I N considering insulin tumours of the pancreas it is interesting to recall the steps by which medical knowledge advanced until it became possible to diagnose these tumours, and to note what a considerable time elapsed between some of those steps. In 1869 Langerhans, while still a medical student, wrote a thesis in which he described the islands of cells scattered through the pancreas which have since then born his name. Naturally he did not guess their function. V. Mering and Minkowski in 1899 showed that diabetes mellitus could be produced in dogs by extirpation of the pancreas. In 1902 Ssobolew showed that ligature of the ducts caused atrophy of the acinar cells but not of the islets, and failed to cause diabetes. Their experimental work laid a sound foundation for the theory that diabetes was caused by the lack of an internal secretion produced in the islets of Langerhans, but in spite of many attempts by many workers, stimulated by Murray's introduction of thyroid extract in 1891, it was not until 1922 that Banting and Best isolated insulin and made it available for the treatment of diabetes. Soon after this it was found that an overdose of insulin would produce convulsions and coma associated with a low blood-sugar level, and that the prompt administration of glucose cut short these attacks. A little later cases of spontaneous hypoglyczmia were recognized (Harris, 1924), though none of Harris's three cases appear to have had insulin tumours, if one may judge by their symptoms ; and occasionally pathologists described tumours of islet origin. It was not until 1927 that Wilder, Allan, Powers, and Robertson (1927) reported a case of hypoglyczmia due to a malignant islet-cell tumour which was verified by operation. In 1929 Roscoe Graham excised the first insulin adenoma and cured his patient, though the growth was suspected of malignancy (Howland, Campbell, Maltby, and Robinson, 1929). Since then something over 400 cases of insulin tumours have been reported, of which roughly 90 per cent have been benign and 10 per cent malignant. As the main interest of these tumours lies in their diagnosis, I will describe the clinical histories of 5 cases of benign insulin adenoma that I have had under my care, with one dubious case and one case of malignant insulin tumour.