The ideal method of blood transfusion seems to me to require that it shall be absolutely certain of success, that the blood shall not necessarily be injected into the patient immediately it has been drawn, so that other circumstances besides the demands of the transfusion operation can be considered, and that no injury shall be done to the donor beyond the puncturing of a vein. In addition to this, the method should be so simple and free from special apparatus that it can be easily learnt and carried out by one operator without skilled assistance. All these requirements are fulfilled by the citrate method, and a satisfactory method of performing this will next be described. As will be seen, the blood can be drawn with the minimum amount of injury to the donor; when drawn, it can be put on one side, for several hours if necessary, and then given to the patient at whatever may be judged to be the most favourable moment; the whole process can be carried out by a single operator without any assistance; and finally, but little practice is needed to make success certain every time. The transfusion apparatus known as “Robertson’s bottle,” first described by Oswald Robertson in 1918, is the basis of most citrate methods. This could be easily improvised in a field laboratory, and was extensively used during the last year of the war. The apparatus consisted of a glass bottle of about a litre capacity, the mouth of which was closed by a rubber bung. Through the bung three glass tubes passed. One, connected by a short rubber tube with a wide-bore needle, ended about an inch from the bottom of the bottle; through this the blood flowed into the bottle. A second tube, which reached to the angle between the side and the bottom of the bottle, was connected by a rubber tube with a cannula; through this the blood was injected into the patient. The third tube reached only just beyond the bung, and to this was attached a Higginson’s syringe, by means of which either negative or positive pressure would be produced inside the bottle, according to which end of the syringe was attached.