Considerable uncertainty existed at this time as to whether the growth-promoting substance was to be regarded as identical with antineuritic and antiscorbutic factors. Stepp26 expressed the belief that there is more than a single class of unknown accessory substances necessary for prolonged maintenance of an animal, but he employed only solvents for lipoids in preparing his foodstuff. Voegtlin and Towles27 noted that the extracts of autolyzed spinal cord were antineuritic, yet were unable to reestablish normal metabolism, i.e. restore body weight. Hopkins28 remarks in 1914, "It is uncertain as yet whether the substance necessary for growth is the same as that which can cure the neuritis which develops in fowls fed on polished rice, and the absence of which from a diet is supposed to be responsible for the disease beriberi." Funk and Macallum29 found that butter-fat had no beneficial effect on pigeons suffering from polyneuritis, and that young rats and pigeons were unable to live on diets of pure casein, sugar, starch, agar (for roughage), mineral salts, and butter-fat, even when the latter made up as much as 30% of the diet. Addition of from 2-6% of dried brewers' yeast remedied the defect so that good growth and maintenance were obtained.

They concluded that "the growth-promoting factor is beyond question contained in the yeast," but were uncertain as to whether there were one or more growth-stimulating components in the yeast, or whether the yeast alone without butter would produce normal growth. McArthur and Luckett30 were also unsuccessful in attempting to feed mice on casein, starch, lactose, and mineral salts or protein free milk, all of which had been extracted with alcohol and ether, supplemented with butter to the extent of 25%, but obtained good results when less than 3% of the alcohol-ether extract of egg-yolk was incorporated in the diet. They proved that the essential substance was neither lecithin, cephalin, cerebrosides, cholesterol, or fat, and suggested that it might be Funk's vitamin, but were unable to account for the discrepancy between the results and those of McCollum and Davis and Osborne and Mendel with regard to the efficiency of butter-fat.

26 Z. Biol. 62, 405, 1013.

27J. Exp. Phann. 5, 67,1913.

28 Ann. Repts. on Prog, of Chem. 1914, 210.

29J. Biol. Chem. 1915, 23, 413.

30 J. Biol. Chem. 1915, ao, 161.

McCollum and his associates were now engaged in a study of the nutritive properties of the individual grains.31 In the case of wheat it was found necessary to supplement the grain by addition of protein, inorganic salts, and a growth-promoting fat (butter-fat), but when this was done the diet proved satisfactory for growth, maintenance, and reproduction. They next turned their attention to rice, expecting similar results, but to their surprise they found that polished rice, even when supplemented with protein (casein), butter-fat, and a suitable salt mixture, not only failed to induce growth in young rats, but permitted the development of polyneuritic symptoms. The possibility of a toxic principle in the rice, as suggested by Caspari and Moszkow-ski,32 was eliminated by the fact that no improvement could be detected when the amount of rice in the ration had been reduced to one-half, the deficit being made up by dextrin. Comparison with the earlier experiments in which good health and growth had been secured in animals fed upon purified protein, starch, milk sugar, butter fat, and inorganic salts, suggested lactose as the only missing component constituent in the rice diet, and experiment showed that when 10 per cent of Merck's lactose replaced an equivalent amount of polished rice there was marked improvement.

McCollum and Davis concluded therefore that the lactose carried a second factor essential for normal nutrition during growth, distinct from that present in butter-fat and egg yolk. This new factor, now known as vitamin B, is present in milk, and being soluble in water but apparently not in fats, remains in the skimmed milk where the butter-fat is separated. It is absorbed by the lactose and can only be removed from this by very careful recrystallization, so that rations containing commercial lactose would contain an abundance of this substance, as would also the "protein-free milk" of Osborne and Mendel. It was found to be present also in rice polishings, in water extract of egg yolk, and in.

31 Wise. Ag. Exp. Sta. Res. Bull. 17,1911; McCollum and Davis, Proc. Soc. Biol. Ghem. J. Biol. Chem. 14, XI, 1013; Hart and McCollum, J. Biol. Chem. 19, 373, 1914; McCollum and Davis, J. Biol. Chem. 20, 415, 21, 615, 1915. See also McCollum, Newer Knowledge of Nutrition (1919), pp. 20 & ff. McCollum and Davit, J. Biol. Chem. 23, 181, 1915.

32 Bed. klin. Wehneohr. 50, pt 11, 1515, 1913.

While there is still some difference of opinion as to the part played by vitamins in connection with certain specific diseases, the impossibility of maintaining health on a diet lacking either A, B, or C is generally regarded as firmly established. As an example of the very small minority who, in the face of overwhelming evidence, persist in attempting to controvert the vitamin hypothesis we may quote Roh-mann.*6 .

Rohmann asserts that accessory foodstuffs are not necessary for the continued maintenance of fully grown animals, and that if long familiar I nutrients are suitable in quality and quantity nothing further is essential in the ration.

"The assumption that some unknown substances are indispensable / for growth is a convenient device for explaining experiments that result in failure - a device that becomes superfluous as soon as the ex- periment succeeds."

As evidence against the indispensability of vitamins he cites a new series of experiments on growth and maintenance of white mice fed with artificial mixtures of food supposedly free from these so-called accessory substances. In many of his earlier experiments he used yeast to impart a suitable texture to the food mixtures. Since yeast is now recognized as an efficient source of vitamin, in his latest experiments he has avoided this possible criticism by using starch digested by purified diastase, with baking powder as leavening agent.

33 McCollum and Davit, J. Biol. Chem. 23, 231, 1915.

34 Lancet, 1018,11, 482.

35 J. Biol. Chem. 1916, 24, 493.

36 Kunstliche Enrahrung u. Vitamine, Berlin, 1916.

The diets used by him were as follows:- food:

Diet I

Egg white ...................5 gm.

Egg white iron............. 2

Potato starch (predigested).. 20

Potato starch (raw)......... 25

Wheat starch .............. 90

Dextrose ................... 5

Margarine ................. 10

Salts ...................... 3

Baking Powder............. 5

The foodstuffs in diet 2 were either treated in alcohol without warming, or were heated three hours at 120-150 C.

The Kalzose used was a commercial preparation of casein combined with calcium, and still containing some milk sugar and doubtless other products present in milk. In this diet 0.2 g. of "Merck's purified diastase," washed with alcohol, was used to digest 20 g. of starch.

Rohmann states that with Diet 1 five young mice having an average initial weight of 6 g. were successfully reared. Addition of casein gave still more successful results, for a third generation was obtained. He states that he has succeeded in rearing young mice and subsequently maintaining them, but these results were frequently induced by added extracts of yeast, by small quantities of milk, or with "Filtrateiweiss," a product "principally composed of the proteins which remain in solution after the precipitation of the casein of milk," and which was supposed to have been rendered vitamin free by heating or by extraction with cold alcohol.

Rohmann's work has been severely criticised by Osborne and Mendel37 who point out that it is doubtful whether either of the methods selected by Rohmann can be depended upon to exclude all traces of vitamins.

"The thesis that a successful, i. e. positive, experiment in nutrition is far more significant than a negative one is doubless valid. On the other hand, in dealing with substances which, like the alleged vitamins, fare potent in surprisingly small amounts, the burden of the proof with respect to the complete absence of effective substances so widely Distributed among the natural foodstuffs falls on those who deny the heed of them. " 38

37 J. Biol. Ch. 31, 149, 1917.

38The Vitamine Manual by Eddy (Baltimore, 1921) contains a bibliography on vitamins. See also the Vitamines, by Funk. Translation from second German edition by Dnbin.