A second objection is the extent of the injury which is necessarily inflicted on the donor. His radial artery must be exposed through an incision of considerable length, and must be ligatured at the conclusion of the process. The operation becomes, therefore, a matter of some moment to the donor, who will be permanently scarred, and can under no circumstances be used for transfusion more than twice. A third objection is that the transfusion cannot be done with due regard to the condition of the patient. A delicate and difficult operation has to be performed with the donor and recipient lying side by side, their arms close together. It is therefore almost imperative that both should be on operating-tables of a convenient height. Often, however, with an exsanguinated patient it is very important that he should not be moved from his bed, but as a bedside operation direct transfusion becomes difficult indeed! A final objection is that in some people the radial artery is of very small calibre, so that when all preparations have been made, and the artery exposed, it is found to be quite impossible to proceed. Another element of uncertainty is thus introduced. There is, therefore, little to be said in favour of direct transfusion, and much to be urged against it. This method has, indeed, in my own opinion, come to be of historical interest only. For this reason the different methods have only been very briefly described. For more detailed information, reference must be made to the various original communications, which will be found in the Bibliography. =Indirect Transfusion.=--The methods of indirect transfusion may be divided into those which depend upon the use of an anticoagulant mixed with the blood and those in which the blood is given unaltered. The technique of either process is simple compared with that of direct transfusion, though any method which makes use of whole blood can never be quite as free from uncertainty or difficulty as one which introduces the use of an anticoagulant. If the blood is prevented from clotting, the chief cause of failure in performing blood transfusions is removed. With any whole-blood method of transfusion speed is exceedingly important, frequent practice is a very great advantage, and it is essential, as with direct transfusion, that the donor and recipient should be in close proximity to one another, if not actually side by side. On the other hand, the use of an anticoagulant renders speed and frequent practice of less account. The blood can be drawn, and can then be put on one side until the best moment for giving it has arrived. Due regard may be had to the patient’s condition, since the blood can be carried about and can be given at leisure to the patient in his bed without disturbing him and almost without his knowing it. The donor, too, is not exposed to the mental shock of lying for some time side by side with a patient who may be _in extremis_, or may even expire during the operation.