306 Pellagra-Producing Diets. IV 
As yet very little is known regarding the supplementary rela- 
tionships of the proteins of one seed for those of another. One 
of us (4) has recorded observations on swine which were fed 
mixtures of two of the cereal grains, nitrogen balances being 
kept on the animals during a period of growth. These records 
did not show any marked increase in the value of these protein 
mixtures for growth as compared with the proteins of the indi- 
vidual grains fed singly. Since at that time the necessity of 
adding certain salts and a growth-promoting fat was not appre- 
ciated, the data obtained cannot be now regarded as satisfactory, 
and the subject needs further study. 
We have in former papers emphasized the principle that it is 
possible to assign to a food factor a biological value only when 
the values of all the*other factors in the diet are known and 
remain constant. Thus an animal may be able to groW at the 
normal rate on a diet, the proteins of which are of such a quality 
as to scarcely make possible the retention of the necessary amount 
of nitrogen to support normal growth, if the inorganic content 
and the two unidentified factors are very satisfactory. On the 
same food mixture so modified as to make the mineral content 
less satisfactory, the same protein supply will fail to nourish the 
animal properly. Distinct benefit can be shown with such ra- 
tions to follow the improvement of the protein moiety of the diet, 
when the salt content remains less satisfactory than the optimum. 
In other words, if one takes an ideal diet as 100 per cent, one 
factor in that diet can be lowered on the scale to 60 per cent and 
the animal may make a nearly normal performance, but if two 
factors are lowered to the value of 70 per cent each, the animal 
might make a three-fourths normal growth, but if one factor 
were lowered to 50 and another to 60, signs of malnutrition soon 
become apparent. As was pointed out in a previous paper (1), 
one cannot say what is an adequate supply of a dietary factor 
unless the biological value of each of the-others is known. 
In the past because of the absence of adequate knowledge con- 
cerning the number of factors which are essential in the diet and 
great paucity of data regarding the differences in the quality of the 
proteins from different sources, students of dietetics have placed 
great reliance on the energy content of the food mixture as shown 
by calorimetric studies and the protein content as shown by 
