Revision of the Family Raspailiidae (Porifera: Demospongiae), with description of Australian Species.

John N . A . Hooper

The marine sponge family Raspailiidae Hentschel is revised and referred to the order Poecilosclerida. Of 48 nominal genera, 17 (including one new genus and one new subgenus) are recognised here: Raspailia Nardo, (Hymeraphiopsis, subg. nov.), Ectyoplasia Topsent, Endectyon Topsent, Trikentrion Ehlers, Cyamon Gray, Aulospongus Norman, Raspaciona Topsent, Rhabdeurypon Topsent, Eurypon Gray, Plocamione Topsent, Amphinomia, gen. nov., Lithoplocamia Dendy, Hymeraphia Bowerbank, Ceratopsion Strand, Thrinacophora Ridley, Axechina Hentschel and Echinodictyum Ridley, and three genera are incertae sedis (Tethyspira Topsent, Sigmeurypon Topsent, Cantabrina Ferrer-Hernandez). Fifty-six species are described for the Australian fauna, of which 14 are new to science: Raspailia daminensis, R. desrnonyiformis, R. keriontria, R. melanorhops, R. phakellopsis, R. reticulata, R. wardi, R. wilkinsoni, Ectyoplasia vannus, Endectyon elyakovi, Ceratopsion montebelloensis, C. palmata, Echinodictyum austrinus, spp. nov. and Amphinomia sulphurea, gen. nov., sp. nov. The phylogenetic relationships and biogeographical distribution of the family are discussed. Tropical north-westem Australian-southem Indonesia has the highest diversity of species in the Indo-west Pacific, and altogether the Australasian region has about 20% of the world's known raspailiid fauna.

Invertebr. Taxon., 1991. 5. 1179-418 Revision of the Family Raspailiidae (Porifera :Demospongiae). with Description of Australian Species John N . A . Hooper Northern Temtory Museum of Arts and Sciences. P.O. Box 4646. Darwin. N.T. 0801. Australia Present address: Queensland Museum. P.O.Box 300. South Brisbane. Qld 4101. Australia. . Contents Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Materials and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Abbreviations used in the text . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Analysis of Raspailiid Characters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Major morphological characters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Systematics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Family Raspailiidae: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Keytogenera . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Genus Raspailia Nardo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Key to subgenera . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Subgenus Raspailia Nardo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Raspailia (Raspailia) atropurpurea (Carter) . . . . . . . . . . . . . . . . . . . . . Raspailia (Raspailia) echinata Whitelegge . . . . . . . . . . . . . . . . . . . . . . Raspailia (Raspailia) gracilis (Lendenfeld) . . . . . . . . . . . . . . . . . . . . . Raspailia (Raspailia) phakellopsis. sp. nov . . . . . . . . . . . . . . . . . . . . . Raspailia (Raspailia) pinnatifida (Carter) . . . . . . . . . . . . . . . . . . . . . . . Raspailia (Raspailia) tenella (Lendenfeld) . . . . . . . . . . . . . . . . . . . . . . Raspailia (Raspailia) vestigifera Dendy . . . . . . . . . . . . . . . . . . . . . . . . Raspailia (Raspailia) wilkinsoni. sp. nov . . . . . . . . . . . . . . . . . . . . . . . Subgenus Clathriodendron Lendenfeld . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Raspailia (Clathriodendron) arbuscula (Lendenfeld) . . . . . . . . . . . . . . . Raspailia (Clathriodendron) bifurcata Ridley . . . . . . . . . . . . . . . . . . . . Raspailia (Clathriodendron) cacticutis (Carter) . . . . . . . . . . . . . . . . . . . Raspailia (Clathriodendron) daminensis. sp nov. . . . . . . . . . . . . . . . . Raspailia (Clathriodendron) desmoxyiformis. sp nov. . . . . . . . . . . . . . . Raspailia (Clathriodendron) keriontria. sp. nov. . . . . . . . . . . . . . . . . . . Raspailia (Clathriodendron) melanorhops. sp. nov. . . . . . . . . . . . . . . . . Raspailia (Clathriodendron) paradoxa Hentschel . . . . . . . . . . . . . . . . . Subgenus Raspaxilla Topsent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Raspailia (Raspaxilla) compressa Bergquist . . . . . . . . . . . . . . . . . . . . . Raspailia (Raspaxilla)frondula (Whiteleggge) . . . . . . . . . . . . . . . . . . . Raspailia (Raspailla) reticulata. sp nov. . . . . . . . . . . . . . . . . . . . . . . Raspailia (Raspaxilla) wardi. sp nov. . . . . . . . . . . . . . . . . . . . . . . . . . Subgenus Syringella of authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Raspailia (Syringella) australiensis Ridley . . . . . . . . . . . . . . . . . . . . . . Raspailia (Syringella) clathrata Ridley . . . . . . . . . . . . . . . . . . . . . . . . Raspailia (Syringella) elegans (Lendenfeld) . . . . . . . . . . . . . . . . . . . . . Raspailia (Syringella) nuda Hentschel . . . . . . . . . . . . . . . . . . . . . . . . . Raspailia (Syringella) stelliderma (Carter) . . . . . . . . . . . . . . . . . . . . . . . . . . J. N . A . Hooper Subgenus Hymeraphiopsis. subg.nov. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Raspailia (Hymeraphiopsis)irregularis Hentschel . . . . . . . . . . . . . . . . . Genus Ectyoplasia Topsent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ectyoplasia frondosa (Lendenfeld) . . . . . . . . . . . . . . . . . . . . . . . . . . . Ectyoplasia tabula (Lamarck) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ectyoplasia vannus. sp . nov. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Genus Endectyon Topsent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Endectyon elyakovi. sp . nov. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Endectyon fruticosa aruensis (Hentschel) . . . . . . . . . . . . . . . . . . . . . . . Endectyon thurstoni (Dendy) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Endectyon xerampelina (Lamarck) . . . . . . . . . . . . . . . . . . . . . . . . . . . Genus Trikentrion Ehlers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Trikentrion fEabelliforme Carter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . GenusCyamonGray . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cyamon aruense Hentschel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Genus Aulospongus Norman . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Genus Raspaciona Topsent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Genus Hymeraphia Bowerbank . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . GenusEuryponGray . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eurypon graphidiophora Hentschel . . . . . . . . . . . . . . . . . . . . . . . . . . . Genus Rhabdeurypon Topsent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Genus Plocamione Topsent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Genus Lithoplocamia Dendy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Genus Amphinomia. gen.nov . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Amphinomia sulphurea. sp . nov. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Genus Ceratopsion Strand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ceratopsion dichotoma (Whitelegge) . . . . . . . . . . . . . . . . . . . . . . . . . . Ceratopsion axifera. (Hentschel) . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ceratopsion montebelloensis. sp . nov. . . . . . . . . . . . . . . . . . . . . . . . . . Ceratopsion palmata. sp . nov. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Genus Thrinacophora Ridley . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Thrinacophora cervicornis Ridley & Dendy . . . . . . . . . . . . . . . . . . . . Genus Axechina Hentschel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Axechina raspailioides Hentschel . . . . . . . . . . . . . . . . . . . . . . . . . . . Genus Echinodictyum Ridley . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Echinodictyum arenosum Dendy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Echinodictyum asperum Ridley & Dendy . . . . . . . . . . . . . . . . . . . . . . Echinodictyum austrinus. sp . nov. . . . . . . . . . . . . . . . . . . . . . . . . . . . . Echinodictyum cancellatum (Lamarck) . . . . . . . . . . . . . . . . . . . . . . . . Echinodictyum carlinoides (Lamarck) . . . . . . . . . . . . . . . . . . . . . . . . . Echinodictyum clathrioides Hentschel . . . . . . . . . . . . . . . . . . . . . . . . . Echinodictyum conulosum Kieschnick . . . . . . . . . . . . . . . . . . . . . . . . . Echinodictyum costiferum Ridley . . . . . . . . . . . . . . . . . . . . . . . . . . . . ? Echinodictyum fruticosum Hentschel . . . . . . . . . . . . . . . . . . . . . . . . Echinodictyum lacunosum Kieschnick . . . . . . . . . . . . . . . . . . . . . . . . . Echinodictyum mesenterinum (Lamarck) . . . . . . . . . . . . . . . . . . . . . . . Echinodictyum nidulus Hentschel . . . . . . . . . . . . . . . . . . . . . . . . . . . . Echinodictyum rugosum Ridley & Dendy . . . . . . . . . . . . . . . . . . . . . . Incertae sedis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Genus Tethyspira Topsent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Genus Sigmeurypon Topsent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Genus Cantabrina Ferrer-Hernandez . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phylogenetic relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Biogeography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Australian Raspailiidae 1181 Abstracf The marine sponge family Raspailiidae Hentschel is revised and referred to the order Poecilosclerida. Of 48 nominal genera, 17 (including one new genus and one new subgenus) are recognised here: Raspailia Nardo, (Hymeraphiopsis, subg. nov.), Ectyoplasia Topsent, Endectyon Topsent, Trikentrion Ehlers, Cyamon Gray, Aulospongus Norman, Raspaciona Topsent, Rhabdeurypon Topsent, Eurypon Gray, Plocamione Topsent, Amphinomia, gen. nov., Lithoplocamia Dendy, Hymeraphia Bowerbank, Ceratopsion Strand, Thrinacophora Ridley, Axechina Hentschel and Echinodictyum Ridley, and three genera are incertae sedis (Tethyspira Topsent, Sigmeurypon Topsent, Cantabrina Ferrer-Hernandez). Fifty-six species are described for the Australian fauna, of which 14 are new to science: Raspailia daminensis, R. desrnonyiformis, R. keriontria, R. melanorhops, R. phakellopsis, R. reticulata, R. wardi, R. wilkinsoni, Ectyoplasia vannus, Endectyon elyakovi, Ceratopsion montebelloensis, C . palmafa, Echinodictyum austrinus, spp. nov. and Amphinomia sulphurea, gen. nov., sp. nov. The phylogenetic relationships and biogeographical distribution of the family are discussed. Tropical north-westem Australian-southem Indonesia has the highest diversity of species in the Indo-west Pacific, and altogether the Australasian region has about 20% of the world's known raspailiid fauna. Introduction Sponges in the Raspailiidae are usually branching or lobate, cylindrical or flattened, stalked or vasiform, and with holdfasts by which they attach to the substratum. They typically have hispid (spiky) surfaces and a harsh consistency. Live coloration is highly variable, although not as diverse as in other Poecilosclerida, such as Microcionidae or Mycalidae. Colours range from bright reds and oranges to yellow, brown and black. Raspailiids are found in many habitats, from clear-water coral reefs to silt-laden estuaries, but they appear to be predominant in turbid waters, on submerged rock reefs or coral-rubble. Raspailiidae occur in all Australian marine provinces, from antarctic to tropical waters, but they are particularly abundant in the tropics. They are known from intertidal regions to depths of nearly 1300 metres. No publications on Australian Raspailiidae have appeared since the early 1900s. Prior to the present work, 40 named forms had been described from the continent and the Australian Antarctic Territories, although some of these are synonyms of other species, and over 80% of these were thought to be endemic to the region. Many of these species are poorly known or barely recognisable from their brief published descriptions, many lacking illustrations. Major works for this region include Lamarck (1813, 1814 [Topsent 1932]), Carter (1882, 1885), Ridley (1884), Ridley & Dendy (1886, 1887), Lendenfeld (18876, 1888), Dendy (1896), Kieschnick (1900), Whitelegge (1907) and Hentschel (1911, 1912, 1914). Several other publications also provide records of species with known Australasian distributions (including New Zealand, the western Pacific and southern Indonesia) (Whitelegge 1889, 1901; Dendy 1897; Hallmann 1912, 1914a, 1914b, 1916a, 1916b; Brondsted 1934; Burton 1934a; Bergquist 1961, 1970; LCvi 1967; LCvi & LCvi 1983; Wiedenrnayer 1989). This paper revises the Australian raspailiid sponge fauna, bringing the total number of species known to 56, including 14 new species. Worldwide, 48 nominal genera of Raspailiidae are recorded, with over 250 nominal species described, but only 17 are recognised here (including one new genus). The Australian fauna represents approximately 20% of the world's known species. This paper is the first attempt at revising the Raspailiidae on the basis of a widely distributed fauna, which is not surprising considering that the family was erected as late as 1923, but there has been some previous discussion on the higher systematic relationships of the family. Vosmaer (1912) and Wilson (1921) attempted to revise typical raspailiids (i.e. the genus Raspailia), but both revisions were grossly inadequate and restricted to few European species. Vosmaer (1912) followed the pioneering authors Ridley & Dendy (1887) in linking Raspailia with genera of the family Axinellidae (Axinella, Phakellia and Acanthella), indicating that skeletal architecture was the most important character for him. Topsent (1894) and Dendy (1905) had previously suggested that these features were 1182 J. N. A. Hooper more likely to be convergences, choosing instead megasclere and microsclere geometry as principal diagnostic features. Wilson (1921) maintained skeletal architecture as the primary systematic character, and subsequently most authors have also done so by including the family within the (polyphyletic) order Axinellida. Wilson's (1921) paper was valuable in documenting numerous instances of interspecific variability amongst a number of characters in species of Raspailia, noting that characters appeared and disappeared amongst species in all combinations, and concluding that Raspailia was a 'comprehensive, heterogeneous' group. The genus remains so today. Vosmaer (1935) made a number of valuable observations, being one of the earliest authors to document intra- and interspecific variability in morphological and morphometric characters. Unfortunately, Vosmaer's work suffered from excessive lumping of species: he synonyrnised 21 species under the name Raspailia hispida (Montagu). Many of these lived in diverse localities and habitats, and many had morphological characters which would otherwise clearly differentiate them from sibling species. Bergquist (1970) provided the most recent review of the Raspailiidae, describing nine New Zealand species of Raspailia, Clathriodendron, Eurypon, Ceratopsion and Thrinocophora. She placed primary emphasis on skeletal architecture, but also drew attention to other features such as ectosomal specialisation and geometry of echinating acanthostyles, which in some cases highlighted obvious similarities between raspailiids and other poecilosclerids. The conclusions of all these revisions serve to illustrate the difficulties faced in sponge systematics, in choosing important diagnostic features over more variable characters and in deciding on levels of taxonomic divergence. The present study is a worldwide revision of the genera of Raspailiidae, with de'scription of Australasian species. The family is referred here to the order Poecilosclerida, with a unique (apomorphic) feature being the possession of a specialised ectosomal skeleton (consisting of groups of ectosomal spicules surrounding single protruding subectosomal or choanosomal spicules). Similarities with the family Microcionidae are emphasised, with a shared primitive feature (plesiomorphy) being the possession of echinating acanthostyles. Genera are differentiated by a combination of characters: acanthostyle geometry and localisation of megascleres within particular regions of the skeleton. Skeletal architecture is included in analyses but not given primary diagnostic importance. Materials and Methods Collection of Material Most of the material examined in this study was obtained from northern and western Australia. These collections were made by the author and his associates between 1982 and 1990, from approximately 5000 km of coastline extending from the Houtman-Abrolhos Is, W.A., in the south-west, to the Gulf of Carpentaria, Qld, in the north-east, and to Ashmore Reef, Territory of Australia, to the west. Intertidal material was collected by hand, subtidal (5-30 m depth) material by scwa, and deeper water (32-800 m depth) material by dredging and trawling. In addition, seasonal sampling was conducted over 2 years in the Darwin region. All localities north of Northwest Cape, W.A., are subjected to large tides, up to 10 m in amplitude (ELWS-EHWS) in some areas, and consequently most sponges studied are peculiar to areas of high sedimentation and turbidity. However, other habitats were also investigated, including clear-water atolls and fringing coral reefs from the Houtman-Abrolhos and Exmouth regions, W.A., Rowley Shoals, Ashmore Reef, Territory of Australia, and New Year I. and Wessel Is, N.T. The sponge fauna of north-westem Australia is particularly rich and diverse, with over 780 species collected since 1982 (personal observations, NTM collections). This fauna is reportedly the richest in tropical Australia (J.E.N. Veron, personal communication; L6vi 1979), but this diversity is also apparently comparable with the Bass Strait region (F. Wiedenmayer, personal communication). Selected references on the shallow-water physical marine and intertidal topography, marine climatic information, and some aspects of the marine benthos of this region include Pope (1967), Semeniuk (1985), Michie (1987), Hanley (1988) and Hooper (1988~).Chiffings (1983) gives further bibliographic details and other information on characteristics of the Northwest Shelf region of W.A. Specimens collected by SCUBA were usually photographed in situ with Nikonos II and V camera systems, and/or photographed on deck soon after collection. Peter Young and Keith Sainsbury (CSIRO Fisheries, Hobart) kindly gave me access to their photographic collections from the Northwest Shelf, made by CSIRO RV Soela remote-sensing cameras mounted on trawls. Additional photographic Australian Raspailiidae 1183 material of deeper-water sponges was taken on deck soon after collection by the author, W. Houston, B.C. Russell and A. Mussig (N.T. Museum), and T. Ward (CSIRO Fisheries). Wherever possible, notes were made on the live coloration of sponges, and subsequent colour changes occurring after collection and exposure to air. Other data, such as shape, consistency, mucus production, surface features and microhabitat distribution in situ, were noted wherever possible. Northern and western Australian material described in this investigation has been accessioned into the collections of the N.T. Museum (registration prefix NTM Z). The author was also granted access to various collections in other Australian institutions (QM, AM, NMV, TM, SAM, WAM, NCI). Preparation and Examination of Material Specimens were usually fixed in 80-100% methylated ethanol soon after collection, although some material was initially frozen. Sponges were later transferred to 70% alcohol for preservation. Where possible, fragments or entire duplicate specimens were kept frozen for biochemical analyses. Nitric acid spicule preparations and thick-section mounts were routinely made as follows. Fragments of each sponge, including ectosomal and choanosomal regions were boiled directly on a glass microscope slide in nitric acid, and mounted in Canada balsam. Thick, hand-cut sections were made perpendicular to the surface, soaked in a saturated solution of phenol and xylene, and mounted in Durcupan (ACM Fluka Products). Microtome sections, cut at 30-35 Dm, were also made for each species. Fragments were embedded in wax, cut, and stained with basic fuchsin. Fragments of dry specimens (e.g. type material) were reconstituted in 5% buffered formalin for 12 h, which produced rehydration of the mesohyl and enabled cleaner histological sections to be made. Morphometric Analyses Spicules were measured from nitric acid histological preparations, using a camera lucida and calibrated stage micrometer. Twenty-five spicules, of each spicule category, were measured for all specimens. Line drawings of spicules and sections were made with a calibrated camera lucida, and photomicrographs were taken with a Leitz MPS microphoto system attached to a Wild-Leitz Dialux compound microscope and a Wild macroscope. Where possible, identifications of all Australian species were checked by comparison with type material. Taxonomic keys were constructed using unordered multistate characters, utilising the DELTA computer software system (Dallwitz & Paine 1986). Scanning Electron Microscopy Two methods were used to prepare sponge samples for scanning electron microscopy (SEM). Nitric acid preparations of spicules were made directly onto microscope slide cover-glasses, mounted on stubs, and sputter-coated with gold for up to 60 s. Thick hand-cut sections of sponges were soaked in chlorine bleach until all loose (type B or collagenous) spongin had dissolved (determined microscopically). Sections were then etched in hydrogen peroxide and removed before any disintegration of skeletal structure occurred. Sections were washed several times in ethanol, dehydrated in absolute ethanol, and dried flat between hvo glass slides in an embedding oven for several hours. Thick sections were mounted on stubs with double-sided adhesive tape, sputter-coated with gold for up to 140 s, and examined with Phillips SR500 and Joel scanning electron microscopes (University of Queensland and Northern Territory University EM Units). Phylogenetic Analysis of Morphometric Characters Morphometric data were tabulated and coded numerically for each taxon. Data matrices were constructed for groups of taxa, usually in multistate format. Phylogenetic analyses of species groups were undertaken using a numerical computer-generated cladistic routine for infemng phylogenies (PAUP: Swofford 1985), which produced minimum length trees under the principle of maximum parsimony, and determining plesiomorphy by outgroup comparisons. These analyses utilise the Wagner method, taking the preferred phylogenetic tree as the most parsimonious one, i.e. the one with the fewest evolutionary steps, based on the principle that minimising homoplasy is the same thing as maximising parsimony in evolution. A consensus tree was derived from a number of minimum length trees obtained through PAUP, and this final tree was produced using CONTREE (issued with PAUP). Abbreviations used in the Text AAT, Australian Antarctic Territories; AFZ, Commonwealth Fisheries Observer Program, Australian Fishing Zone; AHF, Alan Hancock Foundation, University of Southern California, Los Angeles; AIMS, Australian Institute of Marine Science, Townsville, Qld; AM, Australian Museum, Sydney; BMNH, The 1184 J. N. A. Hooper Natural History Museum [formerly the British Museum (Natural History)], London; CPMNP, Cobourg Peninsula Marine National Park, N.T.; CSIRO, Commonwealth Scientific and Industrial Research Organisation, Division of Fisheries, Hobart; DELTA, Description Language for Taxonomy computer system (see Dallwitz & Paine 1986); EIS, Environmental Impact Study; EPMFR, East Point Marine Fish Reserve, Darwin, N.T.; GBR, Great Bamer Reef, Qld; ICZN, International Code of Zoological Nomenclature (see Anonymous 1985); IM, Indian Museum (Zoological Survey of India), Calcutta; JCU, James Cook University of North Queensland, Townsville; LFM, Merseyside County Museums (formerly Liverpool Free Museum), Liverpool; LMJG, Abteilung fur Zoologie am Landesmuseum Joanneum (Landes Museum Jubileum Graz), Graz; MHNG, Museum d'Histoire Naturelle de Geneve, GenCve; MMBS, Mukaishima Marine Biological Station, Faculty of Science, Hiroshima University, Onomichi; MNHN LBIM, MusCum National d'Histoire Naturelle, Laboratoire de Biologie des InvertebrCs Marins et Malacologie, Paris (DT, Topsent collections; DCL, LCvi collections; DJV, Vacelet collections; DNBE, Boury-Esnault collections); MOM, MuseC OcCanographique de Monaco, Monaco; NCI, United States National Cancer Institute (Townsville contract), Marine Bioactivity Unit, Australian Institute of Marine Science, Townsville, Qld; NM, Natal Museum, Pietermaritzburg; NMB, Naturhistoisches Museums zu Basel, Basel; NMNZ, National Museum of Natural History (formerly Dominion Museum), Wellington; NMV, Museum of Victoria (formerly National Museum of Victoria), Melbourne; N.S.W., New South Wales; N.T., Northern Temtory; NTM, Northern Temtory Museum of Arts and Sciences, Darwin, N.T.; NWS, Northwest Shelf, W.A.; PAUP, Phylogenetic Analysis Using Parsimony (see Swofford 1985); PIBOC, Pacific Institute of Bio-organic Chemistry, Far East Scientific Centre, Academy of Sciences of the U.S.S.R., Vladivostok; PMJ, Phyletisches Museum, Jena; PNG, Papua New Guinea; QFS, Queedsland Fisheries Service; Qld, Queensland; QM, Queensland Museum, Brisbane; RRIMP, Roche Research Institute of Marine Pharmacology, Sydney (now defunct, collections in AM); SA, South Australia; SAM, South Australian Museum, Adelaide; SM, Musee Zoologique, Strasbourg; SMEM, Station Maride d'Endoume, Marseille; SMF, Natur-Museum und Forschungsinstitut Senckenberg, Frankfurt; Tas, Tasmania; TM, Tasmanian Museum and Art Gallery, Hobart; USNM, National Museum of Natural History, Smithsonian Institution, Washington; Vic, Victoria; W.A., Western Australia; WAM, Westem Australian Museum, Perth; ZMA, Instituut voor Taxonomische Zoologie, Zoologisch Museum, Universiteit van Amsterdam, Amsterdam; ZMB, Zoologisches Museum fiir Naturkunde an der Humboldt-Universitat zu Berlin, East Berlin; ZMH, Zoologisches Institut und Zoologisches Museum der Universitat Hamburg, Hamburg. Analysis of Raspailiid Characters Terminology In classifying anatomical or morphological features in the Demospongiae, three major regions are delineated (Fig. 1). These are based on the distribution of epithelial cells (pinacocytes) and inorganic skeleton. The ectosome consists of an outer layer of flattened epithelial cells that cover the external (exopinacocytes) and basal surfaces (basopinacocytes). In the strict sense, the ectosome is only one cell thick, but traditionally sponge taxonomists refer to the ectosomal skeleton as those groups of spicules that originate from, or just below, the ectosomal region, which may or may not project through the surface or lie paratangential to it. The choanosome is the internal region containing choanocytes, or collar cells, bounded by the exo- and basopinacocytes on the exterior surface, and the endopinacocytes, which line the incurrent canals (opening to the exterior by small pores or ostia), and excurrent canals (opening to the exterior through larger pores or oscula). The choanosomal skeleton includes a silicious mineral (spicule) component, an organic (fibre or type A spongin) component, and an organic intercellular region termed the mesohyl (comparable to the mesenchyme of higher metazoans), which usually contains deposits of type B or collagenous spongin. In the Raspailiidae the choanosomal skeleton may be divided into two major regions. The axial choanosomal skeleton consists of a central core that typically contains a condensed or reticulate arrangement of spongin fibres and/or spicule tracts. The extra-axial subectosomal skeleton refers to the fibre and/or mineral skeleton that arises from the axis, ascends to the ectosomal region and often protrudes through it. In some genera (e.g. Echinodictyum) there are only rudimentary differences between axial and extra-axial skeletons, consisting of one or few subectosomal spicules. In other genera (e.g. Ceratopsion) there is well marked axial and extra-axial differentiation, and the extra-axial skeleton may be plumose, plumo-reticulate, reticulate or radial. In encrusting species it may be reduced to a plumose-hymedesmoid, or even a hymedesmoid Australian Raspailiidae mg Fig. 1. Diagrammatic representation of a cross section through a branch of a sponge, illustrating a typical raspailiid skeleton and showing the major morphological features. a, Ectosomal skeleton, composed of ectosomal auxiliary spicules; b, axial skeleton; c, extra-axial skeleton; d, choanosomal skeleton; e, ectosome;f , exhalant pores (oscula); g, inhalant pores (ostia); h, fibres (composed of spongin type A); i, mesohyl (containing spongin type B); j, fibre mesh; k, choanocyte chamber; I, subectosomal extra-axial spicule; rn, choanosomal axial spicules coring fibres; n, echinating acanthostyles standing perpendicular to fibres, with at least their bases embedded in spongin type A. structure, consisting of single spicules standing erect on a basal layer of spongin lying on the substrate (Fig. 2). Spicules may be classified on the basis of their size (megascleres or microscleres), their geometry and patterns of spination (Figs 4, 6; and refer to glossary in Wiedenmayer 1977), and also on their distribution within the skeleton (after Hallmann 1912 et seq.). Thus, oxeas, which form brushes in the ectosomal skeleton, may be termed ectosomal auxiliary oxeas. Similarly, anisoxeas, which core spongin fibres in the choanosomal axial skeleton, are called choanosomal axial anisoxeas. Long styles that have their bases embedded in the axial skeleton, and protrude outwards to form radial tracts in the subectosomal region are called subectosomal extra-axial styles. Spined styles that stand perpendicular to fibres, with their bases embedded in the spongin, are echinating acanthostyles. Another term used by contemporary authors is principal spicules, referring to the main structural spicules that core spongin fibres. For some groups, such as the Microcionidae, which can have more than one sort of spicule coring fibres, this term is appropriate, but for the Raspailiidae, where only choanosomal megascleres core spongin fibres, the distinction is unnecessary. Major Morphometric Characters used to Differentiate Raspailiids Five morphological characters (skeletal architecture, ectosomal specialisation, modifications to echinating megascleres, geometry and size of other megascleres and presence of microscleres) J. N. A. Hooper Fig. 2. Examples of skeletal types found amongst the Raspailiidae. a, Condensed axis, radial extra-axis (e.g. Raspailia vestigifera, NTM 22640); b, condensed axis, plumose extra-axis (e.g. Raspailia pinnatifrda, NTM 21501); c, condensed axis, reticulate extra-axis (e.g. Raspailia wilkinsoni, NTM 22734); d, reticulate axis and extra-axis (e.g. Echinodictyum rugosum, NTM 21945); e, halichondroidreticulate axis, radial extra-axis (e.g. Raspailia darwinensis, NTM 21259); f, renieroid-reticulate axis and extra-axis (e.g. Amphinomia sulphurea, NTM 21809); g, axial or basal reticulation, radial or plumose extra-axis (e.g. Plocamione histrix, MNHN LBIM DT1416); h, plumose-hymedesmoid (compressed basal skeleton, radial extra-axis) (e.g. Hymeraphia pilosella, MNHN LBIM DT933). Australian Raspailiidae 1187 are most obvious within the Raspailiidae, and as such they have been used most frequently in the systematics of the group. Genera of Raspailiidae are not easily differentiated by any single feature (or apomorphy), since characters appear throughout the groups in all combinations. The present study places importance on a combination of features to define genera. Skeletal architecture A wide diversity of skeletal types is known to occur in the Raspailiidae, and four groups of genera may be differentiated on that basis. (1) The typical condition, termed 'classical raspailiid skeleton' by authors, is represented by a compressed or reticulate axial skeleton, which occurs together with a well differentiated radial, plumose or plumo-reticulate extra-axial skeleton (e.g. Raspailia s.s., Endectyon, Ectyoplasia, Trikentrion, Ceratopsion and others) (Fig. 2a-c). (2) The second group contains genera with simple reticulate architecture, without marked axial and extra-axial differentiation [e.g. Raspailia (Clathriodendron), Echinodictyum] (Fig. 2d-e), and in one genus (Amphinomia) this reticulation is regularly renieroid (Fig. 2f). (3) The third group includes the 'plocamiid' genera Plocamione and Lithoplocamia, which have a basal or axial reticulation of acanthose diactinal, quasidiactinal or monactinal megascleres, together with a plumose or plumo-reticulate extra-axial skeleton (Fig. 2g). (4) The final group of genera has a reduced hymedesmoid basal skeleton, with microcionid plumose or radial hymedesmoid extra-axial skeletal construction (e.g. Eurypon, Hymeraphia, Cyamon) (Fig. 2h). Included amongst this group of skeletal morphs is Aulospongiella (=Raspailia s.l.), which has a microcionid plumose skeletal architecture (Fig. 65b). Differences in skeletal construction are usually the most prominent features contrasted between species of Raspailiidae, and, as such, those features have in the past sewed as the primary, or even ultimate, diagnostic criterion for earlier authors (e.g. Topsent 1928; de Laubenfels 1936). However, the use of skeletal architecture to the exclusion of other features (e.g. ectosomal skeletal development, spicule geometry) may be completely misleading since that system splits taxa that otherwise have obviously close affinities (e.g. Raspailia s.s. and Clathriodendron). Furthermore, these skeletal types are certainly not unique to this family, and the complete range of structures cited above is also recorded for the Microcionidae, with species showing (1) classical raspailiid construction [e.g. Clathria (Axociella) canaliculata (Whitelegge)], (2) strictly reticulate [e.g. C. (Thalysias) vulpina (Lamarck)] and renieroid skeletons [e.g. C. (T.) nuda Hentschel], (3) plocamoid basal skeleton [e.g. Antho (Navicu1ina)'ridleyi (Hentschel)], (4) plumose [e.g. C. (T.) coralliophila (Thiele)] or simply hymedesmoid architecture [e.g. C. (T.) robusta (Dendy)] (Hooper, unpublished data). Ectosomal specialisation The presence of special ectosomal megascleres grouped into brushes on the surface surrounding one or more large, central, protruding extra-axial megascleres is unique for the Raspailiidae (Fig. 3a). Previously, this feature had also been considered to be completely consistent for a particular species, but there are now documented examples to suggest that ectosomal specialisation may vary intra-specifically (see remarks below for Raspailia arbuscula). Similarly, the presence or absence of ectosomal specialisation is not necessarily consistent for all species within a particular genus. This feature occurs in typical Raspailia, for example, but there are many other species (e.g. R. Jlaccida Bergquist, 1970: 27, R. inaequalis Dendy, 1924: 355) that do not have ectosomal specialisation but otherwise are obvious members of the genus. Similarly, typical Echinodictyum lack any form of ectosomal specialisation, and most species have a membraneous skin-like covering that overlays the peripheral fibres. There is at least one species now known for the genus that does have a typical raspailiid ectosome (i.e. E. nidulus). These anomalies illustrate the impossibility of defining higher taxonomic groups on the basis of one or few morphological characters. The variability in the ectosomal skeleton in the Raspailiidae is illustrated in Fig. 3. 1188 J. N. A. Hooper Fig. 3. Ectosomal features of the Raspailiidae. a, Typical raspailiid condition, with a specialised skeleton consisting of single extra-axial spicules protruding through the surface, each bearing brushes of ectosomal spicules surrounding their points of insertion through the surface (Raspailia vestigifera, NTM 23479); b, typical condition (Ectyoplasia vannus, NTM 22497); c, reduced condition, in which brushes of ectosomal spicules are perched on the surface, overlying the larger extra-axial spicules in the subdermal region (Plocamione pachysclera, NTM 23495); d, specialised ectosomal skeleton is absent, but choanosomal megascleres form dense surface brushes (R. desmoxyiformis, NTM 21259); e, specialised ectosomal skeleton is absent, ectosome consists of a darkly pigmented membraneous skin-like covering, through which choanosomal or subectosomal megascleres may protrude (Echinodictyum mesenterinum, (NTM 256);f, ectosomal skeleton is absent, with thick subdermal collagenous spongin, with no or few protruding choanosomal megascleres (Amphinomia sulphurea, NTM 21787). Echinating acanthostyles The presence of acanthostyles in raspailiids, and the geometry and patterns of spination of these spicules in many species (e.g. Raspailia viminalis) indicate affinities with other poecilosclerid groups, especially the Microcionidae, if that character is attributed any systematic value. Some microcionids, such as Clathria (Thalysias) ridleyi (Lindgren), C. (T.) mutabilis (Topsent) and C. (T.) topsenti (Thiele), also have modified acanthostyles which are otherwise characteristic of specialised raspailiid groups, such as ~ c t y o ~ l & i a tabula (Hooper, unpublished data). Those shared characteristics suggest a number of Australian Raspailiidae 1189 possibilities: that the raspailiids and microcionids may have diverged only relatively recently from common ancestral stock; or that acanthose echinating megascleres may be persistent, ancestral, and relatively conservative throughout the evolution of those groups; or that geometric modifications are independently acquired, convergent, and a consequence of (some unknown) functional adaptation, and these do not give any clues to phylogeny. Whichever explanation is supported, undoubtedly many species in both families show close similarities in the geometry of their acanthostyles. These features-the geometry and pattern of spination on acanthostyles, and unusual modifications to these spicules-have been used as important diagnostic criteria for genera of Raspailiidae (e.g. Topsent 1928), several genera being erected solely on that basis (e.g. Trikentrion, Endectyon, Hemectyon). The variability in the morphology of acanthostyles amongst the Raspailiidae is shown in Fig. 4. Spine morphology also may be useful in differentiating particular taxa. This is best studied using scanning electron microscopy, and it has proved most useful in distinguishing between sibling species of Echinodictyum. Raspailiidae have four spine types, viz. (1) cylindrical, sharply pointed but not recurved spines (Fig. Sa), (2) spicules with sharply pointed and cylindrical spines which have a recurved point (Fig. 5b), (3) spines which are recurved, spatulate and have rounded margins (Fig. 5c), or (4) spines which are recurved, spatulate and bear serrated margins (i.e. secondary spination) (Fig. 5 4 . Geometry and size of megascleres The variety of other megasclere types (other than acanthostyles) found in the axial, extra-axial and ectosomal skeletons of the Raspailiidae is illustrated in Fig. 6. Geometric modifications to these spicules have not been used previously as taxonomic characters of any great importance in the higher systematics of the Raspailiidae, although they are extensive. This is undoubtedly due to the variability of spicule types that may occur in sibling species (e.g. Wilson 1921). However, some genera can be defined by their constituent structural megascleres. These genera include Echinodictyum (with axial oxeas, extra-axial styles, and usually without ectosomal spicules), Amphinomia (with axial styles bearing terminal spines, extra-axial styles, but no ectosomal spicules), Plocamione (with axial or basal acanthostrongyles, extra-axial styles or anisoxeas, and ectosomal styles or anisoxeas), Axechina (with oxeas bearing terminal spines and anisoxeas both forming the axial skeleton, extra-axial styles and flexuous, toxa-like ectosomal styles with spined ends) and Aulospongus (with axial rhabdostyles, echinating rhabdostyles, and no extra-axial spicules). By comparison, other genera cannot be well defined by the possession of one type of structural megasclere over another. The genus Raspailia shows greatest heterogeneity in its megasclere complement: in most species the most common structural spicules are styles, but in otherwise closely related species these spicules may vary from true styles, styles with oxeote modifications, to strongyles or true oxeas (e.g. R. viminalis). Although intra-specific variability in spicule geometry for the Raspailiidae is undocumented, most species have been clearly defined by the form and diversity of their spicules. Examination of many individuals of some species from different geographical localities indicates that spicule geometry varies minimally between samples. Similarly, spicule dimensions have also been used to define species, and their greatest taxonomic value is in helping to distinguish between sibling species. For the Raspailiidae no studies have examined populations of species, and consequently there has been no attempt at defining the variability of spicule dimensions in relation to biotic and abiotic factors. Present results show that within broad ranges it is possible to characterise a particular species by its spicule size, and this character is most useful in conjunction with other features. There is some indication, however, that over large geographical ranges, absolute spicule dimensions for particular species may vary considerably (see below for Echinodictyum mesenterinum). Presence of microscleres Unlike most other Poecilosclerida, which usually have diverse forms of microscleres (isochelae, anisochelae, sigmas, toxas, microxeas, etc.), and in which family-level systematics 1190 J. N. A. Hooper Fig. 4. Range of acanthostyle morphology found in the Raspailiidae. a, Acanthostyle with barren base (Raspailia wardi, NTM 21319); b, evenly spined acanthostyle with erect spines (R, phakellopsis, NTM 2610); c, evenly spined acanthostyle (R. cacticutis, BMNH 1886.12.15.120); d, vestigial acanthosubtylostyle (R. vestigifera, NTM 22356); e, acanthotylostyle with barren base (R. irregularis, SAM S537); f , acanthorhabdostyle with barren base (R. compressa, NTM 21748); g, clavulate acanthostyle with barren base (Ectyoplasia tabula, NTM 21383); h, slender acanthostyle (Eurypon polyplumosa, MNHN LBIM DCL1296); i, vestigial acanthorhabdostyle (Aulospongus tubulatus, BMNH 1931.11.28.18); j, acanthosubtylostyle with bulbous point (Echinodictyum glomeratum, BMNH 1881.21.283); k, sharply pointed, microcionid-like acanthostyle (E. rugosum, NTM 21945); 1, acanthorhabdostyle with recurved spines (Endectyon fruticosa aruensis, SMF 984); m, clavulate Australian Raspailiidae 1191 Fig. 5. Variation in spine structure found on the acanthostyles of Raspailiidae. a, Cylindrical, sharply pointed, erect spines (Raspailia phakellopsis, NTM 21950); b, cylindrical, sharply pointed, recurved spines (Echinodictyum cancellatum, NTM 22359); c, spatulate, recurved spines (E. mesenterinum, NTM 22291); d, spatulate, serrated, recurved spines (E. clathrioides, NTM 21709). relies heavily on microsclere geometry (e.g. de Laubenfels 1936), the Raspailiidae have only raphides (or rnicroxeas) occurring singly or in bundles (trichodragmata). Furthermore, these are present in only five genera [Trikentrion, Thrinacophora, Eurypon (which includes Tricheurypon and Protoraspailia), Aulospongus (including Rhaphidectyon) and Rhabdeurypon] and occur in only certain species. This is one reason why most authors have not included the Raspailiidae with Poecilosclerida: phylogenetic comparisons between these groups cannot be made on the basis of microsclere geometry. This is also an argument for not merging the Raspailiidae with the Rhabderemiidae (e.g. Hentschel 1923), which have rhabdostyle megascleres similar to those found in the raspailiid Aulospongus, but also have a diversity of unusual microsclere types (toxas, sigmas, microxeas, some or all of which may be spined) (Hooper 1!Bob). Other characters Other morphological features used extensively in the description and systematics of demosponges include growth form, texture or consistency, surface characteristics (including acanthostrongyle (E. thurstoni, NTM 21761); n, acanthostrongyle (E. pilosus, SMEM 1705(1)); o, acanthorhabdostyles with barren bases, two size categories (Aulospongus involutum, BMNH 1902.11.16.33); p, stellate acanthotylostyle with barren base (Hymeraphia stellifera, MNHN LBIM Dn501); q, vermiform acanthostrongyle (Lithoplocamia lithistoides, BMNH 1921.11.7.68); r, 1: venniform acanthotylostrongyles and acanthostyle, 2: club shaped derivative of the acanthostongyle (Plocarnione clopetaria, MNHN LBIM DT1416); s, club-shaped acanthostyle with peculiar spination (Plocarnionepachysclera, MNHN LBIM DCL2948); t, acanthose pseudoxeas (Rhabdeuryponspinosum, MNHN LBIM DJV4); u, acanthoplagioquatriaene or echinating sagittal quadract (Cyamon aruense, SMF 1618); v, acanthoplagiodiaene or echinating sagittal diact (Trikentrionjlabelliforrne, NTM 21265); w, acanthostrongyles with perpendicular spines (Tethyspira spinosa, BMNH 1877.5.21.394) (a dubious raspailiid). J. N. A. Hooper Fig. 6. Examples of the major geometric forms of megascleres found in the axial, extra-axial, and ectosomal skelktons of the ~ a s ~ & i d a e .a-m, Axial skeleton: a, fusiform style; b, hastate style; c, subtylote style with stepped point; d, rhabdostyle; e, acanthose style; f, fusiform oxea; g, hastate oxea; h, anisostrongyleoxea or styloid; i, strongyloxea; j, hastate anisoxea; k, anisoxea; I, centrangulate or sinuous oxea; m, acanthose oxea. n-q, Extra-axial skeleton: n, fusiform style; o, sinuous style; p, subtylostyle; q, oxea. r-x, Ectosomal skeleton: r, fusiform style; s, hastate style; t, anisostrongyle oxea or styloid; u, sinuous acanthose style; v, fusiform oxea; w, hastate oxea; x, anisoxea. Australian Raspailiidae 1193 the distribution of oscula and ostia, subdermal drainage canals and any hispid features of the surface), live coloration and solubility (or stability) of pigments, mucous production upon collection and preservation, amount of skeletal spongin (both fibre type A spongin and collagenous type B spongin), proportion of the silica component to organic material in the skeleton, and size and shape of choanocyte chambers. These and other non-morphological features that may be useful in sponge systematics are discussed in general texts on Porifera (e.g. U v i 1973; Bergquist 1978; Simpson 1984). In addition, Hooper (1990a) and Hooper et al. (1992) investigate the relationships between various Poecilosclerida (including Raspailiidae and Microcionidae) and Axinellida using biochemical methods, and these authors provide additional references for these data. Systematics Class DEMOSPONGIAE Sollas Order POECEOSCLERIDA Topsent Family RASPAILIIDAE Hentschel Raspailiidae Hentschel, 1923: 407 (in part).-LBvi, 1973: 608; Bergquist, 1970: 26; 1978: 167; Hartman 1982: 647. Euryponidae Topsent, 1928: 59.