Having convinced himself that pellagra could be prevented by a suitable dietary, Goldberger planned an experiment to determine whether a faulty diet of the type common among pellagrins would produce the disease in man. The plan involved restricting men to a diet similar to those which had been supplied to the institutions where pellagra had been endemic, and where it had been relieved by the changes in the food supply described above. This was, indeed, the type of diet characteristic of the homes of the cotton mill workers throughout the section of the South where pellagra was very common. This experiment is of extraordinary interest because of the care with which it was planned and executed, and because of the unquestioned skill of the observers.

The Governor of Mississippi was induced to offer pardon to any of the healthy white men in the state prison who would submit themselves as subjects of experiment. Twelve men offered themselves and of this group eleven actually underwent the test. White, male adults were selected because this group of the population in pellagrous districts had shown a lower incidence of the disease than any other race, sex or age. It was believed that the experiment, if it was successful in inducing pellagra in these men, would be the more conclusive because carried out with this selected group of subjects.

The fact that attacks of pellagra occur most frequently in the spring and early summer, was also taken into account, and the experiment was started at a season of the year when, if pellagra was successfully produced after a period corresponding approximately to the winter season, during which the people of the South restrict themselves to a simple and monotonous diet, the disease would appear out of season, i. e., in the autumn.

The men were placed on the experimental diet on April 19, 1915, and continued with it until October 31st of the same year.

The experiment was carried out on the Rankin farm of the Mississippi penitentiary. On this farm were a considerable number of convicts who were regarded as sufficiently trustworthy to be permitted to work on the farm in some capacity. These men were under observation as controls. The time during which these men were observed varied greatly, but eight were observed for periods of eight to nine months. Of one hundred and eight convicts on the farm at the beginning of the experiment, thirty were present until it was terminated. No case of pellagra occurred on the farm during the experiment, or had been noted among the convicts or attendants at the institution in previous years, although it was fairly prevalent in the country surrounding. There would seem to be no question of the fact that the eleven experimental subjects were entirely protected against exposure to pellagra, for they were under guard throughout the experiment.

The diet of the men in the experimental group consisted of white wheat flour, degerminated corn meal (maize), polished rice, starch, sugar, molasses, pork fat, sweet potatoes, collards, turnip greens and coffee. From the data furnished by Dr. Gold-berger, I calculated that less than 4 per cent of the total energy value of the diet was derived from cabbage, collards, turnip greens and sweet potatoes. The diet was, therefore, essentially derived from degerminated cereals, and products made from these, together with molasses and fat pork. It will be recalled that it has been thoroughly demonstrated that such a diet is not adequate for the maintenance of health in an adult for a very long period. The average protein intake was 41-54 grams; of fat, 91-134 grams, and of carbohydrates, between 387 and 513 grams per man per day (12).

At the end of five and a half months six of the eleven men were diagnosed as showing the skin lesions characteristic of incipient pellagra. The diagnosis was concurred in by a number of dermatologists who were familiar with pellagra. The first evidence of skin eruption was seen on all the men who became affected, as a bilaterally symmetrical scrotal eruption. Erythema appeared on the backs of the hands of two of the men, and gastro-intestinal and nervous symptoms were noted in all. Goldberger and Wheeler (12) hold the view that it is probable that the first appearance of skin lesions in their experimental men on the scrotum, is to be looked upon as the result of some peculiarity in the diet employed. They regard it as probable that the several clinical manifestations of pellagra, such as the appearance of the initial lesions on the back of the hands, or back of the feet, as due to specific peculiarities in the diet which brought on the pathological state. Thus they suggest that, although their group of subjects was served the same diet, they did not all eat the same proportions of different foods. Some traded dishes, and so one may have eaten more green leafy foods; another more sweet potatoes, etc. On this basis they would account for the fact that one man lost much weight during the experiment (124 to 99.5 lbs.), whereas another lost much less (126-118 lbs.), during the same interval. The idea that there is in such a condition as pellagra an intimate relationship among the several components which go to make up the diet, and the specific clinical manifestations which the disease may assume, involves a degree of specificity in the effects of faulty diet, which has as yet scarcely been acknowledged by investigators.