The administration of the liver and thymus preparations to pellagrins was followed by an improvement in their condition apparently comparable to that produced by the consumption of a diet rich in fresh animal proteins.

The investigators conclude that the evidence presented clearly indicates that the dietary defect responsible for pellagra is distinctly (qualitatively) different from and perhaps more complex than the one causing fowl polyneuritis and human beriberi, a conclusion in accord with that reached by Clementi from his observations on fowls.52

It might be seen that the experiment of Chick and Hume referred to above (p. 254), in which the symptoms of pellagra were produced in monkeys, would eliminate the possibility of any connection between the known vitamins and this disease, since their ration included butter-fat as a source of A, yeast as a source of B, and orange juice as antiscorbutic. The ration used was, however, like all pellagra-producing diets, ill balanced, being characterized by low protein and excessively high carbohydrate and fat. According to Funk and Dubin,53 such a diet requires the addition of extra vitamin (B) in order to maintain health. It follows, therefore, that vitamin deficiency is probably invariably a complicating factor in the disease.

McCollum has shown that the deficiency in Chittenden and Underbill's experimental diet was A, mineral salts, and poor quality of protein, since his experiments have demonstrated that all the important seeds can be rendered dietetically complete by addition of these factors.54 The same deficiencies are commonly observed in the diet of pellagrins. McCollum believes, however, that pellagra is caused by an infectious agent to which individuals may be rendered susceptible through lowered vitality due to faulty diet. The experiments with rats are said to indicate that " health and vigor are promoted by a liberal intake of protein of good quality better than by any diet in which there is a tendency towards parsimony with respect to this dietary factor".

All recent work tends to emphasize the quality rather than the quantity of protein used. Boyd, in the report already referred to on pellagra among the Turkish prisoners in Eygpt, states that: Examination of the diet showed no connection between incidence of pellagra and either total calorific value or fat content or total protein content of the diets. It was only when the conception of the biological value of the protein was applied that any light was thrown upon the connection between diet and incidence of pellagra. The human body must have a certain amount of specific amino acids, and the different proteins have different biological values according to their power to supply these essential building materials, and so maintain the body in nitrogen equilibrium. If meat is regarded as having a biological value of 1.0, milk is found to have a value of 0.96, beans of 0.8, and corn (maize) of only 0.298. It is only when the biological value of the protein in a diet falls below 40 that pellagra begins to appear. Failure to absorb may raise this value as does also the performance of heavy work.

For treatment, increase in the diet of protein of high biological value was found to cause the most rapid improvement, and administration of hydrochloric acid was helpful.54a A similar conclusion was reached by Wilson.55

52 Clementi, Rub. Clin, terapi. Sd. Amu. 16, 121, 1917. 53 punk and Dubin, Sd. 52, 447, 1020.

54 McCollum and Simmonds, J. Biol. Ghem. 32, 29, 181, 847, 1917; lb 83, 303, 1918. McCollum, Simmonds, and Parsons, lb., 33, 411, lb., 38, 118, 1919.

Observations have been made by Wilson,56 upon the etiology of pellagra, and go to prove that the disease is to be attributed principally if not entirely to a defective protein supply in the diet. This conclusion emerged after a careful analysis of a large series of diet known to have produced pellagra, together with others, proved to be preventive or curative. Special attention was paid to the protein provided, not only as regards the actual amount, but also as regards the "biological value." The figures expressing the latter are taken from the work of Thomas, who investigated the nutritive properties of protein derived from various foodstuffs, and estimated their relative value in maintaining nitrogenous equilibrium in the human subject. These values were found to vary within wide limits, the "biological value" of vegetable protein (and especially of the protein derived from the corn grain) being less than that of milk or meat proteins. Equilibrium was not obtained with corn diet, but Thomas calculated from the data obtained that the amount of " tissue repair " effected by 30 grm. protein from meat or milk would require over 100 grm. of corn protein. Other foodstuffs, such as fish, rice, potatoes, occupied an intermediate position.

The reason for this difference lies doubtless in the fact that animal proteins are more approximately constituted for animal nutrition as regards the amount and variety of their constituent amino-acids. The inferiority of corn proteins would be explained by the large proportion contained of sein, a protein which is devoid both of tryptophane and lysine, two amino-acids which are known to be essential for animal nutrition.