Up to this point we have been dealing with alimentary glycosuria, and, in point of fact, it may be said that we have scarcely touched upon the conditions of the proposition stated at the outset, according to which we suppose that our patient has been transformed into a diabetic while upon the same diet upon which he subsisted prior to this change. It is evident, too, how this milder form of diabetes may be aggravated if, in addition, the assimilating office of the liver is deranged and all the absorbed glucose passes directly through the liver into the general circulation. Thus, again, the sugar from ingestion is not stopped from reaching the general circulation as it ought to be. As militating somewhat against this dominating influence of the liver in health should be mentioned the fact that glycosuria does not seem to be caused by such destructive diseases of the liver as acute yellow atrophy and carcinoma. On the other hand, such unnatural transit through the liver may be reasonably ascribed to vasomotor influence causing the blood to move too rapidly through the liver to permit the primary conversion of carbohydrate into glycogen, which seems to be a necessary condition to its proper metabolism.

So much for the presence of sugar in the blood due to the imperfect conversion of the ingested carbohydrate into glycogen. It is also evident that glycosuria may be the result of an abnormally rapid transformation of glycogen into sugar. Glycogen is a product of assimilated carbohydrate, whence it is derived through the operation of living protoplasm. On the other hand, by enzymic and other means it is readily broken down and hydrolyzed into the sugar mole cules from which it was constructed, and if this conversion is too rapid to permit a complete consumption of the sugar, glycemia and glycosuria may be produced. It has been said that the liver and muscles are the chief storehouses of glycogen accumulation. Pavy has found as much as 16.67 per 1,000 in the general muscles of the body of a dog-7.72 in the heart and 29.59 in the diaphragm. Inactivity favors the storage of glycogen; forced exercise its virtual disappearance from the

liver and marked reduction in the muscles. As to the method in which glycogen becomes stored in the muscles, it is thought that the muscle cells, like the liver cells, may by their "protoplasmic action" construct glycogen out of glucose absorbed from the intestine, and perhaps also absorb from the blood glycogen constructed in the liver.

Pflüger and his assistant Nerking, on the one hand, and Pavy on the other, apparently quite independently, suggested that glycogen exists in feeble combination with nitrogenous matters in the shape of a nitrogenous nucleus, from which the sugar may be oxidized through the operation of a diastatic enzyme without disturbing the nucleus. Under normal conditions the storage of glycogen is maintained without leading to a formation of sugar resulting in glycemia and glycosuria. Bernard's puncture may be regarded as one of the deranging influences which disturbs the normal equilibrium. The same is true of operations on the sympathetic. On this source of glycosuria Bernard placed especial stress, believing that in diabetes the amyloid substance is too rapidly converted into glucose to be consumed in the ordinary processes of oxidation, and therefore appears in the urine; so that in all instances of diabetes except of the cases of alimentary glycosuria referred to, he held the excess of sugar in the blood to be derived from glycogen, the result of an exaggerated sugar formation rather than a diminished destruction. At the present day, however, the tendency of thought seems to be to ascribe the glycosuria of essential diabetes to agencies which diminish its oxidation or destruction rather than those which cause its overproduction.

I have alluded to some influences of the nervous system upon the production of glycosuria. Let us follow them a little further. Bernard early discovered that by puncturing the medulla oblongata in the floor of the fourth ventricle anywhere between a point four or five mm. above the nib of the calamus scriptorius, and another about four mm. higher up, the urine in a day or two acquired a considerable amount of sugar, and was increased in quantity. This is the celebrated

picûre experiment. The amount of sugar is larger the better fed the animal, while if all food is removed the sugar is trifling in amount or disappears altogether from the urine. The point thus punctured corresponds with the origin of the roots of the pneumogastric nerves, hence it was supposed that the diabetes was the direct result of irritation of these nerves; but it was found on section of the vagus that galvanization of the distal end of the cut nerve produced no effect, while the same irritation applied to the central end resulted in glycosuria. Stimulation of the depressor nerve by Lafont and Filehne also produced glycosuria. Whence it was learned that the pneumogastric was not an excitor, but a sensory nerve of diabetes, and that the glycosuria produced by irritation of the roots of the vagi or their central cut ends is the result of a reflex action called into play through the pneumogastric as a sensory, and some other nerve-probably the sympathetic-as a vasomotor nerve.

