Botanical Journal of !he Linnean Sock@ ( 1 982) 84: 355-377. With 15 figures The delimitation of the tribe Anthospermeae and its affinities to the Paederieae (Rubiaceae) CHRISTIAN PUFF Institute of Botany, University of Vienna, Rennweg 14, A-1030 Vienna, Austria Received September 1981, accepted for publication February 1982 A critical comparison of many characters suggests that the Rubiaceae tribe Anthospermeae is closely allied to the tribe Paederieae. The delimitation of the tribe Anthospermeae from other tribes is redefined as to include only wind-pollinated genera, and characters of fruit structure, pollination biology and distribution patterns support the subdivision of the tribe into the three subtribes: Anthosperminne, Operculariinae and Coprosminae. All insect-pollinated genera previously placed in the Anthospermeae are transferred to the Paederieae. It is shown that the genus Neogaillonia Linchevskii (=Gaillonia A. Rich. ex. DC.), previously included in the Spermacoceae, also belongs to this tribe; the genera Pteragaillonia Linchevskii, Pseudogaillonia Linchevskii, Jaubertia Guill. and Choulettia Pomel are considered synonyms of .Neogaillonia. The Anthospermeae are believed to be closely, and the Paederieae more remotely, allied to the tribe Theligoneae. KEY WORDS :-anemophily~Anthospermeaeclassificationentomophily ~Paederie;ie~Theligoneae CONTENTS Introduction . . . . . . . . . . . . . . . Historical survey . . . . . . . . . . . . . . Circumscription of the Anthospermeae . . . . . . . . . Wind-pollinated flowers of the Anthospermeae . . . . . . . Insect-pollinated flowers of the Anthospermeae, Paederieae and .Neoguillonia Further characters of the Anthospermeae, Paederieae and .Neogaillonia. . Conclusion . . . . . . . . . . . . . . . Taxonomic conspectus . . . . . . . . . . . . . Relationships to other tribes . . . . . . . . . . . Acknowledgements. . . . . . . . . . . . . . Appendix. Synonyms of.Neogaillonia; new combinations . . . . . References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355 356 357 357 359 364 369 370 373 373 374 375 INTRODUCTION During several years of intensive investigation on African and Madagascan Rubiaceae Anthospermeae, evidence has accumulated to suggest that the tribe Anthospermeae as circumscribed in the widely accepted classifications of the + 0024-4074/82/040.355 23 S03.00/0 355 0 1982 The Linnean Society of London 356 C. PUFF Rubiaceae (Hooker, 1873; Schumann, 1891) is a highly unnatural, inflated group with only vague relationship to other tribes. An attempt is made to redefine the tribe on the basis of all available data. HISTORICAL SURVEY Shortly after Chamisso & von Schlechtendal (1828) described the tribe Anthospermeae, Richard (1830) also studied the group and included the genus Coprosma. His remarks concerning the Anthospermeae were noteworthy but unfortunately their significance was either not appreciated at the time or overlooked by later authors. While Chamisso & von Schlechtendal distinguished the members of the tribe by their two-seeded fruits usually dehiscing into two oneseeded mericarps, Richard (1830: 56) pointed out that “ . . . le vrai caractere qui la distingue . . . consiste uniquement dans le forme et la structure des stigmates. . . . on trouve deux trks-longs stigmates filiformes, subults, poilus”, the first time that an indirect reference was made to the anemophilous flowers of this group (see p. 357). Richard ( 1830) also described the tribe Opercularieae (Opercularia, Pornax) and stated that the genus Opercularia, upon which the tribe is based, should not be excluded from the Rubiaceae and placed in a family of its own (cf. de Jussieu, 1804).He also noted its affinity to the Anthospermeae “par ses deux longs stigmates subults et poilus . . . ”. De Candolle (1830) accepted Richard’s classification and also listed (de Candolle, 1830: 343) the elongated hairy stigmas as being a diagnostic character of the Anthospermeae. During the 43 years which elapsed between the publication of the accounts by Richard (1830) and de Candolle (1830) and Hooker’s (1873) system, many new genera were described. Hooker added seven new genera to the Anthospermeae semu Richard and de Candolle and also included the Opercularieae and another five genera previously placed in two different tribes by de Candolle (cf. Table 1) . The subfamilial classificationof the Rubiaceae by Baillon ( 1880),although never widely accepted, deserves special consideration. He drastically reduced Hooker’s 25 tribes of Rubiaceae to 15 and included the tribe Paederieae in the Anthospermeae. In Die natiirlichen Pjanzenfamilien, Schumann ( 1891) accepted Hooker’s classification and added only the genus Cremocarpon. But the latter as well as other genera described after Schumann’s publication were sooner or later excluded from the Anthospermeae (Table 1). It is noteworthy that, in his remarks concerning the Paederieae, Schumann pointed out the very close affinity to the Anthospermeae and stated that “the berry-like or indehiscent fruits, in at least most cases, allow a distinction” between the two tribes. Thus both Hooker and Schumann placed the Paederieae next to the Anthospermeae. For reasons unexplained, both Verdcourt (1958) and Bremekamp (1966) did not follow suit and placed the two tribes far apart from each other without mentioning or discussing possible affinities between them. Bremekamp merely stated that “in its present delimitation this tribe (Paederieae) makes a rather unnatural impression”, For a satisfactory discussion of tribal delimitation, yet another genus, .Neogaillonia Linchevskii, has to be considered. Linchevskii ( 1973), correctly substituting the name Reogaillonia for the illegitimate name Gaillonia A. Rich ex DC., also created the segregate genera Pterogaillonia and Pseudogaillonia and upheld the genera Jaubertia and Choulettia which were, in the past, considered synonyms of RUBIACEAE 35 7 Gaillonia. In my opinion, these four genera are not sufficiently different from .Neogaillonia to warrant their separation ; they should be considered synonyms (see Appendix for the necessary new combinations). The genus had been placed in the tribe Spermacoceae by all of the authors mentioned above. Both Hooker (1873) and Schumann (1891) categorically stated that Neogaillonia has an “ovule attached to the middle of the septum”, although Richard (1830: plate 5, Fig. 3F) showed clearly, and correctly, that the solitary, anatropous ovules are basally attached as in the Anthospermeae. The basal attachment of the ovules can also be seen in a figure of .Neogaillonia putorioides (A. R. Smith) Linchevskii (as Jaubertia putorioides in Smith, 1971). Qarar. (1973) also suggested an affinity of Neogaillonia to the Anthospermeae but stated that its position within the Rubiaceae is uncertain. Hemsley (in Aitchison, 1882: 162) pointed out affinities between Aitchisonia (Paederieae) and Neogaillonia. A re-investigation of the genera traditionally placed in the Paederieae and a carefiil study of the available literature indicates a clear affinity with the Anthospermeae. Genera placed in both the Anthospermeae and Paederieae sensu Hooker (1873) and Schumann (1891) are discussed below. CIRCUMSCRIPTION OF T H E ANTHOSPERMEAE From the tribes with one, basally-attached, anatropous ovule per locule, Hooker ( 1873) distinguished the Anthospermeae by the following