1112 EXPEEIMENT STATION EECORD. 



the medium ration was compared with the heavy ration. The weights of about two- 

 thirds of the cows increased or decreased as the amount of grain fed was increased 

 or decreased. From a financial standpoint the results for 3 years were unfavoraljle 

 to the high grain ration. For continuous dairy feeding the author favors a medium 

 grain ration, 4 lbs. being considered lower than is advisable with cows of the better 

 class. 



Four cows were used in comparative tests of malt sprouts with (1 ) cotton-seed 

 meal, linseed meal, and bran, and (2) oats and bran. As comiwired with mixture 

 No. 1, malt sprouts decreased the yield of milk 10 per cent, but did not affect the 

 quality of the milk nor the yield in relation to the amount of dry matter consumed. 

 The use of malt sprouts was less economical. As compared with mixture No. 2, 

 malt sprouts decreased the yield of milk 4 percent, the quality remaining unchanged. 

 The dry matter consumed was 9 per cent less, making the yield in proportion to the 

 dry matter 5 per cent greater than when oats were fed. The malt sprouts were not 

 readily eaten by the cows in either experiment. 



The relative feeding value of dried brewers' grains and a mixture of cotton-seed 

 meal and linseed meal was tested with 1 cow. There was no marked difference in 

 the yield and (juality of milk. The use of l)rewers' grains was the more economical. 



(ilnten meals containing, respectively, 29.2 and 36.7 per cent of protein were com- 

 pared with 4 cows, the results showing practically no difference as regards the yield 

 and (juality of milk and the amount of dry matter consumed. 



A third trial of apple pomace was made with 12 cows. As much milk and 3 per 

 cent more butter was produced on apple pomace as on corn silage. The yield per 

 unit of dry matter was 7 per cent greater in the case of the ])omace ration. No 

 unfavorable effects were observed. A ton of apj>le pomace is considered as having 

 practically the same feeding value as a ton of silage. 



Additional data concerning the experimental error involved in feeding trials served 

 to confirm the view already advanced that if 2 or 3 animals are used the experi- 

 mental error inherent in the alternation method may be quite disregarded. 



A comparison of feeding trial methods, J. L. Hills ( Vermont tSta. Rpt. 1902, 

 pp. 318-327). — In earlier experiments (K. S. R., 14, p. 79) tM'o methods of feeding 

 were compared, one in which different rations were fed to the same group of animals 

 during alternate periods, and one in which the animals were divided into 2 groups, 

 one groupbeing fed continously on one ration and the other group alternately on 

 different rations. While the results of the previous work seemed to show but little 

 difference in the two methods, the present trials, made on a more extensive scale, 

 favor the plain alternation system. The work is to be contiiuied in the hojx' of arriv- 

 ing at more definite and satisfactory results. 



Records of the station herd for 1900-1 and for 1901-2, J. L. IIiu.s ( IVr- 

 mo)d Sta. Rpt. 1902, pp. 337-342). — These records are similar in character to those 

 previously reported (E. S. R., 13, p. 586). The average results for the 2 years are 

 respectively as follows: Number of cows 47 and 49, yield of milk 5,814 and 5,166 lbs., 

 fat content of milk 5.07 and 5.03 per cent, yield of butter 344 and 308 lbs., <'ost of 

 food $49. 97 and $49.13, and proceeds from sales of l)utter $94.56 and $S5.3."5. 



A comparison of score card judgments and of milk and butter records, 

 J. L. Hills {Vermont Sta. Rpt. 1902, pp. 342-344). — The station herd and 2 private 

 herds were scored, the results being compared afterwards with the milk and butter 

 records of the cows. The work was done by R. R. Strait, under the author's direc- 

 tions. In all 62 records were used in the comparison. The results seem to show 

 that in the long run cows scoring high were but little if any superior as milk and 

 butter producers. On the other hand, a relationship between lack of conformation 

 to dair}' types and lessened production was considered as jilainly established. 



The effect of feed on the quality of butter, J. L. Hills ( Vermont Sta. Rpt. 1902, 

 pp. 34-5, 340). — In connection with the feeding experiments noted above and in con- 



