This theory as to the distribution of the anti-beriberi factor in seeds is confirmed by other experimenters,45 and McCollum 46 asserts in general that the cell-rich tissues of the plant such as the germ are better sources of B than are the entire seeds, but these in turn are better than is the endosperm.

All these investigators made use of the commercial embryo of the grain for their tests. Osborne and Mendel47 have shown that whereas commercial wheat embryo is rich in B, if the embryo be very carefully separated by hand from the remainder of the seed it is a much less efficient source of this factor than has been supposed. On diets including as much as 150 mg. of pure wheat embryo per day as the only source of B they secured perfect maintenance of rats but no growth. On the other hand the residues from which the embryo had been carefully removed proved quite as effective in promoting growth as did the whole wheat kernel. When the embryo-free grain was divided into sections and fed separately it appeared that the end near the embryo is more efficient than the remainder of the seed, although the latter is by no means devoid of activity. They conclude that the vitamin is located in the endosperm, but not uniformly distributed through it, the high vitamin activity of the commercial embryo being probably due to the adherence of a considerable proportion of the softer parte of the endosperm which surround the embryo proper.

According to Osborne and Mendel, this conclusion is not at variance with their experiments with wheat flour or commercial embryo meal, nor with Voegtlin, Lake and Myers' discovery that patent flour is practically free from vitamin. In making flour by gradual reduction the grain is broken into successively smaller particles, a bolting process being interposed between the breaks. The finer particles, which represent the more friable part of the endosperm, are thus sifted out and the larger and harder particles which remain on the sieves are ground into patent flour. If the vitamin is in fact concentrated in the softer parts of the endosperm the patent flour may be expected to be nearly free therefrom and the lower grades to be correspondingly richer in this food factor.

43 Coward and Drummond, Biochem. J., 15, 630, 1021.

44 Chick and Hume, J. Roy. Med. Corps, 29, 121, 1917.

45 Voegtlin, Lake and Myers, U. 8. Pub. Health Reports 33, 647, 1918; Am. J. Physiol. 48, 504, 1919.

46 McCollum, Simmonds and Parsons, J. Biol. Chem. 37, 287, 1920.

47 Osborne and Mendel, J. Biol. Chem. 37, 657, 1918.

An attempt has been made to correlate the presence of vitamins with the distribution of phosphorus in cereals. Fraser and Stanton 48 concluded that a diet consisting mainly of rice with a phosphorus pentoxide content of less than 0.4 per cent would not furnish sufficient antineuritic body for safety. Voegtlin and Myers assert that:

"The phosphoric anhydride determination of wheat and corn products yields fairly satisfactory information as to the content of these products in accessory foods. A low phosphoric anhydride content indicates that the product is poor in vitamins.49

Voegtlin, Sullivan and Myers 50 proposed a minimum phosphorus pentoxide content of 0.5 per cent as a standard for corn products, and of 1 per cent for wheat flour. Green objects to this standard.51 He admits that the examination of maize-milling products by dietetic experiments using the pigeon as discriminant, indicates that the distribution of vitamin in the corn kernel follows the distribution of phosphorus pentoxide whenever any given sample of grain is taken into consideration, but claims that the parallelism does not hold between different samples of grain, and that no difference in vitamin content could be detected by pigeon analysis in a series of samples of whole corn varying in phosphorus pentoxide content from 0.35 to 0.71 per cent. In these samples the "indicator limit" of phosphorus pentoxide for milled meals on the border-line of efficiency, would vary from 0.23 to 0.46 per cent. From this it would seem impossible to use phosphorus pentoxide in milled products as indicator of vitamin efficiency unless the phosphorus pentoxide content of the original mother-grain is known.

This information is rarely available, and the determination of phosphorus pentoxide as a general analytical guide to efficiency, as advocated by Voegtlin, Sullivan and Myers would be impractical according to Green. Their standard of 0.5 per cent for phosphorus pentoxide corn meal would condemn more samples than it passed, including the majority of perfectly efficient South African meals. Simple microscopic examination of a meal, to gage the extent of milling, would be a safer guide than the phosphorus pentoxide standard.

48 Fraser and Stanton, Studies from the Institute of Medical Research, Federated Malay States, 1909. The etiology of beri-beri.

49 Voegtlin, Lake and Myers, U. S. Public Health Repts., 33, 647, 1918; Voegtlin and Myers, Ibid. 83, 911, 1918.

50Voegtlin, Sullivan and Myers, Ibid. 31, 936,1916.

51Green, 8. Af. J. 8cL 14, 519,1918, Ghem. Abs. 13, 9000,1918.

By taking "average pigeon requirements " as standard for comparison, and stating this as 100, Green has assigned "vitamin indices" to certain cereal products. On this basis whole corn (maize) works out at about 160 to 180 - i. e., contains over 60 per cent more vitamin than is actually required in metabolism - and whole maize can therefore stand depletion of vitamin phosphorus pentoxide to the extent of about one-third before deficiency is likely to be manifested. The following vitamin indices represent determinations on an average series of milling products to an estimated accuracy of about 10 per cent either way:

The actual value in any given case depends, of course, upon the mode of milling. The average fine meal is not deficient. The more highly milled high-class table products and breakfast foods are almost invariably notably deficient, and their vitamin indices may vary from the border-line 100 down to 30. Since, however, these products are more expensive, and only used by white people living on a mixed diet, their low vitamin content is of minor importance.