Abstract. During preparation of accounts for the genera in the Convallariaceae for the Flora of China, Volume 24, it was decided that Polygonatum ginfushanicum (F. T. Wang & Tang) F. T. Wang & Tang (Smilacina ginfushanica F. T. Wang & Tang) should be transferred to Heteropolygonatum, as H. ginfushanicum (F. T. Wang & Tang) M. N. Tamura, S. C. Chen & Turland.

Smilacina ginfushanica F. T. Wang & Tang was described in 1978 (Wang & Tang, 1978: 249), based on specimens collected from the mountain Jinfo Shan in Nanchuan County, Chongqing Province, China (T. H. Hsiung & T. L. Chou 90841, PE). Later, after examining additional specimens from Jinfo Shan (T. H. Hsiung & T. L. Chou 91094, PE; Phytogeographical Research Team 239, PE) and also from Jiangkou in northeastern Guizhou Province (Z. S. Zhang et al. 400642, PE), Wang and Tang (in Wang et al., 1983: 261) transferred the species into Polygonatum, as P. ginfushanicum (F. T. Wang & Tang) F. T. Wang & Tang. They considered that the existence of a terminal inflorescence in this species allied it with the genus Smilacina. However, they also observed an axillary inflorescence in addition to the terminal one (except in some specimens), tepals connate to a high degree, papillae present on the abaxial side of the tepal apex, and ovaries with four ovules per locule, all of which they considered as characteristics of Polygonatum. Hara (1987: 158) treated the species under Smilacina, and cited the combination in Polygonatum as a synonym, but noted "This species is till now doubtful to belong to Smilacina [sic]." Liu (1991: 236--237) considered it a member of Polygonatum. The species has also been collected in southwestern Hubei (Lichuan, 1300 m, 19 Apr. 1973, H. G. Li 11080, MO).

Recently, the genus Heteropolygonatum M. N. Tamura & Ogisu has been described (Tamura et al., 1997). It is similar to Polygonatum, but differs in possessing both terminal and axillary inflorescences (sometimes only a terminal inflorescence) and imbricate tepals. Originally, dimorphic stamen length had been considered characteristic of Heteropolygonatum, but this later turned out to be of no diagnostic value (Bao et al., 1998). However, the present authors still recognize Heteropolygonatum, and a molecular phylogeny of the tribe Polygonateae confirms this recognition (Yamashita et al., in press).

As Wang and Tang (in Wang et al., 1983) observed, Polygonatum ginfushanicum (Smilacina ginfushanica) has both terminal and axillary inflorescences (sometimes only a terminal inflorescence) and also imbricate tepals. These characters clearly indicate that the species belongs to Heteropolygonatum, so the combination in that genus is made here. Now Heteropolygonatum includes four species: H. roseolum M. N. Tamura & Ogisu (the type species), H. pendulum (Z. G. Liu & X. H. Hu) M. N. Tamura & Ogisu, H. xui W. K. Bao & M. N. Tamura, and H. ginfushanicum (F. T. Wang & Tang) M. N. Tamura, S. C. Chen & Turland, all of which are endemic to China (Chongqing, Guizhou, Hubei, and Sichuan Provinces and Guangxi Zhuang Autonomous Region). The generic description, key to the species, and specific descriptions will be provided in the forthcoming Flora of China, Volume 24 (Chen & Tamura, in press).

The specific epithet was spelled "ginfushanica" in the part of the protologue accompanying the Latin description and type designation (Wang & Tang, 1978: 249). However, in the accompanying key to species of Smilacina (p. 29), the Chinese description (p. 38), and the illustration caption (p. 43), the spelling "ginfoshanica" was used. Hara (1987) used the latter spelling. When the species was transferred to Polygonatum by Wang and Tang (in Wang et al., 1983), the epithet was spelled "ginfoshanicum," and likewise by Liu (1991). The type locality in the protologue was cited as "Gin-fu-shan," which uses the Wade-Giles system of transliteration from Chinese. The alternative, pinyin transliteration is "Jinfo Shan." One might argue that either "ginfushanicum" or "jinfoshanicum" would be correct, but not "ginfoshanicum." Therefore, the present authors adopt "ginfushanicum" and reject "ginfoshanicum"; see Article 61.3 of the Tokyo Code (Greuter et al., 1994).

Literature Cited