The dextrose taken up in the intestinal capillaries passes by the portal vein into the liver. During digestion the portal blood contains from '2 to .4 per cent of sugar, whilst during fasting the amount is the same as in the general circulation, .05 to .2 per cent. The sugar taken to the liver is deposited there as the polysaccharide glycogen. The muscles also contain glycogen, and must have the power of forming it from the sugar carried to them in the blood, for no glycogen can be found in the plasma. The liver synthesizes dextrose, levulose, and probably galactose, that is the monosaccharides, to glycogen, which appears to be an identical product in each case. It is assumed that levulose is not first converted into dextrose and then into glycogen because in pancreatic diabetes levulose can form glycogen whilst dextrose cannot. Protein has long been regarded as a source of glycogen, although careful feeding experiments do not prove the point satisfactorily. Nevertheless, even if food proteins cannot be shown to increase the quantity of glycogen in the liver, it appears that the body protein can; for if rabbits are subjected to strychnine poisoning, which leads to the disappearance of all the glycogen from the liver and muscles, and are then put under the influence of chloral, glycogen is found to re-accumulate.

It is certain that sugar can be formed from protein, and though it is not known how far this takes place in health, or whether glycogen is a necessary intermediate stage, in diabetes a large part of the dextrose in the urine is known to be derived from the nitrogenous tissues.

There is a difference of opinion as to whether sugar is formed in the body from fat. We have already seen that when fat is being formed from sugar the respiratory quotient is above unity. Pembrey has shown, conversely, that in hibernating animals respiratory quotients as low as .3 or .4 may be observed. Since the normal oxidation of fat gives a quotient of .7, such a figure as .4 means that much oxygen is retained in the body, and it is reasonable to suppose that this is being used in the formation of carbo-hydrate, which contains much more oxygen than fat. Leathes suggests as another explanation that fat is imperfectly oxidized in the hibernating state. Further argument in favour of the formation of sugar from fat is derived from the study of diabetes. In this disease the ratio between the dextrose and the nitrogen in the urine, known as the D/N ratio, when the patient is on a strict diabetic diet, there being no carbo-hydrate in the food, has frequently been found to be 11 or 12. Now supposing that all the carbon of protein could be excreted in the form of sugar, excepting that of the urea arising from the nitrogenous part of the protein molecule, the theoretical D/N ratio could not be greater than 7. A higher figure, such as 11, means, therefore, that sugar has been derived from some other source than protein: as no carbo-hydrate is being taken, and the glycogen in the body is insufficient to explain the large excess of sugar, it is concluded that it is formed from fat.

The glycogen stored in the liver is discharged as dextrose into the hepatic blood to supply the carbo-hydrate needs of the body. The conversion of glycogen to dextrose takes place very rapidly after death, and is effected by an intra-cellular ferment, which can be extracted by suitable methods. The liver may be made to discharge glycogen experimentally by cold, by fasting, and by severe exercise, all these being conditions which call for a good supply of energy, and this carbo-hydrate food can furnish. The liver also loses its glycogen under abnormal experimental conditions such as after the diabetic puncture and under the influence of various poisons. We possess very little direct evidence as to how and where dextrose, whether it be formed from glycogen, or directly from protein or fat, is oxidized; the blood itself contains a weak glycolytic ferment, but it is improbable that this is responsible for the great destruction of carbo-hydrate which takes place in the body. The muscles are certainly the seat of oxidation of non-nitrogenous matter, but no glycolytic ferment can be extracted from muscle. Very suggestive, however, is the observation of Cohnheim that a mixture of muscle juice and pancreatic extract gives a strong glycolytic ferment. This would explain the occurrence of diabetes after excision of the pancreas, for then the combustion of sugar would presumably fail owing to the absence of the pancreatic constituent of this ferment. If Cohnheim's observations were fully confirmed we should conclude that sugar is oxidized in the muscles under the influence of an activating ferment from the pancreas. As the matter stands at present, however, those who have repeated these experiments have obtained conflicting results.

It is evident that our knowledge of the fate of food-stuffs in the tissues is very imperfect. Nevertheless great strides have been made in the last ten years in the study of this intermediate metabolism. These researches are of vital importance to the science of dietetics. Only as we understand the exact uses to which foods are put can we expect to adapt the diet wisely to all the varied needs of the individual, and to guide public opinion, both in respect to food and to general hygiene, along such paths as shall lead to the perfect development of the body and of the mind.