Such experiments were greatly facilitated by the fact, discovered by Eijkmann in 1897, that fowls develop a diseased condition closely resembling beriberi in man, when they are fed exclusively upon polished rice for 3 or 4 weeks. Ohler (1914) finds that an exclusive diet of white bread, especially when made without yeast, has the same effect as the polished rice diet. This experimental beriberi of fowls (or "polyneuritis gallinarum") does not occur when whole rice or even rice which has been partially milled so as to retain the inner bran coat (pericarp or "silverskin") is fed. It was soon found that rice polishings (bran) when added to the polished rice diet not only protected the fowls but also cured those which had already developed the disease. Aqueous and alcoholic extracts of the rice polishings also served to prevent or cure the disease. The same was found true of many ordinary foods such as meat, potatoes, beans, peas, and peanuts, the legumes being especially efficient.

*Little (1914). The prevalence of beriberi in Newfoundland and Labrador appears to be due to a diet too largely restricted to white bread.

Aron working on Oriental beriberi in the Philippines, and Schaumann in Europe, centering his interest more particularly in ship beriberi, were both impressed with the fact that diets which cause beriberi are poor in phosphorus and that foods of good curative and preventive properties are rich in phosphorus. They were therefore inclined to regard beriberi as connected with a deficiency of phosphorus in the diet. Their attempts to prevent or cure the disease by adding definitely known phosphorus compounds to the polished rice diet gave, however, for the most part negative results. In Aron's experiments the deleterious effects seemed to be reduced though not excluded when phytin was fed. In Schaumann's experiments yeast lecithin and yeast nucleic acid seemed effective, but egg lecithin, phytin, simple phosphates, and glycerophosphate showed no beneficial results. The direct evidence for the "phosphorus theory" is therefore weak and somewhat conflicting. (This does not exclude the possibility that deficiency of phosphatids may be at least a factor in the nerve degeneration as argued by Schaumann.)

Furthermore, Fraser and Stanton showed that an extract of rice polishings which contained only 15 per cent of its total phosphorus was capable of preventing the neuritis, while the residue containing the other 85 per cent of the phosphorus was ineffective; and soon afterward Chamberlain and Vedder showed that an alcoholic extract of rice polishings which was highly protective contained only 0.0007 per cent of phosphorus or less than one part in one thousand of the phosphorus originally present in the polishings.

Chamberlain and his associates also tried the effects of various inorganic salts, of sugar, phytin, lecithin, allantoin, choline, and many of the amino acids, all of which proved insufficient to prevent the development of polyneuritis in fowls kept on a polished rice diet. On the other hand, they added to the knowledge of the properties pi the antineuritic substance by a study of the effectiveness of rice bran extracts after different treatment. The antineuritic substance was found to be insoluble in ether but soluble in alcohol or in water and dialyzable. It was not volatile but was destroyed by heating or by alkali; in the presence of acid, it was more stable. It was not precipitated by lead acetate. They held the curative substance to be an organic base but not an alkaloid. Bean extracts were found to contain one or more substances having similar properties. Fresh milk, meat, and potatoes were also found to have antineuritic properties.

Later in 1911 and early in 1912, several investigators independently and almost simultaneously succeeded in isolating what appeared to be specific antineuritic substances.

Funk's experiments, begun about the middle of 1911, have attracted special attention since he was the first to announce (December, 1911) the isolation of a definite chemical substance possessing the antineuritic property. Pigeons paralyzed by neuritis induced by a polished rice diet were able to run and fly within a few hours after administration of 2 to 8 milligrams of this substance, which appeared to be an organic nitrogenous base related to the pyrimidines and to which Funk gave the name vitamine. He described the preparation of such substances from rice bran and from yeast, and inferred the existence of the same or a similar vitamine in all foods which have antineuritic properties. Funk's view of the relation of vitamine to the phenomena of beriberi is as follows: The lack of vitamine in the food forces the animal to get this substance from its own tissues (with the result that there is wasting of the muscles causing emaciation unless accompanied by oedema). After the stock of vitamines available in the muscles begins to be scarce, there results a breaking down of the nerve tissue and the appearance of nervous symptoms such as are observed in beriberi.

