

 frequently counted under several headings. If these factors are to be evalu- 

 ated on a bacteria count basis for each producer, it is necessary that they 

 be combined. However, the producers with a single factor are kept separate 

 wherever possible. 



In Table 5, the principal factors and combinations are arrayed according 

 to size of bacteria counts in five classes. In all, 1,697 counts ranging from 

 25,000 to 810,000 are allocated. Any one count appearing in Classes II to 

 V inclusive would be high enough to prevent a producer from obtaining first 

 premium during any one pay period. 



The importance of thorough sterilizing to prevent high bacteria counts 

 is again emphasized. Over 38 per cent of the total counts appear under this 

 heading. When combinations are made of sterilizing and other factors, a 

 larger percentage of the counts usually are found in the classes representing 

 higher counts. The highest counts in the series resulted from a combination 

 of lack of sterilizing, gargety milk, and slow cooling. 



Wrong cooling method was the second most important single factor, 

 closely followed by lax methods of hired men and lazy, inconsistent proprie- 

 tors. The latter proved hkely to land counts in any of the five classes and 

 on the whole did not show up as well as the hired help. 



Factors and Premium Losses 



Carrying the analysis of this problem one step farther, what is the cost 

 to Grade A producers in premiums lost for each of the above named factors 

 or combinations? In answering this question the pay periods of each pro- 



