Muelleria Volume 5 Numbers 2 & 3 March, 1983 NATIONAL HERBARIUM OF VICTORIA DEPARTMENT OF CROWN LANDS AND SURVEY Muelleria Volume 5, Number 2 September to December, 1981 CONTENTS Page New Australian species of erechthitoid Senecio (Asteraceae) — Robert O. Belcher 119 Two Western Australian Hydatellaceae — D. A. Cooke 123 A revision of the genus Platylobium Sm. (Papilionaceae) — J. H.Ross 127 A revision of Angianthus Wendl., sensu lato (Compositae: Inuleae: Gnaphaliinae), 1 — PS. Short 143 Editor: Helen I. Aston Published by the National Herbarium of Victoria (MEL). Royal Botanic Gardens, South Yarra, Victoria 3141, Austriia. D. M. Churchill, Director and Government Botanist. 57184/82 The date of distribution of Volume 5, number 1, was 22 March 1982. NEW AUSTRALIAN SPECIES OF ERECHTHITOID SENEQO (ASTERACEAE) by Robert O. Belcher* ABSTRACT Two new species of erechthitoid Senecio from Australia, 5. macrocarpus and S. cahillii, are described. This validates manuscript names which have been applied to certain specimens on determinavit slips since 1967. DESCRIPTION Senecio macrocarpus E Muell. ex Belcher, sp. nov. Erechthites hispidula sensu Benth. FI. Austr. 3:660 (1866), pro parte, non (A. Rich.) DC. (1838); sensu Black FI. S. Aust. 4:610 (1929), pro majore parte. CaudexhvtVxs perennis lignosus, radice palari debli et radicibus adventitis numerosis. Herba caulibus erectis 20 ad 40cm altis, saepe dense congestis; caulis arachnoideus superne glabratus, internodiis inferioribus brevibus, internodiis superioribus plus elongatis. Folia alterna linearia ad 10 cm longis et 24(-5) mm latis, acuta vel mucronata, marginibus revolutis, paginis abaxialibus arachnoideis vel hispidulis; folia inferiora numerosa, dense congesta, infima interdum denticulata; folia superiora parviora laxiora, saepe minute auriculata, apiculata. Inflorescentia raro simplex plerumque cymosa capitulis 6-8; pedicelli ad 5-30(-60) mm longi, ascendentes glabrati; bracteae et bracteoleae ciliolatae longiacuminatae apiculatae, bracteoleae numerosae prope apices pedicellorum saepe congestae, arachnoideae vel glabratae, 4-8 mm longae. Capitulum magnum 15(-18) mm longum, 15-20 mm latum ubi compressum in siccitate; phyllaria 16-21(-33), glabrata vel glabra, 10-13 mm longa, circa 1 mm lata, linearia, acuminata, marginibus anguste hyalinis, apicibus saepe subroseis. Flosculi numero 50-100(-150) varians ad 9-12(-15) mm longi; flosculi extime filiformes (3-)4(-5)-fidi pistillati vel staminodiis rudimentaris nonantheriferis, flosculi intermedii filiformes 4- vel 5-fidi staminodiis antherascentibus nonpolliniferis, flosculi medii anguste infundibulares 4- et 5-fidi staminibus 1-5 polliniferis. Achenia 4.5-5(-6) mm longa, rostrata, dense pilifera, pili brevissimi cinerascentes. Rootstock short, perennial, woody, with a weak primary root and numerous adventitious roots. Herb erect, 20^ cm tall; stems often densely congested, arachnoid, glabrate above, with lower intemodes short and upper intemodes more elongated. Leaves alternate, linear, to 10 cm long and 2-4(-5) mm wide, acute or mucronate, with revolute margins, abaxial surfaces arachnoid or hispidulous; lower leaves numerous, densely congested, the lowermost sometimes denticulate; upper leaves smaller and more lax, often minutely auriculate, apiculate. Inflorescence rarely simple, for the most part cymose with 6-8 capitula; pedicels 5-30(-60) mm long, ascending, glabrate; bracts and bracteoles ciliolate, long-acuminate, apiculate; bracteoles numerous, often crowded toward the apices of the pedicels, arachnoid or glabrate, 4-8 mm long. Capitulum large, 15(-18) mm long, 15-20 mm wide when compressed in drying; phyllaries 16-21(-33), glabrate or glabrous, 10-13 mm long, c. 1 mm wide, linear, acuminate, with margins narrowly hyaline, apices often subroseus. Florets 50 to 100 (to 150), 9-12(-15) mm long; outermost florets filiform, (3-)4(-5) fid, pistillate or with rudimentary staminodes not bearing anthers; intermediate florets filiform, 4- or 5-fid, with staminodes becoming antheriferous but non-polliniferous; central florets narrowly funnel-shaped, 4- and 5-fid, with 1 to 5 polliniferous stamens. Achenes 4.5-5(-6) mm long, brown, rostrate, densely hairy; hairs very short, greyish. The achenes of this species are most similar to those of S. quadridentatm Labill. but are larger and more hairy. They are very distinct from the short (2 mm long), black, thick- cylindric, densely white-haired achenes of S. squarrosus A. Rich. ♦Professor Emeritus of Biology, Eastern Michigan University, Ypsilanti, Michigan 48197, United States of America. Muelleria 5(2): 119-122 (1983). 119 120 Holotype: Victoria — Wimmera district, Walmar Station, 18. ix. 1860, Dallachy 23 (MEL 22987). Selected Specimens Examined: South Australia — Near Gooiwa, 23. xi. 1935, Cleland (AD 96736163). Werrabara Forest, 1890, Gill (MEL 22974). In plain at foot of mountain. Lofty Ranges, Mueller (MEL 22970, exserted styles trifid!). Vicinity of Clare cemetery, 20.x. 1961, Simon 1598 (ADW 24370, rather immature). Androssan, Tate (AD 96744130). Victoria — Backhous (K, det. by Bentham as Erechthites hispidula). Werribee, 24. ix. 1892, Morrison (PERTH, CANB 133472). Wendu Vale to Glenelg River, Robertson (NSW 27857). Skipton, Whan 22 (MEL 22982, NSW 27856). Geelong, ix.l905, Williamson 27855). Tasmania — Woodhole, Dietz (MEL 22976). South Esk River (MEL 22977, very immature). Distribution: South-eastern South Australia, south-western Victoria, and northern Tasmania. Discussion; This species is the taxon referred to as unnamed in my revision of Erechtites and erechthitoid Senecio {Ann. Mo. Bot. Gard. 43:49 (1956)), in the discussion of the confusion by earlier authors over S. squarrosus. I there suggested that it might be a polyploid of S. quadridentatus. This remains an attractive hypothesis, although still untested cytogenetically. In 1956 (1. c.) I decided from Black’s description that his treatment of Erechthites hispidula applied only to S. macrocarpus and excluded S. squarrosus (as well as S. hispidulus). Since then a careful examination of several copiously annotated sheets of specimens from his herbarium (in AD) has shown me that these sheets include specimens of both of these large-headed Australian taxa. S. squarrosus appears in his description only in regard to the leaves: “sometimes . . distantly toothed”. The phrases, “near the preceding” and “achenes as in E. quadridentata' \ clearly apply only to S. macrocarpus, hence my present revised citation as “Black pro majore parte”. Additionally, the synonymy I gave under S. squarrosus for E. hispidula (in 1956) should be changed from "'non Black” to “Black pro minore parte (ex specimina in herb.)”. The holotype is apparently the specimen cited by Bentham, l.c., p. 661, as “Wimmera, Dallachy”. I saw no such specimen at Kew, whereas the several labels on the holotype sheet at NffiL have all been initialled by Bentham. The specific epithet recognizes a very appropriate manuscript name given to the holotype by F. von Mueller. “Walmar” is apparently a misspelling for “Wilmer”, as a Wilmer Station previously existed north-east of Horsham, in the Wimmera district of western Victoria, and other collecting labels of Dallachy associate “Walmar” with locations close to Horsham (H. Aston, MEL, in litt.). S. macrocarpus most closely resembles S. squarrosus, especially in number of phyllaries, but the latter has lanceolate-linear leaves toothed to sublobular rather than linear entire or denticulate, smaller capitula (10-12 mm long rather than 15-18 mm), florets all 5-fid rather than 3- to 5-fid, and different achenes (see above). S. macrocarpus is also approached in number of phyllaries by S. pyrophilus Zoll. & Mor. ex Zoll., of Java and Timor, but the capitula of the latter are more slender and shorter, and are intermediate in size between 5. quadridentatus and S. macrocarpus, as are its achenes. S. pyrophilus, unlike S. macrocarpus, also has phyllaries pubescent on their adaxial faces. No other quadridentatoid species (including S. gunnii (Hook, f.) Belcher and S. runcinifolius Willis), exhibits mucronate leaf apices. Senecio cahillii Belcher, sp. nov. Caudex perennis; caulis suffruticosus erectus 60-120 cm altus, glabrescens, ramis grandis infra inflorescentium paucis vel nullis, saepe ramulis brevibus foliosis congestis e foliorum summonum axillis ortae. Folia sessilia, auriculis nullis vel simplicis vel bidentatis, supra auriculas subpetiolata cuneata glabrescentia, laminae foliorum costis subtus manifeste tumidis; folia basilia 10-13 cm longa, 1.5-2. 5 cm lata, oblanceolata, grosse dentata, vulgo marcida vel abscissa ante anthesin; folia media 6-10 cm longa, 0.4-0. 8(-1.2) cm lata, anguste oblanceolata vel linearia, irreguliter dentata vel aequaliter denticulata; folia superne similaria sed reducta. Inflorescentia corymbosa 121 erecta glabra vel puberula, capitulis 10-25, pedicelli ad 5-10 mm longi, bracteae et bracteolae late subulatae apicibus fuscatis. Capitula 5-6 mm longa, phyllaria involucri 8 or 9(-ll), 4.5-5 mm longa, ca. 0.5 mm lata, binervia, valde carinata in alabastro, applanata per anthesin, demum rigescentia late extendentia post fructificantem, apicibus fuscatis ciliolatis. Flosculi numero 8-16(-21); flosculus pistillatus 4.5-5 mm longus, basi 0.2-0.25 mm diametro, tubo filiformi, apice tenuiter infundibuliformi; flosculi pistillati extimi 3-fidi, intermedii 4-fidi; flosculi medii perfecti 5-5.5 mm longi infundibuliformi 4- vel 5-fidi, ramis stylorum 0.5 mm longis. Achenia 2.5 mm longa subcylindracea non rostrata ferruginea, porcis decern latis humilibus planis, trichomatibus O.I mm longis subappressis in sulcis perangustis; pappus pilis, tenuibus niveis, flosculos aequans, demum phyllaria excedens ad 2 mm. Similis ad 5. biserratum sed foliis angustioribus minus grosse dentatis, flosculis extimis 3-fidis vice 5-fidorum, et ramis stylorum magnitudinedimidiatis. AbS. minimo differt in achenio longiore trichomatibus in sulcis non insidens porcio, in folio angustiore et magis irreguliter dentato, et in phyllario breviore et grosso. Rootstock perennial. Stems'^oody below, erect, to 60-120 cm tail, glabrescent, with few or no major branches below the inflorescence, often with short branchlets congested with leaves and borne in the axils of the uppermost leaves. Leaves sessile, with auricles lacking or simple or bidentate, above the auricles subpetiolate and cuneate, glabrescent, the blades of the leaves with midribs obviously swollen on the undersides; lower leaves 10-13 cm long, 1.5-2. 5 cm wide, oblanceolate, coarsely dentate, generally withering or abscissing prior to anthesis; middle leaves 6-10 cm long, 0.4-0. 8(-1.2) cm wide, narrow- oblanceolate or linear, irregularly dentate or equally denticulate; upper leaves similar but reduced. Inflorescence corymbose, erect, glabrous or puberulent, with 10-25 capitula per corymb on pedicels 5-10 mm long, bracts and bracteoles broadly subulate with apices darkened. Capitula 5-6 mm long, phyllaries of the involucre 8 or 9(-ll), 4.5-5 mm long, 0.5 mm wide, two-nerved, strongly keeled in bud, flattening at anthesis, eventually becoming rigid and widely extended after fruiting, the apices darkened and ciliolate. Florets 8-16(-21) in number; the pistillate floret 4.5-5 mm long, 0.2-0.25 mm in diameter at the base, with the tube filiform, its apex slenderly infundibuliform; outermost pistillate florets 3-fid, intermediate pistillate florets 4-fid, the central florets perfect, 5-5.5 mm long, infundibuliform, 4- or 5-fid, branches of the styles 0.5 mm long. Achenes2,5 mm long, subcylindrical, non-rostrate, reddish-brown, with ten broad low flat ridges, with trichomes 0.1 mm long, subappressed in the very narrow grooves; pappus with hairs slender, snowy white, equalling the florets, eventually exceeding the phyllaries by 2 mm. Similar to S, biserratus Belcher but with leaves more narrow and less coarsely toothed, the outermost florets 3-fid instead of 5-fid, and the branches of the styles only half as long. Differs from S. minimus Poir. in the slightly longer achene with trichomes in the grooves instead of on the ridges, in the narrower and more irregularly toothed leaf, and in the shorter and coarser phyllary. Type Collection: Victoria — East Gippsland, Buchan River near junaion of Reedy River, 6.ii.l973, A. C. Beauglehole 41406 (Holotype: MEL 501429. Isotype: MEL 501428). Selected Specimens Examined: Western Australia — Wellington District, State Forest #15, approx. 16 km SE. of Harvey, in jarrah forest dieback area partially burned in 1965, 25.viii.1967, Ca/i/// 22 (PERTH). Ibidem, I3.X.1967, R. & R. Belcher 245, 250 (EMC). Pasture on Roelands — Collie road about 13 km east of South Western Highway 30.ix.I967, R. & R, Belcher 183 (EMC). Queensland — Bunya Mountains, between Noobler’s Lookout and Burton’s Well, 29. xi. 1967, Belcher 810-A (EMQ. Gympie, Kenny (BRI 073659). New South Wales — Warragamba Dam area, half a mile S of Nattai Junction, 23.ii.1967, Briggs 1118 (NSW 95691). Central Tablelands, Mt. Victoria, 15.i.l892, Fletcher {NSW 27821). Maryland, Wylie Creek road turn-off, 10.i.l956, M Gray (CANB 97679). Central Coast, HQl Top, i.l903. Maiden (F; NSW 27819) New England, Goff’s Gully, iii., C. Stuart (MEL 22750). Victoria — East Gippsland, Timbarra River, Old Ensay Track, 14.xii.l970, A. C. Beauglehole 35491 (EMC; MEL 540861). Distribution: Eastern slopes of the Great Dividing Range from south-eastern Queensland abundantly through New South Wales to East Gippsland in Victoria; recently locally adventive in south-western Western Australia. 122 Discussion: The specific epithet recognises Mr. C. O. Cahill, Weed Control Officer at Bunbury, Western Australia, in 1967. He was the first to refer material of this taxon from that State for identification and also kindly directed me to oecurrences in his area. The species had become locally common as an aggressive invader in paddocks. It also occurred in jarrah forest on dieback areas produced by the root parasite, Phytophthora cinnamonii. Ecological data on the labels of the many New South Wales collections held at the National Herbarium of that State strongly suggest a preference for moist stream sides and disturbed areas. S. cahillii is best recognised in the field by its very erect habit, lack of lower branches above the basal stool (from which it may sprout very prolifically to form large clumps) and especially by the densely crowded branchlets in the axils of the upper leaves, the leaves of these branchlets being short and linear or nearly so. The isotype (MEL 501428) is unusual in that the right-hand specimen has 3 large flowering branches from fairly low on the stem. This, however, is due to the main stem having been broken off earlier, just above the origin of the branches. Manuscript received 5 November 1981. TWO WESTERN AUSTRALIAN HYDATELLACEAE by D. A. Cooke* ABSTRACT Trithuria bibracteata and Hydatelta dioica, two species from Western Australia previously lacking valid names, are described. INTRODUCTION In 1903, material of two species of Hydatellaceae collected by W. V. Fitzgerald in seasonal swamps at Midland Junction near Perth was sent to the Royal Botanic Gardens, Kew, for identification. The names Trithuria bibracteata and T macranthera were assigned to these species by O. Stapf but never published by him. They have passed into literature as nomina nuda. As an account of the family is being prepared for the Flora of Australia, this opportunity is taken to publish valid names for the two species. DESCRIPTIONS IVithuria bibracteata Stapf ex D. A. Cooke, sp. nov. Taxonomic Synonym: T. bibracteata Stapf ex W. V. Fitzgerald in J. W. Aust. Nat. Hist. Soc. 2(1):36 (1904) nomen nudum; Blackall & Grieve 1:58 (1954), sans descr. Lat. Herba annua perpusilla rubescens caespituli foliosi ad 1 cm diametro formans. Caulis brevissimus pilibus numerosis usque ad 2 mm longis. Folio basalia linearia 5-6 mm longa usque ad 0.4 mm lata glabra, basibus subhyalinis parce dilatis, nervis mediis inconspicuis, apicibus acutis. Scapi absentes. Capitula numerosa sessilia unumquidque bracteis 2 involucratum, flosculis masculis 1-2, flosculis foeminis 6-10. Bracteae lanceolatae 2-3 mm longae subhyalinae, basibus dilatis ad c. 1.2 mm latis vaginantibus. Stamen filamento usque ad 2.5 mm longo, anthero oblong-elliptico 0.4-0.6 mm longo c. 0.15 mm lato. Ovarium breviter stipitatum, ovoideum c. 0.3 mm longum, pilis stigmaticis 2-5 c . 2 mm longis terminalibus caducis. Fructus in stipite fragili usque ad 0.4 mm longo, ovoid-trigonus c. 0.5 mm longus, superficiebus 3 delicatis pallidis inter costas vasculares 3 aeque dispositas, semen liberandum fatiscens. Semen ovoideum 0.4-0.5 mm longum; testa brunnea retifoveata. (Descriptio typi.) Very small annual herb, becoming red-tinted, forming leafy tufts to 1 cm in diameter. Stem very short, with numerous hairs up to 2 mm long. Leaves basal, linear, 5-6 mm long and up to 0.4 mm wide, glabrous with slightly dilated subhyaline bases, faint midvdns and acute apices. Scapes absent. Heads numerous, sessile, each with an involucre of 2 bracts, 1-2 male florets and 6-10 female florets. Bracts lanceolate, 2-3 mm long, subhyaline, with sheathing bases dilated to c. 1.2 mm wide. Stamen with filament up to 2.5 mm long, anther oblong-elliptic 0.4-0. 6 mm long and c. 0.15 mm wide. Ovary shortly stipitate, ovoid, c. 0.3 mm long, with 2-5 caducous terminal stigmatic hairs c. 2 mm long. Fruit on a fragile pedicel up to 0.4 mm long, ovoid-trigonous, c. 0.5 mm long, with 3 pale delicate panels between 3 equally spaced ribs containing vascular bundles, disintegrating to release the seed. Seed ovoid, 0.4-0. 6 mm long; testa brown, reticulate-foveate. Type Collection: Boyanup, 15.x. 1947, R. D. Royce 2265 i}\o\o\ PERTH) Selected Specimens Examined (total 8): Perup River, E. of Manjimup, x.1948, H. Butler s.n. (PERTH); Midland Junction, ix.l901, W. V. Fitzgerald s.n. (PERTH); Midland Junction, x.1903, W. V, Fitzgerald s.n. (NSW 148478, PERTH). *9/51 Marne Street, South Yarra, Victoria 3141. Muelleria 5(2): 123-125 (1983). 123 124 Distribution: Scattered in seasonally wet habitats in the Darling district (Beard, 1980) of Western Australia, where widespread in the Drummond subdistrict and also recorded from the Menzies subdistrict. Notes: Trithuria bibracteata is closely related to T. lanterna D. A. Cooke (1981), which it resembles in habit, foliage and inflorescence. The fruit is a morphological link between the hyaline, indehiscent fruit of T. lanterna and the dehiscent capsule of T. submersa Hook. f. In T. bibracteata the pericarp is thin and fragile, generally crumbling irregularly as the whole plant dries out, but sometimes splitting along the vascular ribs as in T. submersa. The thick sculptured testa confirms that the seed, rather than the fruit, is the disseminule. The hairs at the base of the plant each consist of a single row of up to 6 hollow, elongated, thin-walled cells. Similar hairs have been observed on the stem among the leaf- bases in all Australian species of Hydatellaceae, but are greatly reduced or absent in many specimens. Being characteristic of the family, they are thus of little diagnostic value. Hydatella dioica D. A. Cooke, sp. nov. Taxonomic Synonyms: Trithuria micranthera (misspelling of macranthera) Stapf ex W. V. Fitzgerald in J. W. Aust. Nat. Hist. Soc. 2(1):36 (1904) nomen nudum; Blackall & Grieve 1:58 (1954), sans descr. Lat. T. macranthera BortQnschlsLgev et al. in Bot. Not. 119:161 (1966) nomen nudum. Herba annua dioica rubescens usque ad 4 cm alta. Caulis brevissimus. Folia basalia linearia usque ad 25 mm longa et 1 mm lata, glabra, basibus hyalinis parce dilatis, nervis mediis prominentibus, apicibus acutis. Capitula mascula pluria; unumquidque bracteis 2 involucratum, super scapo erecto nonramoso tereti nudo usque ad 3 cm alto terminans. Bracteae oppositae erectae lanceolatae 7-8 mm longae arete vaginantes glabrae subhyalinae nervis mediis prominentibus. Stamina 8-10 alium ex alio exserta, filamentis usque ad 10 mm longis flexuosis persistentibus, antheris linearis c. 3 mm longis et 0.2 mm latis caducis. Capitula foemina non vidi. (Descriptio typi.) Annual dioecious herb to 4 cm tall. Stem very short. Leaves basal, linear, to 25 mm long and 1 mm wide, glabrous, with slightly dilated subhyaline bases, prominent midveins, and acute apices. Male heads several; each with an involucre of 2 bracts, terminating an erect unbranched naked terete scape up to 3 cm tall. Bracts opposite, erect, lanceolate, 7-8 mm long, closely sheathing, glabrous, subhyaline with prominent midveins. Stamens 8-10, exserted one after another, with flexuose persistent filaments up to 10 mm long and caducous linear anthers c. 3 mm long and 0.2 mm wide. Female heads not seen. Type Collection: Midland Junction, 16. xi. 1898, A. Morrison s.n. (Holo: PERTH). Also Examined (total 5): Midland Junction, ix.l901, W. V. Fitzgerald s.n. (NSW 148484, PERTH); Midland Junction, x.1903, W. V. Fitzgerald s.n. (NSW 148483). Distribution: Known only from seasonal swamps at Midland Junction, Darling district. Western Australia, where possibly now extinct due to development. Notes: This is the only dioecious species known in the Hydatellaceae and is therefore placed with the other species having homogamous inflorescences in the genus Hydatella, rather than in Trithuria with heterogamous inflorescences. The specimens examined have leaves and male heads similar to those of Hydatella australis Diels and differ primarily in the greater size and numbers of organs. The pollen grains of this species were described and illustrated by Bortenschlager et al. (1966). 125 ACKNOWLEDGEMENTS I would like to thank the Directors of PERTH and NSW herbaria for the loan of collections and also the staff of the National Herbarium of Victoria (MEL), where this paper was prepared. This work was carried out while the author was acting as a consultant to the Bureau of Flora and Fauna, Department of Home Affairs and Environment. REFERENCES Beard, J. S. (1980). A new phytogeographic map of Western Australia. W. Austral. Herb. Res. Notes '^'.'^>1 -5%. Blackall, W. E. & Grieve, B. J. (1954). ‘How to Know Western Australian Wildflowers*. Part 1. (University of Western Australia Press: Perth). Bortenschlager, S., Erdtman, G. & Praglowski, J. (1966). PollenmorphologischeNotizen ubereinige Bliitenp- flanzen incertae sedis. Bot. Not. 119:160-168. Cooke, D. A. (1981). New Species of Schoenus (Cyperaceae) and Trithuria (Hydatellaceae). Mue/leria 4:299-303. Fitzgerald, W. V. (1904). Additions to the West Australian Flora. J. W. Aust. Nat. Hist. Soc. 2(l):3-36. Manuscript received 13 January 1982. A REVISION OF THE GENUS PLATYLOBIUM Sm. (PAPILIONACEAE) by J. H. Ross* ABSTRACT The endemic Australian genus Platylobium Sm. is revised. Four species are recognised and P. gracile Dum.-Cours. and P. rotundifolium Colla are rejected as names of uncertain application. Formal infraspecific taxa in P. formosum are discussed but not upheld. Descriptions, a key to the identification of species, illustrations and distribution maps are provided, together with notes on ecology and relationships. INTRODUCTION Platylobium, a small genus of four species described by J. E. Smith, Spec. Bot. New. Holl. 1:17, t.6 (1793), is confined to eastern Australia occurring from the vicinity of Wide Bay in south-eastern Queensland southwards to Tasmania and westward to Kangaroo Island in South Australia. The present centre of distribution of the genus is in Victoria where all species occur. The generic name is taken from the Greek and alludes to the broad pods found in species of the genus. Platylobium is a member of the tribe Bossiaeeae (Benth.) Hutch, and belongs to the Bossiaea group of genera which are characterised by having uniform dorsifixed anthers with a broad connective, seeds with generally hooded cap-like arils and slender curved radicles exserted from the cotyledons, and plants with a tendency to accumulate canavanine (Polhill, 1976, 1981). The Bossiaea group comprises the genera Bossiaea Vent., Platylobium, Goodia Salisb., Aenictophyton A. Lee, Ptychosema Benth. and Muelleranthus Hutch.. Platylobium and Bossiaea are readily distinguished from the other genera in having the flowers subtended by a series of papery brown scales. Platylobium was described prior to Bossiaea and a number of taxa initially described ^ species of Platylobium are referable to Bossiaea. This is not surprising as Platylobium is unquestionably very closely allied to Bossiaea and the view has been expressed (Polhill, 1976) that Platylobium is scarcely distinct from Bossiaea at generic level. Platylobium differs from Bossiaea in the development of a distinct wing beyond the upper sutural nerve of the pod and in the valves being very thin and revolute on dehiscence, in the extreme enlargement of the two upper calyx-lobes, and in having a diploid chromosome number of 16 as opposed to the 18 found in Bossiaea. In addition, Ferguson & Skvarla (1981) found that the pollen of Platylobium has a well-defined endoaperture whereas that of Bossiaea has a poorly defined interruption in the endexine. ♦National Herbarium of Victoria, Birdwood Avenue, South Yarra, Victoria 3141. Muelleria 5(2): 127-141 (1983). 127 128 The extreme enlargement of the two upper calyx-lobes in Platylobium does not distinguish the two genera absolutely as the two upper calyx-lobes in some species of Bossiaea are greatly enlarged as in Platylobium, and the pod- valves in some species of Bossiaea are narrowly winged beyond the upper sutural nerve and occasionally are slightly revolute on dehiscence. The four species of Platylobium have a distinctive “look” and appear to represent a fairly natural group. Provided that the close relationship between the genus and Bossiaea is acknowledged, the differences between the two provide a workable distinction and no advantage is seen in including the few species of Platylobium in the much larger and rather heterogeneous Bossiaea. As Platylobium was described first, its amalgamation with Bossiaea would necessitate the conservation of the latter name. In habit, all species of Platylobium are shrubs or subshrubs with woody rootstocks. The leaves in P. alternifolium are alternate whereas in the other species they are opposite except for the occasional occurrence in P. formosum of a variant in which some of the leaves are borne alternately. Leaf size and shape in all species except P. alternifolium varies considerably. The variation in leaf shape within P obtusangulum and P. triangulare is such that unfortunately the character cannot be employed to differentiate the two species. The range of variation in leaf size and shape in P formosum so great that the extremes look quite different but when the range of variation is inspected it is clear that the extremes are linked by numerous and varied intermediates. Floral characters provide useful means of differentiating the species but the pods and seeds are relatively uniform within the genus. TAXONOMY Platylobium Sm., Spec. Dot. New Holl. 1:17, t.6 (1793); Trans. Linn. Soc. Lond. 2:350 (1794); Trans. Linn. Soc. Lond. 9:302 (1808); Willd., Sp. PI. 3:921 (1802); DC., Prodr. 2:116 (1825); G. Don, Gen. Syst. 2:127 (1832); Benth., FI. Austr. 2:152 (1864); Benth. in Benth. & Hook.f., Gen. PI. 1:473 (1865); Taub. in Engl., Pflanzenfam. 3, 3:216 (1893); Hutch., Gen. H. PI. 1:348 (1964); Polhill, Bot. Syst. 1:311 (1976); Polhill in Advances in Legume Systematics 1:395 (1981). Type: P. formosum Sm. Shrubs or Subshrubs, sometimes prostrate or occasionally semi-scandent, with one to several slender, terete, unarmed stems arising from a woody rootstock. Stipules ovate to narrow-ovate, striate, persistent; stipellae absent. Leaves opposite or, less frequently, alternate, simple or unifoliolate, sessile or petiolate, narrow-ovate to broad-ovate- cordate, triangular-ovate, hastate, trilobate, cordate-hastate or occasionally almost orbicular, the apex and usually the angles pungent-pointed, reticulately veined. Flowers 1-several from the axils, almost sessile or on long slender pedicels, orange-yellow and red or purplish-brown, subtended by a series of distichous brown papery scales, the bract indistinguishable from the scales; bracteoles paired, at or near the apex of the pedicel, similar to the scales but often larger. Calyx sparingly to densely pubescent throughout except in P alternifolium, the two upper lobes greatly enlarged, broadly rounded, usually united for less than half their length, the three lower lobes narrow and acute, persisting in fruit. Standard orbicular or reniform, with a well developed claw; wing petals rounded or obtuse apically, much shorter than the standard; keel petals almost as long as the wings. S^am^/7-filaments united in a sheath split open on one side; anthers uniform, dorsifixed, with a broad connective. Ovary sessile or stipitate, style slender, curved, with a small terminal stigma. Pods sessile or stipitate oblong, flat, with a conspicuous thin wing 2-5 mm wide beyond the upper sutural nerve, transversely striate- nerved, several-seeded; valves thin, elastically revolute, revealing a glossy inner surface. Seeds ovate-ellipsoid or ellipsoid, plump, with a small hilum on a long side covered by a hooded cap-like aril; radicle, slender, curved. Key to Species 1. Leaves alternate; calyx glabrous externally except for a conspicuous fringe of hairs on the margins of the lobes P. alternifolium 129 1. Leaves opposite, very infrequently some leaves alternate {P. formosum) but then the calyx clothed externally with spreading or appressed hairs throughout 2. Flowers and pods subsessile, the pedicel concealed by a series of conspicuous brown papery bracts and scales 3. P. obtusangulum 2. Flowers and pods distinctly pedicellate, the pedicels exceeding and exserted from the bracts and scales 3 . Leaves broad-triangular or cordate-hastate, the lateral angles and apex usually pungent 2. P. triangulare 3. Leaves narrow-ovate to broad-ovate -cordate, the apex pungent but the lateral angles usually rounded or obtuse and not pungent 4. P formosum 1. Platylobium alternifolium F. Muell., S. Sci. Rec. 3:99 (1883); J. H. Willis, Handb. PL Viet. 2:277 (1973). Syntypes: See under notes. Lectotype, here selected: Victoria, Mt. William, 1883, D, Sullivan (MEL 569729!). Small subshrub with one to several prostrate spreading or trailing stems, the young stems sparingly to densely clothed with spreading hairs. Stipules ovate or narrow-ovate, up to 3.5x2 mm. Leaves alternate, unifoliolate: petiole 0.3-2 cm long, sparingly to densely clothed with spreading hairs; lamina cordate to ovate or triangular-ovate with an acute apex or almost orbicular and rounded apically, (0.6)l-2.6 xO.6-2.5 cm, upper and lower surfaces sparingly to densely clothed with spreading hairs when young but becoming ^abrous and minutely scabrid with age, venation fairly promiment. Flowers solitary, axillary, on glabrous to sparingly pubescent pedicels 1.5-2 cm long, the pedicels with a series of distichous scales and bracts along their length which are glabrous except for marginal cilia, the paired bracteoles below the calyx 5. 5-7. 5 X2.5-3.5 mm, glabrous apart from marginal cilia. Ca/yx pinkish-red, glabrous outside except for a conspicuous fringe of hairs on the margins of the lobes: 2 upper lobes 7-9.5 mm long (including the basal tubeofupto3.5 mm), 4-5 mm wide, the 3 lower lobes up to 3. 5x1. 2 mm, narrow- triangular. Corolla: standard oblate, 13.5-15.5 mm long including a claw up to 3.5 mm long, 15-18 mm wide, deeply emarginate apically, brick-red outside, yellow inside with a deep yellow basal horse-shoe shaped throat surrounded by a reddish-purplish zone; wings up to 10.5 mm long including a claw up to 2 mm long, up to 3.5 mm wide, auricled, yellow except for a central dark red or purplish band; keel petals up to 9.5 mm long including a claw up to 1.5 mm long, 4-4.6 mm wide, auricled, dark purplish in upper half. Stamensu^ to 9.5 mm long. Ovary shortly stipitate, up to 5 mm long, margins with long white hairs but otherwise glabrous; style with long white basal hairs, glabrous above. Pods oblong, 1.3-2 xO.7-2 cm, with a thin conspicuous wing up to 2 mm wide beyond the upper sutural nerve, (l)2-4-seeded, with scattered hairs on the margins but otherwise glabrous. Seed ellipsoid, 3.6 x 2.5 mm (only one seen) (Fig. 2). P. alternifolium is confined to the Grampians and Bolangum Ranges in western Victoria where it occurs on moist east-facing slopes in closed Eucalyptus woodland (Fig. 3). Representative Specimens: Victoria — Grampians, near crest of Mt. Difficult Range, E. of Wartook Reservoir, lO.x.1962, T. B. Muir 2607 (MEL 99432). Grampians, Mt. Difficult Range (East), 2.vi.l969, A. C. Beauglehole 30754 (MEL 99431). Grampians, Mt. Difficult Range road, 12 km N. of its junction with Halls Gap — Zumstein’s road, 5.x. 1980, M. G. Corrick 6801 & P S. Short (MEL 576033). Notes: Mueller based his description of R alternifolium on the following specimens: “On Mount Disappointment, F.v.M.; on Mount Ben Nevis, Ch. Green; on Mount William, Sullivan and Miller.” Four syntypes are housed in the National Herbarium of Victoria (MEL), namely, Mt. Disappointment, Mueller (MEL 569726), Mt. Ben Nevis, Ch. Green (MEL 569727) and Mt. William, Sullivan (MEL 569728, MEL 569729). I have not succeeded in locating the syntype from Mt. William collected by Miller although the possibility exists that the flowering specimen received at Kew Herbarium in Oct. 1884 from MEL bearing the locality ‘ ‘Grampians”, but without any indication of the collector, is the Miller syntype. The two syntypes collected by Sullivan have a different facies from that of the specimens collected by Mueller and by Green which suggests that Mueller inadvertently 130 Fig. 2. Platylobium alternifolium. a — flowering twig, x 1; b — flower, with a series of distichous scal^ and bracts along the length of the pedicel, x 2; c — calyx opened out (upper lobes on left), x 4; d standard, x 3 ; e — wing petal, x 3; f— keel petal, x 3; g — staminal tube opened out, x 3; h — gynoecium, x 3; i — pod, side view, x I/ 2 ; j — pod, after dehiscence, x F/i; k — seed, side view, x 6; 1 — a-h from M. G. Corrick 6801 & P. S. Short (MEL 576033); i-1 from A. C. Beauglehole 30754 (MEL 99431). 