Funk called this beriberi vitamine. It constituted only 0.05 per cent of the rice polishings corresponding to about 0.01 per cent in the whole grain.

In March, 1912, Edie, Moore, Simpson, and Webster (working independently of Funk) described the isolation from yeast of a base which promptly cured pigeons suffering from polyneuritis. This base they described as having composition corresponding to the formula C7H17N2O5. They called it toruline.

Schaumann (June, 1912) reported the preparation of a phosphorus-free nitrogenous crystallizable base corresponding in general to the description given by Funk and exerting a marked restorative action upon polyneuritic pigeons. This base he considers the "activator" in the cure of polyneuritis, holding that it "mobilizes" the phosphatid substances which must be rebuilt into the degenerated nerve tissue in order to effect a permanent cure.

In July, 1912, Suzuki, Shimamura, and Odake, reported an extended investigation of experimental beriberi in which they had prepared from rice polishings by an independent method a base of high curative power which they called oryzanine. In preparing oryzanine they precipitated an alcoholic extract of rice polishings with tannin, decomposed the tannate by baryta, removed the barium by sulphuric acid, and precipitated the base as a picrate. Only 0.005 to 0.01 gram of oryzanine was required to make the daily diet of polished rice adequate for a pigeon. Since the pigeons ate 25 to 30 grams of rice per day this means that the oryzanine was only 1/2500 to 1/5000 of the (dry) weight of the food eaten. Feeding 0.3 gram cured a dog that was already paralyzed by experimental beriberi.*

It will be seen that these independent investigations all indicate that the antineuritic property shown by rice polishings, yeast, and other natural food materials is due to some basic nitrogenous substance or substances. Much work published since 1912 confirms this general view without establishing the chemical identity of either "Funk's base," or "toruline" or "oryzanine." Pending chemical identification of the naturally occurring antineuritic base or bases the term "vitamines" is commonly applied to them.

* Apropos of the small quantities of vitamine or oryzanine necessary for pronounced effects, Lusk calls attention to the fact that epinephrine (adrenaline), an essential of life, is present in the blood to the extent of only 1 part in 100,000,000.

While the antineuritic property of such "vitamine" has been demonstrated usually by experiments upon animals, Williams and Saleeby have used a vitamine preparation, made from rice polishings, in a case of human beriberi with good results. In connection with this work it was found that acid hydrolysis renders the antineuritic substance of rice polishings more active, or more rapid in its action. It is possible that in natural food materials or simple water extracts the vitamine may exist, either wholly or in part, in combination. This would account for the greater activity and also for the instability of the free "purified" vitamine as compared with the natural form.

Seidell * has devised a method for obtaining a stable preparation of the antineuritic vitamine by precipitating it with hydrous aluminum silicate (Lloyd's reagent).

While the general view has been that a given organism requires a given amount of vitamine to maintain health (presumably a larger amount to effect recovery from disease induced by a previous deficiency), it was suggested by Braddon and Cooper (1914), and a few simultaneous experiments by Funk, that there is a connection between the metabolism of carbohydrate and of vitamine, so that the amount of antineuritic substance required by the organism increases with the quantity of carbohydrate metabolized.

It has also been suggested that the neuritis of beriberi is due to a toxic effect, upon the nerves, of some substance formed in, or absorbed into, the system and that the vitamine, when present in normal amounts, acts as a protection or antidote against such toxicity.

* Reprint No. 325 from the Public Health Reports, U. S. Public Health Service.

This hypothesis is difficult to test and does not seem to have been much studied. Investigations designed to connect the physiological property with some definite chemical substance or type of molecular structure have, however, been continued and are yielding most interesting results.