MANUAL OF CAITLE-FEEDING, 100 



with eacli other. The same is of course true of the analy- 

 sis of the excrement, which is purposely made after the 

 same method. Eemembering these facts, we comprehend 

 that the determinations of digestibility are likewise only 

 approximations. More accurate results are greatly to be 

 desired, but at present we have no means of obtaining 

 them and must be content with our present methods, 

 which, though confessedly imperfect, have yet been of the 

 greatest service in placing the practice of cattle-feeding on 

 a rational basis. We can understand, then, that the pres- 

 ence of these biliary and other products in the excrement 

 is not so gi*eat a source of inaccuracy as it might at jBrst 

 thought seem, since their quantity is relatively small and 

 is comparatively constant in the same animal, so that tlia 

 results of digestion experiments are fairly comparable. 

 At any rate, we may be sure that, if we base our calcula- 

 tions of the amount of fodder to be given for any particu- 

 lar purpose on results obtained by the above methods, the 

 animals will not get less than the calculated amount of 

 imtrieuts, though they may receive slightly more. 



§ 3. DETEEMIHATI0I7 OF THE IlTXJTEITlVB EFFECT OF A EaTIOH. 



Production of Flesh. — The method of determining the 

 gain or loss of flesh in an animal, which has been already 

 indicated, is b^ed on the well-established fact that the 

 nitrogen of the urine is an accurate measure of the amount 

 of protein decomposed in the body. 



If in a digestion experiment, carried out as described in 

 § 1, the urine of the animal be also accurately collected and 

 measured, and the quantity of nitrogen which it contains 

 determined, we have all the data necessary to determine 

 the gain or loss of flesh. 



From the determinations of the digestibility of the fodder 



