New York Agricultural Experiment Station. 69 



nent as cheese-producing materials, or, stated diflereutly, the 

 cheese-produciiig power of milk is almost entirely measured by 

 two of its solid constituents, so far as the composition of the 

 milk is concerned. These two cheese-producing constituents are 

 fat and casein. The other constituents of the milk, such as albu- 

 men, sugar, etc., pass into the whey for the most part and are lost. 

 These two cheese-making constituents of milk vary much in dif- 

 ferent milks. If the amount of cheese made from milk depends 

 upon the amount of fat and casein in milk, why would it not be 

 well to use the fat and casein together as a basis in paying for 

 milk? One serious objection lies in the fact that we have no 

 simple method for determining the amount of casein in milk 

 which is practicable in the hands of anyone but a trained chemist. 

 Milk fat, on the other hand, can be readily determined; but does 

 it alone furnish a satisfactory guide as to the amount of cheese 

 that can be made from milk? It was held that milk-fat cannot 

 be an accurate guide in regard to cheese-yield, because when fat 

 increases in milk, the casein does not increase in anything like 

 the same proportion, or, expressed in another way, milk poor in 

 fat contains more casein for a pound of fat than does milk richer 

 in fat, and will, therefore, make more cheese for each pound of fat 

 than will richer milk. The difference existing was asserted in a 

 broad way, but not in detail, and no evidence was offered. Our 

 former work showed conclusively that, while fat and casein did 

 not preserve an absolutely uniform relation, the relation varied 

 within such limits as not to affect seriously the value of the 

 method of paying for milk on the basis of its fat-content, especi- 

 ally when we take into consideration the influence of fat and 

 casein on the quality of cheese produced, and also the relative 

 market values of fat and casein. 



In this report we propose to consider more in detail than ever 

 before just to what extent the relation of fat and casein in milk 

 varies in the milk of different herds of factory cows. 



We secured separate samples of milk from each of 50 herds 

 every alternate week continuously for six months, obtaining in 

 all 032 samples. 



