We must not forget the very important fact, that even when we find no sugar we cannot at once exclude diabetes. There are many cases of mild diabetes which do not show any sugar on an ordinary diet. To recognize these cases we must give them a test meal of either about three ounces of grape sugar two hours after a light breakfast of coffee and a roll, or, as we have recommended in our book on the rational treatment of diabetes,[131] we must give a test meal of much starchy food, such as rice and cakes, and examine the urine two hours later. If then no sugar is found, or not exceeding 0.2 per cent., diabetes may be excluded. Persons with 0.1 per cent. to 0.3 per cent. of sugar after such a starchy test meal may be considered as on the border line of diabetes. Such persons show a diminished tolerance of carbohydrates, and they have lost the power to consume all the sugar they introduce into their body. Therefore, especially if they come of diabetic parents, they should avoid eating much starchy food so as not to develop further this dangerous tendency. Footnote 131: Second edition, Berline, 1909. Diabetes, in the majority of cases, considerably shortens the patient’s life. Mild cases of diabetes may however live for fifteen or twenty years, or often even longer. In predicting the chances of longevity in diabetic subjects everything depends on the question whether diacetic acid has been found in the urine or not. This can be easily ascertained by putting a few drops of a solution of perchloride of iron into a test tube with the diabetic urine. If diacetic acid is present, the liquid will become Burgundy red in color, and the deeper this red color, the greater is the percentage of diacetic acid present. In such cases the chances are very bad, such patients living on an average for only one to two years. Exceptions to this rule, however, are not infrequent, and we know of a case of acromegaly with severe diabetes, who has passed large quantities of diacetic acid for the last fourteen years. Cases of severe diabetes, with diacetic acid, can often be easily recognized by their appearance. They are thin, the cheeks are often flushed, and they show nearly all the signs of a myxœdematous condition, as already described, whereas cases of mild diabetes present often symptoms of hyperactivity of the thyroid gland, as in exophthalmic goiter. As we have shown, diabetes is often due to an overactivity of the thyroid gland,[132] as well as to degeneration of the pancreas, these glands being antagonistic to one another. This is confirmed and amplified by Sajous, who ascribes one form of diabetes to overactivity of the adrenal system, which includes the thyroid gland. Overactivity of the latter organ may also be followed by its exhaustion, with symptoms of myxœdema, as is the rule in severe diabetes. Footnote 132: “Die Entstehung der Zuckerkrankheit,” Berlin, 1903, and its French translation, Maloine, Paris, 1904.