878 EXPERIMENT STATION RECORD. 



worthy touching the quality of the milk given on the Atlas ration. The 6-week 

 period, however, and those yet longer in duration gave satisfactory results. Com- 

 bining the conclusions on both quantity and quality, it would appear that a 4- week 

 feeding period accurately indicated, in the present case, the character of and prob- 

 ably the extent of quantity changes, but that the slight improvement observed in 

 quality was temporary and not to be relied upon as likely to continue under the 

 conditions of prolonged feeding on a single ration. Since this latter effect is but 

 slight and entirely without practical significance, and since the results of longer 

 feeding discount it, one need not lose faith in the essential accuracy of past observa- 

 tions in which t-week periods have been used." 



The relative feeding rallies of 2 rations of equal balance (pp. 162-104). — 

 In this experiment rations consisting of hay and silage, wheat bran, 

 corn meal, and either 1J lbs. each of cotton-seed meal and linseed meal 

 or 4 lbs. of Atlas gluten meal were fed. The aim was to feed as nearly 

 as possible the same amounts of digestible protein in the two rations, 

 but, as a matter of fact, the amounts fed differed by about 5 per cent 

 in favor of the oil-meal ration. The nutritive ratio of the oil-meal 

 ration was 1:6.3 and of the Atlas meal ration 1:7.3. ]Sme cows were 

 used in the experiment. They were alternated on the 2 rations for 5 

 periods, varying from 4 to 7 weeks each in length. Based on the 331 

 days of feeding, the following conclusions are drawn: 



"Four per cent more dry matter eaten in the form of the Atlas ration than was 

 eaten in the form of the cotton-seed-linseed ration produced 1 per cent more milk, but 

 no more solids or fat. The milk made on the cotton-seed-linseed ration was some- 

 what the richer of the two, but the enrichment could hardly be termed one-sided, the 

 ratio of fat to solids not fat narrowing very slightly (Atlas 1:1.71, cotton-seed- 

 linseed 1:1.68). 



"line hundred pounds of dry matter, both in the entire ration and in the grain 

 portion thereof ( experimental portion), proved more efficient in the cotton-seed-linseed 

 than in the Atlas ration, making on the average about 6 per cent more product. 

 This increase is not sufficient to warrant great stress heinglaid upon it, yet, because 

 of the large number of cows used and the almost perfect uniformity of the trend in 

 each individual case, it is thought that the difference lies outside the pale of possi- 

 ble experimental error and that it is mainly a food effect. The dry matter of the 

 rations being of nearly equal digestibility (cotton-seed-linseed 75, Atlas 76). the con- 

 clusions in the last paragraph may properly he applied likewise to digestible dry 

 matter. 



"It has already been pointed out that the excess of protein in the cotton-seed- 

 linseed ration was but 5 per cent. In these cases of essentially equal consumption of 

 protein one ration proved somewhat more effective than another. The ration con- 

 taining the most digestible carbohydrates, the most digestible fat, and the most 

 digestible dry matter gave the smallest returns." 



Experimental error in feeding tests (pp. 165-167). — 



"The question has been raised whether differences obtained in feeding various 

 rations, one against another, are not frequently due to causes other than changing 

 food, even where every precaution is taken to render all other conditions uniform. 

 In order to test this point a repetition of previous experiments made at this station, 

 consisting f the uniform feeding of cows for several months, was carried out with 

 9 cows. Practically no change occurred in quantity or quality of product on a uni- 

 form ration. A pound of total dry matter produced as much milk, total solids and 

 fat, at one time as at another, lactation stages being equalized. It is thought, how- 

 ever, that if but lew animals are used it is unsafe to claim that changes in product 

 of less than 5 per cent are of necessity due to changes in feeding." 



