In 1916 Kaznelson1 recommended splenectomy for the cure of essential thrombocytopenic purpura; since then many reports have appeared substantiating the excellent results obtained by him. In the last few years, however, cases in which bleeding continued after splenectomy have been reported by various observers, for example, McLean and his associates,2 Kasdoba,3 Lautier, Oberling and Woringer,4 Bloomfield,5 Jones and Tocantins6 and Greenwald.7 In addition, Whipple8 reported a series of cases in which a high rate of mortality (87 per cent) resulted because the spleen was removed during the acute stage of the disease; in his report he strongly advised against splenectomy during the stage of active bleeding. Transfusions, too, while they undoubtedly are life-saving measures in cases in which the bleeding is marked and leads almost to exsanguination, frequently have no effect in preventing subsequent hemorrhages; this was true in one of the