Agricultural Chemistry — Slieep- Feeding and Mayiure. 283 



variations mainly depended on the amount of the matters of the 

 food retained at the time in the stomach and intestines of the ani- 

 mals, an irreg-ularity which was guarded against as far as seemed 

 practicable without imposing unnatural restraints upon the animal, 

 the plan adopted being, to weigh them always about the same hour 

 of the day, and just before their second meal of dry food, their 

 troughs being, however, constantly supplied with turnips. This 

 variation is indeed a source of error which it is very difficult to 

 control, and it is probable that many of the published results of 

 very rapid increase are subject to objection on account of it. The 

 fluctuation, as would be supposed, seems to occur nearly as promi- 

 nently with those sheep which in the main show a good result as 

 with the rest; and, although there is nevertheless considerable 

 difference, yet there is, excluding the extreme cases of loss or 

 gain of individual sheep, to an extent an uniformity throughout 

 each pen at the several periods, and even between pen and pen, 

 as a view of the columns of weekly average gain or loss in each, 

 placed side by side, will show. 



SERIES I. 



Table 3. — Showing the Weekly Average Gain or Loss per Sheep in 

 each Pen throughout the course of the experiments. 



Thus it is seen that during the 1st, 4th, 6th, 8th, and lOlh 

 periods, there is throughout the jpens a general disposition to more 

 than the average increase, especially at periods 1, 8, and 10, 

 excepting in pen 1, but on reference to the details it will be 

 found that the small average of pen 1 at the two latter periods, de- 

 pended upon a considerable loss of one single animal which was 

 so unwell as to require removal and change of food for a time. 

 Again, at periods 5, 7, 9, and 11, there is throughout the pens a 

 gain much below the average, or even a loss, the cases rather ex- 



