In discussing the formation of body fat from carbohydrate it was shown that often the greater part of the fat stored is manufactured in the body from carbohydrate. So striking were the results of some of the experiments demonstrating the synthesis of fat from carbohydrate, that physiologists came to question for a time whether any of the fat deposited in the tissues comes from the fat of the food. Abundant evidence that food fats may be directly deposited in the body has been obtained by feeding characteristic fats to dogs and showing that these fats can afterwards be recognized in the tissues of the animals. Experiments of this kind have been made, using linseed oil, rapeseed oil, or mutton tallow, any of which is easily distinguishable by its chemical and physical properties from the fat normally found in the body of the dog. For example, Munk starved a dog for 19 days, and then for 14 days fed a mixture of the fatty acids obtained from mutton tallow, as a consequence of which about one half of the weight lost by fasting was regained. The dog was then killed and yielded on dissection 1100 grams of fat melting at 400, which is about the melting point of mutton tallow, whereas normal dog fat melts at about 200. In another experiment by Munk rape oil was fed and the fat obtained from the dog was found to contain 82.4 per cent of oleic and erucic acids and 12.3 per cent of solid acids, whereas normal dog fat had only 65.8 per cent oleic, no erucic, and 28.8 per cent of solid fatty acids.

The occurrence in the body fat of properties usually characteristic of some particular fat which has been fed is now very well known and is recognized in establishing standards of purity for fats of animal origin. Thus, the lard obtained from swine which have been fed cottonseed meal shows the characteristic color reactions of cottonseed oil, and more elaborate tests must be made in order to determine whether cottonseed fat has actually been mixed with the lard.

* Journal of the American Chemical Society, Vol. 21, page 823 (1899).

European food officials sought to establish an easy method of distinguishing between butter and its substitutes by requiring manufacturers of any butter substitute to use a certain proportion of sesame oil in the preparation, sesame oil having a characteristic color reaction which can be very easily demonstrated without the use of laboratory facilities. It was found, however, that the same sesame oil reaction might be exhibited by a perfectly pure butter fat from cows which had been fed upon sesame meal.

Evidence of the formation of body fat from food fat has also been obtained by experiments upon the total amount of fat formed in the body when the amount and composition of the food eaten was accurately known. Hoffmann starved a dog until its weight had decreased from 26 to 16 kilograms, so that it must have been almost devoid of fat. He then fed small amounts of meat and large amounts of fat for five days, after which the dog was killed and analyzed. The body contained 1353 grams of fat, of which not over 131 grams could have come from proteins, and only a few grams at most from the small amount of carbohydrates in the meat fed, so that about nine tenths of the fat which the animal had laid on must have come from the fat of the food.

Thus there is abundant experimental evidence that both the carbohydrate and the fat of the food may serve as sources of body fat. In a later chapter it will be shown that protein also may contribute to the production of fat in the body.

A question naturally arises as to how, if proteins, fats, and carbohydrates of food may all contribute to the production of body fat, the nature of the fat can still be to any significant degree characteristic of the species in which it is found. A partial explanation appears to be furnished by the recent work of Bloor, who finds that when the fat of the food has been split to glycerol and fatty acids in the course of digestion and these digestion products are taken up and resynthesized to fat in the intestinal wall, there may go into the resynthesized fat not only the fatty acid radicles of the food fat but also fatty acid radicles formed in the body. These latter, entering into the constitution of the absorbed fat, tend to give it some of the proper-' ties characteristic of the species while at the same time some of the characteristics of the food fat may be retained. Thus when a dog is fattened by feeding mutton tallow which contains more stearin and less olein than ordinary dog fat the organism may, if the fattening is gradual, furnish enough oleic acid radicles to bring the resynthesized fat to the consistency ordinarily found in dog fat, or if the fattening is more rapid the oleic acid radicles may not be supplied at a sufficiently rapid rate to yield this result and the dog will then lay on fat of a character somewhere between that of mutton tallow and ordinary dog fat, the influence of the food fat upon the character of the stored fat being more pronounced the more rapidly the fattening is carried out. It will be noted that, even if the fatty acid radicles synthesized in the body are built into the absorbed fat to such an extent as to bring its consistency and other physical properties to what is characteristic for the species, yet such body fat may still contain some radicles of fatty acids characteristic of the experi-mental food and not ordinarily found in the fat of the animal, as in the case of erucic acid in the experiment cited above (page 32).