242 THE FEEDING EXPERIMENTS 



practical side Avas thus prominent : they were trying to give a 

 scientific basis to the work of the grazier by ascertaining to 

 what source the increased weight of an animal was due, and 

 how it might be produced most rapidly and economically. 



The first set of feeding experiments at Rothamsted dealt 

 with the relation between food consumed and live weight 

 increase produced. Selected pens of the various animals were 

 fed upon specified rations of different foods, one of which was 

 always fed ad libitum, so that the exact composition of the 

 ultimate ration was determined by the animal itself. The 

 nitrogen and dry matter in the food was determined, and the 

 weight of manure produced both in a fresh and dry condition 

 was ascertained. In all, about 600 sheep w^ere employed in 

 the experiments, 160 pigs, and 200 oxen, many of the latter 

 being fattened on the Duke of Bedford's farm at Wobm-n. 



The experiments with sheep came first, and tended to shoA^- 

 that the prevaihng impression of the special importance to be 

 attached to the nitrogenous constituents of food was not 

 correct, but that it was rather the supply of non-nitrogenous- 

 food which regulated both the amount of food consumed 

 by a given live weight in a given time and also the 

 increase in five weight produced. Of course, at that time 

 it was not possible to distinguish between the digestible and 

 the indigestible portions of the food, nor was any attempt 

 made to estimate in the different foods the relative proportions 

 of albuminoid nitrogen and of such nitrogen compounds as the 

 amides, etc., which are abundant in roots, but of whose feeding- 

 value nothing was known. That the exactitude of their 

 experiments was Umited by these considerations, was pointed 

 out by Lawes and Gilbert ; at the same time, the corrections 

 necessary would not invalidate the soundness of the general 

 conclusions they drew. The experiments on pigs indicated 

 still more clearly that the carbohydrates were chiefly concerned 

 in both the maintenance and in the increase in five weight of 

 the pigs, also that very great variations in the amount of 

 albuminoids consumed were without much effect on the result. 



