It is, indeed, now a recognized form of treatment for this disease, though the numerous reports upon results that have been published have not pronounced unanimously in its favour. Variability in results probably depends to some extent upon the difficulty of distinguishing true pernicious anæmia from some forms of secondary anæmia. It is hardly to be expected that much benefit would follow blood transfusion in the undiagnosed secondary type, since the destruction or loss of corpuscles is continuous until the cause has been removed. In true pernicious anæmia, on the other hand, there may be remissions in the disease, and it is quite clear that these may be initiated or prolonged by blood transfusion. The largest number of consecutive cases that has been recorded was treated in the Mayo Clinic in the years 1915 to 1918 (Archibald, Pemberton, Hunt). It was estimated that in about 60 per cent. of the patients with pernicious anæmia a definite improvement followed transfusion. It is generally agreed that the best results are seen in those who have not yet reached the last stages of the disease, though sometimes patients who are actually _in extremis_ will also show great improvement. A remarkable instance of this has been reported in Norway (261). A man, aged thirty-three, was dyspnœic, semi-conscious, and moribund when admitted to hospital. His red cells numbered 850,000 per cmm., and his hæmoglobin percentage was 19. Immediate improvement followed the transfusion of 900 cc. of citrated blood, the red cells rising quickly to 2,000,000 and later to 3,000,000. Twelve days after admission he was walking about. No case must therefore be regarded as hopeless, though disappointments must be expected. As a general rule blood transfusion should be given before the more serious secondary manifestations of the disease have shown themselves, that is to say, some time before the condition has become dangerous to life. Probably the disappointing results of this treatment have partly been due to the fact that it has been regarded as a last resort and has often been given at too late a stage. No rule can be laid down as to when transfusion should be given, but common sense suggests that it should be tried as soon as it is evident that the disease is progressing in spite of other methods of treatment. One authority (Anders) even advises that transfusions should be given as soon as an assured diagnosis has been made, but he weakens his case by adding that other methods of treatment should be used at the same time. If the patient is already seriously ill when first seen, the blood transfusion should be tried at once, as its effect, if beneficial, is likely to be more rapid than that of any other form of treatment.