I1 recently reported a case in which intramuscular injection of approximately 8 cc. of antistreptococcic serum given in fractional doses over a period of three hours failed to protect the patient against the intravenous injection of less than 10 cc. of the same serum, death resulting from an accelerated serum reaction. In reviewing the literature, two other instances of failure of attempted desensitization were noted. In one case, reported by Koch2 in 1915, 5 cc. of serum given subcutaneously five hours before the final injection failed to protect the patient against a fatal outcome. In the second case, reported by Blankenhorn3 in 1925, seven doses of serum beginning with 4 minims (0.25 cc.) and ending with 5 cc. were given at two hour intervals, and yet in spite of this 5 cc. of serum given intravenously produced a severe and almost fatal reaction. These experiences seemed to