characters: “flowers more often unisexual, stamens inserted at the base of the corolla, rarely at the throat, ovary with one to four locules, style fused or branches more often filiform, fruits baccate or indehiscent”. Schumann (1891) used “ovary two- or seldom more-locular, stamens usually attached to the base of the corolla tube, mostly dioecious plants”. These circumscriptions are, in part, incorrect and unsatisfactory. The description of filament attachment is particularly misleading. A reinvestigation has shown that only certain genera (genera 1-12 in Table 1) have the filaments attached to the base or seldom just below the middle of the corolla tube (Carpacoce species with unusually long corolla tubes). All the others have the filaments attached to the throat or always above the middle of the corolla tube. Both Hooker (1873) and Schumann (1891) used these characters in their keys to divide the tribe into two groups and as a result contradicted themselves as regards to filament attachment characters in the tribal key of the Rubiaceae. They included the genus Serissa in the 12 genera (Table 1) with filaments attached at or near the base of the tube, although its filaments are joined to the tube near the throat. They were misled by the fact that the stamen vascular system remains more or less distinct and prominent below the actual point of insertion of the filaments, and can therefore be followed to the base of the corolla tube. WIND-POLLINATED FLOWERS OF THE ANTHOSPERMEAE The same group of 12 genera (Table l ) , discussed in the previous section, is characterized by having anemophilous flowers. The stigmas are long, filiform and the receptive surface is increased by unicellular hairs of varying length (Figs 2 4 , below) ; hairs are distributed over the whole surface of the stigmas and not confined to their inner portion. The style is united, either (sub-)obsolete or, if welldeveloped, never longer and commonly much shorter than the stigmatic areas. Richard - Guettardeae: “gen. obsc.” CofFeeae Baillon I I “Operculariaceae”, cf. de Jussieu (1804) Operculariaceae, to Dipsacales? cE Bremekamp (1966) Anthospermeae of Chamisso & von Schlechtendal (1828) Comments Anthospermeae - - Guettardeae subtribe Cuettardinae Anthospermeae Ant hospermeae Guettardeae subtribe Cuettardinae Anthospermeae Paederieae to Dipsacales?, cf. Bremekamp (1966) The correct name of the genus is Spermadictyon (cf. Backer & Backhuizen van den Brink, 1965: 347) now generally considered synonyms of Paederia (cf. Verdcourt, 1976:175) Spermacoceae subtribe Putoriinae Anthospermeae Spermacoceae subtribe Putoriinae Anthospermeae Anthospermeae Guettardeae subtribe Guettardinae Chiococceae Ant hospermeae Spermacoceae subtribe Putoriinae Anthospermeae Anthospermeae cf. Robbrecht & Puff (1981); to Hedyotideae (Pu& 1982) Anthospermeae to Psathureae (= Psychotrieae), Bremekamp (1958) to Knoxieae, Verdcoun (1958) Anthospermeae Anthospermeae Anthospermeae Anthospermeae Anthospermeae Anthospermeae Anthospermeae Guettardeae subtribe Guettardinae Anthospermeae Anthospermeae Anthospermeae Opercularieae Anthospermeae Opercularieae Anthospermeae Anthospermeae An thospermeae An thospermeae Anthospermeae Anthospermeae de Candolle * = I n Schumannonly; ** =generadescribedata laterdate butassociatedwith the Anthospermeae. Authorsofgeneraaccording to Index NorninurnGenericorum (Plantarum) (Farretal.. 1979). 24. I.cptodermis Wall. 25. Pseudopyxis Miq. 26. Aifchisonia Hemsl. ex Aitch.* Paederieae 22. Paed& L. (syn. Lygodisodea Ruiz & Pav. honfea A. Rich. ex DC.) 23. Spermadupon Roxb. (syn. Hamiltonia Roxb.) Anthospermeae 2. Ncnax Gaenn. (syn. Ambraria Cruse)Anthospermeae Anthospermeae 3. Galopina Thunb. 4. Plyllis .I An thospermeae 5. Carpacme Sond. 6 . CoprosmaJ. R. & J. G. A. Font. Anthospermeae 7. Nertera Banks & Soland. ex Gaertn.“Guettardaceae dubiae” 8. Coynula Hook. fil. 9. Nomandia Hook. fil. 10. Opprcularia Gaenn. Opercularieae Opercularieae 1 1. Pmax Soland. 12. Ehfheranthus F. Muell. (“Eleuthranthes”) 13. Plocmno Ait. Spermacoceae subtribe Putoriiae 14. Putoria Pea. Spermacoceae subtribe Putoriiae 15. Crqllis E. Mey. ex Hook. fil. 16. Mikhella L. “Guettardaceae dubiae” 17. Wlogia Torr. ex Bentham Spermacoceae subtribe Putoriinae 18. SnisSo Comm. ex Juss. 19. Otwphora Zucc. (syn. Mericocabx Bamps**) 20. Crnnocarpon Baill.* 21. Calanda K. %hum.** (syn. Ofocephalur Chiov.**) 1. Anfhospmum L. Anthospermeae Hooker, Schumann Table 1. Genera belonging to tribe Anthospermeae and tribe Paederieaeaccording to Hooker ( 1873)and Schumann ( 1891), and the positionof these genera in the classification systems of Richard ( 1830), de Candolle (1830) and Baillon (1880) RUBIACEAE 359 There are usually two stigmas, even in species in which one of the two carpels or ovules is reduced (except in Carpacoce). In a few species of Coprosma, the number of stigmas and carpels is increased from typically two; in C. lucida, for example, there are usually two stigmas, but flowers with up to five stigmas also occur (Wild & Zotov, 1906). The long slender filaments are basally attached and bear relatively large, dangling, dorsifixed anthers (Fig. 1). The flowers are odourless and have no trace of a nectar-producing disk ; the corollas are inconspicuously coloured, commonly greenish to greenish yellow, and, in African genera in particular, the corolla lobes of hermaphrodite and male flowers are frequently conspicuously rolled back (Fig. 1). In most species there is a marked dimorphism in corolla size and, to a certain extent, in corolla shape between the hermaphrodite and male flowers and purely female flowers. The corollas of the female flowers are much smaller and, sometimes, even absent as compared with the corollas of the distinctive hermaphrodite and male flowers (Figs 1, 2). Hermaphrodite flowers are either protandrous (Anthospermurn, Carpacoce, Galopina, Phyllis) or protogynous (Coprosma, .Nertera, .Nomzandia, Corynula, Pomax). Nenax is the only genus in which all species are strictly dioecious. Often, other genera also have some dioecious species, but include hermaphrodite species, some gynodioecious species and some species with much more complicated sex distributions. Corynula and Eleutheranthus seem to have hermaphrodite flowers only. It must be stressed that, in contrast to Verdcourt’s (1958) and Bir Bahadur’s (1968) views, heterostyly is absent in these genera. Dioecism in the species of the wind-pollinated group is, furthermore, not to be considered “the ultimate consequence” of heterostyly. Wind-pollination in this tribe was first pointed out by Thomson (1881) for Coprosma and Nertera; Cheeseman (1887) cofirmed this for Coprosma. Delpino ( 1889) independently mentioned these two genera as being anemophilous, in addition to Phyllis, Opercularia, Pomax and Anthospermurn. Schumann (1891) must have overlooked these publications as he did not refer to wind pollination in his discussion of general characters of the Rubiaceae or in connection with the Anthospermeae. Although Loew (1905), and later Werth (1923) and Wettstein (1935), did refer to the anemophilous flowers of at least some of these genera, this was not mentioned at all by Verdcourt (1958) or Bremekamp (1966). INSECT-POLLINATED FLOWERS OF T H E ANTHOSPERMEAE, PAEDERIEAE AND .NEOGAILLO.NIA In the majority of taxa (Table 1, genera 13-18 and 22-26), the flowers are hermaphrodite and frequently