196 PLANT PRODUCTS 



the general principles that to obtain one pound live 

 weight increase in the weight of oxen, thirteen pounds 

 of dry food material were necessary, whilst about nine 

 pounds of dry food sufficed in the case of sheep, and 

 five pounds in the case of pigs, the foods fed being of 

 a mixed kind common to the diet used in most parts of 

 England. The rate of increase of an animal is, however, 

 much greater in proportion to its food in the early 

 stages of its growth. Some of the early experiments of 

 Lawes and Gilbert on pigs are convenient evidence on this 

 point. In the first month they found that four pounds of 

 food produced an increase of one pound, and in the second 

 month it took five pounds, and in the last month of fattening 

 it took as much as six and a quarter pounds to produce this 

 increase. There is here no resemblance between the 

 objects in fattening and the objects in obtaining work, 

 since a young horse is not capable of putting forth much 

 energy in return for its food, being occupied chiefly in growing. 

 One method of attempting the assessment of foods is to 

 merely take the dry matter, which is an advance on the 

 crude methods commonly adopted. The next advance on 

 that is to deduct the fibre or indigestible matter. A further 

 advance is to utilize the complicated tables given by Kellner, 

 and a further method is to deduce Kellner's starch equivalent 

 or Hanson's milk unit. Another system consists of having 

 standard rations, tabulated for all kinds of stock, giving so 

 much digestible oil and carbohydrates. The latter method 

 has the objection that it requires rather complicated sets 

 of tables, but is perhaps the most comprehensible to the 

 ordinary practical feeder, who finds starch equivalents rather 

 a little beyond him. At the present time the knowledge of 

 feeding is not sufficiently advanced to reduce the question 

 of feeding to a scientific basis, and probably all these systems 

 will remain in vogue. The difference between individual 

 animals is always very great, and individuality must be 

 allowed for, and hence great precision on the theoretical 

 side is not of first-rate importance. In many instances, a 

 study of Kellner 'stables will show what big variations occur, 



