This genus, like Anadendrum and Heteropsis, is intermediate between the subfamilies Pothoideae and Monsteroideae as constituted by Engler (1920, p.63). When Schott originally described the genus he placed it next to Anadendrum (Pothoideae). Engler (1879, p. 100) left it in the Pothoideae next to Heteropsis which followed Anadendrum. Later, Engler (1908, p. 118) transferred Amydrium to the Monsteroideae as a monotypic genus. In the same publication (p. 1) Engler established a new genus, Epipremnopsis, which he placed next to Anadendrum in the Pothoideae. In this treatment I propose to unite Epipremnopsis and its three species with the hitherto monotypic genus Amydrium. The species concerned are rather different from each other. Study of them in the wild and the herbarium suggested that they were congeneric. In a recent survey of aroid floral anatomy (Eyde et al., p. 481, 486, fig. 9—15, 27) it was found that Amydrium humile and Epipremnopsis media have virtually identical unilocular ovaries with a single, deeply intrusive placenta (since found in E. magnifica and zippeliana). This striking condition has arisen from a bilocular ovary by abortion of the opposing placenta. Among the other aroids this character in known only in the genus Epipremnum and is the fundamental difference separating it from its near relatives, Rhaphidophora and Scindapsus Amydrium is distinguished from Epipremnum by its lack of needlelike trichosclereids which are abundant in the flowers and fruits of Epipremnum.