There are other diets that have been given for this purpose, e.g., Holst and Fröhlich induced scurvy by restricting animals to an exclusive diet of cereals (oats or rye or barley or corn). Hess and Unger have used hay, oats and water given ad libitum. All of these and others are subject to criticism on the basis that they are not necessarily adequate in other food factors and may therefore not be fair bases for testing the antiscorbutic powers of the unknown combined with them. Abels has recently shown that scurvy increases susceptibility to infections and believes that the scurvy hemorrhages are brought about by the toxic effects of infection. It is therefore desirable in testing for antiscorbutic power that the basal diet be itself as complete as possible in all factors except the absence of "C."
The study of rickets has already progressed to the stage of calculating rickets-producing diets and the methodology is identical with that for scurvy but this phase of testing still lacks evidence of an antirachitic vitamine and in that uncertainty it is hardly worth while to elaborate these diets here. The British diets are all based on Mellanby's contention that the "A" vitamine is the antirachitic vitamine. This view is not yet accepted by American workers.
In concluding this chapter it is sufficient to state that with our present methodology the accumulation of data for evaluating the vitamine content of various foods is still far from satisfactory and from the chemist's viewpoint the methodology is most unsatisfactory as a means of testing fractional analyses obtained in the search for the nature of the substance, both because of the time consumed in a single test and from the difficulty of using the fractions in feeding experiments when these fractions may themselves be poisonous or otherwise unsuited for mixture in a diet. It is obvious therefore that interest is keen in any possibility of devising a test that will be specific, quick and not require modification of the material tested, because of its unsuitability for feeding. In 1919 Roger J. Williams proposed a method that seemed to offer promise in these respects but which is not yet in the form for quantitative use. It offers promise that entitles it to a special chapter for discussion and the next chapter presents the present status of the so- called yeast test for vitamine "B."