96 Breeding Plants and Animals. 



year records of food eaten and of product for individual 

 cows in the Minnesota Experiment Station, for example 

 would yield data on the frequency necessary to make 

 tests. Experiments inaugurated by Prof. Carlyle at the 

 Wisconsin station would give valuable facts as to the 

 variation in food requirements of different cows. Mr. 

 Glover's experimental studies of dairy herds in North- 

 ern Illinois would give valuable information as to how 

 to make tests and keep records in the herd. Possibly 

 weighing the milk daily and analyzing for fat once a 

 week, with monthly checks by a disinterested inspector 

 would be a sufficient basis for records of the product. 

 The cost of daily milk weighing and monthly testing 

 for fat would be small. Where the milk is weighed 

 daily hired milkers are more careful to get all the milk, 

 and the owner can more readily discover carelessness 

 in milking, as shown by reduced yields, an4 a better 

 knowledge can be had daily of the health and require- 

 ments of each cow. Weighing the milk often pays 

 merely in securing cleaner milking, thus better keeping 

 up the flow. The feeding experiments would neces- 

 sarily need to be long conducted with uniform food in 

 winter time, and might be better if the cows could be 

 taken to a central test barn. These various tests will 

 not appear so expensive when once it is fully realized 

 that only the best cows and their female progeny would 

 need to be subjected to them. On the other hand, large 

 returns from future sales of breeding stock and superior 

 animals for the dairy herd of the county would war- 

 rant a large expenditure in establishing a county strain 

 or breed with its record based mainly on the solid 

 ground of performance. Herfordshire has reaped prof- 

 its which would have paid dividends on a large invest- 

 ment in working up the blood df the "white-face" breed 

 A* number of other counties in England have done like- 

 wise with their respective breeds <of cattle, sheep, horses 

 and swine. 



To give the records actual value and to have them 

 accepted by the public it would be necessary to i-af- 