131 included discordant elements in the protologue of P. alternifolium. The Mueller and Green syntypes are rather poor sterile specimens and were no doubt included by Mueller as they have alternate leaves, an occurrence not then recorded in any other Platylobium species and to which much significance was attached. One of the Sullivan specimens (MEL 569728) is sterile but the other (MEL 569729) is fertile and shows the series of scales and bracts along the length of the pedicel, the calyx which is glabrous except for a conspicuous fringe of marginal hairs, and a young fruit with long white marginal hairs but otherwise glabrous. The specimen from the Grampians in K referred to above matches the two Sullivan collections. P. formosum typically has opposite leaves but a variant of this polymorphic species in which the leaves are sometimes alternate occurs sporadically in Victoria from Blackwood in the west to Gippsland in the east. When fertile, such specimens may be distinguished from P. alternifolium without difficulty as the calyces and young pods are densely pubescent throughout and the pedicels are exserted from the basal bracts and scales. Sterile specimens are more difficult to place with certainty but I am reasonably satisfied that the specimens collected by Mueller and Green are in fact referable to P formosum and not to P. alternifolium. In view of this, I now select Sullivan (MEL 569729) from Mt. William as the lectotype of P, alternifolium. As a result of the exclusion of the variant of P. formosum with petiolate alternate leaves from the circumscription of P. alternifolium, P alternifolium now has a much more restricted distribution in Victoria being confined apparently to the Grampians and Bolangum Ranges. The only specimen (Audus s.n., MEL 569725) from the Bolangum Ranges is sterile and, because of the uncertainty attached to sterile specimens in this genus, fertile material from this locality is required to confirm the occurrence there of P alternifolium. Audas (1921) made special mention of the unexpected occurrence of P alternifolium in the Bolangum Ranges and there is no reason to suppose that his specimen was not collected there. Audas gave no indication of the route he took after leaving the Kingston mine so it is not possible to re-trace his steps. Unfortunately a recent attempt by my colleague Mrs M. G. Corrick to locate P. alternifolium in the Bolangum Ranges was unsuccessful. P. alternifolium is the only species in the genus in which the calyces are not clothed with appressed or spreading hairs throughout. The consistently alternate leaves distinguish P. alternifolium from the other species except for the occasional variant of P. formosum. 2. Platylobium triangulare R.Br. in Ait.f., Hort. Kew. ed. 2, 4:266 (1812); DC., Prodr. 2:116 (1825) pro parte excl. ref. Sims in Curtis’s Bot. Mag. t.l508; Benth., FI. Austr. 2:152 (1864); J. M. Black, FI. S. Austr. ed. 2:445 (1948); W. Curtis , Stud. FI. Tasm. 1:141 (1956); J. H. Willis, Handb. PI. Viet. 2:278 (1973); non Sims in Curtis’s Bot. Mag. t.l508 (1812); non Hook.f., FI. Tasm. 1:% (1856). Syntype: See under notes. Lectotype, here selected: Tasmania, King’s Island, 23.iv.1802, P. Brown (BM; MEL, photo.). P. murrayanum Hook, in Curtis’s Bot. Mat. t.3259 (1833); Hook.f., FI. Tasm. 1:96 (1856). Type: Curtis’s Bot. Mag. t.3259 (iconotype). Small shrub with prostrate or ascending, slender, glabrous to densely villous stems up to 50 cm long. Stipules ovate, up to 2x1 mm. Leaves opposite, simple, broadly triangular or cordate-hastate, the lateral angles and apex usually pungent or the lateral angles of some lower leaves sometimes rounded, (l)1.3-3.2x(0.6)l-2.6 cm, glabrous above and minutely scabrid or with few scattered hairs especially on the midrib, lower surface usually with scattered hairs especially on midrib, larger veins and margins but sometimes glabrescent. Flowers usually solitary but 1-3 per axil, on villous pedicels 0.6-1. 8 cm long, the pedicels exserted from the basal bracts and scales which are glabrous except for marginal cilia, the paired bracteoles below the calyx 2-3.7 x 1.5-2 mm, glabrous except for the margins, usually reflexed. Calyx densely clothed with long appressed or slightly spreading hairs: 2 upper lobes 9.5-12.5 mm long (including the basal tube of 4.5-6 mm), 5-6.5 mm wide, the 3 lower lobes 4.5-6. 5 x 1.2-1. 5 mm, the centre lobe longest and narrowest. Corolla: standard oblate, 12-16 mm long including a claw up to 3 mm long, 15-18.5 mm wide, deeply emarginate apically, dark purplish-brown or 132 pinkish-red outside, orange-yellow inside with a yellow basal horse-shoe shaped throat surrounded by a reddish-purple zone; wings 10-13 mm long including a claw up to 3 mm long, 4. 5-5. 5 mm wide, dark reddish basally, orange-yellow apically, auricled; keel petals 9.7-12.5 mm long including a claw up to 3 mm long, 4.5-5 mm wide, dark purplish or reddish apically, white basally. Stamens up to 11mm long. Ovary 5.5-9 mm long including a stipe 0.5-1.75 mm long, margins with long white hairs but otherwise glabrous; style with long basal hairs, glabrous above. Pods oblong, 2.2-3. 3 x 1.2-1. 8 cm, on a stipe 2^.5 mm long, with a thin conspicuous wing up to 4 mm wide beyond the upper sutural nerve, 3-5-seeded, with scattered hairs on the margins but otherwise glabrous. Seeds ellipsoid, 2.5-3. 6 x 1. 6-2.3 mm (Fig. 4). P. triangulare occurs in southern Victoria and in Tasmania where it is found most frequently in open woodland and heathland. There is a curious distributional discontinuity in Victoria from Wilson’s Promontory in the east to Orford in the west from where the species extends its range westward as far as the Glenelg River. The area between Wilson’s Promontory and Orford is reasonably well collected and the discontinuity appears to be genuine rather than the result of inadequate collecting. According to the label accompanying MEL 572090, the specimen was collected by L. Henry from the Upper Darling near the Queensland border in 1884. However, as this locality is so far removed from the nearest known populations of P triangulare, it seems probable that the label does not belong with the specimen (Fig. 3). Representative Specimens: Victoria — Portland Distr., 4.8 km S.E. of Gorae West along main Portland-Nelson Rd., 22.x. 1960, H. I. Aston 726 (MEL 570858). 5 km E. of Casterton-Dartmoor main road along Moonlight Rd., 2.xi.l980, P. S. Short 1222 (MEL 1522094). Wilson’s Promontory, Waterloo Bay, 6.xi.l980, M. G. Corrick 7078 (MEL 576042). Tasmania — St. Helen’s, x.1945, W. M. Curtis {HO 115500). 3.2 km N.W. of Coles Bay, 14.x. 1967, J. H. Hemsley 6297 {HO 11557, MEL 570862). Track above Botanical Creek, Freycinet Peninsula, 21.1.1980, A. M. Buchanan 152 (HO 36184). Notes: It is not clear whether R. Brown based his description of P triangulare on a plant raised at Kew Gardens from seed introduced by himself in 1805 from Tasmania, whether the description was based on his specimens collected in Australia, or whether it was based on both. I have not succeeded in locating a specimen in BM from a plant cultivated at Kew. In BM there are, however, three R. Brown collections from Australia, namely, an unnumbered specimen collected on King’s Island, Bass Strait, on 23 Apr. 1802, an unnumbered specimen collected at Arthur’s Seat, Port Phillip, Victoria, in Jan. 1804, and no. 5073 . Unfortunately doubt surrounds the provenance of the latter specimen: it is not clear whether it was collected on King’s Island on 23 Apr. 1802, at Port Dalrymple in Jan. 1804 or at Arthur’s Seat in Jan. 1804. I now select the specimen in BM collected by R. Brown on King’s Island, Tasmania, on 23 Apr. 1804 as the lectotype of P. triangulare. 133 Fig. 4. Platylobium obtusangulum. a — flowering twig, x 1; b — flower, showing the short pedicel concealed by imbricate bracts and scales, x 2; c — calyx opened out (upper lobes on right), x 2; d — gynoecium, x 3; e — pod,x \ .P. triangulare. f — flowing twig, x 1; g — flower, showing the long pedicel exserted from the basal bracts and scales, x 2; h — calyx opened out, x 2; i — gynoecium, x 3; j — pod, x 1. a-d from J. H. Ross 2475 (MEL 531009); e from A. C. Beauglehole 30037 (MEL 100004); f-i from H. I. Aston 726 (MEL 570858); j from L. Renfrey (MEL 570664). 134 P. obtusangulum P. triangulare Pedicels less than 0.5 cm long, completely concealed by the imbricate bracts and scales 0.6-1. 8 cm long, exserted from the basal bracts and scales Bracteoles 7.5-9 X 4-5.5 mm, remaining erect 2-3.7 X 1.5-2 mm, usually reflexed 3 lower calyx lobes 8-9.5 mm long 4. 5-6. 5 mm long Ovary densely villous throughout glabrous apart from long white hairs on the margins Pods subsessile and densely pubescent throughout, at least when young on a stipe 2-4.5 mm and glabrous apart from scattered hairs on the margins Table 1. Comparison of the diagnostic differences between P, obtiisangulum and P. triangulare. As R. Brown’s description of R triangulare was brief, it is not surprising that the species was confused with the taxon later described by W. J. Hooker as R obtusangulum. The two species are superficially similar and the leaves of R. obtusangulum are often as triangular as are those of R. triangulare. Diagnostic differences are given in table 1 and these should enable most fertile material of the two species to be distinguished without difficulty. However, on Wilson’s Promontory and sporadically elsewhere in Victoria, specimens which have some characteristics of R triangulare and some of R obtusangulum are occasionally encountered. As the relationship of these anomalous specimens appears to be with R. obtusangulum rather than with R. triangulare, they are discussed under the former species. 3. Platylobium obtusangulum Hook, in Curtis’s Bot. Mag. 60:t.3258 (1833); Benth., FI. Austr. 2:153 (1864); J. M. Black, FI. S. Austr. ed. 2:444 (1948); W. Curtis, Stud. FI, Tasm. 1:141 (1956); J. H. Willis, Handb. PL Viet. 2:277 (1973). Type: Curtis’s Bot. Mag. t.3258 (iconotype). R. triangulare sensu Sims, Bot. Mag. t.l508 (1812); sensu Hook.f., FI. Tasm. 1:96 (1856), non R.Br. R. macrocalyx in Lehm., PI. Preiss. 1:80(1844). Lectotype, here selected: Victoria, Port Phillip, 1842, C. LaTrobe (NEU 266741!). Isolectotype: NY!. R obtusangulum var. spinulosum J. H. Willis, Muelleria 1:126 (1967); Handb. FI. Viet. 2:277 (1973). Holotype: Victoria, Otway Range, on Airey’s Inlet to Wensleydale road, ± 3.2 km N. of Forestry Tower at Peter’s Hill, 12. xi. 1961, M. Allender (MEL 1522466!). Shrub with slender, weak, glabrous to densely villous, trailing, decumbent or erect stems up to 1 m long. Stipules ovate, up to 4 x 1.5 mm. Leaves opposite, unifoliolate or simple, broadly triangular, hastate or trilobate to ovate-cordate, pungent apically, the lateral angles pungent or rounded and obtuse, 1-3 x 0. 6-3(3. 8) cm, glabrous or minutely scabrid above except when young, lower surface glabrous to densely appressed- pubescent. Flowers 1-3 per axil, subsessile, the pedicels very short and completely concealed by the imbricate bracts and scales which are glabrous except for the margins or sparingly to densely pubescent throughout, the paired bracteoles below the calyx 7.5- 9 X 4-5.5 mm, overlapping the base of the calyx, glabrous except for the margins. Calyx densely clothed with long appressed hairs: 2 upper lobes 12-15 mm long (including the basal tube of 3.54 mm), 7-8 mm wide, the 3 lower lobes 8-9.5 mm long, all ± the same length, 3-3.5 mm wide. Corolla: standard oblate to almost reniform, 12-14 mm long including a claw 3-4 mm long, 16-20 mm wide, emarginate apically, pinkish-red or brownish outside, orange-yellow within with a basal yellow horse-shoe shaped throat surrounded by a reddish zone; wings 11-13 mm long including a claw 2-3 mm long, 3-4 mm wide, orange-yellow apically, pinkish-red basally, auricled; keel petals 10.5- 11 mm long including a claw 2. 5-3. 5 mm long, reddish apically, white basally, auricled. Stamens up to 11 mm long. Ovary 5-6 mm long, sessile, densely villous 135 throughout; style pubescent basally, glabrous above. Pods oblong, 1.4-2.5x1-1.45 cm, subsessile, with a thin conspicuous wing up to 5 mm wide beyond the upper sutural nerve, 1-5-seeded, pubescent throughout. Seeds ellipsoid, 3 x2.1 mm (Fig. 4). P. obtusangulum occurs in south-eastern South Australia, southern Victoria and Tasmania where it is found most frequently in open woodland and heathland (Fig. 5). Representative Specimens: South Australia — Kangaroo Island, near Strepera Falls, Middle River, 50 km W. of Kingscote, 14.X.1963, G. i30(AD 96349102). E. of Penola, 50 km N. of Mt. Gambier, 23.xi.1963, A Hunt 1691 (AD 96405081). Fleurieu Peninsula, Normanville sand dunes, 65 km SSW. of Adelaide, 11 ix 1969 D J E Whibley 2913 (AD 97025079). Victoria — Grampians, Mirranatwa Gap, on Dunkeld Rd. 32 km S. of Hall’s Gap, 27.ix.1959, T. B. Muir 874 (MEL 1522465). Narre Warren East, Wellington Rd., I km before turn-off to Cardinia Creek Reservoir, 19.xi.l977, J. H. Ross 2475 (MEL 531009). Wilson’s Promontory, Tongue Point, on track from Derby Saddle near its junction with track from Derby Beach, 2.xi.l980, M. G. Corrick 7071 (MEL 576039) Tasmania — George Town, 23. xi. 1842, R. C. Gunn (HO 11545). Near Mt. Direction, 23.xi.1842, R. C. Gunn (HO 11546). Near Launceston, x.l943, W. M. Curtis (HO 11543). Notes: Leaf shape in P. obtusangulum is extremely variable and quite unreliable as a means of differentiating the superficially similar P. triangulare. The differences between the two species are given in table 1. Willis, Muelleria 1:126 (1967), recognised var. spinulosum to accommodate a variant from the Otway Ranges, Victoria, with almost rotund leaves which bear 3-7 slender spine-like teeth on the margins. This variant has a very restricted distribution although a similar, but not identical, variant occurs near the Glenelg River. In view of the considerable range of morphological variation encountered within the species, I do not consider this local variant worthy of formal recognition. On Wilson’s Promontory and sporadically elsewhere in Victoria there occur occasional specimens with the pedicels exserted from the basal bracts and scales as in P. triangulare, but with large erect bracteoles, large lower calyx lobes and pubescent ovaries as in P. obtusangulum. The anomalous specimens, of which Corrick 7069 (MEL 576037) and 7072 (MEL 576036) from Wilson’s Promontory are examples, have the general facies of P. obtusangulum and differ from it most obviously in the exserted pedicels. Corrick 7069 and 7072 were growing in a population of P. obtusangulum and the relationship of these specimens appears to be with P. obtusangulum rather than with P triangulare. 4. Platylobium formosum Sm., Spec. Bot. New HoU. 1:17, t.6 (1793); Trans. Linn. Soc. Lond. 2:350(1794); Curtis, Bot. Mag. 14:t.469 (1800); Willd., Sp. PI. 3:921 (1802)- Sm Trans. Linn. Soc. Lond. 9:302 (1808); DC., Prodr. 2:116 (1825); Paxton, Bot. Mag! 13:195 (1846); Hook.f., FI. Tasm. 1:96 (1856); Benth., FI. Austr. 2:153 (1864)- F M Bailey, Queensland FI. 2:362 (1900); Curtis, Stud. FI. Tasm. 1:141 (1956); Burbidge & 136 Gray, FI. Austr. Cap. Territ. 213 (1970); J. H. Willis, Handb. PL Vic. 2:278 (1973); Beadle et al, FI. Sydney Region 300 (1973). R formosumwar, typicum Domin, Biblioth. Bot. 89:728 (1925); P. formosum subsp. formosum — A. Lee, Contrib. N.S.W. Nat. Herb. 4:97 (1970). Lectotype: here selected: New South Wales, 1793, J. White (LINN, sheet 1188.1); see under notes. P. parviflorum Sm., Spec. Bot. New Holl. 1:18 (1793); Willd., Sp. PL 3:921 (1802); Sm., Trans. Linn. Soc. Lond. 9:302 (1808); R.Br. in Sims, Bot. Mag. 37:t.l520 (1813); DC., Prodr. 2:116 (1825); Loddiges, Bot. Cab. 13:t.l241 (1827). P. formosum var. parviflorum {^m.) Benth., FI. Austr. 2:154 (1864); Domin, Biblioth. Bot. 89:728 (1925). P formosumswh^^. parviflorum {Sm.) A. Lee, Contrib, N.S.W. Nat. Herb, 4:96 (1970). Lectotype, here selected: New South Wales, 1793, J, (LINN, sheet 1188.3, no. 1); see under notes. P. ovatum sensu DC., Prodr. 2:116 (1825) non R ovatum Andr. (1802). P, formosum var. cordifolium Wawra, Itinera Principum S. Coburgi 1:11 (1883). Holotype: Victoria, Dandenong, Wawra coll. 1, no. 587 (W!). Shrub or subshrub up to 2.5 m high with prostrate, trailing, scrambling or erect stems, the stems glabrous or sparingly to densely clothed with appressed or spreading villous hairs or minutely scabrid from the persistent bases of the hairs. Stipules ovate or narrow-ovate, 3-5x0.8-2mm, reddish to dark brown, usually becoming reflexed. Leaves invariably opposite but occasionally some or all leaves borne alternately, sessile or on distinct villous petioles up to 2.4 cm long, extremely variable in size and shape, mostly broadly ovate or ovate-cordate to narrowly ovate -lanceolate, apex pungent or mucronate but without lateral angles, (l)2.5-5(7.2) x(0.7)l-2.5(4) cm, coriaceous, conspicuously reticulately veined, upper surface sparingly to densely pubescent when young but glabrous or minutely scabrid when mature, the lower glabrous or densely clothed with long villous hairs, margins flat or revolute. Flowers M per axil, on sparingly to densely villous pedicels 0. 6-3.4 cm long, the pedicels exserted from the basal bracts and scales which are usually glabrous except for the margins but are occasionally pubescent throughout, the paired bracteoles below the calyx usually narrow-ovate, 3. 5-5. 5(8. 5) x 1.5- 3 mm, glabrous except for the margins or the outer surface partly or entirely clothed with long villous hairs. Calyx densely clothed with appressed or slightly spreading hairs: 2 upper lobes 6-12 mm long (including the basal tube of 2.5-6 mm), 3-5.5 mm wide, the 3 lower lobes 3.5-6 x 0.7-1. 2 mm, the centre lobe often longest. Corolla: standard oblate, 9.6- 16 mm long including a claw 2-4 mm long, 12-21 mm wide, emarginate apically, yellow within with a basal yellow horse-shoe shaped throat surrounded by a reddish zone; wings 8-13 mm long including a claw up to 2.75 mm long, 3.5-5 mm wide, yellow throughout or red basally and yellow apically; keel petals 8-12 mm long including a claw up to 2.5 mm long, 3.5-6 mm wide, red apically. Stamens 7.5-10 mm long. Ovary 5. 5-8. 5 mm long including a stipe up to 1.75 mm long, densely pubescent throughout or pubescence confined to the sutures or to the sutures and portion of the valves or occasionally entirely glabrous. Pods oblong, 1.84.3(5.5) xO.9-2 cm, on a stipe 0.5-1 .6 cm long, with a thin conspicuous wing up to 4 mm wide beyond the upper sutural nerve, mostly 4-8-seeded, pubescent throughout or pubescence confined to the sutures or to the sutures and portion of the valves or entirely glabrous. Seeds ellipsoid, 2.6-4 x 1. 8-2.4 mm (Fig. 6). P. formosum is the most widespread species in the genus occurring from the vicinity of Wide Bay in Queensland southwards along the coast and in the Blue Mountains, Southern Tablelands and South Western Slopes of New South Wales, the A.C.T., Victoria as far west as the Grampians and Tasmania. A common understorey shrub, especially on moister sites (Fig. 7). Notes: J. E. Smith’s original description of P formosum (1793) drew attention to the diagnostic distinguishing characters of the species, namely, leaflet shape and the degree of pubescence of the ovary. Although described as “cordato-ovate”, the leaves illustrated in the plate accompanying Smith’s description are hardly cordate basally but this character may be seen in specimens 1188.1 and 1188.2 in Smith’s herbarium (LINN) collected by John White in New South Wales in 1793. 137 Fig. 6. Platylobium formosum. a — flowering twig of typical ''formosum'\ x 1; b — ovary showing pubescence on surface of valves and on the sutures, x2; c — flowering twig of typical ''parviflorum"\ x 1; d — ovary showing the pubescence confined to the sutures, x 2; e — flowering twig of the Victorian variant with alternate leaves, x 1. a & b from T. B. Muir 3653 (MEL.602925); c & d from N. Ford (NSW 4418); e from H. B. Williamson (MEL 584537). 138 J. E. Smith (l.c.) distinguished his P. parviflorum from P. formosum by its “lanceolato-ovate” leaves and glabrous ovaries but the two are conspecific and for a long time have been regarded as such. P parviflorum was accorded varietal rank by Bentham (1864) who lamented that although the variety was usually distinguished by “the narrow leaves, shorter pedicels, smaller flowers, more glabrous bracts, and by the ovary villous near the sutures only and not all over”, he found none of the characters constant. Subsequently var. parviflorum was raised to subspecific rank (Lee, 1970). P. formosum is a polymorphic species and is particularly variable in leaf shape and size, petiole and pedicel length, and the degree of pubescence, especially of the ovary and pod. In their typical forms, ^formosum"' and ^'parviflorum'' are distinctive largely on account of the obvious differences in leaf shape: the former with cordate-ovate leaves less than twice as long as broad and ovaries and pods pubescent on the surfaces of the valves and on the sutures, and the latter with ovate-lanceolate leaves usually more than twice as long as broad and ovaries and pods with pubescence on the sutures only or entirely glabrous. However, the typical forms are linked by numerous and varied intermediates, especially along the coastal belt from Queensland to Tasmania. It is unfortunate that the material available to Smith when he described P. formosum came from coastal New South Wales or was raised in England from seed collected in coastal New South Wales as this is an area in which intermediates abound. The leaves illustrated in the plate accompanying the desaiption of P formosum show a definite approach to some of the intermediates. In response to a request, M. D. Crisp examined the type material of P. formosum and P, parviflorum housed in LINN and reported that the ovaries in the type of P. parviflorum (sheet 1188.3 specimens no. 1) are glabrous except for a few hairs along the sutures, and the ovaries in the type material of P formosum (sheets 1188.1 and 1188.2) are sparsely villous on the surfaces of the valves over the seeds and on the sutures. The two immature pods in specimen 1188.1 are sparsely villous on the surfaces of the valves over the seeds and have a few basal hairs on the upper suture, a condition also found on the two smaller pods in sheet 1188.2. However, the largest pod on sheet 1188.2 is glabrous apart from some hairs on the sutures, a condition frequently met with in P. parviflorum. Of the two sheets, 1188.1 (LINN) is here chosen as the lectotype of P. formosum. As reliance is placed on both leaf shape and on the degree of pubescence of the ovary in differentiating "formosum" from "parviflorum", and, as these characters to some extent vary independently, the two groups are not homogenous within themselves and a number of specimens cannot be placed satisfactorily in either. Attempts to find additional characters to differentiate "formosum" from "parviflorum" have been unsuccessful. 139 The variation within P. formosum is to some extent regional. Over much of Victoria (E. Gippsland excepted), the interior of New South Wales and the A.C.T. the leaves are invariably distinctly ovate-cordate, less than twice as long as broad, and the ovaries are consistently densely pubescent throughout. These specimens with densely pubescent ovaries are perhaps as different from the lectotype of P. formosum as are the specimens with glabrous ovaries referred to ''parviflonim'\ Elsewhere, especially in the coastal belt from Queensland southwards to East Gippsland and in Tasmania, a different range of morphological variation is encountered and it is here that difficulties in naming specimens arise. Although typical ''parviflorum^^ occurs in the central coastal area of New South Wales, specimens with the facies of ''parviflorum'' but with the ovaries pubescent throughout or pods with hairs on the surfaces of the valves and on the sutures occur quite commonly along the coast and in Tasmania (for example, E. E Constable (NSW 30259) from Green Cape Lighthouse Rd., N.S.W.; A. H. S. Lucas (NSW 42888) from near Bicheno, Tasmania; F. H. Long 1203 (HO 11535) from near Orielton, Tasmania). The degree of pubescence of the ovary (pod) is a less reliable character than previous authors believed. Consequently, in some instances it is difficult to know when naming specimens whether to place most emphasis on leaf shape or on pubescence to avoid the unsatisfactory situation where specimens with a similar facies are referred to different infraspecific categories. An example of this difficulty is provided by a distinctive entity with narrowly ovate-lanceolate sessile leaves which occurs from the vicinity of Twofold Bay in New South Wales southwards to East Gippsland, Victoria. Among the specimens from East Gippsland are Muir 1934 (MEL 602809) from near Cann River, Buckland (MEL 584536) from Club Terrace, Beauglehole 62463 (MEL 1508185) from the vicinity of Gelliondale and Lumley Gl/3 (MEL 91913) from Mallacoota. All four specimens have a similar facies but the pods of Muir and Buckland have pubescence on the sutures and on the surfaces of the valves whereas the pubescence on the pods of Beauglehole and Lumley is confined to the sutures. On the basis of pod pubescence, Muir and Buckland are thus referable to "formosum'' and Beauglehole and Lumley to ""parviflorum" which is most unsatisfactory. It is only by ignoring the degree of pubescence of the pods in these specimens that a meaningful solution is arrived at. However, if the pubescence of the ovary (pod) is discounted always and reliance is placed solely on leaf shape to distinguish ""formosum" and ""parviflorum" the placement of many specimens becomes arbitrary and a different set of intermediate specimens is encountered. The inclusion within P, formosum of a slender prostrate or scrambling variant in which some or all of the leaves are borne alternately increases the range of morphological variation encountered within the species. The variant occurs sporadically in Victoria from Mt. Cole in the west to Newby in East Gippsland and typically the specimens have broadly ovate leaves with petioles 0.6-2. 8 cm long, flowers borne on pedicels which usually exceed 2 cm in length, and the young grov^h is often densely clothed with spreading rusty hairs. Examples include H. B. Williamson (MEL 584538) from Mt. Cole and P. R. H. St John (MEL 569732) from between Bayswater and Vermont. Superficially the specimens bear a fairly close resemblance to P alternifolium and in the past have been confused with and referred to this species. However, this variant is readily distinguished from P. alternifolium as the calyces and young pods are densely pubescent throughout and the peduncles are exserted from the basal bracts and scales. As the variation within P. formosum is to some extent regional, the usefulness of formally recognising infraspecific categories appears to vary over the distributional range of the species. The Tasmanian material is relatively uniform as far as leaf shape is concerned but the ovaries vary from glabrous to densely pubescent throughout and many of the specimens are difficult to refer to either ""formosum" or ""parviflorum" with certainty. In view of this, the recognition of infraspecific categories in Tasmania does not appear to be warranted or particularly useful. On the other hand, in the central coastal areas of New South Wales recognition of infraspecific categories is undeniably useful. I have been fortunate in having discussed this species with A. T. Lee, National Herbarium of New South Wales, on several occasions and have had access to the material in NSW which she has annotated. Although acknowledging the existence of differential 140 tendencies in P, formosum I prefer not to recognise formal infraspecific taxa. In reaching this decision I have been influenced more by the difficulties encountered in naming many specimens than in the benefits derived from according the extremes formal taxonomic status. The variation within P. formosum is imperfectly understood but I believe that it is more complex than is implied by the recognition of two infraspecific taxa. Representative Specimens of “typical formosum"': Queensland — Glass House Mountains, Mt. Tunbubudla, 10.viii.l930, C. E. Hubbard 3621 (BRI 270963). New South Wales — between Binda and Bigga, N.W. of Crookwell, 14.X.1953, C. W. E. Moore 2665 (CANB 28812, NSW 42927). A.C.T.: Two Sticks Rd., above Brindabella Valley, 8.xi.l961, N. T. Burbidge 7280 (CANB 126420, MEL 602825, NSW 76864). Victoria — Sassafras Gap, 40 km N. of Benambra on road to Corryong, 25. xi. 1954, H. /. Aston 1273 (MEL 602924). Tasmania — Port Sorell, X.I943, W. M. Curtis {WO 11538). Representative Specimens of “typical parviflorum": New South Wales — Pennant Hills, 19. ix. 1936, J. Vickery (NSW 42947). Epping, I4.ix.l947, N, Ford (NSW 4418). Lindfield Fire trail towards Lane Cove National Park, 28. ix. 1975, J. G, Sew/* 500 (NSW). Representative Specimens of “intermediates”: New South Wales — Green Cape Lighthouse Rd., 9.x. 1954, E. F. Constable (NSW 30259). Fiona Beach, 8 km S. of Forster, 10.x. 1961, E. E Constable 1289 (NSW 100827). Tasmania — near Mt. Direction, 19.xi.l842, R. C. Gunn 1016/1842 (NSW 42887). Black Charles Opening, near Orielton, 13. xi. 1933, F. H. Long 1203 (HO 11535). The following key is provided for those who wish to recognise the two subspecies as defined by Lee (1970): Leaves usually ovate-cordate, length up to twice the breadth; ovary pubescent on the surfaces of the valves and on the sutures; pods pubescent on the sutures and retaining some pubescence on the valves subsp. formosum Leaves usually narrow-ovate, not cordate basally, length more than twice the breadth; ovary glabrous throughout or pubescent on sutures only; pods glabrous or almost so or with pubescence confined to the sutures snhsp. parviflorum SPECIES INCERTAE Platylobium gracile Dum.-Cours., Le Botaniste Cultivateur ed. 2, 7:314 (June 1814). Dumont de Course! provided the following description: “Cette espece a un joli feuillage. Ses tiges et ses rameaux sont tres-menus. Ses feuilles rondes avec une pointe courte particuliere, sont parsemees en-dessus de polls rares, et portees sur de courts-petioles. Elies n’ont que 2 a 3 lignes di diametre. Les fleurs sont petites, jaunes, pedonculfes, solitaires, axillaires.” The description is inadequate to positively identify the plant and it is uncertain whether it is even a species of Platylobium. No specimen appears to have been preserved and consequently P. gracile is rejected as a name of uncertain application. Platylobium rotundifolium Colla, Hortus Ripulensis 1:110 (1824). The brief description given by Colla is as follows: “Sub hoc nomine missum ab H. sedy nullibi enumeratum inveni: parum differre videtur a P. formosoQN: sp. Ill 921). folia tamen sunt orbiculata nec cordataP The description is inadequate to positively identify the plant and I have not succeeded in tracing a specimen in BR or TO, the herbaria alleged (Stafleu & Cowan, 1976) to house Colla’s herbarium, on which the name was based. P rotundifolium is rejected, therefore, as a name of uncertain application. EXCLUDED SPECIES Platylobium lanceolatum Andr., Bot. Repos. 3:t.205 (1802) = Bossiaea heterophylla Vent., Descr. Plant. Nouv. 1:7, t.7 (1800). Platylobium microphyllum Sims, in Curtis’s, Bot. Mag. 22:t.863 (1805) = Bossiaea obcordata (Vent.) Druce, Rep. Bot. Soc. Exch. Club, suppl. 2, 1916:610 (1917). Platylobium obcordatum Vent., Jardin de la Malmaison l:t.31 (1804), non DC. (1825) = Bossiaea obcordata (Vent.) Druce, Rep. Bot. Soc. Exch. Club, suppl. 2, 1916:610 (1917). 141 Platylobium ovatum Andx., Bot. Repos. 4:1.266 (1802), non sensuDC. (1825) = Bossiaea heterophylla Vent., Descr. Plant. Nouv. 1:7, t.7 (1800). Platylobium reticulatum Sieb. ex Spreng., Syst. Veg. ed. 16, 3:231 (1826) = Mirbelia platyloboides (DC.) J. Thompson, Proc. Linn. Soc. N.S.W. 83:123 (1959). Platylobium scolopendrium Andr., Bot. Repos. 3:t.l91 (1801)=Bossiaea scolopendria (Andr.) Sm., Trans. Linn. Soc. Lond. 9:303 (1808). Platylobium spinosum Turcz., BuU. Soc. Nat. Mosc. 26:284 (1853) = Bossiaea spinosa (Turcz.) Domin, Vestn. Krai. Ceske Spolecn. Nauk., Tr. Mat.-Prir. 1919-22, 2:39 (1923). ACKNOWLEDGEMENTS I am most grateful to Mr M. I. H. Brooker, CSIRO Division of Forest Research, Canberra, for photographing several type specimens in BM, K and LINN while serving as Australian Botanical Liaison Officer at Kew Herbarium, Royal Botanic Gardens, England, ^d to his successor. Dr M. D. Crisp, National Botanic Gardens, Canberra, for providing details of the type material of P. formosum and P. parviflorum housed in LINN; to Mrs A. T. Lee, National Herbarium of New South Wales, for answering a nuiTiber of enquiries and for several valuable discussions; to Miss A. M. Podwyszynski, National Herbarium of Victoria, for preparing the illustrations that accompany the text; to the Directors/Curators of AD, BRI, CANB, HO, NEU, NSW, NY and W for the loan of specimens or for working facilities in their institutions; and to Mrs R. Parsons for typing the manuscript. REFERENCES Audas, J. W. (1921). Through the Balangum Ranges and at Rose’s Gap (Grampians). K/cr. Nor. 38:4-8- 11-16 Bentham, G. (1864). ‘Flora Australiensis’. Vol. 2 (Lovell Reeve & Co.: London). Ferguson L K. & Skvarla, J. J. (1981). The pollen morphology of the subfamily Papilionoideae (Legumi- nosae). In R. M. Polhill & P. H. Raven (eds) ‘Advances in Legume Systematics’. 2:859-896 (Royal Botanic Gardens: Kew). Lee, A. T. (1970). Taxonomic notes on Platylobium, Bossiaea and Templetonia in New South Wales Contrib N.S.W. Natl. Herb. ^\96-\05. Polhm, R M. (1976). Genisteae (Adans.) Benth. and related tribes (Leguminosae). In V. H. Heywood (ed.) Bot. Syst.’ 1:143-368. (Academic Press: London). Polhill, R. M. (1981). Tribe 26. Bossiaeeae (Benth.) Hutch. In R. M. Polhill & P. H. Raven (eds) ‘Advances in Legume Systematics’. 