heterostylous (consult Table 2). Crocyllis is an exception in that it is gynodioecious, but there is no corolla size or shape difference between hermaphrodite and female flowers (Puff & Mantell, 1982). Paederia, according to Schumann ( 1891) , has hermaphrodite and polygamous-dioecious flowers. All species of Neogaillonia are, according to Qarar ( 1973), hermaphrodite and homostylous. Flowers have corollas with funnel-shaped to more or less cylindrical, often very long corolla tubes (Figs 5, 6). The tubes are frequently hairy inside, ranging from sparsely hairy to densely pilose. Hairs may be restricted in distribution, occurring only at or below the throat, or the tube may be hairy to the base. The stamens are inserted at the throat or above the middle of the tube and have relatively short filaments so that the anthers are only slightly exserted or are enclosed in the Anthospmnum Anthosperminae Anthospermeae PAQ/J/O 4(5) 4 5 4-7 - F +M (creeping) perennial herbs 5 - - small trees, shrubs (creeping) perennial herbs +M+ F FF 5 4-5 shrubs dwarf shrubs (perennial herbs) - +v +M +M dwarf shrubs perennial herbs small trees, large shrubs+ perennial herbs small trees-rdwarfshrubs D D D D +M;- +M FF 2 2 2 1(2) D - 4-5(10) - - 4-5 - + + - - + - + + - + + + South Africa, Namibia southeastern Afiica Canary Islands and Madeira southAfiica:swtmvestancapeplwina 66 22 22 22 22 44 40 220 c. 154 New Caledonia as Coprosma, but further north and south and occurring in South America western South America 22 (south temperate) amphi-transpacific, ?32 but not reaching South America 44 -68) Africa, Madagascar; Yemen Arab Republic 44 22 Table 2. Summary of the most important characters of the genera of the tribe Anthospermeae Cham. & Schlecht. ex DC., subtribe Anthospermineae, subtribe Copmsminae Puff, and subtribe Operculariinae Bentham, and the tribe Paederieae DC + + + + + + -? Aitchisonia Serissa Paederia + + 4 4-5 ‘ H I 5 4-5 5 4-5 4-5 3-6 5 5-7 +,- + + + + + + + + + + 2-3 2-3 2 5 FF D D D D D 4-5 5 4 F FF D D FF D 2 2-3 2 2 2 D D D +V +M+ +V shrubs climbing shrubs (or perennial herbs) dwarf shrubs (dwaqshrubs perennial herbs shrubs +V +V - +M - *M dwarf shrubs subsrubs, perennial herbs (sub)shrubs, short-lived perennial herbs annual(?) dwarf shrubs shrubs shrubs perennial herbs (creeping) perennial herbs +M +M;- +V(?) +O +O + + + ?- ?- + + + + + + ?- + + + + + + + + + + USSR, Afghanistan, Pakistan-Arabia-northwestern Africa ; Socotra Mediterranean; E. to Iraq Canary Islands southern Namib desert southwestern China; western U.S.A. Korea and adjacent parts of USSR, Japan; eastern North America to Mexico Japan eastern India, Malaysian Peninsula to China southwestern Australia (south)eastern Australia Australia, Tasmania + 22 24 26 22 Himalayas, East Bengal to northeastern 44 China, Japan >60 Afghanistan and adjacent parts of Iran and Pakistan 22 China?; widely cultivated in East Asia 44 southeast Asia to Japan; to Mauritius, Madagascar. trop. and subtrop. eastern Africa; South America 22 22 44 22 22 + + FLOWERS: Insertion of stamens above the middle of the corolla tube = +, below the middle (usually at base) = - ; style much longer than the stigma lobes = ,stigmatic area much enlarged, style very small or *O= -. FRUITS: Very fleshy=FF, ffleshy=F, dry=D; dehiscence: fruits separating into indehiscent m e n c a p = + M ; Mericarps themselves dehiscing = M ;exocarp splitting into valves = +V. Inhctescence opening by means ofan operculum = +O; fruits indehiscent = -. PHYTOCHEMICAL DATA: hperubsidic glycosides and foetid odour: definite absence = - (blanks: no data available). s + + + Pseudopyxis Spermdutyon Mitchella - -? ? + + + Kellogia cr~vllis Puloria Plocama + + + + + + .Nmgaillonia Paederieae (415 1 less 4-5 Eleutheranthus - 1 Q 1 5 4-5 or Pmax Opercularia Operculariinae 362 C. PUFF Figure 1-4. SEM. Figs 1, 2. Male and female flower ofNenax divaricatn (South Africa, western Cape Prov., Puff 790712-2/1, WU); note corolla size and shape differences and insertion of filaments and rudimentary ovule (ro) in male flower, t. x 26. Fig. 3. Portion of inflorescence of female plant (st: stigmas) of Anthospermurn herbaceum (South Africa, Transvaal, Puff 791201-1/2, WU), c. x 2. Fig. 4. Stigma hairs of Anfhospemum randii (Malawi, Chongoni For. Res., Puff 780215-2/1, WU),c. x 165. RUBIACEAE Figures 5, 6. Fig. 5. Flower pair of Mikhellu repens (U.S.A., Mississippi, photo PuJJ 710724-2/4) subtended by a pair of leaf-like bracts (br);short styled flower with exserted stamens (Fa: filaments); the inside of corolla lobes and tube is densely pilose, c. x 6. Fig. 6. Part of inflorescence of Paelria sp. (Madagascar; photo Professor S.Vogel), c. x 0.5. 363 364 C. PUFF uppermost part of the tube. It is only in short-styled flowers that the anthers may be more or less conspicuously exserted (Fig. 5). In Aitchisoniu the stamens are inserted at different levels, a feature which led Bremekamp (1966) to suggest a placement of the genus in the Dipsacales rather than the Rubiaceae. Loew (1905) and Qarar (1973) recorded the same pattern in various species of Puederiu and Neogailloniu respectively. While there are usually only one or two ovules per flower in the wind-pollinated Anthospermeae, flowers of the insect-pollinated taxa more often have five or four ovules and carpels. There is a long, more-or-less filiform, united style which is terminated by two to five papillose, but never distinctly hairy, short stigma lobes, depending on the number of carpels. The stigmatic surfaces are always much smaller than the stylar portions of the gynoecium. Certain Puederiu species are aberrant in that the styles are sometimes split nearly all the way to the base, but the papillose stigmatic area remains considerably shorter than the free, nonreceptive stylar segment. Disks are present in all genera and are sometimes massive (cf. Puederiu pospischilii K. Schum., in Verdcourt, 1976: fig. 19). In conclusion, it can be said that all the insect-pollinated genera of the Anthospermeae and Paederieae, and Neogailloniu are so similar in floral morphology that it becomes impossible to draw a clear demarcation line between them. FURTHER CHARACTERS OF THE ANTHOSPERMEAE, PAEDERIEAE AND NEOCAII.LO.NIA From the previous two sections it is clear that, on the basis of floral biology and morphology, genera 1-12 of the Anthospermeae and the remaining genera form two clear-cut groups. It is now necessary to determine (1) whether there are any other characters which could test the close affinity between the two groups and (2) whether the insect-pollinated genera presently placed in different tribes share features other than floral characters which would group these genera together. Fruits: At first sight, the different genera seem to exhibit a bewildering range of different fruit types. However, one character common to all genera is evident : the seed itself is never released .but remains enclosed in an indehiscent endocarp (Figs 7, 13-14), a mericarp (Fig. 15) or an entire fruit. The fruit wall is either Figure 7. Cross section ofmencarp of Nomandia neocaledonica (New Caledonia, White 2007, K ) ; the exocarp (ex) with five vascular bundles (vb) is about to separate from the thick, sclerenchymateous endocarp (en) which encloses the seed (te: testa, a:endosperm, co: cotyledons); camera lucida drawing of microtome section, c. x 55. RUBIACEAE 365 Figures 8 12. SEM. Figs 8 1 I . fomax urnbellah (SW Australia, Zerny s.n., W). Fig. 8. Side view of infructescence with calyx-like bracts, c. x 22. Fig. 9. Removed operculum, c. x 22. Figs 10, I 1. Dorsal and ventral view of endocarp containing the seed, c. x 22. Fig. 12. Ventral view of mericarp of Nomandia neocakdonzca (New Caledonia, White 2007, K ) ; exocarp (ex) separated from ‘stone’ (endocarp plus seed), c. x 22. differentiated into two layers with an exo- and endocarp, or three layers with an exo-, meso- and endocarp. The fruits are distinctly fleshy, more-or-less fleshy or dry. Distinctly fleshy fruits are indehiscent and are drupes containing two to five, depending on the number of carpels present, ‘stones’or pyrenes (e.g. Plocama, Coproma).The more-or-less fleshy fruits, more-or-less transitional to dry fruits, split into two indehiscent mericarps. In Putoria, for example, the fleshy exocarp dries up and remains on the endocarp as a skin and in Normandia the exocarp separates from the endocarp (Figs 7, 12). 