-LCvi, 1973: 608; Hartman, 1982: 648 (in part); Uriz, 1988: 54. Definition 'Qpically branching, less often massive, flabelliform or encrusting sponges. Surface typically hispid, even, but forms with microconules or entirely smooth surfaces also known. Ectosomal skeleton unique to family, typically well developed but with great variation between species, with ectosomal spicule brushes (fine styles, oxeas or anisoxeas) grouped around one or more long, central, extra-axial spicule (styles, oxeas or anisoxeas). Skeletal construction usually well developed, consisting of at least 2 differentiated structures: axially or basally condensed choanosomal skeleton, and extra-axial subectosomal skeleton, which may be radial, plumose or plumo-reticulate. Taxa with predominantly reticulate skeletal choanosomal architecture preserve at least some rudiments of extra-axial skeleton, in form of single (or groups of) subectosomal spicules. Spongin fibres usually echinated by acanthostyles which may be secondarily lost. Acanthostyle geometry extremely diverse. Structural megascleres vary from true styles to true oxeas, with intermediate forms. Microscleres occur in only few taxa, and when present consist of raphides (or microxeas). Remarks This emended definition of Raspailiidae was developed from Hentschel's (1923) original concept of the family, which initially allowed for the inclusion of taxa bearing chelate and sigmoid microscleres. It also incorporates Bergquist's (1970, 1978) more recent discussions on raspailiids, in which she merged Euryponidae with Raspailiidae. Both families are certainly closely related, but some authors prefer to separate them on the basis of encrusting growth form, presence (Euryponidae) or absence (Raspailiidae) of raphide microscleres, and the presence or absence of (tetractinal) modifications to echinating megascleres (e.g. Uriz 1988). However, those features are known to vary between genera and species in other poecilosclerid families (e.g. Microcionidae, Myxillidae), and are not here considered to be important at the family level of classificiation. Furthermore, the Euryponidae includes both encrusting (e.g. Eurypon) and flabelliform species (e.g. Trikentrion), and the range of skeletal construction in euryponids is as diverse as in other Raspailiidae and other Poecilosclerida (e.g. Microcionidae), with both axially condensed and reticulate forms. On that basis the family Rhabderemiidae could also be included in Raspailiidae, but it is maintained here as a separate taxon possessing a diversity of unusual microscleres (cf. Hooper 1990b). U v i (1973: 608) restricted the definition of Euryponidae to encompass encrusting forms only, at the same time including genera such as Trikentrion which are habitually flabellate. Nevertheless, U v i formally differentiated Euryponidae from Raspailiidae by its encrusting J. N. A. Hooper 1194 growth form and presence of triactinal and tetractinal modifications to megascleres in two genera (Cyamon and Trikentrion). Euryponids characteristically have a skeletal architecture similar to typical raspailiids, and most species have ectosomal specialisation that is also reminiscent of Raspailiidae. l k o groups of euryponids are evident: one with remnants of classical raspailiid and microcionid architecture, albeit encrusting habit (Euryponidae, sensu L6vi 1973), the other which is encrusting or flabellate, with raspailiid structure and with triactinal or tetractinal megasclere modifications. The derivation of these peculiar megascleres is not definitely known, but it is likely that they represent highly derived echinating acanthostyles. Other raspailiid genera (e.g. Ectyoplasia, Endectyon) are also known to have modified acanthostyles, although modifications are never as pronounced as in euryponids. The present interpretation is that the Euryponidae (s.s.) (including the genera Eurypon, Acantheurypon and Tricheurypon) are probably no more than encrusting raspailiids. Those genera are confidently placed with the Raspailiidae (e.g. Bergquist 1970, 1978), whereas the affinities of the second group (Cyamon and Trikentrion) are slightly more obscure. The present study follows Bergquist (1970, 1978) in treating those species as derived raspailiids, and this decision is supported by their possession of other skeletal characters typical of the Raspailiidae. This is also supported by the independent occurrence of cladotylote modifications to megascleres in Acarnus [of the family Microcionidae (e.g. L6vi 1973) or Myxillidae (e.g. van Soest 1984; Hooper 1988b)l. Euryponids could be conveniently included as a subfamily of the Raspailiidae, differentiated from the nominotypical subfamily in having microscleres and/or tetractinal echinating megascleres. However, that division is artificial and internally inconsistent (Eurypon does not have sagittal macts; Cyamon does not have microscleres). Homologous structures in other poecilosclerid sister groups (Microcionidae and Myxillidae) appear to have greatest significance only at the generic level of classification. Key to the Genera of Raspailiidae Echinating spicules are microcionid-like, club-shaped, with rounded or sharp points, subtylote bases, and with evenly or unevenly distributed spines ........................ 2 Echinating spicules are acanthose, club-shaped or strongylote, with strongly curved hooks on the base and shaft (cladotylote), and these spicules are usually confined to a particular region within the skeleton ............................................... Endectyon Echinating spicules are acanthostyles with smooth rhabdose bases, and large recurved spines are distributed over the shaft ............................................ Aulospongus Echinating spicules are acanthostyles with bulbous tylote bases, with or without spines on the points and other modifications to the distal portion ..................... Hymeraphia Echinating spicules are club-shaped with clavulate points; axial and extra-axial skeletons are composed of a single category of undifferentiated choanosomal megascleres ........ ......................................................................................... Ectyoplasia Echinating spicules are absent, but diactinal acanthorhabds are dispersed throughout the skeleton ........................................................................... Rhabdeurypon Echinating spicules are saggittal monact-, diact- or tetractinal (acanthoplagiotriaenes) with only one spined ray ................................................................. Trikentrion Echinating spicules are saggittal tetract- or pentactinal (rarely with fewer rays) (acanthoplagiotriaenes) with all or most rays spined .............................. Cyamon Echinating spicules are absent .................................................................... 5 Choanosomal skeleton consists of slightly axially condensed reticulation of spongin fibres and/or spicule tracts ............................................................ Raspailia (s.s.) Choanosomal skeleton consists of basally condensed skeleton lying on the substrate ....... ..................................................................................................... 3 Choanosomal skeleton consists of a reticulation of spongin fibres and/or spicule tracts, without any trace of axial compression ..................................................... 4 Choanosomal skeleton consists of loosely aggregated or plumose axial fibres ............... ......................................................................................... Raspaciona Spicules in the axial skeleton are acathostrongyles or acanthotylostrongyles ................. ......................................................................................... Plocamione No true choanosomal megascleres occur in the axial skeleton, although large extra-axial spicules and echinating acanthostyles may be embedded in spongin fibres .... Eurypon Australian Raspailiidae 4(2). 1195 Spicules in the axial skeleton are choanosomal styles or subtylostyles with spines on the basal and distal ends .............................................................. Amphinomia Spicules in the axial skeleton are choanosomal oxeas ......................... Echinodictyum Spicules in the axial skeleton are acathostrongyles or acanthotylostrongyles ................. ...................................................................................... Lithoplocamia Choanosomal skeleton consists of dense axial compression of criss-crossed spicules, without axial fibres ............................................................. Thrinacophora Choanosomal skeleton consists of axially condensed reticulation of spongin fibres and/or spicule tracts .................................................................................... 6 Spicules in the axial skeleton are sinuous styles or anisoxeas .................. Ceratopsion Spicules in the axial skeleton are both spined oxeas and smooth anisoxeas ...... Axechina - 5(1). 6(5). ~ Genus Raspailia Nardo Raspelia Nardo, 1833: 522.-Burton, 1938: 33. Raspai1ia.-Nardo, 1847: 3 (nom. emend.); Schmidt, 1862: 59; 1866: 14; Ridley & Dendy, 1887: 188; Hanitsch, 1889: 161; Lendenfeld, 1890: 401; Topsent, 1894: 16; Pick, 1905: 19; Hentschel, 1912: 413; Vosmaer, 1912: 313; Wilson, 1921: 54-60; Vosmaer, 1935: 766; Bergquist, 1970: 26. Rasalia.-Gray, 1867: 522[lapsus]. Raspalia.-Gray, 1867: 523[lapsus]. Abila Gray, 1867: 522 [preocc.] (type species Raspailia freyerii Schmidt, 1862: 60, by monotypy; holotype unknown, possible SM or Graz, from Muggia Bay, Adriatic Sea). Not Abila Gray, 1867: 539. Abilana Strand, 1924: 33 [replacement name]; de Laubenfels, 1936: 102. Axinectya Hallmann, 1917: 393 (type species Axinella mariana Ridley & Dendy, 1886: 480, by original designation and monotypy; holotype BMNH 1887.5.2.28, from Marion I., western Pacific Ocean) (Fig. 7d-f). Clathriodendron Lendenfeld, 1888: 215.-Kirk, 1911: 579; Hentschel, 1911: 383; Hallmann, 1912: 295; Topsent, 1894: 19; de Laubenfels, 1936: 102; Bergquist, 1970: 30 (type species Clathriodendron arbuscula Lendenfeld, 1888: 215, by subsequent designation (Hallmann, 1912: 295); lectotype (here designated) AM G9045, from Port Jackson, N.S.W.) (Figs 19-20) . Dictyocylindrus Bowerbank, 1862: 1108 (in part).-Gray, 1867: 519; Topsent, 1890b: 289; de Laubenfels, 1936: 102 (type species Spongia hispida Montagu, 1818: 81, by original designation; holotype not found BMNH, from coast of Devon, England). Not Dictyocylindrus.-Carter, 1879b : 297. Echinaxia Hallmann, 1916a: 543.-Hallmann, 1917: 391; de Laubenfels, 1936: 102; Bergquist, 1970: 30 (type species Axinella frondula Whitelegge, 1907: 509, by original designation and monotypy; holotype AM G4349, from the south coast of N.S.W.) (Fig. 34). Parasyringella Topsent, 1928: 2 8 7 . 6 e Laubenfels, 1936: 102 (type species Raspailia (Syringella) falcifera Topsent, 1892a: 124, by typonymy; holotype MOM (not seen), schizotype MNHN LBIM DT 901, from San Jorge, North Atlantic) (Fig. 717. Raspailopsis Burton, 1959: 256 (type species Raspailopsis cervicornis Burton, 1959: 256, by original designation; holotype BMNH 1936.3.4.604, from the South Arabian coast) (Fig. 7g-i). Raspaxilla Topsent, 1913: 616.-Bergquist, 1970: 28-30 (type species Raspaxilla phakellina Topsent, 1913: 617, by monotypy; holotype MOM (not seen), schizotype MNHN LBIM DT 1614, from Antarctica) (Fig. 7k-I). Syringella of authors.-Ridley, 1884: 460; Pick, 1905: 18; Topsent, 1892a: 123; 1904: 138; 1928: 42; Burton, 19346: 42; de Laubenfels, 1936: 121. Not Syringella Schmidt, 1868: 10 (type species Raspailia syringella Schmidt, 1868: 10, by monotypy; holotype unknown, possibly SM or Graz, from the coast of Algeria, Mediterranean). Valedictyum de Laubenfels, 1936: 102 (type species Raspailia vestigifera Dendy, 1896: 47, by monotypy; holotype NMV G2468, from Port Phillip, Vic.) (Figs 15-16). Type species: Raspailia typica Nardo, 1833: 522, by monotypy (holotype unknown, fragments of Schmidt's specimen from the Adriatic Sea, MNHN LBIM DCL 1237L, BMNH 1867.3.11.8) (Schmidt, 1862: 59; Vosmaer, 1935: 778); senior synonym of Raspalia viminalis Schmidt, 1862: 59 [invalid subsequent designation of type species by Vosmaer, (1912: 313)] (holotype from Sebenico, Adriatic Sea, possibly extant in SM, schizotype MNHN LBIM DCL 1238L) (Fig. 7a-c). J. N. A. Hooper Fig. 7. a-c, Raspailia viminalis Schmidt (schizotype MNHN LBIM DCL 1238L from the Adriatic Sea) (nominal type species of the genus Raspailia Nardo): a, axial style; b, extra-axial style; c, extra-axial styles. d-f, Axinella mariana Ridley & Dendy (holotype BMNH 1887.5.2.28 from the western Pacific Ocean) (type species of the nominal genus Axinectya Hallmann): d, echinating acanthostyle; e, section through peripheral skeleton; f , specimen. g-i, Raspailia cervicornis Burton (specimens from the South Arabian coast) (type species of the nominal genus Raspailiopsis Burton): g, specimen BMNH 1936.3.4.521; h, specimen BMNH 1936.3.4.522; i, section through peripheral skeleton of latter specimen. j, Raspailia falcifera Topsent (schizotype MNHN LBIM DT 901 from the North Atlantic) (type species of the nominal genus Parasyringella Topsent), section through peripheral skeleton. k-1, Raspaxilla phakellina Topsent (schizotype MNHN LBIM DT 1614 from Antarctica) (type species of the nominal genus Raspaxilla Topsent): k, section through peripheral skeleton; I, acanthorhabdostyles echinating skeletal tracts. Scale line = 100 pm, scale = 30 mm. Australian Raspailiidae Diagnosis Arborescent, lobo-digitate to massive growth form, typically with cylindrical branches and basal holdfast. Surface even or rugose, often optically hispid. Choanosomal skeleton always with fibres and spicules in distinct tracts: axial skeleton typically condensed with widely spaced reticulate fibres cored by styles, but degree of compression varies considerably between species. Extra-axial skeleton typically plumo-reticulate, with extra-axial spicule tracts standiig perpendicular to axial skeleton, cored by large styles or oxeas which ascend to and poke through surface in uni- or paucispicular brushes, although this may be reduced to simple plumose structure. Peripheral spicule tracts may be multispicular or reduced to brushes of subectosomal spicules embedded in the subdermal region. Ectosome typically with specialised skeleton of small styles grouped into brushes standing perpendicular to surface, surrounding bases of protruding extra-axial megascleres, but these may be secondarily lost. Fibres usually echinated by acanthostyles or modified forms, sometimes secondarily lost. Structural megascleres styles, oxeas or anisoxeas, typically 3 sometimes 2 distinct size categories; echinating acanthostyles morphologically similar to microcionid sponges. Microscleres absent. Remarks The genus Raspailia was revised most recently by Vosmaer (1912, 1935) on the basis of exclusively European forms. Many new taxa have since been described, and that brief revision is now grossly inadequate in characterising the genus. Furthermore, Vosmaer's (1935) contribution must be accepted with caution. He synonymised about 20 nominal species, including Nardo's type species, under Raspailia hispida (Montagu), an act which very few authors subsequently accepted. Wilson (1921) made a better attempt at redefining the genus. From an examination of a number of species he arrived at a broad definition for Raspailia s.l., and simultaneously succeeded in conveying the enormous difficulty involved in defining this large and diverse group. There is obviously a high diversity of character states found in the genus, encompassing most characters. Growth form ranges from massive (e.g. R. desmoxyiformis; Fig. 26a), to cylindrical branching (e.g. R. compressa; Fig. 33a), or lobate (e.g. R. darwinensis; Fig. 24a-b). Skeletal structure varies from the typical condition, with a condensed axis and radial extra-axial fibres (e.g. R. pinnatifida; Fig. 13e,g), a reduced extra-axial skeleton (e.g. R. bifurcata; Fig. 219, a plumo-reticulate extra-axial skeleton (e.g. R. atropurpurea; Fig. 8d,f), or a reduced plumose extra-axis (e.g. R. stelliderma; Fig. 45e,h). Structural megascleres may be exclusively stylote (e.g. R. typica; Fig. 7a-c), stylote and oxeote together (e.g. R. paradoxa; Fig. 31) or exclusively oxeote (e.g. R. darwinensis; Fig. 23). Ectosomal structure varies from the typical condition, consisting of protruding extra-axial (subectosomal) megascleres surrounded by special dermal brushes (e.g. R. vestigijiera; Fig. 16g), extra-axial megascleres protruding through the surface but without any specialised ectosomal megascleres (e.g. R. gracilis; Fig. 10e,g), or with only special dermal spicules scattered over the surface (e.g. R. arbuscula; Figs 19d, 20h). The geometry and spination of echinating acanthostyles forms the basis for further subgeneric differentiation. Species of the nominotypical subgenus Raspailia may have typical microcionid-like acanthose subtylostyles (e.g. R. arbuscula; Fig. 20g), myxillid-like acanthostyles (eg. R. cacticutis; Fig. 22f) or thin vestigial acanthostyles (e.g. R. vestigijiera; Fig. 15c); those referred to the subgenus Raspaxilla have rhabdose acanthostyles with smooth bases (e.g. R. compressa; Fig. 33e); those placed in the new subgenus Hymeraphiopsis have acanthostyles with smooth and very swollen tylote bases (e.g. R. irregularis; Fig. 46i); and species that have secondarily lost their echinating megascleres are referred to the subgenus Syringella (e.g. R. nuda). The presence or absence of one or more of those characters has been used to differentiate numerous raspailiid genera at some time or another. For example, the absence of a special dermal skeleton, the presence of a plumose (non-anastomosing) extra-axial skeleton, the presence of rhabdose acanthostyles (Echinaxia, Raspaxilla), the absence of echinating megascleres (Raspailopsis and Syringella), or the atrophy of axial and extra-axial skeletal differentiation to a simply reticulate skeleton (Clathriodendron) may be of lesser importance than is currently accepted. All those characters must be carefully assessed. 1198 J. N. A. Hooper The major difficulty in attempting a revision of any group of Porifera is to clearly define apomorphies for particular genera. In the Raspailiidae most characters intergrade between one or more genera (transformation series): there does not appear to be any clear boundary between any character. The genus Raspailia illustrates this best, and the diagnosis above barely distinguishes between species that have reticulate extra-axial skeletons and lack ectosomal specialisation (e.g. R. wilkinsoni) and many species of Echinodictyum, or between those Raspailia that lack echinating megascleres (i.e. subgenus Syringella) and the genus Ceratopsion. Sirnilariy, the genus could be further subdivided on the basis of other important features (e.g. with or without ectosomal specialisation; with or without axial condensation; with acanthose rhabdostyles, acanthose tylostyles, or with conventional acanthostyles). However, there seems to be little concordance between all these and other morphological characters, which appear in all combinations throughout the many species. Nevertheless, in this revision the geometry and spination of acanthostyles is chosen above other characters to distinguish between genera and subgenera because it is considered that this feature is more prominent and relatively more consistent between species groups than are other characters. This does not necessarily enable some genera, such as Raspailia and Echinodictyum (both of which have a similar diversity of acanthostyle forms), to be easily differentiated. In this situation, therefore, the presence or absence of axial compression, and the differentiation of the axial and extra-axial skeletons become important in placing species in one genus or another. Echinodictyum is a relatively homogenous group (having only reticulate skeletons, and choanosomal oxeas), whereas there is a much greater diversity of characters in Raspailia. Skeletal characteristics, such as the presence or absence of axial compression, axial and extra-axial differentiation, and the geometry of the megascleres which construct these skeletons are relegated to secondary importance, unlike previous systematics for the family. Discussion of Nominal Genera Placed in Raspailia On the basis that Raspailia s.1. presents a diverse array of varying and discordant features, Dendy (1924: 355) and Bergquist (1970: 28, 30) suggested that Echinaxia Hallmann and Raspaxilla Topsent were junior synonyms of Raspailia. For reasons listed below, that synonymy is accepted here, and several other nominal genera are also merged here with Raspailia. (1) Echinaxia is used in the sense of Hallmann (1916a), which includes Raspailia folium (Thiele) and Raspailia horsuta Thiele (cf. Bergquist 1970). These species lack ectosomal specialisation and long extra-axial megascleres (Fig 34), and their retention in a separate genus, E c h i w i a , leaves them barely differentiated from Raspailia (s.1.). This decision is based on an evaluation of whether the absence of a special ectosomal skeleton, characteristic of classical Raspailia architecture, and a reduction in extra-axial skeletal development is of sufficient diagnostic importance at the generic level of classification. Bergquist (1970) showed from a study of Pacific species that ectosornal specialisation in Raspailia was highly variable, and its presence or absence did not correlate with similarities in axial and extra-axial architecture. Of the five species described or redescribed by her, only two [R. arbuscula (Lendenfeld) (described as R. agminata Hallmann) and R. compressa Bergquist] had characteristic dermal brushes surrounding the protruding choanosomal megascleres, one (R. topsenti Dendy) had vestigial ectosomal specialisation consisting of light brushes or tangential dermal megascleres, without protruding structural spicules and two species (R. inaequalis Dendy and R. Jlaccida Bergquist) lacked dermal specialisation entirely. In the present study, similar evidence is provided. Of the 26 species of Raspailia known to occur in Australian waters, 10 (i.e. R. compressa, R. darwinensis, R. nuda, R. paradoxa, R. phakellopsis, R. pinnatijida, R. reticulata, R. stelliderma, R. tenella and R. vestigifera) had well developed, classical ectosomal specialisation. By comparison, two species (R. australiensis and R. echinata) had rudimentary ectosomal specialisation, with very few ectosomal megascleres surrounding protruding extra-axial spicules); four species (R. atropurpurea, R. elegans, R. gracilis and R. irregularis) had ectosomal spicules but these spicules were scattered over the surface and not restricted to basal tufts surrounding Australian Raspailiidae 1199 protruding extra-axial megascleres; one species (R. arbuscula) had a variable ectosomal development; and nine species (R. bzfurcata, R. cacticutis, R. clathrata, R. desmoxyiformis, R. ffondula, R. keriontria, R. melanorhops, R. wardi and R. wilkinsoni) lacked any sort o f specialised ectosomal skeleton. Similarly, o f these 26 Australian Raspailia species, 15 did not have any well marked axial compression, with only reticulate or halichondroid axial skeletons instead, although these species did have differentiated axial and extra-axial components, and the latter consisted o f reticulate, plumo-reticulate, plumose or halichondroid structures (e.g. R. arbuscula, R. cacticutis, R. darwinensis, R. wardi). By comparison, 11 species had well developed axial compression, with well developed axial and extra-axial differentiation (e.g. R. atropurpurea, R. clathrata, R. pinnatifida, R. vestigifera). (2) Topsent (1913: 618) noted that Raspaxilla was like Raspailia but had acanthose rhabdostyles echinating fibres instead o f conventional acanthostyles, and also possessed unusually long and thin ectosomal styles (Fig. 7k-I). Several Indo-Pacific Raspailia species that have smooth rhabdose bases on echinating acanthostyles, and could be appropriately referred to Raspaxilla (or Echinaxia or Axinectya) :R. topsenti Dendy (1924: 354) (BMNH 1923.10.1.135), R. flaccida Bergquist (1970: 27), R. inaequalis Dendy (1924: 3 5 3 , R. folium Thiele (1898: 60), R. frondula (Whitelegge), R. horsuta Thiele (1898: 59), R. reticulata, R. wardi and R. compressa. Since these rhabdose modifications can range from very slight to very well developed, and because this feature also occurs in other genera (e.g. Aulospongus), it is not given primary importance at the generic level, although it does appear to serve as a convenient subgeneric classification. Rhabdose modifications to megascleres appear to have been acquired by sponges independently on at least two occasions: in the families Raspailiidae and Rhabderemiidae (Hooper 1990b). (3) Hallmann (1917) suggested that Axinectya was similar to Echinaxia in having lightly (or partially) spined rhabdose echinating megascleres, but other features are quite different. Axinectya differs from Raspailia (s.s.) in lacking ectosomal specialisation, in possessing rhabdose bases on smooth choanosomal megascleres, and in having echinating rhabdostyles localised at the junction o f axial and extra-axial skeletons. The genus appears to have its closest affinities with Raspailia, particularly in the possession o f huge subectosomal extra-axial spicules, and like Echinaxia and Raspaxilla it is most appropriately placed there. Axinectya is only poorly known from Ridley & Dendy's (1886) holotype (Fig. 7d-$3. (4)Burton (1959)created Raspailopsis for raspailiid species lacking echinating megascleres, with an axially condensed skeleton o f spongin fibres cored by choanosomal styles, a plumose extra-axial skeleton o f subectosomal styles projecting through the surface, and a plumose ectosomal skeleton consisting o f brushes of oxeas or anisoxeas surrounding protruding extra-axial megascleres. It is curious that the genus was established, since Syringella was already in use by earlier authors for much the same purpose. In its original conception, Raspailopsis supposedly differed from Raspailia only in the absence o f echinating megascleres. However, re-examination o f some o f Burton's (1959) additional material o f the type species R. cewicornis [BMNH 1936.3.4.521 and BMNH 1936.3.4.522 from the South Arabian coast (Fig. 7g-i)], showed that echinating megascleres are in fact present on the peripheral fibres o f the axial skeleton, and Raspailopsis is undoubtedly a synonym o f Raspailia. The taxon Syringella is also used here as a convenient subgenus, to delineate a group of Raspailia lacking echinating megascleres. (5) Clathriodendron is superficially quite different from Raspailia, in that it lacks axial condensation, any marked axial and extra-axial skeletal differentiation, and often the specialised ectosomal skeleton typical o f other raspailiids (Figs 20-30). It shows closest similarities to Echinodictyum, and should also be compared with the microcionid genera Echinochalina Thiele and Echinoclathria Carter. Some species o f Clathriodendron (e.g. R. melanorhops; Fig. 30c-e) are also known to incorporate detritus into their skeletons, and in that respect they are homologous to the relationship demonstrated in the Microcionidae between Clathria and Clathriopsamma (Hooper, unpublished data). Hentschel (1911) suggested that Clathriodendron was synonymous with Raspailia, due to the existence o f species such as R. paradoxa, that have a reticulate choanosomal skeleton. A similar skeletal structure is now known to occur in many other Raspailia species (e.g. R. darwinensis; 1200 J. N. A. Hooper Fig. 24c-e). In contrast, Hallmann (1912) rejected Hentschel's arguments and considered that C . arbuscula was sufficiently different from Raspailia to maintain the two taxa, and this decision was followed most recently by Wiedenmayer (1989). For reasons described below (see remarks for R. arbuscula), Hentschel's (1911) decision is accepted here, and Clathriodendron is considered to be a reduced, reticulate Raspailia. However, the genus is conveniently used at the subgeneric level, although this division is obviously arbitrary, and the possession of a reticulate architecture over a differentiated axial and extra-axial skeleton cuts across a primary classification based on echinating megasclere geometry. (6) The action of de Laubenfels (1936) to create a separate genus (Valedictyum) for R. vestigifera is puzzling. In all respects the species is typical of Raspailia (Figs 15-16). His comparison with Echinodictyum is entirely misleading, and his observations that the type species is characterised by a plumose architecture are not entirely correct. That feature is visible only in the peripheral extra-axial skeleton, whereas the main skeleton is a very well developed condensed reticulation of heavy spongin fibres (Figs 15e, 16e-fl. In any case, Echinodictyum is completely reticulate and its supposed relationship with Valedictyum is erroneous. (7) Parasyringella is referred here to Raspailiidae on the basis that it has classical raspailiid ectosomal specialisation, and not for the reasons that de Laubenfels (1936: 102) included the genus in this family; namely, on the spurious assumption that Topsent had defined the genus with spined spicules. Topsent (1928: 287) merely refers to the hispidating (extra-axial) megascleres that protrude through the surface being surrounded at their bases by 'de spicules monactinaux 6pineux' or brushes of fine 'spines' (i.e. ectosomal auxiliary subtylostyles). The type species of Parasyringella, R. (Syringella) falcifera Topsent, is a Raspailia that lacks echinating megascleres and has extra-axial megascleres bearing recurved points (Fig. 7j). Key to the Subgenera of Raspailia With echinating acanthostyles ..................................................................... 2 Without echinating acanthostyles ....................................................... Syringella Choanosomal skeleton consists of an axially condensed reticulation of spongin fibres and/or spicule tracts, with at least some degree of differentiation between axial and extra-axial skeletons ......................................................................................... 3 Choanosomal skeleton consists of a reticulation of spongin fibres and/or spicule tracts, without any trace of axial compression, and with reduced differentiation of the axial and extra-axial skeletons ........................................................... Clathriodendron Echinating spicules are club-shaped acanthostyles, with subtylote bases and straight shafts ........................................................................................... Raspailia Echinating spicules are acanthose rhabdostyles ..................................... Raspaxilla Echinating spicules have smooth, swollen, tylote bases ...................... Hymeraphiopsis Subgenus Raspailia Nardo Raspailia (Raspailia) atropurpurea (Carter) (Fig. 8) Axinella atropurpurea Carter, 1885: 359. Raspailia atropurpurea.-Carter, 1886a: 289; Dendy, 1896: 47; Whitelegge, 1901: 92; Pick, 1905: 23. Material Examined Holotype. BMNH 1886.12.15.1 (AM G2792 slide; NMV sponge archives 36/28): Port Phillip Heads, Vic., 38 m depth, coll. J.B. Wilson. Paratype. BMNH 1886.12.15.36: same locality. Substrate and Depth Range Substrate unknown, depth recorded as 38 m. Australian Raspailiidae Fig. 8. Raspailia atropurpurea (Carter) (holotype B M N H 1886.12.15.1): a, subectosomal extra-axial subtylostyle and style; b, ectosomal auxiliary style; c, echinating acanthostyle; d, section through peripheral skeleton; e, known Australian distribution; f , skeletal structure; g, holotype (scale = 30 mm). J. N. A. Hooper Geographical Distribution Port Phillip Heads, Vic., and Port Stephens, N.S.W. (Fig. 8e). Description Shape. Stipitate, with basal holdfast, short thick stalk (10-15 mm long, 15-20 mm diameter), with short compressed di- and polychotomously divided branches, some fused, branching in more than 1 plane: with bulbous rounded bifurcated or single ends (21-36 mm long, 4-9 mm diameter). Specimens approximately 48-65 mm high, 65-90 mm broad. Colour. In life and preserved state coloration dark purple. Oscula. Small pores up to 2 mm diameter occur irregularly over surface. Texture and surface characteristics. Texture compressible, elastic, with fairly rigid axial skeleton, and surface even, unomamented, relatively fleshy. Ectosome and subectosome. Ectosome only slightly hispid where tips of occasional ectosomal spicule brushes protrude through surface, but this is probably an artifact of preservation. Most features of ectosomal skeleton partially obscured by very heavy purplepigmented collagenous spongin. Ectosomal skeleton with close-set brushes of ectosomal styles-anisoxeas forming continuous palisade on surface. Surface spicule brushes arise from ends of extra-axial skeletal tracts, which extend out from axis in radial arrangement, cored by 3-8 long subectosomal styles fully enclosed in heavy spongin fibres. Extra-axial fibres occasionally anastomose with adjacent fibres, some divide towards periphery. Choanosome. Choanosomal axial skeleton condensed, with well differentiated axial and extra-axial components. Axial skeleton consists of tight-meshed reticulation of very heavy spongin fibres, largest running longitudinally along branches. Fibres sparsely cored by only bases of subectosomal megascleres and sparsely echinated by acanthostyles. Entire mesohyl matrix invested with very heavy granular spongin. Megascleres. Choanosomal spicules of axial skeleton absent. Subectosomal extra-axial megascleres long, thick styles, with evenly rounded or slightly subtylote bases and rounded tips bolotype 986-(1228 5)-1364 x 18426.6)-35 pm]. Ectosomal auxiliary megascleres short, thin, usually slightly curved centrally, with sharp points and fusiform bases. Geometry ranges from styloid to anisoxeote with asymmetrical ends Bolotype 106-(138 .4)-159 x 2 5-(4.3)-6 pm]. Echinating acanthostyles small, cylindrical, evenly spined, with slightly subtylote or rounded bases and rounded tips Bolotype 65472 -8)-97 x 6 4 8 4)-11 pm]. Microscleres absent. - - Remarks According to Carter (1885) there were three specimens for which this species was originally erected, but only two have been found. Dendy (1896) also lists only two specimens, but it is possible that a dry specimen exists somewhere in the BMNH collections. Similarly, Whitelegge's (1901) material from Port Stephens has not yet been rediscovered, but probably exists in the AM general collections. The description above is based mainly on the holotype ('chief type' noted on specimen label), although the paratype is very similar in dimensions and features. This species is remarkable for its persistent heavy purple coloration, globular lobate branches with bifurcated tips, even fleshy surface, heavy collagenous ectosome through which no extra-axial megascleres protrude and ectosomal spicule brushes only barely protrude, the heavy fibrous skeleton, absence of choanosomal axial megascleres, sparsely dispersed echinating spicules and evenly distributed granular spination on echinating spicules. These features easily differentiate this species from other raspailiids. The ectosomal skeleton, with brushes of ectosomal megascleres perched on the ends of extra-axial spicule tracts below the surface, rather than surrounding the bases of protruding extra-axial megascleres, is also unusual in Raspailia. Australian Raspailiidae Raspailia (Raspailia) echinata Wtelegge 0%. 9) Raspailia echinata Whitelegge, 1907: 514-515, pl. 46, fig. 37. Material Examined Lectotype (here designated). AM (343.54: Off Botany Bay, N.S.W., 110-132 m depth, date of collection unknown (FRV 'Thetis') (NMV sponge archives 1/32). Substrate and Depth Range Mud substrate, 110-132 m depth. Geographical Distribution Deeper coastal waters of N.S.W. (Fig. 9f). Description Shape. Stipitate, arborescent, digitate sponge, branching in one plane (124 mm long, 90 mm maximum breadth), with basal holdfast and long slender stalk (39 mm long, 4-6 mm diameter), short branches which bifurcate several times (22-59 mm long, 4-6 mm diameter), branches tapering towards their tips. Colour. Dark grey in dry state. Oscula. Not observed. Texture and sueace characteristics. Texture hard and incompressible in dry state. Surface contracted and minutely honeycombed when dry, with numerous raised irregular ridges and prominent microconules: slightly optically hispid. Ectosome and subectosome. Ectosome membraneous, with only rudiments of specialised skeleton protruding from tops of surface microconules. Sparsely scattered long subectosomal spicules and points of numerous echinating acanthostyles may also poke through surface in vicinity of microconules. Microconules prominent in subectosomal region, consisting of small digitate columns, 200-700 pm high and 250-400 pm apart. Microconules composed of central core of one to several sub&tosomal styles standing perpendicular to surface. Styles may protrude 100-200 pm through ectosome, although dry lectotype has most broken off at surface level. Extra-axial styles originate from peripheral fibres of axial skeleton, with bases embedded in spongin fibres. ~ e n s eplumose brushes of acanthostyles surround basal 113 of extra-axial spicules, and extend outwards perpendicular to the surface. Ectosomal and subectosomal spongin heavy, slightly granular and pigmented dark brown. Choanosome. Choanosomal axial skeleton not compressed, regularly reticulate, subrenieroid, with heavy spongin fibres 25-90 pm diameter, cored by uni- or paucispicular tracts of spicules. Spicules more-or-less divided into primary radial and secondary transverse elements, but with no definite pattern to division. Primary fibres usually contain large subectosomal megascleres, ascending to surface; secondary fibres contain smaller axial spicules. Echinating acanthostyles moderately common in axial region, but far fewer in axial skeleton than at periphery. Fibre meshes oval to rectangular, widely spaced, 130-410 pm diameter, containing abundant but lightly pigmented granular spongin. Choanocyte chambers oval, 90-120 pm diameter. Megascleres. Choanosomal axial spicules short, slender or thick styles, usually slightly curved at centre, with rounded or slightly subtylote bases, tapering to hastate or stepped points [275-(330.2)-395 x 13419.4)-26 pm]. Subectosomal megascleres of extra-axial skeleton, and also occurring in peripheral fibres of axial skeleton long, thick, robust styles, straight or slightly curved at centre, with rounded or slightly subtylote bases, and with hastate, abruptly pointed or stepped tips [548-(745 -6)-1141x21-(35 -8)-58 pm]. Ectosomal megascleres short or long, slender, sinuous or simply curved at centre, with evenly rounded or hastate bases (i.e. anisoxeote) and long fusiform points [217-(308 -6)368 x 1 -542.9)-4 pm]. J. N. A. Hooper Fig. 9. Raspailia echinata Whitelegge (lectotype AM G43.54): a, choanosomal axial styles; b, subectosomal extra-axial subtylostyle; c, ectosomal styles/anisoxeas; d, echinating acanthostyle; e, section through peripheral skeleton; f , known Australian distibution; g, skeletal structure; h, echinating acanthostyle; h, lectotype (scale = 30 mm). Australian Raspailiidae 1205 Echinating acanthostyles short, very thick, cylindrical, with evenly rounded bases, abruptly pointed tips, covered with exceptionally large spatulate and recurved spines [107-(118 -8)-138 X 14-(18 2)-23 pm]. Microscleres absent. Remarks Contrary to Whitelegge's (1907) description, the spongin fibres in the axial skeleton are well developed and contain abundant spongin; there is a definite division of axial and extra-axial megascleres; ectosomal megascleres are present, although ectosomal structure is vestigial; and measurements cited in the description, including spicule dimensions, also differ substantially from those cited above. This species is unusual within the genus in its extra-axial skeletal structure and spiculation. Axial and extra-axial spicules are very thick and robust, and ectosomal spicules are usually sinuous. The geometry and spination of acanthostyles are the most distinctive features of this species. Spines are strongly recurved and spatulate, similar to those discovered in several species of Echinodictyum, described below. Spatulate spines are also found in Raspailia cacticutis, R. gracilis and R. tenella, but these are much more poorly developed than those of R. echinata. The skeletal architecture is also diagnostic for this species, whereby the reticulate axial component occupies the major portion of branch diameter, and the extra-axial skeleton is perched on the outer edge of this core, forming the surface microconules. Similarly, the vestigial ectosomal skeleton is perched on the top of these rnicroconules. Raspailia (Raspailia) gracilis (Lendenfeld) (Fig. 10) Axinella hispida (Montagu) var. gracilis Lendenfeld, 1888: 235.-Whitelegge, 1889: 187. Raspailia graci1is.-Hallmann, 1914a: 268; Hallmann, 1914b: 417; Bergquist, 1970: 27. Material Examined Holotype. AM G9083: Port Jackson, N.S.W., depth and date of collection unknown (schizotypes BMNH 1887.4.27.40, AM G3791 slides, NMV sponge archives 19/19-21). Substrate and Depth Range Rock substrate, depth of collection unknown. Geographical Distribution Port Jackson, N.S.W. pig. 10f). Description Shape. Stipitate, arborescent sponge branching in more than one plane (75 mm high, 64 mm maximum breadth), with short subcylindrical stalk (7 mm long, 4-5 mm in diameter), and series of slightly flattened branches (11-35 mm long, 4-6 mm diameter) which bifurcate several times. Branches spatulate at apex, with rounded or sometimes pointed margins. Colour. Grey-brown with purple tinge in ethanol. Oscula. Not observed. Texture and surface characteristics. Texture tough but compressible. Surface even, slightly microconulose, prominently hispid. Ectosome and subectosome. Ectosomal skeleton rudimentary, with sparse brushes of ectosomal styles or anisoxeas usually parallel to surface, lying tangentially to protruding extra-axial megascleres. Extra-axial region with long slender subectosomai styles, singly or in bundles of up to 10, standing perpendicular to surface, with bases embedded in fibres of axial skeleton. Extra-axial styles extend partially through surface, protruding from tops of low surface conules. Conules produced by peripheral spongin fibres and associated mesohyl of choanosomal skeleton. Peripheral fibres contain bases of protruding extra-axial styles, ascending to surface for short distances. Mesohyl matrix in peripheral skeleton contains abundant slightly granular spongin. J. N. A. Hooper Fig. 10. Raspailia gracilis (Lendenfeld) (holotype AM G9083): a, axial styles; b, ectosomal styles/anisoxeas; c, subectosomal styles; d, echinating acanthostyle; e, section through peripheral skeleton;f , known Australian distribution; g, skeletal structure; h, echinating acanthostyle; i, holotype (scale = 30 mm). Choanosome. Choanosomal skeleton not axially compressed, consisting of small-meshed irregular reticulation of heavy spongin fibres, 2 0 6 0 p m diameter. Fibre anastomoses show superficial similarities with Spongiidae (Dictyoceratida). Fibres divisible into primary elements running mainly longitudinally through branches, cored with paucispicular tracts of thick choanosomal styles, and interconnected by aspicular secondary fibres. In axial region fibres more compact forming lattice-like reticulation, whereas near periphery fibres more radial and interconnected by single spicules running transversely. Fibre meshes 30-70 p m diameter in axis, 90-160 pm near periphery. Echinating acanthostyles sparsely distributed Australian Raspailiidae 1207 in axis, more abundant in peripheral skeleton, found mainly or exclusively on exterior surface of fibres. Mesohyl contains abundant light brown spongin and oval choanocyte chambers 30-70 pm diameter. Megascleres. Choanosomal spicules in axial skeleton long thick styles, occasionally anisoxeas, usually curved towards basal end, occasionally sinuous, with evenly rounded or very slightly subtylote bases, tapering to sharply pointed tips [465-(751 -1)-1580 X 6(12.2)-17 pm]. Subectosomal extra-axial spicules geometrically similar to axial spicules but more slender, often sinuous or with prominent central curvature, with evenly rounded bases, subtylote, or subterminal swellings near base, tapering to long fusiform points l6624974.6)-1246 X 4(7 -6)-12 pm]. Ectosomal spicules oxeas, anisoxeas or less commonly styles, slightly curved or straight, very thin, occasionally raphidifom, with fusiform points, hastate or evenly rounded bases [223-(322 6)-402 x 1 4 3 3-4 pm]. Echinating acanthostyles small, cylindrical, straight, with slightly subtylote bases and hastate points, with evenly distributed and well developed recurved spines; spines spatulate in shape [58-(7 1 -3)-84 x 4 4 6 - 4)-9 pm]. Microscleres absent. - - Remarks This species has been comprehensively redescribed by Hallmann (1914b), and the description presented above is provided mainly for comparative purposes. One of the unusual features of the species is the presence of a vestigial ectosomal skeleton, consisting of scattered bundles of ectosomal styles or anisoxeas that are directed mostly parallel to the surface, unlike most Raspailia. This ectosomal skeleton is never disposed in spicule brushes around bases of extra-axial spicules, as is typical of the family. Other unusual characters of the species include the tendency for most of the spicules seen in cross sections to run longitudinally through branches, the exceptional length and geometric similarities between the axial and extra-axial spicules, and the preponderance of echinating acanthostyles on peripheral fibres. This species is most closely related to R. tenella, and the two species are contrasted further in the description of that species given below. Raspailiu (Raspailiu) phakellopsis, sp. nov. (Figs 11, 12, 109b; Table 1) Material Examined Holotype. NTM Z1950: Stephen's Rock, Weed Reef, Darwin, N.T., 12" 29.2'S., 130" 47.1 'E., 12 m depth, 27.iv.1984, coll. J.N.A. Hooper, SCUBA (sm WR1). Paratypes. NTM 2610: Cootamundra Shoals, N. of Melville I., Arafura Sea, N.T., 10" 50-2'S., 129' 13-15'E., 36 m depth, 7.v.1982, coll. R. Lockyer (stn Don. 7, SCUBA). NTM 22158: 'Bommies', N. edge of Weed Reef, Darwin, N.T., 12" 29.2'S., 130" 37.6'E., 8-10 m depth, 5.x.1984, coll. J.N.A. Hooper, SCUBA (stn WR5). Other material. Scott Reef, W.A.: PIBOC unreg.: Near Scott Reef, 16"41.4's.. 121" 09.6' E., 51-54 m depth, 4.xi.1990 (stn 26). PIBOC unreg.: 16" 36.3/S., 121' 05 .glE., 40-60 m depth, 7.xi.1990 (stn 39). PIBOC unreg.: 16" 32 -2/S., 121" 10-9'E., 43-44 m depth, 4.xi.1990, coll. V. Krasochin, U.S.S.R. RV 'Akademik Oparin', dredge (stn 28). Substrate and Depth Range Rock and dead coral reef, 8-60 m depth range. Geographical Distribution Darwin region, N.T., and Northwest Shelf, W.A. (Fig. llf). Description Shape. Flabellifom sponges (72-120 mm high, 90-150 mm maximum breadth), with basal attachment, very short flattened or cylindrical stalk (10-16 rnrn high, 10-14mm J. N. A. Hooper Fig. 11. Raspailia phakellopsis, sp. nov. (paratype NTM 22158): a, choanosomal axial style; b, ectosomal auxiliary oxea; c, subectosomal extra-axial style; d, echinating acanthostyles; e, section through peripheral skeleton; f, known Australian distribution. diameter), from which arise several flat frondose lobate branches (38-110 mm long, 28-85 mm broad, 1-2.5 mm thick), branching in more than one plane, with rounded digitate and convoluted margins. Colour. Dark red alive (Munsell 5R 4/10) (Fig. 109b), beige-brown in ethanol. Oscula. Not observed. Texture and surface characteristics. Texture h, slightly compressible, harsh to touch, with flexible branches. Surface even, slightly optically hispid. Two specimens have zooanthids dispersed over branches. Ectosome and subectosome. Ectosome profusely hispid, produced by both subectosomal and choanosomal spicules. Longest hispidating spicules are subectosomal extra-axial styles, protruding up to 1 8 mm from surface but relatively sparsely distributed, surrounded at their points of insertion into ectosome by light brushes of ectosomal megascleres usually forming stellate tufts. Numerous choanosomal styles also stand perpendicular to axis, protruding through ectosome, forming a radial relatively dense palisade of spicules extending only 3 0 0 4 0 0 p m from the ectosome, much like species of Phakellia. Ectosome has light brown Australian Raspailiidae Fig. 12. Raspailia phakellopsis, sp. nov.: a, holotype (NTM Z1950) side view; b, holotype top view; c, echinating acanthostyle (scale = 50 pm); d, SEM of skeletal structure, showing the closely reticulate axial skeleton and radial extra-axial skeleton; e, SEM of junction between axial and extra-axial skeletons, showing the radial distribution of echinating megascleres; f , g , geometry and spination of acanthostyles. layer of spongin, significantly darker at surface than in choanosomal mesohyl. Zooanthids perched over ectosome rather than embedded in it, appearing to be attached to sponge only by sponge's protruding choanosomal megascleres. Choanosome. Choanosomal skeletal tracts run longitudinally through branches, without obvious axial condensation, but with well developed axial and extra-axial differentiation. Choanosomal axial skeleton reticulate but relatively disorganised, with spicules orientated longitudinally or in slightly plumose tracts ascending to surface. Spongin fibres extremely light, and spicule tracts are merely aggregated and enveloped by light granular mesohyl 1210 J. N. A. Hooper spongin. Echinating acanthostyles more predominant in axial region of choanosomal skeleton than at periphery. Megascleres (refer to Table 1 for measurements). Choanosomal axial megascleres long, relatively thin, invariably stylote, slightly curved towards centre, with fusiform points and rounded non-tylote bases, without any apparent difference between styles which form axial (longitudinal) skeleton and those standing perpendicular to them, forming radial extra-axial skeleton. Subectosomal extra-axial megascleres very long, relatively thin styles, with rounded non-tylote bases, sharply fusiform points, straight or only slightly curved, very rarely sinuous. Ectosomal auxiliary megascleres extremely thin, rhaphidiform, straight or very slightly curved centrally, with fusiform points, varying from oxeote or anisoxeote to true styles. Echinating acanthostyles relatively long, thin, straight or very slightly curved near basal end, with prominent subtylote bases often bearing recurved hooks, and sparsely dispersed, but relatively large recurved thorny spines along midsection of shaft only. 