DAIRYING DAIRY FARMING. 877 



Studies of methods of experimental feeding trials, J. L. Hills 



( Vermont Sta. Bpt. 1896-' l 97 i pp. 134-169, 193-217).— This includes experi- 

 ments on the proper length of feeding periods, the relative feeding value 

 of rations with like nutritive ratios, the experimental error in feeding 

 trials, and a comparison of Atlas gluten meal with other grain feeds. 

 Full data are given in these experiments in the body of the article and 

 in an appendix, and the results are summarized and discussed. 



The proper length of feeding periods (pp. 146-161). — "So far as the 

 writer can discover by thorough search and through consultation with 

 others, no tests have been made showing what is and what is not a 

 reliable period length. The short period of 1 or 2 weeks is clearly 

 to be condemned as likely to furnish erroneous results, since the animal 

 can hardly be said to have begun to feel the effects of the change in 

 ration when the period is over. While such results are surely untrust- 

 worthy, it is not yet clear whether 3 or 4 weeks or more are necessary 

 to furnish reliable data." 



The experiments were made with 15 cows, divided into 5 lots of 3 each. 

 Two different rations were fed containing corn meal and wheat bran 

 with or without Atlas gluten meal. Lot 4 was alternated on these 2 

 rations for 6 periods of 4 weeks each, lot 5 for 5 periods of 5 weeks, lot 

 for 4 periods of 6 weeks, and lots 7 and 8 for 3 periods of 7 and 8 weeks, 

 respectively. Owing to a misunderstanding lot 5 had to be reduced to 

 periods of 4 weeks. The periods were divided into preliminary and 

 experimental portions, the preliminary part of the period ranging from 

 10 days in the 4-week periods to 20 days in the 8- week periods. The 

 nutritive ratio of the Atlas gluten meal ration averaged 1:6.6, and that 

 of the other ration 1 : 10.2. 



"It seems safe to assert that in these trials results essentially similar in character 

 and extent were obtained when the Atlas ration was led in periods from li 1<> S weeks 

 long as had been observed when it was fed for hut 4 weeks. In other words, in this 

 particular comparison, which it will he recollected is between a 'medium' and a 

 'wide' ration, feeding periods 4 weeks long gave trustworthy results as regards 

 quantity changes. This deduction applies to the yield of milk alone and not to its 

 quality. The effect of the Atlas ration upon the quality of the milk is considered 

 under the next heading." 



As to the effect of food on the quality of milk, an extensive resume 

 is given of experiments conducted in this country and abroad upon this 

 subject, and the results in this particular experiment are summarized. 



"When the Atlas ration of medium balance replaced the wide corn and bran ration 

 for a short time (4 weeks or less) a somewhat richer milk was made, one in which 

 the fat was increased and the solids not fat remained essentially unaffected. This 

 effect was not observed, however, when the ration was fed 6 weeks or more, the 

 quality remaining unaltered, or, if anything, growing a shade poorer. Nine cows 

 fed in periods 4 weeks long testify that the quality of milk is slightly bettered, and 

 the ratio of fat to solids not fat narrowed by the substitution of Atlas gluten meal 

 for corn and bran in the ration. Eight other cows fed in periods from 6 to S weeks 

 long testify to the contrary. . . . The upshot of these series of comparisons 

 appears to be that the 4-week period did not give results which are entirely trust- 



