SPICILEGIA FLOR^ SINENSIS. 9 



Hai-nan, coll. Kev. P. Delavay; juxta Hoi-hau, ejusdem insula, 

 legg. Bullock et Stuhlmann. 



8. Camellia Sasanqua, Thunb. Supra Teli-king, secus fl. West 

 River, prov. Cantonensis, die 7 Nov., 1877, copiose et sponte in 

 collibus nascentera, invenit Rev. R. H. Graves. 



9. Camellia (EucamilUa) Grijsii, sp. nov. Ramulis petiolisque 

 glabris, foliis coriaceis glaberrimis elliptico-lanceolatis basi cuneatis 

 apice caudato-acuminatis crebre argute serrulatis supra vix lucidulis 

 subaveniis subtus opacis costa valida nervisque primariis promi- 

 nentibus secundariis minus conspicuis 2^3^ poll, longis 10-15 lin. 

 latis petiolo 3-4 lineali, alabastris lanceolatis, sepalis ovatis, petalis 

 profunde obcordatis basi in annulum concretis 15 lin. longis, 

 staminibus conspicue monadelphis 3-4 lin. tantum longis glabris, 

 ovario lanato, stylis ad medium usque coalitis crassiusculis glabris. 

 In prov. Tokien, a. 1861, leg. am. C. F. M. de Grijs. (Herb, 

 propr., n. 7308.) 



When this was first sent me I supposed it might be Abel's 

 C. oleifera, but I believe there is now no doubt that this is referable 

 to the preceding species, which certainly does yield oil in Kwang- 

 tung. The present plant I consider to stand between C. Sasanqua 

 and C. Kissi, Walt., referred by Seemann, without sufficient proof, 

 I think, to C. drupifera, Lour. ('Trans. Linn. Soc' xxii., 345.) It 

 is nearest and looks more like the latter in its subopaque acuminate 

 leaves ; but the nervation is still more conspicuous, and the 

 serratures more copious, divergent, and very acute ; the Indian 

 species, too, has but slightly emarginate petals, and slender quite 

 free styles. By its deeply bilobed petals and semiconnate styles it 

 agrees with the former species ; but the leaves are not shining, 

 the stamens are shorter, and more distinctly united into a ring. 



10. Hibiscus inutabilis, Linn. Ad Lung-mun-hap, h. e. ''fauces 

 oris draconis," secus fl. West River, prov. Cantonensis, d. Novem- 

 bris, 1877, coll. Rev. R. H. Graves. The only truly wild specimens 

 I have ever seen. 



11. Hibiscus syriacus, Linn. Secus fl. North River, prov. Can- 

 tonensis sponte crescentem invenerunt Sampson, Julio, 1864, et 

 Rev. J. C. Nevin, Oct., 1876. I do not know why, in the ' Flora 

 of British India,' Dr. Masters has placed this and the preceding in 

 the section Paritium. They certainly do not belong there, their 

 capsule being quite destitute of false dissepiments, but to Ketmia, 

 where DeCandolle originally stationed the latter, placing the former 

 under Abelmoschus. B. Rosa-sinensis apparently never fruits here, 

 but I have little doubt that it also falls under Ketmia, and not 

 Paritium, where it is likewise placed by Masters. 



12. Oxalis {Biophytum) Reinwardtii, Zucc. Prope Macao, ipse 

 legi, 1860; in umbrosis juxta ripas fl. West River, prov. Can- 

 tonensis, d. 21 Julii, 1872. DcCandolle's genus Biophytum, united 

 with Oxalis in the ' Genera,' has been again separated by Kurz, 

 and also in the ' Flora of British India,' but it seems to have little 

 claim to distinction, especially when the wide differences of habit in 

 various species of Oxalis are considered. 



13. Xanthoxylon alatum, Roxb., var. Circa Amoy, ipse legi, 



I 