1:393-395. (Royal Botanic Gardens: Kew). Stafleu, F. A. & Cowan, R. S. (1976). ‘Taxonomic Literature’. Vol. 1. (Bohn, Scheltema & Holkema* Utrecht). Manuscript received 21 April 1982. J A REVISION OF ANGIANTHUS WENDL., SENSU LATO (COMPOSITAE: INULEAE: GNAPHALIINAE), 1 by P. S. Short* ABSTRACT Angianthus Wendl. (sensu Bentham, 1867) has been revised. Eight genera, Cephalosorus A. Gray (monotypic), Chrysocoryne Endi. (6 species), Dithyrostegia A. Gray (monotypic), Epitriche Turcz. {monoiypic), Hyalochlamys K. Gray (monotypic), Muell. (monotypic), Pogo/io/ep/^Steetz (6 or 7 species) and Siloxerus Labill. (3 species) are reinstated. Three species with uncertain affinities, A. axilliflorus W. V. Fitzg. ex Ewart & J. White, A. burkittii (Benth.) J. M. Black and A. connatus W. V. Fitzg. are excluded from the above genera. Fifteen species are recognised within Angianthus s. str. Four of these, A. cornutus, A. cyathifer, A. glabratus and A. prostratus are described as new. A. brachypappus var. conocephalus J. M. Black is raised to specific rank. Three of the 6 species of Chrysocoryne, C. multiflora, C. tridens and C trifida are also described as new. The following new combinations are made: Cephalosorus carpesioides (Turcz.) Short, Epitriche demissus (A. Gray) Short and Pogonolepis lanigera (Ewart & J. White) Short. The majority of genera dealt with have no obvious close affinities with one another but all are referred to the subtribe Gnaphaliinae. CONTENTS Generic history 143 Relationships of A ngianthus and segregate genera 144 Materials and methods 146 Morphology 146 Presentation of data 152 Key to segregate genera and species of sensu lato 153 Genera A ngianthus sQnsustncto. . . . . ... 153 Dithrostegia 201 Pleuropappus . . . . 179 Hyalochlamys 202 Epitriche . ... 181 Pogonolepis 203 Cephalosorus , . . . 182 Siloxerus 204 Chrysocoryne . ... 185 Species of uncertain affinity 209 Names of uncertain application 210 Acknowledgements References 211 Index to genera and species of Compositae 212 GENERIC HISTORY Wendland (71808 or 1809, see Stafleu, 1967), when describing the new genus Angianthus recognised a single species, A. tomentosus. No further species were attributed to the genus until Mueller described the closely related A. brachypappus in 1855. Then in 1867 Bentham, as well as describing several new species of Angianthus, placed the following genera in synonymy (arranged chronologically in taxonomic and nomenclatural groups): Siloxerus Labill. (1806), including the emendations Styloncerus Spreng. (1826) &Ogcerostylus Cass. (1827); CylindrosorusBenth. (1837); Phyllocalymma Benth. (1837); Skirrhophorus DC (1838); Eriocladium Lindl. (1839); Chrysocoryne Endl. (1843); Pogonolepis Steetz (1845); Piptostemma Turcz. (March 1851), Cephalosorus A. Gray (April \85\)-, Hyalochlamys A. Gray (April \8S\)-, Dithyrostegia A. Gray (April 1851), Gamozygis Turcz. (Oct. 1851); Epitriche Turcz. (Oct. 1851) and Pleuropappus F. Muell. (1855). In justifying this procedure Bentham (l.c., p. 561) stated that the genera had ‘ ‘been established chiefly upon minute distinctions in the pappus which appear to me to afford a much better specific than generic character”. ♦National Herbarium of Victoria, Birdwood Avenue, South Yarra, Victoria 3141. Muelleria 5(2): 143-183 (1983). 143 144 Bentham realised that the same Siloxervs had priority over Angianthusbut because the former name was at complete variance with the etymology given by the author and both Cassini’s (1827) and Sprengel’s (1826) emendations of Ogcerostylus and Styloncerus respectively were more recent names than Angianthus, he chose the latter name. Ostenfeld (1921) and Ising (1922) made a number of new combinations with Siloxerus but in 1940 the International Botanical Committee on Nomenclature approved the conservation of Angianthus against Siloxerus. Prior to Bentham (l.c.) the most important work on the group was by Gray (1851). As well as describing a number of new taxa based on collections made by James Drummond in Western Australia Gray produced a conspectus of the Division Angiantheae (== subtribe Gnaphaliinae p.p., see below). He recognised 22 genera within the group. Seven of these, namely Cephalosorus, Chrysocoryne, Dithyrostegia, Hyalochlamys, Phyllocalymma, Siloxerus and Skirrhophorus were ultimately referred to Angianthus by Bentham (l.c.). Despite Bentham’s above-mentioned statement it is clearly evident that Gray did not place undue emphasis on pappus type but also differ- entiated the genera on characters such as size shape and number of capitular bracts and general involucral bracts, the shape of the compound heads or glomerules and the number of florets per capitulum. Such characters have also been used to a large extent in the current work and indeed in this revision only Phyllocalymma and Skirrhophorus sect. Skirrhophorus (sensu Gray) are retained in Angianthus s.str. Although in some cases modified, Chrysocoryne, Cephalosorus, Dithyrostegia, Hyalochlamys^nd Siloxerus reinstated. The genus PogonolepisSiQtiz, treated by Gray as a section of Skirrhophorus, is also reinstated whereas Skirrhophorus demissus A. Gray, referred to section Psuedopappus A. Gray, is referred to the genus Epitriche Turcz. It is clear from Turczaninow’s (1851) work within the group that he too had similar generic concepts to Gray. In fact several genera described by Turczaninow and Gray are nomenclatural synonyms. This is a result of both workers basing many of their descriptions on duplicate collections made by Drummond. Several workers (e.g. Mueller, 1889; Ostenfeld, 1921) were critical of Bentham’s treatment of Angianthus but none the less his concepts have been generally followed in current Australian literature. Ewart & White (1909), attributed two further taxa, A. lanigerus and A. axilliflorus to the genus. Neither species remotely resemble Angianthus s.str. as circumscribed in the current work. A. lanigerus is a species of Pogonolepis while A. axilliflorus, like Fitzgerald’s (1905) species A. connatus, lacks affinity with any of the segregate genera of Angianthus s.L On the other hand the only other species described since Bentham and up to the present time, A. acrohyalinus Morrison (1912), undoubtedly belongs to the genus. A single species, Gnephosis burkittii Benth. was transferred io Angianthus by Black (1929). Only Gardner (1931), in a census of Western Australian plants, attempted to divide the genus. He recognised three sections. Two, namely section Spicatae Gardn. and section Pyramidatae Gardn., are invalid names while the remaining one, section Skirrhophorus (DC.) Gardn., is misapplied. Gardner merely listed the various species under each section, placing A. cunninghamii, the type species of Skirrhophorus DC., under section Pyramidatae Gardn. In this revision the following genera, commonly placed in synonymy with Angianthus, have been reinstated: Cephalosorus, Chrysocoryne, Dithyrostegia, Epitriche, Hyalochlamy, Pleuropappus, Pogonolepis and Siloxerus. Three species, namely^, axilliflorus, A. burkittii and A. connatus avQ excluded from Angianthus s.str. as their relationships are obscure. RELATIONSHIPS OF ANGIANTHUS AND SEGREGATE GENERA In his description of the subtribe Angianthinae Bentham (1867, p. 556) wrote: “Flower-heads small, sessile or nearly so on a common receptacle in dense clusters or compound heads, often closely surrounded by imbricate bracts or by a few floral leaves forming a general involucre. Florets all tubular and hermaphrodite.” Within Australia he 145 recognised eight genera within the subtribe, namely Angianthus Wendl., Calocephalus A. Gray, Chthonocephalus Steetz, Craspedia Forst.f. (also occurs in New Zealand), Gnaphalodes A. Gray ( =Actinobole Fenzl ex Endl., see Eichler 1963), Gnephosis Cass, and Myriocephalus Benth.. Subsequently Bentham (1873a) also referred the Indian Caesulia Roxb. and the African Eriosphaera Less, to the Angianthinae. A similar procedure was also followed by Hoffman (1894) although he excluded Eriosphaera and included monotypic Dimeresia A. Gray (North America) and Decazesia F. Muell. (Australia). Black (1929) also included Eriochlamys Sond. & F. Muell. ex Sond. and Basedowia Pritzel in the subtribe. Prior to Bentham (1867, 1873) the majority of the genera listed above were similarly grouped together on the basis of their compound head and usual general involucre in the “Division Angiantheae” of the “Subtribe Gnaphalieae” (De Candolle, 1836, 1838; Gray, 1851). Recently Merxmuller, Leins and Roessler (1977), making much use of new data derived from investigations of pollen (Besold, 1971; Leins 1971, 1973), the stylar structure of hermaphroditic florets and chromosome numbers, have suggested that only 3 subtribes, i.e. Inulinae, Gnaphaliinae and Athrixiinae be recognised within the tribe. This contrasts sharply with the 9 subtribes recognised by Bentham (1873a) and commonly used in current works. Merxmuller etal.{\.c.) have referred the above-mentioned Australasian genera of the Angianthinae to an '^Angianthus group’’ within the Gnaphaliinae. They have also tentatively included the Australian genera Stuartina Sond. and Nablonium Cass, in the group. Although the alignment of the various genera with others such as Helichrysum Miller and Helipterum DC. seems reasonable the "Angianthus group” is nevertheless artificial. Merxmuller et a/.(l.c.), like past workers, have, I believe, placed too much emphasis on the presence of pseudocephalia or compound heads. Thus they stated that “their [the pseudocephalia] distribution does not look irregular [within the Inuleae], as most of the taxa concerned can be arranged in well delimited taxonomic groups. This has been accepted for a long time in the Angianthus group . . .” (p. 589). Studies of the variation exhibited in the general receptacle, achene and bract morphology of Angianthus s.l. alone make it very clear that many of the segregate genera, e.g. Pogonolepis and Siloxerus, are but distantly related to Angianthus s.str. A similar situation appears to occur in the apparent unnatural genera Calocephalus^ Chthonocephalus, Gnephosis dSid Myriocephalus, (in the sense of Bentham, 1867). For example Blennospora A. Gray, commonly placed in Calocephalus but clearly with little affinity with that genus, should be reinstated (Short, 1981a, b). The artificiality and indeed impracticality of the use of the compound head as a criterion for subtribal recognition may also be seen in the monotypic genera Eriochlamys and Cephalipterum, Eriochlamys behrii Sond. & F. Muell. ex Sond., like the majority of the "Angianthus group” has tubular, hermaphroditic florets but the capitula may be single or aggregated into compound heads. Bentham (1867, 1873a) placed the genus in subtribe Gnaphaliinae but Black (1929) referred it to the Angianthinae. The complex species (Short, 1981c) Cephalipterum drummondii A. Gray contains capitula with tubular female and male florets. It has obvious affinities with species of Helipterum, particularly H. sterilescens F. Muell., but on the basis of the presence of compound heads is placed in the Angianthinae or "Angianthus group” thus suggesting only a distant relationship with members of Helipterum. The occurrence of compound heads in so many species is perhaps the result of selection for more efficient pollination or for a shorter life cycle. A reduction in life span is perhaps the most likely factor as perhaps all but four of the c. 80 species in the "Angianthus group” (excluding Craspedia) are annuals living in semi-arid to arid conditions. Such a hypothesis not only helps to explain the observed aggregation of capitula in the "Angianthus group” but may also help to explain the presence of varying degrees of aggregation of capitula in the inflorescences of many closely related species of the ill-defined genera Gnaphalium L., Helichrysum and Helipterum. Hence, although agreeing with Merxmuller et al. (1977) that the subtribe Angianthinae be abolished and that Angianthus and other genera included in the 146 Angianthinae by Bentham (1867) be placed within the Gnaphaliinae, I cannot, at least under current definition, agree with the maintenance of an "'Angianthus group’'. Further revisions within this group are required before the relationships of the many genera can be determined. MATERIALS AND METHODS Descriptions of taxa were made from dried herbarium material and to a lesser extent from specimens stored in 70% alcohol. In many cases measurements were made from ail available specimens, these usually being fewer than 30 per species. Where more specimens existed collections were selected on a regional basis, care being taken to include the full range of morphological variation. Shapes were defined using the terms given by the Systematics Association Committee for Descriptive Terminology (1962). Specimens were obtained from the following herbaria: AD, BRI, CANB, CBG, GH, K, KP (Kings Park, Western Australia), KW, MEL, NSW, NT, PERTH, and UWA (abbreviations after Holmgren & Keuken, 1974). Much material was also obtained by the author on field trips to Western Australia in August-September 1977 and November 1979 and on a number of trips in South Australia and Victoria during 1977-1980. A complete set of collections from these trips is housed in AD. When possible, publication dates of all works have been checked with Stafleu (1967) and Stafleu & Cowan (1976, 1979). Publication dates for the Bull. Soc. Imp. Naturalistes Moscou were provided by Dr Hj. Eichler (pers. comm., 1978). Turczaninow (1851) described new species in a paper published in volume 24 of that journal. The volume consists of 2 parts and 4 numbers. Part 1, number 1 contains the first part of the above paper and according to the published records (Bull. Soc. Imp. Naturalistes Moscou 24(1): 705) was presented to the meeting of the Society on 15 March 1851. This is accepted as the date of publication. Part 2, number 3 contains the second part of the paper and was presented to the Society on 25 October 1851 (Bull. Soc. Imp. Naturalistes Moscou 24(2):625). As in the above case this is also accepted as the date of publication. Accurate publication dates had to be obtained as both Turczaninow and Asa Gray (1851) published new species, based on duplicates of the same collections, in the same year. Gray published in three fascicles of Hook. J. Bot. Kew Gard. Misc. The fascicles were published in April, May and June of 1851 (Stafleu & Cowan, 1976). MORPHOLOGY As pointed out by Heywood (1971) and Heywood et al. (1977) the uniformity of inflorescence, floral and fruit structure within natural families such as the Umbelliferae, Quciferae and Compositae tend to make the recognition of genera difficult. In some cases apparently trivial characteristics have been used to define generic limits. For example various studies (e.g. De Jong, 1965; King & Robinson, 1966) within the Compositae have shown that the pappus type has frequently been given too much weight in the recognition of genera. It is essentisJ to Took at a wide range of characters before deciding generic limits. In the current work the segregate genera and species of Angianthus s.l. have been recognised on the basis of a combination of a number of morphological characteristics, some of which are unique. For example the oblique scale- like pappus of Pleuropappus is, within Angianthus s.l., unique to that genus. Furthermore the oblique achene and the number and arrangement of capitula- subtending bracts are also unique to Pleuropappus (fig. Ih). In general a single, unique, generically diagnostic character, or a unique combination of characteristics, is used in the key to the segregate genera and species of Angianthus s.l. The characteristics of each genus are also briefly outlined following each generic description. The use of various characters in Composite taxonomy has been discussed by a large number of workers. Perhaps the most useful and best known study was made by Bentham (1873b). The various features used in delimiting genera and species of Angianthus s.l. are discussed below. 147 Fig. L Characteristics of Angianthus and segregate genera, a — capitulum-subtending bracts of Chrysocoryne pusilla [Short 902). b — capitulum of C. pusilla [Short 902). c — capitulum and cC bract (inner view) of C. drummondii [Short 1085). d-e - capitular bracts of c. 902). f — bracts subtending compound head of Hyalochlamys globifera [Short 635) g -habit of Dithyrostegiaa^ x 1 [Short 344). h - floret of Pleuropappus phyllocalymme^ [Wilhelm s.n MEL 541617). i — capitular bract of Pogonolepis stricta [Short 1&3). j — general [Short 1075). k — arrangement of capitular bracts mA. glabratus bract 905). m- capitulum-subtending 148 a. Habit: With the exception of Angianthus cunninghamii, a perennial shrub, all species examined in the current work are annuals. Habit characteristics are useful for distinguishing a number of closely related species. For example within Chrysocoryne one species, C. tridens, never branches from the upper nodes of major axes but other species commonly do so. b. Leaves: The majority of taxa have alternate leaves but there are exceptions. Within Angianthus some species possess a few opposite basal leaves whereas some specimens of A, preissianus have entirely opposite leaves. The distinctive A. connatus and the monotypic Cephalosorus are characterised by possessing some opposite leaves. The latter taxon is also the only genus with petiolate leaves. Leaf shape is more or less constant within any one genus although specific differences exist, e.g. the conduplicate leaves and hyaline apices of Angianthus acrohyalinus are unique within the genus. At the generic level the monotypic Dithyrostegia and Cephalosorus have unique leaf types. The nature of the leaf indumentum is of value at both generic and specific levels. For example in all species of Chrysocoryne and in monotypic Hyalochlamys at least some scale-like glandular hairs are found on the leaves. Such trichomes are absent from other genera. The glabrous or nearly glabrous nature of the leaves of some species of Angianthus is also useful for distinguishing some species. c. Compound Heads: All species included by Bentham (1867) in Angianthus are said to possess compound heads or glomerules, i.e. the individual capitula are clustered together to form a compact inflorescence usually resembling either a capitulum or a spike. Such inflorescences often possess a general involucre of bracts. Recent workers such as Leppick (1977) have referred to such structures as synflorescences or conflorescences but the term compound head, used in current Australian floras and readily understood, is retained in this work. The ancestal nature of the inflorescence or compound head in all genera examined is unknown and probably can only be ascertained following anatomical studies at different stages of development. At least in Angianthus s. sir. the compound heads appear to have been derived from a compound inflorescence, perhaps a compound corymb. This is supported by the fact that in most species of Angianthus W\i\i oblong to cylindrical heads distinct minor receptacular axes (fig. Ij) are more or less evenly distributed along a single major axis. Two or more capitula are usually attached to each minor receptacular axis and, although difficult to ascertain, it also appears that the outermost capitula of each minor axis commonly possess more than one capitulum-subtending bract. The arrangement of capitula in the compound head and the nature of the general receptacle furnish very useful, often diagnostic, generic characteristics. For example the general receptacles of Siloxerus and Dithyrostegia are covered with long silky hairs whereas the receptacle in other genera is glabrous. In Chrysocoryne and Pleuropappus the capitula are arranged in a regular spike-like fashion along the axis but most species of Angianthus have distinct minor receptacular appendages distributed along the major axis. The size and shape of the compound heads provide useful characters for distinguishing many species of Angianthus, Chrysocoryne and Siloxerus. d. Bracts: The bracts of the general involucre, the capitulum-subtending bracts, capitular bracts and the paleae provide some of the best diagnostic characters for distinguishing the genera and to a lesser extent species. d.l. General Involucre (bracts subtending compound heads). The number and shape of the bracts of the general involucre are of much use in distinguishing the genera Cephalosorus, Dithyrostegia, Hyalochlamys and Pogonolepis. For example in Dithyrostegia the compound head is enclosed by two overlapping or connate leaf-like bracts (fig. Ig), a condition not found in other genera. In contrast all species of 149 Chrysocoryne lack a conspicuous involucre. In Angianthus the bracts provide useful characters for distinguishing species. d.2. Capitulum-subtending Bracts. The term capitulum-subtending bract appears to have been applied by Bentham (1867) to a usually singular bract occurring on the adaxial side (relative to the major axis of the general receptacle) of a capitulum of many species of Angianthus sA. It appears that such bracts, under Bentham ’s definition, could resemble the capitular bracts. For example he recognised such bracts as occurring in both A. humifusus ( = Siloxerus humifusus) and Angianthus strictus ( =Pogonolepis stricta). However the present studies suggest that no such bracts should be recognised. In Siloxerus the capitula themselves are difficult to recognise and both receptacular and capitular bracts are similar in appearance. In Pogonolepis all bracts within the centre of a compound head have similar morphologies and it would appear that the inner papillose bracts of the general involucre have been interpreted by Bentham as capitulum- subtending bracts. The capitulum-subtending bracts are therefore best defined as a bract or several bracts which occur on one side of a capitulum (usually very obviously the adaxial surface relative to the centre of the general receptacle) only and differ in their appearance from at least some of the capitular bracts. Thus in Chrysocoryne (figs, la-e; lOc-f, j) the capitulum-subtending bracts are extremely dissimilar to all capitular bracts. In contast the more or less flat capitulum-subtending bracts in both Angianthus and Pleuropappus are similar to the inner flat capitular bracts but are quite distinct from the outer pair of concave capitular bracts of these genera (fig. 11-m). The capitulum-subtending bracts are undoubtedly modified bracts which originally subtended the capitula in ancestral plants with only loosely arranged inflorescences of capitula. Their role in the compound head is presumably one of protection of developing florets and to some extent the attraction of pollinators. The function of protecting developing florets and fruit is perhaps best illustrated in Chrysocoryne where the bracts more or less enclose the capitula. d.3 Capitular Bracts. The capitular bracts, i.e. the involucral bracts of a capitulum, furnish many characters which are important in distinguishing genera. For example the bracts of Pogonolepis are unique in that they possess papillae at their apex (fig. li). Similarly the rigid, opaque, thick-cell walled bracts of Siloxerus are unique to that genus. In other cases the number and arrangement of the bracts is important in distinguishing genera. With the exception of Angianthus microcephalus, all species of Angianthus have four capitular bracts arranged so that two outer, variably concave bracts surrounded two inner flat bracts (fig. Ik). The same arrangement occurs in monotypic Pleuropappus. Although the majority of segregate genera have constant and unique features in relation to the capitular bracts the genus Chrysocoryne is distinctive in the degree of variation exhibited in the number, arrangement and morphology of the bracts. Three of the six species have similar bracts, i.e. C. drummondii (figs. Ic; 9; lOj), C. tridens and C. uniflora each have two, concave, capitular bracts with shortly ciliate margins. In the remaining species the bracts are quite different. C. trifida also has only two capitular bracts but they have long hairs on the upper margins (figs. 9; 101) and are smaller than those found in the three preceding species. Both C. multiflora and C pusillahave from 2 to c. 10 bracts per capitulum. The laminae on the bracts of C. pusilla (figs, le; 9; lOf) distinguish that species from all others in the genus. Within Angianthus the 4 capitular bracts, particularly the inner flat bracts, frequently exhibit specific differences. For example in several species a wing-like extension extends from the adaxial surface of the midrib of the inner bracts (fig. 3e) while in others the inner bracts are abruptly attenuated in the lower Vi to V 3 of the bract (fig. 3d). It seems reasonable to suggest that the major role of the capitular bracts in all species examined is the protection of the developing florets and achenes. Presumably the colour and number of these bracts, plus the capitulum-subtending bracts, are also important in making the capitula more attractive to likely pollinators. However \n Angianthus s.l. the general lack of large, coloured laminae, which are found in many species of genera such as Helipterum, Helichrysum and Calocephalus, suggests that in many species this latter 150 role is only minor. It seems likely that the role of attracting pollinators is more a function of the compound head. Given that protection is probably the major function of the capitular bracts it is not surprising that in many genera the number, morphology and arrangement of bracts is more or less constant. That is, a successful combination of bract characters which ensures floret and achene development has a high selective value and only comparatively minor modifications (such as the degree of development of the wing- like extension in Angianthus) have occurred. The variation found in the capitular bracts of Chrysocoryne obviously does not support this argument, at least at first sight. However, this variation is readily explained. Firstly, there is a marked trend in the genus to a reduction in the number of florets per capitulum (see proposed phylogeny in generic treatment). This decreases the need for a large number of capitular bracts and in fact a decrease in bract number is correlated with a decrease in the number of florets per capitulum. Secondly, and perhaps more importantly, the capitulum-subtending bracts are extremely well developed in this genus (figs, la, c; 9; 10c; h, j, m). In all species they more or less completely cover the capitula and obviously give great protection to the developing florets, thus decreasing the importance of the capitular bracts in this function. d. 4. Paleae. Within Siloxerus the florets in all species are subtended by a single bract which resembles the capitular bracts. Paleae are not found in other genera. e. Florets: The florets of all species are more or less morphologically uniform, i.e. they are all tubular, bisexual and yellow. The most important difference occurs in the number of corolla lobes per floret. Gardner (1977) noted that approximately 80 genera of Compositae include species with tri- and/or tetramerous disc florets. He suggested that a change from a 5 to 3 or 4-lobed state is either related to a change in breeding system, i.e. from chasmogamy to autogamy, or to selective pressure for increased seed production. As previously outlined (Short 1981a, b) in all members of the Australian Gnaphaliinae so far examined species with predominantly or entirely tri- and/or tetramerous florets are inbreeders. Closely related species with pentamerous florets have been classed as outbreeders. Indeed lobe- number is an extremely important character for distinguishing closely related species in Chrysocoryne and Angianthus, The length of the corolla tube, i.e. the length from the base of the floret to the base of the lobes, the diameter of the floret tube and the presence or absence of a basal tubular swelling have been recorded for all species but appear to be of little or no taxonomic significance. In many cases the degree of swelling is related to the age of the floret, the swelling becoming pronounced only in mature florets. f. Stamens: The anthers of all members of the Inuleae are said to be tailed and this appears to be the case in all members of Angianthus s.l. Various workers (e.g. Robinson &^g, 1977) have found that characteristics of the filament collar and the anther appendage provide taxonomicaily useful characteristics. Such differences have not been closely examined in the current work but the only apparent differences in anther morphology are found between related outbreeding and inbreeding species. Outbreeding species have normal tetrasporangiate anthers whereas related inbreeding species have bisporangiate anthers. Furthermore the anther sacs of inbreeders are about one-half the length of those in outbreeders. g. Pollen: It is clear from the works of Besold (1971), Leins (1971, 1973), Merxmuller et aL (1977) and Skvarla et al. (1977) that detailed studies of pollen structure will be of considerable use in clarifying generic relationships within the Inuleae. In the current work pollen grains of Angianthus s.l. have not been anatomically examined. However it has been found that the number of pollen grains per anther or floret is a very useful character for distinguishing species. As previously pointed out (Short 1981a, b) the pollen-ovule ratio (P/O), which in the hermaphroditic species examined is equal to the number of 151 pollen grains per floret, reflect a plant’s breeding system. In the genera examined there are a number of closely related outbreeding and inbreeding taxa with vastly different P/ O values. For example the average P/O of Chrysocoryne tridens is 48.6 but that of the closely related C. uniflora is 1,775.5. Indeed pollen grain number, along with the number of lobes per floret and the size and morphology of the anthers, provide the best, if not the only characteristics for reliably distinguishing some closely related taxa. h. Pappus: Pappus characteristics are of much use at both the generic and specific level. For example the oblique scale-like pappus of Pleuropappus readily distinguishes it from all other genera included in Angianthus by Bentham (1867). In Siloxerus the three species have a pappus of jagged scales joined at the base or a jagged ring with more or less indistinct scales. In contrast an array of pappus types occurs in Angianthus s,str. Here some species such as A. milneiy A. prostratus and A, pygmaeus completely lack a pappus but others such as the apparently closely related taxa A. tomentosus, A. cyathifer and A. acrohyalinus havQ very distinctive types. The pappus is generally regarded as a modified calyx commonly made up of scales or variably plumose bristles which occur at the apex of the achene. Normally, it is not difficult to discern the pappus but difficulties arise in both Epitriche demissus and Angianthus eriocephalus. In the former case it is difficult to tell from the few specimens available whether in fact a pappus should be recognised. Long hairs appear to occur over much of the apex of the fruit and possibly the lower portion of the floret. The apparent irregular, scattered distribution of the hairs suggests that they are not to be regarded as a pappus but merely as part of the indumentum of the achene. A similar case occurs in A. eriocephalus {=A. preissianus). Bentham (1867) recorded that a minute pappus occurs in this taxon. However the ‘pappus’ is nothing more than a cluster of papillae at the apex of the achene. The small papillae are found scattered over much of the fruit. i. Achenes: The majority of Australian works refer to the Composite fruit as an achene. However in recent literature, for example in the various papers included in Heywood, Harborne and Turner (1977), some workers refer to the fruit as a cypsela. Originally the term achene was defined by Necker (1790) to include all dry, indehiscent, one-seeded fruits, irrespective of their derivation from an inferior or superior ovary. De Candolle (1813) redefined the term to include only those fruits derived from an inferior ovary. Subsequently Mirbel (1815) introduced the term cypsela for the same type of fruit. That is an achene is, in the sense of De Candolle’s definition, equivalent to a cypsela. Wagenitz (1976) has proposed that De Candolle’s definition, which is generally accepted in German texts, be commonly adopted. It is accepted in the present work. Features of achene morphology promise to be of great value in defining generic limits within the Australian Gnaphaliinae. Detailed anatomical studies have not been made by the present author but examination of achenes at even low magnification (10"40x) clearly shows that at least some genera possess distinctive, often unique fruit. Thus both Chrysocoryne and Siloxerus are readily distinguished from other segregate genera of Angianthus s.l. by the small, more or less obovoid, variably papillate, pink or purple achenes which lack obvious epidermal mucilaginous cells. In contrast others such as Pogonol^is and Cephalosorus are surrounded by mucilaginous cells. The achene of monotypic Dithyrostegia is readily distinguished by its clothing of long papillae or hairs. The monotypic genera Hyalochlamys and Pleuropappus also have distinctive, unique achenes. In the former genus the carpopodium (the point of attachment to the receptacle), is the most distinctive feature, being represented by a distinct white band of cells. In Pleuropappus Ihe fruit is obliquely attached to the corolla tube (fig. Ih). Features such as size and shape, colour, presence or absence of mucilaginous cells and shape of the carpopodium are thus useful in distinguishing genera. However detailed anatomical studies of the achenes of Angianthus s.l. are highly desirable. Heywood and Humphries (1977), for example, have found in the Anthemideae that characters such as the presence or absence of epicarpic glands and secretory canals, the number of layers and the degree of thickening of the epicarp and mesocarp and the orientation of the 152 cotyledons are all characteristics which show systematically useful variation. Although anatomical studies in Anginathus s.l. may only support the current classification based on more obvious features such studies may be useful in determining relationships of the various genera throughout the Gnaphaliinae. Detailed anatomical studies of achenes will be essential before the many genera of Australian Gnaphaliinae are satisfactorily delimited and their relationships understood. The function of the mucilaginous cells in the fruit of many species of Angianthus s.l. is open to question. Some authors (e.g. Schodde, 1963; Roth 1977) have suggested that the sticky nature of many wet fruits is important in fastening the achenes to the soil particles during germination. This may be so for some species but as Roth (1977) has suggested slime cells may also constitute a water storage tissue. This could be very important in germination, particularly in view of the semi-arid to arid environment inhabited by many of the species of Angianthus s.l. Another possibility is that the cells and their degree of water uptake and/or period of hydration are associated with some mechanism in breaking dormancy. That is, the seeds will probably only germinate when sufficient external moisture is available for growth. In such a situation the water retained by the slime cells is of little importance. PRESENTATION OF DATA When describing new species a number of workers, particularly F. Mueller, cited locality details which differed to some extent from those occurring on labels accompanying specimens which are (as indicated by notes etc.) obviously type material. When citing type specimens in the current work both the published data and data on original collection labels are usually cited. Usually type material of most taxa was available and generally when necessary it has been possible to select lectotypes. However this is not the case for material used by Turczaninow. Holmgren and Keuken (1974) list KW, CW and LE as having important Turczaninow collections. A request for type material of seven species from CW has not been acknowledged and none exists in LE. Only four of the required type specimens exist in KW. Although duplicates of all collections cited by Turczaninow exist outside Russia there is no reason to believe that any of these were examined by him. Thus types housed in KW are tentatively referred to as holotypes. Steetz (1845) described a number of taxa, including Phyllocalymma filaginoides, Skirrhophorus preissianus, Styloncerus cylindraceus and Styloncerus suberectus, in Lehmann’s “Plantae Preissianae”. Type specimens of these taxa occur in GH, LD, MEL and S. The only type collection in GH is Preiss 37, the type of P. filaginoides. It is in F. W. Klatt’s herbarium. Types of all taxa are in MEL, having come from two main sources (possibly the only sources but this in unclear), namely from O. W. Sonder’s herbarium purchased toward the end of the nineteenth century (Court, 1972) and from Steetz’s herbarium purchased in about 1859 (Mueller, 1888). Specimens in LD and S presumably come from Lehmann’s herbarium. The major part of Lehmann’s collection was acquired by Andersson for Stockholm in 1860-61 (Nordenstam, 1980). Types of S. preissianus, P. filaginoides and S, cylindraceus, are in S and all have collection labels in Preiss’s hand and the name in Steetz’s hand. Type collections of all four taxa are housed in LD and are similarly labelled. When offering Lehmann’s herbarium for sale Klatt stated in an advertisement in Bonplandia{%{%)\\A3, 15 April 1860) that “the Preiss herbarium contains the originals of the Plantae Preissianae with the labels of the botanists who worked on this flora, as well as the ‘unica’ of the Preiss collections” (English translation by Mrs D. M. Sinkora). Nordenstam (1980) similarly referred to “the original specimens of Plantae Preissianae"' (p.282) as being sold to J. Agardh in Lund. Such statements suggest that the principal types of all species are to be found in LD. However it is clear that this is not the case for the above-mentioned taxa described by Steetz. The collections from his herbarium now at MEL undoubtedly should be selected as the lectotypes. Steetz, in his own hand, stated on the sheets that he purchased the specimens of each taxon in 1843. He also recorded Preiss’s collection data on the sheets and clearly designated them as new taxa by adding 153 “nobis!” after each name. Furthermore the sheets from Steetz’s herbarium contain as good, or much better material than those in LD, S, or the collections at MEL obtained with Sonder’s herbarium. Unless it is otherwise indicated, by reference to microfiche, photographs or the use of the abbreviation ‘n.v.’, it should be assumed that all specimens, types or otherwise, cited in this paper have been seen by the author. Key to Segregate Genera and Species of Angianthus Sensu Lato 1 . General involucre consisting of 2 leaf-like, overlapping or connate bracts which more or less enclose the capitula(fig.lg) 6. Dithyrostegia 1. General involucre absent or inconspicuous or if well developed consisting of more than 2 bracts 2. Capitular bracts with papillae at the apex (fig. li) Pogonolepis 2. Capitular bracts lacking papillae at the apex 3. Leaves opposite (at least in lower Vi of plant) and distinctly petiolate; laminae 1-2.5 cm long, 0.4-1 cm wide 4. Cephalosorus 3. Leaves alternate or if opposite then lacking petioles and less than 0.3 cm wide 4. Achene obliquely attached to floret; pappus an oblique scale (fig. Ih) 2. Pleuropappus 4. Achene not obliquely attached to floret; pappus absent or not an oblique scale 5. Paleae present, the bracts resembling the capitular bracts, whitish, ± opaque and with thick cell walls 9. Siloxerus 5. Paleae absent 6. Plants prostrate; compound heads woolly; pappus consisting of 8-12 barbed bristles united in a short ring at the base 1. Gnephosis burkittii 6. Plants not with the above combination of characters 7. Base of floret or apex of achenes with long hairs 8. Capitular bracts 4-5; capitulum-subtending bract distinct, rigid and opaque .... \ . Angianthus connatus 8. Capitular bracts 2 or 3; capitulum-subtending bracts absent 3. Epitriche 7. Base of florets or apex of achenes without long hairs 9. Bracts of general involucre with a leaf-like midrib and broad, hyaline, wing- like margins (fig. If), the bracts about the length of the compound heads . . . 7. Hyalochlamys 9. Bracts of general involucre absent or not as above 10. Compound heads with c. lOcapitula; capitulum-subtending bracts rigid and leaf-like \ . Angianthus axilliflorus 10. Compound heads usually with more than c. 10 capitula (commonly 30-several hundred); capitulum-subtending bracts primarily hyaline 11. Capitulum-subtending bracts morphologically ± similar, except for the concave nature of some, to the capitular bracts (fig. Ik-m); achenes brown A ngianthus 11. Capitulum-subtending bracts totally unlike the capitular bracts (fig. la-e); achenes pink or purple 5. Chrysocoryne 1. Angianthus Wendl., Collect. PI. 2:31 (71808); DC., Prodr. 6:150 (1838); Steetz in Lehm. PI. Preiss. 1:438 (1845); Benth., FI. Austr. 3:560 (1867) p.p.; Benth. in Benth. & Hook, f.. Genera PI. 2:319 (1873) p.p.; Hoffman in Engler & Prantl, Natlirl. Pflanzenfam. IV (5):193 (1890) p.p. Type: A. tomentosus Wend. Cassinia R. Br. in W. & W. T. Aiton, Hort. Kewensis 2nd ed. 5:184 (1813), non Cassinia R. Br., Trans. Linn. Soc. London 12:126 (1818) nom. cons. Type: C. aurea R. Br. { = A. tomentosus Cylindrosorus Benth., Enum. PI. 62 (1837); DC., Prodr. 6:151 (1838). Type: C. flavescensBenXh. {=A. tomentosus Phyllocalymma Benth., Enum. PI. 61 (1837); DC., Prodr. 6:150 (1838); Steetz in Lehm. PI. Preiss. 1:436 (1845). Type: P. micropodioides Benth. {=A. micropodioides (Benth.) Benth.) Skirrhophorus DC. in Lindl. ex DC., Prodr. 6:150 (1838); DC. in Lindl., Nat. Syst. Bot. 2nd ed. 260 (1836) nomen nudum; DC. in Deless., Icon. Select. PI. 4:22,t.51 (1840) Skirrophorus'); Steetz in Lehm. PL Preiss. 1:438 (1845); A. Gray, Hook. J. Bot. Kew 154 Card. Misc. 3:147 (1851) p.p. (as to sect. Skirrhophorus). Type: 5. cunninghamii DC. {^A, cunninghamii {DC,) Benth.) Eriocladium Lindl., Edwards’ Bot. Reg.: Swan River Append. 24 (1839). Type: E, pyramidatum {=A. cunninghamii (DC.) Benth.) [Cephalosorus auct. non A. Gray: see synonymy of A. microcephaius.] [Siioxerus auct. non Labill.: see synonymy of A. brachypappus & A, tomentosus.] [Styloncerus auct. non Spreng,, nom. illeg.\ various species, as to combinations of Kuntze, Rev. Generum PI. 367 (1891).] Annual herbs (or perennial shrub, A. cunninghamii only). Major axes prostrate, decumbent, ascending or erect, glabrous or hairy; stems simple or forming major branches at basal and/or upper nodes; major axes often developing minor shoots. Leaves usually alternate (sometimes opposite), sessile, entire, hairy (sometimes almost glabrous), often with a small hyaline appendage at the apex. Compound head ± ellipsoid or lanceoloid to depressed ovoid; bracts subtending compound heads commonly inconspicuous and much less than c, 14 the length of the head (sometimes the bracts c. the length of the head), the outer ones leaf-like, the inner ones with hyaline apices; general receptacle cylindrical to ± oblong or ovoid to broadly depressed ovoid, consisting of a single major axis with minor receptacular axes distributed ± evenly over it. Capitula 20-200(c. 1500) per compound head, each capitulum with 1(4) abaxial, hyaline subtending bracts that overlap the capitular bracts. Capitulum-subtending bracts narrowly elliptic to elliptic or narrowly oblong to oblong or lanceolate to ovate or oblanceolate to obovate; laminae rarely with a distinct constriction in the upper parts; midrib usually conspicuous, opaque, extending c. Va to Vi the length of the bract, glabrous or variably hairy, sometimes with a few glandular hairs present. Capitular bracts 4 {A. microcephaius with 2 or 3 only), hyaline, with an opaque midrib, arranged so that 2 outer, variably concave bracts (always present) surround 2 (absent or 1 only in A. microcephaius) inner flat bracts. Concave bracts with the lamina variably constricted in the upper half; midrib usually conspicuous and extending c. 14 to Vi the length of the bracts, glabrous or variably hairy, sometimes a few glandular hairs present. Inner flat bracts narrow elliptic to elliptic or narrowly oblong or oblanceolate to obovate, tapering gradually to the base or conspicuously attenuated in the lower /2 to Vs; lamina not or variably constricted in the upper half; midrib usually conspicuous, elongate and extending c. ^3 to Vi the length of the bract, glabrous or variably hairy, sometimes a few glandular hairs present, sometimes with an entire or dissected wing-like extension from the adaxial surface. Florets (1)2(3) per capitulum; corolla (3, 4)5-lobed; style branches truncate; stamens (3, 4) 5, with tailed anthers, Achenes dlipsoid or ± obovoid, glabrous or variably papillose or pubescent. Pappus setose, paleaceous, coroniform or absent. Distribution (Fig. 2): Of the 15 species of Angianthus only >1. brachypappus and A. glabratus are absent from Western Australia and 11 are endemic to that state. Ecology: Four or five species are restricted to saline depressions but the majority occupy a wider range of habitats. Affinities/Generic Characteristics: The genus is characterised by the usual presence of two inner flat bracts and two outer concave bracts per capitulum (fig. Ik), the presence of one, rarely two or three, capitulum-subtending bracts per capitulum, the usual occurrence of two florets per capitulum and the usual presence of minor receptacular axes on the general receptacle (fig. Ij). The affinities of the genus appear to be with the monotypic genera Pleuropappus and Epitriche. Both Pleuropappus and Angianthus have morphologically similar capitular and capitulum-subtending bracts, but P phyllocalymmeus is readily distinguished by the oblique nature of the pappus and achene. Furthermore the capitula in this species are arranged in a spike-like fashion (i.e. they do not have distinctive minor receptacular appendages) and there are four distinctively arranged capitulum-subtending 155 A.milnei 0 A. prostratus ♦ A.glabratus a A. micropodioides • V A. preissianus • A.pygmaeus o Fig. 2. Distribution of species of Angianthus. bracts per capitulum. The capitular bracts of Epitriche 3XQ also morphologically similar to those of Angianthus but Epitriche is readily distinguished by the lack of minor receptacular appendages, the apparent absence of capitulum-subtending bracts and the distinctive ring of hairs at the apex of the achene. Evolution/Reproductive Biology: Within the genus a number of species are very closely related. The annual species with large, usually more or less oblong compound heads mostly differ from each other only in aspects of the pappus, size and morphology of the capitular bracts and habit. Thus A. acrohyalinus. A, brachypappus, A. cornutus, A. conocephalus. A, cyathifer, A. glabratus, A. milnei and A. tomentosus appear to form a species cluster. Not far removed are A, micropodioides, which has a somewhat reduced compound head and pubescent achenes, and A. microcephalus. The latter species also has a somewhat reduced compound head but unlike other taxa the two inner capitular bracts are absent. In all other aspects the species undoubtedly belongs to Angianthus. A further species, A. prostratus, also has affinities with species such as A. cornutus and A . conocephalus. The remaining annual species, namely A. drummondii,A. pygmaeus and /I. preissianus also form a small species cluster (see under respective species). The only perennial species, A. cunninghamii, is a coastal shrub found in Western Australia. The perennial habit plus the occasional presence of three florets per capitulum suggest that the species may be a rather primitive member of the genus. With the exception of A. preissianus all members of the genus have pollen-ovule ratios of several thousand, i.e. they all appear to be outbreeders. In contrast the average P/O of 119.7 recorded for A. preissianus clearly reflects its inbreeding nature (Short 1981a, b). A number of species, e.g. A. tomentosus, A. milnei and A. acrohyalinus, are capable of producing compound heads containing from 500 to about 2,000 capitula, although under natural conditions they produce relatively few seeds. For example 156 A. tomentosusy a species known to be self-compatible, may only set about 100 seeds per compound head. Although actual seed set may be dependant upon a number of parameters it is possible that the low actual: potential seed ratio is a result of selection for maintenance of a showy inflorescence, the prime purpose of which is to attract pollinators. In self-incompatible and partially self-incompatible species it is obviously important to attract potential pollinators. Similarly in predominantly self-pollinated species it is also often advantageous to cross-poUinate. A compact inflorescence may be more conspicuous to insect pollinators than an open inflorescence or very smdl, compact inflorescences, particularly if plants are sparsely spread throughout a population. Furthermore any increase in the time spent and/or a change in foraging pattern by insects when on large, compact inflorescences as compared to small and/or more open inflorescences may also increase the likelihood of pollination, whether it be self- or cross- pollination. Thus the cost of maintaining a large inflorescence may be off-set by a higher degree of pollination. Alternatively, considering the semi-arid to arid environment inhabited by the above species the actual seed set may be regulated primarily by moisture availability and hence longevity of a plant. Thus it could be expected that actual seed set will only approach potential seed set in years of high rainfall. In dry conditions many seeds and in fact many florets may fail to develop because of water stress. In contrast to the outbreeding species, A. preissianus appears, as expected, to produce a full complement of apparently mature, viable seed. Keighery (1981) has noted that meat ants (Iridomyrmex spp.) are common flower visitors and pollen vectors of A. conocephalus. Flies and ants are commonly found on flowering plants of all species and are possibly the main pollen vectors in the genus. Key to Speces of Angianthus Sensu Stricto 1. Perennial shrub; major axes 20-50 cm long (west coast of Australia, normally on fore-dunes, sometimes in coastal saline flats) 15. cunninghamii 1. Annual herb; major axes 5-30(44.5) cm long 2. Florets 1 per capitulum, flat capitular bracts absent or rarely 1 per capitulum . . . 11. A. microcephalus 2. Florets 2 per capitulum; flat capitular bracts 2 per capitulum 3. Pappus absent 4. Midrib of capitular bracts with hairs Vz to Vi the length of the bract \0. A. prostratus 4. Midrib of capitular bracts glabrous or with hairs less than c. VS the length of the bract 5. Bracts subtending compound heads inconspicuous or less than c.VS (rarely to cAA) the length of the head (if up to cAA then the inner capitular bracts with horn-like basal appendages); compound heads ± ovoid or narrowly ellipsoid to ellipsoid 6. Flat capitular bracts usually abruptly attenuated in the lower VS and with horn-like basal appendages (fig. 30; compound heads ovoid 1 9. A. cornutus 6. Flat capitular bracts gradually tapering toward the base and lacking basal appendages (fig. 3c); compound heads narrowly ellipsoid to ellipsoid 7. Capitulum-subtending bracts with the lamina constricted in the upper part and the midrib ± denselyhairy toward the apex (fig. 3b) 2. A. milnei 1. Capitulum-subtending bracts without a constriction in the upper part and with the midrib glabrous or sparsely hairy toward the apex 1. A. milnei* 5. Bracts subtending compound heads c. equal to or exceeding the length of the head; compound heads broadly ovoid to broadly depressed-ovoid 8. Flat capitular bracts lacking an entire wing-like extension from the adaxial surface of the midrib 8. micropodioides* 8. Flat capitular bracts with an entire wing-like extension from the adaxial surface of the midrib (fig. 3e) or if absent then the florets 3- or 4-lobed 9. Florets 3- or 4-lobed; pollen grains 16-60 pollen grains per anther 14. preissianus 9. Florets (4)5-lobed; pollen grains c. 350-500 per anther 10. Major axes prostrate; compound heads broadly depressed to ovoid .... A. pygmaeus 10. Major axes erect; compound heads broadly to very broadly ovoid 12. >1. drummondii* 157 3. Pappus present 11. Bracts subtending compound heads c. equal to, or exceeding, the length of the headf 12. Pappus of 5 or 6 jagged scales, each scale terminating in a single smooth or minutely barbellate bristle (Fig. 3k) 8. micropodioides 12. Pappus a cup of scales or a small ring 13. Flat capitular bracts with a wing-like extension from the adaxial surface of the midrib (fig. 3e) \1. A. drummondii 13 . Flat capitular bracts lacking a wing-like extension from the adaxial surface of the midrib 8. micropodioides* 11. Bracts subtending the compound heads inconspicuous or less than c. ‘/4 the length of the head (sometimes reaching c. !4 the length of the head in A. brachypappus) 14. Leaves (at least the upper ones) conduplicate, often incurved at the apex and with a distinct hyaline appendage; pappus of 4-6 bristles, barbellate in the lower Vz, united into a small. slightly toothed ring at the base (fig. 3g) 1. ^4. acrohyalinus 14. Leaves not conduplicate; pappus not as above 15. Pappus of 2 or 3 jagged scales, each scale terminating in I or 2 terminally subplumose bristlesextendingthelengthofthecorolla(fig. 3h) 5. A. tomentosus 15. Pappus a jagged cup (of ± distinct scales) or a ring 16. Leaves almost glabrous, succulent and cylindrical when fresh 4. A. glabratus 16. Leaves conspicuously hairy, usually not succulent 17. Flat capitular bracts tapering gradually to the base; compound heads ± narrowly ellipsoid to ellipsoid 18. Pappus a small, jagged ring (Cliff Head — Jurien Bay region, Western Australia) 2. ^4. milnei* 18. Pappus cup-shaped, jagged, often appearing as 2-4 distinct scales (SW. corner Northern Territory and possibly central Western Australia) 3. /4. cyathifer 17. Flat capitular bracts abruptly attenuated in lower V^-Vi (fig. 3d); compound heads usually narrowly ovoid to ovoid, sometimes ± narrowly ellipsoid to ellipsoid 19. Leaves usually oblanceolate, sometimes linear or narrowly elliptic, 1-3(3. 2) cm long, 0. 1-0.5 cm wide; pappus a jagged cup, 0.15-0.7 mm long, often with 1 or 2 bristles extending Vi-Vi the length of the floret (fig. 3j) (eastern South Australian Queensland, New South Wales, western Victoria) 6. A. brachypappus 19. Leaves ± linear, rarely oblanceolate, 0.5-1. 5(1.7) cm long, c.0.1 cm wide; pappus ajagged ring 0.1-0. 3 mm long (fig. 3i)(Nullarbor Plain region) 7. y4. conocephalus A number of collections, referred to in the key above as A. milnei*, A. micropodioides and >1. drummondii*, are to some extent atypical of the species to which they are referred and under which their diagnostic features are outlined. They possibly represent distina taxa but further collections are required to substantiate this view. 1. Angianthus acrohyalinus Morrison, J. Bot. 50:167 (1912); Grieve &Blackall,W. Aust. Wildfls 812 (1975). Type: “Globe Hill Station and Minderoo, Ashburton River, October.'’ Lectotype (here designated): Morrison s,m, Globe Hill, Ashburton River, 6.X.1905 (PERTH). Syntype: Morrison s.n., Minderoo, Ashburton R., ll.x.1905 (K). Annual herb, (6)10-30 cm high. Major axes erect or ascending, flexuose, hairy; stem often simple in the smaller plants, to c. 7 cm high, but usually forming major branches at basal and/or upper nodes. Leaves alternate, lanceolate, flat to conduplicate, apex often incurved, (0.3)l-6(7) cm long, (0.2)0. 3-0. 6(0. 8) cm wide, the lower ones slightly mucronate, the rest usually with a distinctive hyaline appendage at the apex, all leaves tSince the completion of the manuscript for this revision a collection Newbey 9154 from c. 72 km NNW. of Bullfinch, Western Australia, has been drawn to my attention. It appears to represent a distinct species of Angianthus. It is an annual, has 2 florets per capitulum, a pappus of 2-3 scales, each of which terminates in a single, minutely barbellate awn extending the length of the corolla, and the bracts subtending the compound heads are about equal to or exceeding the length of the compound head. It keys out to lead 12 in the key to species. 1mm 158 Fig. 3. Characteristics of Angianthus. a — general arrangements of capitular bracts {A. glabratus. Short 905). b — capitulum-subtending bract with the lamina constricted in the upper part {A. milnei. Short 473). c — capitulum-subtending bract lacking constriction {A. glabratus. Short 905). d — flat capitular bract abruptly attenuated in lower V^-Vi {A. conocephalus, Ising 1529). e — flat capitular bract with wing-like extension {A. preissianus, Short 716). f — flat capitular bract with horn-like basal appendages (A. cornutus, Chinnock 4692). g — pappus of A. acrohyalinus {Short 484). h — pappus of A. tomentosus {Short 1075). i — pappus of A. conocephalus {Ising 1529). j — pappus of A. brachypappus {Bruchner MEL 84877). k — pappus of A. micropodioides {Short 613). 159 hairy. Compound heads ellipsoid or ovoid, 0.8-2.5(2.8) cm long, 0. 7-1.1 cm diam.; bracts subtending compound heads not forming a conspicuous involucre but several to c. 10, leaf-Uke hairy bracts with hyaline apices present, grading into capitulum- subtending bracts; general receptacle cylindrical to narrowly oblong. Capitula 300-1,000 (c. 1,500) per compound head; capitulum-subtending bracts 1(2, ?3), if more than one then extra one(s) abaxial to, and overlapping, the inner, all bracts narrowly elliptic or obovate, sometimes ± oblanceolate, 2. 2-3. 5(3. 9) mm long, 0.6-l(1.2) mm wide, the midrib glabrous or variably hairy toward the apex and often some glandular hairs present. Capitular bracts with the two concave ones 2. 3-3. 6(3. 9) mm long, the midrib glabrous or sparsely hairy toward the apex and sometimes with some glandular hairs; flat bracts 2, oblanceolate to obovate, ± gradually tapering toward the base, 2.3-3.5(3.75) mm long, 0. 6-0.8 mm wide, the midrib usually glabrous but a few glandular hairs sometimes present. Florets 2; corolla 5-lobed, the tube tapering ± gradually to a variably swollen base, 1.65-2.7(3.2) mm long, c. 0.2-0. 3 mm diam. Achenes ± obovoid or ovoid, 0.45-0.85 mm long, c. 0.2 mm diam., papillose. Pappus of 4-6 bristles c. the length of the corolla tube, barbellate in the lower '/i , sub-plumose in the upper Vz, united into a small slightly toothed ring at the base. Fig. 3g. Distribution (Fig. 2): North-west Western Australia between latitudes 22°S and 27°S and west of longitude 118°E. Common. Ecology: Commonly associated with chenopod dominants in clay soils but also grows in loam or red sand. Collectors’ notes include “On red Triodia sand-hills”, “Reddish loam, with limestone rock. Chenopod, Acacia dominants” and “Saline depression . . . growing amongst Arthrocnemum \ = Halosarcia\ on powdery white-brown clay loam”. Note: 1. A. 6rcro/i>'fl///7iishas obvious affinities with A. tomentosus ecs\d A. m/7ne/ and the 3 species all occur in the north-west of Western Australia. However the flexuose axes, conduplicate leaves and the sub-plumose to plumose pappus bristles distinguish A. acrohyalinus from all other species. Selected Specimens Examined (8/15): Western Australia — Barnett s.n., Carnarvon, s. dat. (PERTH); Beard 2974, Onslow, 23.viii.1963 (PERTH); Chadwick 1473, Cape Range, s. dat. (PERTH, UWA); Demarz 2468, Hamersley Station, 19.viii. 1970 (PERTH); Short 435, 57 km from Denham along Overlander Roadhouse road, 21.viii.l977 (AD)- Short 444, c. 43 km N. of Overlander Roadhouse, 21.viii.l977 (AD); Short 483, c. 3 km N. of Lyndon River on Mimlya-Cape Range road, 27.viii.1977 (AD); Short 1000, c. 19.3 km SE. of Nanga, 17.xi.l979 (AD). 2. Angianthus milnei Benth., FI. Austr. 3:564 (1867). — Styloncerus milnei (Benth.) Kuntze, Rev. Generum PI. 367 (1891). Type: “Shark’s Bay and Dirk Hartog’s Island, abuiidant, Milne.” Lectotype (here designated): Milne s.n.. Shark Bay, New HoUand, plants, s. dat. (K). Isolectotype: PERTH (fragments only). Syntype: Milne s.n., Dirk Hartog Island, planis, plentiful, s. dat. (K, mounted with lectotype). Annual herb, 6-25(26) cm high. Major axes erect or ascending, hairy; stem often simple in the smaller plants, to c. 10 cm high, but usually forming major branches at basal and/ or upper nodes. Leaver alternate, narrowly elliptic, lanceolate or oblanceolate, 1-6. 5(7. 2) cm long, 0.1-0.4(0.9) cm wide, sometimes slightly mucronate, the uppermost ones with a hyalirie appendage at the apex, all leaves hairy. Compound heads narrowly ellipsoid to ellipsoid, 0. 5-2(2. 2) cm long, 0. 5-0.9 cm diam.; bracts subtending compound heads not forming a conspicuous involucre but several leaf -like, hairy bracts with hyaline apices present, grading into capitulum-subtending bracts; general receptacle cylindrical to narrowly oblong. Capitula c. 100-400 per compound head; capitulum-subtending bracts 1, ± elliptic, obovate or ovate, 2.7-3.6 mm long, 0.7-1.2 mm wide, the lamina with a prominent constriction in the upper part, the midrib ± densely hairy toward the apex, the simple hairs with a length up to c. 1-1.5 times the width of the bract, glandular hairs sometimes present. Capitular bracts with the two concave ones 2.4-3. 4 mm long. 160 the midrib sparsely hairy toward the apex and sometimes with a few glandular hairs; flat bracts 2, ± oblanceolate, obovate or narrowly elliptic, gradually tapering toward the base, 2. 3-3. 5 mm long, 0. 6-1.1 mm wide, the midrib usually sparsely hairy toward the apex, glandular hairs sometimes present. Florets 2; corolla 5-lobed, the tube tapering gradually into a variably swollen base, sometimes a ± abrupt taper in the lower Vi, 1.7-2. 3 mm long, 0.3-0.5 mm diam. Achenes obovoid, 0. 6-0.9 mm long, c. 0.3 mm diam., papillose. Pa/?/7w;yabsent. Fig. 3. Chromosome number: n = 13 (Turner, 1970 — as tomentosus\ T5382), Distribution (Fig. 2): North-west Western Australia between latitudes 21°S and 2S°S and west of longitude c.ll9°E. Common. Ecology: Occurs on the margins of saline depressions or in open scrubland, shrubland or tussock grassland formations. Collectors' notes include “Wattle scrub on stony plateau”, “In red sand on Spinifex plain” and “Reddish loam, with limestone rock. Chenopod, Acacia dominants”. Notes: \. A. milnei in many respects resembles A. tomentosus and A. acrohyalinus and commonly grows with the latter. However it is readily distinguished from them by the absence of a pappus and by the presence of the distinctive lamina and hairs on the midrib of the capitulum-subtending bract. The conduplicate leaves of A. acrohyalinus are distinctly different from those of A. milnei. 2. Collections from the Cliff Head-Jurien Bay region (see list below) lack the distinctive lamina and long hairs on the capitulum-subtending bracts and at least one collection. Burns 128, contains some florets with a small, jagged, ring-like pappus. Their distribution falls outside that of typical A. milnei and collection data suggest that the populations tend to grow in saline regions. It appears that the collections represent a distinct taxon, possibly a subspecies of A. milnei. More collections should be examined before any formal status, if any, is conferred upon this taxon. Selected Specimens Examined (8/20): Western Australia — Beard 6020, 10 miles W. of Gascoyne Junction, 18.viii.l970 (NSW, PERTH); Demarz 4689, Lake Austin, 23.x. 1973 (KP, PERTH); Gardner 6007, No. 2 tank between Geraldton and Shark Bay, 17.ix.l941 (PERTH); Gardner 7836, Tuckanarra, 13.x. 1945 (PERTH); George 1125, 20 miles E. of Onslow, 27.viii.1960 (PERTH); Short 483A, c. 3 km N. of Lyndon River on Minilya-Cape Range road, 27.viii.1977 (AD); Speck 678, 15 miles E. of Berringarra, 6.ix.l957 (CANB, PERTH); Turner 5382, 50 miles SE. of Gascoyne Junction, 22.viii.1965 (PERTH). Specimens Examined, A. milnei variant Western Australia — Burns 128, Cliff Head, S. of Dongarra, 25.x. 1967 (PERTH); Keighery578, 3 km E. of Jurien Bay, 20.X.1975 (KP); Paust 1158, 3.2 km NE. of Jurien Bay, 3.X.1972 (PERTH); Short 1012, c. 2.1 km from Jurien Bay, I9.xi.l979 (AD). 3. Angianthus cyathifer Short, sp. nov. [Angianthus tomentosus met. non Wendl.: Chippendale, Trans. Roy. Soc. S. Aust. 84:103 (1961).] Herba annua. Axes maiores plerumque ascendentes decumbentesve, raro erecti, (4)8-18(24) cm longi, variepilosi; caulis plerumque ramis maioribus vix clarus. altema, sublinearia velelliptica vei ita anguste, (0.3)0.5-2.5 cm longa, c. 0.1 cm lata, pilosa interdum paulum mucronata, mucrone foliorum superiorum hyalino. Glomeruli ellipsoidei vel ita subanguste, 1.2-2.5(2.9) cm longi, c. 0.5 cm diametro; bracteae glomerulos subtendentes paulum clarae; receptaculum cylindraceum usque subanguste oblongum. Capitula c. 100-500; bracteae capitulum subtendentes l(2-?3), ± ellipticae, obovataevel ± oblongae, (1.7)2.1-2.7 mm longae, 0.8-1.3 mm latae; costa plerumque pilosa, interdum pilos glandulosos ferens. Bracteae intra capitulum: duo concave (1.7)2-2.7 mm longae, costa ad apicem pilosa, interdum pilos glandulosos ferenti; duo planae anguste ellipticae usque ellipticae obovataeve, (1.65)2.2-2.6 mm longae, 0.6-1 mm latae, gradatim attenuatae, costa plerumque glabra, interdum pilos glandulosos ferenti. Flosculi 2; corolla5-lobata./lc/7e/;/fl'subobovoidea, 0.5-0.75 mm longa, c. 0.2-0.3 mm diametro, papillosa. Pappus subeyathiformis, laceratus saepe 2-4 squamis Claris ad basem coniunctis similis, 0.2-0.55 mm longus. 161 J-.i ! i. HERBARIUM OF NORTHERN TERRITORY 1^, -rlauiida ?=• ?5»:: Coll. ' •" • '-^ts Dote " ' 1,-74. Note* >sat^:^ -lilt aJ>Dv<> ssfiall s«lt lafee, fi< r- , /eUfw fl-araTs, Del. -. .. "i-sssU f,i75. iljti^p! 89*5«'?8 8l0i*S-‘S^7S's20’t3 Fig. 4. /4. cyathifer Short. Holotype {Latz 4961, NT). Holotypus (fig. 4): Latz 4961, Erldunda Station, 25°23'S, 132°39'E. Dominating sandy flat above small salt lake. Decumbent herb, bright yellow flowers, 26. iv. 1974 (NT 43752). Isotypus: BRI 013705 (NT label indicates that further duplicates occur in CANB, DNA & NT reference colln.). 162 Annual herb. Major axes usually ascending or decumbent, rarely erect, (4)8-18(24) cm long, hairy; stem usually not distinguishable from major branches. Leaves alternate, ± linear or ± narrowly elliptic to elliptic, (0.3)0. 5-2.5 cm long, c. 0.1 cm wide, sometimes slightly mucronate, the upper most ones with a hyaline appendage at the apex, all leaves hairy. Compound heads ± narrowly ellipsoid to ellipsoid, 1.2-2. 5(2.9) cm long, c. 0.5 cm diam.; bracts subtending compound heads not forming a conspicuous involucre but several leaf-like bracts with hyaline apices present, grading into capitulum-subtending bracts; general receptacle cylindrical to ± narrowly oblong. Capitula c. l(X)-500 per compound head; capitulum-subtending bracts 1(2, ?3), if more than one then the extra one(s) abaxial to, and overlapping, the inner, all bracts ± elliptic, obovate or ± oblong, (1.7)2. 1-2.7 mm long, 0.8-1. 3 mm wide, the midrib usually hairy and glandular hairs sometimes present. Capitular bracts with the two concave ones (1. 7)2-2. 7 mm long, the midrib hairy toward the apex and glandular hairs sometimes present; flat bracts 2, gradually tapering to the base, narrowly elliptic to elliptic or obovate, (1.65)2.2-2.6 mm long, 0.6-1 mm wide, the midrib usually glabrous, sometimes glandular hairs present. Florets 2; corolla 5-lobed, the tube tapering gradually to the base, (1)1. 4-1.9 mm long, c. 0.5 mm diam., a few glandular hairs sometimes present on the tube. Achenes ± obovoid, 0.5-0.75 mm long, c. 0. 