We have applied the term agranulocytosis to the condition in which there is complete or almost complete absence of granular leukocytes, accompanied by leukopenia and relative increase but, in most instances, absolute decrease of lymphocytes. The term hypogranulocytosis, which was first suggested by Weiss,1 we have applied to that condition in which there is less marked reduction in the number of granular leukocytes and well marked leukopenia, without absolute increase in the number of lymphocytes but with relative lymphocytosis. These terms are not applied to such distinct clinical entities as acute aplastic anemia, acute leukemia with normal or reduced total leukocyte count and agranulocytosis produced by known poisons such as benzene and roentgen rays; for the purposes of this discussion, however, they are applied to the peculiar blood picture associated with various types of infection. Brown2 seems to have been the first to report a case of pharyngeal