4o8 Journal of Agricultural Research voi. iv, No. s 



tive value of the hay and prickly-pear and the effect of adding prickly- 

 pear to a ration of dry material. 



The groups to be compared were balanced as nearly as possible with 

 reference to body weights and yields of milk. In order to do this, 

 records of the milk and fat were kept for a period of lo days before 

 the cows were divided into groups. The 8o-day experimental periods 

 were divided into subperiods of lo days each, and lo days were allowed 

 for making the changes in rations at the end of the first 80 days. 



The grain mixture consisted of equal parts by weight of com chop, 

 wheat bran, and cottonseed meal; the hay was sorghum hay of average 

 quality, and the prickly-pear was a very spiny two-years' growth. The 

 prickly-pear was singed in the field with a gasoline torch, then cut off 

 near the ground, and hauled to the cow lot. After being placed in the 

 mangers, the heavier stems were chopped into small pieces with a sharp 

 spade or hoe. The grain was fed according to the amount of milk fat 

 produced, about 10 pounds being supplied for each pound of fat. 

 The milk fat rather than the milk itself, the milk solids, or the energy 

 value of the milk was taken as the basis of feeding, as previous investiga- 

 tions (2) have shown it to be a more reliable guide for this purpose 

 than the whole milk, and perhaps just as reliable as the milk solids 

 or energy value. 



As much prickly-pear was fed to the cows of the heavy ration prickly- 

 pear group as they would eat; this varied with the different individuals 

 from 100 to 150 pounds a day. Each cow in the two groups receiving 

 the medium prickly-pear ration received 60 pounds a day, and as 

 much hay was fed to the cows of the four groups as they would consume 

 without undue waste. The quantity varied from 3.5 to 20 pounds, 

 depending upon the individual and the amount of prickly-pear in the 

 ration. In addition to these four groups, one cow received all the 

 prickly-pear she would eat and, in addition, 4 pounds of cottonseed 

 meal daily. The object in this case was to ascertain the physiological 

 effects of feeding large quantities of prickly-pear for a long period. 



The body weights of all the cows were controlled by reducing or 

 increasing the roughage so that the gains or losses in weight of the 

 groups to be compared were approximately the same. By conducting 

 the experiment in this manner many of the variable factors were con- 

 trolled and a direct comparison of the amounts of prickly-pear and 

 hay required to produce a pound of milk fat was made possible. 



The body weight of each cow was taken every morning at about 

 the same hour, so that the conditions from day to day as regards fill 

 both of feed and water were as nearly uniform as possible. 



The milk was weighed at each milking, and composite sam.ples for 

 fat analysis and specific-gravity determinations were taken for 5 days 

 near the middle of each lo-day period, the samples being preserved 

 with formaldehyde. The milk solids were estimated from the fat and 



