Attempts to produce the syndrome of pellagra in animals by confinement to faulty diets have not been wanting. Chittenden and Underhill (13) fed dogs, exclusively on vegetable foods, and observed that they did not thrive. One diet, consisting of crackers (bolted wheat flour), peas and cotton-seed oil, produced in dogs restricted to it for several months, a condition regarded by these investigators as strikingly suggestive of pellagra in man. The animals developed inflammation of the mouth with sloughing of the mucosa, diarrhea, and skin changes of a nature regarded as analogous to those seen in pellagra.

McCollum and his co-workers had by this time applied the biological method for the analysis of a food-stuff to all the more important cereals and legume seeds. Their results showed that a diet of peas, wheat flour and cotton-seed oil such as Chittenden and Underhill had employed, could be supplemented with purified protein, mineral salts and fat-soluble A, so as to be made adequate for the maintenance of growth and health in the rat. After the appearance of the paper of Chittenden and Underhill, McCollum, Simmonds and Parsons (14) tested this theory by feeding the diet of Chittenden and Underhill supplemented with purified food-stuffs in various ways to young rats. The results justified the prediction that these three types of corrections were all that was necessary to make the diet complete, for the rat. From these results they drew the conclusion that the diet used by Goldberger and his associates could not possibly be deficient in any unidentified dietary essential analogous to the anti-neuritic or anti-scorbutic substances. At that time there was not available as there is to-day a demonstration of the fact that one species of animal may have decidedly different nutritive requirements than another, with respect to at least one of these dietary factors. The classic example of this is seen in the rat which does not require the anti-scorbutic substance, and which is incapable of developing scurvy, whereas the guinea pig is very susceptible to this disease. It no longer seems warranted to assume that because the rat does not suffer injury from confinement to a certain diet, that this is sufficient evidence that the diet is complete for another species. The possibility remains that man may require a substance for protection against pellagra, which the rat is able to synthesize or to dispense with. Further investigations are required to settle this question.

McCollum, Simmonds and Parsons (14) were inclined to interpret their results as supporting the view that pellagra is an infectious disease, and that the role of diet in its etiology involves only increased susceptibility to infection due to lowered resistance caused by faulty diet persisted in during the winter months by many people in the South. Their results supported incontrovertibly the contention of Goldberger, that the type of diet which he fed to the prison squad was incomplete, and would lead to steady loss of vitality, and they made clear the exact nature of the deficiencies of the prison dietary in so far as they relate to the nutrition of the rat.