stock, and its feeding value to the animal when eaten. On 

 these points there is little reliable information, owing to the 

 difficulty of such inquiries. At the close of the Pumpherston 

 experiments (some of the results of which have just been 

 given), the whole of the area of experiment was grazed ; and 

 several times during the season the extent to which the 

 different plots were fed down was estimated. The following 

 were the averages of these estimates ; 100 representing that 

 the crop was fed down as close as possible, 50 that 50 per 

 cent, had been eaten, and so on : 



TABLE XLV. 



Unmanured 45 p c. 



Superphosphate 65 



Potash salts 78 



Sulphate of ammonia, phosphates, and potash . . 47 



Nitrate of soda, phosphates, and potash .... 27 



The stock evidently preferred the crops grown without 

 nitrogen ; but from a farmer's point of view, the small 

 bulk usually obtained from the use of mineral manures 

 alone makes such dressings undesirable. The difference 

 between sulphate and nitrate is considerable, and, taken 

 in conjunction with the yields from these plots (nitrate 

 35-4 cwt. of hay, sulphate 32-8 cwt.), shows a decided 

 superiority in the effect of sulphate of ammonia. 



One other experiment bearing on this point may be 

 mentioned, though it has not been carried far enough yet 

 to give very reliable results. On the Northumberland 

 County Council Farm, plots of poor grass land manured 

 in various ways were grazed in 1897 by sheep, and the 

 increase in live weight noted on each plot. In the four 

 months of experiment, the following live weight increase 

 was obtained per sheep and per acre : 



TABLE XLVI. 



LIVE WEIGHT INCREASE. 



This is an experiment which obviously needs to be re- 

 peated frequently before it would be safe to generalize 