366 C. PUFF Figures13-15.SEM.Figs 13,14,Pomaxumbcllufa. (S.W.Australia,<mys.s.n.,W).Sectionofendocarp (en) plus seed (te: testa, es: endosperm, c. x40. Fig. 14. Endocarp (en) removed to expose testa (te) ofseed, c. x 40.Fig. 15. Cross section offruit ofAnfhospenumsp. near A. acfhwpicum (South Africa, eastern Cape Prov., Puff 79011+2/1); the mericarps have a distinct endocarp (en) made up of sclerenchyma cells, however, the exocarp (with a massive, tannin containing epidermis; vb: vascular bundle) does not separate from it-the mericarps themselvesremain indehiscent; the thick arrow points to the abscission zone between the two mericarps; te: testa; c. x 110. Dry fruits are most common. They are indehiscent or dehiscent in various ways. Most frequently, fruits separate into two indehiscent mericarps (e.g. Phyllis, Anthospermurn, Neogaillonia) . The mericarps show a distinct differentiation into an endocarp and a thin exocarp (Fig. 15). In Anthospermurn and some Nenax species, the mericarps sit on variously-shaped carpophores. Similar structures occur in some Neogaillonia species (see for example, Jaubert & Spach, 1843: fig. 77), but are RUBIACEAE 367 absent in all other genera. Transitions to indehiscent fruits occur in Nenax and Neogaillonia; also the fruits of Crocyllis do not always dehisce readily. In some genera, such as Carpacoce and Pseudopyxis, the exocarp splits into several valves, the number equalling the number of carpels or calyx lobes. The released diaspores are the endocarps containing seeds. The fruits of Paederia belong to this group, but the endocarps are often conspicuously winged. In Leptodennis, and perhaps Spermadictyon, the hard endocarp is surrounded by a net-like mesocarp. Among the genera with dry fruits, Pomax and Opercularia are unusual in that they produce ‘infructescences’ which develop from fused ovaries of flowers of a partial inflorescence. These infructescences open by means of an operculum (Figs 8, 9). The diaspores are endocarps enclosing the seeds (Figs 10, 11, 13, 14), and not the seeds themselves as suggested by Schumann (1891). Similarly, in Mitchella, the four-carpellate ovaries of fused flower pairs give rise to ‘double drupes’ with two groups of four pyrenes. Inzorescences: The basic and most common “central type” of rubiaceous inflorescence in the sense of Weberling (1977) is a typical thyrse or more accurately a ‘thyrso-paniculate’ inflorescence (Troll, 1964: 95). It is found in a number of genera across the present tribal boundaries, in Phyllis, Plocama and Spennadictyon for example. In other genera, the inflorescences are variously modified into three main groups: ( 1) a congested pseudoverticillate arrangement of flowers (e.g. Anthospermurn), (2) more-or-less head-like inflorescences (e.g. Aitchisonia), and (3) much reduced inflorescences with one-flowered partial inflorescences (e.g. Nutera, and Nenax species). These modifications are considered to be derived from typical thyrsus inflorescences. In the larger genera (e.g. Anthospermurn and Coprosma) there is a wide range of inflorescence organization incorporating all the variations described above. The inflorescences of Pomax and Opercularia deserve a special mention. In these genera, the ovaries of partial inflorescences are fused, thus forming single units of three or more flowers. In Pomax, these units are arranged into a n umbel-like inflorescence, which according to Weberling ( 1977), represents a modified thyrse. Growth forms: There is a very wide range in both growth form and habit. For example in the large genus Anthospennum, there are small trees, large shrubs, erect or cushion forming dwarf shrubs and perennial herbs, some of which can be rather short-lived. Coprosma exhibits a similar range of growth forms, although there are no perennial herbs (Oliver, 1935). In general, woody plants predominate, although there is a conspicuous absence of larger trees. Amongst these, Paederia is the only genus with climbing vine-like species. Some genera are purely herbaceous, for example Mitchella, Pseudopyxis, .Nertera, and all of them are perennial or sometimes biennial. There may not be any true annuals although material of the little-known genus Eleutheranthus still awaits re-examination. It is listed as an annual by both Bentham (1866) and Schumann (1891). Leaves and stipules : Broad and more-or-less large leaves are characteristic for most herbaceous genera. Woody genera tend to have smaller, often narrower, lanceolate to linear, and tougher leaves. It is particularly noteworthy that various kinds of xeromorphic leaves such as convolute or revolute rolled leaves and equifacial 368 C. PUFF round leaves ( = “kon- und revolute Rollblatter, aquifaziale Rundblatter” of Napp-Zinn, 1973) occur in all of the African genera, except Galopina, and Neogaillonia (Qarar, 1973). In species with leaves round in cross section, stomata are dispersed over the whole surface, and in certain species of Carpacoce, a species of Coprosma (Foweraker, 1916) and in Neogaillonia (Qarar, 1973) the more-or-less flat leaves may be amphistomatic. Numerous species of Coprosma are known to have domatia (Oliver, 1935), but there are no records of their occurrence in any other genus of this group. Aitchisonia is unusual in having peculiar, multicellular, stalked glands, perhaps homologues of colleters, occurring on leaves, bracts and stems. They are quite conspicuous on new growth, but disappear on the older parts. Stipules of herbaceous genera tend to be much-dissected and relatively large except in Nertera with its more reduced, more-or-less triangular stipules. Woody genera tend to have stipules with two or one more-or-less triangular lobes, or show a reduction of stipules to stipular sheaths. In the larger genera (Anthospennum in particular) there is much variation in stipule structure. Stipules are deciduous in species of Paederia, but persistent in all other genera. Phytochemical data: Available data are few. According to Kooiman (1969) and the literature cited therein, all investigated genera of the Anthospermeae and Paederieae sensu Hooker and Schumann possess asperulosidic glycosides, although these glycosides do not occur in a number of Coprosma and Nertera species. Several taxa are known to produce an unpleasant foetid odour when the leaves are crushed. Paederoside, a compound structurally related to asperulin (Hegnauer, 1973), appears to be responsible for the odour. It was first isolated from plants of Paederia scandens (Lour.) Merrill by Inouye et al. ( 1968).Foetid odour or paederoside has been also recorded in species of Nertera (Allan, 1961), Spermadictyon and Serissa (Backer & Backhuizen van den Brink, 1965), Leptodermis (Ohwi, 1965), Coprosma and Opercularia (Hegnauer, 1973), Plocama and Putoria (Mendoza-Heuer, 1977), Aitchisonia (Hemsley, in Aitchison, 1882), Kellogia (Dempster, personal communication), Carpacoce and Crocyllis (Puff & Mantell, 1982). There is little doubt that remarks as to the “rather unpleasant odour” of the plants on herbarium sheets of Neogaillonia (for example Lamond I 3 9 in Edinburgh herbarium) are also a clear reference to the presence of paederoside. Outside the Anthospermeae and Paederieae, this distinct foetid odour or paederoside is known only to occur in Theligonum (Theligoneae ;Wunderlich, 1971), Oldenlandia hirsuta L.f. (Hedyotideae; Makino, 1909) and Pentanisia foetida Verdc. (Knoxieae; Verdcourt, 1958). Chromosome numbers: Published counts (compiled from Fedorov, 1969) and my own unpublished karyological investigations confirm a base number of x = 1 1 for most genera. Hamiltonia (Spenadictyon) suaveolens Roxb., with numbers of 2n = 22, 24, 26, and Nertera setulosa Hook.fil., with 2n =40, seem the only exceptions. The chromosome number of 2n= 32 for Coprosma montana Hillebr. appears doubtful. So far only diploid species have been detected in Nenax, Galopina, Carpacoce, Crocyllis (Puff & Mantell, 1982), Phyllis, Putoria, Mitchella, Spermadictyon and Serissa (Vij, Shekhar & Kuthiala, 1981; Fagerlind, 1937, with 2n = 22 for S.foetida Lam.). Tetraploids alone have been found in Nertera and Plocama. Diploids and higher polyploids have been found in Anthospermum, Coprosma and Leptodennis. RUBIACEAE 369 Pollen : Pollen grains are generally 3-colpate or 3-colporate (Robbrecht, 1982). Mature grains are trinucleate in Anthospermum, Nenax, Galopina, Carpacoce, Phyllis, Paederia and Serissa, and binucleate in Leptodermis (see Wunderlich, 1971, and references included therein). Embryology : Little information is available. The archesporium is unicellular in Phyllis (Fagerlind, 1936a) and Anthospermum, but multicellular in Putoria (Fagerlind, 1936b) and Crocyllis (Puff & Mantell, 1982). Nuclear endosperm has been recorded for Phyllis, Putoria (Fagerlind, 1936a, b) and Paederia (Hashmi & Siddiqui, 1974). The suspensor is non-haustorial and uniseriate in Leptodermis and Putoria (Fagerlind 1937, 1936b), biseriate in Paederia (Hashmi & Siddiqui, 1974), and multiseriate in Phyllis (Fagerlind, 1936a). The obturator of Phyllis is a small, finger-like projection (Fagerlind, 1936a); the obturators of Putoria, Leptodemzis (Fagerlind 1936b, 1937) and Crocyllis (Puff & Mantell, 1982) are very similar to each other in having special elongated hair-like cells. Distribution : The genera of the Anthospermeae and Paederieae are essentially tropical or subtropical, with extensions to both the north and south. They are relatively poorly represented in South America, but there is a marked concentration of taxa in both southeast Asia and Africa. Widely distributed genera are Paederia, occurring from southeast Asia to Mauritius, Madagascar and tropical and subtropical eastern Africa, and to South America, and the transpacific genus Coprosma (Oliver, 1935: map, fig. 3). Nertera is even more widely distributed than the former and extends further away from the tropics (van Steenis, 1962: map, fig. 14; in van Steenis, 1965: map, fig. 6, Madagascar is incorrectly included in the distribution map). Anthospmum is widely distributed in Africa and Madagascar and also occurs in the Yemen Arab Republic. Neogaillonia occurs from southwest Asia to northwest Africa (Qarar, 1973: maps 1-4). Opercularia is found over most of Australia (Bentham, 1866). Spermadictyon, Leptodermis and Serissa occur from southeast Asia to Japan. Putoria occurs in the Mediterranean region, and east to Iraq (Ehrendorfer & SchonbeckTemesy, 1980). Kellogia and Mitchella, consisting of two species each, show interesting southeast Asian-North American disjunctions. K . chinensis Franch. occurs in the mountains of the Yunnan Province in China (Smith, 1921; Handel-Mazzetti, 1936; Dempster, 1975), K . galioides Torr. occurs in the western U.S.A. (Hitchcock & Cronquist, 1973); M . undulata Sieb. & Zucc. occurs in Japan, Korea and adjacent parts of the U.S.S.R. (Ohwi, 1965), M . repens L. in the eastern U.S.A. (Fernald, 1950) and south to Mexico. The other genera have more limited distributions. In the northern hemisphere, Phyllis and Plocama occur in the Canary Islands and Madeira, Aitchisonia occurs in Afghanistan and adjacent parts of Iran and Pakistan and Pseudopyxis occurs in Japan. In the southern hemisphere, .Nenax, Carpacoce, Galopina and Crocyllis are found in southern and southwestern Africa, Pomax and Eleutheranthus in southeastern and western Australia, Normandia in New Caledonia and Corynula occurs in western South America. CONCLUSION From this comparison of the characters it seems that the genera of the Anthospermeae and the Paederieae form a related, interconnected group. Most of 370 C. PUFF the characters overlap between genera and show much variation, but the floral structures and pollination mechanisms allow a clear distinction, and a separation into two groups. Rather than including all genera in a single tribe (i.e. the Anthospermeae; viz. Baillon, 1880), it would appear most reasonable to place all wind-pollinated genera in the tribe Anthospermeae s e w strict0 and all insectpollinated genera in the Paederieae. In this context, it is important to emphasize that there is a clear demarcation line between wind- and insect-pollinated taxa; there are no taxa showing gradual transitions from insect- to wind-pollinated flowers or vice versa. Darwin (1976) should be consulted for detailed information on correct names and authors of the tribes and subtribes dealt with in the following section. Table 2 gives a summary of the most important characters of the genera. TAXONOMIC CONSPECTUS Tribe Anthospermeae Cham. & Schlecht. ex DC. The tribe occurs in tropical, subtropical and temperate regions of the southern hemisphere, with a few extensions to the north. Details of fruit structure, pollination biology, particularly differences in protandry and protogyny, together with a consideration of distribution patterns can be used to make subdivisions within the Anthospermeae. Subtribe Anthosperminae Genera Anthospermurn, Galopina, Nenax, Carpacoce, Phyllis This subtribe is restricted to Africa, Arabia in the Yemen Arab Republic, Madagascar, the Canary Islands and Madeira and is characterized by dry, dehiscent fruits with the exception of dry, indehiscent fruits in a few Nenax species, and unisexual and/or protandrous hermaphrodite flowers. There are many indications that both Nenax and Galopina (endemic to southern Africa) may be closely allied to the large, widely distributed genus Anlhospermum, in contrast to the Cape endemic genus Carpacoce which is more likely to have ancient links with non African genera. The same is, perhaps, true for Phyllis, although it might have closer links to Anthospermurn than Carpacoce. Subtribe Coprosminae Puff, Subtribus nova Plan tae cum Anthosperminis et Operculariinis floribus