SummaryThe male flower ofHedyosmum(Chloranthaceae) has been described by all authors for almost two centuries as unistaminate, naked, and ebracteate. It is not so at all, but a flower that bears up to several hundred stamens, spirally arranged along an axis. It is an extraordinary structure in the angiosperms, precisely a gymnosperm cone‐like male flower and closely associated with many other primitive characters (in particular the pollen that is known by a similar monosulcate type as early as the lower Cretaceous time). Hence, there is a very strong probability in favour of the primitiveness of this structure, and its adaptation, which seems to be wind pollination. In particular, theHedyosmumstamen is considered here not according to the general view as a reduced flower, but as a transformed bract. The female flower pattern is like the male one, but more reduced; it theoretically is an axis with several phyllomes, one of which being the carpel, the others, fused with one another and with the ovary, representing the (generally trimerous) perianth which is homologous to the more or less fused sterile phyllomes at the bottom of the male flower. The partial female inflorescence is basically an ordinary spike and the male one a cyme, but there are reasons for believing that the two inflorescence patterns are closely related with one another. There is a “spectacular” evolutionary trend from the common ancestor of bothHedyosmumandAscarina(another Chloranthaceae), the latter genus being possibly insect‐pollinated and having male flowers only 1‐5‐staminate. The new interpretation of theHedyosmumflower reported here leads me to think very seriously that some (many?) angiosperm trends must have had a chloranthoid origin.