Two centuries have elapsed since Werlhof originally described a case of purpura haemorrhagica, but the advances made in our knowledge of the disease have all occurred in the past fifty years. Its clinical manifestations remain as he so vividly described them, though little is known, even today, about the mechanism of their production. Its characteristic blood picture has become familiar. It is well known that many cases are closely associated with the factor of infection, that some are caused by toxic agents and that others occur in the course of diseases affecting the hematopoietic and reticulo-endothelial systems. Some of these cases can be treated effectively. There is, however, a group of cases in which the etiology remains entirely obscure. Strangely enough, it is this group that can be treated with the greatest assurance of success. For many centuries, physicians have recognized the existence of hemorrhagic diseases. It was not until