The statement is sometimes made by physiologists that the instinct furnishes the best guide to an adequate ration. The experience of diverse species is cited as an illustration of what is believed to be some fundamental principle. How erroneous this may be has lately been pointed out by McCollum1 and his collaborators at the University of Wisconsin. A group of animals which have shown themselves singularly suited to the study of problems in nutrition were offered an exceptionally liberal variety of vegetable foods. They thus had the opportunity to select a varied and well appointed ration from the following plant products always offered to them: corn, wheat, oats, barley or rye kernel, polished rice, peas, beans, cottonseed meal, cottonseed flour, alfalfa, corn germ, wheat germ, flaxseed oil meal, peanuts, onions, cabbage, corn gluten, or wheat gluten. The result of this experiment was a failure to obtain adequate nutrition.