1. Numerous diabetics living on a diet containing at least 60 to 80 grm. of carbohydrate and not passing any sugar on this diet, do not excrete more acetone than a healthy individual. As soon as such cases are placed upon a rigid diet the excretion of acetone increases and diacetic acid also appears in the urine. If the strict diet is continued, the excretion of acetone decreases again and the ferric cloride reaction also disappears.

2. In another group of cases of diabetes of medium severity or of severe diabetes in its initial stages the excretion of acetone amounts to as much as one grm. a day on a diet containing more carbohydrates than these patients can tolerate. At the same time, there is no diacetic acid, or only traces of it, and no oxybutyric acid. In these cases, too, a strict diet causes an increases of the acetonuria, at the same time that the excretion of diacetic acid and oxybutyric acid is also increased. This excretion of acetone bodies may persist if the rigid diet is continued, or may even become greater; or, again, may, as in the first class of cases, be decreased in the course of a few days.

3. The third group of cases comprises patients who show all the symptoms of so-called severe glycosuria; the peculiarities of this group are, however, also occasionally seen in patients in whom the amount of glycosuria, according to our ordinary terminology, is slight, i. e., the glycosuria disappears as soon as the carbohydrates are withdrawn. In this group large quantities of acetone and of oxybutyric acid are continuously formed, and finally so much of these bodies accumulates that diabetic coma may develop. At the same time numerous fluctuations occur in these cases, so that the quantity of acetone bodies may suddenly decrease without determinable cause and then increase again after a time until the same amount as before is excreted.

All these different varieties of acidosis can be greatly influenced by the diet. The carbohydrates play the same role in diabetic acetonuria as in the other forms (see above, Hirschfeld). with this difference, however, that in diabetes the conditions are so complicated that the effect of the diet cannot be determined so readily as in healthy subjects or in patients with fever or with some other disease than diabetes.

The following important points may be enumerated, and they agree essentially with the facts discovered in other forms of acetonuria (see above).

As the tolerance for carbohydrates increases, the excretion of acetone bodies decreases, so that on the same diet not only less sugar but also less acetone, etc., is excreted. We believe that under these conditions more carbohydrate is burnt and consequently the oxidation of acetone bodies is favored.

As the tolerance for carbohydrates decreases, exactly the reverse is seen.

If large and increasing quantities of carbohydrate are administered the excretion of acetone and the other symptoms of acidosis can be reduced, at least for a time; for in nearly all diabetics a certain proportion of the increased quantity of carbohydrates that is administered is destroyed. The oxidation of this part of the food consequently decreases the acidemia. There are certain cases in which this factor can be utilized to advantage. In other cases, again, such treatment would do great harm as far as the primary diabetic taint is concerned.

In those rare cases in which all the carbohydrate that is added to the food reappears in the urine, the acetonuria is not influenced in any way by the addition of carbohydrates to the food.

The fact, furthermore, that in no other disease are even approximately so great quantities of acetone bodies found in the blood and urine, etc., agrees fully with the experience we have in regard to acetonuria in general, for in diabetes mellitus precisely those cells are functionally damaged that are expected to aid in the disassimilation and oxidation of the acetone bodies (see page 66). The combustion of the carbohydrates is reduced partly because the function of these cells is inhibited or destroyed, partly because we administer as little carbohydrate as possible in order to influence the actual condition of the patient or to prevent further trouble in the future. In those cases in which really very large quantities of acetone bodies are formed, the carbohydrates of the food are not destroyed; in addition, moreover, only very little of those carbohydrates undergoes this disintegration that is derived from other substance (albuminoids, fats). This occurs in no other disease, and consequently the values for acetone are never so high as in diabetes.

Although diabetic acetonuria follows the same rules as acetonuria from inanition in all these respects, there are, nevertheless, certain exceptions and peculiarities in diabetes that cannot readily be explained; consequently we must for the present have recourse to hypothetical considerations.

One such exception, for instance, is that only a slight decrease in the acetonuria may be seen in certain cases, in which the utilization of the carbohydrates is greatly increased. We showed above that acetone is not excreted in a healthy person if 80 to 100 grm. of carbohydrate are absorbed and assimilated every day. In cases of pneumonia, moreover, with high fever and in patients suffering from gastric ulcer who receive no other nourishment than 100 to 200 grm. of dextrose, the same fact can be determined (see above). In a number of diabetics, on the other hand, who really disintegrate (not only absorb, but actually assimilate) 120 to 150 grm. of carbohydrates, the excretion of acetone bodies may be found to remain very abundant. Another peculiarity that is difficult to explain is the fact that a diabetic excretes such different quantities of acetone bodies at different times, even though no change is made in the character of the food and no change occurs in the excretion of dextrose (in other words, even if the carbohydrate metabolism remains unchanged). The following is one of the many examples of this kind that I have on record: