The results of a blood transfusion upon a patient suffering from acute anæmia are, indeed, amongst the most dramatic effects to be obtained in the whole range of surgery. Within a few minutes of its commencement the whole aspect of the patient alters. His respiration becomes deep and regular, his restlessness disappears, colour returns to his face, his pulse rate falls, and he begins to take an intelligent interest in his surroundings. These changes taking place within a period of fifteen minutes may well strike an onlooker as little short of miraculous. Shortly afterwards the patient may fall into a natural sleep, a sure sign that the normal circulation has been restored to the exhausted central nervous system. In considering how much blood should ordinarily be given in the treatment of acute anæmia, experience is a safer guide than any theoretical considerations. Nevertheless, it is worth while to inquire briefly into the experimental and theoretical basis upon which the treatment of acute anæmia rests. It is difficult to estimate accurately the total quantity of blood in the body of an adult, but it has been variously stated by physiologists to be from a twentieth to a tenth part of the body weight, or, in liquid measure, from 3 to 6 litres (approximately 5 to 10 pints). This has been estimated in several ways, the results of which show some discrepancy. A figure approaching the higher one was obtained long ago by the direct method of washing out the blood from the bodies of executed criminals. Recently it has been claimed by Haldane that these determinations were inaccurate; by means of his carbon monoxide method, with the details of which we are not concerned here, he has estimated that the blood volume is but one-twentieth of the body weight, or in very stout persons is even as low as one-thirtieth. Still more recently Haldane’s estimation has been challenged in its turn by observers who have injected a dye into the circulation and have then determined its degree of concentration in the blood by means of colorimetric comparisons. It is evident that if the dilution which occurs when a known quantity of dye is injected can be accurately estimated, then the total volume of circulating fluid can be calculated. This method could not be used until a non-toxic, non-diffusable dye had been discovered, but it was found in 1915 that “vital red” fulfilled these requirements (143). The results obtained in this way show that those originally given by the direct method were substantially correct. The blood volume was found to vary from 1/13 to 1/10·5 of the body weight; on the average it amounted to 5,350 cc., or 85 cc. per kilogram of body weight. These observations have been in their turn criticized (114), but only to the extent of reducing the amount by 1/10. It may therefore be assumed that, according to the most recent work, the blood volume is from 5 to 6 litres, or, approximately, 8 to 10 pints.