MUELLERIA VOL. I, No. 2. DECEMBER 15, 1959. 4761/69. 4 * ' 'Vj - y'*J * ‘^^'T fV , J “ -vfe '■ . ■■ -;■' Page Australian Species of the Fungal Genus Cordyceps — J. H, Willis . , 67 Orthography of Certain Specific Epithets — J, H. Willis . . , . 90 Reduction of the Lichen Genus Bibhya J. H. Willis — J, H, Willis . . 91 Notes on the Vegetation of Eucla District, W.A. — J, S'. Willis . . 92 Plants of the Recherche Archipelago, W.A. — J, H. Willis . , . . 97 New Species and Varieties of Ptilotus R.Br. (Amaranthaceae ) — Gerhard Beni . . . . . . . . . . . . . . . . 101 Reinstatement of Calotis sujfruticosa Domin (Oompositae ) — Gwenda X. Davis . , . . . . . . . . . . . . . . 109 Two New Australian Species of Brachycome Cass. (Oompositae ) — Gwenda L, Davis . . . . . . . . . . . . . . Ill Studies in Mimosacece Part 1 — A, B. Court 114 MUELLERIA An Australian Journal of Botany VOL 1, No. 2 - DECEMBER 15^1959. NATIONAL HERBARIUM ROYAL BOTANIC GARDENS, MELBOURNE VICTORIA, AUSTRALIA R. T. M. PESCOTT, Director and Government Botanist 4751/59. CONTENTS Page Australian Species of the Fungal Genus Cordyceps — J. H. Willis . . 67 Orthography of Certain Specific Epithets — J. H. Willis . . . . 90 Reduction of the Lichen Genus Bihbya J. H. Willis—./. H. Willis . . 91 Notes on the Vegetation of Eucla District, W.A. — /. H. Willis . . 92 Plants of the Recherche Archipelago, W.A. — J. H, Willis . . . . 97 New Species and Varieties of Ptitotus R.Br. (Amaranthaccae ) — Gerhard Idenl . . . . . . . . . . . . . . . . 101 Reinstatement of Calotis suffridicosa Domin (Compositae ) — Gwenda L. Davis . . . . . . . . . . . . . . . . 109 Two New Australian Species of Brachycome Cass. (Compositae ) — Gwenda L. Davis . . . . . . . . . . . . . . Ill Studies in Mimosacece Part 1 — A. B. Court . . . . . . . . 114 STANDARD iVBBREVIATIONS FOR HERBARIA. MEL National Herbarium, Royal Botanic Gardens, Melbourne. NSW National Herbarium, Royal Botanic Gardens, Sydney. BRI Botanic Museum and Herbarium, Botanic Gardens, Brisbane. PERTH State Herbarium of Western Australia, Perth. AD State Herbarium, Botanic Garden, Adelaide. HO Herbarium of Tasmanian Museum, Hobart (at Botany Dept., University). CANB Herbarium, Division of Plant Industry, C.S.I.R.O., Canberra K Herbarium, Royal Botanic Gardens, Kew, England. BM British Museum of Natural History (Dept. Botany), London. AUSTRALIAN SPECIES OF THE FUNGAL GENUS CORDYCEPS (Fr.) Link (with critical notes on collections in Australian herbaria) by J. H. Willis, National Herbarium of Victoria. SUMMARY This paper embraces: a short history of the discovery of Cordyceps species in Australia; an artificial key to the 15 taxa (known to occur in the Commonwealth and believed to be specifically distinct); the synonymy for each species, list of collections represented in Australian herbaria, location of types, names of host-insects (where known) and sundry comments on identities and nomenclature; a bilbliography concerning Australian collections of Cordyceps. INTRODUCTION Specimens of Cordyceps are no less intriguing to the modern botanist than when, two centuries ago, they were regarded as a lusus naturce — the transmutation from animal to plant kingdom. The permanency, with which “ mummified ” bodies of the host-insect retain their outward form, heightens the interest of these parasitic fungi. Another remarkable fact is that certain species may occur either in the perithecial (perfect) phase or in one or more quite dissimilar conidial states. Some kinds of entomogenous fungi produce only conidial fruiting bodies; these are classified under such form-genera as Isaria, Hirsutella, Hymenostilbe, Sporotrichum, Stilbum, &c., although their relationship (and perhaps identity) with undoubted members of Cordyceps is suspected. The only comprehensive world revision of this genus since the 19th century is by Y. Kobayasi (1941) who recognized 137 species. Japan (34 spp.). North America (31), Brazil (31), Belgian Congo (24) and Java (17) appear to be the richest centres of development. Australia (including Tasmania) is credited with only ten species, according to Kobayasi; but these include C. taylori (Berk.) Sacc. which is most probably the largest “ vegetable caterpillar ” fungus in the world, its branched sporophores often attaining heights up to 1 foot. Kobayasi notes (1941, p. 216) that almost half the total species of Cordyceps are known solely by the original collection, which would suggest that most species are of quite rare occurrence; however, the lack of systematic collecting doubtless accounts for an astonishing paucity 4751/59.— 2 Pi 68 J. H. Willis: Australian Species of the Fungal Genus Cordyceps I of herbarium specimens — at least in Australia. One or more species is known to occur in each of the Australian States except Western Australia, I whence material of an isarioid growth on large cicada larvae (but apparently not Isaria cicadce Miq.) reached me during 1953. The eventual discovery I of several Cordyceps species in such a vast region as the Western State is only to be expected. I At least three different Isarice have been recorded from insect larvae in eastern Australia, viz.: /. surmatodes McAlpine (Sept., 1895) on a cockchafer beetle at St. Kilda, Vic.; /. cicada; Miq. [in Bull. Sci. phys. et nat. Neerl. 1: 85, t. 1, fig. A (1838), also^nn. Sci. nat. ser. 2, 10: 378 ^838)— type from Bahia, Brazil] on a Victorian cicada specimen, but without definite locality, and /. suffruticosa Cooke & Massee (1889) on a hairy caterpillar in New England district, N.S.W. The two latter records were published by M. C. Cooke (Aug., 1892, p. 383), and H. T. Tisdall (1893, p. 94) discusses an unidentified Isaria found on a moth cocoon under a charred log at I Nar-Nar-Goon, Vic. * It is now possible to extend the range of /. cicadce to Queensland, where F excellent fruiting material was collected on the ground in eucalypt forest at Mt. Nebo, Manorina National Park, by Mr. and Mrs. A. B. Cribb in February, 1956. The highly branched conidiophores all emerge from the head of the small host-cicadas (apparently Melampsalta sp.) and the flesh-pink to , orange-reddish fertile heads (each about 2 mm. long) are aggregated like , bunches of grapes. Under I. cicadce. Fetch (1933, p. 65) lists as a synonym F sinclairii (Berk.) Lloyd, 1923, which is not uncommon in New Zealand, If' its type (at Kew) having come from Tauranga on Poverty Bay. Dingley » (1953, p. 339) is not correct in writing “ Fetch stated that the perfect stage M was Cordyceps soblllfera (Hill.) Sacc.” Fetch’s actual words (1933, p. 64) P were: "" Isaria sinclairii is generally accepted as the conidial stage of fjj Cordyceps sobolifera, though no evidence of that relationship has been H adduced, except that both grow on Cicada nymphs ”. m Type of Cordyceps sobolijera (Berk.) Berk. & Broome, 1875, came H from the West Indies and is, presumably, in the cryptogamic section of the R Natural History Museum at Paris. This species, with usually solitary clubs ll' surrounded by sterile, conidiophore-like “ soboles ”, extends to Mexico, Japan, China, Ceylon and Madagascar; so it might well be represented in the I Australian region. i H. Tryon (1893, p. 54) reported the occurrence of a hyphomycetous R fungus — probably Isaria sp. — on a mummified scarabaeid beetle larva at Yandilla, Queensland, while the present writer has found white but barren isarioid growths (to 1 cm. high) all over the