INTRODUCTION The pathogenesis of mycosis fungoides appears to be one of the most perplexing subjects in dermatology. Its nosologic status is far from settled, and there is no complete agreement as to the proper name for the condition. The present classification of mycosis fungoides as one of the lymphoblastomas places it in a group consisting of vague manifestations of reticuloendothelial system disease not readily classifiable. Another confusing aspect is whether or not mycosis fungoides should be considered as a disease sui generis or merely as a clinical manifestation of leukemia, Hodgkin's disease, or lymphosarcoma. The source of confusion probably lies in the improper classification of cases, namely, the inclusion of mycosis fungoides "d'emblée"1 and the "erythroderma"2 types of mycosis fungoides with the classical (Alibert) form of mycosis fungoides. In my opinion, the d'emblée type is always lymphosarcoma, except for rare cases which