588 EXPERIMENT STATION RECORD. 



proper is followed by an intermediate feeding- of 10 days and a post- 

 experimental i^eriod lasting from 40 to 80 days, introduced to study the 

 residuary effects of tlie rations fed in the preceding period on tlie pro- 

 duction of milk and its components. During this time tlie cows in all 

 3 (or 4) lots again receive the same food that was fed to all the cows 

 during the preliminary period, and to one of the lots throughout the 

 experiment. The milk is weighed and analyzed as during the other 

 periods. 



The animals included in the experiments at each estate are in charge 

 of a feed master or superintendent paid by the experiment station, 

 Avhose only business on the estate is to feed and take care of the cows, 

 and to conduct the experiment according to directions. He weighs the 

 morning and evening milk of etich cow and also, as a check, the mixed 

 milk from each lot. He furthermore samples the milk, makes examina- 

 tions by Fjord's centrifuge, weighs out the rations for each lot and dis- 

 tributes among the cows according to the requirements of each animal. 

 He is superintended by an assistant of the exx)eriment station, who 

 visits the estate at intervals and is present regularly on several days 

 during each 10-day period of the experiment proper to make fat deter- 

 minations according to Soxhlet's method, wlien such are made, and to 

 perforni other work that can not safely be intrusted to the feed master, 

 such as sampling of the mixed milk of the different lots and of feeding 

 stuffs eaten, analyses of which are made in the chemical laboratory of 

 the station. The assistant also sees to it that the exiieriment is being- 

 conducted according to the i)lan laid down. 



Each assistant has the supervision of the experiments at two estates. 

 The series of experiments conducted every year have therefore as a rule 

 required for tlieir management 8 feed masters, 4 assistants, and 4 

 chemists, besides the chief chemist of the station. Tlie data obtained 

 are recorded and tabulated by the station bookkeeper and clerk, and 

 the director of tlie station brings the results and the statement of the 

 experiments before the public, making in all 19 men whose time is 

 largely, and as far as IG men go entirely, taken up with the experiments 

 when these are in progress. 



The following synopses of the earlier cooperative feeding experiments 

 ■will show the main results obtained. 



The first year's experiments made (1887-'88) were a study of the value 

 of adding roots to rations containing a good supply of concentrated 

 feeding stuff's. The exjieriments were conducted on 5 estates, with 14G 

 cows in all. While the average quality of the milk produced remained 

 the same for all lots, the milk yield was increased on the root feeding, 

 viz, nearly 3 lbs. per head per diiy on rutabagas and 1^ lbs. on turnips, 

 feeding 40 lbs. of the former and 2G.4 lbs. of the latter. The average 

 live weights of all lots increased during the experiment, the lots 

 receiving- roots gaining more than the lot receiving no roots. 