anemophilis congruentes, sed fructibus carnosis vel carnulosis differt. The subtribe agrees with the Anthosperminae and the Operculariinae in having wind-pollinated flowers, but differs in having fleshy or more-or-less fleshy fruits. Genera Coprosma, ,Normandia, Nertera, Corynula This subtribe has a more-or-less south temperate amphi-transpacific distribution (cf. van Steenis, 1962). It is characterized by the very fleshy to more-or-less fleshy indehiscent to dehiscent fruits and the unisexual and/or protogynous hermaphrodite flowers. RUBIACEAE 37 I The woody, monotypic, New Caledonian, genus Normandia is closely allied to the large, woody genus Coprosma. The herbaceous genus .Nertera is probably also closely allied and related to Coprosma. It has a wider distribution than the latter. The South American endemic genus Corynula is most likely a segregate and very close ally of Nertera. Subtribe Operculariinae Bentham Genera Opercularia, Pomax, Eleutheranthus The subtribe is restricted to Australia and Tasmania and is characterized by the umbel-like or compound head-like inflorescences. The partial inflorescences, whose flowers have fused ovaries, give rise to infructescences opening by means of an operculum. The flowers are unisexual and/or protogynous hermaphrodite. I do not support the placement of Opercularia and Pomax into a separate tribe Opercularieae as was done by Richard (1830) and de Candolle (1830) on the basis of their infructescences. Details of floral morphology and protogyny rather suggest a relatively close affinity to the Coprosminae. The little known, monotypic genus Eleutheranthus probably belongs here. It is supposed to have head-like inflorescences, in which the ovaries of partial inflorescence are not fused. A thorough re-investigation is needed which may well provide evidence for its inclusion in Opercularia. There is no support for the assertion (Bremekamp, 1966) that these three genera should be removed from the Rubiaceae and placed in the Dipsacales as a separate family, the Operculariaceae (cf. de Jussieu, 1804). Tribe Paederieae DC. The basic floral structure of all genera in this group is quite uniform, although size and shape of corolla tubes and number of carpels present, ranging from two to five in number, are variable. The fruits, in parallel to the Anthospermeae, are distinctly fleshy, more-or-less fleshy or dry. The tribe occurs in (sub)tropical southeast Asia, with extensions to the south, the Mediterranean and North America. Its genera are concentrated in the northern hemisphere with the exception of Crocyllis, an endemic genus in the southern Namib desert. A subdivision of the Paederieae into discrete groups does not seem as simple and clear-cut as in the Anthospermeae since less data are available for the tribe. Therefore, it is preferred to only delimit and discuss a number of “species groups” without giving them formal taxonomic recognition. Group 1 : ..Veogaillonia, Crocyllis, Putoria, Plocama, Kellogia .Neogaillonia, occurring in the Saharo-Sindian and Irano-Turanian regions (Qarar, 1973), shows undeniable and, in some characters, very close morphological similarities to the southern Namib genus Crocyllis (Puff & Mantell, 1982). This suggests a close affinity between the two genera and probably an ancient xeromorphic palaeotropic common ancestor. The genus pair provides another example of the close floristic links between the arid regions ofsouthwestern Africa and the north (cf. Burtt, 1971; de Winter, 1971). The Mediterranean genus Putoria is morphologically quite similar to 572 C. PUFF Neogaillonia, but many xeromorphic features found in Neogaillonia are absent. Flowers are similar, fruits are slightly fleshy but, as in Neogaillonia, separate into two indehiscent mericarps. Distribution patterns of other (non-rubiaceous) taxa (cf. Davies & Hedge, 1971 : 3. Irano-Turanian-Mediterranean connections) seem to support, from a phytogeographical point of view, a connection between the genera via a southwest Asiatic ancestral group. The tetraploid, monotypic genus Plocama, although diffcying in its fleshy fruits, appears to be affiliated to Neogaillonia on the basis of vegetative characters, including xeromorphic features, and to Crocyllis on the basis of its short, more-orless funnel-shaped corolla tubes. The flowers, particularly the short, more-or-less funnel-shaped corolla tubes, of the disjunct herbaceous genus Kellogia are similar to those of both Plocama and Crocyllis; Neogaillonia has more-or-less long-tubed flowers. Fruits of Kellogia, Crocyllis and Neogaillonia are almost identical except for the presence of carpophores in some species of Neogaillonia. These morphological similarities seem to point to a relationship with the other genera of this group, although Kellogia is, on account of its distribution (southwest China-western U.S.A.), more-or-less isolated. Group 2 : Mitchella, Pseudopyxis A purely herbaceous group, in which Mitchella shows an east Asian-North American disjunction similar to that of Kellogia, while Pseudopyxis is endemic to Japan. The two genera are similar both vegetatively and in their floral structure. Mitchella has flowers in pairs (partial inflorescences) with fused ovaries. In Pseudopyxis, the partial inflorescences are similarly few-flowered, but the ovaries of adjacent flowers remain unfused. Mitchella has fleshy, indehiscent ‘double fruits’, Pseudopyxis dry, dehiscent fruits. The position of these two genera within the Paederieae and their links to other genera is not entirely understood at present. Perhaps there is an ancient connection to the genera of group three. Group 3 : Spermadictyon, Leptodermis, Aitchisonia, Serissa This group comprises woody genera occurring in central and southeast Asia with the exception of Aitchisonia, which is endemic to Afghanistan and neighbouring areas to the west and south. The group is characterized by flowers with more-or-less long, funnel-shaped to cylindrical corolla tubes, more-or-less large and broad-leaved foliage frequently with more-or-less triangular, single stipular lobes. The fruits are variable. Spermadictyon and Leptodermis have fivecarpellate fruits dehiscing into valves; Aitchisonia has bicarpellate fruits splitting into two dehiscent mericarps. Serissa has two- to three-carpellate fleshy indehiscent fruits. Spermadictyon and Leptodermis appear to be closely related on the basis of their fruit characters, similar flower structure and general appearance. However, in Leptodermis, the flowers and fruits are subtended by an often very conspicuous cuplike structure consisting of a pair of fused bracts which are not to be confused with the carpophores of some Neogaillonia species. This cup-like structure, although not as conspicuous, is also present in Aitchisonia; it may point to a closer affinity between the two genera. Aitchisonia, in spite of its unique stalked glands, also seems to have ‘cross links’ to Neogaillonia with its bicarpellate flowers and anther filaments inserted at different levels in the corolla tube. RUBIACEAE 373 The diploid genus Serissa seems to be isolated because of its fleshy, two- to threecarpellate fruits. However, the flower structure, vegetative characters and general appearance are similar to Spermadictyon and Leptodemis. Spermadictyon is diploid, but also shows aneuploidy ( n = 1 1, 12, 13) ; Leptodemis has di-, tetra- and hexaploid