Substitution of Roots for Concentrated Foods. 457 



Such an arrangement admits of the following comparisons in each 

 group: (i) the dry matter in silage with the dry matter in mangels; 

 (2) the dry matter in mangels with the dry matter in grain when silage 

 is used for succulent food in both cases; and (3) the dry matter in mangels 

 with the dry matter in grain, when silage is fed with the smaller grain 

 ration and when mangels are used for the succulent food with the larger 

 grain ration. By carrying the experiment through three full feeding 

 periods, the comparisons are made at different times in the lactation 

 periods, so that the effect is shown in a sufficient number of cases to 

 offset any error due to the difference in the periods of lactation. Also, 

 sufficient comparisons are made so that decreases in the production 

 of individual cows due to their being "off-feed" or being similarly in- 

 fluenced by other conditions, can be safely ignored in the averages. 



The feeding 



The several foods were weighed for each cow at each feeding. The 

 cows were fed twice a day. The foods of ration I were fed as follows 

 each year : One-half the grain the first thing in the morning, and the 

 silage directly after milking. The cows were then turned out into a 

 covered yard for the day, where they had access to fresh water. They 

 were brought into the barn again about 3 p. m., when they received 

 the remainder of the grain ration. The hay was fed after milking in 

 the evening. The cows receiving ration II were treated the same way 

 except that one-half the mangels was fed after milking in the morning 

 and the other half with the grain at 3 p. m. Ration III was fed similarly, 

 one-half the grain and the silage being fed in the morning, as in ration 

 I, the other half of the grain and the mangels at 3 p. m. and the hay 

 after milking in the evening. Two ounces of salt were fed each cow 

 each day in her grain, 4 pounds of salt being mixed each day with the 

 grain for 30 cows. The rations were fed so as to get the most economical 

 production. 



Dry matter in mangels and silage 



In computing the rations for each year, the dry-matter content for 

 mangels and silage was taken from the feeding tables; but since these 

 foods vary so much in content of dry matter, it was thought important 

 to determine the dry matter of each as fed. A sample of silage and one 

 of mangels was taken during each feeding period, which would represent 

 the average quality of the silage and mangels fed during that period. 

 Dry matter determinations were made from these samples, and the 

 average of these determinations for each year was used in computing 

 the general tables which are to follow. By using these computations 



