MAINTENANCE REQUIREMENTS OF MATTER 325 



the non-protein has some value for maintaining the protein 

 tissues of the body. 



It appears to have been demonstrated by recent experimental 

 results, especially by those of Kellner, Morgen, and the Labora- 

 tory for Agricultural Research in Copenhagen, that the non- 

 protein of ordinary feeding stuffs is available for the maintenance 

 of ruminants, probably indirectly through a conversion to 

 protein by means of micro-organisms in the digestive tract 

 (141). On the other hand, investigations have thus far failed 

 to demonstrate that non-protein has any material value for 

 other species or for production purposes (786-789). In the 

 computation of rations for productive feeding, therefore, it 

 appears desirable for the present to consider ordinarily only 

 the digestible true protein, ignoring the non-protein. This 

 implies, however, that the results of experiments upon the 

 protein requirement shall be expressed in the same manner. 



This will have two effects : First, it will make the protein require- 

 ment appear smaller than it really is. Suppose, for example, that 

 a series of trials in which the ratio of digestible non-protein to 

 digestible protein is i : 10 shows that nitrogen equilibrium is reached 

 with a ration supplying 500 grams protein and 50 grams non- 

 protein. Regarding the true protein only, the maintenance require- 

 ment is 500 grams, while including the non-protein it is 550 grams. 



In the second place, however, this error will be largely compen- 

 sated for when the actual computation of rations is also based on the 

 true protein. Thus in the case just supposed, if a maintenance ration 

 be computed from any feed or mixture in which the ratio of non- 

 protein to protein is the same as in the experiments from which the 

 maintenance was deduced, viz., i : 10, it is obvious that the same final 

 result will be reached whether the maintenance requirement be con- 

 sidered to be 500 grams of true protein or 550 grams of crude protein. 

 Only when the proportion of non-protein to true protein varies widely 

 from that existing in the rations used in determining the protein re- 

 quirement will any significant error arise in computing rations. 



In the results considered on succeeding pages, both the crude 

 protein and true protein of the rations are stated when these 

 are given in the reports of the experiments. 



414. Computation to unit weight. It was shown in Chap- 

 ter VII (345) that the energy requirement for maintenance is 

 substantially proportional to the body surface of the animal. 



