304 Bulletin 39. 



the herds that failed to pay expenses there may have been some 

 good cows while in the herds showing fair returns there were un- 

 doubtedly many unprofitable ones. In all probability there are 

 few herds reported above from which the profit could not have 

 been increased by selling some members of the herd for beef. One 

 must go farther than to simply determine the gain or loss of his 

 herd ; he must know which particular cows are responsible for 

 gain and which for loss. That this is essential is demonstrated 

 in the following part of the bulletin. 



A DAIRY HERD RECORD. 



For a year the writer kept a record of the individual cows of 

 two herds. The milk was weighed and sampled at every milk- 

 ing and the samples tested twice a month, the writer testing the 

 milk and keeping the record as his share of the work. The own- 

 ers of the herds state that the extra time required to weigh the 

 milk, record the weight and take the samples, did not exceed one 

 minute per cow. The samples tested every two weeks were com- 

 posite samples consisting of a part of the milk from each milking 

 during the two weeks, and were kept in condition for testing 

 by the addition of bichromate of potash and bichloride of mercury 

 in about equal parts. The time required for testing the samples 

 for the two herds was about a half day. The Babcock test was 

 used. 



It was the original idea to secure herds fairly representing 

 the different breeds used for dairy purposes, but the men owning 

 Shorthorn and Holstein herds failed to co-operate when the time 

 came to begin the test. Of the two herds of which records were 

 kept, one consisted of twelve full-blood Jerseys. The other con- 

 sisted of thirty-five cows of mixed breeding ; some high grade 

 Shorthorns, some grade Jerseys and others of various admix- 

 tures of blood. Only thirty cows in the latter herd completed 

 the year's record. 



In the following tables the record is given as of a single 

 herd, except that the figures relating to the cows of the full blood 

 herd are in italics in Table III. The record as it was kept 

 shows the amount of milk and butter fat given by each cow of 



