When treating diabetes with insulin, one has to replace the regulated continuous secretion from the normal pancreas into the portal vein by a few daily injections into the subcutaneous tissue. This method is of course only a poor imitation of nature's own mechanism. Serious disturbances are avoided only because the organism has other means of regulating the blood sugar than to vary the rate of insulin secretion. This is the reason why treatment with one or two injections daily is found to give fairly good results in many cases. In more severe cases, however, very pronounced oscillations of the blood sugar may occur. During the last few hours before the injection metabolic disturbances develop, which tend to minimize the available deposits of carbohydrate, strongly needed when the blood sugar runs low after the injection. It is obvious that the insulin hydrochloride will rapidly become absorbed into the blood after the