'Neck' of spicule (proximal to base) and point typically aspinose, although sometimes subterminal spines occur near point, producing a pseudoclavulate morphology. Microscleres absent, although raphidifom ectosomal oxeasfstyles may be scattered throughout the mesohyl near periphery. Table 1. Comparisons in spicule measurements between specimens of Raspailia phakellopsis sp. nov. Measurements are given in micrometres, presented as ranges (and means) of lengthxwidth, and represent 25 spicules measurements per specimen for each spicule category Material Choanosomal styles Subectosomal styles NTM Z1950 311-465~ 9 - 1 8 (392.7~13.2) Holotype 820-1835 x 8-17 (1349.6~12.4) 173-302 x 0.5-3 (231 . 0 ~ 1 - 5 ) 125-156 ~ 5 - 9 (133-2~7.1) NTM 2650 412-502x8-18 (452.7~12.7) 419-554 ~ 9 - 1 8 (473-5x 13.8) Paratypes 1045-1715x9-13 ( 1 3 8 1 . 6 ~11.O) 1082-1694~9-15 (1418-5~11.5) 211-272~0.8-2 ( 2 4 3 . 6 ~1.3) 205-268 xO-5-2 ( 2 3 6 . 6 ~1.2) 109-143x47 (126.2~5.8) 122-146x5-7 (132~0~6.2) NTM 22158 Ectosomal oxeasl styles Echiiating acanthostyles Remarks This species is different from all other Raspailia in having a longitudinal axial and perpendicular extra-axial skeleton (in addition to a subectosomal extra-axial skeleton typical of raspailiids). In its Phakellia-like growth form, pigmentation and surface features, the species is superficially similar to Antho (Isopenectya) chartacea (Whitelegge) (family Microcionidae) (Hooper, unpublished data) and Trikentrion flabelliforme (see below). Spicule geometry and spicule dimensions of this species are also quite different from all other raspailiids of this region. Etymology This species is named for its superficial similarities to Phakellia species (family Axinellidae), in both its thinly flabellate growth form and presence of a longitudinal axial skeleton and a perpendicular extra-axial skeleton. Australian Raspailiidae Raspailiu (Raspailiu) pinnatiflda (Carter) (Fig. 13; Table 2) Dictyocylindrus pinnatifidus Carter, 1885: 353. Axinella setacea Carter, 1885: 359. Raspailia pinnatifida.-Dendy, 1896: 4647; Pick, 1905: 22. Material Examined Holotype. BMNH 1886.12.15.60: Port Phillip Heads, Vic., 10 m depth, coll. J.B. Wilson (schizotypes-BMNH 1887.4.27.97, AM G2803, NMV sponge archives 35/27). Holotype of A. setacea. BMNH 1886.12.15.61: Port Phillip Heads, Vic., 14 m depth, coll. J.B. Wilson. Substrate and Depth Range Depth range 10-14 m, substrate unknown. Geographical Distribution Port Phillip Bay, Vic. (Fig. 13f). Description Shape. Digitate branching sponges (50-280 mrn long), stipitate holdfast, short cylindrical stalk (15-40 mm long, up to 12 mm diameter), bifurcate or sometimes polychotomous branching pattern, cylindrical branches variable in length (10-80 mm long, 5-9 mm diameter), tapering towards apex, and occasionally dividing into two. Colour. Live coloration dark brown (Carter 1885), grey-brown in ethanol. Oscula. Not observed. Texture and sulface characteristics. Consistency of branches soft, flexible, surface optically microconulose, producing velvety 'rat's-tail' appearance noted by Dendy (1896). Basal stalk woody, barely flexible, sparsely microconulose. Ectosome and subectosome. Surface microscopically very hispid, with terminal fibre endings forming surface conules, brushes of subectosomal styles arising from these fibres. Discrete brushes of smaller ectosomal styles surrounding ends of terminal fibres and other places on ectosome as well. Subectosomal region extensive, occupying up to 213 sponge diameter. Choanosome. Axial skeleton moderately compressed, composed of an open reticulation of heavy spongin fibres cored by pauci- or multispicular tracts of subectosomal styles, thicker than those in the peripheral regions. Plumose extra-axial fibres arise from axis, repeatedly dividing and loosely anastomosing, cored by multispicular tracts of long subectosomal styles. Echinating acanthostyles rare. Mesohyl matrix moderately heavy, granular, with numerous tracts of subectosomal megascleres dispersed in plumose arrangement towards surface. Megascleres (refer to Table 2 for measurements). No special category of choanosomal spicule present, but subectosomal styles may core axial fibres. Subectosomal styles in axis long, relatively thick, usually slightly curved towards basal end, evenly rounded base, occasionally subtylote, tapering to sharp points or sometimes rounded at apex. Subectosomal styles in extra-axial region long, setaceous, relatively thin, straight or slightly curved, with rounded or slightly subtylote bases, always tapering to sharp points at their tips; no justification found to differentiate 2 sorts of auxiliary styles in axial and extra-axial regions, being geometrically identical. Ectosomal auxiliary styles relatively short, thin, occasionally raphidiform, thickest in middle, tapering to whispy sharp point at apex, tapering slightly towards base. Echinating acanthostyles rare, straight, tapering to sharp points, slightly subtylote, with evenly distributed spines. Microscleres absent. Remarks Raspailia pinnatijida is easily differentiated from other species by the 'rat's-tail' appearance of branches (Dendy 1896: 47), the scarcity of (Crella-like) echinating acanthostyles and the J. N. A. Hooper Fig. 13. Raspailia pinnatifida (Carter) (holotype BMNH 1886.12.15.60): a, b, extra-axial styles; c, ectosomal styles/anisoxeas; d, echinating acanthostyles; e, section through peripheral skeleton;f , known Australian distribution; g, peripheral skeleton; h, holotype (scale = 30 mm); i, echinating acanthostyle. Australian Raspailiidae 1213 Table 2. Comparisons in spicule measurements between specimens of Raspailia pinnati@fa (Carter) Measurements are given in micrometres, presented as ranges (and means) of lengthxwidth, and represent 25 spicules measurements per specimen for each spicule category Material Choanosomal megascleres Subectosomal styles Ectosomal styles BMNH 1886.12.15.60 Absent Holotype 454-1305x6-16 (802.0~9.5) 292-396 x 1-3.5 (33104~2.3) BMNH 1886.12.15.61 Absent Specimen 503-1010x8-17 (726.8~12.5) 186-322x 1-3 (284.4~2.1) Echinating acanthostyles 66-86 x 3-5 (69~7~3.7) not seen absence of true choanosomal megascleres in the axial skeleton. In the latter two features, this species is similar to R. arbuscula. Dendy (1896) gives two other possible synonyms for this species [Axinella chalinoides var. glutinosa Carter (1885: 359) and its replacement name A. cladoJagellata Carter (1886~: 464)], but relevant material (viz. BMNH 1886.12.15.407) was not located for re-examination. That synonymy is therefore unconfirmed. Raspailia (Raspailia) tenella (Lendenfeld) (Fig. 14) Axinella hispida (Montagu) var. tenella Lendenfeld, 1888: 235.-Whitelegge 1889: 187. Raspailia tenella.-Hallmann 1914a: 268; 1914b: 421; Bergquist 1970: 27. Material Examined Lectotype (here designated). AM G9074 (larger of two specimens, with fewer branches): Port Jackson, N.S.W., depth and date of collection unknown (NMV sponge archives 29/22-23). Paralectotype. Smaller of two specimens, with more prolific branching, same locality. Substrate and Depth Range Unknown. Geographical Distribution Known only from type locality, Port Jackson, N.S.W. (Fig. 14f). Description Shape. Stipitate, branching sponge (52-60 mm high, 3 8 4 2 mm maximum breadth), with long, thin, cylindrical slightly compressed stalk (19-23 mm long, 1-2 mm diameter), without holdfast attached, with slightly flattened bifurcating branches (10-28 mm long, 1 -2-4 mm broad, 1-2 rnm thick), branching mostly in one plane, branches widen towards apex, with rounded or chiselled margins. Colour. Grey-beige in ethanol. Oscula. Not observed. Texture and surface characteristics. Texture firm, elastic, incompressible. Surface even, extremely hispid. Ectosome and subectosome. Ectosomal skeleton with specialised spiculation in form of long thin ectosomal styles or oxeas. Ectosomal spicule brushes grouped around bases of protruding extra-axial megascleres, also found in brushes between extra-axial spicules, standing perpendicular to surface. Ectosomal skeleton moderately dense, spicule brushes may protrude from surface for 100-350 pm. Like R. gracilis extra-axial skeleton of R. tenella perched on edges of branches, forming radial peripheral skeleton, composed of one or more long subectosomal styles, but unlike R. gracilis these are not associated with prominent microconules. Subectosomal extra-axial styles protrude through surface for up to 900 pm. Spongin in peripheral skeleton abundant, slightly granular, pigmented light to dark brown. Subectosomal region more-or-less free of any choanosomal, ectosomal and J. N. A. Hooper Fig. 14. Raspailia tenella (Lendenfeld) (lectotype AM G9074): a, choanosomal axial style; b, subectosomal extra-axial style; c, ectosomal auxiliary styles/anisoxeas; d, echinating acanthostyle; e, section through peripheral skeleton; f , known Australian distribution; g, lectotype (scale = 30 mm); h, echinating acanthostyle; i, skeletal structure. Australian Raspailiidae 1215 echinating spicules, containing only bases of long extra-axial styles which are embedded in exterior axial fibres. This region extends from peripheral fibres of axial skeleton to ectosome, up to 300 pm thick. Choanosome. Choanosomal skeleton similar to R. gracilis, without marked axial compression, but with compact reticulation of heavy spongin fibres in axis, cored by choanosomal styles directed both longitudinally through branches and radially towards edges of skeleton. Branch diameter fairly thin, slightly flattened, spicules relatively large so that lattice-like skeletal reticulation mostly obscured by protruding axial and echinating spicules. Echinating acanthostyles large and feature prominently in skeleton, relatively evenly dispersed throughout axial skeleton. No concentration of echinating megascleres in extra-axial region, as in R. gracilis. Fibre anastomoses form irregular oval or rectangular meshes, 60-130 pm diameter. Choanocyte chambers oval, up to 50 pm diameter. Mesohyl matrix in axial skeleton contains abundant quantities of light brown pigmented, slightly granular spongin. Megascleres. Choanosomal axial megascleres moderately long styles, slender, usually slightly curved near basal end, with evenly rounded or very slightly subtylote bases, tapering to sharp fusiform points at apex [392-(481- 6)-594 x 9 4 11 -4)-15 pm]. Subectosomal extra-axial megascleres geometrically similar to choanosomal styles, but much longer and thicker [754-(1480 8)-2015 x 16-(18 .0)-21 pm]. Ectosomal auxiliary spicules styles, anisoxeas, or less commonly oxeas, varying from raphidiform to moderately robust, straight or slightly curved, evenly rounded bases or tapering hastate bases and sharp fusiform points [261-(3 15 2 ) 4 1 2 x 1 . 5 - ( 2 . 8 ) 4 pm]. Echinating acanthostyles conical, with evenly rounded or slightly subtylote bases, hastate points, with spines scattered evenly over surface. Spines well developed, recuwed, spatulate, with sharp edges [66-(74.1)-85 x 6-(8 4)-11 pm]. Microscleres absent. - - Remarks This species has been comprehensively described by Hallmann (1914a, 1914b), and the lectotype is redescribed above for comparative purposes. This description also concentrates on differentiating R. tenella from R. gracilis, since these two species appear to be very closely related. In fact Lendenfeld (1888) first described these two species as varieties of one taxon, Axinella hispida. They appear to be identical in their external appearance and gross skeletal structure, and acanthostyle geometry is also very close, both species having spatulate spines, but upon closer examination they differ significantly in their ectosomal structure, spicule geometry and spicule dimensions. Similarly, Lendenfeld's (1888) comparison with Montagu's Axinella hispida is merely speculative, being based on growth form and the presence of a prominent surface hispidation. Raspailia tenella has a classical raspailiid ectosomal skeleton, whereas R. gracilis has vestigial ectosomal structure, with most spicule brushes lying parallel to the surface. Additionally, in R. tenella the echinating megascleres are dispersed fairly evenly throughout the skeleton, not concentrated near the periphery. There is no difference between fibre meshes in the core of the choanosomal skeleton from those closer to the periphery, whereas in R. gracilis fibre meshes become larger towards the periphery. Hallmann (1914~)also suggests that the oxeote form of the ectosomal auxiliary spicule in R. tenella is shorter and more slender than other auxiliary megascleres, and that they appear to occur only in the choanosome as single or paired spicules, but this feature was not observed in the lectotype. Raspailiu (Raspailiu) vestigifera Dendy (Figs 15, 16, 109a; Table 3) Raspailia vestigifera Dendy, 1896: 47.-Pick, 1905: 36. Valedictyum vestigifera.de Laubenfels, 1936: 102. Material Examined Holotype. NMV G2468: 'Limebumers' Channel, Cape1 Sound, Port Phillip Heads, Vic., 38' lglS., 144"401E., 22 m depth, coll. J.B. Wilson (stn X, dredge) (? schizotyp+BMNH unregistered slide (RN655): possibly a fragment of NMV G2468). J. N. A. Hooper 1216 Other material. (All material collected by the author using scma, unless otherwise indicated). Darwin Region, N.T.: NTM Z910: EPMFR, 12" 25.01S., 130" 48.4'E., 10 m depth, 31.ix.1982 (stn EP8). NTM 2948: EPMFR, 12" 24 -5'S., 130' 48 -OiE., 10-12 m depth, 13.ix.1982 (stn EP9). NTM 22640, 2704: EPMFR, 9-12 m depth, 3.iv.86 (stn EP28). NTM 2481, 516: Fannie Bay beach, 12" 25.0's.. 130' 50 .OiE., storm debris, 9.ii.82 1982 (EP7). NTM 22012: W. side of Weed Reef, 12'29-2'S., 130'47-llE., 6 m depth, 11.v.1984 (stn WR2). NTM 22165: 'Bommies', N. edge of Weed Reef, 12" 29.5IS., 130' 37 - 6/E., 8-10 m depth, 5.x.1984 (stn WR5). NTM 22193: 6 m depth, 16.xi.1984 (stn WR6). NTM 2853: Oyster beds, Channel I., Middle Arm, 12" 32.7lS., 130"52-5'E., 12-13 m depth, 20.viii.1982, coll. P.N. Alderslade (stn CI3). NTM 22742: Parry Shoals, 11" 11 .41f S., 129' 43.01 'E.,18 m depth, 13.viii.1987, coll. A.M. Mussig (stn 87-3, DON-248). Cobourg Peninsula Region, N.T.: NTM 2153: Sandy I. No. 2, 1l005.6'S., 132" lglE., 8-9 m depth, 22.x.1981, coll. P.N. Alderslade (stn CP29). NTM 2545: 11" 05 IS., 132" 16.6l E., 14 m depth, 2.v.1982 (stn CP34). NTM 22534: Coral Bay, Port Essington, CPMNP, 11" 11 3 I S., 132" 03. OIE., 4 m depth, 18.ix.1985 (stn CP82). Werrel Is, N.T.: NTMZ 3937: N. side of Cumberland Strait, 11' 27 6l S., 136" 28 73' E., 32 m depth, 15.xi.1990 (stn WI-7). NCI Q66C 4816-2: W. headland, Rimbija I., Cape Wessel, 11" 00.5IS., 136"43.SiE., 14m depth, 17.xi.1990, coll. NCI. Northwest Shelf Region, W.A.: NTM 21212: W. of Port Hedland, W.A., 19' 29.4l S., 118' 52- 1IE., 39 m depth, 26.iv.1983 (CSIRO RV 'Soela', stn B9 S02, NWS9, beam trawl). NTM 223.56: NW. of Lacepede I., 16' 34. OIS., 121' 27. OIE., 40-46 m depth, 17.iv.1985, coll. B.C. Russell, Taiwanese Pair Trawlers (stn 85-2, NWS35)). NTM 23479: NW. end of Long I. (Semer I.), Exmouth Gulf, 21" 34.5/S., 114' 39-O'E., 18 m depth, 20.viii.1988, coll. D. Low Choy (stn NWS94). NTM 23049: N. of Amphinome Shoals, 19" 19.7-20.6lS., 119' 08 - SIE., 50m depth, 19.vii.1987 (RV 'Akademik Oparin', stns. 79-81, NWS55, beam trawl). Scott Reef, W.A.: PIBOC 012-202: Near Scott Reef, 16" 33 -4/S., 121" 07 1'E., 4348 m depth, 8.xi.1990 (stn 43). PIBOC unreg.: 16' 36 -33lS., 121" 05.93'E., 40-60 m depth, 7.xi.1990, coll. V. Krasochin, U.S.S.R. RV 'Akademik Oparin', dredge (stn 39). Central Coast, W.A.: NTM 22952: W. of Sunday I., Blind Strait, Dirk Hartog I., Shark Bay, 26" 07.SiS., 113" 14.0iE., 8-9 m depth, 13.vii.1987 (RV 'Akademik Oparin', stn 63, SB5). - - Substrate and Depth Range Invariably attached to rock or coral fragments, never simply embedded in soft substrates; depth range from 3-60 m. Geographical Distribution Confirmed geographical distribution along north-west coast of Australia, from Shark Bay, W.A. to Cape Wessel, N.T. (Fig. 15f). However, the holotype of this species was supposedly collected from Port Phillip, Victoria. It is possible that the species is widely distributed in eastern and western Australian waters, although no other specimens of this species have been found in collections of the NCI or other Australian museums, nor has the species ever been sighted in Queensland, New South Wales, South Australian or Victorian waters (personal observations). Description Shape. Arborescent, digitate, with thin cylindrical even or unevenly bifurcating branches, or with simple branches which taper towards apex. Stalk of variable length (15-45 mm long, 9-22 mm diameter), with enlarged basal plate at point of contact with substrate. Dimensions range from 185-321 mm from base to tip of branches, with branch length and diameter ranging from 12-150 mrn and 3-6 mrn respectively. Colour. Black alive and in ethanol (Fig. 109a). In situ some material silver-grey in appearance due to prominent closely hispid inorganic skeleton protruding through surface, whereas other specimens more fleshy, with ectosomal skeleton virtually entirely covered by black organic (skin-like) dermis. Oscula. Pores not observed in preserved specimens, only seen in few live specimens. In live material, with flaccid organic dermis over protruding inorganic skeleton, oscula up to 4 mm diameter on lateral margins of branches. Texture and suqace characteristics. Surface typically extremely hispid, with long subectosomal megascleres protruding up to 1.5 mm from surface, forming even, closely packed palisade. Organic ectosome optically even but microscopically hispid from smaller Australian Raspailiidae Fig. 15. Raspailia vestigifera Dendy (specimen NTM 22534) (type species of the nominal genus Valedictyurn de Laubenfels): a, choanosomal axial styles/anisostrongyles; b, ectosomal auxiliary style; c, echinating acanthostyle; d, subectosomal extra-axial style; e, section through peripheral skeleton; f , known Australian distribution. ectosomal megasclere brushes dispersed between protruding subectosomal spicules. Branches firm, flexible, whereas basal stalk more woody. Ectosome and subectosome. Ectosomal region pierced by 2 forms of spicules: larger subectosomal extra-axial megascleres forming optically hispid surface, and smaller ectosomal auxiliary spicule brushes surrounding protruding megascleres. Extra-axial skeleton regularly radial, with bases of long styles (or anisoxeas) embedded in axial skeleton, surrounded by very heavy brown spongin matrix. Subectosomal spicules piercing surface surrounded by plumose brushes of styles (or anisoxeas), usually not forming continuous dermal palisade. Choanosome. Choanosomal axial skeleton heavily condensed, with spicules usually running longitudinally through branches. Spongin fibres very heavy, dark brown, forming tight-meshed reticulation, cored by uni- or paucispicular tracts of choanosomal oxeas (or anisoxeas or strongyles). Echinating acanthostyles relatively uncommon, almost vestigial in some specimens, occurring most frequently at junction of axial and extra-axial skeletons. Megascleres (refer to Table 3 for measurements). Choanosomal spicules coring axial fibres thick, relatively short, slightly curved centrally, often asymmetrical, with stylote, oxeote or strongylote tips. Subectosomal extra-axial spicules very long, thick, straight or very slightly curved, asymmetrical, with stylote or anisoxeote ends. Ectosomal auxiliary spicules (forming dermal brushes) relatively short, thin, slightly curved towards basal end, with stylote or anisoxeote points. J. N. A. Hooper Fig. 16. Raspailia vestigifera Dendy: a, holotype (NMV G2468); b, thinly arborescent specimen (NTM Z910); c, thickly arborescent specimen (NTM Z2952), side view; d, top view of same (scale = 30 mm); e, SEM of skeletal structure (left magnified 2 3 . 2 times, right magnified 120 times); f, SEM of extra-axial peripheral skeleton; g, specialised ectosomal skeleton (scale = 1 mm). Australian Raspailiidae 1219 Echinating acanthostyles small, claviform, subtylote, fusiform, with evenly distributed but vestigial spination. Microscleres absent. Remarks This species has typical features of the genus in its spicule complement and skeletal structure, and the creation of a new genus (Valedictyum) by de Laubenfels (1936) was completely unnecessary. Raspailia vestigifera is, in fact, well characterised in having clearly differentiated axial (reticulate) choanosomal, extra-axial (radial) choanosomal, and plumose ectosomal skeletons, and not simply 'plumose architecture' as supposed by de Laubenfels (1936). The species is also typified by its vestigial and claviform acanthostyles, fitting somewhere between the nominal taxa Raspailia and Syringella. Table 3. Comparisons in spicule measurements between specimens of Raspailia vestigiferaDendy Measurements are given in micrometres, presented as ranges (and means) of lengthxwidth, and represent 25 spicules measurements per specimen for each spicule category Material Choanosomal stylotest oxeotes Subectosomal stylotesl anisoxeas NMV G2468 338-864 x 6-22 (597.6~14.7) 853-1805 x 22-72 (1285 - 2x 43 1) NTM various (n = 19) 443-775 x 9-21 (558~5~16.6) 910-2656 x 28-69 (1461-1~36.6) Ectosomal stylotest anisoxeas Echinating acanthostyles 308-414 x 3-9 (358.6~5.5) 66-82 ~ 3 - 45 (74-4~4.1) 132-377 x 1-5 (311.4~3.3) 65-87 x4.5-7 (78.2~5.3) Holotype Specimens The discontiguous distribution of this species (widespread in north-westem Australia and one record from Victoria) is curious. It could be argued that the similarities between populations merely represent the convergence of several characters. In fact there are subtle differences between north-westem Australian and Victorian forms. Megascleres of north-westem Australian specimens are predominantly monactinal, with liberal diactinal modifications, whereas those of the holotype from Victoria are predominantly diactinal, with fewer true monactinal forms. Furthermore, the live coloration of R. vestigifera from Victoria was reported to be 'bottle green with a wash of sepia' (Dendy 1896: 48), turning black in preservative. By comparison, north-westem Australian material is simply black in both the live and preserved states. Nevertheless, on the basis of their mophological features (growth form, surface details, spicule geometry, spicule size, fibre characteristics and skeletal architecture), the two 'populations' are clearly related and they are considered here to be conspecific. The biogeographic implications of that synonymy are uncertain, and only chemical or direct genetic evidence will support or refute this synonymy. It is possible that the Port Phillip population represents a relict or enclave of a once much more widely distributed species. Since R. vestigifera has a thinly arborescent growth form which it shares with many other raspailiids, and these species are likely to be identified initially by their superficial features (surface details and growth form), it is worthwhile comparing this species with others that share some of those features. Raspailia pinnatifida (Carter 1885: 353) from Port Phillip is perhaps cIosest to R. vestigifera in skeletal structure and spiculation. However, R. pinnatifrda is easily differentiated by the 'rat's-tail' appearance of branches (Dendy 1896: 47), and spicules in R. pinnatifida are significantly smaller. In both species the echinating acanthostyles are relatively scarce. Raspailia arbuscula (Lendenfeld 1888: 215) from northem and eastem Australia also resembles R. vestigifera to a certain extent. It is digitate, with a solid stalk and bifurcate branches, but the branches are flattened rather than evenly cylindrical, and the surface is shaggy and more fleshy, rather than simply hirsute. It also has a more evenly reticulate skeletal architecture, without an obvious condensed axis, its acanthostyles are much more robust, and spicule measurements differ. Raspailia bifircata Ridley (1884: 459) from Torres Strait has similar growth form and 1220 J. N. A. Hooper surface details to R. vestigifera, but differs in spicule geometry, spicule dimensions and live coloration. Raspailia rhaphidophora Hentschel (1912: 371) from Aru I., Indonesia is a species of Thrinacophora and is similar to the present species only in details of its radial and plumose peripheral skeleton. Raspailia hirsuta Thiele (1898: 59) from Japan has rhabdose acanthostyles, and spicule dimensions and spicule geometry are also quite different. Raspailia freyerii Schmidt (1862: 60) from the Adriatic Sea has almost identical gross morphology and hispid surface details as R. vestigifera, whereas other skeletal characters appear to be different. Raspailia humilis (Topsent 1892a: 123) from the north Atlantic resembles R. vestigifera only in external morphology, whereas that species has only a single category of megascleres and lacks ectosomal specialisation completely. Topsent (1928: 42) referred the species to the axinellid genus Axinegella. Raspailia falcifera Topsent (1892~: 124) from the north Atlantic has reticulate branches, lacks echinating megascleres, and differs in spicule dimensions and spicule geometry. Raspailia australiensis Ridley (1884: 460) from Darwin is white, and the surface has minute ridges, with three categories of geometrically similar styles present as megascleres. It also lacks echinating acanthostyles, and belongs to the nominal raspailiid genus Syringella. Raspailia clathrata Ridley (1884: 461) from Torres Strait also lacks echinating megascleres, is grey, with laterally flattened branches, without an obvious hispid surface skeleton, and without echinating acanthostyles. Raspailia mariana (Ridley & Dendy 1886: 480, 1887: 180) from Marion I. has similar choanosomal and ectosomal skeletal structure to R. vestigifera (and many other raspailiids), and possesses a strongly hispid surface, but that species is greyish yellow, with an uneven surface, echinating rhabdostyles and different spicule dimensions and spicule geometry. Raspailia Jlagelliformis Ridley & Dendy (1886: 482, 1887: 190) from the Cape of Good Hope has only a minutely hispid surface, is yellowish grey, lacks echinating spicules, has a more disorganised, less obviously longitudinally directed axial skeleton, and lacks a specialised ectosomal skeleton. Raspailia rigida Ridley & Dendy (1886: 483, 1887: 191), also from the Cape of Good Hope, has comparable extra-axial and ectosomal features as R. vestigifera, but lacks echinating megascleres and the dense horny fibres characteristic of this species. Raspailia profunda Ridley & Dendy (1886: 480, 1887: 181) from both the north and south Pacific Oceans lacks echinating megascleres, is dark reddish brown, and has different spicule dimensions, but is otherwise similar to the present species. It lacks echinating megascleres and thus is a member of the nominal genus Syringella. Raspailia nuda Hentschel (1911: 383) from Shark Bay, W.A., also resembles R. vestigifera in surface details, but is orange-red in life, has flattened lobate branches, lacks echinating acanthostyles, has an extra-axial plumo-reticulate skeleton, and differs in spicule geometry and spicule dimensions. Raspailia dichotoma Whitelegge (1907: 515) from the New South Wales coast, referred here to Ceratopsion, has a grooved, uneven surface, but lacks an obvious hispid ectosomal skeleton, axial core, or echinating megascleres. Raspailia vestigifera is the most prevalent species of this genus living in tropical north-west Australian waters. Handling the species is also potentially dangerous, producing, upon contact with skin, severe dermititis similar to irritation produced by contact with fibreglass. Raspailia (Raspailiu) wilkinsoni, sp. nov. (Figs 17, 18) Material Examined Holotype. NTM 22734: Back reef, Davies Reef, off Townsville, Great Barrier Reef, Qld, 18' 50fS., 147"39/E., 15 m depth, ll.viii.1989, coll. C.R. Wilkinson (AIMS ref. RA22). Substrate and Depth Range Coral rubble, 15 m depth. Geographical Distribution Central zone, Great Barrier Reef (Fig. 17f). Australian Raspailiidae a b c Fig. 17. Raspailia wilkinsoni, sp. nov. (holotype NTM 22734): a, choanosomal axial oxeas; b, subectosomal extra-axial style; c, ectosomal auxiliary style; d, echinating acanthostyle; e, section through peripheral skeleton; f , known Australian distribution. Description Shape. Stalked, digitate, branching sponge, 175 mm high, 112 mm breadth, with cylindrical, anastomosing, occasionally fused branches, 5-8 mm diameter, and 2 points of attachment to substrate with 2 stalks, 27-32 mm long, 7-9 mm diameter. Colour. Live coloration dark orange-brown (Munsell 5YR 5/8), beige-brown in ethanol. Oscula. Few large exhalant pores scattered on lateral sides and near tips of branches, 2-3 mm diameter, whereas inhalant pores evenly dispersed over branches, 300-500 pm diameter, located between surface conules. Texture and surface characteristics. Texture of digits firm, flexible, whereas basal stalks more rigid. Surface of branches covered with numerous close-set, sharply pointed rnicroconules, 600-2000 p m high, evenly distributed over branches, approximately 2-3 mm apart, whereas basal stalk relatively smooth. Surface conules interconnected by ridges, reminiscent of Callyspongia (Haplosclerida). Ectosome and subectosome. Ectosomal skeleton membraneous, without specialised spiculation, although both choanosomal oxeas and subectosomal styles protrude through 1222 J. N. A. Hooper surface for 50-100 pm. Projecting spicules occur most frequently on ends of surface conules. Ectosome also contains light covering of detritus embedded in superficial ectosomal spongin. Subectosomal extra-axial skeleton with 2 distinct regions: ascending radial skeleton and transverse reticulate skeleton. Ascending radial skeleton with very long tracts, extending up to 1500 pm from surface producing abundant surface conules. Tracts contain horny spongin fibres, varying from 36-95 pm diameter, cored by multispicular bundles of oxeas for most of their length, but spicules occupy only between 50-80% of fibre diameter. Near tips of conules fibres become thicker, 80-130 pm diameter, fully cored by spicules. ~ndivid&lspicules or plumose bundles of extra-axial styles embedded in ends of ascending fibres, protruding less than 80 pm through surface. Transverse reticulate skeleton contains regular or irregular meshes, with widely spaced light spongin fibres, 20-75 pm diameter, forming cavernous rectangular meshes, 110-570 pm diameter. Fibres cored by uni-, paucior multispicular tracts of choanosomal oxeas, occupying less than 80% fibre diameter. ~chinatingacanthostyles only lightly scattered over fibres near axial core, but heavier towards periphery, especially on exterior of fibres. Fibre meshes lined by abundant lightly pigmented type B spongin, enclosing oval choanocyte chambers, 45-110 pm diameter. Few spicules found scattered between fibres. Choanosome. Choanosomal skeleton axially condensed core of very heavy spongin fibres, 80-165 pm diameter, forming tight reticulation of oval meshes, 75-155 pm diameter. Spongin fibres stratified, containing pauci- or multispicular tracts of oxeas occupying less than 30% fibre diameter. Choanosomal mesohyl without obvious type B spongin or loose spicules, whereas spicules and spongin more common in subectosomal reticulate skeleton. Few echinating megascleres found in axial skeleton. Megascleres. Choanosomal oxeas generally long, slender, slightly curved at centre, tapering to long fusiform, sharply pointed tips [151-(214.0)-346 x4-(6 6)-10 pm]. Subectosomal extra-axial styles long, straight, slender or stout, with fusiform tips and evenly rounded bases [267-(353.3)-542 x 3 4 5 6)-10 pm]. Ectosomal megascleres absent. Echinating acanthostyles short, slender, straight, with slightly subtylote bases, tapering to rounded, blunt tips. Spines occur on all parts of spicules, but most abundant on bases and tips. Spines conical, sharply pointed [48-(58 1 ) 4 7 x 4-(5 -4)-7 pm]. Microscleres absent. - - - Remarks This species is placed in the genus Raspailia because of its well differentiated axial and extra-axial skeletons, but in lacking an ectosomal skeleton, having a predominately reticulate extra-axial skeleton and with fibres cored by oxeas the species could as easily be placed in Echinodictyum. The anastomosing digitate growth form, Callyspongia-like surface ornamentation, and minute acanthostyles serve to differentiate this species from other Raspailia. Etymology This species is named in appreciation of the collector of the holotype, and in recognition of his substantial work on Australian sponges, Dr Clive Wilkinson of AIMS. Subgenus Clathriodendron Lendenfeld Raspailia (Clathriodendron) arbuscula (Lendenfeld), comb. nov. (Figs 19, 20; Table 4) Clathriodendron arbuscula Lendenfeld, 1888: 215.-Whitelegge, 1889: 186; Hallmann, 1912: 296; Hallmann, 1914a: 267. Halichondria rubra var. digitata Lendenfeld, 1888: 8 1. Clathriodenderon nigra Lendenfeld, 1888: 216 (in part). Echinonema anchoratum var. ramosa Lendenfeld, in part (as Echinonema ramosa).-Whitelegge, 1901: 81; Hallmann, 1912: 296. [Ceraospina arbuscula Lendenfeld, MS name.] Raspailia agminata Hallmann, 1914b: 438. ? Raspailia agminata.-Bergquist, 1961: 183; Bergquist, 1970: 26. Raspailia nigra.-Hallmann, 1912: 296. Australian Raspailiidae Pig. 18. Raspailia wilkinsoni, sp. nov.: a, holotype (NTM 22734) (scale = 30 mm); b, SEM section through branch; c, SEM section through peripheral skeleton; d, SEM of fibre characteristics; e, SEM of echinating acanthostyles. Material Examined Lectotype (here designated). AM G9045: Port Jackson, N.S.W., 33" 511S., 151" 16'E., depth and date of collection unknown, Paralectotypes (5). BMNH 1887.1.24.29: same locality (wet; slide AM G3534). AM 2107: same locality (NMV sponge archives 111-2). BMNH 1957.8.30.4: same locality (dry; NMV sponge archives 49/27). BMNH unregistered: same locality (dry; as C. nigra var. arbuscula; NMV sponge archives 49/28). AM Z109: locality unknown (holotype of Ceraospina arbuscula Lendenfeld (MS name ) and Echinonema anchoratum var. ramosa; NMV sponge archives 13/14-16). Holotype of C. nigra var. jacksoniana. BMNH 1887.1.24.64: Port Jackson, N.S.W. (slide AM G3535). Other Material. Specimen of C. nigra var. jacksoniana. BMNH 1887.4.27.112: Port Jackson, N.S.W. 'Holotype' of C. nigra var. microspina -BMNH 1988.9.23.2.b: New Zealand (slide only). Specimen of Clathriodendron arbuscula -BMNH 1886.12.12.22: Precise locality unknown, N.S.W. (Ramsay Collection). J. N. A. Hooper 1224 Recent Material. NTM 21429: Gove region, N.T.,vicinity of 12" S, 137" E, unknown depth, ll.i.1971, coll. N.T. Fisheries ('mud fauna', trawl). NTM 23949: W. headland, Rimbija I., Cape Wessel, Wessel Is, N.T., 11" 0-5'S., 136" 43 -SfE., 15 m depth, 16.xi.1990, coll. J.N.A. Hooper (stn WI-9). Substrate and Depth Range Known only from soft mud and shale bottom; depth distribution unknown. Geographical Distribution North and east coasts of Australia (Fig. 19e) and New Zealand waters. Fig. 19. Raspailia arbuscula (Lendenfeld) (lectotype AM G9045) (type species of the nominal genus Clathriodendron Lendenfeld): a, echinating acanthostyles; b, subectosomal extra-axial styles; c, ectosomal auxiliary styles/anisoxeas; d, section through peripheral skeleton; e, known Australian distribution. Description Shape. Series of numerous elongate, laterally flattened, digitate fronds (NTM 21429: 30-70 mm long, 4-8 mm broad, 1-2 mm thick), connected to common base (with total mass measuring 110x70 mm), superficially resembling Ciocalypta (Halichondriidae). Colour. Live coloration unknown, grey in ethanol (Munsell 7 . 5 YR 512). Oscula. Not seen. Tature and sugace characteristics. Surface prominently hispid, with arenaceous particles embedded in ectosome producing uneven conules dispersed over surface. Texture harsh, branches easily flexible, relatively fragile (easily tom). Ectosome and subectosome. Ectosomal skeleton bears moderate quantities of arenaceous particles, few incorporated into skeleton. Protruding long subectosomal megascleres prominent, and may (NTM 21429) or may not (other material) be surrounded at point Australian Raspailiidae 1225 Fig. 20. Raspailia arbuscula (Lendenfeld): a, lectotype (AM G9045); b, holotype of variety jacksoniana (BMNH 1887.1.24.64); c, specimen from the N.T. (NTM 21429); d, paralectotype (BMNH 1887.1.24.29); e, specimen from Ramsay collection (BMNH 1886.12.12.22) (scale = 30 mm); f , SEM of skeleton (paralectotype AM Z109) (left magnified 46.4 times, right magnified 274 times); g, SEM of echinating acanthostyle (NTM 21429); h, skeletal structure (BMNH 1887.4.27.112) (scale = 1 mm); i. SEM of axial skeleton. at which they protrude through surface by very few, sparsely dispersed brushes of thin ectosomal styles (or anisoxeas). Bases of subectosomal megascleres embedded in axial skeleton, enveloped by heavy fibrous spongin, protruding through surface as single spicules or paucispicular brushes, forming extra-axial skeleton. Subectosomal megascleres near centre of axial skeleton smaller than spicules occurring towards periphery. 1226 J. N. A. Hooper Choanosome. Choanosomal axial skeletal not condensed or well differentiated from extra-axial region, consisting of very heavy spongin fibres forming irregular reticulation of longitudinal tracts, without special category of choanosomal megascleres, but with aspicular fibres or echinating spicules secondarily incorporated into fibres. Fibre reticulation forms rhomboidal meshes. Skeletal tracts heavily echinated by acanthostyles, in slightly heavier concentratations towards periphery than at core. Mesohyl matrix with moderately heavy dark brown granular spongin. Megascleres (refer to Table 4 for measurements). No special category of choanosomal megascleres present coring fibres, although subectosomal styles from extra-axial skeleton extend into axial core. Table 4. Comparisons in spicule measurements between specimens of Raspailia arbuscula (Lendenfeld) Measurements are given in micrometres, presented as ranges (and means) of lengthxwidth, and represent 25 spicules measurements per specimen for each spicule category Material Choanosomal megascleres Absent Absent Absent BMNH 1887.1.24.64 Absent BMNH 1887.1.24.29 Absent BMNH 1887.4.27.112 Absent NMNZ CE6* Absent NTM 21429 Absent A Subectosomal styles/ anisoxeas Ectosomal styles/ anisoxeas Echinating acanthostyles Lectotype 820-1545 x 11-26 (1169.8~18.4) Paralectotypes 633-961 X9-19 (753.2~14-6) 860-920 X 15-23 (897.4~ 19.1) 930-2058 x 16-32 (1360-6~23.4) 680-902 x 11-22 (791.8~16.1) Specimens 747-1705 x 5-22 ( 1 0 4 5 . 4 ~10.4) 930-2300 x 17.5-20 ( 1 8 4 0 ~19.2) 571-3857 x 12-36 (1688~2~23.8) Raspailia agminata sensu Bergquist (from Bergquist 1970: 27). Subectosomal extra-axial megascleres long, setaceous, straight or slightly curved towards basal end, with rounded stylote, sometimes subtylote, less commonly asymmetrical oxeote ends. Ectosomal auxiliary styles or anisoxeas, if present, relatively small, thin, slightly curved at centre or base, with fusiform points and rounded stylote or fusiform oxeote bases. Echinating acanthostyles relatively small, subtylote, with large granular spination, an aspinose region near basal region. Microscleres absent. Australian Raspailiidae 1227 Remarks This species is remarkable for its apparently highly variable ectosomal development, absence of a special category of choanosomal spicule, and lack of well marked axial and extra-axial differentiation. In possessing a reticulate skeleton the species would belong to the nominal genus Clathriodendron, if it were valid. The species is also unusual in having Clathria-like acanthostyles (tapering conical, subtylote, with aspinose 'necks'), which may be secondarily incorporated into fibres (as coring spicules). There are some nomenclatural anomalies associated with this type species of the genus Clathriodendron that require further discussion. Firstly, Clathiodendron arbuscula has 5 extant 'syntypes' (nominated above as paralectotypes, 3 from the BMNH and 2 from the AM). There are also several other 'types' associated with this species. One 'syntype' (AM Z109). labelled 'Echinonema anchoratum var. ramosa Lendenfeld, type of Ceraospina arbuscula Lendenfeld' has also been associated with Clathriodendron arbuscula, as indicated by Hallmann (1912: 296). That specimen represents one of the very few examples in which a correlation exists between Lendenfeld's (presently missing) 'key-list' (species No. 307) and an actual Museum voucher specimen (e.g. Whitelegge 1901: 64). That material is conspecific with R. arbuscula. In contrast, other 'syntypes' [spirit specimen BMNH 1886.12.12.22 (plus slides ZMB 1143, 6539, 6541), AMU G652 (spirit), and G9 125 (dry), labelled Clathrissa arbuscula Lendenfeld (1888: 2 17) (with Museum label annotation 'Clathriodendron arbuscula Ldf.')] are obviously different. That material is conspecific with Plumohalichondria caespitosa (Carter) (family Anchinoidae). Secondly, Lendenfeld (1888: pl. 2, fig. 1) illustrated an example of Halichondria rubra var. digitata Lendenfeld, which was subsequently redescribed and re-illustrated by Hallmann (1914b: 438, pl. 23, fig. 4) as Raspailia agminata Hallmann. It appears that the specimen figured by both authors (viz. the holotype of R. agminata) is also a syntype of C. arbuscula (viz. AM G9045), which therefore makes R. agminata an objective synonym of C. arbuscula. Thirdly, Bergquist (1961: 183, 1970: 26) described R. agminat~from New Zealand waters. She described the species as possessing a plumose axial, and plumo-reticulate extra-axial skeletal architecture, together with a specialised ectosomal skeleton consisting of protruding subectosomal subtylostyles and subdermal brushes of small oxeas, which were not arranged around the protruding structural spicules, but nevertheless were reminiscent of the typical Raspailia condition. It is not certain whether there is any justification in separating New Zealand material [viz. Berquist's (1970) specimen NMNZ CIE6 and Lendenfeld's (1888) specimen BMNH 1988.9.23.2.bl of R. agminata from Port Jackson material (viz. AM G9045), given the range of variability now known to be possible for raspailiid taxa. However, from re-examination of R. arbuscula, including the holotype of R. agminata, it is confirmed here that Hallmann's (1912: 297) redescription of the species is accurate: i.e. R. arbuscula s.s. has no obvious ectosomal skeleton of oxeote megascleres, nor is there any evidence of a distinctive plumose architecture. (The dates of collection for the type material are uncertain, but they are at least over a century old.) Fourthly, the specimen described above from the Gove region is almost identical in external morphology to the holotype of R. agminata (AM G9045) and E. anchoratum var. ramosa (AM Z109) (Fig. 20c), and it also has an exclusively reticulate architecture (longitudinal primary and transverse secondary fibres), but with the rudiments of a specialised ectosome (with sparse brushes of anisoxeas-fusiform styles, some of which were grouped around the bases of protruding choanosomal megascleres) (cf. Bergquist 1970: 26). The existence of such material, which also indicates a considerable geographical variability for the species, poses questions concerning the validity of ectosomal characteristics as a diagnostic character for species and genera of Raspailiidae. Lendenfeld's (1888) description of Clathriodendron nigra is erroneous, and the species is taken in the sense of Hallmann (1912: 296). Moreover, Lendenfeld's 'type series' consists of a conglomerate of two species: the identity of C. nigra var. hirsuta from Mauritius is unknown, but it is certainly not conspecific with R. arbuscula, whereas varieties microspina and jacksoniana from New Zealand and Port Jackson, respectively, are clearly identical to this species. 1228 J. N. A. Hooper Raspailia agminata is therefore obviously a synonym of R. arbuscula, and evidence cited above, and by Hentschel (1911) and Bergquist (1970) clearly shows that Clathriodendron cannot be maintained with sufficient generic distance from Raspailia. Raspailia (Clathriodendron) bifurcata Ridley (Fig. 21) Raspailia bifurcata Ridley, 1884: 459.-Whitelegge, 1901: 92; Pick, 1905: 29. Material Examined Holotype. BMNH 1882.2.23.256: Prince of Wales Channel, Torres Strait, Qld, 10' 301S., 142' 13'E., 9-13 m depth, coll. R.W. Coppinger (HMS 'Alert'). Other material. BMNH 1892.2.6.33: Holothuria Banks, NW. of Cape Londonderry, Timor Sea, W.A., vicintity of 13' 201S, 126"E, 22 m depth, 13.ix.1891, coll. Mr Bassett-Smith (HMS 'Penguin'). Substrate and Depth Range Shell and sand substrate, 9-22 m depth. Geographical Distribution Torres Strait, Qld, NW. coast of W.A., and Tuggerah Beach, N.S.W. (Fig. 21f). Description Shape. Sinall branching sponges 42-53 mm high, stipitate with short stalk (8-12 mm long, 3-6 mm diameter), with dichotomous branches (7-25 mm long, 1 - 5 4rnrn diameter) which taper to bifurcated fusiform points. Colour. Preserved and dry specimens with purple stems and lighter branch tips (Ridley 1884; Whitelegge 1901). Oscula. Not observed. Texture and su$ace characteristics. Texture of both stem and base firm and rigid, branches more flexible. Surface optically even, prominently hispid, with close-set microconules bearing spicule brushes. Ectosome and subectosome. Ectosomal skeleton membraneous around microconules, although axial spicules lie just under surface and tangential to it in other parts of branches. No specialised skeleton of small ectosomal spicules present, but subectosomal extra-axial megascleres perched on edges of branches, mainly scattered around microconules, protruding a long way through surface, in bundles or singly, at oblique angles. Only bases of subectosomal extra-axial spicules embedded in peripheral region of branches, enclosed by heavy peripheral fibres. Majority of branch diameter consists of choanosomal fibres and spicule tracts. Choanosome. Choanosomal skeleton lacks any axial condensation, consisting only of moderately heavy, loosely reticulate fibres, cored by choanosomal megascleres criss-crossing in almost halichondroid fashion, running mainly longitudinally through branches. Spongin fibres 180-320 p m diameter. Anastomoses produce oval meshes up to 240 p m diameter. Acanthostyles embedded in spicule tracts and spongin fibres scattered sparsely throughout choanosome. Mesohyl matrix contains minimal spongin, and smaller auxiliary spicules also dispersed between (and within) fibres. Megascleres. Choanosomal megascleres of axial skeleton oxeas or anisoxeas, relatively thin, usually with asymmetrical curvature, with fusiform tips [406-(852.7)-1343 x 8 4 1 0 5)15 pm]. Subectosomal extra-axial styles (or occasionally anisoxeas) relatively thick, invariably slightly curved towards basal end, with evenly rounded bases or tapering slightly towards base, with rounded tapering points [1696-(1790.5)-2046 x 12417 2)-25 pm]. Ectosomal megascleres absent, but small auxiliary oxeas and anisoxeas present, thin, straight or curved centrally, with fusiform raphide-like points [292-(332 3)-402 X 1 .5-(2 8)-4 pm]. Echinating acanthostyles small, cylindrical, subtylote or evenly rounded at basal end and rounded at apex, with even granular spination 159466-5)-73 x 5 4 6 -4)-8 . 