3. Davallia trichomanoides Blume, Enum. Pl. Javae. 2: 238. 1828.
骨碎补 gu sui bu
Davallia bullata Wallich ex Hooker; D. cylindrica Ching; D. mariesii T. Moore ex Baker (1891), not H. J. Veitch (1880); D. mariesii var. stenolepis (Hayata) Hoshizaki; D. petelotii Tardieu & C. Christensen; D. stenolepis Hayata; D. trichomanoides var. bullata (Wallich ex Hooker) Sarn. Singh & Panigrahi; Trogostolon yunnanensis Ching.
Rhizome 3-8 mm in diam. (without scales), not white waxy. Scales brown or red-brown, with pale border from base to apex or not, flat and nearly acicular, narrowed abruptly from a broad base or above much broader base evenly narrowed toward apex, often curling backward or appressed to rhizome, usually crisped, margins recurved, peltate, 4-8 × 1-1.5 mm, without multiseptate hairs, with marginal setae at least in distal part or toothed. Stipe pale, adaxially grooved, 4.5-20 cm, glabrous or with few scales; lamina compound, tripinnate or quadripinnate toward base and in middle part, deltoid and broadest toward base, 10-35 × 9-25 cm, glabrous, not or slightly dimorphic. Longest petiolules 1-6 mm; pinnae deltoid, longest 5-19 × 3-12 cm; pinnules of at least larger pinnae anadromous, narrowly ovate, longest 20-70 × 10-30 mm; ultimate pinnae linear-oblong or narrowly ovate, lobed almost to midrib; ultimate segments 5-27 × 2-6 mm. Rachises and costae glabrous. Margins of each pinna not thickened. Veins in sterile ultimate lobes simple or forked, not reaching margin; false veins present, rarely absent. Sori separate, frequently single on a segment, at forking point of veins; indusium also attached along sides, pouch-shaped, oblong, longer than wide, 1.2-2 × 0.5-1 mm, upper margin not elongated, truncate or slightly rounded, separated from or even with lamina margin; lamina generally extending into a tooth at both sides of a sorus or only at outside of a sorus.
Epiphytic and epilithic on different kinds of rocks, mostly in wet places, sometimes on dry, exposed places; 100-3500 m. Fujian, Jiangsu, Liaoning, Shandong, Taiwan, Yunnan, Zhejiang [Bhutan, India, Indonesia, Japan (including Ryukyu Islands), Korea, Malaysia, Myanmar, Nepal, Papua New Guinea, Thailand, Vietnam].
One of us (Nooteboom) notes that, according to B. J. Hoshizaki (pers. comm.) and personal observations, several of the species treated here as synonyms behave as good species in cultivation. However, after studying over 400 different collections of the entire area, it is clear that they all belong to one species. That does not exclude that different forms from different localities intergrade in nature but behave differently in cultivation. It would be best to give these forms cultivar names (Mariesii and Stenolepis). As soon as a plant is cultivated and vegetatively propagated it forms a clone of similar plants that can be recognized from other clones of the same species. Formally naming the forms according to the rules of nomenclature means that quite a lot of collections cannot be named. As the spores of all the forms are also extremely similar, there is no doubt as to their conspecificity.