Streptogyna is the only herbaceous bamboo genus with an amphi-Atlantic distribution. Streptogyna crinita occurs in tropical Africa, Sri Lanka, and India, and S. americana is found throughout the Neotropics. Although marked differences in habit and in morphology of lodicules and starch grains suggest segregation at the generic level, they are here retained in a single genus because they are quite similar in spikelet morphology and leaf anatomy. Multicellular microhairs are present on the lodicules of S. crinita and mark the first report of microhairs in the genus. Multicellular microhairs are otherwise well developed in the grasses only in the African herbaceous bamboo Guaduella, and they may be a primitive feature in the family as they are common in its putative outgroup, the Joinvilleaceae. Streptogyna shows strong bambusoid affinities in its ligule and leaf anatomy, spikelet structure, caryopsis and embryo morphology, and chromosome number, but differs from the core group of the subfamily in its seedling morphology and lack of epidermal papillae. Autapomorphies in the two species suggest that neither could have been derived directly from the other. The genus Streptogyna was first brought to the attention of Western botanists in the late 18th century, when British and Swedish collectors brought back specimens of rat-catching grass from the forests of West Africa. A gathering from Nigeria by Palisot de Beauvois was used as the basis for Streptogyna in 1812, which he based on the only species known to him, S. crinita. The narrow leaves and many-flowered spikelets of Streptogyna were long taken as indications of pooid (festucoid) affinities. Thus, Bentham (1883), Hackel (1887), and Hubbard (1936) all considered that the proper disposition of this genus from the tropical rainforest lay with this large, temperate-region group. But there were dissenters, and Nees von Esenbeck (1835) and Steudel (1855), for example, suspected the bambusoid affinities of the genus. Streptogyna was briefly revised by Hubbard (1956), who indicated that the group deserved tribal status, but it was not until Tateoka (1958a) and Metcalfe (1960) examined its leaf anatomy that the bambusoid affinities of Streptogyna became clear. Recent workers agree that Streptogyna should be placed in its own tribe in the Bambusoideae (Calderon & Soderstrom, 1980; Clayton & Renvoize, 1986). In a treatment of the herbaceous bamboos of Sri Lanka, Soderstrom et al. (1987) offered a detailed descriptive account of the leafblade anatomy in the two taxa. The present study provides a revision of the genus and attempts to clarify the relationships of the two species by examining characters that have not yet been studied in detail, such as morphology of lodicules, starch grains, and embryos. MATERIALS AND METHODS Specimens of Streptogyna were examined from the following herbaria: AAU, B, BM, BR, CAY, CEPEC, F, G, ISC, K, M, MO, NA, NY, P, PDA, RB, S, US, W, and WIS. For anatomical studies, spikelets, leaves, and embryos (Table 1) were dehydrated in dimethoxypropane, infiltrated with tertiary butanol, embedded in wax, sectioned using a rotary microtome, and stained in chlorazol black E. Lodicules were rehydrated with AerosolOT before examination. Starch grains from caryopses were cut on a freezing microtome and stained with l2KI. Observations of living plants of Streptogyna were made by Soderstrom in Brazil (March 1972, and May 1976) and by Jud1 We are grateful to Alice R. Tangerini for the illustrations, to Stanley Yankowski for the floret sections, to the curators of the herbaria that lent us specimens for study, and to Gerrit Davidse, Lynn G. Clark, and Mary Sangrey. Part of the research leading to this paper was performed while the second author was a predoctoral fellow at the Smithsonian Institution. 2 Department of Botany, Smithsonian Institution, Washington, D.C. 20560, U.S.A. Died on September 1, 1987. 3Department of Botany, Smithsonian Institution, Washington, D.C. -20560, U.S.A. ANN. MISSOURI BOT. GARD. 74: 871-888. 1987. This content downloaded from 157.55.39.179 on Tue, 12 Apr 2016 09:52:13 UTC All use subject to http://about.jstor.org/terms 872 ANNALS OF THE MISSOURI BOTANICAL GARDEN [VOL. 74 TABLE 1. Specimens of Streptogyna for which embryos (e), floret bases (using the scanning electron microscope) (f), lodicules (1), and starch grains (s) were examined.