2-0.3 mm diam., papillose. Pappus ± cup-shaped, jagged, often appearing as 2-4 ± distinct scales united at the base, 0.2-0.55 mm long. Fig. 4. Distribution (Fig. 2): South-west corner of the Northern Territory and central Western Australia. Moderately common. There is a marked disjunction in distribution between the Western Australian collections and those from the Northern Territory but all specimens appear to represent a single taxon. Further collections are required to substantiate this view. Ecology: Grows in sandy or clay soils which surround saline depressions. Collectors’ notes include “Saltpan area encroached by sand, populated by various annuals, perennial grasses and samphires”, . . on margins of salt pan with Arthrocnemum [ = Halosarcia] spp. in deep, white saline sand” and “clayish soil around salt lake”. Notes: r. The specific epithet alludes to the cup-like pappus. 2. This species has close affinities with A. tomentosus. It is readily distinguished by its pappus and the usually ascending or decumbent major axes. Specimens Examined: Western Australia — Speck 1429, 10 miles W. of Yelma, 12.ix.l958 (AD, CANB, PERTH); Ivlken 6089, Lake Miranda, 11. ix. 1979 (AD, PERTH); Wilson 7490, 30 km NE. of Nambi homestead, 29.viii.1968 (PERTH). Northern Territory — Beauglehole 50933, Tanami Desert, 27 km NW. of the Granites, 20.V.1976 (ACB); Dunlop 1816, 11 miles WSW. Granites, 2.vii.l970(NT); Latz4083, Salt Beef Lake, Tanami Sanctuary, 22.vii.1973 (NT, PERTH); iMtz 4248, Lake Neale, 28.viii.1973 (NT); Latz 6547, south of Mongrel Downs, 4.viii.l976 (MEL); Lazarides 6088, 30 miles SSW. of Napperby Station, 28. ix. 1956 (AD, BRI, CANB, NSW, NT); Maconochie 1790, 3 miles SE. Sangster’s Bore, 22.vii.1973 (NT). 4. Angianthus glabratus Short, sp. nov. Herba annua, 6-14(16) cm alta. Axes maiores erecti ascendentesve, glabri parum pilosi; caulis e nodis basalibus superioribusque, ramos maiores efficiens. Folia alterna, in vivo sucosa teretia, 0.4-1. 6(3) cm longa, c. 0.1 cm lata, mucronem carentia sed interdum folius superioribus ad apicem appendicem hyalinam ferentia, subglabra. Glomeruli ellipsoidei vel ita anguste, 1-2. 5(3. 4) cm longi, 0.4-0. 6 cm diametro; bracteae glomerulos subtendentes parum clarae; receptaculum anguste oblongum. Capitula c, 100-500; bracteae capitulum subtendentes 1(2-73), ovatae suboblongaeve, 1.8-2. 5 mm longae, 1-1.6 mm latae; costa glabra vel ad apicem varie pilosa. Bracteae intra capitulum: duo concave 1.6-2. 3 mm longae, costa glabra vel ad apicem varie pilosa; duo planae, subellipticae obovataeve, 1.6-2. 2 mm longae, 0.7-1. 2 mm latae, attenuatae, costa glabra vel ad apicem varie pilosa. Flosculi 2; corolla 5-lobata. Achenia subovoidea, c. 0.5-0. 8 mm longa, c. 0.3 mm diametro, papillosa. Pappus cyathiformis, varie laceratus, interdum c. 5 squamis ad basem coniunctis similis, 0. 2-0.4 mm longus. 163 Fig. 5. A. glabratus Short. Holotype {Short 838, AD). Holotypus (fig. 5): Short 838, c. 5.8 km west of Nectar Brook Station along road to Chinamans Creek (34®42'S, 137°54'E). Associated mih Arthrocnemum [ = Halosarcia], Atriplex, Alzoon. Salty grey-white clay. Very common. 9.xii.l978 (AD). Isotypus: CANB, K, MEL, PERTH. 164 Annual herb, 6-14(16) cm high. Major axes erect or ascending, glabrous or slightly hairy; stem simple or forming major branches at basal and/or upper nodes. Leaves alternate, succulent and cylindrical when fresh, 0.4-1.6(3) cm long, c. 0.1 cm wide, not mucronate but sometimes the upper ones with a hyaline appendage at the apex, all leaves ± glabrous. Compound heads narrowly ellipsoid to ellipsoid, 1-2. 5(3.4) cm long, 0.4-0. 6 cm diam.; bracts subtending compound heads not forming a conspicuous involucre but several leaf-like, hairy bracts with hyaline apices present, grading into capitulum-subtending bracts; general receptacle cylindrical to narrowly oblong. Capitula c. 100-500 per compound head; capitulum-subtending bracts 1(2, ?3), if more than one then the extra one(s) abaxial to and overlapping the inner, all bracts ovate or ± oblong, 1.8-2. 5 mm long, 1-1.6 mm wide, the midrib glabrous or variably hairy toward the apex. Capitular bracts with the two concave ones 1.6-2. 3 mm long, the midrib glabrous or variably hairy toward the apex; flat bracts 2, ± elliptic or obovate, gradually tapering towards the base, 1.6-2. 2 mm long, 0.7-1. 2 mm wide, the midrib ^abrous or variably hairy toward the apex. Floret 2; corolla 5-lobed, the tube tapering ± gradually to the base, 1.1-1. 5 mm long, c. 0.4 mm diam. Achenes ± obovoid, c. 0. 5-0.8 mm long, c. 0.3 mm diam., papillose. Pappus cup-shaped, variably jagged, sometimes appearing to be composed of c. 5 scales joined at the base, 0. 2-0.4 mm high. Figs.: 3a, c; 5. Distribution (Fig. 2): Upper Eyre Peninsula, South Australia between latitudes 31°S and 33°S and longitudes 135°E and 138°E. Moderately common. Ecology: Commonly grows on the margins of saline depressions where usually associated with species of Halosarcia, Atriplex and Aizoon, but also occurs on coastal sand-dunes. Also recorded in an Acacia linophylla association on red sand dunes. Notes: 1. The specific epithet refers to the more or less glabrous nature of the species. This characteristic readily distinguishes it from perhaps its closest relatives. A, brachypappus and A. tomentosus. Selected Specimens Examined (6/14): South Australia — Chinnock 2618, 30 km W. of Kingoonya on the Tarcoola road, 27.ix.1975 (AD); Eichler 18817, SW. end of Pernatty Lagoon, 22.X.1966 (AD); Higginson s.n.. Port Augusta, 1955 (ACB); Lay 547, Kenella Rocks, Wilgena Station, 1.x. 1971 (AD); Short 793, c. 26.7 km S. of Hiltaba homestead, 25.ix.1978 (AD); Specht & Carrodus 96, 40 miles N. of Nonning homestead, 16.xi.l958 (AD). 5. Angianthus tomentosus Wendl., Collect. PI. 2:32; t.48 (71808); Brown, Trans. Linn. Soc. London 12:103 (1817); Cass., Diet. Sci. Nat. 14:483 (1819); DC, Prod, 6:150 (1838); Sond., Linnaea 25:487 (1853); Benth., FI. Austr. 3:562 (1867); J. M. Black, FI. S. Aust. 1st ed. 644 (1926), 2nd ed. 924 (1957); Willis, Handb. PI. Viet. 2:729 (1973); Grieve & Blackall, W. Aust. Wildfls 811 (1975). — Styloncerus tomentosus (Wendl.) Kuntze, Rev. Generum PI. 367 (1891). — Siloxerus tomentosus (Wendl.) Ostenf., Biol. Meddel. Kongel. Danske Vidensk. Selsk. 3:137 (1921). Type: “Botany Bay”. Lectotype (here designated): GOET (ex herb. Wendl., Herrenhausen; photograph only seen). Probable isoLECTOTYPEs: GOET (ex herb. Bartling; photograph only seen), MEL 543^5 (ex herb. Steetz), (see note 2 below). Cassinia aurea R. Br. in W. T. Aiton, Hort. Kewensis 2nd ed. 5:184 (1813); Spreng., Syst. Veg. 16th ed. 426 (1826). Type: “Nat. of the South coast of New Holland. Robert Brown, Esq. Introd. 1803, by Mr. Peter Good”. Type specimen: Brown s.n.. Bay IV, South Coast, s. dat. (K), (see note 3 below). Cylindrosorus flavescens Benth., Enum. PI. Hueg. 62 (1837). — Angianthus flavescens (Benth.) Steetz in Lehm., PI. Preiss. 1:438 (1845). Type: “Swan River (Hugel)”. Lectotype (here designated): Hugels.n., Swan River, s. dat. (K, herb. Benth.). Isolectotype: W. Probable isolectotype: MEL 84773 (see note 4 below). 165 Annual herb, (3)5-30(44.5) cm high. Major axes erect or ascending, hairy; stem often simple in the smaller plants, to c. 10(19) cm high, but usually forming major branches at basal and/or upper nodes. Leaves alternate, oblanceolate or sometimes a few narrowly elliptic, 0.5-3(5.1) cm long, 0.1-0. 3(0.5) cm wide, sometimes slightly mucronate, the upper ones with a small hyaline appendage present at the apex, all leaves hairy. Compound heads ellipsoid or lanceoloid to ovoid, 0.7-3(5.3) cm long, 0.4-0. 8(1.1) cm in diam.; bracts subtending compound heads not forming a conspicuous involucre but several leaf-like, hairy bracts with hyaline apices present, grading into capitulum-subtending bracts; general receptacle ± cylindrical to narrowly oblong. Capitula 150-700(1,000) per compound head; capitulum-subtending bracts 1(2-4), if more than one then the extra one(s) abaxial to, and overlapping, the inner, all usually ovate, sometimes ± lanceolate, elliptic or ± obovate, (2.1)2.3-3(3.3) mm long, 0.7-1(1.45) mm wide, the midrib glabrous or variably hairy toward the apex and sometimes with a few glandular hairs. Capitular bracts with the two concave ones (1.8)2.1-2.7(3) mm long, the midrib usually glabrous but sometimes a few glandular hairs present; flat bracts 2, ± obovate, gradually tapering to the base, rarely a more or less abrupt taper in the lower Vi , (1.8)2.1-2.8(3) mm long, 0.6-l(1.2) mm wide, the midrib usually glabrous but sometimes a few glandular hairs present. Florets 2; corolla 5-lobed, the tube tapering gradually to the base or sometimes an abrupt taper in the lower Vi, (1.2)1. 3-2(2. 2) mm long, c. 0.35-0.4 mm in diam. Achenes obconical, 0. 5-0.8 mm long, c. 0.3 mm in diam., papillose. Pappus of 2-3 jagged scales, 0.25-0.7(0.8) mm long, each scale terminating in 1-2 terminally sub-plumose bristles extending to c. the length of the corolla tube, the total pappus length (1.25)1.35-1.8(2) mm. Figs: Ij, 3h. Chromosome number: n=12 (Turner, 1970, T5565). Distribution (Fig. 2): South west of New South Wales, western Victoria, South Australia and below latitude c. 26°S in Western Australia. Very common. Ecology: Found in both coastal and inland situations around clay pans, saline depressions and granite outcrops or in woodland, scrub and shrubland formations. Collectors’ notes include “Red soil near ^anite outcrop. Assoc, with Waitzia sp. & Sida calyxhymeniaP, “Claypan depression with Casuarina cristata, Heterodendrum and Atriplex vesicariaP, “In Kochia pyramidata [ = Maireana pyramidata] association”, “On sandy loam with Melaleuca and Atriplex on edge of salty flat”, “Clay loam in mallee Eucalyptus, Melaleuca, Casuarina scrubland” and ''Eucalyptus woodland with Eremophila spp. & Cratystylus as dominant shrubs. Brown clay loam.”. Notes: 1. The publication date of volume 2 of Wendland’s Collectio Plantarum is usually cited as 1810. Stafleu (1967) has suggested that pages 16-34 and plates 43-48, which include the description of A. tomentosus, were published in 1808. 2. The type locality of A. tomentosus is cited by Wendland (?1808) as Botany Bay. However the species does not occur in eastern New South Wales. Brown (1817) and Cassini (1819) were both of the opinion that Wendland described it from material acquired from Kew. This material presumably originated from that collected by Brown and/or Good from Petrel Cove (see note 3 below). Wendland’s reference to Botany Bay as the collection site is possibly due to the fact that many of the Australian plants collected by Good and cultivated at Kew were kept in a glasshouse referred to as the “Botany Bay House” (Smith, 1881). Wendland presumably based his description of A. tomentosus on all material available to him. Of the two collections held at GOET the one selected as the lectotype comes from Wendland’s own herbarium in Herrenhausen while the other is a collection originally sent by Wendland to the German botanist Bartling in 1842 (Prof. Dr G. Wagenitz, pers. comm., 1981). The MEL collection, which consists of cultivated material, was presented to J. Steetz by Wendland’s son in 1844. This information was recorded by Steetz on the herbarium sheet. 166 3. The only type material of C. aurea seen is housed at K. It is clearly marked “Bay IV, South Coast” which indicates that it was collected from Petrel Bay, Isle St. Francis (Steam, 1962). 4. Bentham (1837) based his description of Cylindrosorus flavescens on a collection made by Hiigel in Western Australia. According to Stafleu (1967) the Hugel collections were acquired by the Vienna herbarium (W) in 1839. However a specimen was obtained from Vienna by Bentham and is now housed at K (Bentham, 1863). It follows therefore that one should lectotypify. Both the K and W specimens are well preserved and there seems no reason to give preference to either other than that Bentham presumably chose to retain the specimen at K. The sheet MEL 84773 contains a single specimen designated as C. flavescens. It comes from O. W. Sonder’s collection but the label indicates that it originally came from Vienna. Although there is no indication that the specimen was collected by Hiigel it is nevertheless a good match with the specimens from K and W. Selected Specimens Examined (19/76): Western Australia — Allan 183, Fitzgerald River, 50 miles W. of Ravensthorpe, 8.xi.l969 (BRI, PERTH); Chinnock 1068, 30 km NE. of Depot Springs homestead, 15. ix. 1973 (AD); George 3806, Elder Creek, 21.viii.l962 (PERTH); Kenneally 71/289, IVi miles W. of Ballidu, 28.ix.1971 (UWA); Short 431, Hamelin Pool, 20.viii.l977 (AD); Short 562, Edge of Mongers Lake, 18. ix. 1977 (AD); Short 612, Mt Rupert Station, 20.ix.l977 (AD); Short 620, Hines Hill, 21.ix.l977 (AD); Short 704, Newman Rocks, 29.ix.1977 (AD); Vachells.n., Kellerberin, -.xii. 1903 (NSW 138779). South Australia — Crap d52, Koonamore Station, 8. xii. 1973 (AD, CBG); Lfirwg 99-^, c. 14.7 kmSE. of Hiltaba homestead, 14.x. 1977 (AD); S/ 2 or/“ 705, c. 14.7 km W. of Yalata Mission turn-off onmain highway to Perth, 29.viii.1977 (AD); Specht & Carrodus 23, 10 miles N. of Nonning homestead, 14.xi.l958 (AD); Wace 12, Masillon Island, 5.1.1971 (AD). Victoria — D’Alton s.n., Dimboola, 1901 (NSW 138781); Henshall s.n.. Red Cliffs, 21. xi. 1968 (NT). New South Wales — Alehin 332, Wentworth, 28.x. 1975 (NSW); Green 182, Pooncarie, -.x.1974 (NSW). 6. Angianthus brachypappus F. Muell., Trans. Philos. Soc. Viet. 1:44 (1855); F. MuelL, J. Bot. (Hooker) 8:149 (1856); Benth., FI. Austr. 3:563 (1867); F. M. Bail., Qd. FI. 848 (1900); J. M. Black, R. S. Aust. 1st ed. 644 (1929), 2nd ed. 924 (1957), p.p. (excl. A. conocephalus (J. M. Black) Short); Willis, Handb. PI. Viet. 2:729 (1973). — Styloncerus brachypappus (F. Muell.) Kuntze, Rev. Generum PI. 367 (1891). — Siloxerus brachypappus (F. Muell.) Ising, Trans & Proc. Roy. Soc. S. Aust. 46:604 (1922). Type: “On barren plains near Swanhill.” Lectotype (here designated): ?Mueller s.n., Murray plains near Swanhill, s. dat. (MEL 541214). Possible Isolectotypes: GH (ex herb. O. W. Sonder, location given as “Murray”); MEL 541222 (no locality details but descriptive notes in Mueller's hand and specimens resemble those of lectotype); MEL 541212 (ex herb. Sond., resembles lectotype but locality given as “Murray”); MEL 541213 (resembles lectotype but locality given as “Murray”). Annual herb, (3)5-13.5 cm high. Major axes erect or ascending, sometimes decumbent, hairy; stem rarely simple, usually forming major branches at basal and upper nodes. Leaves alternate, usually oblanceolate, sometimes ± linear or narrowly elliptic, 1-3(3. 2) cm long, 0. 1-0.5 cm wide, usually very slightly mucronate, the upper most ones with a small hyaline appendage at the apex, all leaves variably hairy. Compound heads lanceoloid to ± ovoid or narrowly ellipsoid to ellipsoid, 1-2. 5(2. 9) cm long, CI.5-0.8 cm diam.; bracts subtending compound heads usually not forming a conspicuous involucre, rarely c. '/4 the length of the head, usually of c. 5-6(10) leaf-like bracts with hyaline apices present, grading into capitulum-subtending bracts; general receptacle cylindrical or narrowly oblong. Capitula c. 100-300 per compound head; capitulum-subtending bracts 1(2-3), if more than one then the extra one(s) abaxial to and overlapping the inner, all bracts elliptic or obovate, sometimes ± ovate, lamina rarely with a distinct constriction in the upper part, the entire bracts (2)2.3-3(3.25) mm long, 1-1. 7(1. 9) mm wide, the midrib variably hairy toward the apex. Capitular bracts with the 2 concave ones (2.1)2. 3-3. 2 mm long, the midrib variably hairy toward the apex; flat bracts 2, obovate, abruptly attenuated in the lower Vs-Vz, the edge of the bracts often incurved so as to slightly cover the florets, (2)2. 2-3(3. 3) mm long, 0. 8-1.3 mm wide, the midrib variably 167 hairy toward the apex and often with a dissected wing-like extension from the adaxial surface. Florets 2; corolla 5-lobed, the tube tapering ± gradually to a sometimes variably swollen base, (1.3)1. 5-2.2 mm long, c. 0.4 mm diam. Achenes ± obovoid, 0.5-0. 8 mm long, c. 0.3 mm diam., papillose. Pappus cup-shaped, variably jagged, often with 1 or 2 bristles extending Vi-V^ the length of the corolla tube, the cup 0.15-0.7 mm high, including the bristles the total pappus length up to 1.6 mm long. Fig. 3j. Distribution (Fig. 2): North-eastern South Australia, western New South Wales and north-west Victoria, Common. Two collections, Blake 10441 and WhiteB^l 224128, from Yelarbon, Queensland, represent a disjunct locality. Unfortunately the condition of the specimens is poor but there appears to be no reason to exclude them from A, brachypappus. Ecology: Commonly occurs on sandy soils in open areas. Collectors’ notes include “Open plain, sandy loam”, “Sandridge, very common, blue-bush association” and “Very gently undulating gilgaied depressions on brown gibber soils”. Selected Specimens Examined (8/66): South Australia — Lay 577, Balta Baltana Block, 7.x. 1971 (AD); Symon 9478, W. edge of Simpson Desert, NE. of Macumba, 28.ix.1974 (AD); Weber 1443, Andamooka Opal Fields, 8.ix.l968 (AD). Victoria — Willis s.n.,C. 1 mile E. of Berribee Tank, 31.viii.l948 (MEL 84413). New South Wales — Constable s.n., Mundi Mundi Station, 14.x. 1947 (NSW 4543); LeighS30, 50 miles NE. of Hay, 26. ix. 1963 (NSW); Richley F97, Fowler’s Gap, 20. ix. 1973 (AD). Queensland — Blake 10441, Yelarbon, 22. ii. 1936 (BRI, GH). 7. Angianthus conocephalus (J. M. Black) Short, comb, et stat. nov. Angianthus brachypappus v?ix. conocephalus J. M. Black, FI. S. Aust. 1st ed. 645, fig. 300 (1929), 2nd ed. 924, fig. 1224 (1957), basionym. Type: “Ooldea; Nullarbor Plain.” Lectotype (here designated): ? J, M. Black s.n., Ooldea, 25.ix.1920 (AD 97823002, herb. J, M. Black). Syntype (possible isolectotype): J. M. Black s.n., Ooldea, 25.ix.1920 (AD 98103149, herb, J. M. Black). Other Syntypes: J. M. Black s.n., Ooldea, 24.ix.1920 (AD 98103149, herb. J, M. Black); J. M. Black s.n., Ooldea, growing on edge of Nullarbor Plain, 24.ix.1920 (AD 98103149, herb. J. M. Black); J. M. 5/6fcA: 5. n., Ooldea, just W. of siding & near rlyine, 23. ix.l920 (AD 98103149, herb. J. M. Black); Isings.n., Nullarbor Plain, s. dat (AD 98103149), herb. J. M. Black). [Angianthus brachypappus auct. non. F. Muell.: Grieve & Blackall, W. Aust. Wildns 812 (1975).] Annual herb. Major axes usually ascending or decumbent, rarely erect, 3-8 cm long, hairy; stem usually not distinct from the major branches which develop from basal nodes. Leaves alternate, ± linear, rarely oblanceolate, 0.5-1. 5(1. 7) cm long, c. 0.1 mm wide, nor or very slightly mucronate, the upper most ones with a small hyaline appendage at the apex, all leaves variably hairy. Compound heads ± ovoid, 0. 8-1.6 cm long, 0,4-0. 6 cm diam.; bracts subtending compound heads not forming a conspicuous involucre but usually several leaf-like, hairy bracts with hyaline apices present, grading into capitulum-subtending bracts; general receptacle narrowly oblong, the capitula arranged in a spike-like fashion, the minor receptacular appendages small. Capitula c. 30-100 per compound head; capitulum-subtending bracts 1(2-4), if more than one then the extra one(s) abaxial to, and overlapping the inner, all bracts obovate, 2.3-2. 8 mm long, 0.8-1. 2 mm wide, the midrib glabrous or sparsely hairy toward the apex. Capitular bracts with the 2 concave ones 2. 3-2.7 mm long, the midrib glabrous; flat bracts 2, obovate, abruptly attenuated in the lower Vs, the edge of the bracts incurved so as to slightly cover the florets, 2-2.7 mm long, 0.9-1. 4 mm wide, the midrib glabrous or sparsely hairy toward the apex. Florets 2; corolla 5-lobed, the tube tapering gradually to the base, 1.4-2. 1 mm long, c. 0.4 mm diam. Achenes ± obovoid, 0.4-0.7 mm long, c. 0.3 mmdiam., papillose. Pappus a jagged ring, 0. 1-0.3 mm long. Figs: 3d,i. 168 Distribution (Fig. 2): Nullarbor Plain region. Common, Ecology: Occurs on both clay and loam soils. Collectors’ notes include “Common on clayey soils”, “Fine sandy loam over calcrete” and “In loam over limestone”. Note: l.A. conocephalus was originally described by Black (1929) as a variety of A. brachypappus. The var. conocephalus was considered to have a conical compound head and var. brachypappus a cylindrical head. However the shape of the compound head is quite variable. On the other hand both species exhibit distinct differences in habit and leaf morphology and usually pappus morphology. They are also allopatric. Selected Specimens Examined (5/23): Western Australia — ApHn 1656, Forrest, 31.viii.1962 (PERTH); Chinnock 1151, 30 km S. of Rawlinna, 19.ix.l973 (AD); George 8495, 30 miles NW. of Reid, 14.x. 1966 (PERTH). South Australia — Chinnock 1183, 15 km E. of Koonalda homestead, 21.ix.l973 (AD); Ising 1529, Hughes, 8.ix.l920 (AD). 8. Angianthus micropodioides (Benth.) Benth., FI. Austr. 3:565 (1867) {^micropo ides'); Grieve & Blackall, W. Aust. Wildfls 812 (1975) {'micropoides'). — Phyllocalymma micropodioides Benth., Enum. PI. Hueg. 62 (1837); Steetz in Lehm. PI. Preiss. 1:436 (1845). — Styloncerus micropodioides (Benth.) Kuntze, Rev. Generum PI. 367 (1891) {'micropodes'). Type: “Swan River. (Hiigel.).” Lectotype (here designated): Hugel s.n., Swan River, s. dat. (W). Isolectotype: K (see note 1 below). Phyllocalymma filaginoides Steetz in Lehm. PI. Preiss. 1:437 (1845); Steetz in Walper’s Repert. Bot. Syst. 6:229(1846). — Angianthus micropodioides filaginoides Ewart & J. White, Proc. Roy. Soc. Viet. 22:92 (1909) {'micropoides'). Type: “In solo arenoso — turfoso inter frutices ad fluvii Cygnorum ripam prope oppidulum Perth, mense Januario 1839. Herb. Preiss. No. 37.” Lectotype (here designated): Preiss 37, In Nova Hollandia, (Swan-River Colonia) in solo arenoso turfoso inter frutices ad flumis Cygnorum ripam leg. cl. Preiss, s. dat. (MEL 541603). Isolectotypes: LD, MEL 541604, MEL 541605 (ex herb. O. W. Sonder), MEL 583143 (ex herb O. W. Sonder), S, GH (ex herb. Klatt), (see p.l52). Annual herb. Major axes ascending to erect, 4-15 cm long, hairy; stem sometimes simple to c. 10 cm high, but usually forming major branches at basal and/or upper nodes. Leaves alternate, ± linear or lanceolate, 0.5-1. 5(2.8) cm long, 0.05-0.1 cm wide, distinctly mucronate, variably hairy. Compound heads ± depressed ovoid to broadly depressed ovoid, 0.4-0.6 cm long, 0.4-0.5 cm diam., axillary or terminal; bracts subtending compound heads forming a conspicuous involucre exceeding the length of the head, ofc. 10 leaf-like bracts, ± lanceolate to ± ovoid, 0.5-1. 5 cmlong, c. 0.1 cm wide, mucronate, hairy; general receptacle a small convex axis. Capitula c. 10-30 per compound head; capitulum-subtending bracts 1, ± oblong or ovate, 2. 1-2.8 mm long, 0.8-1. 3(1. 5) mm wide, the midrib variably hairy toward the apex. Capitular bracts with the two concave ones 2. 4-3.1 mm long, the midrib hairy; flat bracts 2, obovate, ± abruptly attenuated in the lower Vi, 2. 4-3.1 mm long, (0.75)0.9-1.25 mm wide, the midrib usually variably hairy toward the apex, rarely glabrous. Florets 2; corolla 5-lobed, the tube tapering gradually towards the base in immature florets, a more abrupt taper in the lower V 3 of mature florets which have variably swollen bases, 1.4-1. 9 mm long, c. 0.5 mm diam. Achenes ± obovoid, 0.8-1 mm long, 0. 5-0.6 mm diam., pubescent. Pappus of 5 or 6 jagged scales fused at the base, each sc^e terminating in a single smooth or minutely barbellate bristle, the total pappus c. ‘73-^3 the length of the corolla tube. Fig. 3k. Distribution (Fig. 2): Western Australia, particularly in the South West Drainage Division (Mulcahy & Bettenay, 1972), between latitudes c.28°30'S and 32°S and west of longitude c.l22°E. Locally common. 169 Ecology; Grows on sand or very sandy loam amongst Halosarcia and Melaleuca on the edge of saline depressions. Notes: 1. Bentham (1837) based his description of Phyllocalymma micropodioides on a collection made by Hiigel in Western Australia. According to Stafleu (1967) the Hiigel collections were acquired by the Vienna herbarium (W) in 1839. However a specimen was obtained by Bentham and is now housed at K (Bentham, 1863). It follows therefore that one should lectotypify. The W sheet of P. micropodioides contains three good individual specimens, the K sheet a single specimen. Thus the former sheet has been designated as the lectotype. 2. Pappus characteristics, i.e. the length of the awns and the jagged nature of the scales, were used by Steetz (1845) to distinguish Phyllocalymma filaginoides from P micropodioides. Such characteristics are however quite variable, even within a single plant, and Bentham (1867) reduced the former species to synonymy. There are however a number of specimens in which the typical awned scales of A . micropodioides are absent. For example the collection Short 1008 contains individuals with a small, jagged, ring-like pappus while a pappus is absent in specimens of Short 992 and Short 946. One collection contains some individuals which lack a pappus (referred to as Short 1012A) and others with a distinct ring-like pappus (referred to as Short 1012), Future investigations may show that the latter collections represent a distinct taxon but apart from the variable characteristics of the pappus there appear to be no features by which it can be distinguished from A. micropodioides. It may be that the pappus variation is under simple genetic control. 3. A. rnicropodioides and A. cornutus exhibit a similar habit but the latter species can be readily distinguished by the presence of horn-like basal appendages on the flat capitular bracts. Furthermore the pubescent nature of the achene appears to be unique to A. micropodioides. Selected Specimens Examined (5/19): Western Australia — Chinnock 4417 & Mortlock river just east of Meckering 22 xi 1978 (AD)- Morrison s.n.. Banks of Swan Estuary, 28.xii.1898 (CANB 209968, BRI 078604, MEL 84466 PERTH)’ PreissS6, Swan River Colonia, 1843 (MEL 583144, ex herb. Sond.; MEL 84467, ex herb. Steetz); Short 1024 c. 13.7 km NW. of Balhdu, 20.xi.l979 (AD); Short 1037, c. 8 km W. of Kalguddering, 20.xi.l979 (AD). ’ Specimens Examined, A. micropodioides Variant: Western Australia — Short 946, 3.4 km N. of Boodarocking, 13. xi. 1979 (AD); Short 954 Lake Campion, 14.xi.l979 (AD); Short 968, 45. 1 km N. of Koorda along main road to Mollerin, 14.X.1979 (AD)- S/Jor/ 987, 1 .9 km N. of Latham, 15. xi. 1979 (AD); Short 992, c. 30.4 km S. of Pindar along main road to Lake, c. 12 km N. of Carnamah, 19. xi. 1979 (AD); Short 1021, 1021A, c. 54.5 km from Nugadong along main road to Gunyidi, 19. xi. 1979 (AD). 9, Angianthus comutus Short, sp. nov. [Angianthus milnei auct. non Benth.: Grieve & Blackall, W. Aust. Wildfls 814 (1975) . j Herba annua. Axes maiores decumbentes ascendentesve, 3-10(16) cm longi. Folia alterna, linearia vel hneari-triangularia, 0. 5-1(1. 2) cm longa, 0.1 cm lata, mucronata, pilosa. Glomerulus ovoidms, 0.8-1. 2 cm longus, 0. 3-0.6 cm diametro; bracteae glomerulos subtendentes involucrum conspicuum longitudine c. 'A-V^ glomeruli partes aequante facientes; receptaculum convexum vel oblongum. Capitula c. 3040; bracteae capitula subtendentes 1(2), obovatae ± oblongaeve, 2.4-3 mm longae, 1. 2-1. 6 mmlatae; costa ad apicem variabi liter pilosa. Bracteae intra capitulum- duo concave f 4-2.6 mm longae, costa versus apicem variabiliter pilosa; duo planae, ellipticae ± obovataeye, m infima tertia parte attenuatissimae, 2.4-3 mm longae, 0.9-1 .2 mm latae, costa versus apicem varie pilosa, ad basin 2 appendicibus propritim cornuatis, c. 0. 2-0.4 mm longis. Flosculi 2, corolla 5-lobata. Achenia ± ellipsoidea, papillosa. Pappus carens. ^ Holotypus (fig. 6): Chinnock 4692, Saline depression 3.8 km E. of Carnegie (25 47 S, E). Prostrate herb growing in sand on edge of Arthrocncnium [ =Halosarcia] zone. Associated with A/zoon, Chrysocoryne and Gnephosis Abundant 16.ix.1979 (AD). IsoTYPUs: CANB, PERTH. 170 t A1. preissianus they do however have primarily 5-lobed florets and are outbreeders (Short 1981a, b). The pappus of A. drummondii also readily distinguishes it from both A. pygmaeus and A. preissianus. The variant of A. drummondii and pygmaeus closely resemble each other. However the latter taxon normally has prostrate or decumbent axes and broadly depressed to depressed ovoid compound heads whereas in the variant of A. drummondii the axes are ascending to erect and the compound heads are broadly to very broadly ovoid. Specimens Examined: Western Australia — Morrison s.n., Hotham River, 12. xi. 1904 (PERTH); Mueller s.n., Harvey River, 5.xii.l877 (MEL 85700); Mueller s.n., Preston River, 5.xii.l877 (MEL 85701). Specimens Examined, A. drummondii Variant: Western Australia — Demarz 6640, Lake Muir Swamp, 21. xi. 1977 (KP); Ehrendorfer 181, south coast area — Walpole/ Albany/Stirling Ranges, 14.xii.l%6 (PERTH); George 7293, Lake King, 3.xi.l965 (PERTH); Short 664, c. 20.5 km S. of Lake Grace along road to Pingrup, 24. ix. 1977 (AD); Short 694, Lake King, 26.ix.1977 (AD); Short 1102, Lake King, 26.xi.1979 (AD). 13. Angianthus pygmaeus (A. Gray) Benth., FI. Austr. 3:567 (1867); Diels & Pritzel, Bot. Jahrb. Syst. 35:612, fig. 69A-E (1905); Grieve & Blackall, W. Aust. Wildfls 815 (1975). — Skirrhophorus pygmaeus A. Gray, Hook. J. Bot. Kew Gard. Misc. 3:148 (May 1851) (‘Skirrophorus’). — Styloncerus pygmaeus (A. Gray) Kuntze, Rev. Generum PI. 367 (1891). Type: “South-western Australia, Drummond.” Lectotype (here designated): Drummond 59, S.W. Australia, s. dat. (K). Isolectotypes: GH (ex herb. Klatt), MEL 541610, NSW, PERTH (see note 1 below). Skirrhophorus mucronulatusTuicz., Bull. Soc. Imp. Naturalistes Moscou 24 (2):72 (Oct. 1851). Type: “Nova Hollandia. Drum.v.n.59.” Holotype: ?CW, n.v. (see p.l52) IsoTYPEs: GH, K. MEL 54160, NSW, PERTH. Annual herb. Major axes usually prostrate or decumbent, rarely ascending or erect, c. 0. 5-6(9) cm long, variably hairy; stem sometimes simple and often ± lacking, but usually forming major branches at basal nodes. Leaves alternate or opposite, ± n^rowly elliptic or ± linear, sometimes semi-succulent, c. 0.3-1 cm long, c. 0.1 cm wide, mucronate, glabrous or slightly hairy. Compound heads broadly depressed to depressed ovoid, c. 0.2-0.4 cm long, 0.2-0.6(l) cm diam.; bracts subtending compound heads forming a conspicuous involucre c.Va or about the length of the head, of c. 5-10 leaf-like bracts, ± elliptic or ovate, 0.3-0.5 cm long, 0.1-0.3 cm wide, often with a small hyaline margin, mucronate, variably hairy, a few inner ones with hyaline apices and grading into capitula-subtending bracts; general receptacle convex. Capitula (4)15-50(c.70) per compound head; capitulum-subtending bracts 1, ± obovate or ± oblong, 1.7-2.4 mm long, 0.7-1.5 mm wide, ± white, the midrib glabrous or slightly hairy toward the apex. Capitular bracts with the two outer concave ones 1.6-2. 2 mm long, ± white, the midrib glabrous or sparsely hairy toward the apex; flat bracts 2, obovate, ± gradually tapering toward the base, 1. 6-2.2 mm long, 0.6-1 mm wide, ± white, the midrib glabrous or sparsely hairy toward the apex and with an entire wing-like extension from the adaxial surface. Florets 2; corolla (?4)5-lobed, the tube tapering gradually to a sometimes variably swollen base, 0.9-1. 3 mm long, c. 0.5 mm diam. Achenes ± obovoid, 0.5-0.7 mm long, c. 0.2-0.3 mm diam., variably papillose and often with a fringe of papillae at the apex. Pappus absent. 176 Distribution (Fig. 2): Restricted to the salt lakes of the Avon River System, Western Australia (Short 1981a, b). Uncommon. Ecology: Appears to grow exclusively in sandy soil on the margins of saline depressions. Commonly associated with species of Halosarcia and Disphyma. Notes: 1. The lectotype sheet of Skirrhophorus pygmaeus contains drawings of the species which, according to Gray (1851), were to be illustrated in leones Plantarum. This did not eventuate. The sheet is also clearly inscribed with the words ''Skirrophorus pygmaeus n.sp.” in Gray’s hand. It is possible that the sheet could be regarded as the holotype as there is no clear indication that Gray saw any of the duplicates. 2, As pointed out under the respective species A. pygmaeus has close affinities with A. preissianus and A. drummondii and, in particular, to a variant of A. drummondii. Specimens Examined: Western Australia — Chinnock 4158, c. 3.5 km W. of eastern edge of Lake King, 26.ix.1977 (AD); Chin nock 4359, Eclipse Lake, ll.xi.l978 (AD); Chinnock 4366, small salt pan 0.7 km beyond western edge of Lake King, 12. xi. 1978 (AD, PERTH); Gardner s.n., Mortlock River flats, E. of Meckering, 22.x. 1945 (PERTH); Pritzel 902, Avon district, -.xi.l901 (NSW); Short 617, 3.4 km E. of Meckering in Mortlock River, 20.ix.l977 (AD); Short 674, 1 km E. of Wave Rock, 25. ix. 1977 (AD); Wilson 6386a, 3 km E. of Meckering, 23.xi.1967 (PERTH). 14. Angianthus preissianus (Steetz) Benth., FI. Austr. 3:566 (1867); K, Hoffman in Engler & Prantl., Naturl. Pflanzenfam. 1V5:194, fig. 98A (1890); J. M. Black, FI. S. Aust. 1st ed. 645 (1929), 2nd ed. 924 (1957); W. M. Curtis, Stud. FI. Tas. 344 (1963); Willis, Handb. PI. Viet. 2:730 (1973); Grieve & Blackall, W. Aust. WildHs 814 (1975). — Skirrhophorus preissianus Steetz in Lehm. PI. Preiss. 1:439 (1845). — Styloncerus preissianus (Steetz) Kuntze, Rev. Generum PI. 367 (1891). Type: “In umbrosis madidis inter frutices prope lacum ad Woodman’s point, mense Dec. 1838. Herb. Preiss. No. 38.” Lectotype (here designated): Preiss 38, In Nova Hollandia, (Swan-River Colonia) in umbrosis madidis inter frutices prope lacum ad Woodman’s point, s. dat. (MEL 541608, ex herb. Steetz). Isolectotypes: LD, MEL 541609, S (see p.l52). Skirrhophorus eriocephalus Hook. f. ex. A. Gray, Hook. J. Bot. KewGard. Misc. 3:148 (1851) (Hook. f. in MSS); Hook, f., FI. Tas. 1:198, pi. 53A (1856). — Angianthus eriocephalus (Hook. f. ex A. Gray) Benth., FI. Austr. 