— 93 — 



as it came from the cow. For three days in each week at first, 

 and afterward in each alternate week, the milk was analyzed. 

 As soon as milked and weighed, the milk of the cows in the same 

 lot was mixed and a sample taken. The evening sample was 

 kept on ice until morning and with the morning sample taken to 

 the laboratory where the two were thoroughly mixed and the total 

 solids and fats determined. 



The cows were selected and the milk weighed and analyzed for 

 two weeks before the feeding began so that data might be secured, 

 as to the normal amount and quality of the milk from the two sets. 



THE RESULTS IN MILK AND BUTTER. 



The time of feeding extended from May 25th to Sept. 27th, or 

 eighteen weeks. This time is very naturally divided by the state 

 of the weather and condition of the pastures into three periods. 

 In the first period, extending from the beginning of the experi- 

 ment to July 1 2th or seven weeks there was an abundant rain fall 

 and the grass grew vigorously, affording a "full bite" all the 

 time. The second period, from July 13 to Aug. 16 or five weeks, 

 was characterized by hot weather and a severe drought, during 

 which the pastures became very dry and bare and were supple- 

 mented with second growth clover. In the third period, from 

 Aug. 17 to the end of the experiment, or six weeks, after copious 

 rains the pastures rapidly recovered their luxuriance. 



In Table II is shown the average daily milk yield of each lot, 

 the average percentage of total solids and fats in the milk of 

 each lot on three days of each alternate week, and the average 

 amount of butter fat produced per cow per week by each lot. 

 This last is found by multiplying the average milk yield for the 

 week by the average per cent, of fat in the milk on three days of 

 the week, except in the weeks when no analysis was made, when 

 the average for the preceding and following weeks was used. 



