Diplacntni.] clvi. cypbrace^ (clarke). .Ml 



2. D. longifolium, C. B. Clarke in Ditrand d' Schinz, Conspect. 

 Fl. Afr. V. 669. Nearly glabrous. Stolons slender. Stem 20 in. long, 

 robust, triquetrous. Leaves 24 by J- J in. Inflorescence of '), distant, 

 short-peduncled, axillary, globose, pale heads h in. in diam. ; bracts 6 in. 

 long and upwards. Spiketets numerous, female J in. long. Female 

 glumes elliptic, mucronate, boat-shaped, many-ribbed, not winired on 

 the keel. Nut about ^tq i^i- ^^^^i white, smooth, with obscure longi- 

 tudinal striation? Avhich partially anastomose. — Urban, Symb. An till. ii. 

 153. Pteroscleria longifolia, Griseb. Fl. Brit. West Tnd. 579 ; J>eHth. 

 in Hook. Ic. PI. t. 1347. 



Upper Guinea. Sierra Leone : marshy ground near Mofari, Scott-EUioi, 

 4406! 



Frequent in Trinidad and Brazil. 



Scott-Elliot's 4406 is the only piece of the section Pteroscleria yet obtained in 

 the Old World. 



26. ERIOSPORA, A. Rich. ; Benth. et Hook. f. Gen. PL iii. 1070. 



Spikelets very small, 2-3-flowered, collected in close spikes re- 

 sembling the spikelets of Scirpus, mostly monoecious, about 4-glamed ; 

 lowest flower female, upper 1-2 male or sterile. Glumes ovate, boat- 

 shaped, obscurely distichous, very minutely mucronulate, lowest empty, 

 next female. Hypogynous hairs numerous, very fine, or (in L\ villosnla) 

 bristles much stouter. Stamens 3-1 ; anthers not crested. Nut from 

 an ovoid base, tapering into an elongate conical trigonous beak (style- 

 base) ; linear part of style short, persistent, branches 3, longer. — 

 Perennials, with linear leaves. Stems with nodes their whole length 

 bearing leaves or bracts. Spikes on slender peduncles in an elongate 

 panicle, copious in the typical species, reduced in E. Oliveri to few 

 spikes. 



Species 7 , scattered through Tropical Africa, the Transvaal and Madagascar. 



The majority of the species of this extraordinary genus have the leaf -sheaths 

 exactly like those of grasses ; they are equitant, more or less distichous, deeply split 

 down one side, with a ring of white hairs entirely simulating the ligule of grasses at 

 the mouth. The illusion is so complete that where (as in many herbarium examples) 

 the plant has been broken in half, a botantist dealing with a single sheet has de- 

 scribed the Eriospora and assumed the basal half to be that of some grass 

 accidentally pasted down on the same sheet. This hypothesis would at all events 

 explain to some extent the way in which competent botanists have dealt with species 

 of Eriospora. Still more surprising is it to find in E. pilosa (and its var.) the leaf- 

 sheath and ligule exactly as of Scleria. 



^'SCLEEiiFOLi^. — Sheaths of the leaves triquetrous, with entire mouth closed by 

 a short-Gvate ligule . . . . . . . 1. E. pilosa. 



='"''=Geamixifoli^. — Sheaths of the leaves compressed, split deeply down one 

 side, with a ring of short hairs in the mouth. 



Stems glabrous. Inflorescence copious, spikes exceed- 

 ingly numerous. 

 Spikes 5-J in. long, brown or chestnut , . . 2. E. ahyssinica. 

 Spikes ^ in. long, yellowish-straw-colour . . 3. ^. schiveinfurthiana. 



