Our concept of A. chloranthum covers the same ground as formerly, but after
study of the fruits, in which we found an extraordinary variation in sculpture, both
internal and external, of the endocarp, w e are now impelled to recognize within it
five subspecies. As in the related A. reticulatum, the subspecies are either allopatric
or, where their ranges overlap, are recognizable from leaf-characters, so that a key
based primarily on geographic dispersal is possible, a key especiaUy to be recommended
when so great a proportion of the material is incomplete.
In the Amazon Basin A. chloranthum, represented by the relatively well-known
subspecies chloranthum and confusum, is easily distinguished, in absence of the diagnostic
staminate flower, from sympatric forms of A. reticulatum by its relatively short
and broad, rhombic-ovate leaf-blades, which are usually little longer than wide. On
the Pacific slope of the Andes from Ecuador northward to the Isthmus of Panama there are four or perhaps five geographic races of Anomospermum, all known to us in pistillate condition only. We refer three of these to A. chloranthum and one to A. reticulatum, but we lack proof that these will turn out to have the staminate perianths that characterize their Amazonian kindred. Our arrangement of these races is perforce provisional, for there is nothing in the fruit taken by itself which will divide the whole complex of subgenus .Anomospermum into specific units.