3:567 (1867); W. M. Curtis, Stud. FI. Tas. 344 (1963). — Styloncerus eriocephalus (Hook. f. ex A. Gray) Kuntze, Rev. Generum PI. 367 (1891). Type: “Georgetown, Van Diemen’s Land, Gunn.” Lectotype (here designated): Gunn 1973, George Town, 21. xi. 1842 (K). Isolectotypes: HO, NSW, NSW p.p. (lacks collector’s no. but cites Georgetown and the dates 21.xi.42& lO.i.43, i.e. a mixed collection). Possible Isolectotypes: GH (several collections ex herb. Hook. f. but each lacks collection date, collector’s no. and gives the location only as Tasmania or “VDL”). Annual herb. Major axes erect to prostrate (0.5)4-10(16) cm long, glabrous or variably hairy; stem often simple in the smaller, erect plants, sometimes ± lacking (less than c. 1 cm high) in the prostrate ones, but usually forming major branches at basal and/or upper nodes. Leaves alternate or opposite, usually ± narrowly elliptic or ± linear, sometimes semi-succulent to sucedent and ± terete, 0. 5-1(1. 2) cm long, c. 0. 1-0.2 cm wide, mucronate, variably hairy. Compound heads broadly ovoid to depressed ovoid, 0.4-0.8(l) cm long, 0.4-0.7(l) cm diam.; bracts subtending compound heads forming a conspicuous involucre about the length of the head, of c. 15 bracts, the outer ones leaf-like, ± elliptic, or ovate to lanceolate, 0.5-1 cm long, 0. 1-0.2 cm wide, variably mucronate, hairy, a few inner ones with hyaline apices and grading into capitulum-subtending bracts; general receptacle an expanded, convex axis. Capitula c. 5-100 per compound head; capitulum-subtending bracts 1(2), if more than one then the extra one abaxial to and overlapping the inner, all bracts ± obovate or ± oblong, 1.7-2. 4(2. 6) mm long, 0.7-1. 5 mm wide, ± white, the midrib glabrous or variably hairy 177 toward the apex. Capitular bracts with the two concave ones 1. 6-2.3 mm long, ± white, the midrib variably hairy toward the apex; flat bracts 2, obovate, ± gradually tapering toward the base, (1.5)1. 6-2. 3 mm long, 0.6-1 mm wide, ± white, the midrib glabrous or sparsely hairy toward the apex and usually with an entire wing-like extension from the adaxial surface. Florets 2; corolla 3(4, 5)-lobed, the tube tapering ± gradually to a sometimes variably swollen base, 0.9-1 .4 mm long, c. 0.4-0. 5 mm diam.; anthers 3(4, 5), each with c. 16-44 pollen grains. Achenes ± obovoid, 0. 5-0.8 mm long, c. 0.3 mm diam., variably papillose and often with a fringe of glandular hairs at the apex. Pappus absent. Fig. 3e. Distribution (Fig. 2): Coastal and inland regions of south western Australia, South Australia, Victoria and Tasmania. Very common. Ecology: Favours saline habitats and sandy soils. Commonly associated with species of Halosarcia, Sarcocornia and Frankenia, Notes: 1. The lectotype sheet of S', eriocephalus contains at least two collections. On the lower V 3 of the sheet there are drawings, a label and a blue envelope containing specimens which are regarded as the lectotype material. Above this there are some further specimens attached to the sheet. These may possibly belong to Mueller’s Brighton collection which is in a further envelope at the top of the sheet although they do match material in the type envelope. 2. A. preissianus exhibits a great deal of variation. Commonly collections contain specimens with erect or ascending major axes and variably hairy, non-succulent leaves. However other collections, e.g. Orchard 1439, Royce 9731 and Short 1013, are of more or less glabrous specimens with short axes and succulent leaves. Further variation may also be found in the wing-like extension on the adaxial surface of the inner capitular bracts. Usually a single membranous wing extends from the axis but in some cases, e.g. Royce 9731 and Tepper s.n. MEL 84892, two wings are present. In a further collection, Short 932, an inner wing is absent. Specimens also exhibit variation in the placement of leaves (opposite or alternate) and the number of papillae at the apex of the mature achenes. Much of the variation exhibited by A, preissianus may well be environmentally induced; certainly the habit of the plant is largely dependent on the degree of shading. Some variation may also be a reflection of the ability of inbreeders (Short, 1981a, b) to produce highly adapted biotypes. 3. Bentham (1867) and Curtis (1963), although commenting on the variation exhibited by the taxa, recognised both A. preissianus and A. eriocephalus. Due to the variation exhibited the latter species is regarded as a synonym of A. preissianus. The recognition of an infraspecific taxon is not warranted. 4. A number of authors, e.g. Bentham (l.c.), Curtis (l.c.) and Grieve and Blackall (1975), have described A. preissianus as having a small pappus. However the pappus appears to be best interpreted as a cluster of papillae at the apex of the achene. The same papillae are found over much of the fruit. 5. A. preissianus is very closely related to A. pygmaeus, A. drummondii and the variant of A. drummondii. Indeed all taxa may, under certain conditions display a similar habit to A. preissianus and the taxa can then only be discerned on microscopic characters. A. preissianus is best distinguished from all related taxa by its 3- or rarely 4-lobed florets which reflect the inbreeding nature of the species (Short, 1981a, b). A. drummondii, the variant of A. drummondii and A. pygmaeus all have 5-lobed florets. Selected Specimens Examined (14/158): Western Australia — Drummond 122, W.A., s. dat. (MEL, NSW); Eichler 20305, c. 30 km NNE of the coast at Stokes Inlet, 18.x. 1968 (AD, PERTH); Orchard 1439, c. 38 km N. of the coast at Stokes Inlet 10.x. 1968 (AD, VEKlWy, Royce 9731, E. of boundary of Watheroo National Park, 7.x. 1971 (PERTH)* Short 932, c. 47 km E. of Yellowdine, 13.xi.l979 (AD); Short 1013, c. 14 km from Jurien along main road to 178 Badgingarra, 19.xi.l979 (AD); Short 1052, saline flat running into Leschenault Inlet, c. 3 km from Bunbury, 22.xi.1979 (AD). South Australia — Martinsen 60, Mambray Creek, 12. ix. 1974 (AD); Short 716, 8.6 km S. of Corny Point Lighthouse, 9.ix.l977 (AD); Short 800, c. 10 km south of Streaky Bay, 26.ix.1978 (AD); Tepper sm.. Kangaroo Island, 1886 (MEL 84892). Tasmania — Rodway s.n.. River Derwent, 3.xii.l899 (NSW 138738); Whinray 221, Cape Barren Island, 3.xi.l973 (AD). Victoria — Morrison s.n. , Port Melbourne, 7.xii.l892 (BRI 078641, MEL 225623, PERTH). 15. Angianthus cunninghamii (DC.) Benth., FI. Austr. 3:565 (1867); Grieve & Blackall, W. Aust. Wildfls 815, pi. 13 (1975). — Skirrhophorus cunninghamii DC., Prod. 6:150 (1838); DC. inDeless., Icon. Select. PI. 4:22, t.51 (1840); SteetzinLehm. PI. Preiss. 1:438 (1845); A. Gray, Hook. J. Bot. Kew Gard. Misc. 3:148 (1851). — Styloncerus cunninghamii (DC.) Kuntze, Rev. Generum PI. 367 (1891). Type: ‘Tn arenosis insulae anglis dictae Dirk Hartog’s ad oram occid. Australiae januario flor. legit cl. A. Cunningham.” Holotype: Cunningham s.n., Sandy downs, Dirk Hartog's Island. West Coast Australia, -.i.l822 (G in herb. DC., ex microfiche IDC), Isotypes: K (excl. illus., ex herb. Allan Cunningham), MEL 541221 (see note 1 below). Eriocladium pyramidatum Lindl., Edwards’ Bot. Reg.: Swan River Append. 24 (1839). Holotype: Toward 15, Swan River, s. dat. (CGE, herb. J. Lindley), (see note 2 below). Perennial shrub, 20-50 cm high. Major axes ± erect and densely hairy. Leaves alternate, often recurved, oblanceolate or ovate, 0.5-2(2.6) cm long, 0.2-0.3 cm wide, densely hairy. Compound /zeals' broadly to broadly depressed ovoid, 0. 5-0.9 cm long, 0.45-0.8 cm diam.; bracts subtending compound heads forming a conspicuous involucre extending c. Vi the length of the head, of c. 20 bracts, the outer ones leaf-like, ± ovoid, 0.2-0. 3 cm long, 0. 1-0.15 cm wide, densely hairy, the inner ones with hyaline appendages and grading into capitulum-subtending bracts; general receptacle ovoid to very broadly ovoid, c. 2-3 mm long, c. 2 mm diam. Capitula c. 25-50 per compound head; capitulum-subtending bract 1, obovate to ± oblanceolate, sometimes ± narrowly oblong to oblong, (2.6)3. 1-3. 8(4,1) mm long, (1)1.2-1.5(1.65) mm wide, with the upper part of the lamina yellow and with a prominent constriction, the midrib usually sparsely hairy toward the apex and some glandular hairs always present. Capitular bracts with the two concave ones (2.3)2.9-3.5(3.7) mm long, with the upper part of the lamina yellow and with a prominent constriction, the midrib usually with a few glandular hairs; flat bracts 2, oblanceolate or ± narrowly oblong, gradually tapering to the base, (2.8)3-3.6(3.75) mm long, (0.6)0.7-l(1.2) mm wide, the lamina with a prominent constriction in the upper part, the midrib usually with a few glandidar hairs. Florets 2 (3); corolla 5-lobed, the tube tapering ± gradually to the base which is distinctly swollen in mature florets, 2-2.5 mm long, c. 0.5 mm diam., glandular hairs often present. Achenes ± obconical, 0.9-1.4 mm long, 0.5-0. 6 mm diam., papillose. Pappus absent. Distribution (Fig. 2): Western coastline of Australia between latitudes 2(FS and 32°S. Common. Ecology: Commonly grows in the unconsolidated calcareous sands of coastal foredunes but also grows in saline flats. Collectors’ notes include “Low salt flats with mangrove and Salicornia” and “Growing on unconsolidated foredunes”. Notes: 1. The sheets referred to as isotypes of 5. cunninghamii have slightly different wording. On the K sheet there is a reference to “sandy plains” rather than “sandy downs” as on the holotype. The MEL sheet has the words “Frequent on desert plains of sand”. Despite these discrepancies both probably can be regarded as isotypes although the number “288” which also appears on the MEL label suggests that this may not be correct. 2. Lindley (1839) based his descriptions of new species from the Swan River Colony on specimens he obtained from Drummond, Mangles, Toward and Ward. No particular 179 collection was specified for Eriocladium pyramidatum but Toward 15 is the only one of the species in Lindley's herbarium and is thus regarded as the holotype. Drummond also collected this species, i.e. Drummond 125 (MEL, GH), Drummond ? 159 ox 245 (GH ex herb. Klatt) and Drummond s.n. (MEL 541216), but there is no reason to believe that Lindley saw any of these collections. 3. A, cunninghamii is the only perennial species of Angianthus and the only one to occasionally produce 3 florets per capitulum. Both characters suggest that the species is a somewhat primitive member of the genus. Selected Specimens Examined (6/36): Western Australia — Allender s.n.. Shark Bay, 18. ii. 1969 (UWA 2493); Demarz 2890, Vlaming Head, 4.xi.l970 (PERTH); Kenneally 1014, Dorre Island, 15.xii.l973 (PERTH); Morrison s.n., Claremont, 28.iii.1900 (BRI 086974); Serventy s.n., Bernier Island, 5.viii.l947 (PERTH); Serventy s.n., Stewart Island, s. dat. (PERTH). 2. Pleuropappus F. Muell., Trans & Proc. Viet. Inst. Advancem. Sci. 37 (1855). Type: P. phyllocalymmeus F. Muell. [Angianthus auct. non Wendl.: see synonymy of P. phyllocalymmeus.] [Styloncerus auct. non Spreng., nom. illeg. \ see synonymy of P. phyllocalymmeus.] Annual herb. Major axes erect or ascending, sometimes decumbent, hairy; stem simple or forming major branches at basal and/or upper nodes. Leaves mainly alternate but opposite at the base of the stem, sessile, entire, linear, mucronate, hairy. Compound heads narrowly ellipsoid or lanceoloid to ovoid; bracts subtending compound heads forming a conspicuous involucre c. the length of the head, the outer ones leaf-like, the inner ones with hyaline apices; general receptacle cylindrical to narrowly oblong, consisting of a single major axis lacking minor receptacular axes, the individual capitula distributed ± evenly along its entire length. Capitula 40-100 per compound head, each capitulum with 4 (5, 6) abaxial, hyaline subtending bracts that overlap the inner capitular bracts. Capitulum-subtending bracts arranged so that an outer bract covers 2 middle bracts which in turn cover a single inner bract, sometimes 1-2 additional bracts covering the inner 4, all bracts flat, ovate or elliptic; midrib usually conspicuous, opaque, c. V^-Vi the length of the bract, variably hairy. Capitular bracts 4, hyaline, with an opaque midrib, arranged so that 2 outer concave bracts surround 2 inner flat bracts. Concave bracts with the midrib conspicuous, c. Vi the length of the bract, variably hairy. Inner flat bracts obovate, abruptly attenuated in the lower V^-V ^\ the midrib conspicuous, c. Vi the length of the bract, glabrous or hairy. Florets 2 per capitulum; corolla 5-lobed; style branches truncate; stamens 5, with tailed anthers. Achene obliquely attached to the floret, ellipsoid, papillose. Pappus an oblique jagged scale. Fig. Ih. Distribution (Fig. 8): A monotypic genus confined to southern Eyre Peninsula and southern Yorke Peninsula. It is poorly collected and Jessop (1977) recorded P. phyllocalymmeus as an endangered species. However field observations suggest that the species, although geographically restricted, is locally common. Affinities/Generic Characteristics: Pleuropappus phyllocalymmeus superficially resembles many species of Angianthus with similar shaped compound heads. Furthermore, there are 4 capitular bracts which are arranged in the same manner as those in Angianthus. However Pleuropappus is readily distinguished by the presence and arrangement of 4 or more capitulum-subtending bracts, by the obliquely attached achenes and by the absence of minor receptacular appendages on the general receptacle of the compound head. Evolution/Reproductive Biology: Although accurate determinations of pollen-ovule ratios (P/Os) have not been made it is apparent that a P/O value of several thousand will be found in this species. Such a value suggests that the species commonly cross-pollinates (Short, 1981a, b). Ants have been observed on flowering compound heads and are possibly important pollen vectors. 180 Fig. 8. Distribution of Pleuropappus phyllo- calymmeus (South Australia), Cephalo- sorus carpesioides and Epitriche demissus (Western Australia). Pleuropappus phyllocalymmeusR Muell., Trans & Proc. Viet. Inst. Advancem. Sci. 37 (1855). — Angianthus phyllocalymmeus (F. Muell.) Druce, Bot. Soc. Exch. Club Brit. Isles 4:604 (1917); Domin, Mem. Soc. Sc. Boheme 2:121 (1923) i!;phyllocalymneus')\ J. M, Black, FI. S. Aust. 1st ed. 645 (1929), 2nd ed. 924 (1957); Hj. Eichl., Suppl. to J. M. Black’s FI. S. Aust. 326 (1965). — Angianthus pleuropappus Benth., FI. Austr. 3:563 (1867) nom, illeg. — Styloncerus phyllocalymmeus (F. Muell.) Kuntze, Rev. Generum PI. 367 (1891) phyllocalymneus^). Type: “On sterile plains of the Port Lincoln district. — C. Wilhelmiy Lectotype (here designated): Wilhelmis.n,, Port Lincoln, s. dat. (K). Probable Isolectotypes: MEL 541617-541619, MEL 84469 (see note 1). Annual herb, 4-8(15) cm high. Leaves 0. 7-1(1. 3) cm long, c. 0.1 cm wide. Compound heads O.S-l,5{2) cm long, c. 0. 3-0.5 cm diam.; bracts subtending compound heads c. 10, the outer ones leaf-like, narrowly elliptic or lanceolate, 0.5-1 cm long, 0.1-0.15 cm wide, ± mucronate, hairy, the inner ones with hyaline apices and grading into capitulum-subtending bracts. Capitula 40-100 per compound head; capitulum- subtending bracts ovate or elliptic, 1. 8-2.2 mm long, 1-1.2 mm wide. Capitular bracts with the two outer concave ones c. 2 mm long; flat bracts abruptly attenuated in lower ‘A- Vi and the edges sometimes incurved so as to slightly cover the florets, 2-2.3 mm long, 0.9-1. 3 mm wide. Florets 2; corolla 5-lobed, the tube usually tapering gradually to the base but sometimes an abrupt taper occurring in the lower Vs, 1. 3-1.7 mm long, c. 0.5 mm diam. Achene obliquely attached to the floret, ellipsoid, 0.7-0.8 mm long, 0.3-0.4 mm diam., papillose. Pappus an oblique jagged scale about the length of the corolla tube. Distribution: See generic treatment. Ecology: Grows exclusively in sandy or clay loam on the margins of saline depressions. Associated with Halosarcia. Note: 1. Following his description of P. phyllocalymmeus Mueller (1855, p.37) cited a single collection, “On sterile plains of the Port Lincoln district. — C. WilhelmiP None of the Wilhelmi collections from MEL & K are designated in this manner but a K collection is recorded as coming from “Port Lincoln”. 181 Because of confusion with the labels of the MEL collections (see annotations on the sheets), the K material, which contains 2 individual specimens in good condition, has been designated as the lectotype. The same sheet dso contains Wilhelmi material designated as coming from “between the Fountain & Long Lake” but this material has been clearly separated from the lectotype. A further label “Victoria, South Australia, July 26/55, Mueller” occurs on the sheet but both the location and the name, ''Chrysocoryne tenella Muell.” ( = C. drummondii A. Gray) suggests that it has been erroneously placed with this material. Selected Specimens Examined (6/13): South Australia — Alcock 2801, Lower Eyre Peninsula, Hundred of Lake Wangary, 14.x. 1969 (AD, CANB); Cleland s.n,. Coffin Bay Reserve, I0.xi.l960 (AD 96404182); Lang 1082, c. 33.7 km WNW. of Cummins on road to Mt. Hope, 20.x. 1977 (AD); Short 806, c. 34 km NW. of Cummins on road to Mt. Hope, 26. ix. 1978 (AD); Short 822, c. 13.5 km W. of Yorketown along main Warooka road, 28.x. 1978 (AD); Wilhelmi s.n.. Lake Greenly, 1855 (NSW 138697). 3. Epitriche Turcz., Bull Soc. Imp. Naturalistes Moscou 24(2):74 (Oct. 1851). Type: E. cuspidata Turcz. (=E. demissus (A. Gray) Short) Skirrhophorus DC. in Lindl. ex DC. sect. Psuedopappus A. Gray, Hook. J. Bot. Kew Gard. Misc. 3:149 (May 1851). Type: S. demissus A. Gray {=E. demissus (A. Gray) Short) [Angianthus auct. non Wendl.: see synonymy of E. demissus.] [Styloncerus auct. non Spreng.: see synonymy of E. demissus.] Annual herb. Major axes erect, glabrous or sparsely hairy; stem simple or forming major branches at upper nodes. Leaves opposite, sessile, the base ± stem clasping and with hyaline margins, the entire leaf glabrous or sparsely hairy. Compound heads broadly depressed ovoid; bracts subtending compound heads forming a conspicuous involucre c. equal to or longer than the head; general receptacle an entire, convex, ± smooth axis, the capitula distributed evenly over its surface. Capitula c. 10-20 per compound head. Capitular bracts 2 or 3 , hyaline, ± flat to concave, with a conspicuous, sparsely hairy midrib extending c. Vi the length of the bract, the bracts overlapping one another. Florets 1 per capitulum; corolla 5-lobed; style branches truncate; stamens 5, with tailed anthers. Achenes ? ± obconical and papillose, the apex beset with long hairs. Pappus absent. Distribution (Fig. 8): A monotypic genus endemic to the south-west of Western Australia. Known only from the type collection and Wilson 8314. Affinities/Generic Characteristics: The lack of collections has made it difficult to ascertain certain characteristics of this genus and the full range of variation exhibited by the species is unknown. For example characteristics of the achene are difficult to ascertain and the number of capitula per compound head has been estimated for only 2 or 3 individuals. At least superficially the genus appears to be allied to Angianthus s.str. However the apparent lack of minor receptacular appendages, the absence of capitulum-subtending bracts and the distinctive ring of hairs at the apex of the achene all suggest that the genus should be reinstated. There is some doubt whether or not the hairs at the apex of the achene should be regarded as a pappus (see morphology section). Epitriche demissus (A. Gray) Short, comb. nov. Skirrhophorus demissus A. Gray, Hook. J. Bot. Kew Gard. Misc. 3:149 (May 1851), basionym. — Angianthus demissus (A. Gray) Benth., FI. Austr. 3:567 (1867); Greive & Blackall, W. Aust. Wildfls 815 (1975). — Styloncerus demissus {A. Gray) Kuntze, Rev. Generum PI. 367 (1891). Type: “South-western Australia, Drummond, 1850.” Lectotype (here designated); Drummond 58, S.W. Australia, 1850 (K) (label in Gray’s hand, plus drawings). Isolectotypes: GH (ex herb. Klatt), K (ex herb. Benth.), KW, MEL 541627, MEL 84428, NSW, PERTH (2 sheets). 182 Epitriche cuspidata Turcz., Bull. Soc. Imp. Naturalistes Moscou 24(2):75 (Oct. 1851). Type: “Nova Hollandia. Drum.V.n.58.” Possible Holotype: KW (see p.l52). Isotypes: GH (ex herb. Klatt), K (2 sheets), MEL 541627, MEL 84428, NSW, PERTH (2 sheets). Annual herb, 2-5 cm high. Leaves ± lanceolate, 0.5-1 cm long, c. 0.3 cm wide. Compound heads 0.3-0.5 mm long, 0.4-0.6 cm diam.; bracts subtending compound heads c. 10-15, the outer ones leaf-like, glabrous or sparsely hairy, the inner ones densely hairy. Capitula c. 10-20 per compound head. Capitular bracts oblanceolate, 2-2.8 mm long, c. 0.5-0.7 mm wide. Florets 1; corolla 5-lobed, the tube bulb-like at the base, 1.7-1. 9 mm long, c. 0.3 mm diam. Achenes ? obconical and papillose, the apex beset with long hairs which are c. Vi-V^ the length of the floret. Distribution: See generic treatment. Ecology: Recorded growing in clay soil {Wilson 8314). No other information available. Specimens Examined: Western Australia — Wilson 8314 , c. 5 km S. of Three Springs, [c. 29^2'S, 115°46'E], 25.vii.1969 (PERTH). 4. Cephalosorus A. Gray, Hook. J, Bot. Kew Gard. Misc. 3:98 (April 1851), 152 (May 1851). Lectotype (here designated): C. phyllocephalus A. Gray { = C. carpesioides (Turcz.) Short). Piptostemma Turcz., Bull. Soc. Naturalistes Moscou 24 (1):191 (March 1851) nom. illeg., [later homonym of Piptostemma Spach., Hist. Veg. Phan. 10:34 (1841).] Type: P carpesioides Turcz. ( = C. carpesioides (Turcz.) Short) [Angianthus auct. non Wendl.: see synonymy of C. carpesioides.] [Styloncerus auct. non Spreng., nom. illeg.: see synonymy of C. carpesioides.] Annual herb. Major axes erect, variably hairy; stem distinct, hollow, simple or with opposite branching from upper nodes. Leaves opposite or alternate, petiolate or sessile, entire, lamina ± elliptic or oblanceolate to obovate, variably hairy. Compound heads broadly depressed to depressed ovoid; bracts subtending compound heads forming a conspicuous involucre Vi to c. the length of the head; the outer ones leaf-like, the inner ones hyaline toward the base, all bracts glabrous to variably hairy; general receptacle an entire, broadly depressed ovoid axis, the capitula sessile and distributed evenly over the surface. Capitula c. 30-60 per compound head. Capitular bracts 3 or 4(5), hyaline, ± flat, or concave, the laminae rarely with a distinct constriction in the upper part, the bracts ± overlapping each other; the midrib ± conspicuous and extending c. V 2 the length of the bracts, variably hairy at or near the apex. Florets 1 per capitulum; corolla 5-lobed; style branches truncate; stamens 5, with tailed anthers. Achenes ± obovoid, with a cellular, diaphanous pellicle. Pappus a jagged cup. Distribution (Fig. 8): A monotypic genus confined to the south-west of Western Australia between latitudes 28°S and c.31°S and west of longitude \\6°E. Nomenclatural Problems: 1. Gray (1851) described two species, namely C, phyllocephalus and C. gymnocephalus, in his new genus Cephalosorus. The generic description supplied covers the major characteristics of both species and there is nothing to suggest that either one should be given preference when selecting a lectotype. C. gymnocephalus clearly differs from C. phyllocephalus and is allied to species in the genus Gnephosis s.l., to which Bentham referred the species in 1867. Thus it is convenient to designate C. phyllocephalus as the lectotype species of Cephalosorus. C. gymnocephalus is excluded from the genus. Affinities/Generic Characteristics: Cephalosorus has no obvious affinities with other segregate genera of Angianthus. 183 It is readily distinguished by the opposite, petiolate leaves which occur in at least the lower half of the plant. Achene morphology and the morphology, number and arrangement of capitular bracts are unique. Cephalosorus carpesioides (Turcz.) Short, comb. nov. Piptostemma carpesioides Tmvqz., Bull. Soc. Naturalistes Moscou 24(1):192 (March 1851), basionym. Type: “Nova Hollandia. Drum. coll. IV. n. 200.” Possible Holotype: KW (see p.l52). Isotypes: GH (ex herb. Klatt), K, MEL 541595, MEL 541596. Cephalosorus phyllocephalus A. Gray, Hook. J. Bot. Kew Gard. Misc. 3:152 (May 1851). — Angianthus phyllocephalus {A. Gray) Benth., FI. Austr. 3:565 (1865); Grieve & Blackall, W. Aust. Wildfls 812 (1975). — Styloncerus phyllocephalus (A. Gray) Kuntze, Rev. Generum PI. 367 (1891). Type: “Swan River, Drummond, 1846, 1848.” Lectotype (here designated): Drummond 200, S.W. Australia, 1848 (K). Isolectotypes: GH (ex herb. Klatt), MEL 541595, MEL 541596 (see note 2 below). Cephalosorus brevipapposus F. Muell., Fragm. 3:159 (1863). — Skirrhophorus phyllocephalus F. Muell., l.c., pro syn., (? as to collections of F. Muell.). Type: “Ad flumen Murchison; Oldfield. Ad sinum Champion Bay; Walcott'' Lectotype (here designated): Oldfield s.n., Murchison R., W.A., s. dat. (MEL 541597). Probable Isolectotype: PERTH (ex MEL, referred to as Angianthus phyllocephalus on label). Syntype: None seen, the only specimens of this species seen from Champion Bay were collected by Oldfield. No Walcott specimens of the species have been seen. Annual herb, 15-25(29) cm high. Leaves opposite and distinctly petiolate in at least the lower half of the plant, the uppermost ones frequently ± sessile and alternate; petiole ± absent to c. 2 cm long, variably hairy; laminae ± elliptic or oblanceolate to obovate, 1-2. 5(3. 4) cm long, 0.4-l(1.3) mm wide, sometimes with a very small mucro at the apex, almost glabrous (particularly the lower surface) to densely hairy. Compound heads 0.5-1. 4 cm high, 0.7-1. 5 cm diam.; bracts subtending compound head c. 10-20, the outer ones ± ovate or ± obovate, 0. 5-1(1. 4) cm long, 0. 3-0.8 cm wide. Capitulac. 30-60 per compound head. Capitular bracts 3. 3-4.2 mm long, (0. 7)1-1. 8 mm wide. Florets 1; corolla tube with a conspicuously swollen base, the tube 1.5-2 mm long, 0.5-0.8 mm diam. Achenes ± obovoid, 1.9-2.5 mm long, 0.9-1 mm diam. Pappus a jagged cup c. 0.7 mm long. Distribution: See generic treatment. Ecology: Little information is available. Collectors’ notes include “Common on rocky ironstone knoll” and “Ironstone gravel”. Note: 1. The lectotype sheet of C. phyllocephalus contains three good, entire specimens, plus drawings of the species. According to Gray (1851) the species was to be illustrated in Incones Plantarum but this did not eventuate. A label attached to the sheet has the words ''Cephalosorus phyllocephalus n. gen.” in Gray’s hand. Specimens Examined: Western Australia — Alpin 56, 1-2 miles North of Carnamah, 4.ix.l958 (PERTH); Burns 24, Port Gregory road, 20. ix. 1970 (PERTH); Gardner 12831, Arrino, 27. ix. 1960 (PERTH); ?Mueller s.n.. Port Gregory, -.x.1877 (MEL 84472); ?Mueller s.n., upper Irwin River, s. dat. (MEL 84473); Oldfield s.n.. Champion Bay, s. dat. (MEL 84471). Paust 1267, 1 mile N. of Northampton-Port Gregory road on Yerina Springs road, 6.X.1972 (PERTH); Wilson 3829, 15 km N. of Badgingarra, 2.ix.l965 (AD, GH, PERTH). (To be continued in Muelleria 5(3): 185) Muelleria Volume 5, Number 3 January to August, 1982 CONTENTS Page A revision of Angianthus Wendl., sensu lato (Compositae: Inuleae; Gnaphaliinae), 2 — P.S. Short 185 Storckiella australiensis sp. nov. (Caesalpiniaceae) fromnorthern Queensland: a new generic record for Australia — J. H. Ross and B. P. M. Hyland 215 Aspienium terrestre and two Asplenium hybrids: new fern records for Australia — P. J. Brownsey 219 Grevillea montis*? Fig. 11. C. multiflora Short. Holotype {Chinnock 4411 & Wilson^ AD). 195 [Chrysocoryne pusilla auct. non (Benth.) End!.: Endl., Bot. Zeitung (Berlin) 1:457 (1843) (see note under generic treatment of Chrysocoryne).] Annual herb, 3-5(c. 6) cm high. Major axes erect with scale-like glandular hairs; stem simple or forming major branches at basal and/or upper nodes. Leaves opposite at the base, the upper ones alternate, all leaves ± oblanceolate to obovate or narrowly elliptic to elliptic, 0.2-0.8(l.l) cm long, 0.04-0. 08(c. 0.1) cm wide, a small hyaline appendage sometimes present at the apex, all leaves densely covered in scale-like glandular hairs. Compound heads narrowly oblong, c. 0.5-2(2.5) cm long, c. 0.2-0.25(0.35) cm diam., with a single head occurring at the apex of an unbranched major axis or several occurring on minor axes which branch from the upper nodes of a major axis. Capitula c. 40-150 per compound head; capitulum-subtending bracts widely elliptic to ± oblate or widely ovate to widely depressed ovate, (1.65)1.7-2.1(2.25) mm long, (1.8)2-2.5(2.65) mm wide; midrib entire, variably hairy, often with a few scale-like glandular hairs. Capitular bracts!, concave, (1.2)1.3-1.65 mm long, 0.5-0.85 mm wide, the upper margins variably dilate, the hairs c. 0.1 mm long; midrib inconspicuous. Florets (1)2(4) per capitulum; corolla 3, 4 (5)-lobed, the tube tapering ± gradually to a thickened base, 0.65-0.8 mm long, 0.25-0.35 mm diam., sometimes with a few glandular hairs on the lower Vi; anthers 3, 4 (5), each with c. 20-60 pollen grains. Achenes ± obovoid, 0.4-c. 0.5 mm long, 0.25-c. 0.3 mm diam., papillose, purplish. Pappus absent. Figs: 9; lOi-j. Chromosome number: not known. Distribution (See Short 1981a, fig. 4): South-west Western Australia, southern South Australia and south-west Victoria. Locally common. Ecology: Found in both coastal and inland situations around salty depressions and granite outcrops or in open woodland. Collectors’ notes include “Open sites between shrubs, in Wandoo woodland or in shrubby areas”, “In open woodland with Eucalyptus cladocalyx as dominant. Greyish sand.”, “Growing in granitic depressions with very small annual grasses, composites etc. Sandy loam”, “Growing on white to greyish sand amongst Carpobrotus, Arthrocnemum [ =Halosarcia] & Melaleuca'' and “Near sea on granite rock”. Notes: 1. Gray (1851) realised that a figure published by W. J. Hooker in leones Plantarum and referred to as Crossolepis pusilla was in fact not of that species. Consequently he referred the figure and specimens collected by Drummond to Chrysocoryne drummondiL In their publications both Gray and Hooker failed to indicate Drummond’s collection number for specimens from the Swan River. At K there is a single sheet bearing apparently two distinct collections. The upper p^t of the sheet contains specimens and the label, '^Crossolepis pusilla. Hugel. Swan River, 16, DrummoncT . The label appears to be in Hooker’s hand. The lower part of the sheet contains specimens clearly designated as "Chrysocoryne DrummondiiGxec^" This label appears to be in Gray’s hand. The original drawings of 'Crossolepis pusilla' published in leones Plantarum are attached to the left hand side of the sheet. From Gray’s information there seems no reason to assume that he did not examine the upper specimens and in any case it would appear that they are the ones used by Hooker in his illustration which was directly referred to by Gray. Thus the Drummond 76 collection has been designated as the lectotype. The lower most collection is regarded as a syntype or, in the event that the specimens are from the one gathering, as an isolectotype. Specimens in MEL, i.e. MEL 541601 with the label “W. A., J.Dr.” and MEL 84756 herb. O. W. Sonder) with the label “Chrysocoryne Drummondii A. Gray, Swan River-Drummond 1844. n.356'\ are also considered to be syntypes or, as in the former case a possible isolectotype, because of the lack of, or Gray’s failure to cite a collector’s number. 2. The only Wilhelm collection of C. tenella accompanied by locality information 196 which agrees precisely with Mueller’s type citation is in K. The accompanying label indicates that the specimens were seen by Mueller and this sheet is selected as the lectotype. Although not correctly designated a further collection, MEL 541620 with the label ''Chrysocoryne tenella Ferd. Muller, Port Lincoln, Wilhelmi” is probably an isoiectotype or syntype. In Mueller’s case it is not uncommon to find that collection details accompanying specimens do not agree entirely with the published information. In this case there would appear to be a major discrepancy in the locality details but similar labels have been found on possible type material of Pleuropappus phyllocalymmeus. 3. Despite its wide distribution C. drummondii exhibits little morphological variation although floret number has been observed to differ in some collections. Most collections have 2-flowered capitula but some collections with 3- and/or 4-flowered capitula (e.g. Cronin MEL 84705) and apparently 1-flowered capitula (e.g. Andrews, PERTH s.n.) have been found in Western Australia. A further collection, Eichler 20312, also from Western Australia, contains specimens with longer, narrower compound heads than those normally found in the species but other attributes suggest that it is best referred to C. drummondii. Selected Specimens Examined (9/73): Western Australia — Andrews s.n., Cannington, -,x.l902 (PERTH); Burbidge 7892, Dryandra State Forest, 22.xii.1971 (CANB, PERTH); Short 943, Yorkrakine Granite Rocks, 13.xi.l979 (AD); Short 1085, c. 8.6 km W. of Lake Grace, 24. xi. 1979 (AD); Short 1110, western margins of Lake Gilmore, 26.xi.l979(AD); Wilson 10,009, c. 1 km W. of Lucky Bay, 30.ix.l970 (PERTH). South Australia — Hunt 414, c. 25 km north-west of Naracoorte, 18.xi.l961 (AD); Short 807, c. 15.2 km from Edilillie along main road to Pt. Lincoln, 26. ix. 1978 (AD). Victoria — Phillips 406, between Apsley & Booroopki, 2.xi.l971 (CBG). 4. Chrysocoryne trifida Short, sp.nov. Chrysocoryne sp. B, Short, Muelleria 4:402 (1981). Herba annua, (2)3-7 cm alta. Axes maiores ascendentes erective pilos peltiformes glandulosus ferentes; caulis interdum simplex sed plerumque ex nodis basalibus superioribusve vel ubique ramificans; ramuli maiores ipsi saepe surculos efficientes. Folia alterna, plerumque anguste elliptica oblanceolatave, raro elliptica, 0.2-0. 8 cm longa, c. 0. 1-0.2 cm lata, pilos glandulosus peltiformes ferentia. Glomeruli cylindracei usque anguste oblongi, 1-c. 4 cm longi, 0.1-0. 