— 243 — 



9. Calyx hirsute, membranaceous Flower red C. reticulafo^). 



Calyx hirsute, rather thick. Flower red. . . C. shinkoensis{ED)'y)- 

 Calyx hirsute, l'ignaceous. Four styles, glabrous. Flower red. 



C. Iwngkongensis^). 



Calyx glabrous, thin. Flower white C. speciosa^) (ED). 



10. Leaves rather large and thin, acuminate. Flowers usually small, white 



or yellow. Ovary hairy ^^• 



Leaves usually coriaceous. Flower ordinarily large, white or red. 

 Ovary hirsute or glabrous ^^ 



11. Fruit with irregular impressions. Three short recurved styles. CoroUa 



very soon deciduous. Branches and leaves pubescent. 



'■'■ C. iniquicarpa. 



Fruit round or triangular. Three free styles. Flowers yellowish . . 12. 



Fruit round or three-lobed. Three to four styles 3), more or iess 



united. Flowers white '"^• 



12. Styles hirsute. Stamens nearly free. Fruit triangular C ^o/z/^mé'ns/s (E D). 

 Styles glabrous, short, recurved. Stamens irregularly united. Fruit 



j-Qu„cl ^- liitescens"^). 



13. Flowers large. Petals emarginate. Fruits 3-lobed. Leaves broadly 



lanceolate ■ ^- confusa. 



Flowers large. Petals obovate. Fruits? Leaves obovate. Cortex brick- 



. coloured *C. Crapnelliana. 



Flowers small. Petals emarginate. Fruits round. Leaves broadly lan- 

 ceolate C. drupifera. 



Flowers medium-sized. Petals obovate. Fruits? Leaves ovate (?). 



C. biflora^) (E D). 



14. Ovary and style glabrous.- Petals obovate. Branches and leaves pubes- 



cent. Sepals glabrous with membranaceous margin. C. yunnanensis. 

 Ovary? style? Petals obovate. Shoots and young leaves pubescent. 



Sepals somewhat puberulous with ciliated margin. 



C. tenuiflora (E D). 

 Ovary hirsute. Petals emarginate. Branches, leaves and calyx hairy or 



not ^^• 



') E D^e descriptione. 



-) C. rcticiilata, shinkoensis, fwngkonrensis and speciosa differ so little from C.japonica, 



that they might be united with tne latter, or preserved under tlie common name of 



C. reticiilata. KOCHS, in his détermination table (1900, p. 581) cannot find sharp 



characteristics either, aiid employs tlie leaf size as such, which is scarcely a reliable 



standard. 

 3) Since the niimber of styles (i. e., carpels) varies from 3 to 4 in the tea plant, I do 



not think that C drupifera may be said to be distinguished from its allies merely 



by having 4 styles. 

 ^) In the Kew herbarium S. KuRZ and D. Prain hâve put soine annotations on asheet 



from Griffith's collection, comparing this species with C lanceolata, to which it 



bears some resemblance indeed in several respects; I hesitate, however, to reduce 



the plant to the second section (Calpandria) as Pierre did. 

 5) Hayata adds this to his very incomplète diagnosis: „There is nothing like this at 



Kew". I cannot judge therefore of its affinity either to this group or f. i. to 14. 



