Summary. Two new families, Cyvclocheilaceae and JVesogenaceae are described for the nonAustralian genera previously included in the Chloanthaceae. One new genus, Asepalum, is described, and the combination A. eriantherum (Vatke) Marais is made. The genus Nesogenes is reviewed and a key to the species provided. The genera Nesogenes A.DC. and Acharitea Benth. have for years been placed with uncertainty in several families of the Gamopetalae, namely Scrophulariaceae, Myoporaceae and Verbenaceae. Briquet (1897) formalized the situation by placing them in Verbenaceae-Chloanthoideae-Achariteae (sic). Oliver (1895; 1896) described Cyclocheilon in the Scrophulariaceae-Gerardieae. Stapf (1901) moved the genus to the Verbenaceae, placing it next to Nesogenes mainly on the character of the reticulately veined nature of the enlarging fruiting 'calyx'; he admitted that no mature fruit, hence no seeds, of Cyclocheilon s.l. were known to him. Apparently he did not consider the basic difference in form of the outer envelope of the flower as being of any consequence. Airy Shaw (1966) placed Nesogenes and Acharitea, with some doubt, in the Dicrastylidaceae (the correct name for this family is Chloanthaceae, see Munir (1979) and previous papers). Moldenke (1971) states in the key, on p. 750 where he keys them out as a tribe of the Dicrastylidaceae, that the seeds of these two genera, as well as of Cyclocheilon s.l., contain endosperm-but perhaps he is best ignored since in the same work, on p. 738, he states that the tribe Achariteae (sic) belongs to the Stilbaceae. He apparently did not study any specimens. Munir, l.c.: 1-2, and in previous publications in which he has monographed all the Australian genera of the Chloanthaceae, restricts the family to these Australian genera. This conclusion is supported by the work on the morphology of the gynoecium carried out by Junell, l.c.: 214 (1934) who excluded Acharitea, Nesogenes and Cyclocheilon from the Verbenaceae while retaining the Australian genera there in a distinct tribe Chloantheae. Junell, l.c.: 60 also remarked that Nesogenes and Acharitea were very closely related, but did not suggest any affinity. He did suggest that Cyclocheilon s.l. might be related to the Acanthaceae. I had, independently, come to the same conclusion from the gross morphology, before reading this particular remark. The material in Kew Herbarium (which is cited below without indication of location) enabled me to confirm that the herbaceous genus Nesogenes, of which fruiting material of most species is available, does indeed have seeds containing endosperm. I have also examined seeds of Cyclocheilon s.l., a genus of perennial woody plants; these contain no storage tissue around the