EDITORIAL. 587 



now at liand, but it unquestionably adds to the value of tlie experi- 

 ments. 



Eacli experiment is divided into a preliminary period, the experiment 

 proper, and a post-experimental i)eriod. 



The preliminary period generally lasts from 3 to 5 weeks. The ani- 

 mals are in each case selected from herds of 100 to 200 milch cows, 

 and are healthy, fresli caws of any age above 2 years. A somewhat 

 larger number of cows than is strictly needed for the experiment is 

 set apart at each estate. These are weighed and their age, time of 

 calving, and milk yield noted. From this number the cows desired 

 for the experiment are selected, either 30 or 40 being generally chosen 

 at each estate. These cows are then j)ii|, on a uniform ration, and the 

 morning's and evening's milk of each cow weighed and tested by the 

 control centrifuge and by Soxhlet's aerometric method. After some 

 days the cows are separated into lots of 10 each, in such a manner 

 that the average age, live weight, number of days in milk, yield of 

 milk and per cent of fat in the milk of the cows in each lot at each 

 estate are so nearly alike that it may be safely assumed that when fed 

 on the same food they will continue to give approximately the same 

 quantity of milk of similar quality for the greater portion of the i^eriod 

 of lactation. The daily weighings and examinations of the milk are 

 now continued for some time as before, and in addition the mixed 

 milk from each lot is analyzed. If the lots thus formed do not prove 

 to ttll the demands made, a second grouping of the cows is made, and 

 the feeding and examinations continued until it is found that the lots 

 formed meet the requirements. During an intermediate period of 10 

 d;iys the animals in the different lots are then gradually accustomed to 

 the food they are to receive in the experiment proper. 



During the experiment proper, one lot of cows is fed as in the pre- 

 liminary and the post-experimental periods, while the rations of the 

 other lots are changed so as to make the former lot intermediate 

 between them. To illustrate: In the fourth and fifth years' experi- 

 ments barley and oats were compared with mixtures of oil cakes. If 4 

 lbs. each of grain and oil cakes were fed on a certaiu estate to lot B 

 (intermediate lot), lot A received 6 lbs. of grain and 2 lbs. of oil cakes, 

 and lot C 2 lbs. of grain and 6 lbs. of oil cakes. The other part of the 

 rations is the same in all cases, except the quantity of straw, this being 

 always fed ad libitum. 



The different dairymen cooperating are left some latitude as regards 

 the quantity of feeding stuffs to be fed their cows, but the relative pro- 

 portions of each and the character of the feeding stuffs or rations to be 

 compared is the same in case of all herds. 



The experiment proper includes from 4 to 10 10 day periods at the 

 different estates and during different years, the duration of the experi- 

 ment depending on a variety of causes, the decrease in milk yields of 

 tlie cows, the supply of feeding stuffs on hand, etc. The experiment 