species. Group 4 : Paederia Paederiu appears to be the most derived genus within the tribe because of its climbing habit, deciduous stipules, anther filaments inserted at different levels in the corolla tube, its winged endocarps and tetraploid chromosome number. It also is the most widely distributed genus in the tribe, although the majority of species occurs in southeast Asia. RELATIONSHIP TO OTHER TRIBES There are many indications to suggest that the monotypic tribe Theligoneae is closely allied to the Anthospermeae (as originally suggested by Wunderlich, 1971, on the basis of a careful comparison of a wide range of characters) and, more remotely, to the Paederieae. I am certain that the genus Theligonum belongs to the Rubiaceae in spite ofdoubts recently expressed by Praglowski (1973) and Nowicke & Skvarla ( 1979). Their conclusions are based primarily on palynological characters; the strong arguments in favour of its placement in the Rubiaceae presented by Wunderlich (1971) are largely ignored. Theligonum, like all Anthospermeae, has wind-pollinated flowers with distinct sexual dimorphism. There is only one ovule per flower and the drupaceous fruit has a hard endocarp enclosing the seed. The species are annual, or possibly shortlived perennials. Their leaves show a strong anisophylly in the upper portions of the stems. A clear trend to anisophylly is also found in some species ofAnthospemum and Curpucoce. The plants produce the foetid odour characteristic of many Anthospermeae and Paederieae and the compounds asperuloside and galiumglucoside are also present (Kooiman, 1971 ). The geographical distribution is noteworthy in that it overlaps with that of several genera of the Paederieae; Theligonum occurs in the Canary Islands, the Mediterranean area and east to Japan (Ulbrich, 1934). The similarities pointed out above seem to suggest that the three tribes Paederieae, Anthospermeae and Theligoneae form a related group. Amongst these, the Anthospermeae and Theligoneae appear to exhibit more ‘derived’ characters than the Paederieae. The Anthospermeae and Theligoneae are the only two groups within the Rubiaceae with anemophilous flowers and this certainly must be considered a derived feature in a large family with predominantly zoophilous flowers. This group of three tribes may, in turn, be linked with the large, more or less pantropical, primarily woody tribe Psathureae ( = Psychotrieae) . Similarities in details of floral structure are undeniable, and the drupaceous fruits with two to many pyrenes suggest some relationship. However, available data are insufficient to draw definite conclusions. ACKNOWLEDGEMENTS I am indebted to Dr E. Robbrecht for reviewing the manuscript and making helpful suggestions for its improvment. I thank Professor S. Vogel for providing me C. PUFF 314 with photographs of Paederia; Miss L. T . Dempster for field observations on Kellogia; Miss D. Mantell for critically checking the manuscript, and Miss J. Dawe for numerous chromosome counts. The CSIR/Pretoria (CSR/P75; 1979), the Senate Research Committee of the University of the Witwatersrand/Johannesburg (SRG/P990; 1978-79) and the Hochschuljubilaumsstiftung der Stadt Wien ( 1982) provided research funds. SEM investigations were supported by the Fonds zur Forderung der wissenschaftlichen Forschung in Osterreich (Projekt 3681) . APPENDIX Synonyms of Neogaillonia ; new combinations Neogaillonia Linchevskii, Novosti Sistematiki Q-sshich Rastenij, 10: 226 ( 1973), Gaillonia A. Rich. ex DC., Prodromus 4: 574 (1830), nom. illeg. (non Gaillona Bonnemaison, 1828). Jaubertia Guill., Annales des Sciences Naturelles, Paris, skrie 2, 16: 60 (1841). TYPE: 3. aucheri Guill. Choulettia Pomel Nouveaux Mathiauxpour la Flore Atlantique: 81 (1874) TYPE : C. reboudiana (Coss. & Dur.) Pomel. Pterogaillonia Linchevskii, Novosti Sistematiki Q-sshich Rastenij, 10: 233 ( 1973). TYPE : P. calycoptera (Jaub. & Spach)Linchevskii. Pseudogaillonia Linchevskii, Novosti Sistmatiki 5-sshich Rastenij, 10: 235 ( 1973). TYPE : P. hymenostephana (Jaub. & Spach)Linchevskii. SYNONYMS: LECOTYPE: N , olivieri (A. Rich. ex DC.)Linchevskii. Neogaillonia calycoptera (Decne.) Puff, comb. nova SYNONYMS: Spennacoce caly(co)ptera Decne., Annales des Sciences Naturelles, Paris, skrie 2, 2: 267 (1834). Gaillonia calycoptera (Decne.) Jaub. & Spach, Annales des Sciences Naturelles, Paris, drie 2, 20: 86 ( 1843). Pterogaillonia calyoptera (Decne.)Linchevskii,Novosti Sistematiki Q-sshich Rastenij, 10: 233 ( 1973). TYPE: Eygpt, Mt Sinai, Bout! 208, (P!, W!). Neogaillonia stscherbinovskii (Linchevskii) Puff, comb. nova Pterogaillonia stscherbinovskii Linchevskii, Novosti Sistematiki Vjkshich Rastenij, 10: 234 (1973). SYNONYM : TYPE : Iran, Ab-Gjarm, 40 km north of Bender-Abbas, Stscherbinovski (LE). Neogaillonia crucianelloides (Jaub. & Spach)Puff, comb. nova Gaillonia crucianelloides Jaub. et Spach, Annales des Sciences Naturelles, Paris, skrie 2, 20: 86 ( 1843). Pterogaillonia crucianelloides (Jaub. & Spach) Linchevskii, Novosti Sistmatiki Q-sshich Rastenij, 10: 235 ( 1973). SYNONYMS: TYPE: Iran, Mt. Ghino, Aucher-Eloy 4681 (P!, W ! ) . RUBIACEAE 375 Neogaillonia hpenostephuna Uaub. & Spach)Puff, comb. nova Gaillonia hymenostephana Jaub. & Spach, Annales des Sciences Naturelles, Paris, sirie 2, 20: 85 (1843). Pseudogaillonia hymenostephana (Jaub. & Spach)Linchevskii, Novosti Sistematiki Vjkshich Rastenij, 10: 236 ( 1973). Gaillonia humijiisa Jaub. & Spach, Annales des Sciences Naturelles, Paris, skie 2,20: 85 ( 1843). SYNONYMS : TYPE:(Oman) Maskat, Aucher-Eloy 4677 (K, P!, W!). Neogaillonia aucheri (Guill.)Puff, comb. nova SYNONYMS : Jaubertia aucheri Guill., Annales des Sciences Naturelles, Paris, skrie 2, 16: 60 ( 1841). Gaillonia aucheri (Guill.)Jaub. & Spach, Annales des Sciences Naturelles, Paris, skrie 2, 20: 87 (1843). TYPE:(Oman) Maskat and Iran, province Laristan, Aucher-Eloy 4680 (P!, W ! ) Neogaillonia reboudianu (Coss. & Dur.)Puff, comb. nova SYNONYMS : Gaillonia reboudiana Coss. et Dur., Bulletin de la Sociktt! Botanique de France, 2: 250 (1855). Choulettia reboudiana (Coss. & Dur.)Pomel, Nouveaux Mdteriauxpour la Flore Atlantique: 8 1 ( 1874). TYPE:Algeria, Oued-en-Nsa, Reboud 22 (P!). REFERENCES AITCHISON, J. E. T., 1882. O n the flora of the Kuram valley, etc., Afghanistan. Part 11. Journal of the Linnean Sociep (Bofany),19: 139-292. ALLAN, H. H., 1961. Flora of New zealand, Vol. 1. Wellington: Government Printer. BACKER, C. A. & BAKHUIZEN VAN DEN BRINK, R . C., JR, 1965. Flora of Java, Vol. 2. Groningen: Noordhoff. BAILLON, H., 1880. Histoire desplantes 7. Monographic des Rubiacies, des Valirrianacies el Dipsacacies: 257-534. Paris: Hachette. BENTHAM, C., 1866. Flora aurfraliensis, Vol. 3. London: Love11 Reeve. BIR BAHADUR, 1968. Heterostyly in Rubiaceae: a review. O a a n i a University Journal of Science, 4 : 207-238. BREMEKAMP, C. E. B., 1958. Monographie desgenres Cremocarpon Boiv. ex Baill. et P y g r o Brem. Candollea, 16: 147-177. BREMEKAMP, C. E. B., 1966. Remarks on the position, the delimitation and the subdivision of the Rubiaceae. Acta Botanica .Neerlandica, 15: 1-33. BURTT, B. L., 1971. From the south: and African view offloras ofwestern Asia. In P. H. Davis, P. C. Harper & I . C. Hedge (Eds), Plant l,i$ of South-west Asia: 135-149. Edinburgh: Botanical Society. CHAMISSO, L. C. & VON SCHLECHTENDAL, D., 1828. De plantis in expeditione speculatoria Romanzoffiana observatis. Rubiaceae. Linnaca, 3: 309-366. CHEESEMAN, T. F., 1887. On the New Zealand species of Coproaa. Transactions of fhe .New zealand Instifutt, 19: 2 18-252. DARWIN, S. P., 1976. The subfamilial, tribal and subtribal nomenclature of the Rubiaceae. Taxon, 25: 595-610. DAVIS, P. H. & HEDGE, I. C., 1971. Floristic links between N.W. Africa and S.W. Asia. Annalen des .Naturhistoruchen Mustums Wien, 75: 43-57. DE CANDOLLE, A. P., 1830. Prodromus s y s h a t i s naturalis regni vegetabilis, Vol. 4. Paris: Treuttel & Wiirtz. DE JUSSIEU, A. L., 1804. Mimoire sur I’Opercularia, genre de plantes voisin de la famille des DipsacCes. Annales du Muslum #Histoire Naturelle (Paris), 4: 418-428. DELPINO, F., 1889. Osservazioni e note botaniche. VI. Anemofilia dei fiori di Phyllis nobla. Malpighia, 3: 348-349. DEMPSTER, L. T., 1975. An Asian Kellogia (Rubiaceae). MadroG, 23: 1W101 DE WINTER, B., 1971. Floristic relationships between the northern and southern arid areas in Africa. Milteilungen der Bofanischen Staahammlung Munchen, 10: 424-437. 