THE FEEDING OF MILCH CATTLE 317 



(4) Methods for ascertaining the effect of food on the 

 secretion of milk. 



On account of the powerful influence which the 

 individuality of the animal exercises upon the pro- 

 duction of the milk in general (p. 308), and the 

 effect which change of food also has (p. 309), 

 special care has to be taken in feeding experiments 

 to eliminate individual influences by the use of a 

 large number of animals. Further, the frequent 

 and irregular jumps which the yield of milk makes 

 in course of the period of lactation (p. 310) must 

 also be taken into account and be guarded against 

 in the same way, viz. by taking a number of ex- 

 perimental animals. 



If the quantity and composition of the milk 

 remain constant for some time when the food is 

 unchanged, it would not be difficult to estimate 

 the effect of different kinds and quantities of food 

 upon the milk production. The natural changes, 

 though, which take place during the period of 

 lactation require that in each investigation these 

 must be separately ascertained. To do this there 

 are two ways, which may here be shortly described : 



(a) The period system. 



The use of this method of investigation assumes 

 that the natural changes in the milk of a large 

 number of cows proceed regularly, and that if 