5 pm]. Microscleres absent. - - - - Australian Raspailiidae Fig. 21. Raspailia bifurcata Ridley (holotype BMNH 1882.2.23.256): a, choanosomal axial oxeas; b, subectosomal extra-axial styles; c, auxiliary oxeas;, d, echinating acanthostyles; e, section through peripheral skeleton; f, known Australian distribution; g, holotype (scale = 30 mm); h, echinating acanthostyle; i, axial skeletal structure (scale = 1 mm). 1230 J. N. A. Hooper Remarks It is possible that the smaller auxiliary spicules seen scattered within the choanosomal skeleton are remnants of an ectosomal skeleton, possibly relocated during sectioning, but none of these spicules were seen anywhere near the peripheral skeleton, nor were any associated with subectosomal extra-axial spicule tufts, so it must be assumed that this species lacks a specialised ectosomal skeleton. Raspailia bifurcata is a relatively poorly known eastern Australian species. It is characterised by having a reticulate skeletal structure similar to Echinodictyum spp., fitting with the nominal subgenus Clathriodendron. The species lacks an axially condensed skeleton, but unlike Echinodictyum it retains at least some axial and extra-axial skeletal differentiation. Raspailia (Clathriodendron) cacticutis (Carter) (Fig. 22) Dictyocylindrus cacticutis Carter, 1885: 354. Raspailia cacticutis.-Dendy, 1896: 48; Pick, 1905: 35; Shaw, 1927: 427; Wiedenmayer, 1989: 55-56, pl. 5, figs 6-7, pl. 23, fig. 5, text-fig. 37. Aulospongus cacticutis.-Dendy, 1905: 176. Clathriodendron cacticutis.-Hallmann, 1912: 297. Material Examined Holotype. BMNH 1886.12.15.120: Port Phillip Heads, Vic., 38 m depth, coll. J.B. Wilson (schizotypes-BMNH 1886.12.15.450, AM G2798, NMV sponge archives 38/25-26). Substrate and Depth Range Known only from shallow coastal waters, to 38 m depth. Geographical Distribution Port Phillip Heads, Bass Strait, Vic., and Maria I., Tas (Fig. 224. Description Shape. Stipitate, club-shaped flabellate sponge (68 mm high, 60 mm broad), with short stalk (9 rnrn high, 7 mm diameter) and lobate branches fused together for some or all of their length. Colour. Olive green-brown, grey-brown or black-brown in life, darker in preserved state. Oscula. Small pores scattered on apical margins of lobate branches, up to 1 . 5 mm diameter, flush with surface. Texture and su$ace characteristics. Surface dense with distinct skin-like covering, raised into prominent cactiform microconules scattered over branches or on apex of longitudinal surface ridges. Texture compressible on branches, woody on stem. Ectosome and subectosome. Ectosomal skeleton membraneous, without specialised dermal spiculation, with numerous acanthostyles protruding through heavy granular, darkly pigmented ectosome. Subectosomal skeleton reduced to single, or groups of subectosomal extra-axial megascleres, mostly running longitudinally through branches and protruding through surface for only short distance. Heavy spongin fibres running just below ectosome and major fibre junctions corresponding with surface conules prominent features of both ectosome and subectosomal skeletons. Conules with or without protruding subectosomal megascleres. Choanosome. Choanosomal skeleton reticulate, without trace of axial condensation. Heavy spongin fibres (up to 160 p m diameter) form relatively tight reticulation, with oval meshes (up to 350 pm diameter), heavily echinated by acanthostyles, without choanosomal axial spicules, with bases of long subectosomal extra-axial spicules embedded within and running mostly transversely through branches (towards periphery). No pronounced axial and extra-axial differentiation, with only remnants of extra-axial skeleton present (i.e, long subectosomal styles). Spongin in mesohyl relatively heavy, granular. Australian Raspailiidae Fig. 22. Raspailia cacticutis (Carter) (holotype BMNH 1886.12.15.120): a, subectosomal extra-axial style; b, echinating acanthostyle; c, section through peripheral skeleton; d, known Australian distribution; e, holotype (scale = 30 mm); f , echinating acanthostyle; g, skeletal structure. 1232 J. N. A. Hooper Megascleres. Choanosomal axial megascleres absent. Subectosomal extra-axial megascleres long, relatively thick, gently curved, tapering to sharp points, with subtylote bases [427-(550 0-54 x 9-(12 8)-18 pm]. Ectosomal auxiliary megascleres absent. Echinating acanthostyles short, cylindrical, with subtylote or evenly rounded bases, tapering to rounded tips, with moderately heavy granular spination distributed evenly [85-(91.7)-101 x 749.5)-12 pm]. Microscleres absent. Remarks In many respects the species is close to Echinodictyum (e.g. the possession of a reticulate fibre skeleton, the lack of a specialised ectosomal skeleton, the presence of dense concentrations of acanthostyles echinating fibres, and the protrusion of these echinating megascleres through the ectosome), but it preserves the remnants of an extra-axial skeleton (long subectosomal styles), although it has lost any trace of an axial mineral skeleton. For some reason Dendy (1905) referred the species to Aulospongus, perhaps on the erroneous belief that it had a plumose columnar skeleton, even though in an earlier publication (Dendy 1896) he diagnosed the species as belonging in Raspailia. Hallmann (1912) was more accurate in suggesting that the species had a reticulate skeleton, and thus was a Clathriodendron, closely related to C. arbuscula. The placement of this species in any genus of Raspailiidae is problematic, and illustrates the grey area between species of Raspailia with axial or reticulate choanosomal skeletons. This species has been recently redescribed and well illustrated from the vicinity of the type locality (Wiedenmayer 1989). Raspailia (Clathriodendron) danuinensis, sp. nov. (Figs 23, 24, 109c; Table 5) Material Examined Holotype. NTM 22239: EPMFR, Darwin, N.T., 12" 25. 0' S., 130" 49.1 'E., intertidal, 8.iii.1985, coll. J.N.A. Hooper (stn EP21, by hand). Paratypes. NTM 23210: EPMFR, Darwin, N.T., 12" 24.5/S., 130"48.01E., intertidal, 25.ix.1987, coll. N. Smit (stn EP34, by hand). NTM 2439: Precise locality unknown, Darwin region, N.T., 14.xi.1974, depth unknown, coll. T. Jackson (RRIMP FN0893, stn Don. 2, hooker). Other material. Darwin, N.T.: NTh4 2805: Channel I., Middle Arm, 12" 32 - 3/ S., 130' E, 11 m depth, coll. S. Chidgey, (Channel I. EIS (FN 12), stn Don. 23, hooker). Scott Reef, W.A.: PIBOC 012-1 16: near Scott Reef, 16" 32.2/ S., 121' 10.9' E., 43-44 m depth, 4.xi.1990, coll. V. Krasochin, U.S.S.R. RV 'Akademik Oparin', dredge (stn 28). Substrate and Depth Range Found on intertidal or shallow subtidal rock reefs to 44 m depth, covered by muddy-sand substrates, usually attached to rock or dead coral fragments, and in areas of extremely high sedimentation. Geographical Distribution Darwin region, N.T., and Northwest Shelf, W.A. (Fig. 23jJ Description Shape. Lobate cactiform sponges (50-85 mm high, 65-88 mm maximum breadth), branching in more than one plane, with flattened, thickly lamellate, lobe-like branches (36-42 mrn high, 32-56 mm wide, 3-9 mm thick). Branches with slightly convoluted margins, expanded basal plates and short thick stalks (13-26 mm long, 14-18 mm diameter), Colour. Dark orange-brown or yellowish brown in life (Munsell 5YR 6/10-4/8) (Fig. 109c), grey-brown in ethanol. Oscula. Not observed. Texture and surface characteristics. Surface mostly even, prominently hispid, always silt covered in life. Texture firm, barely compressible, arenaceous, with woody stalk and flexible lobate branches. Australian Raspailiidae Fig. 23. Raspailia daminensis, sp. nov. (holotype NTM 22239): a, choanosomal axial oxeas;. b,. subectosomal extra-axial oxeas/anisoxeas:. c.. ectosomal auxiliary oxeas; d, echinating acanthostyles; e, section through peripheral skeleton; f,known Australian distribution. Ectosome and subectosome. Ectosome with protruding subectosomal extra-axial megascleres, with bases embedded in peripheral choanosomal spicule tracts and shafts surrounded by bundles of ectosomal megascleres at surface. Ectosomal spicule brushes form discrete bundles on surface, not producing continuous palisade; some brushes lie tangential to surface. Heavy deposits of granular dark brown spongin in ectosomal region. Subectosomal region reduced to small cavernous layer immediately below dermal skeleton, containing only bases of protruding subectosomal extra-axial megascleres occurring individually or in bundles. Choanosome. Choanosomal axial skeleton very irregularly reticulate, almost halichondroid, occupying most of branch diameter, without noticeable axial condensation. Choanosomal fibres heavily invested with yellow-brown spongin, very widely spaced, irregularly reticulate, with major proportion of fibres running longitudinally through branches. Primary fibres cored by multispicular tracts of choanosomal oxeas, interconnected by light secondary fibres J. N. A. Hooper Fig. 24. Raspailia daminensis, sp. nov.: a, b, paratype (NTM 20439) side and top views (scale = 30 mm); c, peripheral skeleton (scale = 1 mm); d, SEM of axial skeleton; e, SEM of spiculo-spongin fibre; f, SEM of echinating acanthostyle. with pauci- or multispicular tracts of oxeas, with occasional smaller aspicular tertiary fibres. Fibre meshes triangular or irregularly oval, and choanocyte chambers oval (180-305 pm diameter). Numerous choanosomal megascleres also dispersed within mesohyl, contributing to halichondroid appearance of axial skeleton. Mesohyl matrix contains abundant granular spongin. Echinating acanthostyles abundant, predominant at fibre nodes. Australian Raspailiidae 1235 Megascleres (refer to Table 5 for dimensions). Choanosomal axial megascleres invariably true oxeas, very thick, slightly curved centrally, abruptly pointed. Subectosomal extra-axial megascleres predominantly oxeas but some asymmetrical anisostyles also present, long, thin or thick, straight or slightly curved centrally, with symmetrical slightly fusiform points. Ectosomal auxiliary megascleres very thin, predominantly oxeote, straight or only rarely asymmetrically curved, with sharply pointed fusiform tips; very occasional anisoxeote or stylote forms also occur. Echinating acanthostyles very thin, claviform, usually straight, with prominent subtylote bases, tapering to sharp points, with granular, vestigial, evenly dispersed spination. Microscleres absent. Table 5. Comparisons of spicule measurements between specimens of Raspailia daminensis, sp. nov. Measurements are given in micrometres, presented as ranges (and means) of lengthxwidth, and represent 25 spicules measurements per specimen for each spicule category. Material Choanosomal oxeas Subectosomal oxeas NTM 22239 365-652 x 12-33 (514.4~23.8) 1438-3277 x 8-26 ( 2 5 6 6 - 7 ~18-6) NTM 2439 3 12-648 x 10-33 (494.6~20~4) 434-592 x 8-32 (497.0~24.6) 1632-3520 x 11-33 (2676.7~24.0) 1044-3065 x 12-31 (2139.1 ~ 2 2 . 3 ) 512-957 x 22-45 (673.7~37.3) 1121-3057 x 18-32 (2325.7~25-3) Ectosomal oxeas Echinating acanthostyles 322-576 ~2.5-6 (454.1~4.3) 142-189 ~4.5-12 (153.9~6.9) 492-848 ~4-10 (647-6~8.5) 357-753 x 2-7 (580.4~42) 112-159 x 6-9 (143.4~7.6) 123-161 x 5-8 (142.4~6.4) 348-823 x 1-9 (567.0~6.1) 129-189 ~5-8.5 (164-3~6.6) Holotype Paratypes NTM 23210 Specimen NTM 2805 Remarks This species is well characterised by its almost halichondroid skeletal architecture without any noticeable axial condensation, with an extra-axial skeleton reduced to single spicules or bundles of long subectosomal megascleres poking through the surface, with exclusively oxeote spiculation, including robust oxeas coring axial fibres, and distinctive cylindrical, sharply pointed, delicate echinating acanthostyles. Morphological features, spicule geometry and spicule dimensions are relatively consistent among the five known specimens of this species, but there appears to be some variability in the size of ectosomal oxeas: those in the holotype are considerably smaller than those in the other material (Table 5). Raspailia darwinensis is closest to R. paradoxa Hentschel from Bunbury, SW. Australia, showing similarities in external morphology and surface features; both species are cactiform and hispid, with flattened lobate branches, but R. paradoxa has smaller extra-axial subtylostyles ( 1 6 0 0 ~ 15-17 pm), no oxeas coring fibres, heavily spined and more robust echinating acanthostyles (128-160 x 10-12 pm), ectosomal 'tomotoxae' (496-568 x 6-9 pm), and an open, more regular reticulate skeleton of heavy primary fibres and lighter connecting fibres. The species shows some similarities in skeletal architecture to R. desmoxyiformis, sp. nov., although the two differ significantly in most other features, and it should also be contrasted with the desmoxyid sponge Higginsia massalis from the Darwin region, in having excessively large choanosomal oxeas forming a more-or-less halichondroid axial skeleton. J. N. A. Hooper Etymology The species is named for its type locality, Darwin Harbour. Raspailia (Clathrwdendron) desmoxy~ormt,sp. nov. pigs 25, 26) Material Examined Holotype. NTM 21259: W. of Port Hedland, NWS, W.A., 19" 28.5'S., 118" 55.3'E., 40 m depth, 26.iv.1983, coll. J.N.A. Hooper (CSIRO RV 'Soela' S02183, stn B9-NWS10, beam trawl). Substrate and Depth Range Shallow offshore rock reefs, sand and gravel; 40 m depth. Geographical Distribution Known only from the type locality, Northwest Shelf, W.A. (Fig. 25e). a b F Fig. 25. Raspailia desmoxyiformis, sp. nov. (holotype NTM 21259): a, choanosomal axial and ectosomal oxea; b, subectosomal extra-axial style; c, echinating acanthostyle; d, section through peripheral skeleton; e, known Australian distribution. Description Shape. Massive, tubulo-digitate sponge (110 mm long, 60 mm maximum breadth, 30 mm thick), with 3 large cylindrical lobate digits (40-65 mm long, 15-22 mm diameter) lying in 1 plane, fused at bases and tapering towards points, with rounded margins. Colour. Pale orange alive (Munsell 5YR 7/10), beige-brown in ethanol. Oscula. Single small osculum on terminal or subterminal end of each lobate digit, up to 2 mrn diameter when contracted (preserved). Texture and surface characteristics. Texture soft, compressible, easily tom, with thick detachable skin-like dermis, superficially resembling Crella incrustans. Surface even, optically smooth, not hispid. Australian Raspailiidae 1237 Ectosome and subectosome. Ectosome microscopically hispid, or evenly microvillose, due to presence of heavy continuous dermal palisade composed of close-set plumose brushes of choanosomal oxeas. Specialised ectosomal megascleres absent. Only remnants of extra-axial skeleton present, with long subectosomal styles lying singly or in paucispicular groups, forming radial brushes extending from deeper halichondroid choanosome up to and inserted below plumose ectosomal brushes. Extra-axial styles barely protrude through dermal skeleton. Spicule brushes on ectosome perched over tight reticulate bundles of choanosomal oxeas in peripheral skeleton. Peripheral skeleton appears narrow radial subdermal band of megascleres. Spongin fibres absent, and only light mesohyl matrix in ectosomal and subectosomal regions. Subdermal cavities relatively numerous at bases of spicule brushes on ectosome. Echinating megascleres rare in peripheral regions of skeleton. Choanosome. Choanosomal skeleton not noticeably axially condensed; axial-extra-axial differentiation prominent at periphery. Choanosomal skeleton plumose at periphery, becoming increasingly reticulate and halichondroid towards core. Main skeletal tracts ascend towards surface, composed of rnultispicular lines of choanosomal oxeas interdispersed with long subectosomal styles; tracts interconnected at irregular intervals and obtuse angles by unior paucispicular tracts of oxeas. Skeletal structure increasingly disorganised towards core. Spongin fibres present but not easily differentiated from heavy mesohyl matrix surrounding skeletal reticulation. Fibre anastomoses produce small oval meshes. Echinating acanthostyles abundant, scattered throughout axial (but not peripheral) skeleton. Echinating megascleres incorporated into skeletal tracts. Megascleres. Choanosomal oxeas of axial skeleton also form ectosomal brushes, thick, almost perfectly straight, with nearly fusiform symmetrical points [395-(422.2)-544~ 10(14-5)-17 pm]. Subectosomal extra-axial styles relatively long, thick, with fusiform points, rounded non-tylote bases r3924781.3)-1053 x 14-(17 6)-22 pm]. Ectosomal spicules absent. Echinating acanthostyles large, thick, straight or very slightly curved towards basal end, with prominent subtylote bases covered with large spines, fusiform tips with spines extending up to 213 of spicule length, typically with aspinose necks [223-(283 3)-3 18 x 12421 .0)-26 pm]. Microscleres absent. Remarks Were it not for the presence of echinating acanthostyles and organised brushes of spicules on the ectosome, this species would probably have been referred to the family Desrnoxyidae. The geometry of choanosomal oxeas (relatively long, straight, thick, heavily silicified), the presence of well marked skeletal organisation only at the periphery, and the radial extra-axial skeleton also suggests superficial similarities with the halichondriids. Nevertheless, this species is included with the Raspailiidae in having echinating acanthostyles and the remnants of an extra-axial skeleton. The species is atypical and its specific affinities are unknown. The species can be differentiated from all other Raspailia by the abnormally large acanthostyles, the presence of apparently undifferentiated robust oxeas in both the choanosomal and ectosomal skeletons, an extra-axial skeleton reduced to individual or paucispicular brushes of styles which do not pierce the ectosome, and a halichondroid choanosomal skeleton. Etymology This species is named for the relatively disorganised, nearly halichondroid, choanosomal skeleton composed of robust oxeas, resembling to some extent similar skeletal types in the family Desmoxyidae. Raspailiu (Clathriodendron) keriontria, sp. nov. (Figs 27, 28) Material Examined Ho1ot)pe. NTM 21184: W. of Port Hedland, NWS, W.A., 19" 30-9'S., 118" 48.7'E., 40 m depth, 26.iv.1983, coll. J.N.A. Hooper (CSIRO RV 'Soela' S02183, stn B7-NWS8, beam trawl). J. N. A. Hooper Fig. 26. Raspailia desmoxyiformis, sp. nov.: a, holotype (NTM 21259) (scale = 30 mm); b, SEM of skeletal structure; c, SEM of axial skeleton; d , SEM of echinating acanthostyles; e, peripheral skeleton (scale = 1 mm); f , spiculation (scale = 500 pm). Substrate and Depth Range Rock and sand substrate, 40 m depth. Geographical Distribution Northwest Shelf, W.A. (Fig. 27e). Description Shape. Elongate lobate, subcylindrical, laterally flattened, honeycombed sponge (60 mm long, 37 mm maximum breadth, 20 rnm maximum thickness), with short basal stalk (10 mm long, now detached). Colour. Live coloration yellowish grey-brown (Munsell 7 - 5 YR 6/2), slightly lighter on points of surface projections, even grey-brown in ethanol. Australian Raspailiidae Fig. 27. Raspailia keriontria, sp. nov. (holotype NTM 21184): a, choanosomal axial and echinating acanthostyles; b, subectosomal extra-axial style; c, second categoIy of echinating acanthostyle; d, section through peripheral skeleton; e, known Australian distribution. Oscula. Surface excavated by numerous large cavities (up to 8 mm diameter) extending completely through sponge, and scattered oscula (up to 3 rnrn diameter) associated with membraneous coverings over surface excavations. Texture and surjiace characteristics. Texture tough, compressible, difficult to tear. Surface bears numerous sharply pointed, tapering, evenly distributed conulose processes, interconnected by ridges, resembling Ircinia (Dictyoceratida) or Acanthella (Axinellida). Surface covered with distinctive shiny membraneous dermis, stretched over surface excavations in places, producing transluscent window-like panels. Ectosome and subectosome. Ectosome membraneous, without specialised mineral skeleton, covered by very heavy darkly pigmented relatively thick layer of spongin (up to 170 pm thick), reminiscent of order Verongida. Ectosome lacks special dermal megascleres, but with sparsely dispersed long subectosomal megascleres protruding through surface or confined to within axial skeleton, running longitudinally through sponge, with bases embedded in peripheral choanosomal fibres. Long subectosomal megascleres represent remnants of extra-axial skeleton. Choanosome. Skeletal architecture lacks axial condensation or any well developed axial and extra-axial differentiation, apart from protruding extra-axial spicules. Axial core with heavy spongin fibres forming irregular reticulation confined entirely to axis of sponge. Fibres sparsely cored by acanthose choanosomal axial styles, heavily echinated by both larger choanosomal axial styles and second category of smaller echinating acanthostyles (probably J. N. A. Hooper 1240 representing single category of megasclere since intermediate sizes between choanosomal and echinating megascleres occur). Acanthose spicules echinate fibres in stellate-microcionid pattern, similar to Clathria coccinea (Bergquist) group of sponges (family Microcionidae) (Hooper 1990~). Spongin fibre reticulation forms elongate oval meshes; mesohyl matrix with very heavy granular spongin. Megascleres. Choanosomal axial spicules, coring and echinating fibres, thick, straight, with prominent subtylote bases, fusiform points, lightly but variably spined shaft and base, or with spines only on base, or occasionally completely smooth [209-(348 4 ) 4 3 x 9 4 1 1 5)-16 pm]. Subectosomal extra-axial megascleres subtylostyles, long, thick, without spination, slightly curved or straight, with tapering fusiform points, rounded prominent subtylote bases [494-(649 -7)-840 x 1 3 418 3)-22 pm]. Ectosomal auxiliary megascleres absent. Second category of echinating megascleres, identical to but smaller than choanosomal axial spicules, claviform in shape, with subtylote bases, lightly and evenly spined [96-(111- 7)-135 x 7410.6)-14 pm]. Microscleres absent. - - - Remarks Raspailia keriontria shows many similarities to the family Microcionidae in its reticulate skeletal architecture and spicule geometry, but it lacks a monactinal ectosomal skeleton characteristic of microcionids. The species has instead an extra-axial skeleton characteristic of raspailiids, composed of long subectosomal megascleres protruding through the surface. This species is therefore a greatly reduced reticulate raspailiid, and homologous to nominal microcionid genera, such as Anaata, which have acanthose choanosomal megascleres. Raspailia keriontria is also unusual in having two size categories (or more probably an extremely variable size range) of echinating acanthostyles, the larger of which both cores and echinates fibres. Etymology The specific name, keriontria, refers to the honeycomb appearance of the sponge, from the Greek kerion (n., honeycomb) and suffix -tria (f., signifiying agent), literally 'made of wax'. Raspailia (Clathriodendron) melanorhops, sp. nov. (Figs 29, 30) Material Examined Hololype. NTM 2660: W. of Cape Latouche Treville, Lacepede I., NWS, W.A., 17"09.01S., 122' 00-5 'E., 18 m depth, 1.v.1982 (CSIRO RV 'Sprightly' SP4/82, stn 56-Don. 16, dredge). Substrate and Depth Range Rock substrate, 18 m depth. Geographical Distribution Northwest Shelf, W.A. (Fig. 29e). Description Shape. Erect arborescent sponge (235 mm high, 110 mm maximum breadth), with rhizomous basal attachment and large thick woody stem (85 mm long, 15 mm diameter). Stalk bifurcates near apex producing numerous dichotomously divided branches, some anastomosing (40-92 mm long, 2-1 1 mm diameter). Branches irregularly cylindrical, slightly flattened, covered with numerous large pointed conules. Colour. Live coloration unknown, purple-black in ethanol (Munsell 5R 314). Oscula. Small oscula (0 -5-1 mm diameter) dispersed over surface at irregular intervals, frequently obscured by surface detritus. Australian Raspailiidae Fig. 28. Raspailia keriontria, sp. nov.: a, holotype (NTM 21184) (scale = 30 mm); b, SEM of skeletal structure; c, SEM of fibre characteristics; d, SEM of echinating acanthostyle; e, peripheral skeleton showing remnants of the extra-axial skeleton (scale = 500 pm). Texture and surface characteristics. Texture firm, barely compressible, flexible, arenaceous. Surface composed of open reticulation of woody fibres with intertwined branches enveloping moderate quantities of detritus. Surface only slightly hispid from protruding megascleres, but prominently microconulose produced by blindly ending spongin fibres. Ectosome and subectosome. Points of echinating acanthostyles and sparsely dispersed subectosomal extra-axial megascleres protrude through surface, producing microscopically hispid ectosome. No specialised ectosomal megascleres. Choanosomal fibres immediately subdermal, and extra-axial skeleton reduced to single subectosomal spicules. Choanosome. Choanosomal skeletal architecture irregularly reticulate, without any sign of axial condensation and only vestige of axial and extra-axial differentiation. Choanosomal fibres heavily invested with spongin, pigmented red-brown, forming loose, open reticulation. J. N. A. Hooper Fig. 29. Raspailia melanorhops, sp. nov. (holotype NTM 20660): a, choanosomal axial styles; b, subectosomal extra-axial styles; c, echinating acanthostyles; d, section through peripheral skeleton; e, known Australian distribution. Arenaceous particles packed around most fibres but not incorporated into organic skeleton. Fibre anastomoses produce widely spaced ovoid or elongate meshes, with little collagenous spongin between fibres but mainly concentrated around fibre nodes. Fibres cored by multispicular tracts of choanosomal subtylostyles occupying small proportion of fibre diameter (except at fibre nodes). Acanthostyles heavily echinating and evenly dispersed over fibres, but frequently obscured by detritus. Megascleres. Choanosomal axial subtylostyles straight or slightly curved towards basal end, entirely smooth, fusiform, thickest midway along shaft, with subtylote bases [295-(3 13 2)-332 x 6 5 4 7 -8)-9 pm]. Subectosomal extra-axial styles smooth, slightly recurved towards basal end, non-tylote, found predominantly near peripheral fibres [215-(233 .6)-242 x 9410.4)-12 pm]. Ectosomal auxiliary megascleres absent. Echinating acanthostyles small, clavifonn, with slightly subtylote bases, with prominent, large, evenly dispersed spines [81-(85 6)-92 x 6-(8 8)-10 pm]. Microscleres absent. - Remarks Raspailia melanorhops is included in the Raspailiidae in having a tree-like external morphology, styles coring fibres and vestigial extra-axial skeleton in the form of single megascleres standing perpendicular to peripheral choanosomal fibres and protruding through the surface. On the basis of these reduced raspailiid features the affinities of R. melanorhops are difficult to determine. The species is a member of the nominal genus Clathriodendron in having a reticulate architecture, without any trace of axial condensation of the skeleton Australian Raspailiidae 1243 " --"- Fig. 30. Raspailia melanorhops, sp. nov.: a, holotype (NTM 20660) (scale = 30 mm); b, fibre characteristics and detritus in skeleton (scale = 500 pm); c, SEM of skeletal structure; d, SEM of skeletal structure; e, SEM of fibre characteristics; f , SEM of echinating acanthostyle. and only the rudiments of an extra-axial skeleton. If the species had oxeas coring fibres instead of styles it could also be included in Echinodictyum, which has exclusively diactinal choanosomal axial megascleres. Similarly, if the species had isochelae microscleres it could be included in the Microcionidae, since it has close affinities with species of Clathria (Clathriopsamma). In fact, the skeletal architecture of this species is closest to C. (Cl.) tuberosa (Bowerbank) (Hooper, unpublished data). Etymology This species is named for its colour and shape, from the Greek melan (black) and rhops (f., shrub or bush). J. N. A. Hooper Raspailia (Clathriodendron) paradoxa Hentschel (Fig. 31) Raspailia paradoxa Hentschel, 1911: 381. Material Examined None. The holotype (ZMH 4449), from an unknown locality in SW. Australia, has not been found. Substrate and Depth Range Unknown. Geographical Distribution Possibly Bunbury, W.A. (Fig. 31e). Diagnosis Stipitate, branching sponge (55 mm high, 30-40 mrn maximum breadth, 25 mm thick), with lobate plate-like branches anastomosing in places to form cup. Surface heavily hispid and slightly microconulose due to close-set brushes of ectosomal and subectosomal megascleres. Pores up to 1 mm diameter scattered over the surface. Colour in ethanol brownish violet. Ectosomal skeleton detachable, with short ectosomal auxiliary oxeas or anisoxeas forming brushes around protruding extra-axial spicules. Subectosomal extra-axial skeleton has long subtylostyles with bases embedded in peripheral choanosomal skeletal tracts, and also scattered between choanosomal fibres. Choanosomal skeleton not axially condensed nor differentiated between axial and extra-axial skeletons. Choanosomal primary fibres well developed, cored by unispicular tracts of long styles ascending to surface in (?) plumose tracts, interconnected by smaller aspicular secondary fibres. Choanosomal fibre skeleton wide meshed reticulation, echinated by acanthostyles; acanthostyles also scattered between fibre meshes. Fig. 31. Raspailia paradoxa Hentschel: a , ectosomal anisoxea; b, base of subectosomal style; c, echinating acanthostyle; d, holotype; e, known Australian distribution (a-d, redrawn from Hentschel 1911: fig. 51). Megascleres. Choanosomal styles apparently undifferentiated from subectosomal megascleres. Subectosomal extra-axial spicules thin, curved towards basal end, with slightly subtylote bases and tapering to sharp points (up to 1600x 15-17 pm). Ectosomal auxiliary oxeas or anisoxeas thin, straight, sharp-pointed, usually asymmetrical (496-568 x 6-9 pm). Echinating acanthostyles regularly cylindrical, with heavy, evenly distributed recurved spination, sometimes with poorly spined region below basal swelling (128-160x 10-12 pm). Microscleres absent. Australian Raspailiidae 1245 Remarks Although the ZMH accession number of the holotype of R, paradoxa has been discovered (F. Knobbe, personal communication through F. Wiedenmayer), that material has so far been inaccessible to loan. The species is known only from Hentschel's brief description and drawings, but it is perfectly recognisable by its external growth form, reticulate skeleton, and spicule geometry and size. It has been compared with R. danvinensis described above. Subgenus Raspaxilla Topsent Raspailia (Raspaxilla) compressa Bergquist (Figs 32, 33; Table 6) Raspailia compressa Bergquist, 1970: 29-30, text-fig. 3a, pls 7b, lla. Material Examined Holotype. NMNZ POR 30: NE. of North Cape, New Zealand, 173" 04/E., 34' 28'S., 54 m depth. Other material. Western Australia: NTM 21748: W. of Port Hedland, NW. Shelf, 19" 05.1 IS., 118"47.7/E., 84 m depth, 28.viii.1983, coll. T. Ward (CSIRO RV 'Soela' S04183, stn 119-NWS20, beam trawl). Substrate and Depth Range The NWS specimen was collected from a shallow offshore rock reef, and the species is known to occur in depths ranging from 54-84 m. Geographical Distribution This species appears to have a disjunct distribution, since it is presently known only from north-east of North Cape, New Zealand, and Northwest Shelf, W.A. (Fig. 32f). Further sampling in deeper offshore waters along the northern coast may show it to have a more widespread distribution. Description Shape. Growth form stipitate, arborescent, bifurcating digitate, with cylindrical or slightly flattened branches which taper to points at apex. Dimensions of NWS specimen: 135 mm long from stipitate base to branch tips, 19 mm basal stalk length, 5 mm basal stalk diameter, 12-46 mm branch length, 1 5-4 mm branch diameter. Colour. Live coloration yellow- or olive-brown (Munsell 2.5Y 6/8), changing to red-brown in ethanol. Oscula. No pores or oscula observed. Texture and sul3Face characteristics. Surface optically and microscopically very hispid, with points of extra-axial subectosomal styles protruding through ectosome. Texture of stalk and branches firm,barely compressible, branches flexible. Ectosome and subectosome. Ectosomal region with dense plumose brushes of ectosomal spicules, surrounding subectosomal megascleres at point where they pierce surface. Ectosomal spicules also intermingled with multispicular plumose brushes of acanthostyles perched on ends of extra-axial skeletal tracts. Extra-axial subectosomal skeleton sharply defined from axial choanosomal skeleton, and each component occupies approximately half of branch diameter. Subectosomal extra-axial tracts plumose, cored by 1-5 long styles, echinated by numerous acanthostyles. Acanthostyles heavily echinating at junction of axial and extra-axial skeletons and also occupying majority of plumose skeletal tracts. Mesohyl matrix in subectosomal region relatively heavy, composed of red-brown granular spongin together with dispersed ectosomal auxiliary megascleres and isolated arenaceous foreign particles. Choanosome. Choanosomal skeleton axially condensed. Axial core closely reticulate, with heavy spongin fibres cored by multispicular tracts of choanosomal styles, mostly orientated along longitudinal axis of branches, interconnected by pauci- or multispicular - J. N. A. Hooper Fig. 32. Raspailia compressa Bergquist (specimen NTM 21748): a, choanosomal axial style; b, subectosomal extra-axial style; c, ectosomal auxiliary anisoxeas; d, echiiating acanthorhabdostyles; e, section through peripheral skeleton; f, known Australian distribution. transverse skeletal tracts. Acanthostyles also echinate axial fibres but precise distribution difficult to detennine because whole region congested with megascleres. Megascleres (refer to Table 6 for measurements): Choanosomal axial spicules moderately short, thick styles, slightly curved towards base, less commonly straight, with tapering fusiform points and rounded non-tylote bases. Subectosomal extra-axial megascleres long, relatively thick styles, slightly curved near base, less commonly straight, with fusiform points rounded or very slightly subtylote bases. Ectosomal auxiliary spicules very thin and flexuous anisoxeas or true fusiform styles, with asymmetrical curvature, less commonly straight, bases oxeote or rounded-stylote. Echinating acanthostyles prominently rhabdose and smooth at basal end, curving at about 10-20% of spicule length and up to an angle of 45",with small granular spination on apical end, tapering to a sharp point. Microscleres absent. Remarks Despite their apparent biogeographical disparities, the Northwest Shelf specimen and the New Zealand holotype have identical morphological features, and on the basis of most Australian Raspailiidae Fig. 33. Raspailia compressa Bergquist: a, specimen (NTM 21748) (scale = 30 mm); b, SEM of skeletal structure; c, SEM of peripheral skeleton; d, peripheral skeleton (scale = 1 mm); e, SEM of echinating acanthostyles. skeletal characters, there is no doubt that they are conspecific. However, the Western Australian specimen differs significantly from the holotype in the length of echinating acanthostyles, which are half as long as in New Zealand material (Table 6), but there appears to be no justification for the erection of a new taxon for the present specimen for that feature. Coloration is also slightly different [recorded as bright yellow in life by Bergquist (1970)l. The affinities of this species, bearing echinating rhabdostyles, have been discussed at length by Bergquist (1970: 30), particularly the relationships with nominal raspailiid genera such as Aulospongus, Echinuxia and Raspaxilla. Those genera 1248 J. N. A. Hooper are discussed elsewhere in this study. This species should also be compared to other Australian species bearing rhabdose acanthostyles, R. fiondula and R. reticulata, described below. Table 6. Comparisons in spicule measurements between specimens of Raspailia compressa Bergquist Measurements are given in micrometres, presented as ranges (and means) of lengthxwidth, and represent 25 spicules measurements per specimen for each spicule category Material NMNZ POR 30A NTM 21748 A Choanosomal styles Subectosomal styles Ectosomal styles 240-360 x 6-10 (290~9.0) Holotype 1020-1400 x 16-20 (1120~ 18.5) 260-360 x2 (278 ~ 2 . 0 ) 220-360 ~6-5-9.0 (260 x 7.5) 234338 ~2-3.5 (283-6~2.7) 93-131 X 4-7 (112.8~5.5) 282-449 X 12-15 (343.0~13.3) Specimen 887-1151 X 17.5-24 (979.2~20-8) Echinating acanthostyles From Bergquist (1970: 30). Raspailia (Raspaxilla) frondula (Whiteleggge) (Fig. 34) Axinella frondula Whitelegge, 1907: 509-510, pl. 46, fig. 32. Echinaxiafiondu1a.-Hallmann, 1916a: 543; Hallmann, 1917: 394-398, text-fig. 1, pl. 21, figs 3 4 , pl. 22, figs 1-2. Raspailia frondu1a.-Bergquist, 1970: 30. Material Examined Holotype. AM G4349: Shoalhaven Bight, S. coast of N.S.W., 34" 511S., 150"45'E., 60 m depth (FIV 'Thetis', stn 50, dredge) (NMV sponge archives 14/25,26). Substrate and Depth Range Soft substrate, 60 m depth. Geographical Distribution Known only from the type locality in shallow coastal waters of southern N.S.W. (Fig. 34e). Description Shape. Stipitate, branching flabellate sponge (75 mm high, 48 mm breadth), with flattened basal attachment and short cylindrical stalk (25 mm long, 5-7 rnm diameter), from which arise thin, flattened lobate branches, wider at apex than at base (16-22 mm long, 8-13 mm breadth, 1-2 mm thick), with rounded apical margins (measurements from Whitelegge 1907; only a fragment of the holotype remains (Fig. 34f). Colour. Colour in spirit olive brown. Oscula. Not observed. Texture and suvace characteristics. Texture soft and flexible, and surface covered with fine close-set microconules. Ectosome and subectosome. Ectosomal skeleton lacks specialised spiculation, although peripheral extra-axial skeletal tracts form distinctive conulose surface structures. Extra-axial tracts perpendicular to surface, composed of prominently radial, non-plumose bundles of choanosomal (axial) styles for a large proportion of their length, extending up to 50 ,um from surface. Larger subectosomal extra-axial megascleres form plumose brushes around Australian Raspailiidae Fig. 34. Raspailia frondula (Whitelegge) (holotype AM G4349) (type species of the nominal genus Echinaxia Hallmann): a, choanosomal axial styles; b, subectosomal extra-axial style; c, echinating acanthorhabdostyle; d , section through peripheral skeleton; e, known Australian distribution; f , surviving portion of holotype (scale = 30 mm); g, skeletal structure (scale = 500 pm); h, peripheral skeleton (scale = 500 pm); i, echinating acanthostyle. 1250 J. N. A. Hooper terminal portions of extra-axial tracts, and scattered throughout reticulate axial skeleton. Extra-axial tracts without spongin fibres or transverse connecting spicules, but bundles of axial styles and plumose brushes of extra-axial styles which form extra-axial tracts are bound together by moderate quantities of light brown granular spongin. Echinating acanthostyles slightly more abundant in periphery than in basal regions of extra-axial skeleton, and also associated with region where larger subectosomal styles are embedded in extra-axial tracts. Choanosome. Choanosomal skeletal architecture only slightly axially condensed, but with well differentiated axial and extra-axial skeletons. Choanosomal axial core composed of very light spongin fibres, forming irregularly subrenieroid reticulate structures, and cored by pauci- or multispicular longitudinal tracts interconnected by mi- or paucispicular transverse fibres. Fibres cored by smaller choanosomal axial styles. Longitudinal multispicular axial fibres lightly echinated by rare acanthostyles, which are only slightly more common towards the periphery. Fibre anastomoses in axial region form irregular triangular, ovoid or rectangular meshes; mesohyl with light spongin and with large subectosomal extra-axial styles dispersed between major spicules tracts, orientated more-or-less longitudinally. Megascleres. Choanosomal axial spicules short, thin, slightly curved towards basal end, rarely straight or symmetrical, predominantly stylote but with occasional oxeote and strongylote forms present, with fusiform tips and rounded non-tylote bases [87-(123 .8)165 x 3 4 4 - 5)-8 pm]. Subectosomal extra-axial styles long, relatively thin, almost hastate, slightly curved towards basal end, with rounded non-tylote bases [278-(469 3)-710 x 4-(8 5)-16 pm]. Ectosomal megascleres absent. Echinating acanthostyles short, thin, with unspined and slightly swollen, rhabdose bases; distal 213 of spicule evenly covered with vestigial spines [72-(104.3)-133 x 4-(6.0)-9 pm]. Microscleres absent. - Remarks A more detailed description of R. ffondula is given by Hallmann (1917), and the above redescription is provided for comparative purposes. The only feature which consistently differentiates this species from many other raspailiids, like species of Raspailia (s.s.), Endectyon, Ectyoplasia, Trikentrion etc., is the geometry of echinating acanthostyles. Nevertheless, for reasons discussed above, R. fiondula is referred to Raspailia, following the arguments presented by Bergquist (1970), and the definition of that genus is expanded to include species beating echinating acanthostyles with smooth rhabdose bases (including R. compressa and R. reticulata). Raspailia ffondula could also be referred to the subgenus Clathriodendron in having a wide-meshed choanosomal reticulation, but there is greater differentiation of the axial and extra-axial skeleton in this species than in typical Clathriodendron. Nevertheless, the presence of both characters (reticulate skeleton and rhabdose acanthostyles) emphasises the artificial nature of the generic divisions Clathriodendron and Raspaxilla. Raspailia (Raspaxilla) reticulata, sp. nov. (Figs 35, 36) Material Examined Holotype. QM GL1982 (fragment NTM 21503): Green I., GBR, Qld, 16'46'S.. 145" 58'E., depth unknown, 10.x.1980, coll. QFS (shot 12-Don. 136, trawl). Substrate and Depth Range Unknown. Geographical Distribution Inter-reef region, northern Great Barrier Reef, Qld (Fig. 35f). Australian Raspailiidae Raspailia reticulata, sp. nov. (holotype QM GL1982): a, choanosomal axial styles; b, subectosomal extra-axial styles; c, ectosomal auxiliary styles; d, echinating acanthorhabdostyles; e, section through peripheral skeleton; f, known Australian distribution. Fig. 35. Description Shape. Arborescent, digitate sponge (66 mm high, 33 mm maximum breadth), with thin, flattened, frondose, bifurcated and anastomosing branches (12-23 mm long, 4 4 mm diameter), surmounted on a short cylindrical stalk (23 mm high, 5 mm diameter), with basal attachment encrusted on rubble. Colour. Live coloration unknown, dark brown in ethanol (Munsell 5YR 412). Oscula. Not observed. Texture and surface characteristics. Surface optically even and hispid, and microscopically conulose in preserved state. Texture firm and compressible. Ectosome and subectosome. Ectosome with heavy layer of dark brown granular spongin. Ectosomal skeleton with specialised spiculation of plumose brushes of thin styles or anisoxeas surrounding long protruding subectosomal extra-axial styles, not forming continuous palisade. Ectosomal brushes and protruding extra-axial spicules most common on raised surface conules. Some choanosomal megascleres protrude through surface also, particularly from peripheral axial skeletal tracts close to surface in vicinity of conules. Subectosomal extra-axial megascleres protrude through surface as individual spicules or in light bundles of 2-3. Some long extra-axial spicules also dispersed throughout reticulate axial skeleton. Choanosome. Choanosomal skeletal architecture barely axially compressed; axial and extra-axial regions poorly differentiated by presence or absence of individual subectosomal spicules in various regions of skeleton. Choanosomal axial skeleton composed of heavy dark brown spongin fibres (26-72 pm diameter), producing more-or-less regular open-meshed J. N. A. Hooper 1252 reticulation, with fibres forming oval or elongate meshes (110-340 pm diameter). Fibres slightly larger and fibre meshes slightly more compacted in axis than towards periphery, but this feature not well developed. Largest axial (=primary) fibres mostly longitudinal in branches; relatively large (=secondary) fibres more-or-less plumose and ascending towards surface; smaller paucispicular (=tertiary) transverse fibres interconnect secondary tracts. Spongin fibres cored by long thin choanosomal axial styles in pauci- and multispicular tracts: Mesohyl with relatively heavy dark brown granular spongin, and both choanosomal and subectosomal megascleres dispersed between fibres. Fibres heavily echinated by acanthostyles, some also incorporated into fibres. Megascleres. Choanosomal axial spicules short or long, thin, usually straight or slightly curved towards basal end, exclusively stylote with evenly rounded bases and fusifom sharply pointed tips [2884376.5)456 x 5 4 7 4)-11 pm]. Subectosomal extra-axial styles very long and thin, slightly curved, sometimes sinuous, with evenly rounded bases and hastate or stepped points [595-(700 4)-851 x 7412 0)-22 pm]. Ectosomal megascleres long and very thin styles or anisoxeas, with rounded bases or tapering hastate bases and raphidiform tips [184--(214 2)-258 x 1-(2.1)-3 pm]. Echinating acanthostyles short, thin, with unspined shaft, bases slightly rhabdose but sometimes straight. Spines vestigial, evenly dispersed over distal 213 of spicules [5 1 4 6 8 6)-84 x 2 - 5 4 3 .9)-5 - 5 pm]. Microscleres absent. - - - Remarks This species is close to R. frondula and R. compressa in the geometry of echinating rhabdose acanthostyles and other aspects of spicule geometry and skeletal architecture, and it could also be included in the nominal genera Echinaxia or Raspaxilla. Those genera differ from R. reticulata in a number of features: a flattened reticulated branching growth form; an open-meshed axial skeleton with poorly developed axial compression; very heavy spongin fibres that ascend all the way to the surface; remnants of an extra-axial skeleton, in the form of individual or groups of subectosomal styles poking through the surface, and poorly developed axial and extra-axial differentiation; and relatively long and slender choanosomal axial megascleres. Raspailia reticulata is included with the subgenus Raspaxilla (rhabdostyles) in preference to Clathriodendron (reticulate choanosomal skeleton), because the presence of echinating rhabdostyles is given higher priority within the genus Raspailia than skeletal structure. Nevertheless, this combination of features illustrates the difficulty of maintaining, and the artificial nature of, the subgeneric divisions between Raspailia, Raspaxilla and Clathriodendron. The status of Raspailia species with echinating rhabdostyles has been discussed above. Raspailia (Raspaxilla) wardi, sp. nov. (Figs 37, 38) Material Examined Holotype. NTM 21319: W. of Port Hedland, NWS, W.A., 19" 03-5'S., 119" 03.6'E., 81 m depth, 28.iv.1983, coll. T. Ward (CSIRO RV 'Soela' S02183, stn B12-NWS15, beam trawl). Substrate and Depth Range Rock reef, 8 1 m depth. Geographical Distribution Northwest Shelf, W.A. (Fig. 37e). Description Shape. Very thin elongate flabellate sponge (175 mm high, 48 rnm maximum breadth), with basal attachment and short cylindrical stalk (32 mm long, 4-5 mm diameter), from which arises single bifurcated, flattened, inverted wedge-shaped fan, 10 mm broad near stalk, expanding to 48 mm broad near apex. Margins of fan significantly thinner (0.5 mm Australian Raspailiidae Fig. 36. Raspailia reticulata, sp. nov.: a, holotype (QM GL1982, fragment NTM 21503) (scale = 30 mm); b, skeletal structure (scale = 1 mm); c, SEM of skeleton; d, SEM of extra-axial skeleton; e, SEM of echinating acanthostyle. thick) than central portions (1 - 5 mm maximum thickness), almost transluscent at margin, and margins only very slightly convoluted. Colour. Live coloration bright red-orange (Munsell 10R 5/12), orange-brown in ethanol. Oscula. Small oscula distributed along lateral margins, extending from just above stalk approximately up to point of bifurcation of fan (where thickness diminishes). In live sponge oscula up to 2 mm diameter, but not visible after preservation. Texture and s u ~ a c echaracteristics. Surface optically even, smooth, microscopically very hispid. Texture firm, barely compressible, flexible. Ectosome and subectosome. No specialised ectosomal megascleres, but ctosomal region with long subectosomal extra-axial megascleres protruding up to 900 pm from surface. Subectosomal extra-axial skeleton with long styles with bases embedded in peripheral spicule bundles; bundles composed exclusively of echinating acanthostyles perpendicular to axial skeleton, forming tightly plumose brushes producing continuous subdermal palisade. J. N. A. Hooper Fig. 37. Raspailia wardi, sp. nov. (holotype NTM 21319): a, choanosomal axial styles; b, subectosomal extra-axial style; c, echinating acanthostyles; d, section through peripheral skeleton; e, known Australian distribution. Ectosomal and subdermal spongin relatively heavy, darkly pigmented, slightly heavier than choanosomal mesohyl. Choanosome. Choanosomal axial skeleton not markedly axially condensed, but with well marked junction between axial and extra-axial skeletons. Axis consists of almost regular renieroid reticulation of relatively heavy spongin fibres; fibres cored by uni- or paucispicular tracts of choanosomal styles. Spongin fibres particularly heavy at nodes of skeletal tracts. Fibre reticulation forms triangular or rectangular meshes (25-55 p m diameter); choanocyte chambers oval. Choanosomal mesohyl matrix relatively light, with light brown pigmented granular spongin. Echinating acanthostyles absent from axial skeleton, occurring only as perpendicular palisade on periphery of axis. Megascleres. ~hoanosomalaxial megascleres predominantly styles, rarely anisoxeas, usually slightly curved towards basal end, sinuous, straight, or occasionally rhadiferous, with slightly fusiform points and rounded non-tylote bases [147-(182.4)-222 x 5 4 6 7)-9 pm]. Subectosomal extra-axial megascleres long, thick styles, slightly curved towards basal end or occasionally straight, with almost hastate points and rounded non-tylote bases [515-(838-3)-1122~7-(11- 1)-16 pm]. Ectosomal auxiliary megascleres absent. Australian Raspailiidae Fig. 38. Raspailia wardi, sp. nov.: a, holotype (NTM 21319) (scale = 30 mm); b, echinating acanthostyles (scale = 100 pm); c, section through skeleton (scale = 1 mm); d, SEM of skeletal structure; e, SEM of junction between axial and extra-axial skeletons; f, SEM of axial skeleton; g, SEM of spine morphology on echinating acanthostyle. Echinating acanthostyles typically slightly rhabdose or merely slightly curved towards basal end, with fusiform points bearing small evenly distributed recurved spines, but spines absent from base [112-(131.4)-165 x 8-(10.3)-12 vm]. Microscleres absent. Remarks This species is well characterised by its extremely thin flabellifonn growth form (ranging from only 0 5-1 - 5 mm in thickness), renieroid axial fibre and spicule skeleton, and its distinctive surface features consisting of echinating acanthostyles (or acanthorhabdostyles) - J. N. A. Hooper 1256 forming a continuous subectosomal palisade just below the ectosome. This feature is reminiscent of the genus Endectyon (see below), and to a lesser extent Aulospongus, but these differ from R. wardi in the morphology of echinating megascleres and specialised growth forms. Irrespective of the unusual distribution of echinating megascleres, the species is included in Raspailia on the basis that species such as R. compressa are also known to have acanthostyles restricted to the extra-axial region. Etymology This species is named in appreciation of Dr Trevor Ward, CSIRO Fisheries, who collected the holotype and greatly assisted the author in studying the sponge fauna of the Northwest Shelf of W.A. Subgenus Syringellu, sensu Ridley Raspailia (Syringellu) australiensis Ridley (Figs 39, 40; Table 7) Raspailia (Syringella) australiensis Ridley, 1884: 460.