576 C. PUFF EHRENDORFER, F. & SCHONBECK-TEMESY, E., 1980. Rubiaceae. In C. C. Townsend & E. Guest (Eds), Flora of Iraq, Vol. IV: 564-628. Baghdad: Ministry of Agriculture. FAGERLIND, F., 1936a. Embryologische Beobachtungen uber die Gattung P!yllis. Botaniska Notiser, 1936: 577-584. FAGERLIND, F., 1936b. Die Embryologie von Putoria. Suensk Botanisk Tiakkrij, 30: 362-372. FAGERLIND, F., 1937. Embryologische, zytologische und bestaubungsexperimentelle Studien in der Familie Rubiaceae nebst Bemerkungen uber einige Polyploiditatsprobleme. Acfu Horti Bergiani, 11: 195-470. FARR, E. R., LEUSSINK, j.A. & STAFLEU, F. A. (Eds), 1979. Indexnominumgenericorum (Plantarum).Regnum vegetabile, 100, 101, 102. Utrecht: Bohn, Scheltema & Holkema. FEDOROV, A. A. (Ed.), 1969. Chromosome numbers offiwering plants. Leningrad: Izd. Nauka. 8th ed. Boston: America Book Company. FERNALD, M. L., 1950. Groy’s Manual ofBo&q, FOWERAKER, C. E., 1916. Mat-plants and cushion-plants of Cass River Bed. Transactions of the New zealand Instituft, 49: 1-45. HANDEL-MAZZETTI, H., 1936. Symbolae Sinuae. Botanischc Ergebnkse der Expedifwn der Akademie der Wissenschaftfn in Wien nach Siidwcst-China 1914/1918, Vol. 7. Wien: Springer. HASHMI, S., & SIDDIQUI, S. A. 1974. Development of endosperm, embryo and seed in Paederia scandens Lour. Geobios (jodhpur), I : 183-184. HEGNAUER, R., 1973. Chotaxonomie der Pjanzen, Vol. 6.Basel & Stuttgart: Birkhauser. HITCHCOCK, C. L. & CRONQUIST, A., 1973. Flora of the PaciJicNorthwesf.Seattle: University of Washington Press. HOOKER, J. D., 1873. Rubiaceae. In G. Bentham & J. D. Hooker (Eds), Genera Plantarum 2(1):7-151. London: Lovell Reeve. INOUYE, H., INOUYE, S., SHIMOKAWA, N. & OKIGAWA, M., 1968. Die Iridoidglucoside aus Paederia scandens (Lour.) Merill var. mairei (Lkveilk) Hara. Tetrahedron Letters, 1968(6):683688. JAUBERT, C. & SPACH, E., 1842-1844. Illurhatwnes Planfarum Orinkdim, Vol. I . Paris: Roret. KOOIMAN, P., 1969. The Occurrence of asperulosidic glycmides in the Rubiacae. Acta Bofanica .Neerlandica, 18; 124-137. KOOIMAN, P. 1971. Ein phytochemischer Beitrag zur Liisung der Verwandtschaftsprobleme der Theligonaceae. ikeweichischr Botanische zeitschrifl, 119: 395-398. LINCHEVSKII, I. A., 1973. 0 rode Gaillonia A. Rich. ex DC. (De genere Gailfonia A. Rich. ex DC.) (Rubiaceae). Nowsti Sistematiki vsshich Rastenij, 10: 223-236. LOEW, E., 1905. Handbuch der Bliitenbwlogip, Vol. 3(2). Leipzig: Engelmann. MAKINO, T., 1909. Observations on the flora of Japan. Bofunical Magazine, ToTbkyo, 23: 81-95. MENDOZA-HEUER, I., 1977. Die Rubiaceen der Kanarischen Inseln. Berichft der Deutschen Bofunischen Gesellschaft, 90: 21 1-217. NAPP-ZINN, K., 1973. Anafomie des Blattes. II. Blattanatomie der Angiosbemen. Berlin-Stuttgart : Borntrager. NOWICKE, J. W. & SKVARLA, J. J., 1979. Pollen morphology: the potential influence in higher order systematics. Annals of the Missouri Botanical Garden, 66: 633-700. OHWI, J., 1965. Flora ofJapan. Washington, D.C.: Smithsonian Institution. OLIVER, W. R. B., 1935. The genus Coposma. Bernice P. Bishop Museum Bulletin. 132: 1-207. PRAGLOWSKI, J., 1973. The pollen morphology of the Theligonaceae with reference to taxonomy. Pollen el Spores, 15: 385-396. PUFF, C., 1982. Studies in Otiophora Zucc. (Rubiaceae). 4. The taxonomic position ofthe genus. Bofhalia, 14: (in press). PUFF, C. & MANTELL, D. E., 1982. The tribal position and affinities of Crocyllis (Rubiaceae). Botanische j’ahrbiicher, 103: (in press). QARAR, M., 1973. Die Gaftung Gaillonia A . Rich. (Rubiaceae-Anfhospeneae)und ihre Differenzitrung im SaharoSindischen und Irano-Turanischen R a m . Ph.D. Thesis, Philosopische Fakultat der Universitzt Wien. RICHARD, A., 1830. M h o i r e sur la famill8 des Rubiacles. Paris: Tastu (= Mhoires de la Socidfl PHistoire Nafurelle Paris, 5). ROBBRECHT, E. 1982. Pollen morphology of the tribes Anthospermeae and Paederieae (Rubiaceae) in relation to taxonomy. Bulletin du Jardin Botanique National de Belgique, 52: (in press). ROBBRECHT, E. & PUFF, C., 1981. Mericocalyx Bamps-synonymous with Ofiophorn Zucc. (Rubiaceae). Bulletin du j’ardin Botanique National de Belgique, 51: 143- 15 I . SCHUMANN, K., 1891. Rubiaceae. In A. Engler & K. Prantl(Eds),Die natiirlichen Pjanzenfamilien IV, 4: 1 156. Leipzig: Engelmann. SMITH, A. R., 1971. Jaubertia putorioides. Hooker‘s Icones Plantarum (5th Series), 7: tab. 3682. SMITH, W. W., 1921. Galium aberrans W. W. Sm., in Diagnoses specierum novarum in herbario Horti Regii Botanici Edinburgensis cognitarum. DI-DL. Notes from the Royal Botanic Garden Edinburgh, 13: I61 -162. THOMSON, G. M., 1881. On the fertilization, etc. of New Zealand flowering plants. Transactions of the New zealand InsfitUte, 13: 241-288. TROLL, W., 1964. Die Inzreszenzen, Band 1. Jena: Borntrager. ULBRICH, E., 1934. Thelygonaceae. In A. Engler & K. Prantl (Eds), Die natiirlichm Pjanzenfamilien, 2. Auj., I&: 368-378. Leipzig: Engelmann. VIJ, S. P., SHEKHAR, N., & KUTHIALA, R., 1981. Rubiaceae. In A. Love (Ed.), Chromosome number reports LXXI. Taxon, 30: 506-519. RUBIACEAE 311 VAN STEENIS, C. G. G. J., 1962. The land-bridge theory in botany. Blumea, 11: 235-372. VAN STEENIS, C. G. G. J., 1965. Plant geography of the mountain flora ofMt. Kinabalu. Proceedings of the Royal Sorirp /.ondon, Series B , 161: 7-38. VERDCOURT, B., 1958. Remarks on the classification of the Rubiaceae. Bulletin du Jardin Botunigue Bruxelles, 28: 209 -281. VERDCOURT, B.. 1976. Rubiaceae, Part 1. In R. M. Polhill (Ed.), Flora of Tropical East Afrua. ImdonTonbridge: Whitejriars Prrss. WEBERLING, F., 1977. Beitrige zur Morphologie der Rubiaceen-Infloreszenzen. Berichte der Deutschen Botanischrn Gesellschaft. 90: 19 1-209. WERTH, E., 1923. u b e r die Bestaubung von Viscum und Loranthus und die Frage der Primitivitat der Windbliitigkeit wie der Pollenblume bei den Angiospermen. I. Berichte der Deutschen Botanisthen Gesellschaft, 41: I5 I - 157. WETTSTEIN, R., 1935. Handbuch der gvtematisthen Botanik, 4. A u j . Leipzig & Wien: Deuticke. WILD, G. V. & ZOTOV, V. D., 1930. Notes on sexual expression in certain species ofNew Zealand Coprosmas. Transactions of the .New zealand Institute, 60: 547-556. WUNDERLICH, R., I97 I . Die systematische Stellung von Theligonum. Osterreuhische Botanische zeitschriff, 119: 329-394.
© Copyright 2026 Paperzz