77 



It is somewhat curious to find that both lots lost weight 

 during the experiment, and very difficult, at first sight, to 

 understand why the cows in lot I. especially should have lost 

 so much. The loss here was due very largely to two cows, 

 Nos. 98 and 115, which decreased in weight during the second 

 five weeks, 63 Ib. and 105 Ib. respectively. There does not 

 appear to have been any special cause for this loss of weight. 

 None of the cows in this lot gained at all in weight during 

 the period of the experiment. In lot II. the loss in live- 

 weight is very much less than in lot I. One cow in this lot, 

 No. 114, actually gained 56 Ib. during the experiment, while 

 two other cows, JN"os. 79 and 112, gained a stone each in the 

 first five weeks, which they more than lost again during the 

 second five weeks. It is obvious from these results that con- 

 centrated food which is not paid for in the milk yield, is not 

 always recovered in the form of live-weight in the case of 

 dairy cows on grass. In neither of the two experiments 

 here described did the cake and meal produce, on the 

 average, any appreciable benefit whether as regards the yield 

 and quality of the milk, or the live-weight of the cows. Profit 

 was, in both cases, out of the question. Indeed, there was a 

 serious loss, financially, which was even greater in Experiment 

 H than in Experiment Gr. 



EXPERIMENT G (1905). 



SET OF 10 COWS ON PASTURE IN TWO LOTS OF 5. 

 PRELIMINARY TRIALS. 



Milk per cow per day. % Fat. 



Lot 1 34 pints ... 3-40 . 



Lot II 33-4 pints ... 3'38 ... 



Results for first 6 weeks, June 21 August 1 



Average live-weight 

 per cow. 



l,1341b. 

 l,1031b. 



