It seems desirable to emphasize the following conclusions:
1. A careful balancing of the normal acids and bases of the urine makes it possible not merely to detect the presence of organic acids in the urine, but also to determine approximately the amount of such acids. The method recently described by Herter and Wakeman can be recommended as securing a greater degree of accuracy, for the amount of labor involved, than any other procedure.
2. The determination of the N of NH3 is a useful procedure for clinical purposes, since it is probably true that a considerable excretion of organic acid (say 15 gm. oxybutyric or more in 24 hours) is always attended by an increased excretion of NH3. As much organic acid as corresponds to 10 gm. oxybutyric acid may be excreted in 24 hours without causing an increased excretion of NH3 (Case IX). We cannot therefore rely on the ammonia output to detect moderate quantities of organic acid.
3. Where organic acids are removed in considerable amount without increasing the excretion of NH3, the acid takes out other alkalies, probably in some instances chiefly K.
4. In cases of diabetic coma the urine always contains a large excess of organic acids and the N of NH3 is usually increased to 18 to 25 per cent of the total N.
5. Crotonic acid can regularly be obtained from the urines of patients in diabetic coma.
6. The condition of diabetic coma is preceded by a period of days, weeks or months, in which there is a large excretion of ß-oxybutyric acid (20 gm. or more in 24 hours), and in which the N of NH3 is largely increased.
7. Patients whose urines show or have shown a large excretion of organic acids are in danger of developing diabetic coma, but the N of NH3 may temporarily rise as high as 16 per cent and yet coma may be delayed for more than 7 months (Case VII). The persistent excretion of more than 25 gm. of ß-oxybutyric acid indicates impending coma.
8. A patient passing 30 gm. of ß-oxybutyric acid in 24 hours may still have enough energy and strength to be about all day and perform considerable muscular work (Case X).
9. A patient who has been excreting very little organic acid and has gained weight may within a few months show the presence of considerable quantities of organic acid, and die in typical diabetic coma (Case VII).
10. When the urine contains little or no organic acid there is no immediate prospect of diabetic coma, but patients with such urine are probably liable to most of the other dangers that threaten diabetic patients. The relation between the degree of acid intoxication and the susceptibility to infection seems worthy of special experimental study.
11. Where the urine regularly contains more than 200 gm. of sugar per day there is usually considerable organic acid in the urine and large amounts of acid, indicative of coma, are invariably accompanied by considerable or great glycosuria.
12. Sometimes there is much sugar and little or no acid in the urine, and sometimes there is considerable acid and little sugar. These facts render it desirable to examine the urine of diabetic patients at least once a month with reference to the amount of acid excreted, for the element of acid intoxication must be clearly separated from the element of glycosuria in our study of the progress of a case. In other words, we must recognize the acid intoxication as an important and sometimes as a dominant factor in the prognosis, and this element should be regarded even in those cases of diabetes which have the clinical indications of a mild type of the disease. We may thus hope to prolong life in many instances by taking precautions, as to diet and out-of-door life, which might not otherwise be deemed necessary.
13. The withdrawal of carbohydrate food frequently leads to a considerable reduction in the quantity of organic acids excreted. The reason for this is not yet clear and the phenomenon deserves careful study.