As indicated in the foregoing, the discovery of vitamins as essentials in the diet of animals followed from attempts to maintain animals on restricted diets of synthetic food mixtures.

Lunin,4 in 1881 compared the growth of mice fed on milk with those on a diet made up of milk sugar, milk fat, milk protein, and the mineral constituents of milk, and found that while adult mice could live in apparently good health for several months on the milk diet they invariably died within one month on the synthetic mixture. He draws the conclusion that "other substances indispensable for nutrition must be present in milk besides caseinogen, fat, lactose, and salts," but this sapient remark was practically ignored for over thirty years.

Henriques and Hansen5 endeavored to maintain white rats on isolated proteins with the addition of sugar, lard, cellulose, and inorganic salts, and noted a gradual loss of body weight in each case.

Hopkins6 studied the nutritive value of purified proteins supplemented by carbohydrates, fats, and mineral salts, and found that neither growth nor maintenance of weight could be secured on these diets. He writes in 1906 (I. c.):

It is probable that our conception of stimulating substances may have to be extended. The original vague conception of such subjects being condiments, chiefly affecting taste, gained in definiteness by the work of the Pavlov School. But the place of specific diet-constituents which, stimulate the gastric secretory mechanism can be taken by the products of digestion itself, and in this connection the stimulant in the diet is by no means indispensable. Most observers agree that the addition to normal dietaries, of meat extracts capable of stimulating the gastric flow does not increase the actual absorption of food, though this point could be properly tested by adding them to artificial dietary known to be free from analogous substances. As was emphasised above, the milk did not affect absorption in my experiments. But such undoubted stimulating effects due to diet constituents as those discovered by Pavlov may quite possibly be paralleled elsewhere in the body on more specific and indispensable lines. Stimulations of the internal secretions of the thyroid and pituitary glands which are believed on very suggestive evidence to play an important part in growth processes, can be legitimately thought of. On the other hand, the influence upon growing tissues may be direct.

If the attachment of such indispensable functions to specific accessory constituents of diets is foreign to current views upon nutrition, so also is the experimental fact that young animals may fail to grow when they are absorbing daily a sufficiency of formative material and energy for the purpose of growth.

4 Ztschr. f. physiol. Ch. 1881, 5, 31.

5 Ztschr. f. physiol Ch. 1904-6, 43, 417, 1908, 54, Henriques, Ibid. 1009, 6o, 105.

6 Willcock and Hopkins, J. Physiol, 1906, 35, 88, Hopkins, Analyst, 1906, 31, 396.

Falta and Noeggerath7 fed rats on various purified proteins, starch, sugar, and fat, with the addition of salts, water, and in some cases lecithin, cholesterol, and sodium nucleate, all obtained as pure as possible. In most cases there was a steady decline in weight throughout the experimental period, followed by death, although occasionally the body weight was maintained fairly well for the first few weeks. In no case did they succeed in keeping the animals alive more than 94 days, and usually death ensued in shorter time. On the other hand they were able to keep rats in health on milk alone for periods of six months or more.

Jacob8 used pure starch, olive oil, casein, sugar, salts, and cellulose as a diet for pigeons and rats, and concluded that the lack of success in this and similar experiments was partly due to the unsatisfactory physical character of the diet and, probably, partly to the difficulty of inducing the animals to eat sufficient of a food which was unpalatable. McCollum9 noted the same difficulty in, feeding rats on complex artificial mixtures of purified foodstuffs, and attempted to overcome it by frequent changes in the combinations of foodstuffs used and by addition of flavors. He concluded that palatability is a most important factor in animal nutrition, and that the failure of previous efforts to maintain animals on mixtures of isolated foodstuffs was due to lack of palatability. Although he was successful in keeping his animals alive for comparatively long periods they failed to maintain their weights. In this connection Osborne and Mendel have pointed out the difficulty of deciding whether the failing appetite invariably observed is a cause or result of imperfect nutrition.

7 Hofmeister's Beitr. Z. Ghem. Physiol. 1906, 7, 320.

8Ztschr. f. Biol. 1906, 48, 19. 9 Am. J. Physiol. 1909, as, 130.

That a monotonous and unaccustomed food may be used successfully over long periods of time without ill-effects was proved by the experiments of Falta and Noeggerath already referred to, in which rats were maintained successfully for six months or more on monotonous diets of milk, milk powder or lean horsemeat; by Socin,10 who kept mice in good health for 90 days on egg-yolk, starch, and cellulose, and by many other experimenters.11

In the light of our present knowledge many of these early experimental diets were obviously lacking in amino acid or mineral content or both, nor were the conditions ideal in other respects, nevertheless the multiplication of evidence seemed to point to an inexplicable difference between natural foods and the synthetic mixtures designed to be qualitatively and quantitatively equivalent to them. Roh-mann12 indeed claimed to have kept mice indefinitely and secured reproduction on food mixtures made up of purified components, but there is reason to believe that his success was due to the fact that the food substances used still contained the essential constituents in no inconsiderable amount.