It has been said that the lower end of the delivery tube reaches into the angle between the side and the bottom of the flask. When therefore the flask is nearly empty, it should be tilted so that very nearly the whole of the blood can be forced up the tube. When the flask is quite empty, the blood in the barrel of the air-lock must be carefully watched, and when its level has fallen to the bottom of this, the clip must be applied to the tube above the cannula. By this means no blood is wasted except the small quantity which remains in the tube below the air-lock. As soon as the tube has been clipped the cannula is withdrawn, the vein is ligatured above the opening into its lumen, and the edges of the skin incision are sutured. Transfusions carried out in this way can be performed with uniform success. The technique is simple and straightforward at every stage, and can be easily demonstrated and learnt. It is, in addition, a perfectly clean process, and not a single drop of blood need be spilt. Any method which involves the injection of blood under pressure is open to the objection that it is possible to overlook the fact that the flask has been emptied and to kill the patient by injecting air into his veins. This can, however, only happen as the result of great carelessness on the part of the operator. The presence of the air-lock affords an additional safeguard, as it can hardly escape the operator’s notice that blood has ceased to flow from the nozzle of the delivery tube. [Illustration: Fig. 13.--INJECTION OF THE BLOOD, SHOWING USE OF AIR-LOCK] The method may also be criticized on the ground that some damage is caused to the corpuscles of the donor’s blood by the shaking which is necessary to mix it with the citrate solution. This objection is, in my opinion, theoretical rather than practical. If, however, it be desired to avoid any such shaking, the apparatus designed by A. E. Stansfeld and described by him in 1918 may be used. This ensures that the citrate and the blood flow into the containing vessel together, so that no further mixing is needed. The apparatus is more cumbrous, more fragile, and less easy to clean and to sterilize than that described above. In the hands of an expert it will give excellent results, but its use requires some little practice, and it is therefore not so well adapted for general use. The whole of my own apparatus, as described above, may be obtained from Messrs. Allen & Hanburys, Wigmore Street, London, W.1, who also provide a convenient box for carrying it.