Hallervorden interpreted the increased excretion of ammonia in diabetes to signify the presence of some acid that was hitherto unknown. Later Stadelmann isolated crotonic acid from diabetic urine and later Minkowski and Kulz succeeded in isolating B-oxybutyric acid itself from diabetic urine. Since that time the acid has been found in a great many other diseased states, as, for instance, in scurvy, in certain infectious diseases, in carcinoma, in typhoid and dysentery. The appearance of this acid in the urine is by no means a rare event. Owing to the fact that this acid is so closely related chemically to acetone and diacetic acid, one is justified in suspecting its presence in the urine whenever these two bodies are excreted in considerable quantities. As a matter of fact, one always succeeds in finding the acid under these circumstances. (Magnus-Levy, Gerhart and Schlesinger, Kraus, and myself).

The highest values in the excretion of the acid are seen in diabetes, particularly before and during coma, and it is precisely in this state, as we know, that the highest values for acetone are also obtained. This fact has led to the theory that diabetic coma is due to the accumulation of acids in the body (acidosis), and it is upon this theory that the alkali treatment of diabetic coma is based.

As a matter of fact, however, B-oxybutyric acid is just as little toxic when taken by mouth as acetone or diacetic acid. This has been demonstrated by numerous experiments in which the acid was administered to healthy subjects and to sufferers from diabetes. In depancreatized dogs with diabetes it is also impossible to produce coma or comatose conditions by the administration of very large quantities of B-oxybutyric acid. The deleterious effect of B-oxybutyric acid in coma as a matter of fact, is not due to any specific toxic properties of B-oxybutyric acid but to its acid character in general and to the power it possesses, as an acid, of withdrawing alkali from the organism. Whether only fixed alkalies are removed in this way or whether, as Magnus-Levy claims, the basic components of the proteids are removed, so that the latter in this way become unfit to carry on the life processes, remains to be determined. In view of the fact that the alkalescence of the blood is unquestionably greatly reduced in diabetic coma, it is probable that all these factors are operative at the same time.

The appearance of acetone bodies in the different pathological conditions that I have just enumerated, and that I will sketch in detail below, is rendered more comprehensible when we study the factors that determine the appearance of these bodies in the urine under otherwise normal conditions. Above all, the concep tion must be abandoned that the secretion of acetone is under all circumstances a morbid phenomenon. This erroneous conception must be made responsible for the formulation of a variety of forms of so-called acetonuria that are more theoretically constructed than well-founded clinically. As our knowledge of the physiology of acid formation in the organism has increased, the old disease pictures that were originally sketched by Peters and Kaulisch, Cantani and von Jaksch have been abandoned. In the first place, it was found that normally the secretions of the body contained a small quantity of acetone that varied within narrow boundaries. The urine in a healthy subject fed on an ordinary diet contains from one to three centigrams of acetone in 24 hours. A still larger quantity (60 to 70 per cent of the total acetone) leaves the body through the expired air (J. Müller), L. Schwarz, Gelmuyden. The amount of acetone excreted varies at different hours of the day, and during the night time a larger quantity is excreted than during the day. This indicates that taking food exercises an influence on the excretion of acetone. As a matter of fact, the excretion of acetone rises considerably during inanition. Fr. Müller found 40 times as much acetone on the seventh day of starvation as before the period of fasting. In addition, diacetic acid and oxybutyric acid are also excreted in the urine if the period of fasting is still further prolonged (Minkowski and others), whereas, as we know, the latter bodies are always absent in the normal secretions of the body. In addition Biermer showed that the excretion of acetone increased in diabetes if the subjects are put upon a meat-fat diet; and here, too, diacetic acid is occasionally excreted, and moreover in a healthy subject living on a diet consisting exclusively of meat and fat, the same phenomenon is observed. The explanation of this phenomenon seemed easy, for, it was argued that if acetonuria from fasting was due to the diminished ingestion of food, then the same cause (underfeeding) must be made responsible for the acetonuria occurring on a meat-fat diet; for, as we know, the caloric value of the food is usually too small to maintain full nutrition if the patients are fed on meat and fats exclusively.