9 



the four cows, showed no change in either volatile acids or iodine 

 number which could not be attributed to the advance of lactation. 

 Grass did not, in general, affect the composition of the butter, which 

 fact corresponds with recent German investigations. 



THE SECOND SERIES 



of experiments was begun on January 2d, 1892, with seven cows, 

 and various kinds of hay or dry fodder were compared. The cows 

 were arranged in four sets or groups as follows : 



„ T ( Duchess. 



Group I. j Du£ 



Shorthorn. 

 Shorthorn. 



Jersey. 

 Jersey. 



Holstein. 



Ayrshire. 

 Avrshire. 



ichess, 2d. 



n TI { Pilot's Lily. 



Group II. j prjncess L > eto 



Group III. Maramee. 



„ T .,r ( Chinchilla, Tr. 



Group IV. j Frost>Jr >■> 



The rations, as before remarked, contained the same kinds and 

 quantities of grain, but varied in the amount and variety of hay. 

 They were calculated from the table in Bulletin No. 4, of this sta- 

 tion, and are given in the following table, where each ration is des- 

 ignated by a number at the top of the column : 



TABLE OF RATIONS. 



Pounds of Food per 1,000 lbs. live weight. 



