710 BOAED OF AGRICULTURE. 



of butter fat in seven consecutive days. Notwithstanding this she gave 

 11,445 pounds of millc and 389 pounds of butter fat, which made 454 

 pounds of butter. While Checli the year before was officially tested and 

 qualified, making 12.2 pounds of butter fat in seven consecutive days, yet, 

 she gave during that year only 6,812 pounds of milk and made only 215 

 pounds of butter fat, and 251 pounds of butter. In other words, Alfrida 

 failed to enter the Holstein-Friesian Advanced Registry, but produced 

 4,G33 pounds more milk and 174 pounds more butter fat in one year than 

 Check, that entered the Holstein-Friesian Advanced Registry. Moreover, 

 Alfrida's average record for the two j'ears is greater by 3,399 pounds of 

 milk and 18 pounds of butter fat. The point is, does one week's record 

 in a whole year give the true value of a dairy cow? 



It is plain to every thoughtful man that we must be guided by yearly 

 records and place but little, if any, confidence in the weekly test. But we 

 must not stop with one year's record and base our selection upon one's 

 year work. 



While the scales and Babcock test can be of gi'eat service in the 

 selection of our dairy animals, they must, however, be used with judg- 

 ment. Dairy cows have their "off years," and this must be considered 

 when the herd is being culled. If we do not bear this fact in mind, we 

 are apt to sell some of the best cows from our herds. The writer has 

 in mind the cow Sweet Briar, of the Minnesota Experiment Station, that 

 produced for ten years an average of 358.07 pounds of butter a year, 

 while in 1898 she produced only 206.G2 pounds of butter, but in 1899 she 

 made 306.53 pounds, and in 1901 370.53 pounds. If the merits of Sweet 

 Briar had been wholly based on the work she did in 1898 she would have 

 been classed as a very ordinary cow, and perhaps sold. The great value 

 of scales and Babcock test lies in their continued use in the dairy herd 

 and not in one year's test. Good heifers usually come from the best dairy 

 cows, but it sometimes happens that a promising heifer may do very 

 poorly for the first year. In such cases the heifer's individuality, together 

 with her breeding, should be considered before she is sold. The testing 

 of cows should, however, be carried on in every dairy if a systematic 

 selection is to be made. A good cow seldom has two "off years" in suc- 

 cession. 



In closing, let me present a few tables comparing the result of differ- 

 ent cows and herds and also comparing the first and second year's work. 



Table 1, comparing the average performance of the best and the poor- 

 est herds; also giving the average performance of all the herds tested: 



Milk, Fat, Fat, Butter, Da)/K in 



Poundf. Per Cent. Pounds. Ponndx. Milk. 



Best herd 6,444 4.17 269 313 330 



Poorest herd 4,613 3.71 171 199 271 



Average of herds 5,261 4.06 214 249 300 