brown-woolly cocoon of a grass moth (Anthela acuta Walk.) at Beenak, Vic. (May, 1935). The latter may possibly represent I. suffruticosa Cooke & Massee, to which I. japonica Yasuda ex Lloyd, 1917, is very closely allied if not conspecific. Mrs. K. Healey has recently found excellent examples of the same parasite on woolly moth cocoons at Tarra Valley National Park, Vic. (June, 1959). J. H. Willis: Australian Species of the Fungal Genus Cordyceps 69 Cordyceps is known to parasitize a wide diversity of hosts, including larvse of numerous genera of Coleoptera (beetles), Hemiptera (bugs and cicadas), Orthoptera (cockroaches and crickets), Lepidoptera (moths and butterflies), Hymenoptera (ants and wasps) and Diptera (flies). It is usual for fruiting bodies to emerge just behind the head of the host, but occasionally they appear between the anal or intermediate segments. Occurrences of Cordyceps on spiders are comparatively rare (e.g., the Japanese C. arachneicola Kobayasi, and tropical American C. caloceroides Berk. & Curt.), while a few boreal species — C. japonica Lloyd, C. jezoensis Imai, C. ophioglossoides (Fr.) Link, C. intermedia Imai and C. capitata (Fr.) Link — are restricted to underground fruiting bodies of the fungal genus Elaphomyces. Frequently each Cordyceps is peculiar to a single host or to several related species, yet the almost cosmopolitan C. militaris (Fr.) Link has been recorded for at least 13 genera of Lepidoptera, as well as on coleopterous pupae and the cocoons of certain Hymenoptera. Conversely, the large Victorian swift moth larva Oxycanus diremptus is known to be parasitized by at least four species of Cordyceps: C. gunnii (Berk.) Berk., C. hawkesii G. R. Gray, C. cranstounii Olliff and C. robertsii (Hook.) Berk, were all found on this grub within an area of a few square yards along Koonung Creek at Doncaster, Vic., by Mr. and Mrs. Paul Fisch in June, 1942, and more recently (June, 1959) the same four species were gathered on Oxycanus at Tarra Valley National Park, Vic., by Mrs. K. Healey {vide specimens in Melbourne Herbarium). The writer has personally examined all Cordyceps exsiccatae housed in the National Herbaria at Melbourne and Sydney, the National Museums of Victoria and South Australia, the considerable material in the Botany Department of Melbourne University (amounting to 22 collections, including types of three species) and, by courtesy of Dr. N. T. Flentje (pathologist), eight of the twelve collections of Cordyceps in the Waite Agricultural Research Institute at Adelaide. He is also indebted to Dr. Winifred M. Curtis., Mr. R. C. Carolin and Mr. A. Musgrave (through Director J. W. Evans) for valuable information on the material housed in the respective herbaria of Hobart University, Botany Department at Sydney University and the Australian Museum (Sydney). There are no Australian specimens in the State Herbaria at Perth, Adelaide or Brisbane. Special thanks are due to Dr. Alan B. Cribb (Botany Department, University of Queensland, Brisbane) who willingly placed at my disposal the four Cordyceps collections (1954-56) of his own fungal herbarium, and gave permission to publish any pertinent details. These collections (all by Cribb and his wife) comprise four distinct species — C. gunnii, C. hawkesii, C. robertsii and Jsaria cicadee Miq. — none of which have been recorded previously for Queensland; indeed, the genus Cordyceps would seem to have been hitherto quite unknown in that State. All three species came from rain forest along Coomera track in the Lamington National Park — presumably on the same large hepialid moth larva (as happened at Doncaster, Vic., in 1942). It is possible that some other important collections exist in private herbaria; but, if so, details have eluded me. 70 J. H. Willis: Australian Species of the Fungal Genus Cordyceps RECORDING OF SPECIES FOR AUSTRALIA Apparently the first species to be discovered in the Australasian region was C. robertsii, diagnosed as a Sphaeria by W. J. Hooker (1836) from New Zealand material, but later recorded for Tasmania (Rodway, 1920), found in Victoria (1942), in New South Wales (1947)* and Queensland (1955). Kobayasi (1941, p. 102) extends its range to Chile. The host in each instance is a large lepidopterous larva in the Swift Moth family, Hepialidce (e.g., Oxycanus diremptus and probably O. fusco-maculatiis) . A second Australasian species was collected on the banks of Murrumbidgee River near Yass, N.S.W., in 1837; it was first described and figured as Sphaeria innominata by Rev. R. Taylor (1842), but has been universally known under the name Cordyceps taylorl for the past century — the inadmissibility of the latter, and probably both epithets, will be discussed elsewhere. In 1896 C. taylori was rediscovered at Queanbeyan, less than 40 miles south-east from the type locality; in the meantime, however, no less than six independent collections had been made in the Otway Ranges, Vic. (1886-94), two between the Ovens and Mitta Mitta Rivers, two in the Strzelecki Ranges of South Gippsland and one labelled “ Caulfield ” (1870). This large and remarkable ascomycete is still known only from Victoria and south-eastern New South Wales. The third Australian taxon to receive recognition was C. gunnii, described under the genus Sphaeria by Rev. M. J. Berkeley (1848) on the basis of a Tasmanian collection from Launceston district — not “ Lancaster " as in the original description. This is the most widely distributed and by far the best known Australian species, occurring in all States except Western Australia and extending also to the North Island of New Zealand. Its large, usually simple, dark olive-green heads appear early in winter, mostly under Acacia trees (e.g., the Silver Wattle, A. dealbata), and fruiting bodies are well represented in herbaria. The first Victorian specimen would seem to be one taken at Studley Park, Melbourne, in 1857 by W. Kershaw and now in the possession of his great-grandson (R. C. Kershaw of West Tamar, Tas.); strangely enough, F. Mueller does not seem to have noticed the species before 1874, when it appears in his Report as Government Botanist (p. 12). Subsequent collections have been made almost throughout the cooler parts of the State — from Portland to Mallacoota. C. gunnii was collected in New South Wales at least as early as 1888 ( Shoalhaven River ) , on the Hunter River, then at several parts of the Blue Mountains, and later still in the National Park (south of Sydney) and in the Riverina district. Tepper and McAlpine (1897) recorded the species for South Australia on the basis of several collections from Kingston and Sellick’s Hill. By courtesy of the Director, South Australian Museum, the present writer has been able to examine three of these collections and he agrees with McAlpine’s identifications of 1897. Almost half a century later (in 1942 and 1943) C. gunnii was collected at Kingston again and * Several collections from the Port Jackson area were recorded under the name of Cordyceps selkirkii by Olliff (1895). J. H. Willis: Australian Species of the Fungal Genus Cordyceps .71 the record published, together with a coloured illustration, by J. R. Harris (1946). No other species of the genus is known to occur in South Australia. The first, and only, Queensland collection would seem to be that of A. B. and J. W. Cribb — from Lamington National Park in the far south-east. May 1955. C. hawkesii was described (as a Sphieria) from Tasmania by G. R. Gray (1858). The type, presumably at British Museum (Natural History), has not been consulted; but similar, or perhaps identical, material was collected on the Snowy River near Orbost, Vic. (1890) — subsequently at Doncaster, Cockatoo, Olinda, &c. — at Coonamble, N.S.W. (about 1894), and lastly in the Lamington National Park, Queensland (May, 1954). F. Mueller and Berkeley (1878) published a description and figure of C. meneristitis (the epithet erroneously rendered “ menesteridis their type coming from the mouth of Yarra River, Vic.; M. C. Cooke (Aug., 1892) reduced it to a variety of the European C. entomorrhiza Link, and Kobayasi (1941, p. 140) made it a doubtful synonym of his own new Japanese species, C. gracilioides. The Stilbum formicarum, described by Cooke and Massee (1889) from an ant collected at Cheltenham, Vic., was considered by T. Petch (1933, p. 67) as referable to a conidial state of the Brazilian C. bicephala Berk. Only three species of Cordyceps were accorded an Australian distribution by Saccardo (1883) and Cooke (Aug., 1892) in their respective fungus floras of the world and of Australia. * Saccardo omitted C. hawkesii— probably not having seen Gray’s private and rather obscure pamphlet of 1858 — and the latter authority, for some inexplicable reason, failed to mention C. taylori. D. McAlpine (1895) listed the four species known from Australia at that date (viz., C. taylori, C. hawkesii, C. gunnii and C. entomorrhiza var. "'menesteridis”)', he also admitted typical C. entomorrhiza for Victoria. The next comprehensive account of Australian species appeared in June, 1895, by A. S. Olliff, Government Entomologist of New South Wales, who died six months later at the early age of 30. He provided illustrations and diagnoses for six species, believed to be new, making a total for the Commonwealth of ten. Although this attempt to name several apparently undescribed entities was praiseworthy enough, Ollitf was not a botanist and his revision called forth adverse criticism. L. Rodway (1900) remarked: Mr. Olliff . . . drew attention to the forms found in Australia, at the same time describing many forms as new species. I doubt if mycologists will accept them all. C. selkirkii and C. coxii are too close to C. larvarum [=: c. robertsii] and C. trictena owes its existence to an unfortunate oversight. . . . And C. G. Lloyd’s opinion (1920) concerning the same revisionist was expressed thus: From the systematic account of Cordyceps, however, Mr. Olliff seems to us to be very local in his view, and his species, we believe, should mostly be referred to others. •Saccardo later listed C. hawkesii in Sylloge Fungorum 9 {Supplementum Universale)-. 1001 (1891). 72 J. H. Willis: Australian Species of the Fungal Genus Cordyceps One of Olliif’s six “new” species (C. selkirkii) is almost certainly a form of C. robertsii, another (C. pieli) seems to be merely C. hawkesii, while a third (C. trictence) is actually based upon the type illustration of C. taylori (published as Sphaeria taylori Berk., 1843)! The remainder (C. scottiana, C. coxii and C. cranstounii) appear to be genuine novelties, worthy of specific rank, but their types have not been located and their circumscription must rest upon the original figures and rather inadequate descriptions. A possible type specimen of C. coxii is preserved in the Insect Gallery at the Australian Museum, Sydney; this has not been examined by the present writer. C. scottiana is not known other than by the type collection (Hunter River, N.S.W., 1861) which may be among Berkeley’s numerous fungi at Kew; the original specimens of both C. coxii and C. cranstounii came from Kurrajong Heights, N.S.W.; but collections conforming well to the descriptions of these entities have been made in Victoria during the past twenty years. Also in 1895 G. Massee published another new species, C. henleyte, the type coming from Ovens River, Vic., in 1893. E. Cheel, as reported by Lloyd (1920), made the suggestion that C. henleyee was merely a branched condition of C. robertsii; it seems to connect this species with C. taylori, and, in my opinion, would be better placed as a form of the latter (with which it shares the same very large hepialid host-larva). In C. G. Lloyd’s Synopsis of the Cordyceps of Australasia (1915), C. henleyee, C. taylori, C. gunnii, C. dovei [see below] and C. gracilis are considered as “ good ” Australian species; to the last-named Algerian species Lloyd referred C. meneristitis F. Muell. & Berk. During the present century five species have been added to the Cordyceps flora of the Commonwealth — four from Victoria and one from Tasmania: C. dovei Rodway (1900) was presented from Mt. Bischoff, western Tasmania, and later noted in the North and South Islands of New Zealand; C. furcata McLennan & Cookson (1923) and C. brittlebankii McLennan & Cookson (1926), both described from Ringwood, Vic.; C. aphodii Matthieson (1949) from Miner’s Rest, near Ballarat, Vic.; and what appears to be a form of the boreal genotype, C. militaris Link, from chrysalids in the Otway Ranges, Vic. (first noted in Sept., 1935, and collected there again during the next two years). From time to time, there are found in Australia Cordyceps fruiting bodies which do not satisfactorily match the existing descriptions, and it is likely that a number of endemic species still awaits recognition — especially those on small or little-known insects. Joan M. Dingley’s monograph (1953) on the Hypocreales of New Zealand includes eight species of Cordyceps (three being new to science) and is the most recent treatment for a large region in the Southern Hemisphere. She assigns Cordyceps, and five other genera, to the family Clavicipitacece. The three species (C. robertsii, C, gunnii and C. dovei) common to Australia and New Zealand are described in detail. Probably one or more of the four species, at present considered endemic in that Dominion, may extend to Australia. J. H. Willis: Australian Species of the Fungal Genus Cordyceps 73 ARTIFICIAL KEY TO AUSTRALIAN CORDYCEPS SPECIES. 1. Perithecia entirely immersed in stroma 2 . Perithecia superficial or very prominently exserted- 6 ^ _ 2 Fertile portion of sporophore terminal, orange, obtuse, unbranched, 2-20 mrn. long (on chrysalids of various Lepidoptera) C. militaris Fertile portion not terminal, never orange; apex of sporophore sterile, often slender (on larvae) — ^ 3 3. Stroma 3-10 mm. long, 1-2 mm. wide, pale ochraceous\ perithecia darker, forming irregular, often lateral pads 3-5 mm. long and occupying the greater part of fructification (on cerambycid, or longicorn, beetle grubs — several to many, rarely only 1-2, sporophores crowded around neck of host) ^ C. dovei Stroma > 30 mm. long (and up to 20 cm.), bay brown', perithecia dark brown to blackish- — ^ 4. Stipe glabrous, very slender, simple or forked; perithecia acute, 0-2-0-3 mm. long (on scarabaeid, or cockchafer, beetle larv^) C. coxii Stipe brown-tomentose toward base; perithecia obtuse (on large moth larvae )-5 5. Fructification simple or slightly branched, slender, never > 3 mm. wide (usually 1-2 mm.) at junction with host; perithecia truncate, ± 4 per mm., 0-3-0-5 mm. long (host < 15 mm. thick) -C. robertsii Fructification usually much branched, antler-like and massive (or, if occasionally simple, then !