METABOLISM OF MATTER AND ENERGY. 1007 



the more technical points in making such experiments. In a feeding 

 experiment, as usually conducted, an animal, or in many cases a lot con- 

 sisting of several animals, is fed for a longer or shorter period a given 

 ration of known composition. The amount consumed and the gain or 

 loss in weight are recorded. Often the cost of the food eaten and the 

 value of the animal at the beginning and end of the test are also noted, 

 to furnish the means of judging of the economical value of the ration. 

 A feeding test is generally preceded by a preliminary period of arbi- 

 trary duration during which tlie ration to be tested is fed, in order that 

 the body may adjust itself to the new ration and during the test proper 

 may be actually living upon the ration tested and not upon material 

 stored from the previous ration. When different foods are compared, 

 either similar animals or lots of animals are fed the selected rations at 

 the same time and under similar conditions, or the different rations are 

 fed to the same animal or lots in different periods, separated by suit- 

 able preliminary periods. The two methods are often combined. If 

 the feces are analyzed the data are obtained for determining the diges- 

 tibility of the food, since it is assumed that the feces consist of the 

 undigested residue of the food eaten. It goes without saying that the 

 feeding experiment is more valuable when the comparative digestibility 

 of the foods tested is also learned. The fact that digestion experiments 

 have increased in number in recent years shows that the station work- 

 ers fully appreciate this fact. That the collection and analysis of the 

 urine affords additional data for drawing deductions from feeding tests 

 and permits the deduction of other important facts is not recognized, 

 or if recognized is not carried into practice. 



Determining the nitrogen balance in connection with the feeding 

 experiment is useful in the following ways: 



It may be very helpful as an indication of the proper length of the 

 preliminary feeding period or the time when the feeding experiment 

 proper may begin, since the existence of nitrogen equilibrium implies 

 that so far as nitrogen is concerned the body is living upon the food 

 eaten. 



It is also helpful as showing whether the ration is sufficient for the 

 needs of the animal. If it is insufficient the body will lose nitrogen or 

 carbon or both. In long feeding experiments this loss would almost 

 certainly appear in loss of weight and in poorer condition of the animal. 

 But if the period is short a chauge in weight is not a sufficient indica- 

 tion of the physiological condition, for change in weight is known to be 

 influenced by different water content of the body and other factors. 



The feeding standards commonly followed have been deduced from 

 the best data available at the time they were proposed and in general 

 are in accord with the best practice. In some instances the amount 

 of nitrogen proposed was influenced by the amount consumed from 

 the tissue of a fasting animal as shown by the excretion of nitrogen in 

 the urine, since it was assumed that when living on its own tissue the 



