In previous communications1 from this laboratory descriptions have been given of apparatus devised for the purpose of making continuous intravenous injections at accurately controlled rates. The present paper deals with the application of this apparatus and the principles of timed intravenous injection to a clinical study of glucose tolerance in normal human individuals and in a series of pathologic conditions. In order that the work may appear in its proper relationship to that of previous writers the more important contributions to the subject may be reviewed. It has long been known that sugars administered by mouth in sufficient amounts cause melituria, but an attempt to learn the size of the dose which would accomplish this was first made by Worm Müller2 in 1884. Two healthy men on a diet free from carbohydrate received weighed quantities of glucose, lactose and saccharose and it was found that 50 gm. of