-Pick, Homuxinella austra1iensis.-Burton, 1934b: 42. 1905: 18, 35; Vosmaer, 1912: 316. Material Examined Lectotype (here designated). BMNH 1882.2.23.253: Port Darwin, N.T., 13-21 m depth, coll. R.W. Coppinger (HMS 'Alert'). Paralectotype. BMNH 1882.2.23.254: same locality. Other material. Queensland: QM GL850 (fragment NTM 21501): Near Upolu Cay, o f f Cairns, Great Barrier Reef, Qld, 16' 43.0's.. 146' 04.3'E., depth unknown, 10.01.1981, coll. QFS (stn B1 Cairns Ground Truth Survey). QM G5200 (fragment NTM 21586): 1 km W. o f Macleay I., Moreton Bay, Qld, 27" 35.3'S., 153' 21 .OIE., 7 m depth, 13.x.1967, coll. R. McKay et al. (stn 365). Substrate and Depth Range 7-21 m depth, mud and sand substrate. Geographical Distribution Darwin, Arafura Sea, N.T., inter-reef region, northern section Great Barrier Reef, and Moreton Bay, southern Qld (Fig. 39e). Description Shape. Unbranched cylindrical digits, or with 1 or more bifurcations near base (104165 mrn long), with short holdfast attachments and long slender woody stalks (9-35 mm long, 2-4 mm diameter), with 1 or more branches arising from stalk (1 branch: holotype; 2 branches: QM GL850; 14 branches: QM G5200); branches thicker in diameter than stalk (4-9 mm maximum), but tapering to fusiform rounded margins. Colour. Live coloration unknown, beige-white in ethanol. Oscula. Not seen. Texture and surjace characteristics. Texture firm, flexible, with woody central stem (=axial skeleton) and fleshy branches (produced by extra-axial region). Surface optically even, but microscopically rugose and hispid, with minute conules and ridges. Ectosome and subectosome. Ectosomal skeleton with sparse brushes of ectosomal styles surrounding brushes of extra-axial subectosomai megascleres at point of insertion into surface. Ectosomal mesohyl heavily invested with lightly pigmented spongin. Extra-axial spicule tracts in thick bundles, arising from axial skeleton at tangential and oblique angles. Extra-axial portion of skeleton makes up only small proportion of branch diameter (20-40%). Extra-axial tracts composed of plumose brushes of styles or anisoxeas, embedded in axial skeleton and protrude only a relatively short distances through surface. Choanosome. Axial core fasciculated, composed of discrete but closely set bundles of choanosomal styles usually running longitudinally through branches but also at more oblique angles. Compressed axial core comprises majority of branch diameter (60-80%), Australian Raspailiidae 1257 producing solid but flexible skeleton. Spicule bundles in axial and extra-axial skeletons not associated with spongin fibres, and axis poorly invested with collagenous spongin. Megascleres (refer to Table 7 for measurements): Three types of structural megasclere have very similar geometry, differing from each other mainly by size and location within skeleton. Choanosomal (axial) megascleres moderately long, straight or slightly curved styles, with evenly rounded or tapering fusiform bases and tapering to sharp points. Subectosomal (extra-axial) megascleres long, thick, relatively straight styles, with fusiform tapering bases and sharply pointed tips. Ectosomal (auxiliary) megascleres range from thin whispy flexuous forms with anisoxeote ends, to thicker straight or slightly curved styles. Echinating megascleres absent. Microscleres absent. Fig. 39. Raspailia australiensis Ridley (lectotype BMNH 1882.2.23.253): a, choanosomal axial style; 6,subectosomal extra-axial style with tapering and rounded bases; c, ectosomal auxiliary styles; d, section through peripheral skeleton; e, known Australian distribution. Remarks The material described above from the Cairns inter-reef region and Moreton Bay is the first record for the species since it was originally described, and also a new record for Queensland waters (Solanderian province). In growth form, surface features, skeletal structure and spicule geometry all three specimens described above are quite close, although the Queensland material has marginally larger megascleres than the type material (Table 7). Despite extensive recollections from the type locality (Port Darwin), R. australiensis has not been rediscovered in N.T. coastal waters, but it is possible that the species exists only in the deeper tidal-scoured channels along the northern coastline, which have not yet been fully explored. 1258 J. N. A. Hooper Raspailia australiensis Ridley: a, paralectotype (BMNH 1882.2.23.254) (smaller specimen) and lectotype (BMNH 1882.23.253) (larger specimen) (scale = 30 mm); b, specimen (QM 850) (scale = 30 mm); c, peripheral skeleton; d, SEM of skeletal structure; e,f , SEM of axial skeleton. Fig. 40. Raspailia australiensis is characterised by its thin cylindrical growth form resembling a gorgonian (Junceela) whip-coral, the fleshy barely hispid surface, the absence of a discernible fibrous skeleton, the fasciculate nature of both the longitudinal axial skeletal spicule bundles and radial extra-axial plumose brushes (similar to R. cervicornis (Burton 1959) from the Arabian Gulf, holo- and paratypes BMNH 1936.3.4.604, 521, 522), and the close geometric similarity between each of the three spicule types, which differ only in size and location within the skeleton. Ridley (1884) and subsequent authors who examined the type material of this species (i.e. Vosmaer 1912; Burton 19343) appear to have overlooked the presence of a third category of spicules, viz. the ectosomal auxiliary styles/anisoxeas. This is surprising since these spicules are relatively common, although they do not form prominent surface brushes like those of many other Raspailia species. Burton (1934b: 42) referred R. australiensis to the genus Hornaxinella in the Axinellidae, but similarities between those two are merely superficial, being based on gross skeletal Australian Raspailiidae 1259 construction, and the affinities of this species clearly lie with Raspailiidae. The type species of Hornaxinella (i.e. Axinella supraturnescens Topsent, 1907, from Antarctica, lectotype MNHN LBIM DT1660, largest of several specimens labelled no. 47 in the same bottle, here designated) has a condensed reticulate axial core of large choanosomal styles running longitudinally through branches, with a radial plumose extra-axial skeleton also composed of choanosomal styles running transversely to the surface of branches. The presence of styles forming tufts on the ectosome was supposed by Burton to be homologous with the specialised raspailiid ectosomal skeleton, but these merely represent a hispid surface crust, similar to that found in Clathria (Thalysias) of the Microcionidae. This surface structure may be merely an analogue, and it does not necessarily imply a close relationship between Hornaxinella and this obvious raspailiid species. Table 7. Comparisons in spicule measurements between specimens of Raspailia australiensis Dendy Measurements are given in micrometres, presented as ranges (and means) of lengthxwidth, and represent 25 spicules measurements per specimen for each spicule category Material Choanosomal styles BMNH 1882.- 232-365 x 6-9 (313.0~7.8) Subectosomal styles Ectosomal styles/ anisoxeas Echinating acanthostyles 145-237 ~0.5-5 (207.5~2.1) Absent 96-245 ~0.5-5 (222.0~3.5) Absent 126-304 ~0.5-5 (268.1 ~ 2 . 8 ) 218-392 x 1-3 (272.4~2.1) Absent Lectotype 2.23.253 543-964 x 9-25 ( 7 5 5 . 5 ~16.2) Paralectotype BMNH 1882.2.23.254 297404 X 5-9 (356.6~6.5) 514-1055 x 11-22 (789.9~14.8) Specimens QM GL850 QM G5200 374524 x 5-8 (452.4~6.8) 287464 x 4-8 (356.3~6.0) 676-1034 x 9-22 ( 9 4 8 . 2 ~16.1) 686-1355 X 11-17 ( 9 6 8 . 4 ~13.3) Absent The species should be contrasted with R. clathrata from Torres Strait, which also lacks echinating megascleres and any obvious hispid surface. That species is grey, with laterally flattened branches and with spicule dimensions and spicule geometry that differ significantly from R. australiensis. It too was referred to Hornaxinella by Burton (1934b). Raspailia flagelliforrnis Ridley & Dendy (1886: 482, 1887: 190) from the Cape of Good Hope also has only a minutely hispid surface and lacks echinating spicules, but is yellowish grey, has a more disorganised, less obviously longitudinally directed axial skeleton, and lacks a specialised ectosomal skeleton or ectosomal spicules. Raspailia rigida Ridley & Dendy (1886: 483, 1887: 191), also from the Cape of Good Hope, is probably closest to R. australiensis, lacking echinating megascleres and any trace of a fibrous skeleton. It has a well differentiated axial core and extra-axial plumose skeleton, but extra-axial tracts form divergent bundles of spicules which protrude a long way through the surface, specialised ectosomal spicules are absent, and spicules are nearly twice as long as those of R. australiensis. J. N. A. Hooper 1260 Raspailia (Syringella) clathrata Ridley (Fig. 41) Raspailia (Syringella) clathrata Ridley, 1884: 461-2, pl. 41, fig.f.-Pick, 1905: 18, 35; Vosmaer, 1912: 316. Homaxinella c1athrata.-Burton, 1934b: 42. Material Examined Holotype. BMNH 1882.2.23.335: Thursday I., Torres Strait, Qld, vicinity of 10' 35'S., 142' 13'E., 12-22 m depth, coll. R.W. Coppinger (HMS 'Alert'). Substrate and Depth Range Bottom sand, 12-22 m depth. Geographical Distribution Torres Strait, Qld (Fig. 414. Description Shape. Stipitate reticulate planar branching sponge (115 mm high, 85 mm maximum width), with short stalk (14 mm long, 4 mm diameter) and dichotomously divided branches which narrow towards apex, and mostly anastomose with adjacent branches. Colour. Pale grey in the preserved state. Oscula. Not observed. Texture and surface characteristics. Stalk woody, branches firm and barely compressible, but branch tips more flexible. Surface uneven but not hispid, with minute ridges running longitudinally and laterally joining to form conules and depressions. Ectosome and subectosome. Ectosomal skeleton a membraneous heavy layer of darkly pigmented granular spongin, lacking specialised spiculation. Brushes of subectosomal extra-axial spicules protrude through this region; brushes consist of compressed bundles of large styles, arising from axial core at oblique or right angles to surface, and forming tight palisade of dermal spicules. Choanosome. Choanosomal fibres and spicule bundles form reticulate slightly condensed axial skeleton, running both longitudinally and transversely through branches. Choanosomal fibres heavily invested in spongin, heavily cored by smaller choanosomal styles; fibre anastomoses form oblong or oval meshes (up to 290 pm diameter). Spongin in mesohyl abundant but lightly pigmented. Echinating megascleres absent. Megascleres. Choanosomal axial megascleres thin, relatively short, usually slightly curved, tapering to sharp points or sometimes stepped, with rounded non-tylote bases [155-(237.8)-348 x 2-(3 1)-5 pm]. Subectosomal extra-axial styles long, relatively thick, straight or slightly curved, tapering to sharp points, with rounded bases [581-(684 4)-832 x 4-(8 -3)-12 pm]. Ectosomal auxiliary spicules absent. Echinating megascleres absent. Microscleres absent. - - Remarks As for R. australiensis Ridley, Burton (1934b) transferred R. clathrata to Homaxinella, but similarities between R. australiensis and Homaxinella (s.s.) are superficial and based solely on the structure of the axial and extra-axial skeletons (see remarks for R. australiensis). The affinity of R. clathrata in the Raspailiidae is less obvious since it lacks a specialised ectosomal skeleton completely, lacks echinating megascleres, and has an axial skeleton more reticulate than condensed. Nevertheless, it is obviously closely related to R. australiensis which is clearly a raspailiid sponge. Australian Raspailiidae Fig. 41. Raspailia clathrata Ridley (holotype BMNH 1882.2.23.335): a, choanosomal axial styles; b, subectosomal extra-axial styles; c, section through peripheral skeleton; d, known Australian distribution; e, holotype (scale = 30 mm); f, extra-axial skeleton (scale = 1 mm). 1262 J. N. A. Hooper Raspailia (Syringella) elegans (Lendenfeld), comb. nov. (Fig. 42) Antherochalina elegans Lendenfeld, 18876: 787, pl. 22, fig. 40. Syringella e1egans.-Burton, 1934a: 558. Material Examined Holotype. BMNH 1886.8.27.452: Torres Strait, Qld, depth and date of collection unknown (schizotype AM G3461, slide). Other material. Northwest Shelf, W.A.: NTM 21798: W. of Port Hedland, 19" 052-06'S., 119' 0-09'E., 85 m depth, 29.viii.1983, coll. T. Ward, CSIRO RV 'Soela', trawl (stn S04183-125). NTM 2715, 728: N. of Adele I., Collier Bay, 15' 58 - 03 IS., 122'239 - 07' E., 59 m depth, 21.iv.1982, coll. CSIRO RV 'Sprightly' (stn SP4/82-40). NTM 2688: W. of Buccaneer Archipelago, 16' 20-OIS., 120' 10.OIE., 35 m depth, 28.iv.1982, coll. CSIRO RV 'Sprightly' (stn SP4182-39). NTM 23337, 3341: Fringing reef, Direction I. National Park, 21' 32 .0' S., 115' 07 -2'E., intertidal, 24.viii.1988, coll. D. Low Choy, by hand (stn NWS61). Substrate and Depth Range On fringing coral reef or deeper offshore reefs, 0-85 m depth, in rubble or shell grit substrata. Geographical Distribution North-west Australia, W.A., and Torres Strait, Qld (Fig. 42e). Description Shape. Regularly symmetrical, planar or biplanar fan (155-190 mm high, 114-140 mm maximum breadth, up to 5 mm thick), with rounded even margins, basal attachment and short cylindrical stalk (13-22 mm long, up to 9 mm diameter). Colour. Colour alive is pale orange (Munsell 5YR 5/8), and in preserved state grey-brown. Oscula. Numerous small pores scattered over surface, up to 1 5 mm diameter, predominant near base of fan. Texture and sugace characteristics. Texture of fan firm, barely compressible, whereas stalk more rigid. Surface even, not optically hispid, with prominently branching stellate subdemal striations, grooves and drainage canals, becoming radial towards margins of fan Ectosome and subectosome. Ectosomal skeleton raised into conules; conules produced by protruding peripheral choanosomal fibres, associated with sparse brushes of choanosomal oxeas erect on surface. Subectosomal extra-axial styles protrude through choanosomal brushes singly or in bundles. Brushes of ectosomal oxeas grouped in vicinity of extra-axial spicules, although not regularly so, lying tangential or erect on surface. Choanosome. Choanosomal axial skeleton not compressed but more-or-less regularly reticulate, with few dense spongin fibres concentrated in axis and running longitudinally through fan; fibres cored by multispicular tracts of choanosomal oxeas. Other fibres and spicule tracts radiate towards surface and anastomose producing more-or-less regular reticulation. Fibre diameter not markedly different between central and peripheral regions, although fibres in core composed of heavier spongin; tendency for slightly thicker fibres to radiate from axial spongin core towards periphery, interconnected by slightly thinner transverse fibres. Fibre anastomoses produce squarish or ovoid meshes; mesohyl matrix contains very light spongin. Subectosomal extra-axial styles also dispersed throughout axial skeleton. Echinating megascleres absent. Megascleres. Choanosomal axial megascleres relatively long, slender oxeas, symmetrically curved, tapering to sharp points 1176-(232 -5)-275 x 4 4 7 7)-11 pm]. Subectosomal extra-axial styles very long, thin, usually curved towards basal end, occasionally sinuous, with evenly rounded bases, tapering to raphide-like points [478(645 .5)-830 x 2 . 5 4 3 6)-5 pm]. Ectosomal auxiliary oxeas short, thin, symmetrically curved or straight, tapering to sharp points [lo84137 8)-164 x 1 - 5 4 2 -4)-3 5 pm]. Echinating megascleres absent. Microscleres absent. - - - Australian Raspailiidae Fig. 42. Raspailia elegans (Lendenfeld) (holotype BMNH 1886.8.27.542): a, choanosomal axial oxeas; b, subectosomal extra-axial styles; c, ectosomal auxiliary oxeas; d, section through peripheral skeleton; e, known Australian distribution;f,holotype (scale = 30 mm); g, skeletal structure (scale = 1 mm). Remarks This species is unrecognisable from Lendenfeld's (1887b) original description, which also incorrectly described the geometry of megascleres as styles and strongyles instead of oxeas and styles respectively. Burton's ( 1 9 3 4 ~ brief ) 'revision' of Antherochalina was more or less accurate in placing this species with the raspailiids (viz. Syringella), and although not 1264 J. N. A. Hooper stated as such, it must be concluded that his revised diagnosis was based on re-examination of the BMNH holotype cited above. Raspailia elegans is a borderline case between Echinodictyum and reticulate species of Raspailia (viz. nominal genus Clathriodendron). The only evidence of an extra-axial skeleton is the presence of long whispy subectosomal styles dispersed throughout the periphery and core of the reticulate axial skeleton. However, these subectosomal spicules do not appear to be exclusively localized in their distribution, unlike most other raspailiids. Similarly, the remnants of an ectosomal skeleton appear to be present and vaguely associated with extra-axial styles, but their distribution on the ectosome is not typical of Raspailia (i.e. not grouped on the surface around protruding extra-axial spicules). Raspailiu (Syringella) nuda Hentschel (Figs 43, 44, 109d; Table 8) Raspailia (Syringella)nuda Hentschel, 1911: 383, text-fig. 52. Raspailopsis nu&.-Burton, 1959: 256. Material Examined Holotype. ZMH 4450 (not seen): NNE. Heirisson Prong, Denham Channel, Shark Bay, W.A., 25'57IS., 113' lglE., 22-25 m depth, 18.vi.1905, coll. W. Michaelsen & R. Hartmeyer, (Hamburg Expedition, sm 15, dredge). Other material (all material collected by the author using SCUBA, unless otherwise indicated). Darwin Region, N.T.: NTM 2943: EPMFR, 12" 24.5IS., 130" 48. OIE., 10-12 m depth, 13.ix.1982, (stn EP9). NTM 22043: same locality, 12" 25.O1S., 130" 48-4'E., 6-10m depth, 10.v.1984, coll. R.S. Williams (stn EP14). NTM 22083: same locality, 12" 24.5'S., 13Oo48.0'E., 7-10m depth, 20.vii.1984 (stn EP15). NTM 22244: same locality, 10 m depth, 12.iv.1985, coll. C. Hood & J.R. Hanley (stn EP22). NTM 22394: same locality, 8 m depth, 29.vii.1985 (stn EP23). Substrate and Depth Range Coral rubble, sand and gravel, overlaying rock substrate, with depth range from 6-12 m. Geographical Distribution Central western and north-west coasts of Australia (Fig. 43e). Description Shape. Stipitate, arborescent digitate sponges (51-135 mm high, 24-150 rnm maximum breadth), with basal holdfast and relatively long cylindrical stalk (14-26 mm long, 8-17 mm diameter) and numerous long, evenly cylindrical branches (21-66 mm long, 3-9 mm diameter) bifurcating regularly and branching in more than one plane; branches taper to rounded points. Colour. Orange-brown alive (Munsell 5YR 6/10) (Fig. 109d), brown in ethanol. Oscula. Not seen. Texture and surface characteristics. Texture firm, barely compressible, branches easily flexible. Surface even and prominently hispid, but hispidating spicules do not extend far from surface. Ectosome and subectosome. Ectosomal skeleton with dermal oxeas or anisoxeas forming tight bundles, rarely plumose brushes, surrounding protruding subectosomal extra-axial styles. Ectosomal region heavily pigmented with dark brown granular spongin. Long subectosomal styles protrude through surface, with bases embedded just below ectosome, in ends of ascending extra-axial plumose tracts. Extra-axial tracts composed of subectosomal styles in radial arrangement, originating from condensed axial core, overlaid or intermingled by reticulate skeleton of choanosomal oxeas, which extends up to ectosome. Choanosome. Choanosomal skeleton clearly differentiated into two regions: axial core condensed, occupying only small proportion (15-30%) of branch diameter; tightly reticulate, consisting of close-set heavy spongin fibres cored by choanosomal oxeas or anisostyles. Extra-axial skeleton, div-erging from axis at right angles, more-or-less regularly plumose (composed of subectosomal styles) and reticulate (formed by choanosomal oxeas). Reticulate component of peripheral skeleton has multispicular tracts of oxeas ascending to surface, interconnected by single or paucispicular transverse reticulate tracts of oxeas. Mesohyl matrix contains relatively heavy and granular spongin. Echinating megascleres absent. Australian Raspailiidae a Fig. 43. Raspailia nu& Hentschel (specimen NTM 22244): a, choanosomal axial oxea; b, subectosomal extra-axial style; c, ectosomal auxiliary oxea; d, section through peripheral skeleton; e, known Australian distribution. Megascleres (refer to Table 8 for dimensions). Choanosomal axial megascleres true oxeas, only rarely with styloid modifications, straight or slightly curved at centre, typically symmetrical, with sharply pointed fusiform tips. Subectosomal extra-axial megascleres long styles, thin, curved and setaceous (dispersed within mesohyl), or thick, mostly straight, with rounded non-tylote bases and fusiform points. Ectosomal auxiliary megascleres typically oxeote, symmetrical, slightly curved centrally, with fusiform points, and with fewer anisoxeote asymmetrical forms. Echinating megascleres absent. Microscleres absent. Remarks This species illustrates the unsuitability of classifying skeletal components solely on the basis of their location within the skeleton (i.e. axial, extra-axial, and ectosomal megascleres), without reference to their derivation (i.e. principal and auxiliary megascleres). The presence 1266 J. N. A. Hooper Fig. 44. Raspailia nuda Hentschel: a, specimen (NTM 22394) (scale = 30 mm); b, skeletal structure (scale = 1 mm); c, d, SEMs of the peripheral and ectosomal skeletons; e, SEM of specialised raspailiid ectosomal skeleton. of an axial core (composed exclusively of choanosomal oxeas), an extra-axial skeleton with both plumose (subectosomal styles) and reticulate components (choanosomal oxeas), a peripheral extra-axial skeleton (composed exclusively of subectosomal styles protruding through the surface), and an ectosomal component (composed of small auxiliary oxeas) is impossible to describe in terms of axial and extra-axial skeletons alone. There is no doubt that the Northern Territory material recorded above is conspecific with Hentschel's (1911) species, having the same growth form, skeletal architecture and similar spicule geometry. Although the ZMH accession number has been traced (F. Knobbe, personal communication, through E Wiedenmayer), it has not yet been possible to borrow material from the HM for re-examination. There are some differences in spicule dimensions between these materials (Table 8), but these are certainly not significant and do not warrant the creation of a new taxon for the N.T. material. Hentschel (1911) reported that dermal spicule bundles in Shark Bay material consisted of ectosomal styles; however, most of these spicules in N.T. material are oxeas or anisoxeas. Australian Raspailiidae 1267 Table 8. Comparisons in spicule measurements between specimens of Raspailia nu& Hentschel Measurements are given in micrometres, presented as ranges (and means) of lengthx width, and represent 25 spicules measurements per specimen for each spicule category Material Choanosomal oxeal anisoxeas ZMH 4450A 488-640 X 9-14 Subectosomal styles Ectosomal oxeast anisoxeas Echinating spicules 256-304 ~2 Absent Holotype 1160-1560 ~9-16 Specimens NTM various (n = 5) A 260-488 x 7-24 ( 3 9 2 . 7 ~14.8) 820-1673 X9-15.5 (1151 1 x 12.2) 243472 x 1 .5-7 (346.3~3.6) Absent From Hentschel (1911: 384), Burton (1959) referred R, nuda to Raspailopsis since it lacked echinating spicules, had an axially condensed axial skeleton, plumose extra-axial skeleton, and a specialized ectosomal skeleton. He suggested further that R. nuda and R. cervicornis were very similar, showing significant differences only in the location of small styles and oxeas/anisoxeas (in the extra-axial skeleton and plumose ectosomal brushes in R. cervicornis, and supposedly reversed in R. nuda). This assumption is incorrect and the two species differ in many other respects also (size and geometry of megascleres, extra-axial skeletal construction, etc.). Raspailia (Syringella) stelliderma (Carter), comb. nov. (Fig. 45; Table 9) Axinella stelliderma Carter, 1885: 360.-Dendy, 1897: 232; Bergquist, 1970: 15. Axinella stelliderma var. acerata Carter, 1885: 360-361. Axinella acerata.-Dendy, 1897: 233. Material Examined Lectotype (here designated). BMNH 1886.12.15.33: Port Phillip Heads, Vic., 20 m depth, coll. J.B. Wilson (schizotypes AM G2789 slide, NMV sponge archives 35/21, MNHN LBIM DCL 182 slide). Paralectotype. BMNH 1886.12.15.111: same locality. Other material. Type of var. acerata -BMNH 1886.12.15.63: Port Phillip, Vic., 22 m depth, coll. J.B. Wilson (2 specimens). Substrate and Depth Range Substrate unknown, 20-40 m depth. Geographical Distribution Port Phillip Heads, Vic. (Fig. 45e). Description Shape. Stipitate, branching growth form (104-169 mm long, 58-110 mm maximum breadth), with enlarged basal holdfast, moderately long, thick cylindrical stalk (33-60 mm long, 9-20 mm diameter), numerous thick, lobate, cylindrical branches arising from apex (21-94 mm long, 4-16 mm diameter); branches bifurcate 1-2 times, some fused and taper to stubby digitate rounded margins. Colour. Live coloration reportedly variable, ranging from 'rich maroon-red mottled with lighter shade' to 'dull ochre-yellow' @endy 1897); colour in preserved state beige. Oscula. Not observed. J. N. A. Hooper Fig. 45. Raspailia stelliderma (Carter): a, choanosomal axial styles; b, subectosomal extra-axial styles; c, ectosomal auxiliary styles/anisostrongyles; d, section through peripheral skeleton; e, known Australian distribution; f, lectotype (BMNH 1886.12.15.33) (scale = 30 mm); g, ectosomal skeleton (scale = 500 .urn); h, skeletal structure (scale = 1 mm). Australian Raspailiidae 1269 Texture and sulface characteristics. Texture compressible, fleshy, flexible, with firm axis. Surface optically even, not hispid, but with minute stellate subdemal drainage canals (now barely visible), once a prominent feature of surface (Carter 1885) for which the species was named. Ectosome and subectosome. Ectosomal skeleton membraneous in most places, with heavy layer of smooth spongin, at regular intervals on surface (80-150 pm apart) are plumose ectosomal spicule brushes protruding only a short distance through ectosome. Each bundle of ectosomal spicules formed by a discrete set of styles, each associated with larger styles from plumose extra-axial tracts, but larger extra-axial styles usually broken off close to surface. Peripheral extra-axial skeleton composed of plumose or simply radial brushes of long styles; extra-axial tracts sometimes associated with spongin fibres but sometimes simply tracts of spicules, most ascending to surface in sinuous tracts. Subectosomal extra-axial skeleton occupies most (70-80%) of branch diameter. Spongin in mesohyl relatively heavy but lightly pigmented. Choanosome. Choanosomal skeleton axially condensed, with well differentiated axial and extra-axial components. Axial core a close-meshed reticulation of heavy spongin fibres, divided into primary (multispicular, running longitudinally through branches) and secondary elements (uni-, pauci- or aspicular, running transversely through branches, interconnecting primary fibres). Fibres cored by thin styles of same diameter but marginally shorter than extra-axial spicules. Echinating megascleres absent. Table 9. Comparisons in spicule measurements between specimens of Raspailia stelliderma (Carter) Measurements are given in micrometres, presented as ranges (and means) of lengthxwidth, and represent 25 spicules measurements per specimen for each spicule category Material Choanosomal styles/ anisoxeas BMNH 1886.- 325-528 X 3-5 (407.2~4.3) Subectosomal styles/ anisoxeas Ectosomal styles/ anisoxeas Echinating megascleres Lectotype 12.15.33 434-714 x4 9 (583.8~508) 345-45 1 ~ 2 . 5 4 (392.3 ~ 3 . 4 ) Absent 205-405 x 2-6 (327.3~3.3) Absent 285-439 x 3-6 (341.1~3.9) Absent Paralectotype BMNH 1886.12.15.111 280-595 x 2-6 (355.5~4.9) 415-811 x4 9 (576-4~6.5) Specimen BMNH 1886.12.15.63 252-5 15 x 3-6 (392.8~4.4) 364-672 ~5.5-10 (549.0~7.6) Megascleres (refer to Table 9 for dimensions). Choanosomal axial spicules long, thin, slightly curved styles or anisoxeas, with tapering points and tapering hastate bases. Subectosomal extra-axial spicules long, relatively thick styles, rarely anisoxeas, sometimes stongylote, with rounded bases and abruptly pointed hastate tips, or points may be evenly rounded. Ectosomal auxiliary styles or anisoxeas thin, short, straight or slightly curved at centre, with rounded or tapering hastate bases, hastate points. Echinating megascleres absent. Microscleres absent. Remarks Both Dendy (1897) and Bergquist (1970) remarked on the similarity in skeletal structure between A. stelliderma and Raspailia. Berquist did not transfer the species from Axinella, but she did remark that the presence of a central extra-axial spicule surrounded by brushes of ectosomal spicules, as described in this species by Carter (1885) and noted by Dendy J. N. A. Hooper 1270 (1897), was not a feature of Axinella but of Raspailia. This species is formally transferred here to Raspailia, but its affinity with Raspailia s.s. is not completely obvious, since its spicule composition consists solely of long styles, which are occasionally anisoxeote, without clear differentiation between axial, extra-axial and ectosomal megascleres. The differentiation of A. stelliderma and A. stelliderma var. acerata at the species level, based on one having styles and one with oxeas as proposed by Dendy (1897), is not supported here. Both morphotypes appear to be identical, at least from the material examined above, and both have examples of styles and anisoxeas; neither have true oxeote spicules. Subgenus Hyrneraphiopsis, subg. nov. m e species: Raspailia irregularis Hentschel, 1914: 121. Diagnosis. Raspailia with echinating megascleres having swollen tylote bases. Raspailia (Hymeraphiopsis) irregularis Hentschel (Fig. 46; Table 10) Raspailia irregularis Hentschel, 1914: 121, pl. 8, fig. 6.-Boury-Esnault & van Beveren, 1982: 50, text-fig. 12f-h. Eurypon miniaceurn.-Burton, 1932: 325; 1934b: 34; Koltun, 1964: 84, pl. 13, figs 1-3; 1976: 192. Not Eurypon miniaceum Thiele, 1905: 446-447, fig. 64a-f. Material Examined 'Syntypes'. ZMB 4832 (not seen): Gauss-Station, Antarctica, 385 m depth, 14.vil9.vii.1902. Other material. Antarctica: BMNH 1928.2.15.244: South Georgia, 53" 51 S., 36' 21 .5'W., 200-236 m depth, 20.i.1927 (RRS 'Discovery', stn 156). SAM S537 (TS 9727): MacRobertson Land, AAT, 66' 2S1S., 72' 411E., 1266 m depth, 25.xii.1929 (RRS 'Discovery', stn 29). SAM S538 (TS 9726): MacRobertson Land, AAT, 66" 4S1S., 71" 24/E., 540 m depth, 27.xii.1929 (RRS 'Discovery', stn 30). Substrate and Depth Range Rock and mud substrate, 35-1266 m depth. Geographical Distribution Antarctic and subantarctic region (Fig. 46e). Description Shape. Small bushy branching sponge (38 mm high, 40 rnrn maximum breadth), with multiple points of attachment to rock substrate, several basal holdfasts and short stalks (up to 7 mm long, 3 mm diameter), from which extend numerous flattened irregularly bifurcate and bushy branches (3-14 mm long, 2-6 mm diameter), with irregular apical margins. Colour. Live coloration unknown, beige-brown in ethanol. Oscula. Not observed. Texture and surface characteristics. Surface of branches aculeate, irregularly conulose and prominently hispid. Texture preserved is compressible and flexible. Ectosome and subectosome. Ectosomal skeleton with paucispicular brushes of ectosomal styles or anisoxeas, dispersed over surface and intermingled with erect brushes of acanthostyles just below ectosome; brushes not associated with protruding subectosomal extra-axial spicules. Ectosome a heavy granular spongin layer. Subectosomal extra-axial megascleres embedded in axial skeleton, protruding up to 2 mm from surface, and dominating skeletal architecture so much that axial skeleton simply resembles a cluster of spicules surrounding bases of subectosomal megascleres. Choanosome. Choanosomal axial skeleton a series of condensed axial clusters of echinating megascleres, without fibres or any specialized choanosomal megascleres, united by granular spongin and forming plumose brushes; echinating spicule brushes near periphery Australian Raspailiidae 1271 directed outwards, and points of acanthostyles protrude up to but not through ectosome. Bases of subectosomal extra-axial spicules embedded in core of axial skeleton. Megascleres (refer to Table 10 for measurements). Choanosomal megascleres of axial skeleton absent. Fig. 46. Raspailia irregularis Hentschel (specimen BMNH 1928.2.15.244): a, subectosomal extraaxial style; b, ectosomal auxiliary styles/anisoxeas; c, echinating acanthotylostyles; d, section through peripheral skeleton; e, known Antarctica distribution;f , plumose skeletal arrangement (scale = 1 mm); g, 'holotype' of Eurypon miniaceaum sensu Burton (BMNH 1928.2.15.244); h, Antarctica specimens (SAM S.537, 538) (scale = 30 mm); i, echinating acanthotylostyle. 1272 J. N. A. Hooper Subectosomal extra-axial spicules very long and thick styles, curved near basal end, with enlarged but not necessarily tylote bases, tapering to sharp points. Ectosomal auxiliary megascleres relatively long thin styles or anisoxeas, sometimes straight but mostly with prominent curvature at centre, with rounded tapering hastate bases, sometimes oxeote, tapering to sharp points. Echinating acanthostyles straight, thin or thick, with grossly enlarged tylote base, evenly cylindrical shaft and sharply pointed at apex, with vestigial spination on shaft or sometimes confined to points. Microscleres absent. Table 10. Comparisons in spicule measurements between specimens of Raspailia irregularis Hentschel Measurements are given in micrometres, presented as ranges (and means) of lengthx width, and represent 25 spicules measurements per specimen for each spicule category Material Choanosomal spicules BMNH 1928.2.15.244 Absent SAM S537 Absent SAM S538 Absent Subectosomal styles Ectosomal styles1 anisoxeas Echinating acanthostyles 1969-3016 x 35-54 (2651.4~44.6) 206&2970 x 23-65 (2512.3~38-4) 2544-3110 x 3548 (2873.7 ~ 4 1 . 3 ) Remarks Raspailia irregularis is not part of the continental Australian fauna, but is the only raspailiid so far recorded from the Australian Antarctic Territories (Koltun 1964, 1976). This species is referred to a new subgenus, Hymerhaphiopsis, within the genus Raspailia, in having smooth, greatly enlarged tylote bases on echinating megascleres, although it is admitted that this subdivision is probably merely one of convenience, based on the similarities in the geometry of echinating spicules with those of Hymeraphia (e.g. Fig. 4p). One of the two syntypes of R. irregularis from Antarctica is apparently extant in ZMB collections (Wiedenmayer, personal communication), although this has not yet been seen. The other syntype may be present in the collections of ZMH, but this has not yet been confirmed. Nevertheless, Hentschel's (1914) species is easily recognizable from his original description and figures. Raspailia irregularis was synonymised with Eurypon miniacea by Burton (1932) on the basis of similarities in structure and spiculation. Eurypon miniacea is supposedly well characterised and differentiated from other raspailiids by the swollen tylote bases of its acanthostyles, which resemble unspined versions of those found in typical Hymerhaphia from the Atlantic (e.g. Bowerbank 1864; Topsent 1904), but these features are also seen in the material of Burton (1932) and Koltun (1976), described above. That material was referred to R. irregularis by Boury-Esnault & van Beveren (1982), but it is possible that it correctly belongs to E. miniaceum. The holotype of E. miniaceum is apparently not in the collections of ZMB, although other types of other species collected from the same locality are present (Wiedenmayer, personal communication). There is a schizotype of the species in the BMNH (i.e. BMNH 1908.9.24.160 from Calbuco, Chile), but this material does not provide any more accurate details of the species beyond those provided by Thiele (1905). For the present, the decision of Boury-Esnault & van Beveren (1982) is upheld, and this has also been followed by Wiedenmayer et al. (in press). It could be argued that the larger echinating megascleres (with smooth tylote bases and vestigial spination on the shaft) represent choanosomal axial spicules, and the smaller Australian Raspailiidae 1273 more-heavily spined versions of these are the true echinating spicules, but there is an obvious continuum of spicule types from heavily spined to unspined, and these spicules are treated as identical. Genus Ectyoplasia Topsent Ectyoplasia Topsent, 1930: 2 3 4 - d e Laubenfels, 1936: 102; Wiedenmayer, 1977: 158. Type species: Spongia tabula Lamarck, 1814: 374 (by original designation and monotypy) (Figs 49-50). Diagnosis Flabellate, flattened arborescent or tubular growth form, with even or slightly corrugated hispid surface. Choanosomal skeleton slightly condensed axial reticulation of spongin fibres cored by styles or anixoxeas, echinated by clavulate acanthostyles; subectosomal extra-axial skeleton radially arranged plumose ascending tracts, composed of undifferentiated choanosomal styles protruding through ectosome, interconnected by uni- or paucispicular transverse spicule tracts producing regular reticulation; ectosomal region with a specialised skeleton of small styles or anisoxeas typically forming brushes around choanosomal styles at surface, but sometimes lying tangential to surface. Structural megascleres styles or rhabdostyles of 2 sizes, sometimes with anisoxeote or strongylote modifications; echinating spicules always acanthostyles with clavulate points; microscleres absent. Remarks The type species of Ectyoplasia shows classical raspailiid ectosomal specialisation, with brushes of ectosomal auxiliary spicules surrounding the bases of protruding choanosomal spicules. By comparison, E. frondosa (Figs 47, 48a-e) and E. ferox (Duchassaing & Michelotti 1864: 81; specimen ZMA POR4629 from the West Indies, kindly sent by Dr R.W.M. van Soest; Fig. 48f-g) have an atrophied ectosomal skeleton containing only sparsely dispersed tangential ectosornal spicules. Topsent (1930) suggested that the genus differed from Raspailia (s.s.) in having less condensed skeletal architecture, clavulate tips on acanthostyles, and 'exceptional differentiation' of the exhalant pores (oscula) along the lateral margins of branches. The latter feature has little systematic importance at the generic level, and in any case it does not occur in either E. frondosa or E. vannus. Similarly, supposed differences in skeletal architecture between the two genera are not upheld upon comparison with other species [e.g. compare R. phakellopsis (Figs 11-12) with E. frondosa (Figs 47-48), and R. darwinensis (Figs 23-24) with E. ferox (Fig. 48f-g)]. Thus, in Topsent's (1930) conception of Ectyoplasia this leaves only acanthostyle morphology, or modifications to acanthostyles, as a primary distinguishing feature. Another character is introduced here: like Raspailia s.s., Ectyoplasia has well differentiated axial and extra-axial skeletons, but in this genus spicules contained within both these skeletons are identical (i.e. the genus lacks specialised subectosomal megascleres). Ectyoplasia frondosa (Lendenfeld), comb. nov. (Figs 47, 48; Table 11) Antherochalina ffondosa Lendenfeld, 18876: 787, pl. 22, fig. 43.-Lendenfeld, 1888: 90; Burton, 1934a: 558. Clathria j?ondosa.-Whitelegge, 1902: 275, 279, 288 (in part); Hallmann, 1912: 237. Antherochalina perforata, in part.-Lendenfeld, 1888: 89-90; Whitelegge, 1902: 275, 279, 287. Not Antherochalina perforata Lendenfeld, 1887b: 788. Material Examined Lectotype (here designated). AM 2962 (dry): Vicinity of Exmouth, W.A. (Lendenfeld's No. 321). Paralectotype. BMNH 1886.8.27.446 (dry): (?) East coast of Australia, no other details known. [Not AM G3462: (Lendenfeld's no. 329, label reads 'Antherochalina levis', =Reniera dendyi; Whitelegge)]. Other material. AM G9043 (dry): dubious locality of 'Port Phillip, Victoria', no other details given; (label reads 'Antherochalina perforata Lend. Eastern Australia'). J. N. A. Hooper Fig. 47. Ectyoplasiafrondosa (Lendenfeld) (paralectotypeBMNH 1886.8.27.446): a, choanosomal axial and extra-axial styles; b, ectosomal auxiliary styles; c, clavulate acanthostyles; d , section through peripheral skeleton; e, probable Australian distribution. Substrate and Depth Range Unknown. Geographical Distribution Type material was apparently collected from the mid-west coast of W.A. (Fig. 47e) (AM specimen label reads 'purchased from J.E Bailey, Melbourne'), whereas published locality data contradicts this (see Remarks). Description Shape. Erect, stipitate flabelliform sponges (290-330 mm high, 90-140 mm wide when folded), consisting of short cylindrical stalk (20-40 mm long, 8-15 mm diameter), with thinly lobate fan (3-5 mm thick) bearing even margins or occasional digitate projections along lateral margins. Colour. Colour grey-brown in dry state. Oscula. Not observed. Texture and surface characteristics. Surface optically smooth, even, microscopically hispid. Texture firm, barely compressible, inflexible in dry state. Australian Raspailiidae Fig. 48. a-e, Ectyoplasia frondosa (Lendenfeld): a, lectotype AM 2962; b, paralectotype BMNH 1886.8.27.446 (scale = 30 mm); c, SEM of skeletal structure (left magnified 93 times, right magnified 744 times); d, peripheral skeleton (scale = 1 mm); e, echinating clavulate acanthostyle (scale = 50 pm). f-g, Ectyoplasia ferox (Duchassaing & Michelotti) (specimen ZMA POR4629): f , skeletal structure (scale = 1 mm); g, echinating clavulate acanthostyle (scale = 50 pm). Ectosome and subectosome. Ectosomal skeleton consists of erect choanosomal spicules from ascending primary tracts, piercing surface and forming small plumose brushes; below surface brushes lying tangentially are uni- or paucispicular tracts of choanosomal spicules embedded in peripheral skeletal fibres, intermingled with relatively rare ectosomal auxiliary spicules. Ectosomal spicules may form surface brushes but they probably do not surround bases of protruding choanosomal megascleres like typical raspailiids. However, type material is dry, ectosomal skeleton is not well developed, and this feature cannot be corroborated from existing material. J. N. A. Hooper 1276 Choanosome. Choanosomal skeleton irregularly reticulate, with clearly differentiated primary and secondary skeletal lines, and less well differentiated axial and extra-axial components. Axial skeleton consists of narrow fibrous core, with irregularly anastomosing, moderately heavy fibres cored by paucispicular tracts of choanosomal spicules; tracts connected by smaller uni- or bispicular fibres interconnecting larger fibres at all angles. Fibre anastomoses form relatively tight but irregular meshes. Extra-axial skeletal architecture radially reticulate, with multispicular, plumose, ascending, primary tracts of choanosomal megascleres diverging at periphery and piercing surface individually. Primary tracts interconnected by uni- or paucispicular secondary lines, more-or-less transverse and anatomosing irregularly, with primary skeletal lines forming larger meshes than in axis. Spongin in extra-axial skeleton minimal. Echinating clavulate acanthostyles moderately common, more abundant in axis than in peripheral skeleton. Mesohyl matrix with abundant, heavily pigmented, granular spongin, with few extra-fibre spicules. Megascleres (refer to Table 11 for dimensions). Choanosomal axial and extra-axial spicules thick, curved centrally or rarely straight, varying from styles with evenly rounded bases, strongyles, anisoxeas, to oxeas with hastate points. Subectosomal megascleres absent, or completely undifferentiated from choanosomal spicules. Ectosomal auxiliary megascleres thin,relatively small styles, oxeas or anisoxeas, rarely straight, usually sinuous or at least slightly curved, with hastate or rounded tips. Acanthostyles club-shaped, with unspined inflated bulbous bases, tapering to thinner bulbous tips on distal region often with clavulate microspines or vestigial spines, and small spines evenly distributed over only half length of shaft. Microscleres absent. Table 11. Comparisons in spicule measurements between specimens of Ectyophsia frondosa (Lendenfeld) Measurements are given in micrometres, presented as ranges (and means) of lengthxwidth, and . . represent 25 spicules measurements per specimen for each spicule category MaEnfipChoanosomal styles/ anisoxeas AM a 6 2 248-324 X 8-22 (290-4~16.2) BMNH 1889.8.27.446 194341 x 11-21 ( 2 7 6 . 