CUTTER. 99 



the other hand, cottonseed meal was known to make a hard but- 

 ter, and when fed in moderate quantities with corn meal and 

 bran, produced an article of the best quality. Consequently, 

 cottonseed meal became a popular feed as a source of protein, 

 almost to the exclusion of the gluten meals among farmers who 

 were feeding for butter production. But in the fall and early 

 winter of 1896 and 1897 the gluten meals were offered in our 

 markets at very low prices furnishing, thereby, protein more 

 cheaply than any other concentrated feed. For this reason it was 

 desirable that farmers use these feeds to the largest possible 

 extent. In view of this fact it was considered advisable to inves- 

 tigate the cause of the unfavorable action of these products and 

 determine, if possible, a way to eliminate it either through the 

 method of manufacturing or feeding them. 



A comparison of the composition of the gluten meals with corn 

 meal showed that the chief variations were in the protein and fat 

 content. The percentages of these substances being much 

 higher in the gluten than in the corn meal, this fact, together with 

 the work done by Morseff in feeding oil to cows, led us to 

 believe that the oil of the gluten products was the disturbing ele- 

 ment. Accordingly some feeding experiments were begun in 

 the winter of 1896-7 to test the following points, viz. : If the 

 high oil content of the glutens caused them to make soft butters. 

 and if so, to what extent they should be freed from it to remedy 

 the difficulty. 



EXPERIMENT I. 



In this experiment, the feeding trials were divided into periods 

 of two weeks each, which is a shorter time for feeding tests than 

 is generally desirable, but as the only object in this experiment 

 was to test the effect of the feed on the composition of the butter 

 fat, the length of time employed was considered sufficient. 



At this time there was offered in the market two kinds of 

 gluten meals, one being rich in fat, containing from 15% to 20% 

 and another containing from 7% to 10% fat. These meals 

 varied sufficiently in their fat content to be desirable for use in 

 this experiment, especially as the one richer in fat was poorer in 

 protein, making it necessary to greatly increase the fat in one 



ttX. H. station Bulletin 16. 

 7 



