Anthospermum is a genus of small shrubs or suffrutescent plants with small, narrow leaves and often an ericoid habit; unisexual or sometimes bisexual, reduced, apparently wind-pollinated flowers borne in the axils of the leaves and with the corolla generally reduced, the stamens exserted, and the stigmas long and feathery; and the fruits schizocarpous with two small, generally smooth, usually indehiscent mericarps. Usually the leaves of the axillary stem buds develop but without elongation of the internodes of the axillary stems, so superficially the leaves often appear densely grouped or verticillate even though they are not. Separation of species of Anthospermum is difficult due at least in part to its reduced vegetative and reproductive structures and usually unisexual and markedly dimorphic flowers. Puff (1986) revised the genus in extensive detail, and recognized 39 species with most in South Africa. Anthospermum is sometimes similar in general aspect to Galium, but can be separated by its opposite or ternate leaves with developed interpetiolar stipules, which have the sheaths fused to the leaf bases or petioles on each side, vs. leaves (apparently) in whorls of 4-8 with foliaceous stipules that exactly resemble the leaves in Galium.
Anthospermum is found at generally higher elevations in Madagascar. Puff (1986) recognized 8 species here, all endemic. He considered A. emirnense, A. isolaense, and A. madagascariense closely related to each other, and all related to A. usambarense of East Africa; A. perrieri closely related to several species from South Africa; and A. ibityense and A. longisepalum closely related to each other, A. palustre and A. thymoides closely related to each other, and all four of these related to the widespread African species A. herbaceum L.f. One specimen, Decary 13415 from Massif de Tsiafajavona, is quite similar to A. palustre except in its fruits, which are ca. 2.2 mm long and puberulous-papillose: thus in Puff's classification, this plant falls in the circumscription of A. herbaceum, which was not documented by him from Madagascar. Puff did not cite this specimen in his treatment. Additional study may confirm this is A. herbaceum, or document additional variation in the fruit morphology of A. palustre.