> 3 mm. wide at point of emergence from host); perithecia often ± apiculate, 5—6 per mm., 0-1— 0*2 mm. long (host rt 20 mm. thick) C. taylon 6. Stipe hair-like, ± 5 cm. long, simple or once-forked, smooth, dark brownish; capitula almost globose, 2—3 mm. long (on ants — the perithecial stage unknown in Australia, where small pink-headed conidiophores spring from various parts of host) C. bicephala Stipe neither hair-like nor associated with ants- —7 7. Fructification very small ( 3 cm. long, frequently single and simple; capitulum at least 5 mm. long (often more) ^ 8 8. Capitulum dark reddish-brown, 5-10 mm. long, with paler and usually acute sterile terminal beak (4—8 mm.); stipe 1—2 mm. wide (on scarabaeid beetle larvae) C. brittlebankii Capitulum fertile up to the obtuse, rounded apex — 9 9. Stipe > 3 mm. wide (often up to 10 mm.), undivided, sometimes 20-30 cm. long; capitulum normally > 10 mm. long (on hepialid moth larvae) 11 Stipe < 3 mm. wide or irregularly much branched; capitulum to 10 mm. long — ^ 10. Mouths of perithecia distinct and widely spaced (4-7 per mm.); stipe very fiexuose, irregularly branched (with several yellow capitula) and lacerate (on hepialid moth larvae) — — C. cranstounii Mouths of perithecia microscopic, > 7 per mm.; stipe undivided, smooth'. capitulum brick-reddish or sepia, larvae) 5-8 mm. long (on tenebrionid beetle C. meneristitis As for the last, but the yellow-brown capitulum ± 10 mm. long and host a lucanid beetle larva — — — ; ^C. scotliana 11. Living capitulum dark olive-green to blackish, ill-defined below where it merges gradually into the pale yellow stipe, usually with small longitudinal wrinkles or creases and protruding perithecial mouths in the dried sitate — — — C. gunnii Living capitulum bay or coffee-brown, well-defined and sharply distinct frorn the paler brown stipe, not wrinkled in drying^ ^C. hawkesii 12. Capitula 3 per stipe, each 4-5 mm. long, red-brown, regularly ovoid, sharply contracting into a nipple-like apical point (identity of larval host unknown )^ C. furcata Capitulum solitary, 3-4 mm. long, ochre-brown, sometimes distorted (on scarabffiid, or cockchafer, beetle larvse) 13 13. Perithecia scattered, obliquely inclined toward axis- Perithecia close and compact, perpendicular to axis — C. aphodii C. sp. 74 J. H. Willis: Australian Species of the Fungal Genus Cordyceps ALPHABETICAL ARRANGEMENT OF SPECIES, SYNONYMY HOSTS, LOCATION OF TYPES, AND COLLECTIONS IN AUSTRALIA The following standard abbreviations have been used for various herbaria (the numbers in brackets indicating collections of Australian Cordyceps material housed in each local herbarium): ADM (3) ADW(12) BDW BM CANTY HO (10) K LPS MEL (40) MELM (12) MELU (22) NSW (9) PR SYD (5) SYDM (5) US South Australian Museum, Adelaide. Waite Agricultural Research Institute, Adelaide. Botany Division, Department of Scientific and Industrial Research, Wellington, N.Z. British Museum (Natural History), London. Canterbury Museum, Christchurch, N.Z. University of Tasmania (Botany Department), Hobart. Royal Botanic Gardens (Herbarium), Kew, London. Institute de Botanica C. Spegazzini, La Plata, Argentina. National Herbarium, Royal Botanic Gardens, Melbourne. National Museum of Victoria, Melbourne. University of Melbourne (Botany School). National Herbarium, Royal Botanic Gardens, Sydney. Botanical Department of National Museum, Prague, Czecho- slovakia. University of Sydney (Botany Department). Australian Museum, Sydney. National Museum and Smithsonian Institute, Washington (including Lloyd Herbarium, now under loan to Plant Industry Station, Beltsville, Md.). C. aphodii J. Mathieson in Trans Brit. mycoL Soc. 32: 134, t. 12-14 (1949). Host: Aphodius howitti Hope {Coleoptera — Scarabasids). Type: Miner’s Rest, near Ballarat, Vic. (/. Mathieson, Oct., 1946— MELU). Other Collections: Miner’s Rest, near Ballarat, Vic. (/. Mathieson, July, 1945 — MELU, asexual stage). The author of the species gives a very detailed account of all stages of the parasite, its artificial culture from both conidiospores and ascospores, secondary parasites which attack it and the life history of the host-insect (a small cockchafer beetle in the family Scarabceid^) . C. aphodii has not been reported beyond the type area in Victoria, but a very similar, still undetermined, species from New South Wales is discussed on page 85. ? C. bicephala Berk, in Hook. J. Bot. Kew Gdn Misc. 8: 278 (1856). C. australis (Speg.) Sacc. SyiL fang. 2: 571 (1883); C. unilateralis Tul. supsp. australis Speg. in An. Soc. dent, argent. 12: 215 (1881) — TYPE (?LPS) on Pachycondyla striata from mossy trunks near Apiahy, southern Brazil (Dr. Puiggari)\ Hyrnenostilbe melanopoda (Speg.) Petch in Trans. Brit, mycol. Soc. 16: 209 (1932); Isaria melanopus Speg. Bol. Acad. Cienc. Cordoba IT. 620 (1889) — TYPE (?LPS) on decayed beetle from mossy trunks near Apiahy, Brazil (Dr. Puiggari); Stilbum formicarum Cooke & Massee in Grevillea 18: 8 (1889) — TYPE (K) on an ant, “ Formica consobrina ”, from sundew plant at Cheltenham, Vic. (C. French, ± 1888). J. H. Willis: Australian Species of the Fungal Genus Cor dy ceps 75 Host: Unknown (probably ant). Type: Panure, near sources of Rio Negro, N.W. Brazil (R. Spruce, ± 1853 — K). The synonymy given above reflects the opinion of T, Fetch (1933, p. 67). It is generally conceded that the two taxa C. bicephala and C, australis (both described from Brazil) are conspecific, and it is practically certain that Hymenostilbe melanopoda (type of which came from the same tree-trunks as that of C. australis) is only a conidial stage of this species; but, since many distinct species of Cordyceps are known to parasitize ants, it is surely assuming too much to identify the single old Victorian collection of Stilbum formicarum with the perfect stage of a South American fungus? They may indeed be the same, but my doubt is expressed by the prefixing query to C. bicephala in this list of Australian Cordyceps. E. B. Mains [Bull. Torrey bot. CL 76: 24-30 (1949)] supports Kobayasi (1941, p. 183) in taking up the name C. australis instead of the earlier C. bicephala Berk., because the latter was presumed to be based on a doubtfully mature specimen; but Berkeley’s diagnosis explicitly mentions both asci and sporidia, and, in any case, the International Code of Nomen- clature does not sanction rejection of any name on a plea of immaturity for the type on which it is based. Fetch, in his belief that only a single species was involved, had every justification for synonymizing C. australis under C. bicephala. Kobayasi has extended the range of this fungus to Uganda in East Africa, and Mains (/.c.) presents it as a common parasite in the Western Province of Liberia, West Africa, “ killing ants by thousands C. brittlebankii E. McLennan & I. Cookson in Proc. roy. Soc. Viet. n. ser. 38: 74, t. 5 fig. 5-6, t. 6 fig. 4-5 (1926). Host: Heteronyx sp. (Coleoptera — Scarabaeids). Type: Ringwood, Vic. (/. Cookson, Apr. 1924 — MELU). Other Collections: Tyabb, Vic. (/. M. Raff — MELU). The species has not been observed outside Victoria. C. coxii Olliff in Agric. Gaz. N.S.W. 6: 412, t. 4 fig. C-D (1895). Host: 1 Lepidiota sp. {Coleoptera — Scarabsidee), also Cicadidae {Hemiptera). Type: In vicinity of turpentine trees at Kurrajong Heights, N.S.W. (T. Cranstoun, ? Mar., 1895 ? SYDM). Other Collections: Concord, N.S.W. {F. C. Lovegrove, Aug., 1910 — NSW); near Gembrook, Vic. {J. H. Willis, 1935 — MEL); Between Dayles- ford and Trentham, Vic. {Judith Thiele, July, 1946 — MEL); Alexandra, Vic. (C. G. Lane, Nov., 1907 — MELM). [At the Entomological Branch of the N.S.W. Department of Agriculture, the writer has seen a collection of Cordyceps-h^diung cockchafer larvae from Taren Point on Botany Bay {E. E. Hellenish, Dec., 1934); the fruiting bodies on these are imperfect, without perithecia, but most probably referable to C. coxii.] Although type of C. coxii has not been examined, the original figures and description (brief as it is) accord well with later collectings of a distinctive Cordyceps on cockchafer larvce, which is here referred to that species. 