II. A similar result, even more striking, was obtained in a woman aged 42. She was treated medicinally for four months, during which time her red cells steadily decreased from 1,250,000 to 429,000 per cmm. She was then transfused with 400 cc. of blood, and her blood count rose immediately to 967,000. The rise continued steadily, and three months later her blood count was 3,690,000 per cmm. Two very small additional transfusions were given during this period, but to what extent these helped in the treatment cannot be estimated. The results in this case also are represented graphically by the diagram above. [Illustration: Fig. 5.--PERNICIOUS ANÆMIA, CASE III] III. A less favourable result is illustrated by the following history: A stores assistant, aged 47, had been ill for two years, and was first treated for pernicious anæmia in April, 1920. He was medicinally treated with arsenic, but no improvement followed. On June 18, 1920, his corpuscles numbered 1,060,000 per cmm. He was transfused with 600 cc. of blood, and his corpuscles increased at once to 1,840,000 per cmm. A month later there had been a further increase to 2,520,000, but this was not maintained, and nine months afterwards he was given a second transfusion of 500 cc. of blood. Immediately after this his red cells numbered 1,800,000 per cmm. (April 14, 1921). There was a further slight rise and then another rapid fall, so that on June 4, 1921, he had only 830,000 red cells per cmm. He was then given a third transfusion of 700 cc. The effect of this was a steady rise, and on June 17 he had 2,112,000 red cells per cmm. A fourth transfusion of 500 cc. was given at this point, and thereafter the improvement was maintained, with slight variations, until, on August 4, 1921, his corpuscles numbered 3,450,000 per cmm. In this case the effect of the two first transfusions was short-lived, but perseverance with the treatment brought him in the course of two months from an extremely serious condition to a state of comparatively good health, in which he could again for a time go about his business. The diagram illustrates well the rise which followed each of the later transfusions. He had again relapsed four months later, but, unless each transfusion had chanced to coincide with the remissions which may occur spontaneously in this disease, it seems clear that the treatment greatly relieved him for a time.