Scliizandracem.^ ceylon plants. 5 



6. DILLENIA, Lin. 



1. D, speciosa, Tlmnb. (H.f. et T., I.e. p. 69, ram sj^n,)— c.p. 2961. 



Hab. Common iu the warmer parts of the island, up to an elevation 



of 20()0 feet; most frequent on the banks of streams. Nom. vulg. 

 "Hondapara." 



Sect. CoLBEETiAj Salish, 



2. D. retusa, Thunb. {Wormia refusa, H.f. el T., I.e. p. 68, cum 



syn.)— c.P. 2960, 



Hab. In similar situations with the preceding species, but by no 

 means so abundant. Nom. vidg. " Godapara," 



Arbor niediocris ; peialis albis ; stawmilms immerosis, 5-6-serialibus (externis 

 gradaalim brevioribus), pallida sanguiucis ; antherariim loculis lateralibus, albis ; 

 or«^Y/.s 5j 30-35-ovulati3 ; stylis filiformi-subulatis, recurvatis ; or;*/?.? 4-serialibus ; 

 fructu maturo pendulo, croceo, \\-\\ poll. diam. ; carpellis indehiscentibus, ealyce 

 valde ancto canioso arete inclusis ; seminibus exarillatis, in pulpa pellacida gelati- 

 nosa nidulantibus. 



There can be no doubt that tbis bandsome species belongs to the Colhertia section 

 of Billeiiiaj though it seems to differ from the others iu having white instead of 

 yellow petals. There is an indication of the stamens being collected into five groups 

 alternating with the carpels. 



t 3. D. Integra, Thunb. {Wormia Integra, H.f. et T., 1. c. p. 68, cum 



syn.) 



I know^ nothing of this plant, and suspect there must be some error in a fourth 

 species of these large Dilleniaceous trees being attributed to Ceylon. 



III. MAGA^OLIACE^. 



1. MICHELIA, Lin. 



1. M. NilagiHca, Zenk. (H.f. et T., I.e. p, 82, cum syn.) 



Var. a, oca^j/bZ^'a, Wight; petalis cum sep. sa^pissimc 0. — cp. 141, 424 



Var. ^. WalkerU H.f. et T. {M. Walkeri and Jf. glauca, Wight) ; 



petahs cum sepalis 8a?pi88ime 12.— cp. 2636 and 576, 299, 338, 423, 617, 

 Hab. Central Provmcc, at an elevation of 3000 to 8000 feet. Nom. 



^Ig. '*Walsappoo." 



In Ceylon the most typical form of this plant is that called M. Walkeri by Dr. 

 Wight, having lanceolate leaves more or less glaucous on the under surface, aud 

 most commouly twelve petals. Between this and the oval-leaved form a gradual 

 transition may apparently be observed, and, although the latter has most connnonly 

 nine petals, and these wider, yet, lite the authors of the * Flora Indica/ I have 

 Been quite unable to discover any permanent characters on which two seppj-ate, 

 clearly-defined species can be founded. The form ocrnrring on the Neilgherries, as 

 shown in a specimen sent tome by Dr. Hooker, differs from any I have seen growing 

 ^ Ceylon in its larger leaves and much larger petals. 



M. Chavipaca, L., is very common in cultivated ground, but I have never seen it 

 undoubtetUy wild in the forests. 



IV. SCHIZANDEACE^. 



1. KADSUEA, Juss. 



"Wightiana, Arn. (H.f. et T., Lc, p. 84, cum 8yn.)~c.P. 1028. 