6 ~16.8) AM G9043 242-3 16 x 9-2 1 (289-6~16.1) -- Subectosomal megascleres Lectotype Absent Paralectotype Absent Specimen Absent -- -- - - - - Ectosomal styles/ oxeas Echinating acanthostyles 151-194 x2 4 (176-8~2.9) 74-88 ~5-10 (81.6~7.2) 143-185 x 2-5 (166-0~3.4) 65-9 1 x 5-9 (75.2~790) 159-185 x2 4 (174-5~2.8) 61-84 x 5-8 (70.4~6-6) Remarks The status of this species is confused, owing to the idiosyncratic and at least partly erroneous systematics of Lendenfeld. Lendenfeld (1887a, 1888) based his original description of Antherochalina frondosa on three specimens. One is housed in the BMNH collections (designated here as the paralectotype), which Whitelegge (1902) incorrectly suggested was fragmented and virtually unrecognisable (Fig. 48b); one is a haplosclerid, identified as Reneira dendyi by Whitelegge (1902); and the third is identical to Lendenfeld's (1887~) figured specimen (AM Z962), which is dry and labelled as 'cotype', and is designated here as the lectotype of A. frondosa (Fig. 48a). The original locality of this species is cited as 'East coast of Australia' by Lendenfeld (1887a, 1888), but both the specimen Australian Raspailiidae 1277 label and the AM register bear the locality 'Western Australia' (cf. Whitelegge 1902: 275; Hallmann 1912: 237). Another specimen, labelled 'Antherochalina perforata' (AM G9043) from Eastern Australia (specimen label) or Port Phillip, Victoria (AM register), or possibly neither, was also referred to A. frondosa by Whitelegge (1902: 275). It is identical to the lectotype in all details. Whitelegge cites AM G9043 as bearing Lendenfeld's number 315, but n o evidence of that notation was seen on either specimen labels or the AM register. Lendenfeld's (1887b, 1888) descriptions are also incorrect. The material described above bears very little resemblance to published descriptions of the species, but it is most appropriately placed with the genus Ectyoplasia. It is easily differentiated from E. tabula in having significantly smaller spicule dimensions and a more markedly reticulate extra-axial skeleton formed by differentiated primary and secondary skeletal lines, whereas in E. tabula the axial component of the skeleton is significantly more condensed. In both these features E. frondosa shows closer similarities to the West Indies species E. ferox (Fig. 48f-g). Ectyoplasia tabula (Lamarck) (Figs 49, 50, 109e, 109f; Table 12) Spongia tabula Lamarck, 1814: 374. Ectyoplasia tabula.-Topsent, 1930: 23-24, text-fig. 2, pl. 3, fig. 6; de Laubenfels, 1936: 102; Wiedenmayer, 1977: 158. Material Examined Holotype. MNHN LBIM DT 553: E of Cape Leeuwin, W.A. ('le long des cotes de Leuwins, S.W. Australie, par Peron et Lesueur, 1803'). Other material (all material collected by the author using SCUBA,unless otherwise indicated). North-west Australia: BMNH 1896.2.28.2a: Unspecified locality from NW. Australia, Saville Kent collection. Darwin Region, N.T.: NTM 2997, 1004: EPMFR, 12" 24.5's.. 130" 48-OIE., 8-10 m depth, 26.x.1982 (stn EP10). NTM 22082: same locality, 7-10 m depth, 20.vii.1984 (stn EP15). NTM 22382: same locality, 8 m depth, 29.vii.1985 (stn EP23). NTM 22677, 2702: same locality, 9-12 m depth, 3.iv.1986 (stn EP28). NTM 21947: Stephen's Rock, Weed Reef, 12'29.2'S., 130" 47.11E., 6 m depth, 27.iv.1984 (stn WR1). NTM 2831, 843: Channel I., Middle Arm, 12' 33.8'S., 130" 51.4'E., 20 m depth, 18.vii.1982, coll. S. Chidgey (stn FN A13, A24, Don. 24, hooker). NTM 2852, 863: NNE. of Channel I., 12" 32.7'S., 130" 52 -5'E., 12-13 m depth, 20.viii.1982, coll. P.N. Alderslade (stn CI3). Cobourg Peninsula Region, N.T.: NTM 266 (3 specimens): Coral Bay, Port Essington, CPMNP, 11" 11.5'S., 132" 03 .OIE., 4-6 m depth, 17.x.1981 (stn CP20). NTM 21353: same locality, 11" 11 -3/S., 132" 03.75'E., 5.5 m depth, 16.v.1983 (stn CP60-1). NTM 21383 (2 specimens), 1390: same locality, 6 m depth, 17.v.1983 (sm CP61). NTM 22492: same locality, 11' 09.4'S., 132' 04/E., 4-7 m depth, 13.ix.1985 (stn CP71). NTM 23309: Inside outer reef slope, Coral Bay, 11"09-41S., 132" 04.01E., 2-7 m depth, 12.ix.1986 (stn CP70). Wessel Is, N.T.: NTM 23928, NCI Q66C4782-0: Bay N. side of Cumberland Strait, 11' 27 -53'S., 131" 28.S1E., 20 m depth, 14.xi.1990 (stn WI-6). Northwest Shelf, W.A.: NTM 21764: W. of Port Hedland, 19" 03.3'S., 118' 49.g1E., 82 m depth, 29.viii.1983, coll. T. Ward, (CSIRO RV 'Soela' S04183, sm 121-NWS21, beam trawl). NTM 21851: same locality, 19' 28 -9/S., 118" 52 3'E., 38 m depth, 31.viii.1983 (stn 127-NWS27). WAM 667-81(1) (fragment NTM 21701): N. of King Bay, Dampier Peninsula, 20" 3S1S., 116"45'E., 8 m depth, ll.viii.1978, coll. L. Marsh (stn ML 15, Don. 163, trawl). WAM 142-82 (fragment NTM 21713): off Port Hedland, 20" 12's.. 118' 2g1E., 16 m depth, 25.vii.1982, coll. J. Fromont (CSIRO RV 'Soela' ,904182, stn 01-Don. 167, trawl). WAM 153-82 (fragment NTM 21719): same locality, 20" 13/S., 118' 28'E., 18 m depth, 5.viii.1982 (stn 13, Don. 169). NTM 21185: N. of Bedout I., 19" 30.g1S., 118'48.7IE., 40m depth, 26.iv.1983 (CSIRO RV 'Soela' S02183, sm B7-NWS8, beam trawl). NTM 21156, 1166: same locality, 19" 29-6/S., 118" 52.2/E., 38 m depth, 26.iv.1983, stn B8-NWS6). NTM 21190: same locality, 19' 29.4l S., 118' 52.1 'E., 39 m depth, 26.iv.1983 (stn B8-NWS9). NTM 21146: W. of Port Hedland, 19" 58.g1S., 117" 51.3IE., 42 m depth, 22.iv.1983 (stn B1-NWS4). NTM 2706: N. of Adele I., Collier Bay, 15" 58.3'S., 122' 39-7'E., 59 m depth, 21.v.1982 (RV 'Sprightly' SP4182, stn 40-Don. 20, dredge). NTM 2647: W. of 80 Mile Beach, 19" 33-5'S., 119" 05.7IE., 35 m depth, 4.v.1982 (CSIRO RV 'Sprightly' SP4182, stn 78-Don. 15, dredge). NTM 22332: NW. of Lacepede I., 16" 29-33IS., 121' 27-29'E., 3 8 4 0 m depth, 17.iv.1985, coll. B.C. Russell (Taiwanese Pair Trawls, stn PT85-4-NWS34). NTM 22462: NW. of Amphinome Shoals, 19" 12IS., 118"36-5'E., 76-80m depth, l.vi.1985, coll. B.C. - J. N. A. Hooper 1278 Russell (Taiwanese Pair Trawl, stn BCR 8515 haul 6-NWS36). NTM 23037: N. of Amphinome Shoals, 19" 19.7-23 -3's.. 119" 08.8-12 -2'E., 50 m depth, 19.vii.1987 (U.S.S.R. RV 'Akademik Oparin', stn NWS5.5, beam trawl). NTM 23393 (2 specimens): 2 km W. of Hermite I., Monte Bello I., 20" 27 - 1IS., 115" 34.2' E., 6 m depth, 29.viii.1988, coll. D. Low Choy and NCI (stn NWS74, DLC-42). NCI Q66C-1341-X (fragment NTM 23473): S. of Muiron I., Exmouth Gulf, 21' 40.01S., 114' 20.01E., 12 m depth, 19-08-1988, coll. NCI (stn NWS90). Scott Reef, W.A.: PIBOC unreg.: near Scott Reef, 16" 36.71%. 121' 11 llE., 50 m depth, 17.xi.1990 (stn 39). PIBOC unreg.: 16' 41.4' S., 121' 09 -6'E.. 51-54 m depth, 4.xi.1990, coll. V. Krasochin, U.S.S.R. 'Akademik Oparin', dredge (stn 26). Central coast, W.A.: NTM 22961: Sunday I., Dirk Hartog I., Shark Bay, 26" 07.5'S., 113" 14.01E., 8-9 m, 13.vii.1987 (U.S.S.R. RV 'Akademik Oparin', sm SB5). - Substrate and Depth Range Attached to rock or dead coral substrate, associated with shallow coastal reef slopes and deeper offshore reefs; 4 to 82 m depth. Geographical Distribution Characteristic component of subtidal and shallow water tropical sponge fauna of NW. Australia; also present in warm temperate waters of W.A., as it was originally recorded from vicinity of Cape Leeuwin, southern W.A. by Pkron and Lesueur (Topsent 1930). From the limited information available the species appears to be endemic to this region (Fig. 49e). Fig. 49. Ectyoplasia tabula (Lamarck) (specimen NTM 22677) (type species of the genus Ectyoplasia Topsent): a, choanosomal axial and extra-axial styles; b, ectosomal auxiliary stylelanisoxea; c, echinating clavulate acanthostyle; d, section through peripheral skeleton; e, known Australian distribution. Description Shape. Stipitate digitate sponges (190620 mm high, 110-390 mm maximum breadth of branches), with basal holdfast and cylindrical stalks of variable length (50-130 mm long, 12-30 mm diameter) sometimes laterally flattened, with flattened branches (120-480 rnm long, 16-84 mm wide, 6-19 mm thick), branching in 1 plane only, usually bifurcating more than once and terminating in rounded or chiselled margins; degree of branch compression varies considerably. Colour. Live specimen coloration variable and relatively unstable, although variability within defined range. Majority of deeper water specimens light brown to grey-brown Australian Raspailiidae 1279 - (Munsell 7 5YR 812) or even lighter; shallow water specimens invariably light orange-pink (10R 818) (Fig. 109e). Oscula. Pores dispersed between surface ridges and conules; larger oscula usually on smoother distal and lateral margins of branches. Oscula may reach 2 mm diameter in life, often slightly raised above surface. Texture and su$ace characteristics. Surface features characteristic, with more-or-less regular corrugation of surface ridges running laterally across branches, becoming smaller and finally disappearing on glabrous rounded margins of each branch. Each ridge composed of minute, closely packed conules, together producing sandpaper-like stiff texture. Ectosome and subectosome. Ectosomal skeleton consists of discrete bundles of ectosomal auxiliary spicules surrounding bases of protruding choanosomal megascleres from extra-axial skeleton. Extra-axial skeleton with more-or-less radial non-plumose bundles of choanosomal axial spicules, in uni- or paucispicular tracts, extending from axis to surface in tracts of 1-3 spicule lengths (varying with branch thickness and presence of surface conules), protruding through surface for only short distances. Choanosomal and subectosomal megascleres undifferentiated, but spicules in peripheral tracts slightly thicker, poking through ectosome. Subectosomal brushes only rarely form plumose subdermal tracts. Spongin in mesohyl of subdermal region relatively heavy, smooth and non-granular. Choanosome. Skeletal architecture typical of most raspailiids with distinct reticulate axial skeleton, radial non-reticulate extra-axial skeleton, and plumose ectosomal spicule brushes. Axial skeleton irregularly reticulate, relatively condensed, with well developed spongin fibres cored by paucispicular lines of choanosomal axial spicules entirely enclosed by spongin. No obvious division of primary or secondary skeletal lines, and spicule tracts become increasingly plumose towards periphery (at junction of axial and extra-axial skeletons). Fibres moderately heavily echinated by clavulate acanthostyles scattered relatively evenly over skeletal lines. Few megascleres occur between spiculo-spongin tracts. Megascleres (refer to Table 12 for dimensions). Choanosomal styles of axial and extra-axial skeletons entirely smooth, straight or slightly curved, with evenly rounded non-tylote bases and slightly fusiform points. Subectosomal megascleres absent, or completely undifferentiated from choanosomal spicules. Ectosomal auxiliary spicules long, very thin, straight or slightly curved towards basal end, with smooth rounded non-tylote bases, and hastate points, predominantly styles but occasionally oxeas or anisoxeas. Raphidiform varieties (approximately 1-4 pm thick) also occur in dermal spicule brushes. Echinating acanthostyles relatively long, thin, straight, with bulbous aspinose bases, and non-bulbous clavulate points. Microscleres absent. Associations Shallow water specimens of E. tabula were remarkable in being able to survive during prolonged periods of complete coverage by epiphytic algae, without showing any noticeable signs of degeneration or extensive necrosis. Seasonal algal blooms in shallow waters of the wet-dry tropics occurred during the late dry and pre-wet seasons (September-January). Heavily infested specimens were re-examined during the dry season and were found to recover from algal infestation. Several specimens were also noted to be infected by parasitic barnacles and zoanthids. Remarks This species is distinguished from other Ectyoplasia by its characteristic flattened digitate growth form, with prominent surface ridges running across branches, pink-orange or yellowish 'pastel' coloration, compressed reticulate axial skeleton, radial non-reticulate extra-axial skeleton, undifferentiated extra-axial and axial spicules but with well developed ectosomal spicule brushes surrounding the protruding choanosomal extra-axial styles, well developed clavulate spines on club-shaped acanthostyles, and its spicule dimensions (Table 12). This species is most closely related to E. vannus (see below). J. N. A. Hooper Ectyoplasia vannus, sp. nov. (Figs 51, 52, 109g; Table 13) Material Examined (All material collected by the author using SCUBA, unless otherwise indicated.) Holotype. NTM 22497: Coral Bay, Port Essington, CPMNP, N.T., 11" 09 -4'S., 132"04.O'E., L 7 m depth, 13.ix.1985 (stn CP71). Australian Raspailiidae 1281 Table 12. Comparisons in spicule measurements between specimens of Ectyoplasia tabula (Lamarck) Measurements are given in micrometres, presented as ranges (and means) of lengthxwidth, and represent 25 spicules measurements per specimen for each spicule category. n =number of specimens examined Material Choanosomal styles MNHN LBIMDT553 438-843 x 17-29 (662-6~22.3) Mid-coast W.A. (n = 1) NW. Shelf region W.A. (n = 18) Darwin region N.T. (n = 10) Cobourg region N.T. (n = 8) 446-805 x 17-29 (637-9~21.4) 618-1016 x 15-35 (752.0~23.7) 601-879 x 8-39 (755.1~26.6) 472-8 17 x 11-33 (686-3x24-3) Subectosomal megascleres Holotype Absent Specimens Absent Absent Absent Absent Ectosomal styles/ anisoxeas Echinating acanthostyles 244-331 x &7 (301.2~5.3) 111-146 x6-11 (12404~8.3) 273-387 x 2-7 (331-1~3.8) 285421 x 2-6 (365.3~3.9) 336-412 x 4-7 (378-7~5.4) 221406 x2.5-6 (301.5~4.1) 93-135 x 5-9 (113.6~7.1) 97-138 x 5-1 l (112.3~8-6) 89-125 x5-10 (105.9~7.3) 93-148 x 6-10 (111.7~7.6) Paratype. NTM 23045: N. of Amphinome Shoals, NWS, W.A., 19' 19.7-23.3'S., 119" 08-812-2IE., 50 m depth, 19.vii.1987 (U.S.S.R. RV 'Akademik Oparin', stn NWS55, beam trawl). Other material. Darwin Region, N.T.: NTM 22048, 2052: EPMFR, 12" 25 .O1 S., 130" 48.4'E., 6-10 m depth, 10.v.1984, coll. R.S. Williams (stn EP14). NTM 2611: Cootamundra Shoals, N. of Melville I., 10" 50.2/S., 129" 13.16'E., 36 m depth, 7.v.1982, coll. R. Lockyer (stn Don. 7). NTM 2842: Channel I., Middle A m , 12" 33.S1S., 130" 51.4IE., 20 m depth, 18.vii.1982, coll. S. Chidgey (stn FN A23, Don. 24, hooker). Wessel Is, N.T.: NTM 23927, 3933, NCI Q66C 4781-N: Bay N. of Cumberland Strait, 11" 27.5'S., 131" 28 .8'E., 20 m depth, 14.xi.1990 (stn WI-6). NCI Q66C 4718-T: NE. tip of Wigram I., English Company Is, 11" 44.4IS., 136" 37-83/E., 18 m depth, 12.xi.1990, coll. NCI. Northwest Shelf, W.A.: NTM 2704, 743: N. of Adele I., Collier Bay, 15" 58.3/S., 122" 39-7/E., 59 m depth, 21.iv.1982 (CSIRO RV 'Sprightly', SP4/82, stn 40-Don. 20, dredge). NTM 2650, 652: W. of 80 Mile Beach, 19" 33.5/S., 119°05-71E., 35 m depth, 4.v.1982 (CSIRO RV 'Sprightly' SP4/82, stn 78, Don. 15, dredge). NTM 2686: W. of Buccaneer Archipelago, 16" 20.01S., 120' 10-OIE., 35 m depth, 28.iv.1982 (CSIRO RV 'Sprightly', SP4/82, stn 39, Don.19, dredge). NCI Q66C1339-V (fragment NTM 23472): S. of Muiron I., Exmouth Gulf, W.A., 21'40.01S., 114" 20.01E., 12 m depth, 19.viii.1988, coll. NCI (NWS90, SCUBA). Substrate and Depth Range Shallow coastal reefs, 4-59 m depth range. Geographical Distribution NW. Australia, W.A. and N.T. (Fig. 5 le). Description Shape. Erect, stipitate, flabelliform sponges (60-305 mm high, 35-288 mm wide), consisting of short cylindrical or laterally flattened stalk (20-72 mm long, 12-20 mm wide, 4-5 mm thick), with thinly lobate fan (3-6 mm thick) bearing evenly rounded unfolded margins, or with ragged, digitate or convoluted margins. Colour. Beige alive (7.5YR 814) (Fig. 109g) and in ethanol. J. N. A. Hooper Fig. 51. Ectyoplasia vannus, sp. nov. (holotype NTM 22497): a, choanosomal axial and extra-axial styles; b, ectosomal auxiliary styles/anisoxeas; c, echinating clavulate acanthostyles; d, section through peripheral skeleton; e, known Australian distribution. Oscula. Minute oscula (up to 2 mm diameter), slightly raised above surface and evenly dispersed over both surfaces of fan, although these pores are more difficult to observe in preserved material. Texture and su$ace characteristics. Surface optically smooth and even, microscopically villose and very hispid. Texture firm,barely compressible, but flexible. Ectosome and subectosome. Ectosomal skeleton consists of erect choanosomal styles from extra-axial skeleton piercing surface, forming small plumose brushes or occurring singly, evenly spaced approximately 700 p m apart, with ectosomal auxiliary styles forming plumose brushes surrounding bases of protruding choanosomal megascleres. Ectosome with heavy layer of darkly pigmented granular spongin, up to 500 ,um thick, through which plumose extra-axial brushes of choanosomal spicules extend separating axial and extra-axial skeletons. Choanosome. Choanosome lacks compressed axial fibres characteristic of E. tabula, although clearly differentiated axial and extra-axial skeletal components present. Axial skeleton plumose, with vestigial reticulation, with central core of choanosomal styles in bundles of up to 10 spicules running longitudinally through branches, sparsely interconnected Australian Raspailiidae Ectyoplasia vannus, sp. nov.: a, holotype (NTM 22497); b, paratype (NTM 23045) (scale = 30 mm); c, echinating acanthostyles (scale = 100 pm); d, SEM of skeletal structure; e, SEM of specialised ectosomal skeleton; f-g, SEMs of echinating acanthostyle geometry and spination. Fig. 52. by uni- or paucispicular tracts of styles; plumose tracts of choanosomal styles diverge from central core and ascend towards surface forming extra-axial skeleton. Echinating acanthostyles occur predominantly in axial region, but these vary in abundance between specimens. Mesohyl matrix in axial skeleton only lightly invested with spongin. Megascleres (refer to Table 13 for dimensions). Choanosomal axial and extra-axial styles are long, thick, curved centrally or rarely straight, with evenly rounded bases and fusiform points, sometimes with strongylote modifications. Subectosomal megascleres absent, or at least completely undifferentiated from choanosomal spicules. Ectosomal auxiliary megascleres thin, relatively small styles, with oxeote and anisoxeote modifications, straight, sinuous or slightly curved, with hastate or rounded smooth points. Echinating acanthostyles relatively long, thin, straight, club-shaped, with smooth bulbous bases, tapering to thinner only slightly bulbous points on distal end bearing clavulate microspines, and small spines cover only 112 length of shaft. Microscleres absent. 1284 J. N. A. Hooper Table 13. Comparisons in spicule measurements between specimens of Ectyoplasia vannus sp. nov. Measurements are given in micrometres, presented as ranges (and means) of lengthxwidth, and represent 25 spicules measurements per specimen for each spicule category Material Choanosomal styles NTM 22497 580-804 x 8-36 (696.4~22.9) NTM 23045 642-853 x 14-30 Subectosomal megascleres Holotype Absent Paratype Absent Ectosomal styles/ anisoxeas Echinating acanthostyles 350-552 x2-5-6 (444.4~3-8) 306561 x 2-6 (722.3 ~ 2 2 . 4 ) NTM various (n = 10) 435-826 x 13-32 (632-8~23.5) Specimens Absent 329-555 X3-11 (413.3~7-0) Associations Phyllidia sp. nudibranchs were collected from this species in the Darwin region, apparently feeding on the sponge mucous (W. Rudman, personal communication). Remarks This species has a growth form similar to E. ffondosa, but has a skeleton and spiculation reminiscent of E. tabula. It is obviously closely related to E. tabula, with similar spicule geometry, spicule dimensions, and extra-axial skeletal construction, but the two species appear to differ significantly in growth form, surface sculpturing and the structure of the axial skeleton (plumo-reticulate v. reticulate, respectively). These differences may be trivial, but I propose to erect this new species on the basis of these cryptic differences, and this conclusion is supported by biochemical differences between E. vannus and E. tabula (i.e. general protein electrophoresis, carotenoid pigments and free amino acid profiles; Hooper et al. 1992). Etymology This species is named for its flabellate fan-shape, from the latin vannus (f., fan used for winnowing grain). Genus Endectyon Topsent Endectyon Topsent, 1920: 25.-Burton, 1938: 34; Cabioch, 1968a: 224; Thomas, 1976: 169. Hemectyon Topsent, 1920: 27 (type species Raspailia (?) hamata Schmidt, 1870: 62, by original designation; holotype from the Atlantic Ocean, LMJG, schizotype MNHN LBIM DT 2161) (Fig. 53d-f). Basiectyon Vacelet, 1961: 37 (type species B. pilosus Vacelet, by monotypy; lectotype from Corsica, Mediterranean, SMEM 1705(1), schizotype MNHN LBIM DNBE 718L) (Fig. 53j-k). Type species: Phakellia tenax Schmidt, 1870: 62 (by original designation; holotype from the Atlantic Ocean, LMJG unregistered, schizotypes MNHN LBIM DCL 2163, 1194, BMNH 1870.5.3.170) (Fig. 53g-i, Table 15). Diagnosis Arborescent growth form; surface prominently hispid due to single long projecting spicules or numerous close-set conulose projections; always with marked axial and extra-axial Australian Raspailiidae 1285 differentiation of skeleton; axial skeleton with well developed spongin fibres forming condensed reticulation, cored by stout choanosomal styles; extra-axial subectosomal skeleton radial or plumose, with multi- or paucispicular tracts of long subectosomal styles (Endectyon) or choanosomal styles (nominal genus Hemectyon), sometimes connected by unispicular tracts forming hexagonal meshes, usually protruding through surface. Ectosomal skeleton varies from typical raspailiid condition, with thin ectosomal styles grouped in brushes around protruding subectosomal styles (Endectyon), to surface brushes composed of subectosomal styles only (nominal genus Basiectyon), to brushes of acanthostyles surrounding choanosomal styles (nominal genus Hemectyon). Erect brushes of echinating acanthostyles located on outer margin of axial skeleton, surrounding boundary between extra-axial and axial regions, or forming plumose brushes along length of extra-axial tracts, or localised exclusively to base of sponge (nominal genus Basiectyon). Structural megascleres smooth styles of 2-3 size categories; echinating megascleres modified acanthostyles and/or acanthostrongyles with peculiar strongly curved (clavulate) hooks on shaft, base, and/or apex. Microscleres absent. Remarks Endectyon differs from typical raspailiids (e.g. Raspailia) in having clavulate modifications to acanthostyle geometry. Moreover, unlike species of Raspailia in which echinating spicules are relatively evenly dispersed throughout the skeleton, Endectyon has these confined to a particular region of the skeleton (i.e. outside the axis). Some species of Endectyon show reduced characteristics from the typical condition. The affinities of these species are not completely obvious, but the most reasonable interpretation considered here is that they are forms of Endectyon. The nominal genus Hemectyon differs in having a more openly reticulate axial skeleton, lacking differentiated subectosomal megascleres in the extra-axial skeleton, and lacking a specialised ectosomal skeleton (Fig. 53d-f). Basiectyon also has a loosely reticulate axial skeleton (Fig. 5351, lacks ectosomal specialisation, and its acanthostyles (Fig. 53k) are localised at the base of the sponge. However, acanthostyle geometry, acanthostyle distribution, and axial and extra-axial skeletal structure are shared features that argue for their inclusion into a single taxon (see Topsent 1920). The genus contains 10 species: E. tenax (Schmidt) (Fig. 53g-i), E. delaubenfelsi Burton, 1930: 490 (Fig. 53a-c), E. demonstrans Topsent, 1892a: 118, E. fruticosa (Dendy, 1887: 160), E. hamata (Schmidt 1870: 62) (Fig. 53d-f), E. lamellosa Thomas, 1976: 169, E. pilosus (Vacelet, 1961: 37) (Fig. 53j-k), E. teissieri Cabioch, 1968b: 224, E. tenuis p d l e y & Dendy, 1886: 482), and E. thurstoni (Dendy, 1887: 161) (see below). Two other species [E. xerampelina (Larnarck, 1814: 443) and E. elyakovi, sp. nov.] are also referred here to this genus. Endectyon elyakovi, sp. nov. (Figs 54, 55; Table 14) Material Examined Holotype. NTM 22957: Sunday I., near Dirk Hartog I., Shark Bay, W.A., 26' 07.5's.. 113" 14.01E., 9 m depth, 13.vii.1987, coll. J.N.A. Hooper (RV 'Akademik Oparin', sm SB5, SCUBA). Paratype. AIMS RN P17 (fragment NTM 22738): Pandora Reef, vicinity of Townsville, GBR, Qld, 18" OOIS., 146" 23-26/E., 15 m depth, 23.vi.1981, coll. C.R. Wilkinson (stn Don. 214, SCUBA). Substrate and Depth Range Sand flat and coral rubble, 9-15 m depth. Geographical Distribution Shark Bay, W.A. and Great Barrier Reef, Qld (Fig. 54f). Description Shape. Dichotomously branched, arborescent sponges (100-230 mm high, 80-210 mm wide), with short cylindrical digits (10-35 mm long, 5-14 mm diameter) and a long or 1286 J. N. A. Hooper Fig. 53. a-c, Endectyon delaubenfelsi Burton: a, holotype (BMNH 1929.8.21.3) (scale = 30 mm); b, peripheral skeleton (scale = 1 mm); c, echinating acanthostyle (scale = 50 pm); d-f, E. harnata (Schmidt) (schizotype MNHN LBIM DT 2161) (type species of the nominal genus Hemectyon Topsent): d, skeletal structure (scale = 1 mm); e, peripheral skeleton (scale = 200 pm); f , echinating acanthostyle (scale = 50 pm); g-i, E. tenax (Schmidt) (schizotype MNHN LBIM DCL 2163) (type species of the genus Endectyon Topsent): g, skeletal structure (scale = 1 mm); h, echinating acanthostyles (scale = 50 pm); i, peripheral skeleton (scale = 200 pm); j-k, E. pilosus (Vacelet) [lectotype SMEM 1705(1)] (type species of the nominal genus Basiectyon Vacelet): j, skeletal structure (scale = 1 mm); k, echinating acanthostyle (scale = 50 pm). Australian Raspailiidae a, choanosomal axial style; b, ectosomal auxiliary style; d, echinating acanthostyles/acanthostrongyles;e, section through peripheral skeleton; f , known Australian distribution. Fig. 54. Endectyon elvakovi, sp. nov.: subectosomal exki-axiai style; c, short stalk (9-27 mm long, 9-16 mm diameter). Branches bifurcate repeatedly and taper distally to fine points. Colour. Live coloration bright red (5R 5/10), brown in ethanol. Oscula. Not seen. Texture and sudace characteristics. In situ surface highly conulose and slightly optically hispid, closely resembling Axiamon foliurn (=Reniochalina stalagmitis) of family Axinellidae. Texture firm,compressible, easily flexible. Surface of paratype infested with white zoanthid. Ectosome and subectosome. Specialised ectosomal skeleton present, with brushes of raphidifom auxiliary styles surrounding bases of protruding extra-axial spicules. Extra-axial styles poke through surface for short distance, usually in groups of several spicules, with bases embedded in peripheral (plumose) skeletal tracts, and extra-axial spicules only slightly larger than more abundant choanosomal structural spicules. Tracts in extra-axial region predominently plumose, although occasional spicules may f o m criss-cross and give reticulate appearance. Extra-axial spicule tracts enclosed within light spongin fibres, forming elongate anastomoses, and mesohyl matrix in this region contains relatively heavy deposits of yellow-brown spongin. Tracts extend outwards from axial core as bundles; bundles correspond to surface papillae. Echinating acanthostyles lightly scattered throughout J. N. A. Hooper Fig. 55. Endectyon elyakovi, sp. nov.: a, holotype (NTM 22957); b, paratype (NTM 22738) (scale = 30 mm); c, peripheral skeleton (scale = 1 mm); d, SEM of skeletal structure; e, SEM of fibre characteristics;f , SEM of the periphery of the axial skeleton; g, SEM of echinating acanthostyle. extra-axial skeleton, slightly more common at edges of fibre bundles, particularly in peripheral skeleton. Choanosome. Choanosomal skeletal architecture axially condensed, with obvious differentiation of axial and extra-axial skeletals. Axial region with heavier spongin fibres and much tighter fibre meshes; all fibres cored only by paucispicular tracts of choanosomal styles. Echinating acanthostyles absent from axial core and mesohyl matrix in axis relatively light. Australian Raspailiidae 1289 Megascleres (refer to Table 14 for measurements). Choanosomal styles of axial core variable in length and thickness, typically slightly curved towards basal end, slightly fusiform, with rounded non-tylote bases. Subectosomal extra-axial styles relatively long, straight or slightly curved towards base, with fusiform points and evenly rounded bases. Table 14. Comparisons in spicule measurements between specimens of Endectyon elyakovi, sp. nov. and related species Measurements are given in micrometres, presented as ranges (and means) of lengthxwidth, and represent 25 spicules measurements per specimen for each spicule category Material Choanosomal megascleres Subectosomal styles Ectosomal styles Echinating acanthostyles Endectyon elyakovi, sp. nov., Holotype NTM 22957 321-394 ~9-11 ( 3 5 2 . 4 ~10.2) AIMS RN P17 262-35 1 X 13-19 ( 3 2 2 . 0 ~15.6) BMNH 1929.- 210-268 x412 (238.6~8.6) 438-594 ~7-17 ( 5 6 0 . 8 ~12.0) 214321 ~1.5-2.5 (247-4~2.1) 104-135 x 9-1 3 (122.8~10.6) 241-274 ~2.0-3.5 (255.2~2.8) 96-131 X8-12 ( 1 1 5 . 4 ~10.0) Paratype 521-630 x 14-22 ( 5 7 4 . 8 ~18-2) E. delaubenfelsi Burton, Holotype 8.21.3 5 18-847 x4-15 (727- 1 ~ 9 . 4 ) Not seen 120-137 x 6-9 (127.8~7.8) 197-228 ~1.5-2.5 (214.6~2.1) 110-132 x 12-15 (122.4~13.2) E. tenax (Schmidt), Holotype LMJG unreg. 239404 x 15-20 (351.2~17.4) 806-1134 x 16-25 (979-1~20.0) Ectosomal auxiliary spicules mostly straight, relatively thick towards rounded base but becoming more raphidiform and hair like towards apex. Echinating acanthostyles short, thick, with slightly subtylote bases bearing few very large recurved hooks, and hastate points bearing terminal or subterminal large, recurved hooks. Shaft may or may not bear hooks, and if present never numbering more than a few; some spicules strongylote with double-clad ends, whereas others modified styles. Microscleres absent. Remarks Endectyon elyakovi is closely related to the type species E. tenax from the Atlantic Ocean. The two species differ in geometry and size of megascleres that core the axial and extra-axial skeletons, and E. tenax has very long extra-axial subectosomal megascleres (Fig. 53h) whereas those of E. elyakovi are much shorter and are only differentiated slightly from the choanosomal axial spicules (Table 14). In gross skeletal structure, growth form and live coloration the present species also shows similarities to E. delaubenfelsi from Plymouth (holotype BMNH 1929.8.21.3), E. demonstrans from the North Atlantic Ocean (schizotype MNHN LBIM DT 1255), and E. thurstoni (see below). Nevertheless, these species are easily differentiated by their acanthostyle geometry (Figs 53-55), spicule dimensions (Tables 14-15), and in E. elyakovi the extra-axial fibre bundles are composed of subectosomal spicules. Both known specimens of E. elyakovi were collected from opposite sides of Australia, from different habitats (AIMS RN P17 from a coral reef lagoon, NTM 22957 from a sandy bottom and limestone rock reef), but there is no doubt that these specimens are conspecific. Some differences were recorded between both specimens in spicule dimensions (Table 14): the Queensland specimen has significantly more robust choanosomal and subectosomal 1290 J. N. A. Hooper styles, but these differences are considered to be relatively minor, and in all other details the species are the same. This species may eventually prove to be distributed around the coast of northern Australia, but despite recent collecting efforts it has not been discovered in Northern Territory waters or in the Gulf of Carpentaria. Etymology The species is named in appreciation of Professor Georgy Elyakov, Director of the Pacific Institute of Bio-organic Chemistry, Vladivostok, who provided the author with the means to gain access to remote collecting localities along the Western Australian, Gulf of Carpentaria and northern GBR coastlines, during the cruises of the RV 'Akademik Oparin', 1987, 1990. Endectyon fruticosa aruensis (Hentschel), comb. nov. (Fig. 56) Raspailia fruticosa Dendy, 1887: 160.-Burton, & Rao, 1932: 347; Thomas, 1976: 172. Raspailia fruticosa Dendy var. aruensis Hentschel, 1912: 372. Material Examined Holotype. SMF 984: Sungi Barkai, AN I., Arafura Sea, 6" S, 134" 501E., 18 m depth, 10.iv.1908, coll. H. Merton (schizotype MNHN LBIM DCL 2181). Other material. NTM 23655, 3846, NCI Q66C-4098-U: NE. tip of KO Yao Hyai, E of Phuket I., S. Andaman Sea, Thailand, 8" 5.5/N., 98" 31.6/E., 8 m depth, 02.vi.1990, coll. J.N.A. Hooper and NCI. Substrate and Depth Range Rock substrate, high sediment, 8-18 m depth. Geographical Distribution Known from Tuticorin and Madras, India, S. Andaman Sea, Thailand, and the Arafura Sea (Fig. 56f). Description Shape. Small digitate sponge 18 mm high, with basal holdfast (4 mm wide) and short cylindrical stalk (4 mm long, 2-3 mm diameter), producing several flattened bifurcated branches (3-9 mm long, 3-5 mm wide, up to 2 mm thick); margins of branches rounded. Colour. Red to red-brown when alive, khaki brown in ethanol. Oscula. Not observed. Texture and s u ~ a c echaracteristics. Surface even but prominently hispid. Texture harsh and flexible. Ectosome and subectosome. Peripheral skeleton greatly reduced, consisting of radial arrangement of long subectosomal styles protruding long way through surface, surrounded by brushes of whispy ectosomal anisoxeas/styles at their points of insertion through ectosome. Subectosomal styles of extra-axial skeleton inserted into axial skeleton almost immediately below surface, surrounded at bases, in contact with axis, by heavy tufts of acanthostyles. Choanosome. Clear distinction between axial and extra-axial regions of skeleton. Axial skeleton dense, closely reticulate, occupying major portion of branch diameter. Fibres absent, but skeleton consisting of criss-cross of choanosomal styles. No echinating megascleres seen in axial skeleton. Spongin in mesohyl of choanosome moderately heavy, granular, and concentrated near peripheral skeleton. Megascleres. Choanosomal axial styles relatively long, slender, straight or slightly curved towards basal end, with evenly rounded smooth bases and slightly hastate points [308-(407 5)-482 x 8-(10.7)-13 pm]. - Australian Raspailiidae a Fig. 56. Endectyon fruticosa aruensis (Hentschel): a, choanosomal axial styles; b, subectosomal extra-axial style; c, ectosomal auxiliary styles; d, echinating acanthorhabdostyles; e, section through peripheral skeleton; f, known Australasian distribution; g, echinating acanthostyles; h, holotype of variety (SMF 984) (scale = 10 mm); i, skeletal structure (scale = 1 mm). Subectosomal extra-axial styles long, thin or thick, usually curved centrally, sometimes recurved and flexuous, with evenly rounded smooth bases and tapering to sharp fusiform points [943-(ll69 5)-1507 x 9 4 1 5 6)-20 pm]. Ectosomal spicules vary from true styles with tapering fusiform rounded bases, to asymmetrical anisoxeas with sharp points at both ends. Pointed ends of all ectosomal spicules whispy, usually curved or recurved, tapering to very fine points [282-(318-2)358 x 1 - 5 4 2 7)-4 pm]. 1292 J. N. A. Hooper Echinating clavulate acanthostyles robust, straight or slightly curved towards basal end, with rounded predominately smooth bases bearing few large recurved spines, terminally or subterminally, whereas apex sharp-pointed bearing sparse evenly distributed large recurved spines [141-(170 .0)-193 x 10-(14.2)-17 pm]. Microscleres absent. Remarks The single specimen known from Aru I. may require a new name, in which case 'aruensis' should be used, because growth form and spicule dimensions differ between it and the Madras form. Material listed above from the Andaman Sea is most similar to the Indian morph. This species has not been recorded from Australian waters, but is probably a member of the Dampierian-tropical marine province, the type locality being approximately 300 km from Australian territorial reefs. This species is characterised by its flattened branching growth form, prominently hispid surface, distribution of echinating spicules on the outer side of the axial skeleton and the clavulate geometry of acanthostyles, which is also typical of other Endectyon species. Dendy (1887: 160) gave measurements for the holotype of E. ffuticosa from Madras as follows: choanosomal styles ( 3 1 5 10 ~ pm), subectosomal styles (800x7 pm), ectosomal styles/anisoxeas (present as 'raphides') and clavulate acanthostyles (140 x 9 5 pm). These dimensions compare closely to Hentschel's variety from Aru I., and indeed the two forms differ mainly in their growth forms. The distinction between the typical form Wendy 1887) and the variety (Hentschel 1912) is maintained here until the Indian material is seen. Endectyon thurstoni Wendy) (Figs 57, 58; Table 15) Raspailia thurstoni Dendy, 1887: 161, pl. 12, fig. 1. Hemectyon thurstoni.-Burton & Rao, 1932: 347. Endectyon thurstoni.-Burton, 1938: 34, pl. 4, fig. 26; Cabioch, 1968b: 222; Thomas, 1976: 169. Material Examined Holotype. BMNH 1887.8.4.9: Madras, Coromandel Coast, India, coll. E. Thurston, depth and date of collection unknown. Other material (all material collected by the author unless otherwise indicated). Northwest Shelf, W.A.: NTM Z721: N. of Adele I., Collier Bay, 15" 58.3' S., 122' 39.7'E., 59 m depth, 21.iv.1982 (CSIRO RV 'Sprightly' SP4/82, stn 40-Don. 20, dredge). NTM 21260: N. of Bedout I., W. of Port Hedland, 19" 28.5'S., 118" 55.3'E., 40m depth, 26.iv.1983 (CSIRO RV 'Soela' S02/83, sm B9-NWSIO, beam trawl). NTM 21761: W. of Port Hedland, 19" 03.3'S., 118' 49.g1E., 82 m depth, 29.viii.1983, coll. T. Ward (CSIRO RV 'Soela' S04/83, sm 121-NWS21, trawl). NTM 21845: same locality, 19' 26.gfS., 118" 54.2'E., 50 m depth, 30.viii.1983 (stn 126-NWS26). Scott Reef, W.A.: PIBOC 012-141: near Scott Reef, 16" 25/S., 121" 10.5'E.. 4749 m depth, 4.xi.1990, coll. V. Krasochin, U.S.S.R. RV 'Akademik Oparin', dredge (stn 29). PIBOC unreg.: 16" 20.9'S., 121' 13-2'E., 50-52 m depth, 8.xi.1990 (stn 44). Substrate and Depth Range Deeper offshore rock and coral reefs, 40-82 m depth. One previous record of this species suggests that it may be restricted to these trawl-depths in deeper coastal waters (Burton & Rao 1932). Geographical Distribution Vicinity of Madras, India and Northwest Shelf, W.A. (Fig. 57e). Description Shape. Arborescent, dichotomously branched sponges, 170410 mm long, with thin, evenly cylindrical bifurcating branches (28-190 mm long, 4-12 mm diameter) surmounted Australian Raspailiidae Endectyon thurstoni (Dendy) (specimen NTM 21761): a, choanosomal axial and extra-axial style; b, ectosomal auxiliary styles; c, echinating acanthostyles/acanthostrongyles;d, section through peripheral skeleton; e, known Australian distribution. Fig. 57. on small cylindrical basal stalk (29-90 mm long, 16-25 mrn diameter) and expanded basal plate of attachment. Branches taper towards distal extremities. Colour. Live coloration dark orange-brown (Munsell IOR 5/10), dark brown in ethanol. Oscula. Not seen. Texture and sudace characteristics. Surface relatively optically smooth and even, not visibly hispid in live state, whereas upon preservation surface breaks up into small compartmentalised sections which superficially resemble conules. Texture soft, compressible, flexible, but difficult to tear. Ectosome and subectosome. Ectosome lacks specialised dermal skeleton, or at least thinner auxiliary spicules not confined to surface but scattered throughout extra-axial region. Nevertheless, paucispicular bundles of ectosomal auxiliary megascleres occasionally found close to and perpendicular with ectosome, and these may be interpreted as remnants of dermal skeleton. Both choanosomal styles from peripheral extra-axial tracts and echinating acanthostyles protrude only short distances through ectosomal skeleton, forming moderately dense plumose structures, but surface cannot be considered to be prominently hispid. Subectosomal extra-axial skeleton occupies major portion of branch diameter (up to 70%), with groups of ascending plumose spiculo-spongin tracts forming bundles. It is these bundles which break up and separate upon preservation, giving surface a microvillose appearance. Overall structure of extra-axial skeleton is radial composed of ascending pauci- J. N. A. Hooper 1294 or multispicular tracts, connected by fewer more randomly orientated transverse uni- or paucispicular tracts. Spicules in extra-axial region not confined to spongin fibres, although most are enveloped within heavy type B spongin. Echinating acanthostyles abundant in subectosomal region, echinating plumose choanosomal spicule tracts at oblique angles. Numerous thinner auxiliary spicules also dispersed between major tracts. Spongin in mesohyl non-granular, mostly dark brown pigmented, heaviest in ectosomal region. Choanosome. Choanosomal skeleton axially condensed reticulation of fibres and spicules, with well marked axial and extra-axial skeletal differentiation. Unlike extra-axial skeleton, spongin fibres clearly visible in core, heaviest at centre. Fibres cored by pauci- or multispicular tracts of choanosomal styles. Axial megascleres usually run longitudinally through branches, with fewer uni- or paucispicular connecting transverse elements. Fibre anastomoses form small elongate or oval meshes, with light mesohyl matrix; echinating megascleres virtually absent from core; no difference in sizes of megascleres between axial, extra-axial and peripheral skeletal tracts. Megascleres (refer to Table 15 for measurements). Choanosomal styles in axial and extra-axial skeletons are variable in size but not divisible into more than 1 category. Styles slightly curved towards basal end, less frequently straight, almost hastate, with rounded non-tylote bases. Subectosomal extra-axial spicules absent. Ectosomal auxiliary megascleres rhaphidiform styles, extremely thin, mostly straight but tapering to sinuous hair-like points, and with only slightly thicker bases. Echinating megascleres more-or-less stylote, with strongylote modifications in form of heavy spines on both ends of spicule. Spines recurved, resembling double-clad cladotylotes, showing superficial similarities to Acarnus of Myxillidae, but these are obviously no more than modified acanthostyles. Spines large and recurved, sometimes distributed all over shaft and both ends, but more often with small aspinose area proximal to (subtylote) basal end. Microscleres absent. Remarks Acanthostyles of both E. thurstoni (Fig. 58f) and E. hamata (Fig. 53f) have a relatively consistent geometry and spination, whereas in many other species of Endectyon these features are more variable. Similarly, both species lack long extra-axial subectosomal styles or any specialised (true) raspailiid ectosomal skeleton, and for these reasons it is possible that the nominal genus Hemectyon should be retained for these two species. This argument is not accepted here, because modifications to acanthostyle geometry and localisation of acanthostyles to particular regions within the skeleton are given higher priority. Burton & Rao (1932) suggested that E. thurstoni and E. fi-uticosa (both from Madras) were synonyms, but these differ considerably in their spicule dimensions (E. fruticosa having very long and thin subectosomal extra-axial styles), the presence or absence of specialised ectosomal features, and degree of axial compression. Specimens from NW. Australia differ from the holotype in the thickness of branches (maximum 5 mm diameter in holotype, 4-12 mm in recent material) and spicule dimensions (see Table 15). In particular, the length of acanthostyles is significantly greater in recently collected material, and choanosomal styles are also marginally more robust. There is considerable variability between specimens from NW. Australia, and differences between this material and the holotype are undoubtedly minor. Endectyon xerampelina (Larnarck), comb. nov. (Fig. 59) 372. Spongia xerampelina Lamarck, 1814: 443.-1815: Raspailia xerampe1ina.-Topsent, 1932: 96-97, text-fig. 2, pl. 4, fig. 7. Material Examined Holotype. MNHN LBIM DT 574: locality not known (schizotype BMNH 1954.2.20.27, desilicified section). Fig. 58. Endectyon thurstoni (Dendy): a, holotype (BMNH 1887.8.4.9) (scale = 30 mm); b, specimen (NTM 21260) (scale = 30 mm); c, skeletal structure (scale = 1 mm); 4 SEM of extra-axial skeleton; e, SEM of skeletal structure; f, SEM of echinating acanthostyles; g, SEM of spine geometry. J. N. A. Hooper 1296 Table 15. Comparisons in spicule measurements between specimens of Endectyon thurstoni (Dendy) Measurements are given in micrometres, presented as ranges (and means) of lengthxwidth, and represent 25 spicules measurements per specimen for each spicule category Material Choanosomal megascleres Subectosomal styles Ectosomal styles Echinating acanthostyles - BMNH1887.8.4.9 176-468 ~7-15 ( 2 9 9 . 8 ~11.3) NTM 20721 307-477 x 9-16 ( 4 0 2 . 0 ~12-8) 251-395 x 9-12 ( 2 9 4 . 4 ~10.2) 285-474 x 12-22 ( 3 5 6 . 2 ~16.2) 268-538 x 9-15 (355.2~12.0) 250-350 x 10-14 NTM 21260 NTM 21761 NTM 21845 Burton (1938) Holotype Absent Specimens Absent Absent Absent Absent Not mentioned 139-205 ~2.5-4 (161.3~2.8) 104-132 x 6-12 (120.6~8.4) 210-272 ~1.0-2.5 (234.3~1.6) 135-242 ~1.0-2.0 (195.5~1-5) 268-321 ~1.5-2.5 ( 2 9 4 . 2 ~1-9) 272-302 ~1.0-2.0 (233.4~1.6) 129-175 ~8-12 (158.3~10.5) 125-151 ~7-11 (141.0~9.2) 14&197 x 10-15 (179.4~12.6) 133-183 ~9-13 (150.8~10.8) 150 x 10 Not mentioned Distribution Unknown. Description Shape. Stipitate, arborescent sponge (95 mm high, 76 mrn broad), with thick basal attachment (9 mm diameter), and tightly anastomosing branches (10-21 mm diameter) which taper distally and lie mostly in 1 plane,. Colour. Beige in dry state. Oscula. Not observed. Texture and sugace characteristics. Surface in the dry state uneven and shaggy, whereas surface of live specimen was probably microconulose. Texture harsh, skeleton inflexible. Ectosome and subectosome. Ectosomal region hispid, with long thick extra-axial spicules protruding through surface originating from subectosomal region, surrounded at base of insertion through surface by brushes of robust ectosomal styles. Choanosome. Choanosomal skeleton slightly condensed axial core of horny spongin fibres forming tight reticulation, fully cored by choanosomal megascleres in pauci- and multispicular tracts. Major tracts run longitudinally, and secondary tracts ascend to surface in plumo-reticulate bundles. Axial skeleton occupies major proportion of branch diameter, up to 80% in places, being clearly differentiated from extra-axial skeleton; extra-axial skeleton with plumose tracts of subectosomal megascleres ascending to periphery, protruding through ectosome. Echinating acanthostyles predominant in peripheral skeleton. Megascleres. Choanosomal megascleres coring axial fibres slender strongyles (modified styles), usually slightly curved at centre, with evenly rounded tips and very slightly subtylote bases, [301-(355 2 ) 4 3 1 x 11-(12 8)-14 pm]. Subectosomal extra-axial spicules similar to axial megascleres, occasionally straight, but significantly longer and much more robust than strongyles in axis [632-(918 -4)-1320~17(26 -4)-32 pm]. - - Australian Raspailiidae Fig. 59. Endectyon xerampelina (Lamarck): a, choanosomal axial styles/strongyles; b, subectosomal extra-axial styles/strongyles; c, ectosomal auxiliary styles; d, echinating acanthostyles; e, section through peripheral skeleton; f , holotype (MNHN LBIM DT 574) (scale = 30 mm); g, structural spicules (scale = 200 pm). Ectosomal styles relatively long and thick, straight, with slightly subtylote bases, tapering to sharp points 13084360 8)-396 x 344.2)-5 pm]. Echinating acanthostyles typical for genus, with strongylote and clavulate modifications, and megascleres may range from subtylote with few spines to strongylote with clavulate spines on both ends [I 114121 -2)-137 x 8 4 1 0 -5)-12 pm]. Microscleres absent. - Remarks The geographic province of this species was not recorded by Lamarck (1814) or Topsent (1932), although Lamarck speculated that it could be American. It is included here as a possible member of the Australian sponge fauna, and in many features it resembles E. 1298 J. N. A. Hooper elyakovi. The most unusual feature about E. xerampelina is the modification of both axial and extra-axial megascleres to stongyles; in this respect the species may be distinguished from all others. Genus Trikentrion Ehlers Trikentrion Ehlers, 1870: 6, 31.