NUTRITION. 325 



less by conducting the steaming process more thoroughly than is the 

 present practice. Since rats have been brought to full maturity from 

 a very early age and have produced vigorous broods of young when 

 this seed furnished the sole source of protein, the high nutritive value 

 of steamed cotton-seed meal is demonstrated. Furthermore, our 

 experiments prove that cotton-seed meal is highly efficient as a 

 supplement to protein-rich by-products from cereal grains. 



Mention was made in our report for 1916 of soy beans as among the 

 liigh-protein food substances with which feeding experiments were 

 being made. In view of the rapid extension of the cultivation of this 

 seed and its increasing use as food for man, as well as for domestic 

 animals, our studies of it have been greatly extended and we have 

 endeavored to learn as much as possible concerning its food value. We 

 have found that (unlike most other leguminous seeds) soy beans fed as 

 the sole source of protein, or as a supplement to corn gluten, promote 

 the normal growth of young rats and maintain adults. They contain 

 sufficient water-soluble vitamine; for diets containing soy-bean flour, 

 butter-fat, starch, and an artificial salt mixture have promoted growth 

 as well as comparable rations containing natural protein-free milk. The 

 presence of some fat-soluble vitamine has been indicated by our long- 

 continued growth experiments, which are still in progress. The mineral 

 constituents of the soy bean are inadequate for growth, but the addi- 

 tion of suitable salts suffices to remedy this defect. Rats fail to eat 

 sufficient food containing untreated soy-bean meal to make normal 

 gi'owth, but when the meal is heated with a little water they readily 

 eat enough to fully satisfy all of their nutritive requirements. Meal 

 heated without the addition of water is usually not eaten any better 

 than the unheated. The raw, or dry-heated, meal is non-toxic; for 

 the few animals which have eaten enough of foods containing such 

 meals have grown well. Unlike cotton seed, raw soy beans extracted 

 with ether are not eaten any more freely than the unextracted. A 

 mixture of unheated soy-bean meal with corn gluten has proved 

 satisfactory as the sole source of protein in the diets of chickens. 



The high nutritive value thus shown by the soy bean, as well as by 

 the proteins isolated from it, is interesting in connection with the 

 results which we have obtained with the conamon kidney bean, Phase- 

 olus vulgaris. Some years ago we showed that phaseolin, the chief 

 protein of this seed, when prepared by extracting kidney beans with 

 sodium chloride and dialyzing the extract, failed to even maintain rats 

 when fed as the source of protein in an otherwise suitable dietary. It 

 was non-toxic, for when it was fed with an equal quantity of edestin 

 the rats made satisfactory growth. We have now found that if the 

 phaseolin is boiled with water for a few minutes and then dried, or dis- 

 solved in alkali and reprecipitated by neutraUzing, it is capable at 

 least of maintaining rats without growth. On the other hand, phase- 



