238 ILLINOIS STATE DAIRYMEN'S ASSOCIATION. 



Different cows differ in their power to make milk from food, and 

 the same cow varies in this respect from time to time. 



It pays to select the individual according to her power to manu- 

 facture milk from food, and according to the character of the product. 



Aside from the influence of food or environment each animal ex- 

 hibits individual variations of her own, and such variations tend to show 

 something like periodicity in the separate functional activities of the 

 animal body. 



HOW THE EXPERIMENT WAS CONDUCTED. 



For a period of ten months, beginning May 1, 1897, the yield of milk, 

 of fat, and of solids not fat, was determined for each of five cows and 

 separately for each milking. Of the cows under experiment Dolly and 

 Janet are high grade Jerseys, Jochemke (called Jock) and Lady Pietertje 

 Veemanr (Called Lady Pietertje) are registered Holstein-Friesians, and 

 Eva is a high grade of the same breed. 



The animals were pastured during the summer, and in fall and win- 

 ter they received the same kinds of food in such amounts as their 

 appetites required. It was in no sense a feeding experiment, and no 

 attempt was made to compare the yield of milk with food consumed. 

 The sole object was to study the daily and periodic variations in the 

 yield and the character of the milk from the same and from different 

 animals kept under conditions as nearly as possible like those in com- 

 mon practice. Throughout the experiment the milking periods were by 

 design unequal, the period from morning until evening being 11 hours 

 and from evening until morning 13 hours. This inequality ir> common 

 on the farm, excepting that the periods are generally reversed in length 

 during the summer season. 



All tests were made in duplicate. Fat was determined by the Bab- 

 cock metiiod with accurately calibrated bottles. The solids not fat 

 were determined with the Quevenne lactometer and the results calcu- 

 lated from Babcock's tables. 



