METABOLISM OF MATTER AND ENERGY. 1009 



To calculate accurately the gain or loss of fat the balance of income 

 and outgo of carbon must also be known. The measurement of this 

 factor necessitates the use of a respiration apparatus. This will be 

 spoken of later. However, a method l for calculating the carbon bal- 

 ance from data furnished by experiments in which the food, urine, and 

 feces have been analyzed has been devised and seems applicable in 

 feeding experiments and experiments on the production of milk. 



The length of time a feeding experiment should be continued is 

 shown, in part at least, by determining the nitrogen balance. In the 

 ordinary feeding experiment the proper length of time is largely a mat- 

 ter of conjecture or convenience, the object being to continue the test 

 until the effect of the ration is manifested with certainty. When the 

 nitrogen balance is determined and nitrogen equilibrium is reached 

 and maintained the subject is known to be living upon the tested ration, 

 and continuing the period for a longer time is a matter to be determined 

 by the special question under investigation. For instance, in feeding 

 a certain ration to pigs it is often desirable to learn how long the ration 

 may be profitably fed rather than to study the physiological effect of 

 the ration. 



Disturbing conditions will be very readily shown by variations 

 in the nitrogen excretion. If the ration is unvarying and nitrogen 

 equilibrium has been reached the excretion of nitrogen in the urine 

 should be practically uniform from day to day. If this is not the case 

 it is certain that there is a disturbing factor. This would often be 

 undetected in the ordinary feeding experiment, and the fact that there 

 is no certain way of recognizing the presence of disturbing factors 

 when the ration is uniform may account for the fact that contradictory 

 results are so often obtained. In other words, the investigator is not 

 certain that the experimental conditions remain uniform or that they 

 are similar in different experiments. Reaching and maintaining nitro- 

 gen equilibrium is an indication that the experiment is progressing 

 regularly, with the experimental conditions under control. This is 

 undoubtedly one of the most important applications of this line of work 

 to feeding experiments. 



It is often desirable in the feeding experiment to test some other 

 conditions than rations; for instance, the effect of housing in warm 

 barns. If the special condition studied is sufficient to produce any 

 marked physiological effect it is very probable that this would be shown 

 by variations in the nitrogen balance; that is, a change in the physi- 

 ological condition would produce a change in the cleavage of protein 

 in the body, and hence in the nitrogen excretion in the urine. Thus, 

 if it should appear in the feeding experiment that drinking large 

 quantities of water increased the excretion of nitrogen in the urine — 

 that is, the cleavage of protein in the body — the deduction would be 



1 Landw. Jahrb., 24 (1895), p. 283 (E. S. R., 7, p. 237) ; 26 (1897), p. 555 (E. S. R., 9, 

 p. 788). 