4751/59.-3 76 J. H. Willis: Australian Species of the Fungal Genus Cordyceps Olliff (/.c.) suggests that C. coxil “may prove to be an extreme variety ” of his species C. selkirkii, the latter always parasitizing large moth larvas and differing in its longer, truncate, less crowded perithecia. The writer unhesitatingly supports Lloyd (1920, p. 912), Kobayasi (1941, p. 99) and Dingley (1953, p. 331) in relegating C. selkirkii to synonymy under C. robertsii, but he does not uphold Kobayasi's treatment of C. coxil as another straight synonym of “ C. larvarum ” [i.e., of C. rohertsii]. C. cranstounii Olliff in Agrlc. Gaz. N.S.W, 6: 408, t, 2 fig. B (1895). (Plate VIII). Host: Oxycanus diremptus, and probably other species {Lepidoptera — Hepialidae). Type: Kurrajong Heights, N.S.W. (T. Cranstoun, Mar., 1895 ? loc.). Other Collections: Bola Creek, National Park, N.S.W. (A. Burges, July, 1947 — MEL); Koonung Creek, Doncaster, Vic. (Mr. & Mrs. Paul Fisch, June, 1942— MEL, ADW No. 3142); without locality, Vic. (MELM); Tarra Valley Nat. Park, Vic. (Mrs. K. Healey, June, 1959 — MEL). The characteristic features of C. cranstounii are its lacerate (or irregularly byssaceous) stipe, multiple heads which are relatively short, obtuse, yellowish, with perithecial orifices large and widely spaced. Type was not available for inspection, but the writer confidently refers the Bola Creek (N.S.W.) and Doncaster (Vic.) collections to this distinctive species. Many fresh specimens from Doncaster were examined over a period of four months; but, although the fruiting heads of these had well-developed perithecia, in no instance could sporidia (or even differentiated asci) be found! C. dovei L. Rodway in Pap. roy. Soc. Tasm. 1898-1899: 101, cum icon. (1900). C. aemome Lloyd MycoL Notes 6*^-: 932, fig. 1695 (1920) — TYPE (US) on Aemona hirta in rotting logs of Melicytiis ramifiorus. Weraroa, New Zealand (G. //. Cunningham, No. 51, Sept., 1919); CO-TYPE (No. 78, ?BDW). Host: ( Coleoptera — Cerambycid^e ) . Type: In decayed trunk of Nothofagus cunninghamii at Mt. Bischoff, Tas. {H. Stuart-Dove — HO; photo, of type MEL). Although six collections have been recorded from New Zealand (widely distributed through both islands), the only one known from Australia is that of the type which came from the high-rainfall area of western Tasmania. In its comparatively very small fructifications (with sterile apices and half- exserted perithecia), more or less crowded near the head of the beetle-host, C. dovei is almost unique. New Zealand populations differ slightly from the Tasmanian in having longer, more slender, less crowded stromata (sometimes even single), with the fertile portions more distinctly lateral and sterile tips less truncated (sometimes acute). Lloyd recognized them as constituting a separate endemic species, C. cemonce, but Kobayasi (1941, p. 103) reduced this to synonymy and pointed out the trifling nature of the differences. J. H. Willis: Australian Species of the Fungal Genus Cordyceps 11 C. furcata E. McLennan & 1 Cookson in Proc. roy. Soc. Viet. n. ser. 35: 157, t. 10 (1923). Host: Unknown larva. Type: Ringwood, Vic. (E. McLennan & I. Cookson, Sept., 1922 — MELU), The species is known only by the single type collection. C. gunnii (Berk.) Berk, in Hook. f. Flor. Tasm. 2: 278 (1859). (Plate IX, figs. 1-3). Sphaeria gunnii Berk, in Lond. J. Bot. 7: 577, t. 22 (1848); C. consumpta G. H. Cunn. in Trans. N.Z. Inst. 53: 377, t. 60 fig. 1 (1921) — TYPE (No. 230, CANTY) on buried larva of Porina sp. [= Oxycaniis sp.] at Rotorua, N.Z. (A. Lush^ June, 1920); 1C. craigii Lloyd Mycol. Notes 4-^^: 511, fig. 718 (1915) — TYPE (US) on Porina enysii [~ Oxycanus enysii] from old and abandoned kumara {Ipomcea batatas) beds at Auckland, N.Z. {E. Craig)', 1C. hum Lloyd Mycol. Notes 1061, fig. 1994 (1921)— TYPE (US) from N.Z.. without details (H. H. Hill). Host: Oxycanus spp. {Lepidoptera — Hepialidie). Type: Franklin Village, near Launceston, Tas. {R. C. Gunn, No. 1800, Apr., 1846— K, NSW). Other Collections: Tasmania — Knockiofty, Hobart (?L. Rodway, June, 1895 — HO); without locality {W. V. Fitzgerald, 1891 — MEL); Strzelecki Peak, Flinders Island (/. H. Willis, Apr. 1954 — MEL). Victoria — ? Dimboola {E. Muir, 1948 — MEL); Bahgallah, near Casterton {R. C. Miller, June, 1885 — MEL); Portland (/. A. Leach, June, 1906 — MEL); Macedon (MELU); Apollo Bay (MELU); Port Phillip (C. French, June, 1869 — MEL); Studley Park {W. Kershaw, 1857 — Herb. R. C. Kershaw; F.M. Reader, June, 1885 — Herb. Viet. Dept. Agric., Burnley); “ Comellia (C. French, May, 1900 — Herb. Viet. Dept. Agric., Burnley); Kew (MELU); Doncaster {P. Fisch, June, 1942 — MEL, MELU); Dandenong (F, Gessner, 1892 — MEL); Kallista {E. I. McLennan, Aug., 1949 — MELU); Dandenong Ranges (MEL); Kalorama (Mrs, Peters, May, 1957 — MELM); Mornington (MELU); Tyabb (Master Blackwood, Nov., 1920 — MELM); Nyora (MELU); Korumburra (MELU); Orbost (MELU); Mallacoota (W. Hunter, Oct., 1955 — MEL); Wangaratta (M. Ferris, June, 1956 — ADW No. 7450); Tarra Valley Nat. Park (Mrs. K. Healey, June, 1959 — MEL). [Noted also as abundant at Emerald and Cockatoo.] South Australia — Penola (C. Barrett, June, 1931 — MELU); Sellick’s Hill (Dr. E. C. Stirling— ADM); Kingston (Dr. A. Engelhardt— ADM; /. B. Cleland, Aug., 1943— MEL; June, 1943 —ADW Nos. 266-268; J. B. Cleland, July, 1942— ADW No. 3143; A. R. Naimes, Aug., 1945 — ADW No. 3145). 78 J. H. Willis: Australian Species of the Fungal Genus Cor dy ceps New South Wales — Shoalhaven River (?W. Bauerlen, June, 1888 — MEL); Kurrajong Heights (SYDM); Bola Creek, National Park (/. McLuckie, A. Burges 3 mm diam.) foliato, angulato-striato, sublsvi, intlorescentiam versus tomentosiusculo, corymboso-ramoso; ramulis summis mono- vel pleiostachyis. Folia caulina, alterna breviter petiolata, laminis (ca. 3,5 cm longis, 1,2 cm latis) primo sparse pilosis, dein glabrescentibus, oblongo-lanceolatis, acuminatis, integerrimis, in petiolum (0,5 cm) alatum angustatis; superiora reducta (0,5-0,7 cm longa) squamiformia. Spicre solitariiE vel composite, pedunculate vel raro sessiles, terminates vel laterales, in paniculam congestse, ovato-hemisph^ricce vel oblongo-ovoidere, 1-2,5 cm longffi, ad 1,5 cm latre; rhachi lanuginosa. 