94 



were finished. This was unavoidable as the necessary selec- 

 tions had to be made from the results obtained in these trials. 

 In the meantime the cows wer^ fed, exactly alike. Each of 

 the Experiments I and J lasted respectively eight weeks. 

 Experiment K was carried on for ten weeks. For purposes of 

 comparison, however, the results are tabulated both for eight 

 and ten weeks. The cows were milked as usual three times a 

 day, and the milk was sampled an$ tested as in all the other 

 experiments. 



A point of interest worth recording is that those cows 

 which received the brewers* grains in their ration had a 

 nicer bloom than those which did not at the time receive any 

 grains. 



At the end of the preliminary trials, the two lots stood as 

 follows : 



Total yield in pints Average yield of inilk Per cent, of butter fat Average 

 per lot per day. per cow per day. in total daily yield. live-weight. 



Lot I. ... 117 23-4 pints. 3'33 10501b. 



Lot II. ... 118 23-6 pints. 3-40 1038*8 Ib. 



Rations. The same plan was followed in the general 

 arrangement of the experiment as in previous experiments, 

 viz., the cows were fed on certain rations for a period, after 

 which the rations were reversed, and the experiment continued 

 for a similar period under the altered conditions. In the 

 present case, the rations used were as follows per 1,000 Ib. 

 live-weight : 



LOT I. 



4 Ib. maize meal. 



2 ,, Bombay cotton cake. 



3 ,, chopped straw. 

 12 hay. 



20 brewers' grains. 

 20 ,, swedes. 



LOT II. 



4 Ib. maize meal. 



2 ,, Bombay cotton cake. 



3 ,, chopped straw. 

 12 hay. 



60 ,, swedes. 



There was practically the same amount of organic matter 

 in each ration, but that given to lot I. was more nitrogenous 

 than the other. The difference, however, was not great, and 

 it probably varied a little from time to time, owing to the 

 variation in the composition of the brewers' grains. 



