227 



Ceylon. Common in rockv 



according to Trimen. Wild specimens were specially collected 

 by Dr. J. C. Willis and sent alive to Kew for the purpose of tliis 



ipli in 1912. The native name for this jjlant in Ceylon 



mono 



is Niyanda. 



It is remarkable that, although this species was one of the 

 earliest to be described, yet it is one of the least known, or rather 

 is practically an unknown plant to science, since in all modern 

 descriptions, it is inextricably mixed up with the Indian S. Ilox- 

 burgliiana and S. lanuginosa, and the South African S. aethi- 

 cpica. The confusion was began by Willdenow himself, who 

 mixed up two other species with this ; but as" the name and 

 description clearly point to the Ceylon plant as here understood 

 and not to either of the South African species included under it 

 by Willdenow, I think the specific name zeylanica should be ex- 

 clusively applied to the Ceylon plant, which, with the exception 

 of the living plants now at Kew, from which the above descrip- 

 tion_ has been made, does not appear to be anywhere in 

 cultivation out of Ceylon, nor have I seen in any British or 

 foreign Herbarium good dried specimens that I should unhesi- 

 tatingly refer to the true Ceylonese plant. Such as I have 

 examined bearing the name of S. zeylanica, belong either to 

 ^. A^oxburghia7ia, S. aethiopica, or some other totally different 

 species and are often from cultivated specimens, in no sinole 

 instance have I seen a localised dried specimen from CeylSn 



If lOQ-'^^'"'' °^ ^' ^^y^'^''^^^'^ gi^^en in the Kew Bulletin for 

 c%' r ' ?' ^' ^^^0^8's to S. aethiopica, and the figure on p. 4 to 

 ^. ho,rburghiana, whilst the accounts of the fibre obtained from 

 plants grown m India, Mauritius and Jamaica probably refer 

 to those or some other species and not to the true Ceylon plant. 



28. S. Roxburghiana, Schultes Syst. Yes^. vol. vii. p. 357 



J±ig. 14, I) ■£.).— i^temless, with a creeping rootstock. Leaves G-24 

 to a growth not 2-ranked, those of juvenile plants and sometimes 

 the outer of the tuft spreading, smooth above, slight] v rough 



rowK L ^^r- ^"^ny^ "^^- ^^■^^'^' fl^*' «trap-shaped or naV 

 rowlj lanceolate, usually abruptly rounded into a stout subulate 



point 5-1 m. louo-; muer or adult leaves ascending and slio-htlv 



fiom 1-2 ft. long 1-1 m. broad, li-2 lin. thick, linear, deeply 

 concave-channelled down the face, rounded or very Obtusely 

 koded on the back gradually tapering into a stout subulati 

 soft green point 1-2 m. ong green, transversely marked with 

 darker green ra her regular bars on both sides and with 6-11 



vnT'of ten 1 tl ^^ ^'^'' '^ /^' '''''''^^ P^^^^ under-«urface 

 <vnd often 1-3 on the upper; edges green, with age becomino- 



.ery narrowly whitish. Florver-sfem 1-21 ft. },iX wTthTS 

 erect acuminate sheaths 1-11 in. long on 'the low?r p. t and a 



1? r lin '?""" 1^'^" ^. ^°^-^ '^ flJwei-clusters Tbo^ve bfacts 



i\f ?°^g:',^^"^e«late-attenuate, membranous. Flowers 



about four m a cluster; redicels 21-4 lin. long, iointed LT iZ 



long, loDes 4-4^ lui. long, Imear, obtuse; Kunth, Enum! PI. 



