208 Experiments on the Feeding of Sheep. 
Before passing from a consideration of the results given in the 
Summary Table, it may be interesting to remark that the propor- 
tion of water drank to the food consumed was the greatest in 
Pen 2, with the hay and beans — that is to say, where the amount 
of nitrogenous substance consumed was the greatest. This is 
quite consistent with the observations of ourselves and others, 
that, under otherwise equal circumstances, the larger the amount 
<)f the nitrogenous constituents in the food, the greater will be 
the amount of urea passed off in the urine, and that, as has recently 
been shown, the greater the elimination of urea, the greater will 
be the demand of the system for water. Again, there was a larger 
proportion of water drank to actual food consumed in Pen 4, with 
the beans, oil, and hay, than in Pen 3, with barley and hay ; but 
it is worthy of remark, that when the amount of oil is multiplied 
by 2 "5, and so reckoned as starch, and the total amount of food 
assumed to be in that degree greater in Pen 4, the proportion of 
water drank to food consumed, as so estimated, is very nearly 
the same in the two cases. 
Whilst referring to the subject of the amount of water drank, 
attention may be directed to the connexion between the food and 
water taken, and the temperature and its changes, as indicated 
by the records given in Table V., which relate to weekly periods, 
commencing April 13, and ending August 23. 
The result indicated is, that there was in every pen a general 
tendency to an increased consumption of food in proportion to a 
given weight of the animal towards the middle of the period, 
and then towards the conclusion a diminution, which was the 
more marked the better the food and the greater the progress of 
the animals. On the other hand, there was a diminution in the 
proportion of water taken towards the middle, and then a slight 
increase towards the end of the period. It is clear, therefore, 
that the amount of water taken had not an undeviating relation 
to the amount of food. 
Nor had either the amount of food, or the amount of water, so 
direct a connexion as might have been anticipated with mere 
height of temperature, so far as this can be judged of by the 
readings of a non-registering thermometer at the fixed hours 
specified. They appear to have been far more influenced by 
changes, as indicated by the range of temperature at the respec- 
tive hours during each weekly period, than by the condition of 
atmosphere as marked by the average actual temperature of the 
periods. 
The very small quantity of water taken in every pen during 
the week from May 17-24, is not to be attributed to conditions 
of atmosphere alone, for, although the range of temperature at 
the specified hours of the day (6 A.M., 12 Ji., and 5 P.M.) was 
