The demonstration of oxydase and peroxydase ferments1in certain cells of bone marrow origin has led to the extensive application of these reactions as diagnostic aids in the differentiation of myeloid and lymphoid leukemia. This is particularly true of the acute leukemias, in which undifferentiated progenitors of the granular and nongranular leukocytes form a conspicuous part of the blood picture. The popularity of the oxydase reaction is largely due to its supposed ability to distinguish, by simple color reactions, primitive blood cells which, by the ordinary methods of staining, are exceedingly difficult, if not impossible, to differentiate. The diagnostic value of the reaction in a given case naturally depends on its specificity for the cells in question, a point on which there is considerable difference of opinion. To those hematologists who regard the oxydase reaction as specific for myeloid cells, the presence of these ferments represents a biochemical difference by