machines. During December each milking- from each individual cow 

 was weighed, and sampled for testing. The composite sample made 

 up of the individual daily samples, was tested at the end of the month, 

 and the pounds of milk fat were obtained by multiplying the pounds 

 of milk given by each cow by her test and dividing by ioo, e.g., cow No. 

 1-5 gave 952 pounds of milk during December. Her composite test was 

 3.4, and 952 multiplied by 3.4 divided by 100 equal 32.36 pounds fat. 



During January each cow's milk was weighed morning and evening 

 and samples were taken for testing every seven days. The pounds of 

 milk and milk fat were calculated from these weights and tests, hence 

 are not so accurate as for December. 



Table Showing Comparison of Y-ields of Milk and Milk Fat for December, 

 1Q05 (hand milking) and January, igoO (machine milking). 



From the preceding table we learn that these fifteen cows gave 1,351 

 pounds less milk in January, 1906, than they did in December, 1905. The 

 percentages of fat were fairly constant, though the tendency was for a 

 slightly higher average test for January (3.97) as compared with De- 

 cember (3.90). The pounds of milk fat, however, decreased 47.22 pounds 

 in January as compared with December. If we allow an increase of one- 

 sixth on the fat for calculating the butter, the decrease in butter on the 



