318 Agricultural Chemistry— -Sheep -Feeding and Manure. 



the food supplied. It was soon discontinued, however, and oil- 

 cake and swedes, a few mangolds only being intermixed, were 

 given to the animals until they were supposed somewhat to have 

 regained their position, after which mangolds and oil-cake were 

 given alone. 



The collected results of the experiments are given in Table II. 



An account of the progress of those animals which either died 

 from illness, or were killed, is excluded from this Table; and it 

 will be seen that there was one such in pen 1, with dry barley ; 

 one in pen 3, with steeped barley ; one in pen 4, with steeped 

 malt ; and one in pen 6, with the beans. No. 1 sheep, in pen 3, 

 also with dry malt, became unwell during the last few weeks of 

 the experiment, and it was found necessary to kill him a few days 

 after the experiment closed. As, however, his increase was fair 

 at the commencement, and he actually lived to the end of the 

 experiment, his weights are admitted into the calculations. This 

 bad result, as to the health of the animals, is sufficiently general 

 in the different pens to show that the explanation of it cannot be 

 sought in the character of the special foods employed. It is 

 indeed more probable that the mangolds, and perhaps not imma- 

 terially the confinement also, were at fault. 



A glance at the top line of results in the Table will show how very 

 general throughout the pens was the loss of weight during the 

 first 22 days. In the bottom one is given the total gain of each 

 animal, inclusive of this preliminary period. This estimate, how- 

 ever, is an under-statement of the effects of the special foods, since 

 it is affected both by the amount of dirt and moisture lost, and by 

 the depreciation due to the inaptitude of the succulent food em- 

 ployed. The results of the experimental period, on the other 

 hand, being, as has been before observed, more probably in excess, 

 depending chiefly on some few cases of individual irregularity. 

 In some few instances, however, the somewhat excessive rates of 

 gain of single animals, after the commencement of the experimental 

 period, are seen not to have been preceded by a corresponding 

 loss, and in such cases the results of the experimental period are 

 not open to the objections referred to above. 



The average weekly gain of the animals is given in Table III. 



Looking across the columns of the Table, we see that whether 

 we calculate from February 28, as in the first division, or from 

 March 20, as in the second, the rates of increase of different 

 sheep upon the same food are very variable, and so general is the 

 irregularity in all the pens, that it cannot be determined that one 

 food was less subject to it than another. The variations are, 

 however, more prominent on the experimental than upon the 

 longer period — a circumstance already explained. 



So far as the average results can be relied upon as providing a 



