The affinity of Taccaceae has been subject to diverse opinions, many authors favouring a place near Dioscoreaceae, but in my view the unisexual flowers, the branching habit, the racemose inflorescences and the 3-celled ovary in that family make this not very probable. I prefer to share the opinion of those who seek its affinity to Amaryllidaceae, because of the habit, the scape-shaped inflorescence, the umbellate flower disposition with an involucre, and the fact that in that family also occasionally a 1-celled ovary is found. Neither the systematic anatomy nor the inadequately known phytochemistry or chromosome number are sufficiently diagnostic to support opinions on affinity. Before my precursory revision (1972) two genera were distinguished, viz Tacca and a monotypic genus, Schizocapsa (S. plantaginea HANCE) ranging from Thailand to Kweichow, which differs only by having dehiscent capsular fruits instead of the indehiscent ones in Tacca. I have seen since that this species has dehiscing fruit indeed, which removes my doubt (l.c. 370) on this point. I have tried to see whether there may be a tardy dehiscence in Tacca by keeping fruit stalks of T. chantrieri in erect position in the Leyden greenhouse, but this had no result. I see no reason to keep Schizocapsa as a separate genus or infrageneric taxon.