ration. This particular change in the composition of the feed used, 

 was made for the purpose of sec%iring whenever desired a closer relation 

 between the digestible nitrogenous and 7ion-nitrogenous organic constit- 

 uents ot the food consumed, as compared with that which served in our 

 previous experiments. The gluten meal was chosen for that end from 

 among tlie various concentrated commercial feed-stuffs, on account 

 of its close relation to corn, of which it constitutes a part, and its 

 higher nitrogenous character, when compared with that of the corn 

 meal and the wheat bran (shorts). One of the principal objects of 

 these more recent investigations has been the same as that of the 

 earlier ones, namely : To study the feeding effect of dried corn fodder 

 (stover), as a substitute, whole or in part, for English hay ; and that 

 of corn ensilage as compared with roots under corresponding circum- 

 stances ; for details in this connection see our IV. Annual Report, 

 page 11. 



Three cows, mixed breed, from five to six years old were 

 selected for our work. They were practically in the same milking 

 period, from three to six weeks after calving, at the beginning of the 

 trial. The observation extended over a j^eriod of seven months, 

 October 1st, 1886 to April 24th, 1887. 



The temporary changes in the diet, wherever decided upon, were 

 carried out gradually, as it is customary in all carefully conducted 

 feeding experiments. At least five days are allowed in every inst- 

 ance to pass by, in case of a change in the character of the feed, 

 before the daily observations of the results appear in our published 

 records. The dates which accompany all detailed reports of our 

 feeding experiments, past and present, furnish exact figures in that 

 direction. This is in particular the case whenever such statements 

 are of a special interest for an intelligent appreciation of the final 

 conclusions presented. 



As our feeding experiments with milch cows were originally under- 

 taken with the intention of cai-ryingout a systematic course of inquiry 

 into the economical relations of the production of milk with reference 

 to the dairy industry as well as to a practical general farm manage- 

 ment, it was thought best for various reasons to begin our work with 

 cows of moderate milking qualities. The effect of different diets, on 

 the quantity and quality of the milk produced, as well as their bearing 

 on the-?ie^ cost of production promised to be of particular interest 

 under the stated circumstances. A beginning of our work at the 

 lower end of the scale of the production of milk offered besides, the 