2 cm diametro. Capitula c. 30-100; bractea capitulum subtendens ± late elliptica obovatave, 1. 6-2.2 mm longa, 1.4-1.9 mm lata; costa varie pilosa, saltern 3 lobis distinctis. Bracteae intra capituluml, concavae, 0.8-1. 3 mm longae, 0.2-0. 3 mm latae, marginibus superioribus pilis c. 0.5 mm longis; costa conspicua, circa dimidium longitudinis bracteae altingens, glabra. Flosculi 1(2); corolla 5-lobata; antherae 5, unaquaeque pollinibus c. 350-5(X). Achenia ± obovoidea, c. 0. 3-0.4 mm longa, c. 0.2-0. 3 mm diametro, papillosa, purpurea. Pappus carens. Holotypus (fig. 12): Short 966, 45.1 km N. of Koorda along main road to Mollerin. Salt lake. c. 30°28'S, 117°31'E. Growing in white to brown sand or very sandy loam amongst Melaleuca, just Arthrocnemum [ =Halosarcia] zone. Chrysocoryne tridens and C. trifida commonly found growing together, 14.xi.l979 (AD 98(X)2348). IsoTYPus: PERTH. Annual herb, 3-7 cm high. Major axes ascending or erect, with scale-like glandular hairs; stem sometimes simple but usually forming major branches at basal and/or upper nodes. Leaves alternate, narrowly elliptic or oblanceolate, rarely elliptic, 0.2-0. 8 cm long, c. 0. 1-0.2 cm wide, a small hyaline appendage often present at the apex, all leaves with scale-like glandular hairs. Compound heads cylindrical to narrowly oblong, 1-4 cm long, 0.1-0.2 cm diam. Capitula c. 30-100 per compound head; capitulum-subtending bracts ± widely elliptic or ± widely obovate, 1.6-2. 2 mm long, 1.4-1. 9 mm wide; midrib with at least 3 distinct lobes, variably hairy, at least the lower bracts with some scale-like glandular hairs. Capitular bracts 2, concave, 0.8-1. 3 mm long, 0.2-0.3 mm wide, the upper margins with long hairs V3-V2 (c. 0.5 mm long) the length of the bract; the midrib conspicuous and c. Vi the length of the bract, glabrous. Florets 1(2); corolla 5-lobed, the tube tapering ± gradually to the base or with a fairly distinct constriction in the lower Vi, the entire tube 0.8-1 mm long, c. 0. 3-0.4 mm diam.; anthers 5, each with c. 350-500 pollen grains. Achenes ± obovoid, 0.3-0.4 mm long, c. 0. 2-0.3 mm diam., papillose, purplish. Pappus absent. Figs: 9; lOk-m; 12. Chromosome number: n = c. 11. 197 96S \ H 0 LO TYPI STATI HiRtAKIUM OF SOUTH AUSTRAUA AOKAmi ,cAhQ\eng\h of the florets, ± hyaline, whitish, with papillae at the apex. Florets 1 per capitulum; corolla 5-lobed; style branches truncate; stamens 5, with tailed anthers. Achenes ± ovoid or ± obpyramidal, covered with mucilagenous cells, brown. Pappus absent. Fig. li. Chromosome numbers: n=4, 5, 6, 7, c. 10, c. 12. The taxonomy of Pogonolepis is yet to be resolved. For comments see Muelleria 4:404-405 (Short, 1981a). Three species normally referred to Angianthus, i.e. A. lanigerus, A. muellerianus (==P. muelleriana (Sond.) Short) and A. strictus ( = P. stricta Steetz) belong to Pogonolepis. The new combination transferring A. lanigerus to Pogonolepis is made below. Pogonolepis lanigera (Ewart & J. White) Short, comb. nov. Basionym: Angianthus strictus var. lanigerus Ewart & J. White, Proc. Roy. Soc. Viet. 22:92 (1909). Synonym: Angianthus lanigerus (Ewart & J. White) Ewart & J. White, Proc. Roy. Soc. Viet. 23:288 (1911). 9. Siloxerus Labill., PI. Nov. HoU. 2:57 (1806); Less., Syn. generum Comp. 270 (1832); Ostenfeld, Biol. Meddel. Kongel. Danske Vidensk. Selsk. 3:134, p.p. (as to S. humifusus & S. filifolius only). — Styloncerus Spreng., Syst. veg. 3:356, 451 (1826), nom. illeg. — OgcerostylusCdiS,^., Diet. Sc. Nat. 49:221 (1827), nom. /7/e^. ; Stuedel, Nom. Bot. 2nd. ed. 242 (1841) {'Oxerostylus'). Type: Siloxerus humifususLdbiW. Chamaesphaerion A. Gray, Hook. J. Bot. Kew Gard. Misc. 3:176 (June 1851). Type: Chamaesphaerion pygmaeum A. Gray ( = 5. pygmaeus (A. Gray) Short). Gyrostephium Turez., Bull. Soc. Naturalistes Moscou 24(2):76 (Oct. 1851). Type: Gyrostephium rhizocephalum Turez. ( = S. pygmaeus (A. Gray) Short). [Angianthus auct. non Wendl.: see synonymy of S. humifusus & S. filifolius.] [Chthonocephalus auct. non Steetz: see synonymy of S. pygmaeus.] [Gnaphalodes auct. non A. Gray, nom, illeg., later homonym of Gnaphalodes Miller (see Hj. Eichler, Taxon 12:295 (1963): as to Gnaphalodes fdifolium Benth. {=^Siloxerus filifolius).] Annual herbs. Major axes ± absent or if present then decumbent to erect, glabrous or hairy; stem simple and minute or forming major branches at basal and/or upper nodes. Leaves in a basal rosette or, if major axes present then opposite to alternate, all leaves entire, sessile, glabrous or sparsely hairy, apex mucronate, the base often with hyaline margins. Compound heads ± ellipsoid to broadly ellipsoid or ± lanceoloid to depressed ovoid; bracts subtending compound heads conspicuous, leaf-like, at least c. *4 to Vi the length of the head, often c. equal to or exceeding the length of the head; general receptacle of a single hairy axis which lacks minor receptacular axes, the axis becoming hollow with age. Capitula ± evenly distributed over the general receptacle, ± indistinct and lacking subtending bracts. Capitular bracts c. 5-15, mainly hyaline but the uppermost portion opaque and often crenulate, with a green, ± glabrous midrib which extends c. Vi-Vi the length of the bract, the bracts arranged in ± 1 or 2 indistinct whorls. Paleae resembling capitular bracts, one bract per floret. Florets A~\5Q.2) per capitulum; corolla 3-5-lobed; style branches truncate; stamens 3-5, with tailed anthers. Achenes ± obovoid, sparsely to densely papillose, purple. Pappus of 5-7 variably jagged scales joined at the base or a jagged ring lacking distinct scales. Fig. 15. 205 0 kms 300 Fig. 15. Distribution and floret characteristics of Siloxerus. a and # — S. humifusus. b and O — S. filifolius, c and A — 5. pygmaens. Distribution (Fig. 15): South-west of Western Australia. Affinities/Generic Characteristics: Siloxerus contains 3 species. It is readily distinguished from all other members of Angianthus s.L by the presence of paleae, a hairy general receptacle, the ± obovoid, purple, papillose achenes and the more or less rigid nature of both the capitular and receptacular bracts. Unlike other members of Angianthus s.l. the capitula in members of this genus are somewhat ill-defined. The genus has no obvious affinities with other members of the ‘'Angianthus group'' (sensu Merxmuller et al., 1977) and any affinities with other members of the Gnaphaliinae are yet to be determined. Evolution/Reproductive Biology: All species, as evidenced by approximate P/O determinations of c. 200 (Short 1981a, b), are inbreeders. They have close affinities with one another and indeed 5. humifusus may be a polyploid directly derived from S. filifolius (see note 1 under S. humifusus). Key to Species of Siloxerus 1. Stem apparently absent; compound heads depressed ovoid, c. 0.4-0.6 cm long, 0.6-1. 1 cmdiam 3. 5. pygmaeus 1. Stem simple or branching, the major axes decumbent to erect, 1-7(9) cm long; compound heads ellipsoid to broadly ellipsoid or ovoid to broadly depressed ovoid, c. 0.6-2. 9 cm long, c. 0.5-1. 5 cm diam. 2. Pappusc. !/2 or rarelythelength of the floret; capitular bractsandpaleae(2)2.5-4.5(6.3) mm long 2. 5. humifusus 2. Pappus c. equal to or slightly exceeding the length of the Horet; capitular bracts and paleae 1.25-1.9 mm long I ^ filifolius 1. Siloxerus filifolius (Benth.) Ostenf., Biol. Meddel. Kongel. Danske Vidensk. Selsk 3:136 (1921). — Gnaphalodes filifolium Benth., FI. Austr. 3:578 (1867). — Angianthus filifolius (Benth.) C. A. Gardner, Enum. PI. Austr, Occ. 135 (1931); Grieve & Blackall W. Aust. Wildfls 811 (1975). Type: “Murray river, Oldfield.” Holotype: Oldfields n ’ Tufts, low wet places, Murray R., W. Aust., s. dat. (K). Isotype: MEL 84436 (see note 1)’ 206 Angianthus humifusus var. minor Benth., FI. Austr. 3:563 (1867). Type: “Kalgan river, Oldfield.’ ’ Lectotype (here designated): Oldfield^.n. , Kalgan River, s. dat. (K) (see note 2). Probable Isolectotypes: K (mounted on lectotype sheet), MEL 84433. Annual herb. Major axes decumbent to erect, l-3(6.5) cm long, glabrous or variably hairy; stem simple or forming major branches at basal and/or upper nodes. Leaves often opposite at the base, the upper ones alternate, ± linear or lanceolate, c. 0.5-l(1.7) cm long, c. 0.1 cm wide, at least the upper leaves mucronate, all leaves glabrous or sparsely hairy. Compound heads ± ellipsoid or lanceoloid to very broadly ovoid, (0.6)l-2.4(2.7) cm long, c. 0.5-1. 5 cm diam. Capitulum with c. 10-13 capitular bracts and paleae, all bracts obovate to widely obovate, 1.25-1.9 mm long, 0.65-1.4 mm wide, crenulate near the apex, white or pale pink. Florets c. 4-5; corolla 4 or 5-lobed, the tube distinctly swollen in the lower Vi, c. 0.35-0.65 mm long, 0.3-0.65 mm diam. Achenes ± obovoid, c. 0.4-0.65 mm long, 0.2-0. 3 mm diam., variably papillose. Pappus of 4-6 jagged scales joined at the base, 0.6-0.9 mm long, c. equal to or slightly exceeding the length of the floret. Fig. 15. Distribution (Fig. 15): South-west of Western Australia being found within an approximately 200 km wide coastal belt. Ecology: Data suggest that S. filifolius, like the closely related S. humifusus with which it commonly grows, is capable of growing in a variety of habitats. Collectors’ notes include “Minor drainage channel in Regelia heath community ... in open sandy sites”, “on soil at base of granite rock in moist situations. Associated with Angianthus sp. under open Acacia scrub”, ''Eucalyptus-Xanthorrhoea community on deep grey sands. Growing c. 10 m from Siloxerus humifusus'" and “Growing in open Eucalyptus woodland on brown sandy loam covered by coarse gravel. Growing with Siloxerus humifusus". Notes: 1. When describing Gnaphalodes filifolium Bentham (1867) cited a collection made by Oldfield from the Murray River, Western Australia. This collection, said by Bentham to consist of a single specimen, is housed at K. A duplicate of this collection, MEL 84436, also exists. Bentham, as indicated by the initial B on the label, saw this specimen but apparently overlooked it when preparing his description of the species. The name G. filifolium does not appear on the sheet. 2. When describing the variety Angianthus humifusus var. minor Bentham (1867) cited a single collection i.e., Oldfields.n., Kalgan River. He described the taxon as having “Clusters of flower-heads and . . . flowers . . . much smaller” than var. grandiflorus ( = S. humifusus) and “a pappus as nearly as long as the floret”. The coUection at K contains, as previously noted by Ostenfeld (1921), individuals belonging to two distinct taxa, namely S. humifusus and S. filifolius. The brief description of var. minor almost certainly refers to the latter species and indeed there is an envelope attached to the K sheet which contains a single specimen of 5. filifolius and is clearly marked “var. minor, Kalgan river, Oldfield”. This collection has been designated as the lectotype. Two further specimens of S. filifolius are mounted with specimens of S. humifusus and these too are marked as “var. w//7or Benth.” However they are mounted on a separate piece of paper which has been subsequently attached to the main sheet and are referred to as probably isolectotype material. The sheet also contains specimens of S. humifusus collected by Oldfield from the Kalgan River and the Gordon River, Western Australia. A further Oldfield collection from the Kalgan River is located in MEL (MEL 84433). Although not named var. minor the specimens were seen by Bentham (as indicated by the initial B) and are undoubtedly the same taxon. Unlike the lectotype collection it is marked with the number 84d. It is referred to here as a probable isolectotype. Selected Specimens Examined (7/43): Western Australia — Burbidge 4977^ 3 miles S. of Bunbury, 2.i.l956 (PERTH); Burbidge 7955, Twin Swamps Wildlife Sanctuary, 28.xii.1971 (CANB, PERTH); Chinnock 4334, Mt. Walker Rock, 10.xi.l978 207 (AD); Demarz 5351, Orleans Farm, 16.X.1974 (KP, PERTH); Short 1056, c. 10 km from Jarrahwood along road to Nannup, 22. xi. 1979 (AD); Short 1058, c. 41 km from Kojonup along main Boyup Brook road, 23. xi. 1979 (AD); Willis s.n.. North Twin Peak Island, 20.xi.l950 (MEL). 2. Siloxems humifusus Labill., PL Nov. Holl. 2:57 (1806); Less., Syn. generum Comp. 270 (1832). — Styloncerus humifusus (Labill.) Spreng., Syst. Veg. 3:451 (1826); DC., Prod. 6:149 (1838); Steetz in Lehm. PI. Preiss. 1:435 (1845). — Ogcerostylus humifusus (Labill.) Cass., Diet. Sci. Nat. 49:222 (1827); Steud., Nomen. Bot. 2nd ed. 2:242 (1841) {"Oxerostylus') (n.v.). — Angianthus humifusus (Labill.) Benth., H. Austr. 3:563 (1867); Grieve & Blackall, W. Aust. Wildfls 811 (1975). Type: “Habitat in terra Van-Leuwin.” Holotype: ILabillardiere s.n., habitat in terra van-Leuwin, s. dat. (FI). Styloncerus cylindraceus Steetz in Lehm., PI. Preiss. 1:435 (1845). Type: “In sinu regis Georgii III. mense Nov. 1840. Herb. Preiss. No. 41.” Lectotype (here designated): Preiss 41, In Nova Hollandia, (Swan-River Colonia) in sinu regis Georgii III, s. dat. (MEL 541624, ex herb. Steetz). Isolectotypes: LD, MEL 54151 (ex herb O. W. Sonder), S.(Seep.l52). Styloncerus suberectus Steetz in Lehm. PI. Preiss. 1:436 (1845). Type: “In arenosis terrae in ferioris, mense Dec. 1839. Herb. Preiss. no. 42.” Lectotype (here designated): Preiss 42, in arenosis terrae inferioris (Swan River Colonia), s. dat. (MEL 541622, ex herb. Steetz). Isolectotypes: LD, MEL 541623 (herbO. W. Sonder). (Seep. 152). Angianthus humifusus var. grandiflorus Btnih,, FI. Austr. 3:563 (1867), type as for S. suberectus. Annual herb. Major axes decumbent to erect, 2-7(9) cm long, glabrous or variably hairy; stem simple or forming major branches at basal and upper nodes. Leaves often opposite at the base of the major axes, the upper ones alternate, all leaves ± linear or lanceolate, (c. 1)1. 5-3 cm long, c. 0.1-0.15 cm wide, glabrous or sparsely hairy, at least the upper ones mucronate. Compound heads ± broadly ellipsoid or ovoid to broadly depressed ovoid, c. 0.6-2(2.9) cm long, (c. 0.5)0.7-1.2(1.3) cm diam. Capitulum with c.8-10 capitular bracts and paleae, all bracts oblanceolate to obovate, (c. 2)2.5-4.5(6.3) mm long, (0.7)0.9-1.7(1.9) mm wide, crenulate near the apex, white or pale pink. Florets c. 5; corolla 4 or 5-lobed, the tube distinctly swollen in the lower Vi, (c. 0.85)1-2(2.25) mm long, c. 0.3-0.5 mm diam. Achenes ± obovoid, c. 0.7-0.95 mm long, c. 0.25-0.4 mm diam., variably papillose. Pappus of 5-7 jagged scales fused at the base, c. 0.95-1.7 mmlong,c. Vi or rarely the length of the floret. Fig. 15. Distribution (Fig. 15): South-west of Western Australia, within an approximately 200 km wide coastal belt. Ecology: Grows in a variety of habitats. Collectors’ notes include “Recently dried muddy depression in sandy swamp under Acacia cyanophyM\ “Rush marsh . . . under shrubs of Astartaea fascicularis with Cotula coronopifolia and Schoenus trachycarpus^\ ''Eucalyptus-Xanthorrhoea community on deep grey sands. Growing c. 10 cm from Siloxerus filifolius'^ and “Growing in open Eucalyptus woodland on brown sandy loam covered by coarse gravel. Growing with Siloxerus fdifolius^\ Notes: 1. S. humifusus is primarily distinguishable from S. filifolius on differences in size of various organs, the achenes, capitular bracts, paleae, pappus scales and florets of S. humifusus being approximately twice the length of the same organs in the latter species. Such features suggest that S. humijmus may be of polyploid origin. 2. Bentham (1867) recognised two varieties of Angianthus humifusus, var. minor Benth. and var. grandiflorus Benth. The former variety is recognised here as a distinct species, Siloxerus filifolius. The latter variety was based on Preiss 42, the type collection of Styloncerus suberectus Steetz, which possesses larger capitular bracts and paleae (c. 4-6.3 mm long) than those of Preiss 41, (c. 3.7-4.2 mm long), the type of Styloncerus cylindraceus Steetz. Furthermore in Preiss 42 the pappus is about one-half the length of 208 the floret whereas in Preiss 41 it is approximately the length of the floret. Thus it is not surprising that in the past S. suberectus and S. cylindraceus have been recognised as different taxa. Initially it was felt that these separate taxa could be maintained. However, although extensive field studies have not been made, examination of other herbarium collections has shown that the recognition of two taxa is apparently not tenable, the size of the bracts and the ratio of pappus length to floret length being quite variable. Selected Specimens Examined (6/97): Western Australia — Abbot 53, Island, Recherche Archipelago, ii. 1976 (MEL); Burbidge7945, Twin Swamps Wildlife Sanctuary, 11.1.1972 (CANB); Burbidge 7962, Twin Swamps Wildlife Sanctuary, 20.i.l972 (CANB, PERTH); Congdon 75034b, Blackwood River Estuary, 29.xi.1975 (PERTH); Short 1055, c. 1 km from Jarrahwood along road to Nannup, 22. xi. 1979 (AD); 5/tor? 7059, c. 41 km from Kojonup along main Boyup Brook road, 23. xi. 1979 (AD). 3. Siloxenis pygmaeus (A. Gray) Short, Muelleria 4:413 (1981). — Chamaesphaerion pygmaeum A. Gray, Hook. J. Bot. Kew Gard. Misc. 3:177 (June 1851). — Chthonocephalus pygmaeus (A. Gray) Benth., FI. Austr. 3:582 (1867); Grieve & Blackall, W. Aust. Wildfls 820 (1975). Type: “South-western Australia, Drummond.” Lectotype: (here designated): Drummond 55, S.W. Australia, 1850 (K). Isolectotypes: GH (ex herb. Klatt), MEL 542228, PERTH (ex herb. K, ex herb. TCD). Gyrostephium rhizocephalum Turcz., Bull. Soc. Naturalistes Moscou 24(2):77 (Oct. 1851). IVpe: “Nova Hollandia. Drum.V.n.55.” Holotype: ? CW, n.v. (see p.OOO). Isotypes: GH (ex herb. Klatt), K, PERTH (ex herb. K, ex herb. TCD). Annual, almost stemless herb consisting of a compound head surrounded by a basal rosette of c. 10-20(30) leaves. Leaves lanceolate, c. 0.5-1 cm long, c. 0.1 cm wide, glabrous or sparsely hairy, mucronate and usually with distinct hyaline margins at the base. Compound heads depressed ovoid, c. 0.4-0.6 cm long, 0. 6-1.1 cm diam. Capitulum with (18)20-30 capitular bracts and paleae, all bracts narrowly elliptic to elliptic or sometimes oblanceolate to obovate, 3.2-4.2(4.5) mm long, (0.75)0.85-1.5 (1.85) mm wide, white. Florets c. 10-20; corolla 3- or rarely 4-lobed, the tube tapering gradually to the base, 1.5-1. 8(2.1) mm long, 0.2-0.25 mm diam. Achenes ± obovoid, 0.6-0.75(0.85) mm long, 0.3-0.5 mm diam., papillose. Pappus a jagged ring c. 0.15-0.45 mm long. Fig. 15. Chromosome number: n = c. 12 or 13. Distribution (Fig. 15): South-west of Western Australia, occurring south of latitude c. 30°S and between longitudes c. 117°E and c. 122°E. Ecology: Generally restricted to saline, sandy soils surrounding inland salt lakes but also found at the base of granite outcrops. Collectors’ notes include “Granite outcrops . . . Sandy loam amongst Eucalpytus woodland at base of rock”, “White to greyish sand between Melaleuca and extending into Arthrocnemum [ = Halosarcia] zone around salt depression” and “Growing in open areas on pale brown, very sandy loam between Melaleuca and Eucalyptus above saline depression”. Note: 1. Apparently mature achenes of this species exhibit marked size differences within any one compound head. Some fruits are c. \ Vi times larger than the majority. It is difficult to ascertain their exact location but they appear to occur on the outer margins of the compound heads. The larger fruits generally appear to germinate several days earlier than the smallest ones in the heads. Such a staggering of germination times may be of adaptive value in areas of low, unreliable rainfall; that is unless sufficient moisture is available for a prolonged period of time the smaller achenes will remain dormant. A better food supply in the larger fruits may ensure their survival in adverse conditions. S. pygmaeus, at least in part of its range, does occur in a low rainfall area. Furthermore southern Australia has experienced greater cycles of aridity in the recent past than have occurred throughout the Tertiary period. 209 Selected Specimens Examined (7/18): Western Australia — Chinnock 4160, c. 3.5 km W. of western edge of Lake King, 26.ix.1977 (AD); Chinnock 4598, Phillips River, 17 km W. of Ravensthorpe, 8.x. 1979 (AD); Newbey 4342, 16 km N. of Needilup, 4.ix.l974 (PERTH); Short 660, Roe Dam, 23. ix. 1977 (AD); Short 678, 1 km E. of Wave Rock, 25.ix.1977 (AD); Short 949, western edge of Lake Campion, 14.xi.l979 (AD); Short 1071, c. 10 km SW. of Pingrup, 23 .xi. 1979 (AD). SPECIES OF UNCERTAIN AFFINITY In this revision a number of species referred to Angianthus by Bentham (1867) are considered to belong to distinctive, segregate genera. However there are three species, namely A. axilliflorus, A. burkittii a.nd A, conwa/us" which clearly have no affinities with Angianthus s.str. or any of the segregate genera. They may represent monotypic genera or have affinities with other members of the Gnaphaliinae not yet examined by the author. For the time being it seems appropriate to refer A. burkittii to Gnephosis, the genus to which the species was originally referred by Bentham (l.c.). Lectotypes have been chosen for the above three species and various attributes of each are noted below. Angianthus axilliflorus W. V. Fitzg. ex. Ewart & J. White, Proc. Roy. Soc. Viet. 22:315, pi. 56, figs. 1-3, {m9){‘axilifloru^)\ W. V. Fitzg., J. Bot. 50:21 (1912); Grieve &Blackall, W. Aust. Wildfls 812 (1975) {^axiliflorus'). Type: Cowcowing, W. Australia. Max Koch, Oct., 1904. No. 1196.” Lectotype (here designated): Koch 1196, Cowcowing, 1904 (MEL 541217). IsoLECTOTYPEs: AD, MEL 541218, MEL 541219, NSW (2 sheets), PERTH (see note 1 below). Notes: 1. Ewart & White (1909), working in Melbourne, noted that the “species was received, marked W. V. Fitzgerald inedit, from both the collector and the Sydney Herbarium” (p. 316). All three MEL sheets bear good specimens but MEL 541217, the sheet containing the specimens received from NSW and the only one marked with the word “ined.”, has been chosen as the lectotype. 2. The rigid, leaf-like capitulum-subtending bracts and the arrangement of the capitular bracts readily distinguish this species from all others included \n Angianthus s.l. 3. The species is apparently rare. Apart from the type material the only other collection of the species seen by the author is Blackall 1276, collected from the edge of a salt lake near Newdegate, Western Australia in 1931. Angianthus connatus W. V. Fitzg., J. West. Aust. Nat. Hist. Soc. 2:24 (1905); Grieve & Blackall, W. Aust. Wildfls 816 (1975). Type: “Minginew.-W. V. E, September, 1903.” Lectotype (here designated): Fitzgerald s.n.,M\n^nt^, W.A., -.ix.l903 (NSW 138682). Iso lectotypes: NSW 138683, PERTH (ex W. E. Blackall) (see note 1 below). Notes: 1 . Both of the NSW collections, unlike the PERTH collection, are clearly designated as coming from Fitzgerald’s herbarium. Of the two sheets NSW 138682 contains the best material and therefore has been designated as the lectotype. The remaining one, NSW 138683, has the word “type” written on the label, possibly in Fitzgerald’s hand, but there is no reason to believe that Fitzgerald did not base his description on all of the material available to him. 2. The species is known only from the type collection. It is readily distinguished from other members of Angianthus s.l. by the presence of long hairs at the base of the florets or apex of the achenes and the rigid, opaque capitulum-subtending bracts. The morphology of the capitular bracts is also unique. Gnephosis burkittii Benth., FI. Austr. 3:570 (1867). — Angianthus burkittii (Benth.) J. M. Black, FI. S. Aust. 1st ed. 645, pi. 53 (1929), 2nd ed. 925, fig. 1227 (1957)- Hi Eichl., Suppl. to J. M. Black’s H. S. Aust. 326 (1965); Willis, Handb. PI. Viet 2*730 210 (1973); Grieve & Blackall, W. Aust. Wildfls 813 (1975). Type: “Lake Gillies, Burkitt.” Lectotype (here designated): Burkitt s.n.y Lake Gillies, s. dat. (MEL 541211). Isolectotype: K (see note 1 below). Angianthus whitei J. M. Black, Trans & Proc. Roy. Soc. S. Aust. 37:122, pi. 4 (1913). Type: “Corunna Station, Eyre Peninsula, August, 1912.’’ Lectotype (here designated): White s,n,, Corunna Statn., W. of Pt. Augt., near Lake Gilles, 27.viii.1912 (AD 98103150). Isolectotypes: MEL 541611, NSW 7831/13. Notes: 1. Both of the sheets containing type material of G. burkittii were seen by Bentham. The MEL sheet contains by far the better collection and thus has been chosen as the lectotype. 2. The reddish, prostrate or ascending major axes, the woolly compound heads and the pappus of 8-12 barbed bristles readily distinguish this species from others included in Angianthus s.l. The capitular bracts are also unique to this species. NAMES OF UNCERTAIN APPLICATION Bentham (1867) described as new a species he called Angianthus plumiger. His description was based on collections made by Oldfield from the Swan and Murchison rivers. It could be expected that type specimens of this species would be housed in K and/or BM but apparently no collections of this taxon exist in either institution (A. A. Munir & J. Lewis, pers. comms, 1980). No specimens have been located in E or any Australian herbaria. From the description it seems that the name should not be applied to any species, old or new, described in the current revision of Angianthus sJ. Cassini (1820) described as new the genus Hirnellia and attributed to it a single species H, cotuloides Cass. De Candolle (1838) regarded //. cotuloides as a possible synonym of Angianthus tomentosus but it was not listed as such by Bentham (1867) and subsequent workers on the Australian flora. It has not been possible to view type material but from the published description it appears that the name is not a synonym of A. tomentosus. ACKNOWLEDGEMENTS Work on this paper was primarily carried out when I was the recipient of a Commonwealth Postgraduate Research Award at the Flinders University of South Austredia. I sincerely thank my supervisor. Dr B. A. Barlow, for his general advice throughout the project and for his comments on the original manuscript. Many other people, in particular the staff at AD and my colleagues in MEL, have contributed to my studies by collecting various species and by providing constructive criticism of the manuscript. I thank them all, especially the following: Dr W. R. Barker for the translation of descriptions of new species into Latin, for collections of various species and for general comments relating to terminology, typification and descriptive format; Mr R. J. Chinnock for his collections, for testing the keys and for allowing me to accompany him on an extensive field trip to Western Australia in 1977; Dr Hj . Eichler for originally suggesting the project and for providing publication dates of the Bull. Soc. Imp. Naturalistes Moscow, Miss H. I. Aston and Dr J. H. Ross for comments on the manuscript. I also thank Miss A. M. Podwyszynski for the illustrations. Much of my working time from January 1977 to January 1980 was spent at AD and I thank Dr J. P. Jessop for making available the facilities of that institute. Field work in 1977 was made possible by the generosity of the Board of the Adelaide Botanic Gardens which allowed me to accompany Mr Chinnock to Western Australia. Thanks are also due to Ms D. Nicholas and Mr M. Tippett for field assistance in Western Australia in 1979, this trip being partly financed by the Flinders University Research Committee. Finally I thank Ms T. Munro and Ms M. James for typing the manuscript. 211 REFERENCES Bentham, G. (1863). ‘Flora Australiensis’ vol. 1. (Reeve: London). Preface, pp. 1-18. Bentham, G. (1867). ‘Flora Australiensis’ vol. 3. (Reeve: London). Compositae, pp. 447-680. Bentham, G. (1873a). Compositae. In Bentham, G. & Hooker, J. D., ‘Genera Plantarum’ vol. 2. (Reeve: London), pp. 163-533. Bentham, G. (1873b). Notes on the classification, history and geographical distribution of the Compositae. J. Linn. Soc. (Bot.) 13: 335-577. Besold, B. (1971). Pollenmorphologische Untersuchungen an Inuleen (Angjanthinae, Relhaniinae, Athrixiinae). Dissertationes Botanicae 14. Bettenay, E. & Mulcahy, M. J. (1972). Soil and landscape studies in Western Australia. (2) Valley form and surface features of the south-west drainage division. J. Geoi Soc. Aust. 18: 359-369. Black, J. M. (1929). ‘Flora of South Australia’ pt. 4, ed. 1. (Govt. Printer: Adelaide), pp. 570-662. Cassini, H. (1820). Description d’un nouveau genre de plantes (Hirnellia). Bull. Sci. Soc. Philom. Paris pp. 57-58. Court, A. B. (1972). Preliminary notice on the Sender collection in the National Herbarium of Victoria. Mueller ia 2:188. De Candolle, A. P. (1813). ‘Theorie elementaire de la botanique’. (Paris). De Candolle, A. P. (1836). ‘Prodromus Systematis Naturalis Regni Vegetabilis’ vol. 5. (Paris). Compositae, pp. 4-706. De Candolle, A. P. (1838). ‘Prodromus Systematis Naturalis Regni Vegetabilis’ vol. 6. (Paris). Compositae, pp. 1-687. De Jong, D. C. D. (1965). A systematic study of the genus Astranthium (Compositae, Astereae). Mich. State Univ. Publ. Biol. Ser. 2: 429-528. Eichler, Hj. (1963). Some new names and new combinations relevant to the Australian flora. Taxon. 12: 295-297. Gardner, C. A. (1931). ‘Enumeratio Plantarum Australiae Occidentalis’. (Govt. Printer: Perth). Gardner, R. C. (1977). Observations on tetramerous disc florets in the Compositae. Rhodora 79: 139-146. Gray, A. (1851). Characters of some Gnaphalioid Compositae of the division Angjantheae. Hook. J. Bot. Kew Card. Misc. 3: 97-102, 147-153, 172-178. Heywood, V. H. (1971). Systematic survey of Old World Umbelliferae. In Heywood, V. H. (ed.)‘The Biology and Chemistry of the Umbelliferae’. (Academic Press: London & New York), pp. 31-41. Heywood, V. H., Harbome, J. B. & Turner, B. L. (eds) (1977). ‘The Biology and Chemistry of the Compositae’. (Academic Press: New York & London). 2 vols. Heywood, V. H., Harbome, J. B. & Turner, B. L. (1977). An overture to the Compositae. In Heywood, V. H., Harbome, J. B. & Turner, B. L. (eds), l.c. 1: 1-20. Heywood, V. H. & Humphries, C. J. (1977). Anthemideae — systematic review. In Heywood, V. H., Harbome, J. B. & Turner, B. L. (eds), l.c. 2: 851-898. Hoffman, O. (1890). Inuleae — Angianthae. In Engler, A. & Prantl, K. (eds), ‘Die natiirlichen Pflanzen- familien’. (Engleman: Leipzig). 4 (5): 192-195. Holmgren, P. K. & Keuken, W. (1974). ‘Index Herbariorum. Part 1. The Herbaria of the World’, ed. 1. Regnum Veg. 92: 1-397. Ising, E. H. (1922). Ecological notes on South Australian plants. Part 1. Trans. & Proc. Roy. Soc. S. Aust. 46: 583,^. Jessop, J. P. (1977). Endangered species in the South Australian native vascular flora. J. Adelaide Bot. Card. 1: 135-139. Keighery, G. J. (1982). Pollination syndromes and breeding systems of Western Australian arid zone plants. In Barker, W. R., Greenslade, P. J. M. & Baverstock, P. R. (eds), ‘Evolution of the Flora and Fauna of Arid Australia’. King, R. M. & Robinson, H. (1966). Generic limitations in the Hofmeisteria Complex (Compositae — Eupatorieae). Phytologia 12: 465-476. Leins, P. (1971). Pollensystematische Studien an Inuleen. 1. Tarchonanthinae, Plucheinae, Inulinae, Buph- thalminae. Bot. Jahrb. 91: 91-146. Leins, P. (1973). Pollensystematische Studien an Inuleen. II. Filagininae. Bot. Jahrb. 93: 603-611. Leppick, E. E. (1977). The evolution of capitulum types of the Compositae in the light of insect-flower inter- action. In Heywood, V. H., Harbome, J. B.. & Turner, B. L. (eds), ‘The Biology and Chemistry of the Compositae’. (Academic Press: London, New York & San Francisco). 1: 61-89. Lessing, C. H. (1832). Crossolepis. ‘Synopsis generum Compositarum’. (Berlin), p. 270. Merxmuller, H. Leins, P. &Roessler, H. (1977). Inuleae — systematic review. /« Heywood, V. H., Harbome, J. B. & Turner, B. L. (eds), ‘The Biology and Chemistry of the Compositae’. (Academic Press: London, New York & San Francisco). 1: 577-602. Mirbel, C. F. B. de (1815). ‘Elemens de physiologic vegetale et de botanique’ vol. 1. (Paris). Mueller, F. J. H. (1888). The Melbourne Herbarium. Card. Chron. 4: 211-212. Mueller, F. J. H. (1889). ‘Second systematic census of Australian plants. Part 1. Vasculares*. (McCarron, Bird & Co.: Melbourne). Mulcahy, M. J. & Bettenay, E. (1972). Soil and landscape studies in Western Australia. (1) The major drainage divisions. J. Geol. Soc. Aust. 18: 349-357. Necker, N. J. de (1790). ‘Corollarium ad Philosophiam botanicum Linnaei spectans’. (Neuwied). Nordenstam, B. (1980). The herbaria of Lehmann and Sender in Stockholm, with special reference to the Ecklon and Zeyher collection. Taxon 29: 279-291. Ostenfeld, C. H. (1921). Contributions to Western Australian botany. Pt. 3. Biol. Meddel. Kongel. Danske Vidensk. Selsk. 3: 1-44. 212 Robinson, H. &King, R. M. (1977). Eupatorieae — systematic review. In Heywood, V. H., Harborne, J. B. & Turner, B. L. (eds), ‘The Biology and Chemistry of the Compositae’. (Academic Press: London, New York & San Francisco). 1: 437-485. Roth. I. (1977). Fruits of Angiosperms. In Zimmerman, W., Carlquist, S., Ozenda, P. & Wulff, H. D. (eds), ‘Handbuch der Pflanzenanatomie’. (Gebruder Borntraeger: Berline & Stuttgart). 10: 1-675. Schodde, R. (1963). A taxonomic revision of the genus Millotia Cassini (Compositae). Trans. Roy. Soc. S. Aust. 87:209-241. Short, P. S. (1981a). Pollen-ovule ratios, breeding systems and distribution patterns of some Australian Gnaphaliinae (Compositae:Inuleae). Muelleria 395-417. Short, P. S. (1981b). ‘Biosystematic studies in Australian Gnaphaliinae (Compositae:Inuleae)’. Ph.D. thesis, Flinders University of South Australia. Short, P. S. (1981c). Cephalipterum . 7/i Jessop, J. P. (chief ed.), ‘Flora of Central Australia’. (Reed: Sydney), p. 391. Skvarla, J. J., Turner, B. L., Patel, V. C. & Tomb, A. S. (1977). In Heywood, V. H., Harborne, J. B. & Turner, B. L. (eds), ‘The Biology and Chemistry of the Compositae’. (Academic Press: London, New York & San Francisco). 1: 141-265. Smith, J. (1881). Botanical collectors. Card. Chron. 