2. mm Fig. 2 . — Ptilotus carinatus G. Beni. Outer perianth-segment, side view. Flores subdensi visu rosei, albido-pilosi (pilis denticulato-nodosis) ; bracteis brac- teolisque scariosis, concavis inaequalibus: bractea atro-fusca conspicua, ovata, acumi- nata, ad 5 mm longa et 3 mm lata, extus laxe villosula, post lapsum perianthii superstiti; 2 bracteolis appressis praeter nervum medium fuscum pilosiusculum ± hyalinis, nitentibus, subglabris, rotundis, 3,5 cm latis, apicem versus subserrulatis, in cuspidem setaceam I mm longam productis. Perianthium 5-partitum campanulato- patens, basim constrictam versus valde indurescens; tepalis linearibus obtusis, basi ima incrassata et indurata in tubum 1 mm longum (1,2 mm diam.), extus hirsutum angustatis et coalitis, inferne conspicue carinatis, rigidis, dorso insequaliter pilosis (pilis ad 2 mm longis apicem baud aequantibus), bicoloratis, limbatis, areola mediana virescenti pubescentia abscondita, marginibus inferne ciliatis, supeme roseis in apicem nudum, roseum, 1,8 longum, dilatatum, interdum recurvatum transeunti- bus, inaequalibus: 2 exterioribus 7,5 mm longis, 1 mm latis, apice truncate, subdenti- culato 1,5 mm lato, intus glabris; 3 interiori- bus 7 mm longis, 0,7 mm latis, apice subspathulato, eroso-denticulato 1,2 mm lato, intus in parte inferiore pilis crispis, intricatis, marginibus praecique orientibus indutis. G. Benl: New Species and Varieties of Ptilotus R.Br. 105 Stamina 5 inaequalia, plerumque 2-3 minora abortiva, basi dilatata (0,5 mm) in anulum minimum (0, 1-0,2 mm altum) membranaceum, glabrum, tubo perianthii partim adnatum coalita; pseudostaminodiis nullis; filamentis fertilibus ligulatis, linearibus ad 4 mm longis, 0,15 mm latis, superne subulatis, antheris bilocularibus obscure flavis, subrotundis (0,6-0, 7 mm diam.), dorso affixis. Ovarium subclavatum longe stipitatum, 1,5 mm latum, stipite incluso 2,8 mm longum (stipite 1 mm), apice villoso; stylo =lz excentrico, gracili ad 3 mm longo, glabro; stigraate distincte capitellato, papilloso. Representative locality: WESTERN AUSTRALIA — near Wittenoom Gorge, about 100 miles south of Marble Bar (HOLOTYPE in NSW — K. McMahon, 1952; No. 33667). P. carinatus is easily distinguished from all those species possessing a visible, naked, blunt and denticulate apex of their perianth-segments by its striking ciliate keel in the lower part of the tepals, a feature not observed until now in any previously described species. 3. P. PSEUDOHELIPTEROIDES G. Beni, species nova, [Fig. 3.] Herba annua (?) pubescens pluricaulis, caulibus (ca. 10) ad 20 cm altis, basi 3,5 mm diam., erectis vel adscendentibus, foliosis, striatulis, prime tomentosiusculis — pilis creispis nodulosis- — dein subglabris, ramosis; ramis (4-10) dz fastigiatis, mono- vel plurispicatis. Folia alterna obscure viridia, pilis argenteis sericeis (3 mm longis) induta, 1-2 cm distantia, petiolata, summa inflorescentiam juvenilem zb involventia; laminis oblongo-lanceolatis ad 3 cm longis, 0,8 cm latis, mucronatis (mucrone ca. 1 mm longo), basi in petiolos ad 0,7 cm longos contractis, integerrimis, nervo medio subtus prominente. Spic^ numeros£e elongato-ovoideas, 2,2 cm longcC, 1,4 cm latas, pedun- culatse, subdense paniculate; rhachi villosa. Flores (ca. 50) conferti straminei; bracteis bracteolisque inequalibus conspicuis, glabris, scariosis, tenuibus, hyalinis, pallide succcineis, lucidis, integerrimis, baud appressis, uninervibus, nervo in cuspidem (0,5 mm) producto: bractea inferiore dz angustata, ovato-lanceolata, 6 mm longa, ad 2,5 mm lata, post lapsum perianthii superstiti, 2 bracteolis lateralibus ovoideis, 5 mm longis, 3 min latis. Perianthium pentaphyllum campanulato-patens, erectum, basim incras- satam, callosam, conicam, extus hirsutam versus indurescens; tepalis lineari- oblongis, distincte uninerviis, ecarinatis, rigidis, parte inferiore sub- cartilagineis, limbatis, parte mediana virescenti, apicem versus divaricatis, usque ad basim fere liberis, intus levigatis extus hirsutis (pilis dorsalibus rectis, albidis ca. 2 mm longis, articulatis, apicem vix excedentibus), inequalibus: 2 exterioribus ad 7 mm longis et 1,2 mm latis obtusis intus glabris (pilis paucis brevibus, nervo medio inferiore orientibus exceptis), marginibus hyalinis in apicem nudum, truncatum, serrulatum, 0,3 mm latum transeuntibus; 3 interioribus 6 mm longis, 1 mm latis acutis, intus opacis, basi pilis crispis, baud numerosis, imprimis marginibus orientibus indutis. 106 G. Benl: New Species and Varieties of Ptilotus R.Br. Stamina 5, omnia ^equalia fertilia vel 1-2 sterilia, longiora (3 mm), basi modice dilatata in anulum brevem (ca. 0,2-0, 4 mm), membranaceum, plerumque fimbriatum, basi callos^e perianthii adnatum coalita; pseudo- staminodiis zb imperfectis, fissis; filamentis filiformibus, antheras fertiles portantibus 2 mm longis, 0,05 mm, basim versus 0, 1-0,3 mm latis, primo dilute fulvis, dein parte mediore zb rufescentibus; antheris bilocularibus, ellipticis, obscure flavis 0,2 mm longis et 0,1 mm latis, dorso affixis. Ovarium subglobosum, sessile ca. 1 mm longum, apice lanuginosum; stylo central! glabro, 1,5 mm longo, basi, 0,1 mm diam.; stigmate papillose inconspicuo. I I I i I 5 rnm Fig. 3 . — Ptilotus helipteroides (F. Muell.) F. Muell., var. helipteroides. {a) Staminal cup spread open, inner view; (6) pistil . — Ptilotus pseudohelipteroides G. Beal, (c) staminal cup spread open, inner view; {d) pistil. G. Benl: New Species and Varieties of Ptilotus R.Br. 107 Habitat: “On hard bare pebbly brown clay loam flats” (fide S. L. Everist). Representative locality: QUEENSLAND (Gregory North District) — Currawilla, about 100 miles west of Windorah: Stallion Paddock (HOLO- TYPE in CANS, ISOTYPES in BRl and K—S. L. Everist, 9th June, 1949, No. 3937). The present species bears in its general appearance some likeness to P. helipteroides (F. MuelL) F. Muell., var. helipteroides. But apart from numerous details in the structure of the leaves and especially of the flowers (Fig. 3) the latter is characterized by the red colour of its perianth- segments. Traces of red may be observed even in plants collected more than 46 years ago, while the three specimens of P. pseudohelipteroides found in 1949 by Mr. S. L, Everist and cited above do not show the least bit of red in their straw-coloured spikes. P. helipteroides is known from Northern Territory, from Western and South Australia, the new and really distinctive species comes from Queensland: the disjunction of the respective localities may be considered as an additional argument for separating the plant from P. helipteroides. 4. P. DISSITIFLORUS (F. Muell.) F. Muell., var. LONGIFOLIUS G. Beni, varietas nova. Perennis baud tomentosa (rhachi excepta), minus ramosa. Recedit a typo specie! foliis glabris ad 7 cm longis, 5 mm latis, spicis floribusque maioribus, tepalis apicibus denticulatis, filamentis cupulam versus conspicuc (ad 0,45 mm) dilatatis. TYPE from Port Denison, QUEENSLAND (North Kennedy) (HOLO- TYPE of variety in MEL — E, Fitzalan, 1874). The variety differs from the typical form primarily in the longer, narrower and glabrous leaves, thus resembling in its habit of foliage P. distans (R. Br.) Poir., from which it is, of course, well distinguished by its floral structure. This had already been stated in 1951 by Miss C. I. Skewes (MEL). 5. P. LANATUS A. Cunn. ex Moq., var. GLABROBRACTEATUS G. Beni, varietas nova, [Fig. 4.] Rhachis minus villosa; bractea 3,5 mm longa lanam rhachis multo excedens, glabra (ima basi interdum excepta), bracteol£e 2 mm longue apicem versus vix pilosulse; pili tepalorum minus densi, crispiores. Representative localities: WESTERN AUSTRALIA — near King Sound (HOLOTYPE of variety in MEL, ISOTYPES in NSW and K— IP. W. Froggatt, 1887); Balmaningarra, Northen Kimberley District (in K and NSW — H. Basedow, April, 1916, No. 74). 108 G. Benl: New Species and Varieties of Ptilotus R.Br. The distinguishing feature of the above variety is the different size of the bract and bracteoles, and their lack of pubescence, as Fig. 4 shows. 