The hemorrhagic diathesis of chloroform and phosphorus poisoning is without doubt due to a deficiency of fibrinogen in the blood and that of a type of melena neonatorum to a retarded rate or complete failure of blood coagulation. The hemorrhages of jaundice and hemophilia can perhaps be accounted for by abnormal blood coagulation, although proof of this is not so clear as in the former diseases. The disease with which I shall deal in this paper, namely purpura hemorrhagica, has apparently an entirely different pathogenesis and, it is believed, is due wholly or partly to an almost complete absence of platelets from the blood. That the number of platelets in the blood is reduced in purpura hemorrhagica was first observed in 1887 by Denys, a histologist, and later by Hayem, the discoverer of blood-platelets. Denys1 reported three cases in which platelets could hardly be found in fresh blood films.