INVESTIGATIONS ON THE METABOLISM OF MILCH COWS. 809 



coal, cork dust, or berry seeds taken with the last meal before and the 

 first meal after the period under investigation. In experiments with 

 Herbivora, owing to the lack of suitable methods of separating the 

 feces it is customary to delay the collection of the excretory products 

 until the food under investigation has been consumed for from 2 to 3 

 weeks, since it may be assumed that digestion and excretion then take 

 place regularly, and that the stomach and intestinal tract of the animal 

 at a given time will contain the same amount of material. It is essen- 

 tial that the animals be fed regularly and given like amounts of food 

 and water each day. In the case of the urine it is assumed that under 

 ordinary circumstances no considerable quantity is retained in the 

 biaddov. 



, I 



r 





■ '':- 





% 



.-.-.-•« 



Fig. 16.— Parts of apparatus for use in digestion experiments with rows. 



A metabolism experiment with milch cows should be conducted as 

 follows : A preliminary period of 3 weeks should precede the experiment 

 proper. The floor of the stall is then thoroughly cleaned. The device 

 described above is attached to the cow and the time noted at which the 

 last urine and feces were excreted. Suppose, for instance, that urine 

 was excreted at 1.20 p. m., March 25, 1898, feces at 2.50, and the appa- 

 ratus was attached at 3 p: m. The urine and feces are then collected 

 for from 8 to 11 days and the time at which the experiment is discon- 

 tinued is noted, together with the time of excretion of the last urine 

 and feces. Suppose this in the case of urine to have been 1.10 p. m. 

 and in the case of feces 2.30. Then the urine has been collected for 10 

 days 3£ hours and the feces for 10 days lacking one-third of an hour. 

 In calculating the daily excretion suitable correction must be made for 

 fractional days. 



