We had treated but a few cases when the report of McGuire and Redden appeared. These observers working in the Naval Hospital at Chelsea, Mass., presented very excellent results in the use of immune serum from convalescent influenza cases in the treatment of pneumonia. They reported 30 recoveries out of 37 cases, with 1 death, and 6 cases still under treatment at the time of their report. This form of treatment began at Chelsea on September 28, 1919. In Texas, on October 15, Brown and Sweet gave two cases of influenzal pneumonia citrated blood from convalescent influenza patients. Their two cases recovered. Our published results, although not showing such excellent figures as from the Chelsea observers, agree very well with their work. Since that time a number of confirmatory reports have been brought forward. Ross and Hund have shown that this method has been of value in their hands, and recently a further statement from McGuire and Redden tends to confirm their first views as to the value of immune serum from convalescent patients. Their last report giving a mortality of 6 in 151 cases of pneumonia cannot be other than positive proof of the value of this method of treatment. As the technical side of the work has been given in several articles, we hardly think it necessary to again review it in detail. A few phases should, however, be recalled. It would seem that either serum or the whole citrated blood may be used. Solis-Cohen and his group of workers believe that whole blood has stronger bactericidal properties than defibrinated blood or the plasma. But yet one cannot complain, even on a theoretical basis, against the results obtained with serum by McGuire and Redden. The use of whole blood increases the detail of the procedure, in that the agglutination reactions must be estimated. Unfavorable results in this regard also naturally cut down the supply of available donors. In a military hospital a dearth of donors does not arise, but in civilian practice the problem is very different. In our work we never gave more than 100 cc. of whole blood; usually the amount varied between 50 cc. and 75 cc. On account of the small amount we felt that isoagglutination would not be a serious factor, and in more than 200 injections we failed to see any evidence of ill results from this source. Giving up to 500 cc., as was done by Ross and Hund, is probably a different affair, and accurate agglutination tests are essential. We feel that if the case is treated sufficiently early in the disease as much good can be shown to occur after 50 cc. as after 100 cc. of blood. We do believe, however, that the pooling of sera, where one is able to carry out this method, as it means a liberal supply of donors, is really the method of choice. Syphilis must be ruled out, both clinically and serologically.