118 MANUAL OF CATTLE-FEEDI^^G. 



undertakes to mate feeding experiments should be aware 

 of the exceeding ambiguity which attaches to hJuaU 

 changes of weight. 



Tlxe foregoing explanations make evident how much 

 labor and care are necessary in order to determine, with 

 any certainty, the nutritive value of even a single article 

 of fodder for a single class of animals, and it is no cause 

 for surprise that the theory of feeding can only reach 

 its complete development in all directions slowly. When 

 the question is only of the gain or loss of flesh, the method 

 of experiment, as we have seen, is much simpler and less 

 laborious and demands less expensive apparatus, than when 

 the eftect of the ration as a fat-producer is to be deter- 

 mined ; and it is therefore natural that the laws of " flesh- 

 building" are already very thoroughly explored, while in 

 regard to the circumstances under which tlie greatest and 

 most advantageous production of fat or work is to be ob- 

 tained we are much more in the dark. 



