Brachyspatha Schott; Candarum Reichenbach ex Schott & Endlicher; Conophallus Schott; Hydrosme Schott; Kunda Rafinesque; Pythion Martius.

Herbs, terrestrial, small to massive. Stem subterranean, tuberous, rarely a chain of tubers or a true rhizome; rhizome ± creeping. Leaves usually solitary, rarely paired or more; petiole terete, rarely angulate, shallowly grooved, or partly rugulose, rarely entirely verrucate or hairy; leaf blade decompound, divided into 3 main segments; main segments equally long or anterior shorter than posterior ones (blade ± subpedate); rachises naked, narrowly or broadly winged and often carrying supernumerary leaflets on proximal parts; bulbils sometimes developing on leaves, either epiphyllar, intercalary (developing inside rachis), or half-epiphyllar (base developing in rachis, apex exposed beyond rachis). Inflorescence 1, rarely 2 or 3 (synflorescence) or more per season (then developing from different buds on stem), epigeal, rarely partly buried, solitary or simultaneous with or directly preceding leaf development, rarely emerging after leaf development. Spathe base convolute, rarely open or connate, not or clearly separated from limb by a constriction, outside variously colored, variously shaped, often cymbiform or campanulate, rarely funnel-shaped; limb erect, spreading, oblique, or arching. Spadix sessile or shortly stipitate; female zone contiguous with male zone or separated from it by a sterile zone; flowers sometimes surrounded by staminodes; female flowers consisting of 1 pistil; ovary sessile or shortly stipitate, 1-4-loculed, with 1 ovule per locule, basifixed, or rarely axillary ca. halfway up ovary; style present or (nearly) absent, clearly separated from ovary or less so, sometimes with apical projections ("branches") extending beyond stigma; sterile zone (when present) covered with staminodes, rarely partly or entirely naked; male zone cylindric, fusiform, conic, or obconic; male flowers consisting of (1-)3-6(-8) stamens; stamens depressed or elongate; filaments present or nearly absent, separated or partly or entirely fused within a flower or rarely fused between adjacent flowers; anthers bithecal; thecae 2-celled (with 2 pollen sacs), rarely 1-celled; pores apical, rarely lateral or subterminal; pollen inaperturate, globose or elliptic, exine rarely absent, psilate, striate, verrucate, echinate, areolate, porate, fossulate, reticulate, or scabrous; appendix rarely absent, contiguous with male zone or separated by a constriction or a short stipe, sometimes with large longitudinal folds or ± irregular deep cracks. Berries crowded or distant, ripening red, rarely blue, globose, ovoid, or narrowly elliptic, smooth or rarely verrucate, 1-4-seeded. Seeds usually with a distinct raphe; endosperm absent.

Amorphophallus stipitatus Engler (Notizbl. Bot. Gart. Berlin-Dahlem 8: 457. 1923) was described based on cultivated material in the Berlin Botanical Garden, originally from Guangdong. The holotype (Mell. s.n., B) has been destroyed, and Engler’s protologue does not give enough clues as to the identity of the species to which the name refers. It is no doubt a member of the A. yunnanensis alliance, but the differences between species recognized in this group are too subtle to leave room for any guessing about the identity of A. stipitatus. So as not to confuse matters, this name is here treated as of uncertain application.

The status of Amorphophallus zengianus C. Long & H. Li (Novon 10: 125. 2000), described from S Yunnan (Jinping), is uncertain. H. Li notes that the morphology of the inflorescence might be misleading because it was damaged while in the bud stage. She considers it to be conspecific with A. yunnanensis, based on morphological similarity of the petiole and leaf blade. On the other hand, Hetterscheid considers A. zengianus to represent a less-than-optimal specimen of A. krausei, based on photographs provided by H. Li of living material of A. zengianus. This identity with A. krausei could be confirmed if there exists a staminodal zone between the male and female zones of the spadix. However, H. Li notes that the living material no longer exists and the holotype specimen cited in the protologue (C. L. Long 98003, KUN) never existed.

"Amorphophallus tienmushanensis" (Y. Z. Tao, Compreh. Invest. Rep. Nat. Resource Tianmu Mount. Nat. Reserve, 130. 1992) was described from Zhejiang (Tianmu Shan). The present authors consider this name to be not validly published because the requirement of Art. 36.1 of the ICBN (Vienna Code) was not fulfilled. The intended validating description given by Tao is a mixture of Latin and English. The code requires "a Latin description or diagnosis." D. H. Nicolson (pers. comm.) is of the opinion that a percentage of non-Latin is admissible; a similar view was held by R. K. Brummitt and D. E. Boufford (pers. comm.; for further discussions see Plant Press 45, Jan. 1994 and Plant Press 45, Feb. 1994). We consider as quite disturbing the discussion about the acceptance of a percentage of non-Latin in a diagnosis where "a Latin description or diagnosis" is required, especially given that non-English users of the ICBN will not be able to read between subtle English lines. The ICBN must be unambiguous at an international level, and the phrase "a Latin description or diagnosis" should be clear to all, as it is to us, to mean 100% Latin.

The following species were recorded in FRPS but are, in fact, not distributed in China: Amorphophallus bangkokensis Gagnepain (FRPS 13(2): 91, 99. 1979), A. mekongensis Engler & Gehrmann (p. 91), A. oncophyllus Prain ex J. D. Hooker (p. 98; A. muelleri Blume), and A. variabilis Blume (p. 95).

About 200 species: paleotropical, W to E Africa, S and SE to E Asia, N Australia, Pacific islands; 16 species (seven endemic) in China.

(Authors: Li Heng (李恒 Li Hen); Wilbert L. A. Hetterscheid)