Another possible complication is introduced into the treatment by the necessity for giving repeated transfusions. It has been noticed that sometimes a serious reaction follows one or more of the later transfusions of a series, even when the blood is taken from the same donor who had been used before without ill effects. A report on several such cases shows that this form of reaction cannot be predicted or eliminated by the most careful testing beforehand for reactions between the patient’s serum and the donor’s corpuscles, though it has occasionally been so severe as actually to hasten the patient’s death (34). This fact suggests that the reaction is not due to the presence of agglutinins, but is rather of the nature of an anaphylactic shock, the patient having been sensitized by a trace of foreign protein introduced in the blood on the earlier occasions. Possibly it may be to some extent avoided by not using the same donor if another is available. It also emphasizes the necessity for giving the blood slowly and cautiously, so that the transfusion may be stopped at the first sign of a reaction in the patient. Very large numbers of transfusions for pernicious anæmia have been given in the past, yet a reaction of a dangerous severity has occurred in but few of them. This need not, therefore, be regarded as a contra-indication for transfusion, but rather as an indication for circumspection in giving it. Transfusion is clearly a therapeutic measure of great value. Very recently it has been claimed by Waag that excellent results have been obtained by the repeated _subcutaneous_ injection of small doses (5 cc.) of whole blood. In an actual case which he reports, nine injections were given twice weekly. If the claim be substantiated by further successes, this method of treatment may eventually supplant the more elaborate process of actual transfusion. TOXÆMIAS _Bacterial Infections_