In his monograph of Dryopteris (1913) Christensen adopted the name Cyclosorus with subgeneric status for a group of tropical American species, most of which are not closely related to C. gongylodes (of which the type specimen came from Guiana). Ching extended the use of the name, with generic rank, to cover species of the Old World some of which are related to the atypical American species of Christensen. Copeland (Gen. Fil. 1947) followed Ching, distinguishing Cyclosorus from Lastrea (= Thelypteris sensu Ching) solely by anastomosing/free venation, but this does not result in a natural division. In my judgement, Cyclosorus (as here restricted) is most nearly related to Ampelopteris (see Holttum, Sen & Mittra 1970) and Thelypteris s. str. (all have the same aquatic habitat and a wide distribution), probably also to Mesophlebion, not closely to any other group. As here restricted, Cyclosorus is a complex which can only be resolved by extensive cytotax- onomic study of plants from the whole of its range. The fronds of W. Indian tetraploids are large, firm and almost hairless on the lower surface; South African plants, originally named Polypodium tottum Thunb., have a similar aspect. Seventeen binomials, with dilferent types, are based on specimens belonging to this complex.
All Indian and African records diploid, with n =36; in the W. Indies and S. America some plants tetraploid.