If the data is then meagre in the field of the "B" vitamine it is still more limited in the case of the "A" and the "C." One of the earliest difficulties encountered in the study of the "A" vitamine was the failure of fat solvents to extract the material from its richest vegetable sources. If butter or egg yolk is extracted with ether, the fat obtained is rich in the "A" vitamine. If, however, ether-extraction is applied to green leaves or seeds it removes the oils but these oils contain little or no vitamine. Pressing methods also fail to remove the substance from vegetable sources. For example, if we press or extract cotton seed we obtain the oil but the vitamine is retained in the press cake. McCollum suggested the following explanation for this behavior. His idea is that the "A" vitamine while soluble in fat is so bound up in the vegetable source that extraction methods fail to loosen it. When these vegetables are eaten the vitamine is set free in the process of digestion and being fat-soluble passes into solution in the animal fats. Hence, when these fats contain it in solution, they retain it in the process of extraction while, lacking this separatory process, ether fails to loosen it from the vegetable binding. Recently, however, Osborne and Mendel have presented data in regard to this binding and shown that if for ether we substitute an ether-alcohol mixture the removal of the "A" with the fat is fairly complete even from vegetable sources. They advance the idea that preliminary treatment with alcohol is a process which will materially assist in breaking the attachment of the vitamine and render its removal with the fat solvent effective. Butter-fat rich in the "A" vitamine has been conclusively shown to be free of nitrogen and phosphorus and it is generally assumed that the "A" vitamine is a nitrogen-free and phosphorus free compound. Further than that however we know nothing of its nature. Concerning the "C" we know only that it is like the "B," water-soluble and we know somewhat of its properties, but nothing of its chemical nature. One of the greatest difficulties still encountered in the study of chemical fractions is the delay in identification of the active portion. For this purpose we must rely on tests that are far from delicate and time-consuming to a degree. As a result the study of only a few fractions must extend over long periods of time with all the cumulation of difficulties in the way of change in material, etc. that this delay implies. An idea of these difficulties can best be obtained by a review of our present methods for vitamine testing and these methods constitute the subject matter of the next chapter. CHAPTER III THE METHODS USED IN TESTING FOR VITAMINES