Description from Flora of China
Diplophragma (Wight & Arnott) Meisner; Exallage Bremekamp; Gonotheca Blume ex Candolle (1830), not Rafinesque (1818); Hedyotis sect. Diplophragma Wight & Arnott; Metabolos Blume; Oldenlandia Linnaeus; Thecagonum Babu.
Herbs, subshrubs, or shrubs, annual or perennial, procumbent to erect or climbing, unarmed. Raphides present. Leaves opposite [or rarely whorled], sometimes clustered at ends of stems, without domatia; secondary venation rarely triplinerved or palmate; stipules persistent, interpetiolar, fused to petiole bases, or united around stem, triangular to truncate, entire or ciliate to laciniate, erose, 1- to several lobed and/or -setose. Inflorescences terminal, pseudoaxillary, and/or axillary, few to many flowered and fasciculate, cymose, paniculate, capitate, or glomerulate or reduced to 1 flower, sessile or pedunculate, bracteate or bracts reduced. Flowers pedicellate or sessile, bisexual and monomorphic or distylous [to unisexual on dioecious plants]. Calyx limb shallowly to deeply (2-)4-lobed (or 5-lobed, Hedyotis hainanensis). Corolla white, pink, purple, or blue, tubular, funnelform, salverform, rotate, or urceolate, variously glabrous or pubescent inside; lobes (2-)4(or 5, H. hainanensis), valvate in bud. Stamens 4(or 5, H. hainanensis), inserted in corolla tube or throat, included or exserted; filaments developed to reduced; anthers dorsifixed often near base. Ovary 2-celled, ovules few to numerous or rarely 1 in each cell on axile placentas; stigma 2-lobed with lobes linear to clavate or rarely undivided, included or exserted. Fruit indehiscent, schizocarpous, or capsular, generally subglobose to ovoid or dicoccous, crustaceous to membranous or leathery, when schizocarpous splitting into 2 mericarps, when capsular splitting partially to entirely septicidally and/or loculicidally, subsequently sometimes splitting other way, apically flattened or with short to well-developed beak (i.e., disk area inside calyx limb), sometimes dehiscent primarily through beak, with calyx limb persistent; seeds few to numerous, small, angular or plano-convex; testa smooth, reticulate, or otherwise variously ornamented; endosperm fleshy; radicle clavate or terete.
This is a very problematic genus or group of genera. Neither the overall identity and limits of this lineage---distributed throughout the tropics and warm temperate regions of the world, with numerous species with often reduced morphology---nor the evolutionary patterns within it are at all understood or delineated. Widely differing taxonomies and species-level characters have long been used in different regions and floras. It is generally accepted now that Hedyotis is closely related to or at least in some cases perhaps includes Houstonia Linnaeus, Kadua Chamisso & Schlechtendal, Kohautia Chamisso & Schlechtendal, Neanotis, Oldenlandia, and a number of smaller segregate genera including Exallage, Oldenlandiopsis Terrell & W. H. Lewis, Pentodon Hochstetter, Stenaria Terrell, Stenotis Terrell, and Thecagonum. The situation is far from resolution or even general consensus. This genus is treated broadly here, as done also by many recent authors working our flora region (Fukuoka, S. E. Asia Stud. 8(3): 305-336. 1970; W. C. Ko in FRPS 71(1): 26-77. 1999; Wang & Zhao, J. Trop. Subtrop. Bot. 9(3): 219-228. 2001; Dutta & Deb, Taxon. Rev. Hedyotis. 2004). Recently, some authors have separated Oldenlandia; but, as outlined by Terrell and Robinson (Taxon 52: 775-782. 2003), recent molecular studies have concluded that the circumscriptions and relationships of these two groups are less well understood than had been thought, and these groups are probably paraphyletic and/or polyphyletic with relation to each other as well as several other genera.
The taxonomy of Hedyotis is further complicated by nomenclatural issues, in particular the designation of the type species. Dutta and Deb (loc. cit. 2004 - a late publication of a 1991 manuscript), following majority opinion of the time, considered H. auricularia as the type of Hedyotis; however, subsequently, H. fruticosa Linnaeus instead was successfully proposed as the conserved type of the genus (Nicolson, Taxon 41: 564. 1992; see Vienna Code, App. III, p. 343). The typification of Hedyotis and corresponding generic names were reviewed in detail by Terrell and Robinson (loc. cit.).
Terrell and Robinson (loc. cit.) also summarized the infrageneric classification and species groups of Hedyotis, including those accepted by W. C. Ko (loc. cit.), but without noting a few differences between Ko’s classifications and theirs, nor the use of some incorrect sectional names by Ko (e.g., H. sect. "Euoldenlandia" would have been called H. sect. "Oldenlandia" if it had been published, but it was not; Ko’s H. sect. "Diplophragma" included the species that is now the type of the genus, thus this should have been called H. sect. Hedyotis, while this particular section was synonymized by Terrell & Robinson). The genus circumscription as well as the infrageneric classification of Hedyotis are very far from understood at present (Groeninckx et al., Scripta Bot. Belg. 44: 33. 2008).
The information available about Hedyotis bodinieri is inadequate to include this species in the key. Because of the complexity of this genus or group of genera and the large number of species in China, the descriptions here are more detailed than in some other Rubiaceae genus treatments here. W. C. Ko (loc. cit.) described the fruit of most species of Hedyotis as dehiscent into 2 mericarps at maturity, with mericarps vertically dehiscent at ventral part, but these fruit are considered capsules by other authors. In some cases, this description was not entirely accurate because the fruit are actually truly schizocarpous (i.e., with indehiscent mericarps) or primarily loculicidal.
About 500 species: tropical and subtropical regions worldwide, most in Africa and Asia, a few in warm temperate regions; 67 species (38 endemic) in China.
(Authors: Chen Tao (陈涛); Charlotte M. Taylor)