278 SCIENTIFIC FEEDING OF ANIMALS 



light, medium, and heavy work, given in Table III 

 of the Appendix, are based, and these standards, 

 it has been proved, are covered by the food given 

 in many establishments where large numbers of 

 horses are kept. Naturally, what in actual prac- 

 tice is termed medium or heavy work differs very 

 considerably. As a rule, those rations which have 

 been found to give good results are continued, 

 and it is only, for example, when the price of oats 

 is very high, or when other reasons make it impera- 

 tive to introduce some other feeding-stuff into the 

 ration, that a recalculation is made. In such cases 

 Table I of the Appendix gives all the necessary 

 data for the calculation. In this table the estima- 

 tions of the starch equivalents are calculated for 

 ruminants, but it has been shown that they can 

 serve equally well for horses. It has been found 

 by means of eighteen separate experiments with 

 different foods how much work the starch equiva- 

 lent of the various rations was able to perform, and 

 the results have agreed very closely with the theory. 

 The horses in these tests had to turn a braked 

 capstan, and the work was gradually increased 

 until the animal began to lose weight. If ioo be 

 taken as the observed maximum of work, it was 

 found that the figures obtained by calculation 

 from the starch equivalents in the several cases 

 varied between 96-5 and 103-1, the average of 

 which is 99-8. It is thus possible to use, until 



