SPICILEGIA FLOE.E SINENSIS. 3 



pl. iii. 56), and Maximowicz (Mel. biol. Acad. St. Petersb. ix. 713), 

 that Acerantlnis cannot be maintained as a genus. 



5. Polyr/ala Waitersii, Hance. — Gregarie crescentem montesqiie 

 secus fl. Lien-cbau prov. Cantonensis, pulchre ornantem, versus 

 Martii exitum a. 1881, invenit rev. B. C. Henry. The detection of 

 this remarkable species in Kwang-tung enables me to add some 

 particulars to the character I formerly drew up. According to 

 Mr. Henry, it forms "a branching shrub, rather round and sym- 

 metrical in shape, from four to eight feet high ;" and he describes 

 the inner petaloid sepals as of a golden yellow, and the petals as of 

 a purplish hue ; and so they dry in his specimens. The racemes 

 are usually solitary in the axils of the crowded upper leaves, and 

 have as many as forty or more flowers ; and these are little more 

 than half the size of those on the I-chang specimen. But I do 

 not in any way doubt that the two are conspecific. Although so 

 utterly different in habit from the other species, which are all 

 delicate herbs, the Chinese plant seems, in every point of structure, 

 to accord with Hasskarl's genus Semeiocardium ; which, however, 

 unless CharnccbiLvus be maintained, cannot, I hold, be separated 

 from Poly gala. 



6. Grewia yarviflora, Bunge. — In collinis ad ripas fl. Yang-toz, 

 asst. 1878, necnon circa Chin-kiang, prov. Kiang-su, Maio 1880, 

 coll. Bullock. 



7. Linum trigynmn, Roxb. — Juxta urbem I-chang, ineunte m. 

 Aprili 1880, leg. am. T. Watters. The specimen is poor, but there 

 is no doubt of its identity, and this is an unexpected addition to the 

 Chinese flora. The discovery of several closely allied species in 

 California, forming Asa Gray's section Hesperolinon (Proc. Amer. 

 Acad. vii. 521 ; Watson Bot. Calif, i. 89), has conclusively shown 

 that Pieiyiwardtia is untenable as a genus. In the Asiatic plant the 

 small appendages of the petals cohere strongly with the adjacent 

 petal, and their function appears to be simply to hold the petals 

 together. 



8. Xantlioxijlo7i Bungei, Planch. — In collibus Feng-wang-shan, 

 prope Shang-hai, Maio 1876, invenit F. B. Forbes ; circa Chin- 

 kiang, prov. Kiang-su, Maio 1880, leg. Bullock; juxta I-chang, 

 prov. Hu-peh, Jun. 1880, coll. Watters. 



9. Vitis bryoniifolia, Bunge. — Circa A-moy, coll. De Grijs, a. 

 1858 ; in ins. Formosa, prope Tam-sui, Aprih 1864 (Oldham n. S5) ; 

 ad cacumina collium, Ningpo, Maio 1877, leg. Hancock. Han- 

 cock's plant is so named by Maximowicz (Ad fl. As. or. cogn. 

 melior. Fragm. 7) — than whom no one should have a better know- 

 ledge of Bunge's j)lants — and it quite agrees with the character ; 

 on the other hand, I have the high authority of Mr. Bentham that 

 De Grijs' s (which is undoubtedly the same), is identical with V. 

 fici/olia, Bge. It is also quite impossible to distinguish the For- 

 mosan plant, though its leaves are much less deeply incised, 

 but the differences are almost as great in some oi De Grijs' s 

 specimens. Miquel (Ann. mus. bot. Lugd.-Bat. i. 93) refers 

 F. Jicifolia to V. Labrusca, Linn., as does also Eegel (Act. hort. 

 Petrop. ii. 896), but he appears to consider T'. brg<mufulia (also 



