The preceding chapter explains how it was that so important a factor in nutrition as the vitamins are now believed to be, could have escaped recognition for so long a time, and follows in some detail the gradual development of the modern conception. In order that we may be better satisfied as to the validity of this theory however, a careful examination of the experimental evidence forming its basis is necessary. The present chapter is devoted to a description of the experimental methods employed.

As we have already pointed out, it was a matter of common observation that certain restrictions in the diet were followed in some cases by definite diseases, beriberi and scurvy, and in others by general failure. As knowledge of the familiar constituents of food became more extensive experimental conditions were improved and causes and results could be connected more accurately. In this way the first information about vitamin A was arrived at by accident, as it were, in the course of investigations designed to solve other problems, while the direction in which the investigation of the antineuritic vitamin should proceed was indicated by Eijkman's observation that the polyneuritis produced in fowls by a diet of polished rice was closely analogous to, if not identical with, the disease of beriberi in man.

Vitamin research has been carried on almost entirely with the use of two or three types of small animals as subjects;1 - rats and occasionally mice 1a for the study of the factors A and B, pigeons and fowls for work on the antineuritic principle (although cats2 have also been recommended as very satisfactory for the purpose), and guinea pigs and in a few cases monkeys 3 for observations on C

The advantages of working with the smaller animals are obvious; the greater ease of caring for them and preparing their food supplies, an arduous task at the best; the simplification of the necessary measurements and analyses; the possibility of working with larger numbers of animals and hence reducing the error from individual variations; and finally the comparatively short time required to enable such animals to reach maturity, making it possible to observe and complete the record of two or three generations within a reasonable time.

1 Hume recommends guinea pigs as a substitute for rats in testing for A in non-fatty foodstuffs. Biochem. J. 1921, 15, 30.

1a McArthur and Luckett, J. Biol. Chem. 20,161,1915. Rohmann, F. Ueber kunstliche Ernahrung und Vitamine, Berlin, 1916.

2 Voegtlin and Lake, Am. J. PhyBiol. 47, 568, 1918.

3 Chick, Hume and Skelton, Lancet 1918, ii, 735.

The rat is particularly satisfactory for the majority of investigations. Its span of life is about three years; a rat of this age being said to be comparable to a man of ninety 4 although Slonaker 5 has kept white rats alive for more than a thousand days. It has a gestation period of only 21 days and the young are weaned at the end of 25 days. The female usually produces her first litter at the age of about 120 days, and will as a rule have five litters by the time she reaches the age of 14 months, which age marks the end of her fertility.6 Moreover it is omnivorous, and hence adapts itself more easily to a variety of diets than do many species, and, from the many records of animals kept in vigorous health on a suitable diet over long periods of time, it appears to suffer less from confinement than larger animals necessarily must. On the other hand the rat appears to be entirely immune from scurvy, and so for observations on scorbutic and antiscorbutic diets the highly susceptible guinea pig is used instead, while the extreme promptness with which pigeons and fowls succumb to polyneuritis and respond to anti-neuritic treatment renders them peculiarly suitable for experimentation in this direction.

In general, it seems justifiable to assume that a factor which is essential in the diet of one species of animal will be highly advantageous if not equally indispensable for another. The close similarity between the symptoms of polyneuritis in fowls and in cats and dogs and of beriberi in man, of scurvy as observed in the guinea-pig, in puppies, and in human beings, and the fact that the eye-disease so often appearing in animals on a diet deficient in A has also been observed repeatedly in children on a diet lacking fat, would appear to confirm this conclusion. Moreover the curative measures proved efficient for the lower animals have produced similar results in human cases. McCollum says:7 "A sufficient number of comparable experiments have now been conducted with several species of animals to make it appear certain that the chemical requirements of one species are the same as that of another among all the higher animals. The requirements with respect to the physical properties of the food vary greatly.

The ruminants must have bulky food with the right consistency, whereas the omniv-ora (man, pig, rat, etc.) cannot, because of the nature of their digestive tracts, consume enough of such foods as leaves and coarse vegetables, to meet their energy requirements".

4 Donaldson: Comparison of the white rat with man in respect to the growth of the entire body. Boas Memorial Volume. N. Y. 1906.

5 Slonaker, J. An. Behav. ii, 20, 1912.

6 McCollum, Newer Knowledge of Nutrition, 1919, 14. 7 McCollum, Newer Knowledge of Nutrition, 1919, 15.

McCollum, Simmonds and Pitz8 assert that their experience of over a decade with rats and swine has convinced them that the nutritive requirements for these two species at least are essentially the same. As mentioned above, rats fail to develop scurvy on a diet on which a guinea pig succumbs promptly, nevertheless such a diet does not usually promote growth and well-being in rats.

On the other hand certain investigators have uttered a word of warning. Stcenbock, Kent and Gross9 say that " though the data (obtained by various investigations) cover a considerable range of nutrients from diverse sources much remains to be accumulated before the human dietitian or the animal feeder - unless willingly assuming the role of a speculator - can state what limiting factors are imposed on an animal by subsistence on a ration compounded from various sources. In this biological method of analysis, results are obtained which are not always directly applicable to animals of different species. For instance, it has been found that the requirements of birds for the water-soluble vitamin are far greater than those of the rat. Specific unequivocal data can only be obtained by direct experimentation on the animals in question. In this connection it must not be forgotten that the difference in the anatomical features of the digestive tracts of different animals entirely precludes the possibility of obtaining comparable data of the nutritive value of rations for the ruminant, for example, as compared with the rat. ... As 'guiding information' for the outlining of future work much of value can, however, be obtained by the use of the rat as the experimental animal.

When once the dietary properties of various naturally occurring foods have been ascertained individually, it will in many cases be possible to forecast with a considerable degree of certainty the nutritive value of combinations of these. In any case the situation is far from being simple as is especially the case of protein supplementation where the eighteen amino-acids must be considered as the physiological unite, many of which by the inability of the body to synthesize them, can act as growth determinants".

8 McCollum, Simmonds, and Pitz, J. Biol. Chem. 30, 14, 1917.

9Steenbock, Kent and Grots, J. Biol. Chem. 35, 62, 1918.