3mm Fig. 4 . — Ptilotus lanatus A. Cunn. ex Moq., van lanatus. {a) bract; (b) bracteole. — Ptilotus lanatus A. Cunn. ex Moq., van glabrobracteatiis G. Beni, (c) bract; {d) bracteole. 6. P. STIRLINGII (Lindl.) F. MuelL, var PUMILUS G. Beni, varietas nova. Planta pumila, caulibus (10) ex rhizomate (4 mm diam.) orientibus dense collocatis, albido-tomentosis, 7—12 cm longis, mono- ad tetrastachyis; foliis pilosis inferioribus ad 3,5 cm, superioribus vix 0,8 cm longis; spicis hemisphsericis 1,5 cm diam., rhachi lanuginosa, bractea 3,5 mm longa et 1,5 mm lata, extus villosa, bracteolis 4 mm longis, 2 mm latis, tubo perianthii 1,2 mm longo; tepalis extus pilis spinuloso-articulatis 4 mm longis obtectis, exterioribus 8,5 mm, interioribus 7,5 mm longis; staminibus fertilibus (2) 4-5 mm longis; ovario glabro 2 mm, stylo 3 mm longo. TYPE from Sharks Bay, WESTERN AUSTRALIA (HOLOTYPE of variety in MEL— F. Mueller, Oct., 1877). This plant looks like a dwarf of the usual form of the species, with a denser indumentum of the perianth-segments. Further differences — regard- ing proportions of floral organs — are brought out by closer examination. The new species and varieties presented above were found out of a total of more than 3,000 specimens of PTiLOTUS/TRTCHINIUM from European, American and Australian herbaria. The study of the Australian material was made possible to me by the courtesy of Mr. R. H. Anderson (Botanic Gardens, Sydney), Miss N. T. Burbidge (C.S.I.R.O., Division of Plant Industry, Canberra), Miss C. M. Eardley (University of Adelaide, Department of Botany). Mr. S. L. Everist (Botanic Museum and Herbarium, Brisbane), Mr. A. W. Jessep (Melbourne Botanic Gardens and National Herbarium of Victoria, South Yarra), Dr. R. T. Patton and Miss E. M. Packe (University of Melbourne, Botany Department), and of Miss C. I. Skewes (National Herbarium of Victoria, South Yarra). A considerable number of sheets were generously sent to our Munich State Collection as a gift by the National Herbarium of New South Wales (Mr. R. H. Anderson), the Herbarium of the Division of Plant Industry, Commonwealth Scientific and Industrial Research Organization (C.S.I.R.O.), Canberra (Miss N. T. Burbidge), the Queensland Herbarium, Brisbane (Mr. S. L. Everist), and by Mrs. H. E. Ramsay, Red Cliffs, Victoria. I am much indebted to all the above-mentioned authorities and institutes, and I take the opportunity of thanking them sincerely for the valuable help received. REINSTATEMENT OF CALOTIS SUFFRUTICOSA Domin (Compositae). by Gwenda L. Davis. (Department of Botany, University of New England, Armidale, N.S.W.) In my revision of the genus Calotis R. Br. [Proc. Linn. Soc. N.S.W. 77: 171 1952] I concluded that C. suffruticosa Domin was conspecific with C- lappulacea Benth. and reduced it to synonymy. No type material was available in Australia and, from the original description, it seemed that the points of difference noted by Domin were insufficient to justify specific status. Since then, however, specimens have come to hand collected not far from the type locality of C. suffruticosa which are undoubtedly this species and quite distinct from, although similar to, C. lappulacea. C. suffruticosa Domin is accordingly removed from synonymy under C. lappulacea and reinstated as a good species. The original description is reproduced below, translated from the Latin, together with supplementary notes and figures based on the recent material. I am indebted to Dr. Adair Dale of the Classics Department, University of Mew England, for this translation. C. suffruticosa Domin in Bibl. bot., Stuttgart 89: 655 (1929). “ A much-branched, erect sub-shrub about half a meter in height with a stem which is thick and woody at the base; the branches slender, erect, close together and almost twig-like, slightly hirsute or entirely glabrous, leafy. Leaves linear, flat, acute, slightly appressed, hirsute, entire or occasionally linear-cuneate, with acute teeth on both sides; the larger ones nearly 2 cm. long and 3 mm. broad. Inflorescences numerous, 1 cm. or more in diameter across the receptacle. Involucre campanulate, with many bracts which are narrow oblong-elliptical, very fiat, smooth and at least 5 mm. long. Fruits muricate with two rigid, elongate, divergent bristles at least 3 mm. long, bearing short recurved spines; the apex of the fruit bears a ring of very short bristles about 0-5 mm. or a little more in length. Queensland: Savannah woodland near Jericho (Domin III., 1929). Related to C. lappulacea Benth. but differing in the involucre, the fact that the capitula are twice the size, and that the fruits bear two elongate awns and many short ones.” 109 1 10 G. L. Davis: Reinstatement of Calotis suffruticosa Domin Supplementary Notes. Material: 12 miles south-east of Muttaburra on the Aramac-road, western Queensland, open grassy plains, flowers bright yellow to orange, 9.9.1956, N,T. Burbidge (CANB. 5521; BRI; MEL). Cauline leaves sessile, acute; the lower leaves up to 5 cm. long, 1 cm. broad, linear-cuneate, acutely toothed distally, becoming entire and broad-linear higher up the stem. Involucral bracts obtuse, torn-ciliate. Ray florets numerous, in several rows, the rays 4-5 mm. long, 0*5 mm. broad. Fruits cuneate, flattened, light brown, finely tuberculate. Calotis suffruticosa Domin. 1. Lower cauline leaf, x 1. 2-3. Fruit. Outer and lateral surfaces, x 15. TWO NEW AUSTRALIAN SPECIES OF BRACHYCOIWE Cass. (Composite) by Gwenda L. Davis. (Department of Botany, University of New England, Armidale, N.S.W.) SUB-GENUS EUBRACHYCOME. BRACHYCOME TETRAPTEROCARPA G. L. Davis, species nova [Fig. 1-3, p. 113.] Herbce adscendentes, pilis glandulosis, ad 30 cm. altae, e basi ramosas, folia radicalia non nisi in herbis immaturis reperienda; folia caulina sessilia, ad 6 cm. longa, 1-5-2 mm. lata; peduncula nuda; capitula ad 50 numero, 1 cm. sine radiis transverse mensa; involucri phylla 1 8, 4 mm. longa, 1 • 5 mm. lata, arte lanceolata, obtuse acuta, glabra, marginibus minute denticulatis; flores radii 25-34; radii 5 mm. longi, 1*2 mm. lati, albi; receptaculum 5 mm. latum, 2 mm. altum, minime convexum; achcenia 2 mm. longa, 2 mm. lata, late oblongata, suffusca, corpore levi, alis lateralibus valde incurvatis, integris, alis dorsalibus ventralibusque quibus margines undulatae adsunt; pappus ininutus albusque. Ascending glandular-hairy herbs up to 30 cm. high, branching from the base. Radical leaves only present on young plants. Cauline leaves sessile, up to 6 cm. long, pinnatipartite with 9—1 1 acute, lanceolate segments about 6 mm. long, 1 *5-2 mm. broad. Peduncles naked. Capitula up to 50, 1 cm. diameter excluding the rays. Involucral bracts 18, 4 mm. long, 1*5 mm. broad, narrow-lanceolate, bluntly acute, glabrous, with minutely denticulate margins. Ray florets 25-34; the rays 5 mm. long, 1*2 mm. broad, white. Receptacle 5 mm. broad, 2 mm. high, slightly convex. Fruits 2 mm. long, 2 mm. broad, broadly oblong, light-brown; the body smooth, with broad, strongly incurved, entire, lateral wings and narrow dorsal and ventral wings with undulate margins; pappus minute, white. Holotype: 15 miles south-east of Muttaburra on Aramac-road, Queensland, “ Open grassy downs, in small water channel ”, 9.9.1956, N. T. Burbidge (CANB. No. 5523). Paratypes: Same collection (MEL; NSW; BRI; ADW.). Specimens examined: Type series; Aramac (Qd), 7, 1950, W. Robertson (CANB. No. 1950). The fruits of this species are unique in that they bear four longitudinal wings at equal distances around the glabrous poorly-defined body. The lateral wings are rather thick but are very similar in form to those of B. curvicarpa, whereas the dorsal and ventral wings are narrow and thin. This resemblance between the two species is seen also in the habit and vegetative features and suggests a close relationship. In view of the fact that B. tetrapterocarpa is known only from the type district, it is possible that it originated as a local variant of B. curvicarpa, whose most northern record is Longreach, approximately 60 miles from both Muttaburra and Aramac. Ill 112 G. L. Davis: Two New Australian Species of Brachycome Cass SUB-GENUS METABRACHYCOME. BRACHYCOME DIMORPHOCARPA G. L. Davis, species nova [Fiij. 4-7, p. 113.] Herbse adscendentes aut inflrme erectse, ad 38 cm. altce, multistipitauc, ramosEe in partibus maturis minutissime septato-pilosae, aliter pilis glandulosis; folia radicalia non nisi in herbis immaturis reperienda; folia caulina ad 3 cm. longa, late sessilia, pinnatisecta, segmentis 9-1 1 lanceolatis, breviter mucronatis, ad 9 mm. longis, 1*5 mm. latis; pedunculi filiformes, nudi, glandulosi; capitula 50-100, fere ad 5 mm. (sine radiis) transverse metientia; involucri phylla 16, 3 mm. longa, 0*9 mm. lata, oblanceolata, acuta, glandulosa, marginibus minutissime serrulatis; flores radii 12-20, ligulis 4-6 mm. longis, 0*8-1 mm. latis, violaceis; receptaculum alte convexum, 1*4 mm. latum, 0*8-1 mm. altum; achfenia dimorphica — radiorum 1*5 mm. longa, 0*5 mm. lata, fusca, clavata, compressa, alte tuberculata, marginibus levibus, pappo albo minutissimoque; discorum autem 1*7 mm. longa, 1*3 mm. lata, corpore valde fusco glabro, pilis glandulosis paucis lineam erectam mediam utrinque formantibus, alis lateralibus latis integris albis aut subluteis apicem versus perimplicatis atque secus marginem a pilis parvis glandulosis instructis, pappo albo minutoque. Ascending or weakly erect herbs, up to 38 cm. high, many stemmed, branching, minutely septate-hairy on mature parts, otherwise with glandular hairs. Radical leaves only present on young plants. Cauline leaves up to 3 cm. long, broadly sessile, pinnatisect with 9-11 lanceolate, shortly mucronate segments, up to 9 mm. long, 1*5 mm. broad. Peduncles filiform, naked, glandular. Capitula 50-100, about 5 mm. diameter, excluding the rays. Involucral bracts 16, 3 mm. long, 0*9 mm. broad, oblanceolate, acute, glandular, with minutely serrulate margins. Ray florets 12-30, the rays 4-6 mm. long, 0*8-1 mm. broad, mauve. Receptacle steeply convex, 1 *4 mm. broad, 0*8-1 mm. high. Fruits dimorphic, those of the ray I *5 mm. long, 0*5 mm. broad, brown, clavate, flattened, deeply tuberculate with smooth margins; pappus white, microscopic. Disc fruits 1*7 mm. long, 1*3 mm. broad, the body dark-brown, smooth with a few glandular hairs down the centre of each face; lateral wings broad, entire, white to pale yellowish-brown, strongly infolded distally, with small marginal glandular hairs. Holotype: Bon Bon Station to Kingoonya, South Australia. “Low wash in red sand country with Myall {Acacia sowdenii Maiden) 1 1.10.1955, N. T. Biirbidge and M. Gray (CANB. No. 4653). Paratypes: Same collection (MEL; NSW; ADW; BRI). The vegetative similarity of this species with B. ciliaris (Labill.) Less., together with the presence of dimorphic fruits, indicates a close relationship, particularly with B. ciliaris var. lanuginosa (Steetz) Benth. The disc fruits of B. dimorphocarpa, however, are quite distinct in the curved and infolded nature of their lateral wings. L G. L. Davis: Two New Australian Species of Brachycome Cass 113 Mr. Gray, in personal communication, described the wings of the fruits as being “ a most effective aid to wind-dispersal as they spin off at a great rate when the wind catches them B. dimorphocarpa is only recorded from the type locality, near Kingoonya, South Australia, but there is a possibility that, owing to its close vegetative resemblance to B. ciliaris, it has not been recognized elsewhere as distinct. I am indebted to Dr. Adair Dale of the Classics Department, University of New England, for supplying the Latin diagnoses of these species. Brachycome tetrapterocarpa. 1. Lower cauline leaf, x 1. 2-3. Fruit. Outer and inner surfaces, x 15. B. dimorphocarpa. 4. Lower cauline leaf, x 1. 5. Ray fruit, x 30. 6-7. Disc fruit. Outer and inner surfaces, .r 30. STUDIES IN MIMOSACEAE— Part 1. by A. B. Court, National Herbarium of Victoria. ACACIA FLEXIFOLIA A. Cunn. ex Benth., A NEW RECORD FOR VICTORIA H. B. Williamson in A. J. Ewart Flor. Viet, 587 (1931) confounded A. flexifolia A. Cunn. ex Benth. in Hook. Lond. /. Bot. 1:359 (1842) with A. lineata A. Cunn. ex G. Don Gen. Syst. 2:403 (1832) although these two species differ very markedly in several important respects. Hitherto, A. flexifolia had been recorded as endemic in New South Wales, but now it must be added formally to the Victorian flora on the basis of the collections listed below. KEY DIFFERENCES BETWEEN THE TWO SPECIES. A. flexifolia: Narrow greyish-green virgate shrub to 1*5 m. high; branches glabrous, with prominent scaly resinous ridges; phyllodes quite glabrous, more or less linear, broadening towards their apices, usually bent sharply near their bases; apices of phyllodes rounded but sometimes with a very small point; peduncles ca. 3 mm. long or less, appearing farinaceous; flowers 3-8 (rarely more) per head, appearing May to October. A, lineata: Spreading greenish, or sometimes greenish-yellow, shrub to 2 m. high and often as broad; extremities (at least) of branches pubescent, not prominently resinous; phyllodes usually pubescent, more or less simply linear, broader near their centres, usually acute; peduncles ca. 8-10 mm, long (sometimes more), usually quite glabrous but never appearing farinaceous; flowers 10-15 (rarely less) per head, appearing July to October. DISTRIBUTION NOTES. The most westerly record for A. flexifolia in Victoria is the Whipstick Scrub near Bendigo and the most easterly locality for A. lineata lies about 15 miles to the west at Inglewood. These two species are represented in the National Herbarium of Victoria by the following collections: A- flexifolia — Whipstick forest near Bendigo, D. J. Baton, May, 1918; Bendigo, C, S, Sutton, s.n., 12/9/1920; Bendigo, D. J, Baton, s.n., October 1922; Central Whipstick near Bendigo, W, Berry, s.n., 1946; Central Whip- stick Scrub near Bendigo, W. Berry, s.n., 27/12/1953; Central Whipstick near Bendigo, W. Berry, s.n., 20/11/1955; In hard soil amongst mallee in Whipstick Scrub ca. 8 km. (5 miles) north from Huntly near Bendigo, A. B. Court, n. 941, November 4, 1958; Earlston via Violet Town, R. A. Black, s.n., 13/9/1944; Forest Sanctuary Reserve, Reef Hill, about 1 mile south of Benalla, H. C. E. Stewart, s.n., 15/6/1959. A. lineata — Mallee District, N. W. Viet., St. Eloy D’Alton, s.n., October, 1899; Kurting, N. W, Victoria, A. Burdie, s.n,, 1894; North Western Desert of Victoria, L. Morton, s.n., s.d.; 15 Miles from Nhill in the direction of Lake Hindmarsh, St. Eloy D' Alton, s.n., June, 1892; Nhill, St, Eloy D' Alton, s.n., 31/3/1897; In sandy soil beside Lillimur-AMP road ca. 25 km. (15 miles) north from Lillimur North, A. B. Court, n. 1310, October 14, 1958. 114 By Authority: A. C. Brooks, Government Printer, Melbourne. . 4 i > -r :