Following these early discoveries of Bernard, instances of the effect of nervous lesions producing glycosuria multiplied rapidly. It would be unprofitable to attempt to detail them. Among them may be mentioned destructive lesions of the pons and posterior crura of the cerebellum, section of the optic thalami and great cruri cerebri, section of the medulla oblongata and of the cord above the second dorsal vertebra-that is, above the lower end of its cervical enlargement. Section of the cord below this point is without effect, while glycosuria again ensues on extirpation of the superior cervical ganglion and of the first thoracic ganglion, and on section of the sympathetic chain ascending thence and accompanying the vertebral artery. Schiff also produced glycosuria by section of the sciatic nerve, while from the clinical standpoint Rosenstein called attention to the combination of glycosuria with sciatica, limited to the peroneal and tibial nerves, ascribed this effect to a hyperemia of the abdominal organs, and says that treatment directed to the latter relieved the former.

It is evident from the above that the glycogenic function of the liver is influenced through the vasomotor nerves transmitted in the upper spinal cord and the sympathetic nerve. It does not matter for our purpose what the exact route of the impressions is beyond what has been stated. It is evident that any lesion in any part of the nervous system affecting this route directly or indirectly, may produce glycosuria. Such influence may be exerted upon the ends of the peripheral nerves as well as upon the nerve centres and great tracts. It may be supposed that the effect is to produce undue dilation of the bloodvessels of the liver, which allow a flow of blood through the organ too rapid to permit the normal conversion of glucose into glycogen. On the other hand, such rapid movement of the blood may overstimulate the reconversion of glycogen into glucose.

Let us now turn to other lesions which produce diabetes. Attention was attracted to disease of the pancreas associated with diabetes as far back as 1788, when Cowley published a case of severe diabetes associated with pancreatic calculus and atrophy of the gland. Since then an enormous amount of evidence has been accumulated from experiments on the lower animals and clinico-pathologic observations on men. For some time previous to 1890, experimenters endeavored to arrive at an explanation of this evident relation of pancreatic disease to diabetes mellitus by experiments consisting in extirpation of the pancreas, section of its nerves, and ligation of the pancreatic duct; but nothing was accomplished until in that year Von Mering and Minkowski took up the subject anew from the experimental standpoint. They showed conclusively that extirpation of the pancreas in the dog, cat, rabbit, toad, and frog produces, without exception, glycosuria with the other symptoms of diabetes mellitus-hyperglycemia, ravenous appetite, thirst, emaciation, and death. The condition is comparable to the severest cases of diabetes in man, because glucose continues to be excreted after all carbohydrate food is withdrawn. In the dog the proportion of sugar in the urine is very large—10 per cent. being not un

53

usual, while 16 and 22 per cent. is sometimes attained. As in the human being, toward the end, when the emaciation is extreme, the proportion of sugar falls, and diacetic acid, acetone, and oxybutyric acid appear in the urine. So, too, the maximum elimination of sugar appears with a carbohydrate diet, and grape-sugar is almost completely secreted as such. Even on an albuminous diet five and six percent is eliminated. Starvation diminishes it still more, but does not remove it. Partial extirpation of the pancreas of animals does not necessarily produce diabetes, but as larger and larger pieces are removed, diabetes appears, first in a mild form, which generally grows worse until it terminates fatally. Like results follow, though not invariably, injection of the pancreatic duct with warm paraffin, which congeals on cooling and leads ultimately to complete destruction of the gland. Mention should also be made of the fact that in experiments by Flexner and Pearce injection of gastric juice into the pancreatic duct was followed in thirty minutes by glycosuria.

As I have already said, the diabetes that succeeds extir pation of the pancreas is fatal. No other organ seems to be able to replace its function. How far are these results sustained by observations in man? The human pancreas has never been totally extirpated, but it has been partially removed, and diabetes or glycosuria has supervened, not always immediately, and in one instance referred to by Flexner not until one and a half years after operation. So, too, necropsy has disclosed lesions of the pancreas; but, unfortu nately for what seems at first a most promising theory, not every case of diabetes that has come to autopsy exhibits recognizable pancreatic disease. Lancéraux was the first to formulate any relation between diabetes and disease of the pancreas. He designated such cases as diabéte pancreatique, diabéte maigre, or lean diabetes, from the fact that its subject was usually lean or emaciated. Hansemann found alterations in the pancreas in 40 out of 54 cases, or in about 75% of the cases of diabetes which came to autopsy. Wind