16: 568-570. Stafleu, F. A. (1967). Taxonomic literature: a selective guide to botanical publications with dates, commen- taries and types. Regnum Veg. 52: 1-556. Stafleu, F. A. & Cowan, R. S. (1976). Taxonomic literature. A selective guide to botanical publications with dates, commentaries and types. Regnum Weg. 94: 1-1136. Stafleu, F. A. & Cowan, R. S. (1979). Taxonomic literature. Regnum Veg. 98: 1-99. Stafleu, F. A. & Cowan, R. S. (1981). Taxonomic literature. Regnum Veg. 105: 1-980. Steam, W. T. (1962). ‘Three prefaces on Linnaeus and Robert Brown’. (Cramer: Weinheim). Systematics Association Committee for Descriptive Terminology (1962). Terminology of simple symmetrical plane shapes (Chart 1). Taxon 11: 145-156, 345-347. Turczaninow, N. (1851). Synanthereae. Quaedam hucusque indescriptae. Bull. Soc. Naturalistes Moscou 24 (1): 166-214; 24 (2); 59-95. Turner, B. L. (1970). Chromosome numbers in the Compositae. XII. Australian species. Amer. J. Bot. 57: 382-389. Wagenitz, G. (1976). Was ist eine Achane? Zur Geschichte eines karpologischen Begriffs. Candollea 31: 79-85. INDEX TO GENERA AND SPECIES OF COMPOSITAE New names are in bold. Synonyms and misapplied names are in italics. Page numbers which refer to the main taxonomic treatment are in bold, those which refer to maps and figures are asterisked and those in brackets refer to keys. Actinobole 145 Angianthus 143-146, 147*, 148-152, (153), 154, 155* (156)-I58, 173, 174, 179, 182, 187, 201,203-205,209,210 sect. Pyramidatae i44 sect. Skirrhophorus 144 sect. Spicatae 144 acrohyalinus 144, 148, 151, 155* (157), 158*-I60 amplexicaulis 202 axilliflorus 143, 144, (153), 209 brachypappus 143, 154, 155*, (157), 158*, 164,166-168 brachypappus 167 var. conocephalus 143, 167, 168 burkittii 143, 144, 209 connatus 143, 144, 148, (153), 209 conocephalusl55*, 156, (157), 158*, 166, 167, 168, 171 cornutusl43, 155* (156), 158* 169, 170* cunninghamii 144, 148, 154, 155*, (156), 178, 179 cyathiferl43, 151, 155*, (157), 160, 161 demissus 1 8 1 drummondii 155*, (156, 157), 174-177 eriocephalus 151, 176 filifolius 205 flavescens 164 glabratus 143, 147*, 154, 155*, (157), 158*, 162, 163* globifer 203 humifusus 149, 207 var. grandiflorus 207 var. minor 206, 207 lanigerus 144, 203, 204 microcephalus 149, 154, 155*, (156), 173, 174 micropodioides 153, 155*, (156, 157), 158*, 168, 169, 171 micropoides 168 \?iX.filaginoides 168 milnei 147* 151, 155* (156, 157), 158* 159,160 milnei 169 muellerianus 204 myosuroides 198 phyllocalymmeus 180 phyllocephalus 1 83 platycephalus 174 pleuropappus 1 80 plumiger210 preissianus 148, 151, 155*, (156), 158*, 175-176, 177 prostratus 143, 151, 155*, (156), 171, 172* pusillus 189 \SLV. polyanthus 190 pygmaeus 151, 155*, (156), 175-177 St rictus 149, 203, 204 var. lanigerus lOA tenellus 193 tomentosus 143, 147*, 151, 153, 155*-(157), 158*-160, 162,164, 165,210 tomentosus 160 whitei 210 213 Angianthusm, 181, 182, 185,201-204 Basedowia 145 Blennospora 145 Caesulia 145 Calocephalus 145, 149 Cassinia 153 Cassinia 153 aurea 153, 164, 166 Cephalipterum 145 drummondii 145 Cephalosorus 143, 144, 148, 151, (153), 182 brevipapposus 183 carpesioides 143, 180*, 182, 183 gymnocephalus 182 microcephalus 173 phyllocephalus 182, 183 Cephalosorus 154 Chamaesphaerion 204 pygmaeum 204, 208 Chrysocoryne 143, 144, 148-151, (153), 185, I86*-(189), 191,193,195,199 sect. Bisquama 185 angianthoides 190, 192 drummondii 147*, 149, 181, 185, 186*-(189), 191*, 193,195,196 huegeln 189 multiflora 143, 149, 186*-(189), 191*, 192, 194* myosuroides 198, 199 pusilla 147*, 149, 185, 186*-(189), 191*-193 pusilla 195 sp. A 192 sp. B 196 sp. C 199 tenellalSU 193, 196 tridens 143, 148, 149, 151, 186*-(189), 196,198, 199, 200* Irifida 143, 149, 186*-(189), 191*, 196, 197* uniflora 149, 151, 185, 186*-(189), 198, 199 Chthonocephalus 145, 204 pygmaeus 208 Craspedia 145 Crossolepis 187 linifolia 187 pusilla ISl, 189, 193, 195 Crossolepis 185 Cylindrosorus \Al , 153 flavescens 153, 164, 166 Decazesia 145 Dimeresia 145 Dithyrostegia 143, 144, 148, 151, (153), 201 amplexicaulis 147*, 201*, 202 Epitriche 143, 144, (153), 154, 155, 181 cuspidata , 182 demissusl43,151,180*, 181 Eriochlamys 145 behrii 145 Eriocladium 143, 154 pyramidatum 154, 178, 179 Eriosphaera 145 Gamozygis 143, 201 flexuosa 201 , 202 Gnaphalium 145 Gnaphalodes 145, 204 filifolium 204-206 Gnephosis 145, 209 burkittiil44,(153),209,210 Gyrostephium 204 rhizocephalum 204, 208 Helichrysum 145, 149 Helipterum 145, 149 sterilescens 145 Hirnellia210 cotuloides210 Hyalochlamys 143, 144, 148, 151, (153), 202, 203 globifera 147*, 201*-203 Myriocephalus 145 stuartii 190 Nablonium 145 Ogcerostylus 143, 144, 204 humifusus 207 Oxerostylus 204 Phyllocalymma 143, 144, 153 filaginoides 152, 168, 169 micropodioides 153, 168, 169 Piptostemma 143, 182 carpesioides 182, 183 Pleuropappus 143, 144, 146, 148, 149, 151, (153), 154, 179 phyllocalymmeus 147*, 154, 179, 180*, 196 Pogonolepis 143-145, 148, 149, 151, (153), 203, 204 lanigera 143, 204 muelleriana 204 strictal47* 149, 203,204 Siloxerus 143-145, 148-151, (153), 204, (205) brachypappus 1 66 filifolius204,(205*)-207 humifusus 204, (205*), 206, 207 pusillus 1 89 pygmaeus 204, (205*), 208 strictus 204 tenellus 193 tomentosus 164 Siloxerus 154, 185, 204 Skirrhophorus\A2>, 144, 153, 181,204 sect. Pogonolepis 203 sect. Psuedopappus 144, 181 sect. Skirrhophorus 144, 154 cunninghamii 154, 178 drummondii 174 demissus 144, 181 eriocephalus 176, 177 mucronulatus 175 muellerianus 204 phyllocephalus 183 preissianus 152, 176 pygmaeus 175, 176 strictus 204 Stuartina 145 Styloncerus 143, 144, 154, 179, 181, 182, 185, 201,202, 204 amplexicaulis 202 brachypappus 166 cunninghamii 178 cylindraceus 152, 207, 208 demissus 181 drummondii 174 eriocephalus 176 globifer 203 214 humifusus 149, 207 microcephalus 173 micropodes 168 micropodioides 168 milnei 159 myosuroides 198 phyllocalymmeus 180 phyllocephalus 183 platycephalus 174 preissianus 176 pusiilus 189 pygmaeus 175 strictus 204 suberectus 152, 207, 208 tenellus 193 tomentosus 164 Manuscript received 30 September 1981. STORCKIELLA AUSTRALIENSIS SP. NOV. (CAESALPINIACEAE) FROM NORTHERN QUEENSLAND: A NEW GENERIC RECORD FOR AUSTRALIA by J. H. Ross* & B. P. M. Hyland! ABSTRACT Storckiella australiensis J. H. Ross & B. P. M. Hyland sp. nov. is described from the Cook District of northern Queensland and the characters that distinguish it from other species are given. This constitutes the first record of the genus Storckiella Seeman for Australia. TAXONOMY The account of Caesalpiniaceae for the Hora of Australia is being prepared by one of us (J. H. R.) and this opportunity is taken of describing a new species of Storckiella Seem, from northern Queensland. Storckiella australiensis J. H. Ross & B. P. M. Hyland, sp. nov., affinis incertae; ab omnibus speciebus staminibus 5(6) filamentis brevibus, et foliolis paucioribus majoribus, differt. Arbor usque ad 35 m alta, caule usque ad 120 cm diametro, anteridibus praedita. Folia imparipinnata: petiolus 2-6 cm longus; rhachis 2-11.5 cm longa; foliola (3)5, obovato-elliptica vel oblonga, 6.5- 21 cm longa, 2.5-9. 7 cm lata, basi cuneata, apice acuta vel acuminata, coriacea, costa et nervis lateralibus satis obviis; stipulae late ovatae, 3-4 x 2-3 mm, imbricatae, gemmam terminalem includentes, mox caducae. Inflorescentiae paniculatae, terminates vel axillares, 15-25 cm longae. Flores hermaphroditi. Sepata 5, late ovata, 5-8 mm longa, 3-6 mm lata, externe et interne ferrugineo-sericea, in alabastro imbricata, sub anthesi reflexa. Petala 5(6), elliptica 7.5- 10 mm longa, 3-4 mm lata, libera, armeniaca. Stamina 5(6), libera, omnia fertilia; filamenta 0.8-1 .2 mm longa; antherae basifixae, loculae rimia brevibus lateralibus apicalibus dehiscentes. Ovarium sessile, usque ad 4.5 mm longum et usque ad 1.5 mm latum, compresso-planum, basi et margine ferrugineo-sericeum. Legumen elliptico-oblongum, compresso-planum, 5-11 cm longum, 2.8-4 cm latum, dorsale late alatum, coriaceum, glabrum. Semina 1-2(5), applanata, ± reniformia, 7-10 mm longa, 10-16 mm lata, albuminosa. Tree up to 35 m tall, stem up to 120 cm diameter at breast height and conspicuously buttressed in mature plants. 5arA:less than 2.5 cm thick, smoothish, with conspicuous pale or rust-coloured mostly horizontally elongated lenticels; outer blaze varying from pink to reddish-brown and texture from fibrous to granular; inner blaze pink, fibrous and with conspicuous ripple marks. Subrhytidome variable but generally showing shades of pink and red. Heartwood (tree 40 cm d.b.h.) pink to pale brown, with conspicuous tangential bands of parenchyma, ripple marks prominent. Young branchlets with numerous oval lenticels. Leaves imparipinnate, spirally arranged, rusty sericeous when young but soon becoming glabrous: i^tiole 2-6 cm long, terete, inconspicuously sulcate adaxially; rhachis 2-11.5 cm long; petiolules 5-15 mm long, wrinkled when dried, sulcate above; leaflets (3)5, opposite or subopposite, without stipellae, lamina obovate- elliptic or obo vat e-oblong, 6.5-21 cm long, 2. 5-9.7 cm wide, cuneate basally, acute or acuminate apically, coriaceous, margins entire but slightly undulate, the lateral veins curved, 8-15 on either side of the midrib and forming loops well inside the blade margin, forming an angle of 50-60° with the midrib. Stipules broadly ovate, mostly 3^ mm long, 2-3 mm wide, overlapping and enclosing the terminal bud, rapidly caducous, glabrous within, rusty sericeous outside and with ciliate margins. Inflorescences paniculate, terminal or axillary, 15-25 cm long, much-branched; lateral racemose branches 3-6 cm ♦National Herbarium of Victoria, Birdwood Avenue, South Yarra, Victoria, 3141. tCSIRO Division of Forest Research, P.O. Box 273, Atherton, Queensland, 4883. Muelleria 5(3): 215-217 (1983). 215 216 Fig. 1 . Storckielia australiensis. a — flowering twig, x Vi; b — flower-bud showing imbricate sepals, x 3; c — open flower showing reflexed sepals and erect stamens, x 3; d — stamen showing short filament and anther dehiscing apically, x 6; e — gynoecium, x 6; f — pod, x 14; g — seed, x 1. a-e from Hyland 6357 (QRS 21157); f and g from Hyland 6603 (MEL 601574). 217 long, sparingly to densely rusty sericeous when young but becoming glabrous. Flower- buds ellipsoid. Flowers bisexual; pedicels of open flowers 8-11 mm long, spai'ingly to densely rusty sericeous. Sepals 5, broadly ovate, subequal or the outer 2 slightly larger, 5-8 mm long, 3-6 mm wide, sparingly to densely rusty sericeous within and without, imbricate in bud, reflexed at anthesis. Petals 5{6), elliptic, 7.5-10 mm long, 3-4 mm wide, apricot, alternating with the sepals. Stamens 5(6), free, all fertile, equal, perigynous, alternating with the petals; filaments 0. 8-1.2 mm long, glabrous; anthers basifixed, 4- 5 mm long, opening by lateral apical slits. Ovary sessile, up to 4 x 1.5 mm, compressed, 1-chambered, 4 or 5-ovulate, thinly sericeous basally and on margins but otherwise glabrous; style curved, ± 4 mm long, stigma small, terminal. Pod elliptic-oblong, flat, 5- 11 cm long, 2.8-4 cm wide, with a wing 6-9 mm wide along the upper suture, apparently tardily dehiscent on the forest floor, brown, leathery, with fairly prominent ± transverse venation, glabrous. Seeds irregularly ± reniform, 7-10 mm long, 10-16 mm wide, 1 or 2(-5) per pod, pale chestnut-brown, flattened, wrinkled, albuminous. Germination epigeal, cotyledons orbicular to reniform. Type Collection: Queensland, Cook Distr., Portion 62 Parish of Alexandra, 16°10'S, 145°10'E, 31.viii.l972, B. P M. Hyland 6557(MEL 601493 holo.; isotypes QRS and to be distributed to BRl, CANB, K, L). Selected Specimens Examined (5/10): Queensland — Noah Creek, I6=10'S, 145°20'E, 11.x. 1967, B. P. M. Hyland 1079 (BRI 126227, QRS 51864 & 51865). Portion 62, Parish of Alexandra, 16°10'S, 145°25'E, 19.xii.l972, B. P. M. Hyland 6602 (MEL 601577, QRS 21059-21065); Ibidem, B. P. M. Hyland 6603 (MEL 601574, QRS 21068-21074). E/P 41, Oliver Creek, V.C.L. Noah, 16°08'S, 145“22'E, 3.xi.l977, K. Sanderson 1236 (QRS 17096). Noah Creek, 16“06'S, 145°27'E, L. J. Webb & J. G. Tracey 13228 (BRl 279480). S. australiensis appears to be confined to a very limited area of extremely wet lowland rainforest from about 15 to 25 kilometres north west to north east of Daintree. The annual rainfall is uncertain but apparently exceeds 3.8 metres. When in flower, S. australiensis is very conspicuous and it is surprising that such a large rainforest tree has escaped formal description for so long. The phyllotaxis in S. australiensis can be confusing. At times the leaves appear opposite but close inspection reveals that this effect results from axillary buds producing shoots which outgrow the parent shoot. The occurrence in Australia of a species of Storckiella, a small genus of four species with a disjunct distribution pattern previously known only from Fiji and New Caledonia, is of phytogeographic interest and provides yet another example of the floristic link between these regions. S. australiensis does not appear to be closely related to any of the other species. S. australiensis resembles S. vitiensis Seem, from Fiji in being a large forest tree but the latter differs in having flowers with 10(12) stamens which have filiform filaments 9-13 mm long and leaves with 9-15 leaflets (15)30-130 mm long and (8)15-50 mm wide. S. pancheri Baill. and S. comptonii Bak.f. in New Caledonia differ from S. australiensis in being large shrubs or small trees up to 8 m high and in having 4(5) stamens with filiform filaments up to 8 mm long, leaves with 8-13 leaflets 40-48 mm long and 10-25 mm wide, and smaller pods. The occurrence of a different stamen number in each of the three centres of distribution of the genus is of interest. ACKNOWLEDGEMENTS We are most grateful to Dr M. D. Crisp, National Botanic Gardens, Canberra, for furnishing some information on S. v/t/e/ 15/5 while serving as Australian Botanical Liaison Officer at Kew Herbarium, Royal Botanic Gardens, England; to Miss A. M. Podwyszynski, National Herbarium of Victoria, for preparing the accompanying illustration, and to the Curator, Herbarium Australiense, Canberra, and the Director, Queensland Herbarium, for the loan of specimens. Manuscript received 19 April 1982. ASPLENIUM TERRESTRE AND TWO ASPLENIUM HYBRIDS: NEW FERN RECORDS FOR AUSTRALIA by P. J. Brownsey* ABSTRACT The presence of Asplenium terrestre Brownsey in the Australian fern flora is recorded for the first time. It occurs in Tasmania and parts of Victoria where it has previously been confused with A . bulbiferum Forst. f. and A . flaccidum Forst f. A natural hybrid between A. terrestre and A . bulbiferum is also recorded, and Australian specimens previously referred to A. scleroprium Hombr. are thought to belong to another hybrid combination, A. flaccidum x obtusatum Forst f. NEW RECORDS Asplenium terrestre Brownsey, New Zealand J. Bot. 15: 71 (1977) Asplenium terrestre was first described in a revision of the New Zealand species of Asplenium (Brownsey 1977a). Plants belonging to this species were previously included within y4. flaccidum Forst. f. In New Zealand, A. terrestre is an octoploid species which c^ be distinguished morphologically from the tetraploid A. flaccidum by its more highly divided frond and its more prominently ridged spore pattern. It also differs from A. flaccidum in growing on the ground and having a more or less erect frond, whereas A. flaccidum is commonly epiphytic and has limp, pendulous fronds. In New Zealand, both A. terrestre and A. flaccidum have two distinct habit-forms, recognised at the subspecific level, which are characteristic of forest and coastal habitats respectively. The distinguishing features of all four taxa are fully described and illustrated in my original paper (Brownsey 1977a). When first recognising terrestre as being distinct from A. flaccidum I regarded it as a species endemic to New Zealand. However, subsequent investigation of collections in AD, HO, MEL, NSW and WELT, together with some field observations in Tasmania, has shown that A . terrestre is also present in the Australian flora in addition to the more widespread A. flaccidum. The plants found in Australia match very closely A. terrestre subsp. terrestre from New Zealand and can be distinguished from^. flaccidum suhsp. flaccidum by the same morphological characteristics as in the latter region. I have seen no Australian specimens referable to either A, terrestre subsp. maritimum Brownsey or A. flaccidum subsp. haurakiense Brownsey which appear to be New Zealand endemics. Unfortunately it has not yet been possible to obtain chromosome counts from plants of Australian A. terrestre but the range of mean spore sizes from five Tasmanian populations was found to be 42-45 x 28-31 /^m which is consistent with New Zealand material. Curiously, the range for four Tasmanian populations of A. flaccidum was found to be 42-48 X 27-30/im which is larger than for subsp. flaccidum in New Zealand (36-44 x 23-27 /.an), though within the range for subsp. haurakiense (43-49x26-31 /mi). It is obviously desirable that chromosome counts should be obtained from Australian material to confirm that the New Zealand and Australian species are the same. A . terrestre is confined in Australia to Tasmania, some of the Bass Strait islands and a few localities in southern Victoria. In Tasmania it appears to be more common than A. flaccidum, which is restricted to the northern half of the island. Unlike A. flaccidum which is normally epiphytic, A . terrestre grows on the ground, on damp rock faces, or at the bases of trees. It occurs in wet forest habitats and becomes progressively rarer ♦National Museum, Private Bag, Wellington, New Zealand. Muelleria5(^): 219-221 (1983). 219 220 northwards. It has been collected from Flinders Island and Cape Barren Island in Bass Strait, and from Mt. Mueller, Byaduk Caves and the Grampians in Victoria. In Australia, plants of A . terrestre seem to have been confused more frequently with A. bulbiferum Forst. f. than withv4. flaccidum. Asplenium bulbiferum and^. terrestre often grow in close association but the latter can be distinguished by its thicker, more leathery fronds, by the absence of bulbils and by its longer sori (2-7 mm cf. 2-4 mm). Representative Specimens Examined: Victoria {^/'^) — Grampians. Victoria Range, Castle Rock, 6.xi.l966, A. C. Beauglehole ACB 15883 (MEL 532857); Byaduk Caves, Church Cave, 2. i. 1956, /I. C. Beauglehole ACB 42827 {MEL 1502290); Mt. Mueller, nr Mt. Baw Baw, 1889. J. Melvin (MEL 99457-8). Tasmania (4/30) — Cape Barren Is., 28.iv.1967, J. S. Whinray 480 (MEL 527961); Weldborough, 17.xii.l962, T. & J. Whaite 2644 (NSW s.n.); Liena to Lorinna Track, W. of Mole Creek, 4. i. 1941, J. Somerville (HO 243); Betts Falls Track, Mt. Wellington, 3.viii.l981, P. J. Brownsey (WELT P11095). Asplenium bulbiferum x terrestre Hybrids between >1. bulbiferum and A . terrestre have been collected at least twice in Australia and can be recognised by their intermediate morphology and aborted spores. The capacity of related Asplenium species to hybridise suggests that this combination is likely to occur wherever the two parent species grow in close proximity. A. bulbiferum x terrestre was not amongst the nineteen different hybrid combinations initially recorded for New Zealand (Brownsey 1977b). However, it has now been found in parts of Canterbury in the South Island. Specimens Examined: Victoria {\/\) — Grampians, Victoria Range, Castle Rock, 12.ii.l960, A. C. Beauglehole ACB 20983 (MEL 532858). Tasmania (1/1) — Betts Falls Track, Mt. Wellington, 3.viii.I981. P. J. Brownsey (WELT PI0972-3). Asplenium flacddum x obtusatum A. scleroprium Hombr. is a species that occurs only in the New Zealand region where it is confined to coastal areas, often growing alongside A. obtusatum Forst. f. on cliffs and in scrub exposed to salt spray (Brownsey 1977a). In Australia it has been recorded from Wilson’s Promontory and adjacent islands in Victoria (Ewart 1931; Wakefield 1940; Jones & Clemesha 1981, ut A, aucklandicum (Hook, f.) Crookes), but these Victorian records now appear erroneous. I have been able to locate only two Australian collections which in any way resemble New Zealand plants of A. scleroprium. Spores from these specimens are all aborted indicating that the plants are almost certainly of hybrid origin. Their morphology suggests that they may be hybrids between A. flacddum and A. obtusatum^ a combination recorded once in New Zealand from Stewart Island (Brownsey 1977b, fig. 11). The sterile hybrids, from both New Zealand and Australia, are very similar to A. scleroprium and can really only be distinguished by their aborted spores. The similarity is so striking that I postulated that A. scleroprium may have arisen in New Zealand as a new allopolyploid species from such sterile hybrids (Brownsey 1977b, fig. 26). If that were the case, there is no reason why the species should not ^so have arisen independently in Australia. However I have seen no evidence that A. scleroprium is present in the latter country. Specimens Examined: Victoria (2/2) — Biddy’s Camp, Wilson’s Promontory , xi.l927, A. J. Ewart (MEL 99454); Rabbit Islands, Wilson’s Promontory, 5. i. 1939, E. Rossiter {MEL 99449). ACKNOWLEDGEMENTS I am extremely grateful to Dr A. E. Orchard (HO) for first drawing my attention to plants of Asplenium terrestre in Tasmania and for considerable assistance with field work around Hobart in July 1981 . I would also like to thank the Directors of HO and MEL for the loan of collections. 221 REFERENCES Brownsey, RJ. (1977a). A taxonomic revision of the New Zealand species of Asplenium. New Zealand J. Bot. 15: 39-86. Browsey, P. J. (1977b). Asplenium hybrids in the New Zealand flora. New Zealand J. Bot. 15: 601-37. Ewart, A. J. (1931). ‘Flora of Victoria’. (Melbourne University Press). Jones, D. L. and Clemesha, S. C. (1981). ‘Australian Ferns and Fern Allies’, ed. 2. (Reed: Sydney). Wakefield, N. A. (1940). A new Victorian fern record. Viet. Nat. 57: 114-116. Manuscript received 19 June 1982. GREVILLEA MONTIS-COLE SP. NOV. (PROTEACEAE) FROM VICTORIA by R. V. Smith* ABSTRACT Grevillea montis-cole^ a new species from western Victoria, is described and its affinities discussed. Two subspecies, montis-cole and brevistylOy are distinguished. DESCRIPTION Grevillea montis-cole R V. Smith, sp. nov. Frutexzd 1.5 m altus x3.6 m latus. Tomentum et ramorum superorum et foliorum et inflorescentiarum plerumque pilorum patentium vel rectorum vel curvatorum vel torquatorum. Folia (l-)2-7(-13.5) cm longa x (l-)2-5(-7.5) cm lata, plerumque longiora quam lata, in (3-)5-12(-22) lobis primariis, ± triangularibus ad ellipticis-oblongis profunde dissecta, qui aut simplices aut 1-5 lobos secundarios breves ± triangulares habent; quisque lobus primarius secundariusque in aculeum gracilem rigidum terminans. Superficies supera foliorum adultorum vivide viridis, sub- nitens, paeneomnino glabra; superficies inferapallide viridis, hebetata, pilis sparsis plerumque in costa et venis principalibus. Inflorescentia racemus secundus, horizontalis ad parum deflexus, 0.6-4 cm longus. Pedunculus teretus ad parum angulatus, 0.5-3. 5 cm longus. Bracteae floris maxime variabiles, vel planae vel curvatae vel undulatae, erectae patentesve, ellipticae ad ovate- rhombeae, 1.5-4.5(-5) mm longae x l-3(-3.5) mm latae. Pedicelli (1.0-)1.5-4.5(-5) mm longi. Perianthium 6-9(-10) mm longum (de medio toro ad summum arcu) x 1.5-3 mm latum, extus pilosum, intus glabrum, viride ad hinnuleum, arcu atropurpureo. TomsvdXdQ obliquus ad paene rectus, nectario semiannulari prominenti. Stipes 1-3 mm longus. Ovarium manifeste stipitatum, dense pilosum, pilis longis. erectis ad patentibus. SO'/ws vel 9-13 vel 17-21 mm longus, vivide ruber, vel rectus vel curvatus vel flexus, glaber praeter in base. Fructus manifeste stipitatus, 8-14 mm longus X 3-6 mm latus, dense pilosus, pilis brevibus longisque intermixtis. Shrub 0.6-1. 5 m high x 0. 9-3.6 m wide, decumbent to semi-erect. Distal parts of branches ribbed or angled, sparsely to moderately hairy with straight, curved, waved or twisted hairs, the hairs sometimes loosely appressed but generally strongly spreading. Tips of branches and young leaves densely hairy with ferruginous or reddish-purple hairs. Leaves: petioles (2-)3-10(-12) mm long, with a sparse to dense indumentum similar to that of the distal branch-parts; blades broad- to narrow-triangular, truncated-ovate or truncated-elliptic in outline, ± truncate to cuneate at the base, (l-)2-7(-13.5) cm long x (l-)2-5(-7.5) cm wide with length (<-)= or > width, deeply divided into (3-)5-12(-22) primary lobes; primary lobes 0.5-2. 5(-4) cm long, ± triangular to elliptic-oblong, symmetrically placed in opposite pairs or arranged asymmetrically, either simple or bearing 1-5 short ± triangular secondary lobes, each ultimate lobe terminating in a slender rigid prickle 1-2.5 mm long. Upper surface of mature leaves bright green, subshiny, almost glabrous except for a few basal hairs; lower surface pale green, dull, sparsely hairy with curved waved or twisted hairs mostly on the midrib and main veins; main midrib and midribs of primary lobes prominently projecting on lower surface. Young leaves mostly glabrous above. Flowers in horizontd to deflexed secund racemes (0.6-)1.5-3.5(-4) cm long x (1.5-)2-3(-4) cm wide, terminating a terete to somewhat angled peduncle (0.5-)0.7-3(-3.5) cm long. Peduncle hairy or partly glabrous, usually with a single bract arising from below to well above the midpoint, occasionally a second bract also present; peduncle usually bent or geniculate at the bract. Bract 2-5(-7) mm long, shortly hairy and gently keeled on back, glabrous ventrally, flattish to infolded (sometimes upper margins infolded and the lower spreading or slightly recurved), broad- to narrow-lanceolate with an acute shortly subulate tip, or the tip occasionally trifid or expanded into an incipient leaf with a sm^ lobed lamina. Rhachis with an indumentum *National Herbarium of Victoria, Birdwood Avenue, South Yarra, Victoria 3141. Muelleria 5(3); 223-227 (1983). 223 224 Fig. 1. Grevillea montis-cole ssp. montis-cole. a — branch showing leaves and flowering racemes, x I. b — young flower with pollen presenter still enclosed and with floral bract at base of pedicel, x 4. c — floral bracts showing variation in shape, x 8. d — fruiting raceme, x 1. e — mature flower, x 4. f — mature fruit, showing dark longitudinal bands and patches on dorsal and lateral surfaces, x 3. g — section through mature flower, showing oblique torus, stipitate ovary, style and pollen presenter, x 4. a-c, e and g from holotype; d and f from Smith 73/66 (MEL). 225 similar to that of the distal branch-parts. Floral bracts very variable, flat curved or undulate, erect spreading or sometimes reflexed with maturity, elliptic to ovate-oblong to broad-rhombic, from almost truncate and suddenly contracted into a blunt tip to broad- or narrow-acute, 1.5-4.5(-5) mm long x l-3(-3.5) mm wide with length < to > width, densely to sparsely hairy dorsally with short ± spreading hairs, glabrous ventrally, greenish-yellow to reddish-tinged, from sometimes caducous (faUing early from flowers) to persistent even in fruit. Pedicels (l-)1.5-4.5(-5) mm long (from base to midpoint of torus), straight to curved, spreading to strongly reflexed, from c. half as long as to much exceeding the floral bracts, resembling the ^stal branch-parts in vestiture. Perianth 6-9(-10) mm long (from midpoint of torus to top of arch) x 1.5-3 mm wide, sparsely to densely hairy outside with a mixture of loosely appressed straight or gently waved longitudinally directed hairs and longer spreading waved or twisted ones, glabrous inside; hairs whitish, or those of the limb reddish or pale ferruginous; body of perianth greenish to fawn, the arch blackish-purple or almost black or sometimes light purple, especially inside; external colour partly obscured by the indumentum. Anthers (0.7-p.8-1.2(-1.3) mm long, lemon yellow to pale creamy yellow. Torus oblique to almost straight with a prominent semi-annular to bluntly V-shaped nectary. Stipe 1-3 mm long, sparsely to densely hairy, with ± straight, appressed to spreading, greyish-white to pale ferruginous hairs; stipe arising close to summit of torus, and from slightly shorter to considerably longer than the ovary. Ovary prominently stipitate, densely hairy with long, erect to spreading, straight to curved or slightly twisted greyish-white to pale ferruginous hairs. Style on flowers (9-)10-12(-13) mm long in ssp. brevistyla or 17-21 mm in ssp. montis-cole, bright red, straight curved or bent, glabrous except at base where sparsely to strongly hairy, the hairs frequently extending up style for several mm; fruiting styles generally 1-6 mm longer than the flowering styles at least in ssp. montis-cole. Pollen-presenter (stigmatic disc) greenish yellow, slightly to very oblique, with a broad undulate-crenulate rim rising to a low oblique cone terminated by a rounded or flattish pale cream-yellow stigmatic tip. Fruit prominently stipitate, (8-)9-12(-14) mm long x (3-)3.5-5.5(-6) mm wide, asymmetrically ellipsoidal, densely hairy with a mixture of ± appressed to spread- ing straightish hairs and much longer spreading, straight to gently twisted ones; hairs greenish-grey to fawn (sometimes light ferrugineous when dry), but blackish-purple in longitudinal bands or patches on the dorsal and lateral surfaces. Seed body elliptic in outline, 6.5-8. 5 mm long x 2-3 mm wide, dark brown to black, glabrous, smooth or minutely wrinkled, gently convex on outer face, ± flattened or depressed on inner face, surrounded by a thin pale yellow wing from 0.5 mm to almost obsolete on the lateral margins, to 1-2.5 mm wide at top and bottom. subspecies montis-cole Stylus floris 17-21 mm longus est. Description and measurements are as given above for the species, except for the restriction of the flowering style length to 17-21 mm. Leaves are mostly longer than wide. Type Collection: Mt Cole State Forest, Glut area, east aspect slopes close to Sandersons Road — Glut Road link, c. 400 m from Glut Road, c. 520 m alt., central-western Victoria (Grid J.14), 9.xi.l965, R. V. Smith 65/183 (Holotype: MEL 611505. Isotypes: AD, CANS, K, MEL 611506 and 611507, NSW). Selected Specimens Examined: Victoria (central-western; Mt Cole State Forest) — North slopes of Mt Buangor, c, 900 m alt., 16.xi.l970, R, V Smith 70/33 (AD, BRI, CANB, K, MEL 611518 and 9, NSW, PERTH); Glut area, Sandersons Road, c. 520 m alt., 5.xii.l973, R. V Smith 73/66 (AD, BRI, CANB, HO, MEL 611516 and 7, NSW, PERTH); Glut area, near Tunbridges Road, c. 580 m alt., 5.xii.l973, R. V. Smith 73/69 (BRI, HO, MEL 611512 and 3, NSW); Saddle between Cave Hill and Mt Sugarloaf, c. 700 malt., 5.xii.l973, R. V. Smith 73/72 (AD, CANB, MEL 611508 and 9, NSW). Distribution and Habitat: Apparently confined to the Mt Cole State Forest, north-west of Beaufort, in central- 226 western Victoria. It occurs on granite formation within an altitudinal range from c. 520 m in the Glut area to c. 900 m on the north slopes of Mt Buangor. This subspecies occurs in a number of different plant associations which include Eucalyptus obliqua, E. sp. (Peppermint), E. nitida, E. st-johnii. Acacia dealbata, A. diffusa, A. myrtifolia, A. verticillata, Bursaria spinosa, Correa reflexa, Epacris impressa, Helichrysum dendroideum, Platylobium obtusangulum, and Pteridium esculentum at the lower elevations; at the higher elevations the associations include Eucalyptus pauciflora, E. st-johnii, E. dives, E. macrorhyncha, E. nitida, E. goniocalyx. Acacia diffusa, A. oxycedrus, Astrotricha sp., Daviesia ulicifolia, Dodonaea cuneata, Epacris impressa, Helichrysum obcordatum, Kunzea parvifolia, Leptospermum juniperinum and Pultenaea daphnoides. subspecies brevistyla R. V. Smith, ssp. nov. Differt subsp. montis-cole, stylo floris (9-)10-I2(-13) mm longo. Description of flowering material essentially as for the species except for the restriction of the flowering style length to (9-)10-12(-13) mm. Fruits and seeds not seen. Leaves vary from about as wide as long to a little longer than wide. Younger and intermediate leaves are somewhat more pubescent below than in the ssp. montis-cole. Measurements of plant parts usually fall within the lower limits of those given for the species. Type Collection: Mt Langi Ghiran, north facing slope on NE. shoulder along water supply pipe track, c. 850-900 m alt., western Victoria, 37°17'S, 143°07'E, 31.x. 1981, M. G. Corrick 7493 (Holotype: MEL 603607! Isotypes: ADI, CANBI, Kl, MEL 6036081, NSW n.v., PERTH!). Specimens Examined: Victoria (western; Mt Langi Ghiran) — Ibidem for type collection, 31.X.I981 M. G. Corrick 7494 (CANB!, MEL 603609!, NSW n.v.). Distribution and Habitat: Apparently confined to Mt Langhi Ghiran in western Victoria, and there known only from a single population on the north-east shoulder. As with the ssp. montis-cole, it occurs on granite formation and apparently occupies a similar habitat to that of the typical subspecies at Mt Buangor. Associated plants recorded are Dodonaea cuneata, Kunzea parvifolia, and Astrotricha sp. DISCUSSION Grevillea montis-cole shows closest affinities with G. steiglitziana N. A. Wakefield and G. floripendula R. V. Smith. These three species have very similar floral bracts, which are either flat, gently curved or undulate, but not concave, and vary in shape from elliptic to ovate, ovate-oblong, or rhombic. Distinguishing characters are given in table 1. The ssp. brevistyla shows some approach to G. floripendula in the short style, the leaves varying from wider than long to a little longer than wide and also in the tendency to show a noticeable pubescence on the under surface of the leaves, but in all other respects it accords well with G. montis-cole montis-cole. G. montis-cole differs from G. dryophylla and G. microstegia primarily in the floral bracts. The latter two species, discussed under G. floripendula in Muelleria 4:426 (1981), have floral bracts which are strongly concave, thickened in the lower part and from ovate-rhombic to broad-rhombic. G. montis-cole ssp. montis