Bunge states that "the food must be more abundant in carbo-hydrates in proportion to the work performed by the muscles, and more abundant in fat according to the lowering of the surrounding temperature." From what has already been said, we know that both work and the maintenance of the internal temperature demand the expenditure of energy. The nitrogenous matter of food can be utilised by the system to some extent for supporting the internal temperature and for forming fat, as it contains the necessary carbon and hydrogen to become oxidised. Its fat-forming and heat-developing power is, however, much more feeble among vegetable feeders, like the horse, than among the carnivora, whose normal amount of fat and internal temperature can be sustained on a diet of lean meat, which consists practically of nitrogenous matter. Starch and sugar, as we have seen, are capable of forming fat in the system, but not so readily as food-fat, which can become absorbed and deposited in the tissues without having to undergo any chemical modification. Vegetable feeders, like the horse, have far less power of absorbing fat from food than carnivora, which in a state of nature consume little or no starch or sugar.

Professor Ugolino Mosso has shown by experiments he made in the University of Genoa on fasting animals, that sugar is far more efficacious than starch or albumen in raising the internal temperature, when, owing to deprivation of food, it has fallen below its normal standard. This good effect of sugar is particularly well marked in cases of extreme heat depression from long fasting, and is in accordance with the fact that energy is much more readily supplied to the body by sugar than by the other two substances. The amount of sugar given with the object in question, varied from .1 to .4 per cent. of the animal's weight, namely, I to 4 lb. for a horse weighing 1,000 lb.

Lawes and Gilbert have proved by experiment, that the feeding values of starch and cane sugar are almost identical. Muntz has shown that starch and sugar are almost entirely digested by horses, and that the carbo-hydrates of grain are much more soluble than those of hay, and especially, of straw.

The carbo-hydrates and fat in the food not only build up fat in the tissues and furnish energy in other ways, but they also retard the waste of nitrogenous tissue (muscle, etc.). We can easily understand that with a food poor in carbo-hydrates and fat, the demands of the system may entail increased waste of nitrogenous tissue.

The fact that fat horses can live far longer than thin ones, when both are deprived of all food except water (p. 127), shows that a food rich in fat-formers supplies much more energy than a food rich in muscle-formers.

It is a matter of common knowledge that the fat of pigs and cattle, used for human food, varies greatly in appearance and flavour, according to the food upon which the animals have been respectively fed. The differences in question seem to be principally due to the fact already mentioned, that a certain amount of the food-fat becomes absorbed and deposited in the tissues while undergoing little or no change. The experiments alluded to in the following extract from Henry's Feeds and Feeding, confirm this view: "Lebedeff and Munk fed a dog mutton suet, and later rape oil, for a long time and in large quantities, and found that the fat deposited in the body was more like mutton suet or rape oil in its chemical properties than normal dog fat. Munk showed that neutral fat may be formed in the animal body resulting from the feeding of free fatty acids. He fed a starved dog lean meat and a large quantity of fatty acids prepared from mutton suet. The animal, which had shrunk 32 per cent. in weight during nineteen days, was fed 3,200 grams of flesh and 2,850 grams of fatty acids in fourteen days, and increased 17 per cent. in weight during this period. When killed a well-developed layer of fat was found under the skin, and also large deposits of fat on the internal organs. By the use of the knife, 1,100 grams of fat were separated, which resembled mutton suet, and according to chemical analysis must have contained at least 96 per cent. of neutral mutton suet. It is well known that fatty acids are formed from the fat of the food in the digestive processes through the action of the pancreatic juice, and the above experiment therefore practically proves that the fat stored in the bodies of animals may be derived from the fat of the food. Munk's findings have recently been corroborated by Walker, also in experiments with dogs."

Woody Fibre in passing through the stomach and small intestine suffers but little change, and although it becomes, as we have seen, partly broken up in the large intestine, the process of its chemical alteration cannot be regarded as digestion, for the products of its decomposition are not nutritious; hence it has little or no nutritive value for a horse. Its presence in food is, however, an indispensable aid to digestion, apparently by separating the nutritive particles from each other, so as to allow the digestive fluids to readily permeate through the entire mass. Bunge remarks that "cellulose [fibre] is absolutely essential to animals with a long intestinal tract. If rabbits are fed on a diet containing no cellulose, the onward movement of the intestinal contents ceases, inflammation in the intestines ensues, and the animals rapidly die. But if horn-parings be added to the same food, nutrition is normal. These horn-parings are, as Knieriem proved by experiments devoted to that purpose, absolutely undigested, and can therefore only have taken the place of woody fibre in so far as its mechanical properties were concerned." To sum up the matter, we may say that the alimentary canal of a horse is not suited for the digestion of concentrated food; in other words, that unless the nutritive matter of a horse's food is diluted by a more or less inert vehicle, of which vegetable fibre is the natural and best form, disturbance of the digestive organs will ensue. The mechanical action of fibre in this respect is generally put down to its bulk-giving properties, on the supposition that a horse is unable to digest a food that is not bulky. We know that absence of concentration is more essential than bulk; for if an animal is put on, say, the starvation and non-bulky diet of 2 quarts of ordinary oats and 2 quarts of chopped hay, he will digest it in a normal manner; although in all probability his digestion would be upset by double that bulk (8 quarts) of oats from which the husk (mostly woody fibre) had been removed. We find that the wholesomeness (p. 72) of different samples of oats varies according to their respective percentages of husk, other conditions being equal, when this grain is given without other food; and that, in such a case, a horse can consume in a healthy manner a moderate quantity of oats which has a fair proportion of husk, but is unable to do so if the husk is absent. Again, maize, which has almost the same composition as oats, except that it contains less fibre and mineral matter, requires for its healthy digestion the addition of more "chop" than do oats. In fact, by adding fibre to maize (for instance, by grinding up the grain and cob together, so as to make the fibre percentage of the mixture about equal to that of oats), we can render maize nearly if not quite as wholesome a food as oats. It is evident that the arrangement by which each kernel of oats is surrounded by a comparatively thick covering rich in fibre, greatly promotes the wholesomeness of that grain. The custom of mixing chopped hay with oats has long been an unconscious recognition of the truth of the principle in question, which principle is also more or less carried out when the animal is allowed hay at feeding time.