378 CARNEGIE INSTITUTION OF WASHINGTON. 



NUTRITION. 



Osborne, T. B., and L. B. Mendel, New Haven, Connecticut. Continuation 

 and extension of work on vegetable proteins. (For previous reports see 

 Year Books Nos. 3-13.) 



Since our earlier experience has shown that marked differences exist 

 between some of the proteins which are abundant constituents of many 

 of the important foods of men and animals, we have devoted much 

 attention during the present year to studying the relative nutritive 

 value of certain proteins, both for growth and maintenance. 



In determining the optimum proportion of food protein for either 

 growth or maintenance it has been the usual practice of students of 

 animal nutrition to observe the effect of feeding mixtures of natural 

 foodstuffs containing different percentages of protein. Experiments 

 of this kind have been planned to furnish a sufficient supply of calories 

 in the non-protein part of the ration and reduce the proportion of pro- 

 tein by adding food substances low in protein or by diminishing the 

 proportion of those rich in protein. All of this has involved changes 

 in so many factors with every alteration in the food mixture that the 

 results have value only in respect to the particular mixture used for the 

 animal in question and are therefore to a very large extent empirical. 

 Furthermore, possible differences in the relative efficiency of differ- 

 ent proteins were formerly never considered in making such experi- 

 ments ; and the possibility of effecting economies in the use of the many 

 available foodstuffs by so combining them that any nutritive deficien- 

 cies of the proteins of one food might be supplemented by proper com- 

 binations with another has heretofore generally escaped notice. 



In studying these questions food was supplied ad libitum to rats in 

 order to first learn the quantity of food naturally eaten by them under 

 the conditions of our experiments. The hmits of variation of food 

 intake between different individuals could thus be determined. If 

 such differences were great the food intake, referred to some definite 

 standard, such as body-weight or body-surface, must be taken into 

 consideration before minor differences between proteins of different 

 origin can be detected. 



These experiments showed that this method of study, though ade- 

 quate to indicate grosser inequahties between the proteins from a 

 nutritive standpoint, is not suitable for making more accurate com- 

 parisons. The Umitations of the method he in the fact that there is a 

 wide difference in the amount of food eaten by different individuals, 

 even of the same size and sex. Thus, by eating larger quantities of a 

 ration low in protein an animal may consume a far greater absolute 

 amount of protein than control individuals which consume less food 

 containing the same percentage of protein. For this reason it became 

 necessary to compare the absolute amount of protein eaten by the 



