As in the early days of transfusion, so at the present time, a considerable proportion of the patients that need transfusion will be met with in the course of obstetrical practice. It has often been remarked how much blood can be lost by a woman following the delivery of her child without any serious result; nevertheless, many deaths are occasioned every year by post-partum hæmorrhage, placenta prævia, and rupture of an ectopic gestation. Sometimes the bleeding is so rapid that there is no margin of time available for a transfusion unless all the facilities be immediately at hand. Short of this, transfusion is the ideal treatment, and the problem is a simple one, the relief of acute anæmia being the only object in view. One interesting modification of the procedure has been recently recommended by German writers, namely, the reinfusion of the patient’s own blood. This is applicable only when the hæmorrhage has taken place into the peritoneal cavity, and is therefore limited to the treatment of a ruptured liver or spleen, a ruptured uterus, or a tubal abortion. With a ruptured uterus the sterility of the blood is not assured, and this condition were better not included. For the other conditions Lichtenstein recommends that the blood should be ladled out of the peritoneal cavity into Ringer’s solution and then strained to remove clots. The resulting fluid is infused into a vein. Judging from my own experience of intraperitoneal hæmorrhage, not much blood would actually be recovered in this way, since usually so much of it has clotted. In any case, the whole procedure is to be looked upon with suspicion owing to the unknown and probably profound changes that have taken place in partially clotted blood. Eberle records that in one case reinfusion was followed by hæmolysis, and among twenty-one cases reported by Schweitzer in 1921, one death was attributed to the reinfusion, which, as in Eberle’s case, was followed by hæmoglobinuria. Transfusion has also been used for the toxæmias of pregnancy, but this will be dealt with under another heading. CHAPTER III INDICATIONS FOR BLOOD TRANSFUSION--_continued_ HÆMORRHAGIC DISEASES It is claimed that blood transfusion provides an efficient means of treatment in most conditions distinguished by symptoms of spontaneous hæmorrhage or by traumatic hæmorrhage which cannot be controlled. All such diseases have the common features that the coagulation time of the blood is abnormally prolonged, and it may be supposed that the transfused blood supplies some missing constituent, so that for the time the blood is enabled to coagulate more normally. Most of the evidence available shows that the claims made for transfusion are not exaggerated.