Description from Flora of China
Shrubs or trees, unarmed; branches sometimes flattened. Raphides absent. Leaves opposite or occasionally in whorls of 3, without or rarely with domatia, with margins rarely denticulate (apparently where scabrous projections form triangular thickenings), infrequently with venation not visible abaxially; stipules persistent or caducous as a whole or by fragmentation, interpetiolar, triangular to pandurate or leaflike (i.e., generally ovate and narrowed to a stipitate base), entire or rarely bilobed, erect and flat to longitudinally folded and/or spreading to reflexed. Inflorescences terminal, cymose, thyrsoid, or paniculiform, many flowered, sessile to pedunculate, bracteate. Flowers sessile or pedicellate, bisexual, monomorphic, often fragrant. Calyx limb 5-lobed. Corolla white, purple, or red, tubular, salverform, or funnelform, glabrous or pubescent inside, with top portion of tube often reflexed at anthesis; lobes (4 or)5, imbricate in bud, strongly reflexed to revolute at anthesis. Stamens (4 or)5, inserted in corolla tube near throat, partially to fully exserted; filaments short to developed; anthers dorsifixed, sometimes with connective prolonged in short apical and/or basal appendages. Ovary 2(or 3)-celled, ovules numerous in each cell on peltate axile placentas; stigma bifid or rarely clavate and shortly emarginate at apex (Wendlandia pendula), exserted. Fruit capsular, subglobose, loculicidally dehiscent across apical portion into 2 valves with valves later sometimes splitting septicidally, papery to woody, with calyx limb persistent; seeds numerous, small, compressed, sometimes narrowly winged; testa membranous, reticulate-striate; endosperm fleshy.
The flowers seem to open generally all at once on an individual plant and perhaps in the population. They are reported in several species to be strongly fragrant. W. C. Chen (in FRPS 71(1): 191. 1999) described the fruit as rarely septicidal, but this has not been reported by other authors and has not been seen on the specimens studied.
Wendlandia has been studied by several authors, notably in a comprehensive monograph by Cowan (Notes Roy. Bot. Gard. Edinburgh 16: 233-313, t. 232-235. 1932, with supplemental notes shortly afterward; Notes Roy. Bot. Gard. Edinburgh 18: 183-188. 1934) and then in treatments of the Chinese species by F. C. How (Sunyatsenia 7(1-2): 32-62. 1948) and later W. C. Chen (Acta Phytotax. Sin. 21: 277-284, 386-403. 1983).
Many Wendlandia species seem to be morphologically variable with much of the variation difficult to partition into distinct clusters, as discussed by Cowan (loc. cit. 1932). Cowan also evaluated characters used by various authors to distinguish species and concluded that many were variable and/or incorrectly interpreted, although he used several of these to distinguish infraspecific taxa. With more species known now, there also seems to be variation within species in some of the characters that Cowan considered reliable to distinguish species. W. C. Chen (loc. cit. 1983) used Cowan’s characters to delimit some of his new species, but circumscribed other species to include a relatively wide range of variation in some of the same features (e.g., W. pingpienensis).
Cowan recognized four series and four subseries, distinguished by stigmas, anther, and stipule morphology as well as habit. These taxa were accepted by W. C. Chen in FRPS (loc. cit. 1999: 195, 196, 200, 202, 208, 218, 221); however, the classification following Cowan’s treatment of several Chinese Wendlandia species described by recent authors is problematic because flowers are needed but are unknown for these. The key published by W. C. Chen (loc. cit. 1999: 192-195) closely follows that of Cowan except some leads were numbered incorrectly while others appear to be missing, and the updated information on morphological variation and geographic distribution was not added to the key so could be problematic to use. The key to Wendlandia species here follows that of W. C. Chen, to outline for reference the species distinctions in that treatment; however, it has been augmented with the new morphological and distributional information.
In addition to the species treated here, F. C. How (loc. cit.: 43) reported Wendlandia ternifolia Cowan provisionally from China based on Tsang 21937, Liang 67941, and Zoo 69290. However, W. C. Chen in FRPS (loc. cit. 1999: 192) commented that study of Tsang 21937 showed the calyx to have stiff pubescence, which is inconsistent with the protologue description of W. ternifolia, while the other two specimens cited have no flowers in adequate condition for identification. Therefore, W. ternifolia was excluded from the Chinese flora; no new or alternative identification was given by Chen for these specimens.
At least 90 species: mainly in tropical and subtropical Asia and a few in the Pacific region; 31 species (21 endemic) in China.
(Authors: Chen Tao (陈涛); Charlotte M. Taylor)