I04 STORRS AGRICULTURAL EXPERIMENT STATION. 



pounds of milk was less in the second test for tzvo herds and zvas 

 ■more for the other two herds. The average food cost of one pound 

 of butter was less in the cases of three herds afid more in the case 

 of one herd hi the second tests. 



When the food cost of producing milk and butter is compared by 

 grojips of cozvs fed different ainotmts of protein in the first and 

 second tests, it appears that the cost of produci^ig 07ie pound of 

 butter was less in nearly all grotips in the second test than in the 

 first, when the amount of protein in the second test was not more 

 than about 2.6 pounds per day. The gains in yields and profit 

 in the second tests seem to be most marked when the results from 

 the use of from i.g to 2.4. potinds of protein are co^npared with 

 those fro7n the use of smaller quantities . 



The economy of feeding according to the yields of milk or b^dtet 

 fat depends much upon what yields are taken as a basis. In 

 these experiments, where about 2 pounds of protein was used as a 

 basal ration for a yield of .5 of a pound of butter fat daily, and 

 the protein was increased from .20 to .2^ of a po2tnd for an in- 

 crease of from .75 to .20 of a p07ind of bictterfat, the food cost of 

 production was generally less, tcntil the protein reached about 2.6 

 pounds, than where iiniform ratio7is with smaller amoiaits of 

 protein were used. The munber of experiments is too small and 

 the variatio7is in the amotaits of protehi fed are too limited to be 

 used as a basis for judging the amounts of p7^otein needed for a 

 given yield of butter fat; but the indications are that the amount 

 of protein fed should be between i.p and 2.^ pounds, varying ac- 

 cording to the productiveness of the cow. 