--Carter, 1879b: 291; Hentschel, 1912: 373; Topsent, 1928: 58; de Laubenfels, 1936: 80; LBvi, 1973: 609. Tricentrium.-Ridley (in Zoological Record 16: 5). Plectronella Sollas, 18796: 17.--de Laubenfels, 1936: 81 (type species Plectronella papillosa Sollas, by monotypy). w e species: Spongia muricata (Pallas) Esper, 1794: 185 (by monotypy; holotype unknown, schizotype ZMB 7160, from West Africa) (Fig. 60a-d). Diagnosis Arborescent, digitate and flabellate growth forms. Surface usually hispid, and even or rough and microconulose. Choanosomal skeleton with poorly developed wide meshed reticulation, with only slightly condensed axial fibres running longitudinally through axial core. Fibres cored by pauci- or multispicular tracts of oxeas. Axial and extra-axial skeletons not markedly differentiated; extra-axial skeleton consists of ascending multior paucispicular primary fibres cored by choanosomal oxeas, interconnected by uni- or paucispicular secondary fibres, together producing a more-or-less regular reticulation. Single or brushes of subectosomal styles embedded at ends of ascending primary fibres, poking through peripheral skeleton. Peripheral fibres moderately heavily echinated by sagittal triacts. Ectosome with specialised skeleton of ectosomal styles in brushes surrounding bases of extra-axial styles. Structural megascleres styles of 2 sizes and choanosomal oxeas; echinating megascleres sagittal triacts (or acanthoplagiotriaenes) with only one spined ray. Microscleres may include raphides, occurring singly or in bundles (= trichodragmata). Remarks Trikentrion is similar to Cyamon in having highly modified echinating spicules called sagittal triacts, usually triradiate or further modified to quadriradiate and other forms. On this basis alone these genera could be eventually merged, but they differ in other significant details: Trikentrion has only one spined ray on echinating spicules, it has diactinal choanosomal spicules, a non-plumose skeleton, and a specialised raspailiid ectosomal skeleton. For those reasons they are maintained separately here (see also remarks for Cyamon). Four species are presently known for Trikentrion: T. muricata (Esper) from West Africa (Fig 6 0 ~ 4 T., helium Dickinson (1945: 15) from the Gulf of California (holotype AHF 7), T. laeve Carter (1879b: 294) from West Africa (holotype BMNH 1848.10.4.6, schizotypes BMNI-I 1954.3.9.127.129, MNHN LBIM DCL 36) (Fig. 60e-g) (not T. laeve var. flabelliforme Carter, 1882), and T. jabelliforme Carter (see below). Hallmann (1914b: 440) mentions another 'undescribed' Trikentrion from northwest Australia but he provides no further clues as to its identity or characteristics; it is likely that he is referring to T. flabelliforme, which is abundant in N.T. coastal waters. Plectronella was merged with Trikentrion in 1879 (by Ridley in the Zoological Record for that year), and again by de Laubenfels (1936), since he found that its type species was synonymous with T. muricata from West Africa. Trikentrion Jlabelliforme Carter (Figs 61, 62, 109h, 109i; Table 17) Trikentrion laeve var. flabelliforme Carter, 1882: 114. Not Trikentrion laeve Carter, 1879b: 294. Trikentrion flabelliforme.-Hentschel, 1912: 373-374, pl. 13, fig. 9, pl. 20, fig. 32; Capon et al., 1986: 6545-6550. ? Trikention sp.-Hallmann, 1914b: 440. Australian Raspailiidae 1299 Fig. 60. a-d, Trikentrion muricata (Pallas) (type species of the genus Trikentrion Ehlers): a, specimen (BMNH 1872.10.19.1) from the River Volta, West Africa (scale = 30 mm); b, structural megascleres (scale = 250 pm); c, sagittal diact (scale = 100 pm); d , skeletal structure (scale = 1 mm). e-g, T. laeve Carter: a, specimen (BMNH 1939.2.20.9) from Pointe Noire, Belgian Congo (scale = 30 mm); f , skeletal structure (scale = 500 pm); g, sagittal diact (scale = 100 pm); h, structural spicules (scale = 200 pm). Material Examined Lectotype. BMNH 1887.5.21.1865 (dry): precise locality unknown, 'SW. Australia', Bowerbank collection (one of two known syntypes; hypotype of Trikentrion laeve var. jlabelliforme Carter). Paralectotype. SMF (not located). Other material (all material collected by the author using SCUBA, unless 0 t h e ~ i s eindicated). Indonesia: BMNH 1931.8.4.57: Jedan, E coast of Aru I., Indonesia, coll. 'Siboga', stn SE BY. AM G580: Precise locality unknown, NW. Australia, coll. unknown (labelled 'Trikentrion laeve Carter, 1879'). Darwin Region, N.T.: NTM 21041, 1052, 1061: EPMFR, 12" 24.7'S., 130" 48/E., 13 m depth, 9.xi.1982 (stn EP11). NTM 22247, 2248: same locality, 12' 24.5/S., 130' 48 O'E., 10 m depth, 12.iv.1985, coll. C. Hood & J.R. Hanley (stn EP22). NTM 22383, 2384, 2396, 2397: same locality, 8 m depth, 29.vii.1985 (stn EP23). NTM 22618, 2639, 2672, 2692, 2705, 2711, 2712: same locality, 9-12 m depth, 3.iv.1986 (stn EP28). NTM 21944, 1949, 1953, 1960, 1964: Stephen's J. N. A. Hooper 1300 - Rock, Weed Reef, 12" 29 2'S., 130' 47.1 'E., 12 m depth, 27.iv.1984 (stn WR1). NTM 22029, 2035: W. side of Weed Reef, 12" 29.2' S., 130" 47 1' E., 19 m depth, 11.v.1984 (stn WR2). NTM 22168, 2169: 'Bommies', N. edge of Weed Reef, 12" 29.2'S., 130" 37-6'E., 8-10 m depth, 5.x.1984 (stn WR5). NTM 2827, 839: S. side of Channel I., Middle Arm, 12" 33.8'S., 130" 51.4'E., 20 m depth, 18.vii.1982, coll. S. Chidgey (Channel I. EIS (FN A9), stn Don. 24, hooker). NTM 2849, 859, 871: vicinity of pearl oyster beds, Channel I., Middle Arm, 12'32.7'S.. 130" 52.5'E., NTM 2232: Indian 12-13 m depth, 20.viii.1982, coll. P.N. Alderslade & P. Homer (stn C13, SCUBA). I., Bynoe Harbour, 12' 35/S., 130" 33/E., 2 m depth, 18.xi.1981, coll. P. Byers (M 'Skeleton', snorkel). NTM 21071: S. reef, Fish Reef, Bynoe Harbour, 12" 26.2'S., 130" 26-2/E., 10-12 m depth, 24.xi.1982 (stn FR1, SCUBA).NTM 22104: N. side of Fish Reef, Bynoe Harbour, 12' 26.2'S., 130" 26.2/E., 9-10m depth, 5.ix.1984 (stn FR2). NCI Q66C-0530-Q (fragment NTM 23083): Pany Shoals, Timor Sea, N.T., 11" 11-411S., 129' 43.01 IE., 18 m depth, 13.viii.1987, coll A.M. Mussig (stn Don. 248/AM 87-3). Cobourg Peninsula Region, N.T.: NTM 2108, 125: Sandy I. No. 2, 11' 05.5'S., 132" 17'E., 7 m depth, 20.x.1981 (stn CP26). NTM 2143: same locality, 10 m depth, 21.x.1981 (stn CP27). NTM 22525: Orontes Reef, mouth of Port Essington, 11" 03.6' S., 132' 05.4'E., 18-20 m depth, 16.ix.1985 (stn CP78). NTM 22528 (6 specimens): same locality, 20m depth, 17.ix.1985 (stn CP80). Wessel Is, N.T.: NTM 23913: NE. tip of Wigram I., English Company Is, 11" 44.4/S., 136'37-83/E., 20 m depth, 12.xi.1990 (stn WI-3). NCI Q66C 4783-P, 4780-M: N. side of Cumberland Strait, 11" 27.6'S., 136" 28.7'E., 30 m depth, 15.xi.1990, coll. NCI. Northwest Shelf Region, W.A.: NTM 21018: NW. of Dampier Archipelago, 20' 10IS., 117" 3g1E., 35 m depth, 20.x.1982, coll. J Blake (CSIRO RV 'Soela' shot 48, stn Don. 28, trawl). NTM 21033: same locality, 20' 201S., 117' 28/E., 26 m depth, 22.x.1982 (CSIRO RV 'Soela' shot 64, stn Don. 35). NTM 2727: N. of Adele I., Collier Bay, 15' 58.3's.. 122' 39.7/E., 59 m depth, 21.iv.1982 (CSIRO RV 'Sprightly' SP4/82, stn 40-Don. 20, dredge). NTM 21126: W. of Port Hedland, 19' 04.3'S., 119" 01 llE., 83 m depth, 19.iv.1983 (CSIRO RV 'SoeIa' S02/83, stn Bll-NWS1, beam trawl). NTM 21232, 1265, 1266: same locality, 19" 28.5'S., 118" 55.3/E., 40 m depth, 26.iv.1983 (stn B9-NWS10). NTM 21804, 1811, 1832: same locality, 19" 26 .9'S., 118" 54.2/E., 50 m depth, 30.viii.1983, coll. T. Ward (CSIRO RV 'Soela' S04/83, stn 126-NWS26, trawl). NTM 22307: NW. of Lacepede I., 16' 29-33IS., 121' 27-29'E., 3 8 4 0 m depth, 17.iv.1985, coll. B.C. Russell (stn PT 85-4-NWS34, Taiwanese Pair Trawl). NTM 22975: Due W. of Camarvon, W.A., 24" 55 6-56 5 S., 112" 50.8-53 5 ' E., 80-85 m depth, 14.vii.1987 (U.S.S.R. RV 'Akademik Oparin', stn SB6, beam trawl). NTM 23339, 3342: Fringing reef, 200 m from Direction I. National Park, W.A., 21" 32. OIS., 115" 07 -2'E.. intertidal, 24.viii.1988, coll. D. Low Choy (stn NWS61/DLC9). Scott Reef, W.A.: (all material collected by V. Krasochin, U.S.S.R. RV 'Akademik Oparin', dredge). PIBOC 012-115, 119: near Scott Reef, 16'32.2/S., 121" 10.glE., 4 3 4 m depth, 4.xi.1990 (stn 28). PIBOC 012-145: 16" 20.01S., 121' 12.5/E., 54 m depth, 4.xi.1990 (stn 28). PIBOC unreg.: 16' 36.7'S., 121' 11. 1' E., 50 m depth, 17.xi.1990 (stn 39). PIBOC unreg.: 16'44.5'S., 121" 14-6/E., 42-48 m depth, 4.xi.1990 (stn 27). PIBOC 012-145: 16" 20.01S., 121" 12-5'E., 54 m depth, 4.xi.1990 (stn 30). - - Substrate and Depth Range Subtidal and shallow offshore rock and coral rubble reefs, 3-82.5 m depth. Geographical Distribution Confirmed distribution extends northwards from Exrnouth Gulf, W.A., to Aru I., Indonesia (Fig. 61g). The lectotype in Bowerbank's collection at the BMNH was supposed by Carter (1882) to come from SW. Australia, but there was apparently no label attached to that specimen, and this distribution is unconfirmed. Description Shape. Growth form varies from very thin to moderately thick flabelliform fans (65-260 rnm high, 38-190 mm wide, 2-14 mm thick) to flattened cylindrical digits (up to 15 rnm wide), with bifurcate or even margins, usually on short cylindrical stalk (15-55 mm long, 5-13 mm diameter). Colour. Shallow water specimens 1 of 2 distinctive colour morphs: predominant form orange-red to blood red alive (Munsell 10R 5110-5R 4/10) (Fig. 109h); other form evenly beige ( 7 - 5 R 812) or light orange-brown (5YR 814) in life (Fig. 109i). Deeper water material invariably beige; in ethanol coloration beige to light brown. Australian Raspailiidae Fig. 61. Trikentrion flabelliforme Carter (specimen NTM 22384): a , choanosomal axial oxea; b, subectosomal extra-axial style; c, ectosomal auxiliary style; d, echinating sagittal mono-, di- and triacts; e, raphides and trichodragmata; f , section through peripheral skeleton; g, known Australian distribution. - Oscula. Minute oscula (approximately 0 5 mm diameter) dispersed evenly over both sides of fans, but these only observed on silt-covered specimens in situ, made visible by several small silt-free drainage canals radiating from each osculum. Texture and surface characteristics. In shallow water material surface in situ optically smooth, usually silt covered, microscopically hispid, with minimal subdermal sculpturing, usually without ridges or conules in flabelliform material, whereas in digitate material surface is slightly raised into (few) irregularly scattered conulose projections. Texture firm, compressible, branches easily flexible but tough. Ectosome and subectosome. Ectosomal skeleton consists of plumose brushes of thin, flexuous ectosomal styles (or anisoxeas) surrounding bases of protruding extra-axial megascleres. Dermal brushes combine to form more-or-less continuous palisade on surface, seen only microscopically, whereas protruding subectosomal extra-axial styles make surface 1302 J. N. A. Hooper optically hispid. Subectosomal extra-axial region relatively plumose or plumo-reticulate; long extra-axial styles embedded in peripheral ends of ascending multispicular choanosomal tracts, composed of axial oxeas and echinating megascleres. Ascending radial subdermal tracts interconnected at more-or-less regular intervals by uni- or paucispicular tracts of choanosomal megascleres; skeletal anastomoses produce more-or-less cavernous subdermal, rectangular or square meshes. Echinating sagittal triacts also contribute to dermal spicule brushes, with long (spiny) axis of triacts usually protmding into ectosomal skeleton, and these spicules also scattered along (echinating) extra-axial tracts. Subdermal skeletal tracts enclosed within very light spongin fibres, whereas mesohyl matrix much heavier in periphery. Within subdermal mesohyl heavy deposits of microxeas scattered singly or in trichodragmas. Choanosome. Choanosomal skeletal architecture only slightly condensed in axis, but with well differentiated axial and extra-axial skeletons. Axial core has significantly heavier spongin fibres than periphery, and fibres enclose multispicular tracts of choanosomal oxeas, forming almost' perfectly regular fibre reticulation, superficially resembling a Petrosia. Fibre reticulation produces square meshes, with spiny arms of echinating sagittal triacts protruding into oval choanocyte chambers lying between fibre meshes. Mesohyl matrix relatively light (compared with peripheral skeleton), but microxeas (and trichodragmata) common in axis. Megascleres (refer to Table 17 for dimensions). Choanosomal axial megascleres thick oxeas, usually greatly curved at centre, symmetrical, rarely straight, with slightly fusiform tips. Subectosomal extra-axial styles long, relatively thin, slightly curved at centre or straight, tapering to very fine points and with rounded non-tylote bases. Ectosomal auxiliary megascleres raphidiform, straight or flexuous, occasionally sinuous, with hair-like point and more-or-less stylote base only slightly thicker than apex, occasionally anisoxeote. Echinating megascleres sagittal triacts (or acanthoplagiotriaenes), with asymmetrical rays; longest ray heavily spined usually with recurved hooks at apex, gradually diminishing towards distal end of spicules. Short axes of spicules consist of 1 (crooked, asymmetrical), 2, or even 3 rays, usually approximately equal in length. Microscleres: raphides (or microxeas) straight, long, often forming bundles (trichodragmata). Associations This species is remarkable in often being very heavily infested by a white zoanthid, and in the Darwin region at least, that infection may be related to (as yet unknown) seasonal events (e.g. freshwater dilution of estuaries, algal blooms, etc.). Approximately 30% of specimens collected from this region were infected. The infection is historically well known for Trikentrion, and the parasitic polyp involved has been recorded as Bergia (e.g. Carter 1882: 115) or Palythoa (e.g. Carter 1879b: 295). The zoanthid incorporates oxeas from the sponge into the velum, and these spicules are orientated in line with opercular rays, presumably used as skeletal support. Another even more interesting chemical exchange appears to be taking place within this sponge, and it is possible that these events are related to the sponge-zoanthid association. Trikentrion flabelliforme contains two very unusual alkaloid secondary metabolites, recently named 'trikentrins' (Capon et al. 1986), which show significant growth inhibitory activity against Bacillus subtilis. It is speculated that those alkaloids may be produced by the sponge in response to zoanthid infections (so-called chemical defence). However, recent work has established that the polyps appear to tolerate the trikentrins without chemical modification (R.J. Capon, personal communication), and their role in chemical defence, if any, is still uncertain. Nevertheless, it would be interesting to investigate whether the production of trikentrins varies seasonally, and whether any variability present correlates with the distribution and abundance of zoanthids. Remarks This species is a prominent member of the shallow subtidal and offshore coastal water fauna in this region, and with its bright coloration and thin (mostly) flabelliform habit it is Australian Raspailiidae Fig. 62. Trikentrion Jabelliforme Carter: a, flabellifom lectotype of T. laeve Carter var. Jlabellifonne Carter (BMNH 1887.5.21.l865); b, flabelliform specimen (NTM 22711); c, flabello-digitate specimen with zoanthids (BMNH 1931.8.4.57); d, digitate specimen (NTM 21804) (scale bar = 30 mm); e, skeletal structure (scale = 1 mm); f,g, echinating sagittal diand triacts (scale = 100 pm); h, SEM of axial skeleton (left magnified 93 times, right magnified 600 times). 1304 J. N. A. Hooper easily recognised in the field (albeit occasionally confused with a few thinly flabelliform Phakellia species). In its skeletal architecture, surface features, spiculation and distribution of megascleres the species has obvious relationships with the Raspailiidae, although the genus has been traditionally associated with the nominal family Euryponidae. The species is easily differentiated from other Trikentrion species in having a longer acanthose ray than other rays on the sagittal triact spicules. Three specimens collected from the Wessel Is region, N.T., lacked sagittal triacts altogether, but were otherwise identical in other features. Genus Cyamon Gray Cyamon Gray, 1867: 546.-Dendy, 1922: 107; de Laubenfels, 1936: 80; LCvi, 1973: 609; Sirn & Bakus, 1986: 17. 5 p e species: Dictyocylindrus vickersii Bowerbank, 1866: 267 (by monotypy; holotype BMNH 1877.5.21.1887, from the West Indies) (Fig. 63a-e). Diagnosis Encrusting to massive growth form. Surface hispid and usually conulose. Choanosomal skeleton basally condensed layer of spongin fibres lying on substrate, with microcionid-like plumose spongin-fibre nodes ascending to surface. Basal fibres cored by pauci- or multispicular plumose tracts of choanosomal styles, with no or few anastomoses. Extra-axial skeleton with few very long thin extra-axial styles, embedded in choanosomal fibre nodes and protruding through surface. Extra-axial spicules also scattered throughout mesohyl. Fibres very heavily echinated by sagittal triacts, tetracts, or forms even further modified, producing almost rigid interlocking secondary skeleton. Ectosomal specialisation absent. Structural megascleres styles or subtylostyles of 2 sizes; thinner forms may have few spines and subtylote swellings on distal and basal ends respectively; echinating spicules sagittal tetracts or pentactinal megascleres (acanthoplagiotriaenes), less commonly diacts or triacts, with all (s.s.) or at least more than 1 acanthose ray. Microscleres absent. Remarks Bowerbank's (1866) original description of C. vickersii is inadequate and barely describes the genus, but Carter's (1879b: 292) redescription of the type specimen is much more detailed. The BMNH holotype is dry but still in good condition (Fig. 63a), and from that material it is confirmed that skeletal structure consists of plumose columns of choanosomal and subectosomal styles, there is no ectosomal skeleton, and there is an interlocking secondary skeleton of sagittal echinating spicules (Fig. 63c). This structure is quite different from Trikentrion, but it is probably related to the growth form of Cyamon, and the character may not stand up at the generic level. Nevertheless, there are also significant differences in the geometry and spination of sagittal echinating megascleres between the two genera: Trikentrion has monact-, diact-, triact- or rarely tetractinal sagittal spicules, with only one spined ray; Cyamon (s.s.) has tetract- or pentactinal sagittal spicules, rarely with fewer rays, and most, if not all, rays are spined. Dendy (1922: 108) and de Laubenfels (1936: 80) also redescribed additional material of C. vickersii from Amirante (BMNH 1931.1.1.19; Fig. 63b) and the West Indies, respectively, although de Laubenfels (1936) questioned the conspecificity of Dendy's material. Nevertheless, from those accounts it appears that Cyamon differs from Trikentrion, and the two genera are maintained here. Seven species of Cyamon are presently known: C. vickersii (Bowerbank) (Fig. 63a-e), C. neon de Laubenfels, 1930 from southern California, C. argon Dickinson, 1945: 15 from the Gulf of California, C. quinqueradiata (Carter, 1880: 43) from the Gulf of Manaar, C. koltuni Sirn & Bakus, 1986: 18 from California, C. catalina Sim & Bakus, 1986: 18 from California and C. aruense Hentschel, 1912: 374 (Figs 63-64), but only the latter species has been recorded previously from waters in the vicinity of Australia. Australian Raspailiidae Fig. 63. a-e, Cyamon vickersii (Bowerbank) (type species of the genus Cyamon Gray): a, holotype (BMNH 1877.5.21.1887); b, specimen (BMNH 1931.1.1.19) (scale = 30 mm); c, skeletal structure (scale = 1 mm); d, echinating sagittal triact with spined rays (scale = 100 pm); e, structural megascleres (scale = 200 pm). f-i, C. aruense Hentschel: f , holotype (SMF 1618) (scale = 30 mm); g, skeletal structure (scale = 250 pm); h, structural megascleres (scale = 100 pm); i, echinating sagittal triact with spined rays (scale = 50 pm). Cyamon aruense Hentschel (Figs 63, 64) Cyamon aruense Hentschel, 1912: 374. Material Examined Holotype. SMF 1618: Straits of Dobo, Am I., Arafura Sea, 6" S, 134' 501E., 40m depth, 20.iii.1908 (schizotypes MNHN LBIM DCL 2330). Substrate and Depth Range Limestone reef, 40 m depth. Geographical Distribution Arafura Sea (Fig. 64e). J. N. A. Hooper Fig. 64. Cyamon aruense Hentschel (holotype S M F 1618): a, choanosomal axial subtylostyle; b, subectosomal extra-axial styles; c, echinating sagittal quatract; d, section through skeleton; e, known Australasian distribution. ' Description Shape. Thickly encrusting on Phloeodictyon (up to 400 pm thick, 60 mm long, 30 mm wide). Colour. Beige in ethanol. Oscula. Not observed. Texture and suflace characteristics. Holotype friable and detachable from substratum. Surface hispid and even. Ectosome and subectosome. Ectosomal skeleton membraneous, lacking specialised spiculation. Subectosomal extra-axial styles lie at oblique angles near surface; longer more-erect examples protrude through ectosome, although most in holotype are broken off at level of surface. Due to relatively poor condition of holotype it is not possible to determine whether extra-axial megascleres standing perpendicular to substrate are embedded in basal spongin layer or merely perched over axial spicules. Acanthose rays of sagittal megascleres also poke through dermal skeleton, but never protruding far beyond surface. Choanosome. Choanosomal skeleton basal layer of spongin lying on substrate, with oxeas from host sponge (i.e Phloeodictyon) embedded below ground substance. Bases of large axial subtylostyles embedded in basal spongin, these also protruding a long distance through ectosome. Choanosomal skeleton dominated by an interlocking skeleton of sagittal echinating megascleres, with longer ray pointing outwards. Spongin in mesohyl heavy and granular. Megascleres. Choanosomal subtylostyles of axial skeleton long, thick, slightly curved near base, with subtylote bases and tapering to sharp points [780-(1040.1)-1373 x 12(17.3)-22 pm]. Subectosomal extra-axial subtylostyles vary considerably in length and thickness, but no justification in distinguishing 2 size categories; spicules relatively thin, usually curved towards basal end, with subtylote bases, tapering to sharp points [185-(366.0)-570 x 2 4 6 8)-11 pm]. Ectosomal auxiliary spicules absent, although extra-axial megascleres lie close to surface which might be construed as dermal spicules. Echinating sagittal spicules (acanthoplagiotriaenes) tetract-, pentact- or hexactinal (no fewer rays observed); rays more-or-less evenly spined, although longer ray with slightly more recurved spines than other rays 1.55472 2)-88 x 6-(9 3)-11 pm]. Microscleres absent. - - - Australian Raspailiidae 1307 Remarks Hentschel (1912) noted that in addition to subectosomal extra-axial spicules in this species there was another category of auxiliary style, and by implication (and geometry of his figured spicules) these might comprise a category of ectosomal megascleres. This feature was not observed in the holotype, and it must be assumed that specialised ectosomal spiculation is not present (the thinly encrusting holotype is not in good enough condition to provide a definitive opinion). This species differs from other Cyamon in having all rays of sagittal spicules heavily and evenly spined. Genus Aulospongus Norman Aulospongus Norman, 1878: 267.-Dendy, 1889: 89; 1922: 61; Burton, 1938: 38. Not Aulospongus.-de Laubenfels, 1936: 100. Aulospongiella Burton, 1956: 141 (type species Axinella monticularis Ridley & Dendy, 1886: 481, by original designation and monotypy; holotype BMNH 1887.5.2.20, from Cape Verde I., North Atlantic) (Fig. 65a-c). Heterectya Hallmann, 1917: 393 (type species Raspailia (?) villosa Thiele, 1898: 60, by original designation; holotype ZMB 2204 from Hakodate, Japan) (Fig. 65d-f). Rhaphidectyon Topsent, 1927: 15 (type species Rhaphidectyon spinosum Topsent, 1927: 15, by original designation and monotypy; holotype MOM, schizotypes MNHN LBIM DT 1139, BMNH 1930.7.1.39, from St. Vincent I., North Atlantic) (Fig. 65g-11. Hemectyonilla Burton, 1959: 254 (type species Stylostichon involutum Kirkpatrick, 1903: 250; by original designation and monotypy; holotype BMNH 1902.11.16.33, from Natal, Indian Ocean) (Fig. 66a-f). Type species: Haliphysema tubulatus Bowerbank, 1873: 29 (by original designation; holotype BMNH 1873.7.2 1.9 from Ceylon, Indian Ocean) (Fig. 66g-k). Diagnosis Clathrous, cylindrical, spherical, lobate, cupriform and digitate growth forms, composed of fused bundles of plumose fibres; surface more-or-less hispid and microconulose, produced by protruding ends of each spongin-fibre bundle. Choanosomal skeleton greatly reduced or lacking any axial condensation; axial and extra-axial differentiation also reduced. In central axis choanosomal megascleres aggregated into longitudinal plumose columns, loose bundles axial fibres, or branching microcionid-like plumose fibres. Extra-axial skeleton without specialised subectosomal spiculation, merely consisting of radial, vertically ascending plumose columns of choanosomal rhabdostyles, which do not anastomose but diverge towards periphery; extra-axial skeletal columns echinated by heavy plumose tracts of acanthorhabdostyles (or acanthostyles). Ectosome with (nominal genus Hemectyonilla) or without (s.s.) specialised skeleton; when present ectosomal spiculation consists of brushes of very long, slender oxeas arising from ends of extra-axial spicule columns; when ectosomal megascleres absent choanosomal rhabdostyles may protrude through surface in bundles. Structural megascleres include smooth rhabdostyles, styles, or subtylostyles of a single category, with or without auxiliary ectosomal oxeas, anisoxeas or more rarely styles; echinating acanthostyles with smooth rhabdose bases and distally spined shafts, or entirely spined (nominal genus Aulospongiella). Microscleres absent, or may include bundles of raphides (nominal genus Rhaphidectyon). Remarks Aulospongus (s.s.) is atypical of Raspailiidae in having an exclusively plumose rather than a condensed reticulate skeleton. It also lacks any marked differentiation between the axial and extra-axial skeletons, or axial compression. These features are most obvious in the type species and can be seen clearly in Dendy's (1922) specimen from Amirante, Indian Ocean (BMNH 1931.11.28.18; Fig. 66h). Other species, such as A. clathrioides U v i (1967: 21; holotype from New Caledonia, MNHN LBIM DCL 823; Fig. 67a-d), have the plumose extra-axial skeletal structure typical of the genus, but also show rudiments of a 1308 J. N. A. Hooper slightly compressed axial skeleton (Fig. 67b). This provides some clues to the raspailiid affinities of the genus, but this relationship is not completely obvious. The growth form of fused tubular fibre bundles (e.g. Fig. 66g-h) is quite unusual amongst Raspailiidae, but it is well known amongst certain Microcionidae [e.g. Microciona microjoanna de Laubenfels (holotype USNM 21469), M. parthena de Laubenfels (holotype USNM 21383), and in particular Clathria toxipraedita Topsent (holotype RSME 1921.143.1400) and C. rhaphidotoxa Stephens (RSME 1921.143.145 1) (Hooper, unpublished data)]. In fact, Aulospongus sensu de Laubenfels (1936: 100) (viz. A. schoenus de Laubenfels; holotype USNM 22404) is a microcionid sponge that Simpson (1968: 56) referred to the genus Thalysias, whereas Aulospongus s.s. belongs to the Raspailiidae. Four nominal genera are synonymised here with Aulospongus. Aulospongiella is bulbousencrusting, with rnicrocionid skeletal architecture composed of plumose non-anastomosing spicule tracts (Fig. 65a-c), without any trace of axial compression or axial and extra-axial differentiation. It is differentiated from most raspailiids in having acanthostyles with evenly spined rhabdose bases, unlike typical Raspaxilla, such as R. compressa (Fig. 33e), which have smooth rhabdose bases. In Aulospongiella these spicules are more or less incorporated into skeletal tracts, along with foreign particles. However, neither character is unique to the family, and the species is considered here to be a synonym of Aulospongus. Interestingly, several species of microcionids are also known to incorporate their echinating megascleres into skeletal tracts ('phorbasijormis' species complex, e.g. Clathria myxilloides Dendy, C. dura Whitelegge) (Hooper, unpublished data). Although the inclusion of detritus into the skeleton is relatively common amongst the Microcionidae (e.g. nominal genera Clathriopsamma and Aulenella), rhabdose aca&hostyles have not been-recorded for that group. Hallmann's (1917) genus Heterectya differs from the typical form of Aulospongus only in having choanosomal styles with rhabdose bases. Thiele (1898) was in error in suggesting that its type species, Raspailia villosa (Fig. 65d-f), had spined choanosomal megascleres (which in any case agrees with the definition of the nominal genus Hernectyonilla). Thus it appears that Hallmann erected Heterectya on spurious grounds. Both nominal genera (Aulospongus and Heterectya) have plumose skeletons, no notable axial compression, no specialised ectosomal or extra-axial megascleres, and echinating acanthostyles have rhabdose bases. Similarly, Rhaphidectyon is also merged with Aulospongus since it too is an atypical raspailiid with a plumose choanosomal axial and extra-axial skeleton, without specialised ectosomal or subectosomal spiculation, it differs from typical Aulospongus only in having raphide microscleres and lackkg the tubular growth form characteristic of typical Aulospongus (Fig. 65g-i). Hemectyonilla has rhabdose choanosomal and echinating megascleres that have smooth bases and acanthose points, and that form plumose tufts along radial microcionid-like plumose skeletal tracts. Like typical Aulospongus, Hemectyonilla species also lack a specialised category of subectosomal extra-axial spicule, but in Hernectyonilla there are extremely long and thin ectosomal megascleres present, much larger than those found in other raspailiids. The absence of axial compression and axial and extra-axial differentiation is atypical of Raspailiidae, and in this respect the genus is identical to Aulospongus. In fact Burton's (1959) specimen of H. involuturn from the Arabian Gulf (BMNH 1936.3.4.118, fragment MNHN LBIM DCL 61) even resembles A. tubulatus in growth form (Fig. 66b), being quite different from the typical growth form of H. involutum (Fig. 66a), whereas skeletal architecture and spiculation of both Hemectyonilla specimens is identical (Fig. 66c-f), and there is no doubt that this material is conspecific. The major differentiating feature between Hemectyonilla and typical Aulospongus is the presence of distal spination on choanosomal megascleres, the presence of a specialised ectosomal skeleton, and-the lack of a specialised tubular growth form. By these features Hemectyonilla s.s. is more distantly related to Aulospongus s.s. than either of Heterectya or Rhaphidectyon, but it is considered here that these differences do not provide sufficient reason to maintain the two genera separately. Another species referred to Hemectyonilla, and which should also be included in Aulospongus, is Plumohalichondria gardineri Dendy (1922: 87) from Amirante (holotype Australian Raspailiidae Fig. 65. a-c, Aulospongus monticularis (Ridley & Dendy) (type species of the nominal genus Aulospongiella Burton): a, echinating acanthostyle (scale = 25 pm); b, section through peripheral skeleton (scale = 50 pm); c, holotype (BMNH 1887.5.2.20). d-f, A. villosum (Thiele) (holotype Z M B 2204) (type species of the nominal genus Heterectya Hallmann): d, skeletal structure (scale = 500 pm); e, fibre characteristics (scale = 250 pm); f , echinating rhabdostyle (spined) and rhabdose structural styles (smooth) (scale = 100 pm). g-i, A. spinosum (Topsent) (type species of the nominal genus Rhaphidectyon Topsent) (schizotype MNHN LBIM DT 1139): g, skeletal structure (scale = 1 mm); h, fibre characteristics (scale = 500 pm); i, structural spicules (scale = 200 pm). BMNH 1921.11.7.74, schizotype MNHN LBIM DCL 1452) (Fig. 67e-h), which Burton (1959) erroneously synonymised with the type species from Natal. Australian Species None. Genus Raspaciona Topsent Raspaciona Topsent, 1936: 49.4a13, 1958: 254; Vacelet, 1961: 36; Pulitzer-Finali, 1977: 41. Type species: Halichondria aculeata Johnston, 1842: 131 (by original designation; holotype BMNH 1877.5.21.956, schizotype MNHN LBIM DNBE 320L, from the Mediterranean) (Fig. 68a-b). J. N. A. Hooper a-f, Aulospongus involutum (Kirkpatrick) (type species of the nominal genus Hemectyonilla Burton): a, holotype (BMNH 1902.11.16.33) (scale = 30 mm); b, specimen (BMNH 1936.3.4.118) from the Indian Ocean (scale = 30 mm); c, skeletal structure of holotype (scale = 1 mm); d, fibre characteristics of specimen (scale = 250 pm); e-f, echinating and stmctural spicules of holotype (scale = 200 pm). g-k, A. tubulatus (Bowerbank) (type species of Aulospongus Norman): g, holotype (BMNH 1873.7.21.9) (scale = 30 mm); h, specimen (BMNH 1931.11.28.18) from the Indian Ocean; i, plumose skeletal structure (scale = 1 mm); j, structural rhabdostyle (scale = 100 pm); k, echinating acanthorhabdostyle (scale = 50 pm). Fig. 66. Diagnosis Lobo-digitate, or ramose-bushy growth forms; surface hispid and prominently rnicroconulose. Choanosomal skeleton plumose, consisting of basally condensed spongin fibres ascending through branches and diverging, but without any special category of choanosomal megasclere. Extra-axial skeleton with weakly developed, ascending, plumose columns of subectosomal styles protruding through ectosome, branching or forming occasional anastomoses, and forming shaggy surface microconules. Ascending extra-axial tracts echinated by acanthostyles, concentrated around bases of plumose skeletal columns. Spongin predominant in basal Australian Raspailiidae 1311 Fig. 67. a-d, Aulospongus clathrioides Lkvi: a, holotype (MNHN LBIM DCL 823) (scale = 30 mm); b, skeletal structure (scale = 1 mm); c, structural rnegascleres (scale = 100 pm); d, echinating acanthorhabdostyle (scale = 50 pm). e-h, A. gardineri (Dendy): e, holotype (BMNH 1921.11.7.74) (scale = 30 mm); f , echinating and structural acanthorhabdostyles (scale = 100 pm); g, skeletal structure (scale = 1 mm); h, fibre characteristics (scale = 1 mm). region, at point of contact between plumose (extra-axial) tracts and condensed (axial) fibres; very little spongin elsewhere in skeleton. Ectosomal skeleton with specialised spiculation of small styles arranged in dermal brushes around protruding extra-axial styles. Structural megascleres long and short styles or subtylostyles of 2 sizes, and 1-2 size classes of acanthostyles ranging from basally spined to entirely spined. Microscleres absent. Remarks There are currently three species assigned to this genus, all from the Mediterranean: R. aculeata (Johnston) (Fig. 68a-b), R. robusta Sarh and R. calva SarL Raspaciona has a spiculation and ectosomal structure typical of Raspailiidae, but a plumose choanosomal and subectosomal architecture reminiscent of Hymeniacidonidae (e.g. Ulosa). Variability in skeletal morphology between specimens of different growth forms has been well illustrated 1312 J. N. A. Hooper -- Fig. 68. a-b, Raspaciona aculeata (Johnston) (schizotype MNHN LBIM DNBE 320L) (type species of Raspaciona Topsent): a, plumose skeletal structure (scale = 500 pm); b, ectosomal features (scale = 200 pm). c d , Hymeraphia stellifera Bowerbank (specimen MNHN LBIM DT 2501) (type species of the genus Hymeraphia Bowerbank): c, skeletal structure (scale = 500 pm); d, structural and echinating spicules (scale = 200 pm). e-f, Eurypon pilosella (Topsent) (schizotype MNHN LBIM DT 933) (type species of the nominal genus Acantheurypon Topsent): e, hymedesmoid skeletal structure showing the unusually long extra-axial skeletal spicules (scale = 500 pm); f , ectosomal and basal skeletal structure (scale = 200 pm). g, E. viridis (Topsent) (holotype MNHN LBIM DT 1838) (type species of the nominal genus Tricheurypon Topsent), skeletal structure (scale = 200 pm). h-k, E. cactoides (Burton & Rao) (schizotype BMNH 1931.1.1.52) (type species of the nominal genus Protoraspailia Burton & Rao): h, skeletal structure (scale = 1 mm); i, schizotype (scale = 30 mm); j, trichodragmata (scale = 10 pm); k, echinating acanthostyle (scale = 100 pm). I-m, E. polyplumosa (L6vi) (holotype MNHN LBIM DCL 1296) (type species of the nominal genus Proraspailia Levi): I, structural and echinating megascleres (scale = 100 pm); m, skeletal structure (scale = 1 mm). Australian Raspailiidae 1313 by Topsent (1925) and Pulitzer-Finali (1977). Those authors showed that megasclere size and ornamentation varied quite considerably, and it has been suggested that Sari's (1958) R. robusta and R. calva are merely different morphs of R. aculeata (e.g. Vacelet 1961). However, the acanthostyles of R. calva are illustrated with swollen tylote bases, much the same as those found in Hymeraphia species (Sari 1958: fig. 21). Although Raspaciona lacks any remarkable characters that can distinguish it readily from other genera, it can be differentiated from other plumose species (e.g. of Aulospongus and Tethyspira) by its Raspailia-like spiculation, and from the typical Raspailia condition by its strictly plumose structure. Australian Species None. Genus Hymeraphiu Bowerbank Hymeraphia Bowerbank, 1864: 189. Not Hymeraphia.-Hentschel, 1912: 377. Mesapos Gray, 1867: 543 (type species: Hymeraphia stellifera Bowerbank, 1864: 189, by monotypy). Type species: Hymeraphia stellifera Bowerbank, 1864: 189 [by original designation; holotype BMNH 1877.5.21.460 (l9lO.l.l.87), from the English Channel] (Fig. 68c-d). Diagnosis Encrusting growth form. Choanosomal skeleton reduced to basal membrane lying on substrate; bases of echinating acanthostyles and extra-axial styles embedded in basally condensed fibres, in radial or hymedesmoid arrangement, standing perpendicular to substrate, not grouped into brushes or other structures. Choanosomal megascleres absent. Extra-axial styles usually protrude a long way through ectosome. Ectosomal megascleres undifferentiated from extra-axial styles, the latter lying tangentially or paratangentially to surface. Structural megascleres consist of a single category of extra-axial styles or tylostyles; echinating megascleres are acanthostyles with prominently swollen tylote bases (those of type species also have modified stellate-acanthose points). Microscleres absent. Remarks Hymeraphia shows some similarities to the nominal Microcionidae genus Microciona (= Clathria), but has only single spicules forming the choanosomal skeleton rather than plumose columns of spicules (Fig. 68c), and so is closest to the microcionid Leptoclathria (=Clathria) condition. Hymeraphia lacks differentiated choanosomal, subectosomal or ectosomal megascleres (i.e. it has only a single category of structural spicules), has modified acanthostyles, and lacks microscleres. Hymeraphia sensu Hentschel (1912) is a microcionid, with chelae, toxas, and ectosomal megascleres. The presence of peculiarly modified acanthostyles (Fig. 6 8 4 and lack of microscleres is therefore the only morphological feature that can reliably differentiate Hymeraphia and thinly encrusting microcionids such as Leptoclathria. Other thinly encrusting raspailiids, such as Eurypon, have plumose skeletal columns, whereas Hymeraphia is strictly hymedesmoid. Mesapos Gray is an objective synonym. Australian Species None. Genus Eurypon Gray Eurypon Gray, 1867: 521.-de Laubenfels, 1936: 107; Bergquist, 1970: 31. Epicles Gray, 1867: 521.-de Laubenfels, 1936: 110; Bergquist, 1970: 31. Acantheurypon Topsent, 1927: 15.-Topsent, 1928: 291 (type species Hymeraphia pilosella Topsent, 1904: 163, by original designation; holotype MOM, schizotype MNHN LBIM DT 933) (Fig. 68e-f). 1314 J. N. A. Hooper Tricheurypon Topsent, 1928: 2 9 5 . 4 e Laubenfels, 1950: 80; Wiedenmayer, 1977: 159 (type species Hymeraphia viridis Topsent, 1889: 43, by monotypy, holotype MNHN LBIM DT 1838) (Fig. 68g). Protoraspailia Burton & Rao, 1932: 342 (type species Protoraspailia cactoides Burton & Rao, 1932: 343, by original designation and monotypy; holotype IM ZSI P78911, schizotype BMNH 1931.1.1.52, from India) (Fig. 68h-k). Proraspailia LBvi, 1958: 27 (type species: Proraspailia polyplumosa LBvi, 1958: 27, by monotypy; holotype MNHN LBIM DCL1296, from the Red Sea) (Fig. 681-m). Ilfrpe species: Hymeraphia clavata Bowerbank, 1866: 143 (by monotypy; holotype BMNH 1877.5.2 1.1556, from Shetland) (Fig. 69e-f). Diagnosis Encrusting, massive or digitate growth forms. Surface hispid, even, granular or conulose. Encrusting species have microcionid choanosomal skeletal architecture, a basally condensed layer of spongin fibres lying on substrate, producing small spongin-fibre nodes echinated by acanthostyles, and radially disposed extra-axial skeleton composed of subectosomal styles standing perpendicular to and embedded in basal fibres. Massive species have slightly axially condensed plumose tracts of extra-axial styles, often forming fan-like bundles, and tracts are lightly echinated by acanthostyles. Extra-axial styles may be partially or entirely spined (Acantheurypon). Ectosomal specialisation present (s.s.) or absent; if present ectosomal skeleton consists of fine monactinal (or diactinal) spicule brushes surrounding single protruding extra-axiah styles. In one group of species (Protoraspailia) long raphidifom oxeote spicules are dispersed throughout mesohyl, not associated with dermal spicule brushes. Structural megascleres consist of 1-2 categories of styles or subtylostyles (rarely modified to oxeas); echinating acanthostyles microcionid-like, typically long and with subtylote bases. Microscleres absent (s.s.), or may include raphides occurring singly or in trichodragmata (Tricheurypon). Remarks The use of the name Eurypon over Epicles, both established by Gray in 1867, stems from an apparently arbitrary decision made by de Laubenfels (1936: 110), even though Epicles has page-line priority over Eurypon (see Topsent 1927: 15). That both genera are synonyms is agreed by most authors (e.g. Bergquist 1970), despite inaccurate arguments by de Laubenfels (1936). However, Epicles has not been used by contemporary authors and Wiedenmayer et al. (in press) consider it to be a nomen oblitum. Topsent (1928) noted that specimens of the type species, E. clavatum, varied in their ectosomal skeletal development, ranging from single megascleres to definite surface brushes, and as such the diagnostic importance of that character is questionable, at the supra-specific level at least. Similarly, the definition of Eurypon provided above is broadened here to include species with spined, as well as smooth, choanosomal megascleres (to include species of Acantheurypon), and with or without raphidiform microscleres (to include species of Tricheurypon and Protoraspailia). This conflicts with earlier definitions of the genus (e.g. Bergquist 1970: 31; Boury-Esnault & Lopes 1985). Eurypon is probably no more than an encrusting Raspailia, with basal rather than axial compression of the choanosomal fibre skeleton, lacking true choanosomal megascleres, and with exclusively plumose architecture (or hymedesmoid-microcionid fibre structure in thinly encrusting forms). However, Eurypon never has more than three categories of megascleres: echinating acanthostyles, ectosomal spicules, and a single category of long spicule that is equated with the long subectosomal extra-axial megascleres found in typical Raspailia. By comparison, typical Raspailia have four categories of megascleres. This is, however, a tentative separation based on skeletal architecture, which is shown elsewhere in this paper to be less reliable for indicating affinities than other characters. The morphology of echinating acanthostyles in Eurypon also points to affinities with the family Microcionidae. Thus the relationship between Eurypon and Raspailia is probably the same as that between Clathria and Microciona (which have been merged by a number of authors) (Hooper, unpublished data). Australian Raspailiidae 1315 Topsent (1928) assigned six species to Acantheurypon, all of which differ from Eurypon only in having partially basally spined, or rarely entirely spined, long extra-axial megascleres [e.g. E. pilosella (Topsent) from the Azore I., north Atlantic (Fig. 68e--1. In the strict sense, Acantheurypon contains species with monactinal ectosomal megascleres that form surface brushes, but some (e.g. A. mucronale Topsent) have modified diactinal megascleres (tomotes, subtylote and polytylote tomotes). Acantheurypon should be merged with Eurypon for the same reason that the microcionids Anaata and Clathria should be synonymised: basal or partial spination of structural spicules varies considerably between species of both families (Hooper, unpublished data). Tricheurypon is similar to Eurypon and Acantheurypon, but lacks dermal specialisation and has raphide microscleres. The genus is monotypic [i.e. H. viridis Topsent from Banc de Campeche, Atlantic (Fig. 68g)l. Protoraspailia has a plumose axial and extra-axial skeleton consisting of tracts of subectosomal megascleres, lightly echinated by acanthostyles, without an ectosomal skeleton and with raphide microscleres (Fig. 68h-k). The genus is monotypic, and Burton & Rao (1932) considered that it was intermediate between the Raspailiidae and other families of Poecilosclerida. Similarly, Proraspailia has a skeletal structure that is exclusively plumose, with long Eurypon-like echinating acanthostyles, also lacking both choanosomal axial megascleres and an axial skeleton, but with long raphidiform oxeote spicules dispersed throughout the mesohyl (Fig. 681-m). LCvi (1958) suggested that the type species showed similarities with Poecilosclerida such as Pronax (Stylostichon) and Raspailiidae such as Aulospongus. Both Protoraspailia and Proraspailia are merged here with Eurypon. This broad definition of Eurypon is consistent with the treatment of the diverse and similar (homologous) features in the Microcionidae, in which genera containing smooth or spined megascleres (e.g. Clathria Schmidt, and Dictyociona Topsent or Anaata de Laubenfels), and with or without microscleres (e.g. Clathria and Abila Gray) should be merged (van Soest 1984; Hooper 1988b; Hooper, unpublished data). This comparison with Microcionidae is most appropriate for Eurypon, and many species have been removed from Raspailiidae (or Euryponidae) and synonymised with various microcionid genera [e.g. Eurypon asodes de Laubenfels, 1930: 27, and E. microchela Stephens, 1916: 240 were transferred to Dictyociona (de Laubenfels 1936); E. rhopalophora (Hentschel, 1912: 380), and E. tenuissima Stephens, 1916: 240 were transferred to Microciona (Burton 1959; LBvi 1960)l. Those species are now recognised as merely being encrusting Clathria-like sponges (Hooper, unpublished data). Another aspect that has been used to differentiate the raspailiid-like taxa (Eurypon) from microcionid species (Clathria s.1.) is acanthostyle morphology (e.g. Berguist 1970: 32), but given the diversity of those megascleres within obvious microcionids (i.e. chelae-bearing taxa), in some instances this difference must also be questioned. It is possible that Eurypon will eventually be divided amongst Raspailia (i.e. species of Eurypon with ectosomal oxeas), Clathria (Clathria) (i.e. species without ectosomal megascleres), and Clathria (Thalysias) (i.e. species with ectosomal styles or subtylostyles). However, the genus is surprisingly poorly represented in Australian waters, and no such revision is presently contemplated on the basis of existing material. Eurypon graphidiophora Hentschel (Fig. 69) Hymeraphia graphidiophora Hentschel, 1911: 350.