454 PRINCIPLES OF AGRICULTURAL CHEMISTRY 



same amount of nitrogen was eliminated as in the first period. 

 When the same amount of protein was being decomposed, it was 

 found that the animal was doing more work. That is, the starch 

 was used for production of work. A similar experiment showed 

 that oil enabled the animal to do more work. An experiment 

 showing that proteids may be used in work was as follows : 

 A horse was fed a ration rich in protein, and put on light work 

 for about ten days. The work was then trebled, being a hard 

 day's work. The excretion of nitrogen immediately increased 45 

 grams per day, and the live weight of the animal gradually 

 decreased. 



Exercise may cause muscular growth. For example, Atwater 

 and Benedict found that a man at rest lost 0.7 gram nitro- 

 gen per day and 7.8 grams fat; but when working, he gained i.i 

 gram nitrogen and lost 48.4 grams fat on the same ration. 



Respiratory Methods. The following is a method 1 for deter- 

 mining the consumption of energy during various kinds of work, 

 which has also been of service in other studies, such as ascertain- 

 ing the amount of energy involved in chewing. The animal to 

 be studied is subjected to a surgical operation, and a tube inserted 

 into its windpipe, so that while air may be inspired freely, the 

 expired air passes through a rubber tube into a suitable vessel for 

 collection. The expired air is measured, and the quantity of car- 

 bon dioxide and oxygen in it determined. The quantity of nitro- 

 gen eliminated in the urine is also determined, and shows how 

 much proteids have been oxidized. When fats are burned, for 

 every i cc. of oxygen which disappears, 0.707 cc. carbon dioxide 

 is formed. With carbohydrates, I cc. oxygen is replaced by i 

 cc. carbon dioxide. Hence the ratio of carbon dioxide to oxygen 

 (corrected for proteids consumed) allows us to calculate the 

 relative proportions of fats and carbohydrates oxidized. This 

 method cannot be considered as highly accurate. By means of 

 it, the expenditure of energy caused by walking or running on a 

 smooth slope, going up-hill, drawing a load, etc., have been 

 studied. 



1 Hagemann, Exp. Sta. Record 10, p. 813. 



