connected ivith tJie Flora of Guiana. 229 



often above the middle, and then suddenly expanded ; the segments are always 

 in a single row, (thongh imbricate in aestivation,) nearly equal in size, and 5 

 in number in the species I have seen, G, or perhaps more, in some described 

 by Bonpland, and the ovary 3- 4- or 5-celled. Of this group I have examined 

 S. marlinicensis, Linn., S.Cipon'ima, S. ^-Jrechea, L'ller., Stenimatosiphon plati/- 

 p/ii/llitm, nifens, and unijioruin, Pohl., Si/mplocos puhescens, Klotsch, and two 

 new species described below; and, judging from Bonpland's figures, I should 

 likewise refer to it his Si/mplocos semilata and rufescens. 



As a second genus, or at any rate as a distinct section of Si/mplocos, I 

 should propose to restore Linmieus's Alstonia, characterized by a more cam- 

 panulate corolla, with an inner row of small corolline segments, which may 

 perhaps be considered as an outer row of sterile stamina. 1 have only seen 

 one species, the AMonia thecvfonnis, Linn., and of tliat I could only dissect 

 one imperfect flower, in which the inner row of petals was very irregular, and 

 certainly took the place of some of the external stamina. I should associate 

 with it Bonpland's Si/?nptocos cerntta and coccinea, judging from the figures, 

 and perhaps also Si/mplocos fomenfosa, Bonpl., and S. octopetala, Swartz. 

 But it would require a re-examination of all these species to determine the 

 importance of the inner row of petals as characterizing a section or a 

 genus. 



In the third very distinct genus, Hopea, Linn, (not Roxb.), the aspect of the 

 flower is very different ; the corolla is almost rotate, constantly uniseriate at 

 the base, though the divisions be imbricate, and 5- or 6-cleft; the stamina are 

 also spreading, their filaments slender, but sligiitly connected at the base, often 

 somewhat pentadelphous, and usually longer than the corolla. I have also 

 never found more than tliree cells to the ovary, (in H.sinica and cn/ta'goides 

 there are but two,) and the species appear much more apt to dry yellow than 

 in the true Si/mplocos. I would refer to Hopea, so characterized, //. tincturla, 

 Linn., and the greater number, if not all the Asiatic species. Amongst these 

 the -S. .v/«/f fl. Bot. Reg., and S. cratcegoides, Hamilt., should form a distinct 

 section, as proposed by Don, but characterized by the bilocular ovarium and 

 comparatively slender stigmata. There appears also to be a considerable di- 

 versity in the fruit, which is pear- or bottle-shaped, and very small in S.spi- 

 cata, Roxb., S. polycarpa, Wall., and .S'. lamina, Wall., small, oblong, and 



VOL. XVIII. 2 H 