-de Laubenfels, 1936: 110; Bergquist, 1970: 32. Material Examined The type material has not yet been traced, but it may be extant in collections of the ZMH. Substrate and Depth Range Unknown. Geographical Distribution Geraldton, W.A. (Fig. 69d). J. N. A. Hooper Fig. 69. a-d, Eurypon graphidiophora Hentschel: a, basal end of a subectosomal extra-axial subtylotyle; b, ectosomal auxiliary styles; c, echinating acanthostyle; d, known Australian distribution. e-f, E. clavatum (Bowerbank) (specimen from the North Atlantic, MNHN LBIM DT 1341) (type species of the genus Eurypon Gray): e, section through peripheral skeleton (scale = 200pm); f , spiculation (scale = 500 pm) (a-c redrawn from Hentschel 1911: fig. 33). Description Shape. Thinly encrusting, up to 0.5 mm thick, on worm tubes and other calcareous material. Colour. Drab grey. Oscula. Not observed. Tature and suflace characteristics. Prominently hispid surface, produced by protruding extra-axial spicules. Ectosome and subectosome. Ectosome with bundles of auxiliary styles surrounding extra-axial megascleres around point of contact with surface. Subectosomal skeleton radially arranged extra-axial styles perpendicular to substrate. Australian Raspailiidae 1317 Choanosome. Choanosomal skeleton reduced to basal layer of spongin lying on the substrate, with bases of extra-axial spicules and echinating acanthostyles embedded and perpendicular to fibres. Megascleres. Choanosomal axial megascleres absent. Subectosomal extra-axial styles variable in length, thin, straight or slightly curved near basal end, with tapering fusiform points, and subtylote or sometimes polytylote bases; larger forms entirely smooth whereas smaller spicules have vestigial spination (280-1500 x 7-1 1 pm).. Ectosomal auxiliary spicules very thin, straight, setaceous, sharply pointed (352-400~23 pm).. Echinating acanthostyles straight, slightly subtylote, sharply pointed, evenly spined, with recurved spines on basal end (48-88 x 5 pm).. Microscleres absent. Remarks The above description is taken from Hentschel (1911). Type material of this species has not been traced to collections of the ZMB or SMF, and it is unknown whether type specimens exist in ZMH since it has not been possible to obtain any material from that institution. Hentschel (1911) divided the subectosomal extra-axial megascleres into two size classes, and noted that the smaller spicules had vestigial spination. He suggested that these spicules may be smaller forms of the larger extra-axial spicules, and they have been included in that category here. However, it is equally as likely that they may represent a larger example of echinating acanthostyles. Since these spicules were not figured it is not possible to decide one way or another. Hentschel suggested that the species was most closely related to Hymeraphia similis Thiele, but differed in the geometry and dimensions of extra-axial megascleres. He also contrasted it with E. miniaceum (see Raspailia irregularis above), although spicule geometry appears to differ substantially between the two species. Genus Rhabdeurypon Vacelet Rhabdeurypon Vacelet, 1969: 188. Type species: Rhabdeurypon spinosum Vacelet, 1969: 188 (by original designation and monotypy; holotype MNHN LBIM DJV4 from Cassidaigne, Mediterranean) (Fig. 70c-d). Diagnosis Thinly encrusting growth form. Surface even and hispid. Choanosomal skeleton a basally condensed layer of spongin lying on substrate. Choanosomal megascleres and true echinating spicules absent, but mesohyl contains acanthorhabds and microxeas scattered throughout basal skeleton, without apparent order. Extra-axial skeleton with long smooth subectosomal subtylostyles embedded in and perpendicular to substrate, protruding through surface. Ectosome with specialised skeleton of oxeas in brushes surrounding extra-axial styles. Structural megascleres are smooth choanosomal styles or subtylostyles of 1 category, ectosomal auxiliary oxeas, and pseudoechinating diactinal acanthorhabds. Microscleres smooth microxeas. Remarks The diactinal acanthorhabds of Rhabdeurypon are unusual (Fig. 70d), they resemble the discorhabds of Latrunculia (cf. Negombo tenuistellata Dendy, 1905), the acanthoxeas of Histodermella (cf. H. ingolfi Lundbeck, 1910), and are also similar to diactinal forms of acanthose spicules of Tethyspira (Fig. 4w). Rhabdeurypon differs from Tethyspira in having diactinal v. monactinal acanthose spicules, and a specialised raspailiid skeleton. Vacelet (1969) notes that acanthorhabds in R. spinosum are not echinating, but merely dispersed throughout the mesohyl in the basal skeleton, and therefore perhaps they should not be regarded simply as modified acanthostyles. However, the presence of a typical raspailiid ectosomal skeleton and a radial or plumose extra-axial skeleton (Fig. 70c), and the presence of acanthose accessory spicules does indicate that the affinities of this genus probably lie with the Raspailiidae. Australian Species None. 1318 J. N. A. Hooper Fig. 70. a-b, Tethyspira spinosa (Bowerbank) (schizotype BMNH 1877.5.21.394) (type species of Tethyspira Topsent): a, section through peripheral skeleton (scale = 250 pm); 6 , spiculation (scale = 200 pm). c-d, Rhabdeurypon spinosum Vacelet (holotype MNHN LBIM DJV 4) (type species of Rhabdeurypon Vacelet): c, section through peripheral skeleton (scale = 200 pm); d , pseudoechiiating acanthorhabds (scale = 50 pm). e-f, Plocamione clopetaria (Schmidt) (schizotype BMNH 1870.5.3.78) (type species of the nominal genus Raspeloplocamia Burton): e, section through peripheral skeleton (scale = 250 pm); f , structural megascleres (scale = 500 pm). g, P. ornata (Dendy) (holotype BMNH 1923.10.1.126) (type species of the nominal genus Axoplocamia Burton), basal skeleton (scale = 250 pm). h, P. dirrhopalina Topsent (holotype MNHN LBIM DT 1245) (type species of the genus Plocamione Topsent), section through peripheral skeleton (scale = 200 pm). i, P. hystrix (Duncan) (specimen MNHN LBIM DT 1416), section through peripheral skeleton (scale = 500 pm). j-I, P. pachysclera (L6vi & Levi) j, holotype (MNHN LBIM DCL 2948) (scale = 30 mm); k, section through peripheral skeleton (scale = 1 mm); I, spiculation (scale = 200 pm). Australian Raspailiidae Genus Plocamione Topsent Plocamione Topsent, 1927: 16; 1928: 63. Raspeloplocamia Burton, 1935: 402 (type species Plocamia clopetaria Schmidt, 1870: 63, by original designation; holotype LMJG, schizotypes MNHN LBIM DCL1106L, BMNH 1870.5.3.78, from Cuba) (Fig. 70e-f). Axoplocamia Burton, 1935: 402.-Bergquist & Fromont, 1988: 122 (type species Bubaris ornata Dendy, 1924: 351, by original designation; holotype BMNH 1923.10.1.126, schuotypes 1923.10.1X.3, from New Zealand) (Fig. 70g). Plocamia Schmidt, 1870: 62 (in part) [preocc. by Plocamium Lamouroux (a seaweed)] (type species: Plocamia gymnazusa Schmidt, 1870: 62, by subsequent designation (Burton, 1935: 401); holotype LMJG (not seen), schizotype MNHN LBIM DCLllOSL, from the Adriatic). Dirrhopalum Ridley in Ridley & Duncan, 1881: 477 (in part) [replacement name for Plocamia under Art. 67h ICZN (Anonymous 1985)l. Type species: Plocamione dirrhopalina Topsent, 1927: 16 (by monotypy) (holotype MOM, (not seen) schizotype MNHN LBIM DT1245, from the Azores) (Fig. 70h). Diagnosis Encrusting or digitate growth forms. Surface even and hispid. Choanosome a basally or axially condensed reticulation of choanosomal acanthostrongyles, forming a thick longitudinal core; echinating acanthostyles embedded in core, standing perpendicular to axis. Extra-axial skeleton composed of radial or plumose columns of long subectosomal styles. Ectosomal skeleton with specialised skeleton of ectosomal styles or anisoxeas forming brushes around extra-axial spicules. Structural megascleres include 2 categories of styles or subtylostyles; echinating spicules evenly spined acanthostyles with subtylote bases, sometimes entirely smooth, partially acanthose, or with smooth, spined or tuberculate bases. Choanosomal axial or basal spicules acanthostrongyles or acanthotylostrongyles, straight or slightly curved, distinctly acanthose, merely tuberculate, or rarely entirely smooth (= 'pegtop' spicules of Ridley in Ridley & Duncan 1881). Microscleres absent. Remarks There is a large number of species and genera of 'plocamiform' sponges that have a basal or axial skeleton of acanthose diactinal megascleres, but most of those taxa are relatively poorly known. Most are obvious members of the order Poecilosclerida related to the Myxillidae, Crellidae or Microcionidae [Plocamia Schmidt, 1870 (=Dirrhopalum Ridley in Ridley & Duncan, 1881), Lissoplocamia Brondsted, 1924, Heteroclathria Topsent, 1904, Plocamiopsis Topsent, 1928, Echinoplocamia Burton, 1959, Plocamionida Topsent, 1927, Plocamissa Burton, 1935, Plocamiancora Topsent, 1927, Damiria Keller, 1891, Damiriella Burton, 1935, and Damiriopsis Burton, 19281. The others [Plocamione, Raspeloplocamia, Axoplocamia, Lithoplocamia (of which the first three are synonyms)] are referrable to the Raspailiidae. There are six species currently placed in Plocamione: P. dirrhopalina Topsent (Fig. 70h), P. carteri @3uncan in Ridley & Duncan 1881: 488, from the North Atlantic (holotype not found in the BMNH)], P. clopetaria (Schmidt) (Fig. 70e-f), P. hystrix Duncan in Ridley & Duncan 1881: 491 (holotype not found in the BMNH, specimen MNHN LBIM DT1416), from the south-west coast of Spain (Fig. 70i)], P. ornata (Dendy) (Fig. 70g), and P. pachysclera (Gvi & G v i 1983: 947) from New Caledonia (holotype MNHN LBIM DCL2948) (Fig. 70j-I). All of these species are typical raspailiids in having specialised ectosomal brushes of spicules clustered around a central extra-axial spicule, and five of these species also posses curved or vermiform 'sausage-shaped' spicules (acanthostrongyles) that are characteristic of 'plocamiform' sponges (sensu Burton 1935). Those megascleres are absent in P. pachysclera (Ltvi & Ltvi), but that species has other features that suggest a close relationship with the group (Fig. 70j-I). Other species included in this genus or one of its synonyms by authors (e.g. Ridley & Duncan 1881) have affinities with the Microcionidae, and are transferred here to the 1320 J. N. A. Hooper genus Antho (Hooper, unpublished data): Dirrhopalum gymnazon (Schmidt, 1870: 62), Dirrhopalum plenum (Sollas, 1879a: 49, Dirrhopalum coriaceum (Bowerbank, 1874: 136), Dirrhopalum manaarense (Carter, 1880: 34) and Dirrhopalum novizelanicum Ridley in Ridley & Duncan, 1881: 483. Dirrhopalum microcionides (Carter, 1876: 390) is a synonym of Clathria ambigua, and most appropriately placed in the genus Plocamionida Topsent, 1927 (Myxillidae or Anchinoidae) (van Soest & Weinberg 1980). Australian Species None. Genus Lithoplocarnia Dendy Lithoplocamia Dendy, 1922: 79. Monectyon LBvi & Vacelet, 1958: 236 (type species Monectyon atlanticus LBvi & Vacelet, 1958: 236, by monotypy). Type species: Lithoplocamia lithistoides Dendy, 1922: 79 (by monotypy; holotype BMNH 1921.11.7.68, from Mauritius) (Fig. 71a-c). Diagnosis Massive or encrusting growth forms. Surface often sculptured by drainage canals. Choanosomal skeleton lacks axial compression, with dense, regular isodictyal (s.s.) or irregular sub-isodictyal reticulation of choanosomal acanthostrongyles. Echinating megascleres Fig. 71. a-c, Lithoplocamia lithistoides Dendy (type species of the genus Lithoplocamia Dendy): a, holotype (BMNH 1921.11l.68) (scale = 30 mm); b, isodictyal skeletal structure (scale = 250 pm); c, acanthostrongyle of the choanosomal skeleton (scale = 100 pm). d-f, L. dolichosclera LBvi & LBvi: d, holotype (MNHN LBIM DCL 2946) (scale = 30 mm); e, subisodictyalskeletal structure (scale = 1 mm);f,choanosomalacanthostrongyle(scale = 200 pm). Australian Raspailiidae 1321 absent, although the type species has 2 categories of acanthose megascleres. Extra-axial skeleton with radial tracts of subectosomal styles, but these are usually masked by dense choanosomal reticulation. Ectosomal skeleton membraneous and typically without specialised spiculation, but when present spicules are long, slender oxeote megascleres, forming radial bundles at ends extra-axial skeletal tracts. Structural megascleres are smooth styles or subtylostyles (extra-axial spicules), oxeas (ectosomal spicules), diactinal or pseudodiactinal acanthostrongyles (choanosomal axial spicules), with or without accessory acanthostyles. Microscleres absent. Remarks Three species, all from relatively deep waters, are referred here to Lithoplocamia: L. lithistoides Dendy (Fig. 71a-c), L. dolichosclera U v i & LBvi, (1983: 948) from New Caledonia (holotype MNHN LBIM DCL 2946) (Fig. 71d-f), and L. atlanticus (LCvi & Vacelet, 1958: 236) from the eastern Atlantic (holotype missing from the MNHN). This genus shows some similarities to some species of Antho (family Microcionidae) and Damiria, lacking chelae, and Dendy (1922) suggested that the type species was most closely related to Plocamia massalis. However, these supposed affinities are probably artificial, being based solely on similarities in geometry of the choanosomal acanthostrongyles. Sponge consistency, extra-axial skeletal structure and spicule morphology appear to be different. Dendy (1922) and U v i & U v i (1983) remark on the close resemblance in the basal skeletal construction between this genus and those 'Lithistida' with monocrepidial desmas (e.g. Monocrepidium Topsent, Lithobubaris Vacelet; family Bubaridae). Intermediate forms also exist (e.g. Cerbaris Topsent). Dendy speculates further that monocrepidial desmas may have developed from the curved acanthostyles characteristic of Lithoplocamia (and other 'plocamiid' sponges), but if any relationship exists between these two forms of megascleres it is probably more likely that the reverse is true. LBvi & LCvi (1983) also compare Lithoplocamia with Endectyon, the former having a massive and non-branching growth form and a compact isodictyal or subisodictyal reticulation of acanthostrongyles, whereas Endectyon has acanthose spicules echinating plumose columns of choanosomal axial styles. In this regard Lithoplocamia must also be compared with the isodictyal raspailiid genus Amphinomia, gen. nov. The nominal genus Monectyon was described as having a skeleton composed only of acanthostrongyles, without other megascleres present at all, but LBvi & LBvi (1983) suggest that additional material of M. atlanticus may eventually show that the species does possess true ectosomal auxiliary and subectosomal extra-axial spicules; Lithoplocamia and Monectyon are synonyrnised here. Australian Species None. Genus Amphinomia, gen. nov. m e species: Amphinomia sulphurea, sp. nov. (Figs 72-73). Diagnosis Massive flabellate-lobate growth form. Surface fleshy, uneven and not hispid. Choanosomal skeletal architecture regularly renieroid-reticulate, without any axial condensation or noticeable axial and extra-axial differentiation of skeletal tracts. Spicules and fibres form ascending multispicular tracts, interconnected by mi- or paucispicular transverse tracts, and fibres lightly echinated by acanthostyles. Mesohyl matrix heavily invested with spongin. Extra-axial skeleton vestigial, with subectosomal styles scattered throughout skeleton, predominant near periphery. Ectosomal skeleton thickly membraneous, without specialised spiculation. Structural megascleres are styles of 2 categories: subectosomal styles short, slender and smooth, choanosomal styles peculiarly acanthose, with large spines on basal and distal extremities; echinating acanthostyles sparsely but evenly spined. Microscleres absent. 1322 J. N. A. Hooper Etymology The generic name Amphinomia (f.) is derived from the name of the shoals near which the type species was first collected. (Amphinome is preoccupied, erected for a group of polychaetes by Bruguiere in 1792). Amphinomiu sulphurea, sp. nov. (Figs 72, 73, 109j, 110a; Table 16) Material Examined Holotype. NTM 21787: NW. of Amphinome Shoals, NWS, W.A., 19'04.11S., 118'57.9'E., 84 m depth, 29.viii.1983, coll. T. Ward (CSIRO RV 'Soela' S04/83, stn 124-NWS24, trawl). Paratypes. NTM 22485: NW. of Amphinome Shoals, NWS, W.A., 19' 07'S., 118' 29'E., 82-86 m depth, 2.vi.1985, coll. B.C. Russell (Taiwanese pair trawler, stn BCR 8518, haul 4-NWS39, trawl). NTM 23128: Pany Shoals, Timor Sea, N.T., 11" 11-411S., 129'43 .OllE., 18 m depth, 13.viii.1987, coll. A.M. Mussig & NCI (stn Don. 248-AM 87-3, SCUBA). Other material. Wessel Is, N.T.: NTM 23943: Bay S. of Sphinx Head, 11" 14'S., 136"40.9'E., 13 m depth, 15.xi.1990 (stn JH-90-029). NTM 23964: S. of Truant I., English Company Is, 11' 40.45'S., 136' 50 - 2' E., 26 m depth, 19.xi.1990 (stn JH-90-034). Northwest Shelf, W.A.: PIBOC 012-154: near Scott Reef, 16' 05.01S., 121" 14.7'E., 73-75 m depth, 15.xi.1990, coll. V. Krasochin, U.S.S.R. RV 'Akademik Oparin', dredge (stn 31). NTM 21809: W. of Port Hedland, NWS, W.A., 19' 26.9' S., 118"54.2/E., 50 m depth, 30.viii.1983, coll. T. Ward (CSIRO RV 'Soela' S04/83, stn 126-NWS26, trawl). Timor Sea, N.T.: NTM 23105: Pany Shoals, Timor Sea, N.T., 11" 12.53'S., 129'42.08'E., 20 m depth, 15.viii.1987, coll. A.M. Mussig & NCI (stn Don. 251-AM 87-6, SCUBA). Substrate and Depth Range Shallow coastal reefs, sand, silt and coral rubble substrate, 13-86 m depth. Geographical Distribution Northwest Shelf, W.A., Timor Sea, Joseph Bonaparte Gulf region, and Wessel Is, N.T. (Fig. 72e). Common in the Wessel Is region. Description Shape. Massive, thickly lobate-lamellate growth form, with flabellate lobes, and rounded or slightly undulating margins (90-150 mm height, 160-220 mm maximum lateral span, with lobes ranging from 8-25 - 5 mm thick); holdfast and basal stalk absent, sponges attached to substrate directly on underside of subcylindrical lobes. Colour. Live surface pigmentation ranges from pink-brown (shallow water material, Munsell 2 -5YR 714) (Fig. 109~) to yellow-brown (deeper water material, 2 -5Y 818); deeper water material with darker orange-brown areas on margins of lobes (2.5Y 7/10) (Fig. 110a). Pigmentation of choanosome slightly darker than surface (beige-yellow-brown). Surface pigmentation of deeper water specimens oxidises after several minutes following contact with air, turning dark blue to black (2.5B 312 or darker), whereas choanosomal coloration remains beige-brown. Oscula areas pale beige white alive (2 5Y 814). Oscula. Large oscula regularly distributed on or near apical margins of lobes, located in surface depressions, 5-5-11 mm diameter. Pores scattered throughout surface, on both faces of lobes, usually located between ridges and surface protniberances. Texture and surjCace characteristics. Surface thickly glabrous, fleshy, prominently crinkled and uneven, produced by irregular, interconnected low ridges, depressions and microconules. Some lobes convoluted with a sinuous outline, but most relatively even. Texture firm, resilient, only slightly compressible and more-or-less rigid and difficult to cut in preserved state. Ectosome and subectosome. Ectosomal skeleton simply membraneous, consisting of a thick layer of darkly pigmented type B spongin, 80-256 p m thick, with very few choanosomal megascleres protruding through surface. Average thickness and density of ectosomal spongin varies considerably between specimens, but essential details of glabrous ectosomal skeleton - Australian Raspailiidae Fig. 72. Amphinomia sulphurea, gensp. nov. (holotype NTM 21787): a, choanosomal axial styles; b, subectosomal extra-axial style; c, echinating acanthostyle; d, section through peripheral skeleton; e, known Australian distribution. are consistent. Specialised ectosomal skeleton absent; in subdermal region tracts of choanosomal styles usually converge upon each other, forming slightly elevated surface conules dispersed at irregular intervals, corresponding with irregular surface projections. Relatively large subdermal cavities, 110-430 p m diameter, lie between convergent spicule tracts; cavities lined by thick non-fibre (type B) spongin. Some subectosomal extra-axial styles scattered throughout peripheral regions of skeleton; subectosomal styles predominant in but not restricted to this region. Choanosome. Choanosomal skeletal architecture more-or-less regularly renieroid, verging on subisodictyal in places. Choanosomal architecture towards axis consists of ascending multispicular tracts interconnected by transverse uni- or paucispicular tracts; towards peripheral skeleton renieroid reticulation more obvious due to less differentiation between ascending and transverse tracts. Spongin fibres thin, 38-91 p m diameter, 70-128 pm at fibre nodes, and very lightly invested with type A spongin. Fibres usually obscured by heavy mesohyl matrix (type B spongin), mostly fully cored by spicules. Some larger and most smaller secondary fibres (25-92 pm diameter) uncored, with faint but optically diffuse pith; fibres have very fine and regular radial striations on surface. Fibre reticulation and spicule tracts produce triangular, hexagonal or square meshes; each side of mesh never longer than 1 spicule length. Echinating acanthostyles dispersed throughout skeleton, moderately abundant, and slightly more common at fibre nodes. Mesohyl matrix typically very heavy and moderately darkly pigmented, but spongin pigmentation varies between specimens. Collagenous (type B) spongin of mesohyl slightly granular, sometimes aggregated into mesentary-like tertiary fibre formations. Choanocyte chambers circular or J. N. A. Hooper Fig. 73. Amphinomia sulphurea, gen.sp. nov.: a, light micrograph of the peripheral skeletal structure (scale = 500 pm); b, paratype (NTM 23128); c, specimen (NTM 23105) (photo NCI, from colour slide); d, holotype (NTM 21787, photo T. Ward, from colour slide) (scale = 30 mm); e, SEM of renieroid skeletal structure;f, SEM of primary (multispicular, spongin coated) and secondary (unispicular) fibres; g, SEM of echinating megasclere in sifu; h, SEM of terminally spined choanosomal styles. ovoid, occasionally elongate, 46-89 pm diameter. Subectosomal extra-axial styles scattered between fibres throughout skeleton, but these are relatively uncommon. Megascleres (refer to Table 16 for dimensions). Choanosomal axial styles thick, usually slightly curved midway along shaft, occasionally straight or sometimes with sigmoid curve, Australian Raspailiidae 1325 with hastate and typically heavily spined points, and with evenly rounded and heavily spined bases, but rare examples with rudimentary spines or completely smooth, shaft, between basal and apical extremities, entirely smooth. Subectosomal extra-axial styles straight or slightly curved, fusiform, entirely smooth, with rounded or very slightly subtylote, or very rarely polytylote bases. Ectosomal auxiliary megascleres absent. Acanthostyles variable in size, with smaller and thinner examples probably being young forms. Spicules rounded or very slightly subtylote, fusiform, with sparse but evenly dispersed spines; spines typically large and recurved. Microscleres absent. Table 16. Comparisons in spicule measurements between specimens of Amphinomia sulphurea, gen. et sp. nov. Measurements are given in micrometres, presented as ranges (and means) of lengthxwidth, and represent 25 spicules measurements per specimen for each spicule category Material Choanosomal styles Subectosomal styles Ectosomal megascleres Echinating acanthostyles Absent 102-138 Absent 92-141 ~6-11 (126.3~8.4) 96121 ~6-10 (106.2~8.6) - Holotype NTM 21787 152-274 NTM 22485 231-288 x 18-23 (261.3~20.1) 2 12-274 x 12-24 (238.9~17.8) 178-248 Paratypes NTM 23128 190-235 x 4-10 (220-7~6.8) 172-228 x 3-8 (193.6~596) Absent Specimens NTM 21809 NTM 23105 205-264 x 1622 (239.3~19.0) 193-282 X 12-22 ( 2 4 1 . 2 ~17.7) 192-222 x4-11 (210-7~7.0) 153-215 X 3-9 (191-7~5.4) Absent Absent 105-138 x4-11 (115-1~7.2) 91-138 X5-10 (116-7~6.3) Remarks Amphinomia sulphurea has a peculiar mixture of characteristics. Although it is not easy to select a single feature that is unique to the genus, it differs from other raspailiids by the presence of a regularly reneiroid skeleton without any axial compression, and the choanosomal megascleres have peculiar spination. It is similar to massive suberitids (Hadromerida) in growth form and texture. It shows superficial similarities in skeletal construction to members of the order Agelasida, the poecilosclerid families Microcionidae [such as Antho (Isopenectya) and Antho (Naviculina)] and the Myxillidae (such as Myxilla and Acarnus). The pigment oxidation that occurs in the three deeper water specimens soon after exposure to the air is reminiscent of the order Verongida, as is the darkly pigmented and aspiculose ectosomal spongin layer and dense mesohyl matrix. However, it is likely that pigment oxidation is due to the high levels of sulphur in the sponge, rather than to the presence of guanidine carotenoid pigments, which are characteristic of the Verongida (e.g. Bergquist 1980). Apically and basally spined choanosomal styles found in A. sulphurea are unusual but not unique amongst the Porifera. Bowerbank (1864: 232, fig. 32) figured similar spicules from an unnamed sponge collected from an unspecified locality. Bowerbank (1874: 149, pl. 58, fig. 31) also erected Isodictya lurida for a poecilosclerid containing J. N. A. Hooper 1326 Table 17. Comparisons in spicule measurements between specimens of Trikentrion flabeI1ifonne Carter Measurements are given in micrometres, presented as ranges (and means) of lengthxwidth, and represent 25 spicules measurements per specimen for each spicule category Material Choanosomal oxeas Subectosomal styles Ectosomal stylesf anisoxeas Echinating sagittal triacts Raphides 135-253 ~5-18 (219.4~14.2) 416-843 ~8-17 (620-8x1301) 212-342 xO-5-4 (263.8~2-5) 122-143 x 11-17 (127.7~14.5) 43-78 ~0.3-0.8 (63.8~0.5) 174312 X7-19 (243.6~12.4) 213-340 x 5-22 (256.3~15-2) 430-814 X9-18 (587-0~12.8) 405-1011 x4-19 (723-8x9-4) 182-265 x 1-4 (224-8~2.3) 184385 x0.54 (288.0~1.9) 122-181 52-80 x 12-19 ~0.3-1 (139-4~15.1) (64.5~0-5) 120-168 39-78 x 8-20 xO.3-1 ( 1 2 3 . 2 ~ 1 3 ~ 3 ) (57.9~0.5) 720-1024 x9 240-344 x3 Lectotype BMNH 1887. 5.21. 1865 Specimens BMNH 1931. 8.4.57 NTM various (n=58) Hentschel (1912) 264-304 x 14-19 136-168 x 12-16 65-88 choanosomal styles with similar spination, but that species also contained isochelae and ectosomal tomotes. Iophon chelifer Ridley & Dendy and some species of Hymerhaphia Bowerbank have been recorded with similarly spined styles, but those species differ quite significantly from the present genus in most other features. Freshwater Spongillidae [e.g. Oncosclera schubarti (Bonetto & Ezcurra de Drago), Radiospongilla crateriformis (Potts)] are known to have basally and apically spined megascleres, but these are usually oxeote or strongylote in geometry. Basal acanthostyles or acanthostrongyles found in microcionid genera such as Antho and Isopenectya frequently have spined extremities, but those spicules represent a separate category of megascleres, and in any case they invariably occur in conjunction with extra-fibre tracts of smooth or only basally spined choanosomal styles. Amphinomia sulphurea shows convergent features with Labacea (=Antho) juncea (sensu Burton) which also has a renieroid skeletal architecture, similar fibre characteristics, and more-or-less similar spiculation (Hooper, unpublished data). The species differs from Labacea [sensu de Laubenfels 1936; including L. oxeifera (Ferrer-Hemindez)] in completely lacking an ectosomal skeleton or microscleres, in lacking well differentiated primary ascending and secondary transverse skeletal tracts, and in having only the bases and points of choanosomal styles bearing microspines. Isopenectya (=Antho) chartacea is also similar, containing small basally and apically (sometimes entirely) spined styles in a renieroid reticulation. However, it also has a subisodictyal and condensed axial, and plumose extra-axial skeleton of larger smooth choanosomal styles, which protrude through the surface in plumose tracts, and a tangential or paratangential layer of subectosomal subtylostyles on the surface. Apart from spicule ornamentation, which links Amphinomia with some Microcionidae and Myxillidae poecilosclerids, the genus shows similarities with some other members of the Raspailiidae. Realistically, the genus could be placed in either family, but because of its affinities with Lithoplocamia and the remnants of an extra-axial skeleton, it is referred to the Raspailiidae. Lithoplocamia has a basal renieroid skeleton of peculiarly spined diactinal megascleres, an extra-axial plumose or radial skeleton of subectosomal styles, with or without echinating acanthostyles and ectosomal specialisation. Nevertheless, Amphinomia is well differentiated from Lithoplocamia by the very different geometry of their basal choanosomal spicules (monactinal and diactinal, respectively). The genus may be compared with, and differentiated from, other raspailiids that have omamentation on choanosomal styles. Axechina raspailioides Hentschel (type species of Australian Raspailiidae 1327 Axechina) also has apically and basally spined choanosomal megascleres, but these spicules are oxeas or anisoxeas; skeletal construction, surface features and spicule geometry are quite different between the two. Stylostichon involutum Kirkpatrick, (type species of Hemectyonilla, and referred here to Aulospongus) has rhabdostyles with distal spination, but with a plumose choanosomal skeleton. Raspailia villosa Thiele (type species of Heterectya, and also referred here to Aulospongus) has a plumose, less obviously plumo-reticulate skeleton, with a simply membraneous ectosomal skeleton. Rhaphidectyon spinosum Topsent (type species of Rhaphidectyon, and also referred here to Aulospongus), has a simple spiculation and membraneous ectosome, but its architecture is plumose and choanosomal styles are entirely smooth. Axinella mariana Ridley & Dendy (type species of Axinectya, referred here to Raspailia), has entirely and lightly spined rhabdostyles, but it also has a condensed axial and plumose extra-axial skeleton. Ectyoplasia tabula (type species of Ectyoplasia) has a condensed reticulate axial skeleton, a radial or slightly plumose extra-axial skeleton, and a plumose ectosomal skeleton. However, two other species referred to that genus (E. ji-ondosa and E. ferox) have more pronounced reticulate skeletons, with less emphasis on axial and extra-axial differentiation, and on paper at least (e.g. Wiedenrnayer 1977), the present genus shows similarities to those (non-typical) Ectyoplasia species. The ectosomal skeleton of E. ferox is fleshy and pierced by plumose brushes of the ascending peripheral choanosomal skeleton. Choanosomal skeletal architecture is isotropic, with prominent ascending multispicular tracts, becoming plumose at the periphery, interconnected by a uni-, bi- or paucispicular transverse reticulation. Those features contrast with the aspiculose ectosome and renieroid choanosomal skeletons of Amphinomia. Moreover, there is no suggestion of clavulate modifications to echinating acanthostyles, which are characteristic of Ectyoplasia. Rhabderemia Topsent (including Rhabdosigma Hallmann) of the family Rhabderemiidae is also known to have spinose choanosomal megascleres, but in other features they bear little resemblance to Amphinomia. Etymology This species is named for the yellow-brown coloration of deeper water type material, and its curious antimicrobial properties. The antimicrobial activity was found to be positive against Escherichia coli and Bacillus subtilis (Hooper et al. 1992), and is apparently due to high levels of sulphur, although it is possible that the chemical is bacterial in origin (R. Capon, personal communication). Genus Ceratopsion Strand Ceratopsis Thiele, 1898: 57 (preocc.).-Hallmann, 1916a: 541. Ceratopsion Strand, 1924: 3 3 . d e Laubenfels, 1936: 132; Bergquist, 1970: 18. Type species: Ceratopsis expansa Thiele, 1898: 57 (by original designation; holotype possibly ZMB (although not found), paratype in SM, schizotype MNHN LBIM DCL 981L, from Sagarni Bay, Japan) (Fig. 74g). Diagnosis Erect, lamellate or arborescent growth forms. Surface granular and hispid. Axial and extra-axial skeletons well differentiated. Choanosomal skeleton a condensed axial reticulation of spongin fibres, deficient in type B spongin, cored by sinuous or straight styles, anisoxeas or strongyles. Echinating megascleres absent. Extra-axial skeleton consists of radially arranged subectosomal styles, anisoxeas or strongyles, including sinuous and/or straight forms, embedded in and perpendicular to axial core, projecting through surface. Specialised ectosomal skeleton present, consisting of smooth oxeas (or styles) usually forming continuous palisade of brushes on surface, or brushes grouped around bases of projecting extra-axial megascleres. Structural megascleres include at least 2 categories of styles, anisoxeas or strongyles (one usually sinuous), oxeas or anisoxeas, and thin ectosomal oxeas or styles. Microscleres absent. 1328 J. N. A. Hooper Remarks This emended definition of Ceratopsion allows for the inclusion of species that have either monactinal or diactinal axial megascleres and extra-axial megascleres. Hallmann (1916~)used Ceratopsion as a catch-all for various Axinellidae with microxeas (excluding sigma-bearing and desmoxyid genera). However, Bergquist (1970) showed that these 'rnicroxeas' were, in fact, ectosomal megascleres, which were raphidifom in some species, and on this basis the genus is referred here to the Raspailiidae. Species placed in Ceratopsion by authors include: C. expansa Thiele (Fig. 74g), C. clavata Thiele (1898: 57) from Sagami Bay, Japan, C. cuneiformis Bergquist (1970: 18) from New Zealand, C. dichotoma (Whitelegge) (Fig. 74a-f), C. erecta Thiele (1898: 58) (schizotype MNHN LBIM DCL980L) from Sagami Bay, Japan (Fig. 74h), C. microxephora (Kirkpatrick, 1903: 242) from South Africa, C. minor Pulitzer-Fiali (1983: 520) from Calvi, Mediterranean, and C. ramosa Thiele (1898: 58) from an unknown locality in Japan. To this list should be added two new species described below and Axinella axifera Hentschel, 1912. Of all Ceratopsion species only C. cuneiformis and C. montebelloensis, sp. nov. have ectosomal megascleres that are sparsely dispersed around the protruding extra-axial spicules (i.e. in classical raspailiid arrangement), whereas in all other species the dermal skeleton consists of dense ectosomal brushes that may or may not always appear to be associated with the extra-axial skeleton. Ceratopsion clavata, C. microxephora and C. minor form a group characterised by the possession of sinuous strongyles in the axial skeleton, reminiscent of the Bubaridae. Spicule dimensions of all these species are compared in Table 18. Ceratopsion lacks echinating megascleres, but has other features found in most raspailiids. On paper the genus is difficult to distinguish from Raspailia (Syringella), but the presence of sinuous axial styles or anisoxeas in some species, and the possession of a distinctly radial extra-axial skeleton of styles, anisoxeas or strongyles may serve to differentiate these two genera. Ceratopsion dichotoma (Whitelegge), comb. nov. (Fig. 74a-f; Table 18) Raspailia dichotoma Whitelegge, 1907: 515, pl. 46, fig. 36. Material Examined Holotype. AM G4353: Off Coogee, N.S.W., 98-100 m depth (FRV 'Thetis', stn 44) (NMV sponge archives 312). Substrate and Depth Range Substrate unknown, 98-100 m depth. Geographical Distribution Deeper offshore waters of N.S.W. (Fig. 744. Description Shape. Stipitate, arborescent, digitate sponge, branching in 1 plane (183 mm high, 80 mm maximum breadth), with basal plate attached to substrate, long stalk (42 mm long, 4-5 mm diameter), long cylindrical branches bifurcated at about 112 their length and again near their tips, tapering towards their tips (maximum branch diameter 3.5 mrn). Colour. Grey in the dry state. Oscula. Few oscula (0-8-1 mrn diameter) scattered over branches, not confined to any particular surface. Texture and s u ~ a c echaracteristics. Texture inflexible and incompressible when dry. Surface contracted from drying, with ridges and irregular reticulated depressions, even surface and not hispid. Ectosome and subectosome. Ectosomal skeleton present, composed of plumose brushes of smaller styles perched on ends of extra-axial tracts. Ectosomal brushes intermingled Australian Raspailiidae 1329 with ends of larger extra-axial spicules, and consequently they are difficult to differentiate in histological sections. Nevertheless, distinct size differences exist between spicules in extra-axial and ectosomal skeletons. Extra-axis markedly radial, non-anastomosing, with large subectosomal styles in dense brushes standing perpendicular to axis, embedded in axis Fig. 74. a-f, Ceratopsion dichotoma (Whitelegge): a, choanosomal axial and ectosomal styles; b, subectosomal extra-axial styles; c, section through peripheral skeleton; d, known Australian distribution; e, holotype (AM G4353) (scale = 30 mm); f , photomicrograph of skeletal structure (scale = 1 mm). g, C . expansa Thiele (schizotype MNHN LBIM DCL 981L) (type species of the genus Ceratopsion Strand), spiculation (scale = 200 pm); h, C . erecta Thiele (schizotype M N H N LBIM DCL 980L), spiculation (scale = 200 pm). 1330 J. N. A. Hooper for nearly a quarter of their length. In addition, subectosomal styles also lie tangential to and in contact with periphery of axial skeleton. Spongin is heavy in mesohyl of ectosomal and extra-axial regions, smooth and non-granular, but only lightly pigmented. Bundles of extra-axial spicule tracts relatively widely spaced (80-130 pm apart), composed of 10-40 spicules per brush. Choanosome. Choanosomal skeleton a dense axial reticulation of closely packed spongin fibres; fibres are only lightly invested with spongin, 30-80 pm diameter, forming small elongate meshes, 25-70 pm diameter; fibres nearly fully cored with axial oxeas. Axial fibre meshes much more compact than in extra-axial region, and fibres run mostly longitudinally through branches. Fibres not easily divisible into primary or secondary elements, but marked differences between fibres in number of coring choanosomal spicules. Echinating megascleres absent. Choanocyte chambers difficult to observe in condensed axial skeleton, whereas in extra-axial region they are circular, 15-33 pm diameter. Megascleres (refer to Table 18 for dimensions). Choanosomal spicules in axial skeleton long, slender, usually straight or slightly curved near point, with prominent subtylote bases, tapering to sharp points at tip. Subectosomal megascleres in extra-axial skeleton long, thick, usually straight subtylostyles, with slight or prominent subtylote swellings on bases, tapering to sharp fusiform points or more abrupt terminations. Ectosomal megascleres identical to those found in axial skeleton. Echinating megascleres absent. Microscleres absent. Remarks Whitelegge's (1907) description of this species is quite misleading. His description of sections through the skeleton are incorrect: skeletal meshes are much more compact than described, by a factor of about three. Similarly, axial megascleres are confined entirely to within fibres and not scattered between fibres as suggested in the original description, and axial megascleres are subtylostyles, not oxeas as stated. Spicule dimensions of the holotype, cited in Table 18, also differ substantially from those published by Whitelegge (1907). This species is very distinctive in skeletal structure, although it is very plain in spiculation. Megascleres in all regions of the skeleton appear to be geometrically identical, differing only in their length and thickness. Similarly, although well separated into different regions of the skeleton, spicules in the peripheral (ectosomal) and axial (choanosomal) regions appear to be identical in size and shape, separated by a large region in between that has much longer and more stout styles. This extra-axial region is the most prominent feature of the skeleton, most notably the bundles of spicules standing perpendicular to the axis, and this species is referred here to Ceratopsion on the basis of these structural features. i Ceratopsion axifera (Hentschel), comb. nov. (Figs 75, 76a-b; Table 18) Axinella mifera Hentschel, 1912: 418-9, pl. 14, fig. 2, pl. 21, fig. 56. Material Examined Holotype. SMF 1666 (Schizotypc+h4NHN LBIM DCL 2253): Straits of Dobo, Aru I., Indonesia, 6'S., 134" 501E., 16 m depth, 20.iii.1908, coll. H. Merton (stn 3). Other material. NTM 23035: N . of Amphinome Shoals, NWS, W.A., 19" 19.7-23.3/S., 119' 08-8-12.2'E., 50 m depth, 19.vii.1987, coll. J.N.A. Hooper (U.S.S.R. RV 'Akademik Oparin', stn NWS55, beam trawl). Substrate and Depth Range Shallow offshore reefs, 16-50 m depth, sand and shell-grit substrate. Geographical Distribution Indonesia and Northwest Shelf, W.A. (Fig. 75e). Australian Raspailiidae 1331 Table 18. Comparisons in spicule measurements between species of Ceratopsion Measurements are given in micrometres, presented as ranges (and means) of lengthxwidth. For material described in this work these measurements were obtained by sampling 25 spicules per specimen for each spicule category Material Choanosomal megascleres Specimen NTh4 23035 354-623 x6-18 (533.0~11.2) (oxea/anisoxea) Holotype NTM 23388 498-65 1 ~5-10 (572.7~7.4) (style) Holotype NTM 23051 229-391 ~4-18 ( 3 4 8 . 4 ~12.9) Paratype NTM 23487 Subectosomal megascleres C . mifera (Hentschel) I: 985-1956 x2148 (1379.7~31.8) (style) 11: 486-675 x 13-26 (598 8-18.1) (oxea/anisoxea) Ectosomal megascleres Echinating megascleres 154-373 X3-12 (283.3~7-3) (style) Absent 232457 ~ 0 ~ 5 4 . 5 (352.3~2.1) (stylelanisoxea) Absent 20242 XI-5-6 (340.1~3-5) Absent 23 1-406 ~3-18 (317.6~12.3) 327448 X I .5-5 '(392-9~3.6) Absent Paratype NTM 2755 341480 x 8-20 ( 4 0 5 . 0 ~15.6) 191-553 xO-5-8 (422.5~4.1) Absent Specimen NTM 21355 273612 x 6-17 ( 4 7 4 . 8 11.7) ~ (oxea) 352-741 x4-15 (545 . O X 10.2) (oxealanisoxea) Absent 60-110 x3 (oxea) Absent C. montebelloensis, sp. nov - 690-1500 x 10-30 (strongyle) 449-560 x 14-28 (490 x 20) (style) Holotype AM G4353 263-465 x4-15 (346.4~9.2) (subtylostyles) 750-2000 x 30-40 (style) 543-782 x 9-2 1 (647.5~14.3) (style) C . palmata, sp. nov. 521-808 x 5-1 8 ( 7 2 2 . 7 ~10.3) C . clavata ~hiele* 550-1500 x 10-50 (style) C . cuneiformis ~ e r g q u i s t ~ 552-780 978-1699 ~3.44-6 ~7-12 (670x3 .O) (1450x 10) (style/oxea) (style) C . dichotoma (Whitelegge) 523-743 x 11-19 (635-8~15.2) (subtylostyles) C . erecta T'hieleA 830-1020 x 7-25 Absent Same as choanomal megascleres (subtylostyles) Absent 70-100 x? (oxeas) Absent 1332 J. N. A. Hooper Table 18. Continued Material Choanosomal megascleres - 250-960 x 13-20 (oxea) - 560-620 x 18-20 (stylelanisoxea) Subectosomal megascleres Ectosomal megascleres Echinating megascleres 65-100 x 2-3 (oxea) Absent C . eqansa ~ h i e l e ~ 1000-1200 x 10-30 (stylelstrongyle) C . microxephora ( ~ i r k ~ a t r i c k ) ~ 670-1250 X 12-16 (strongyle) 70 x3 (oxea) Absent 100-125 ~2.5-5 (oxea) Absent 70-80 x? (oxea) Absent C . minor Pulitzer-~inali~ I:190-500 X 11-19 (style) II:1150 x9.5 (strongyle) - A 1000-1400 ~7-16 (style) C . ramosa T'hieleA -750-11x 25 (style/strongyle) From Thiele (1898: 57-8) and Hoshino (1975: 33). From Bergquist (1970: 18). From Kirkpatrick (1903: 242). From Pulitzer-Finali (1983: 520). Description Shape. Erect, stipitate branching sponge (340 mm high, 370 rnm maximum breadth), with short robust cylindrical stalk (75 mrn high, 28 mm diameter), evenly cylindrical branches (9-16 mrn diameter), which bifurcate repeatedly and taper towards apex (5-9 mm diameter), and branching in 1 plane. Colour. Olive-brown alive (on-deck) and in ethanol (Munsell 7 -5YR 616). Oscula. Large and small pores (0.5-1 a5 mm diameter) dispersed on lateral sides of branches, but these are sparse and located mainly near branch endings. Texture and surjace characteristics. Texture firm and flexible, stalk woody, and branches barely compressible. Surface smooth and not optically hispid, composed of adjoining oval, elliptical or rounded small plates (2-3 mm diameter), which form very low microconules when alive, but flush with surface in preserved state. Surface-plates are produced by discrete groups of subectosomal extra-axial spicule bundles protruding through surface. Ectosome and subectosome. Ectosome slightly microscopically hispid, formed by a nearly continuous dermal palisade of brushes of ectosomal oxeas, which stand perpendicular to surface. Ectosomal brushes associated with groups of subectosomal extra-axial megascleres, but dispersed between extra-axial bundles. Subectosomal skeleton plumose, consisting of radially arranged, long, thick, extra-axial styles, with bases embedded in axial core. Individual extra-axial spicules stand perpendicular to axis, surrounded by tight sheath of subectosomal oxeaslanisoxeas. In choanosomal region these extra-axial oxeaslanisoxeas are arranged radially, whereas closer to surface they fan out to form plumose tracts. Together, extra-axial styles and oxea/anisoxea bundles form discrete spicule tracts, up to 700 p m diameter near axis and up to 1500 p m diameter near periphery. Between these bundles of extra-axial megascleres are halichondroid tracts of subectosomal megascleres and relatively heavy granular collagenous spongin. Australian Raspailiidae Fig. 75. Ceratopsion axifera (Hentschel) (specimen NTM 23035): a, choanosomal axial oxeas; b, subectosomal extra-axial styles/anisoxeas; c, ectosomal auxiliary oxeas; d , section through peripheral skeleton; e, known Australian distribution. Choanosome. Choanosomal skeleton consists of condensed spongin fibres (120-270 pm diameter) forming a tight axial reticulation and producing oval meshes (120-260 pm diameter). Choanocyte chambers eliptical (280x70 pm).. Fibres relatively heavily invested with spongin but virtually unpigmented. Fibres sparsely cored with hastate axial oxeas. Echinating megascleres absent, and extra-fibre spongin in axial skeleton also virtually absent. Megascleres (refer to Table 18 for dimensions). Choanosomal axial megascleres oxeas or anisoxeas, relatively small and thin, usually straight or symmetrically curved, occasionally sinuous, with hastate rounded, stepped or marnmilliform tips. Subectosomal extra-axial spicules of 2 distinct categories: I, long, thick, straight styles with evenly rounded bases and fusiform points; 11, shorter, stout straight or slightly curved oxeas or anisoxeas with rounded hastate tips, occasionally modified to true styles, which form a sheath around larger styles. Ectosomal auxiliary megascleres exclusively styles, ranging from very thin raphidiform examples to relatively thick spicules, straight and tapering to fusiform points. Echinating megascleres absent. Microscleres are absent. J. N. A. Hooper Remarks In growth form, surface features and certain aspects of skeletal construction this species shows superficial similarities to Dendropsis bidentifera Ridley & Dendy (1887: 192). Ceratopsion axifera is quite unusual for the genus in particular, and Raspailiidae in Fig. 76. a-b, Ceratopsion axifera (Hentschel): a, specimen (NTM 2303.5': (scale in centimetres); b, SEM of skeletal structure. c-e, C. montebelloensis, sp. nov.: c, holotype (NTM 23388) (scale in centirnetres); d , ectosomal features (scale = 1 mm); e, SEM of skeletal structure. Australian Raspailiidae 1335 general, by its minute plate-like surface ornamentation produced by the discrete bundles of extra-axial megascleres that protrude through the surface. These bundles also have a peculiar spicule composition, consisting of a single central style surrounded for most of its length by a sheath of oxeas/anisoxeas. Ceratopsion montebelloensis, sp. nov. (Figs 76c-e, 77, 110b; Table 18) Material Examined Holotype. NTM 23388 (fragment NCI Q66C 1528-A): 12 km from Hag I., Monte Bello Is, NWS, W.A., 20" 27.8'S., 115' 30.9'E.,6 m depth, 27.viii.1988, coll. D. Low Choy & NCI (stn NWS72). Substrate and Depth Range Coral rubble substrate, 6 m depth. Geographical Distribution Northwest Shelf, W.A. (Fig. 77e). Fig. 77. Ceratopsion montebelloensis, sp. nov. (holotype NTM 23388): a, choanosomal axial styles; b, subectosomal extra-axial styles; c, ectosomal auxiliary styles/anisoxeas; d, section through peripheral skeleton; e, known Australian distribution. J. N. A. Hooper Description Shape. Stipitate, flabellate vasiform sponge (210 mm high, 240 mm maximum breadth), with short flattened broad stalk (30 rnm high, 41 mm broad, 18 mm thick), and thinly lobate flabellate branches (110-170 mm high, 4-8 rnm thick); branching in more than 1 plane, forming vase-shape, with tapering rounded or slightly digitate margins. Colour. Live coloration grey-brown (Munsell 7.5YR 612) (Fig. 110b), same in ethanol. Oscula. Not observed. Texture and surface characteristics. Texture in ethanol stiff and flexible, barely compressible. Surface even, without conules or other ornamentation, prominently hispid, and it feels villose to touch. Occasional lobate digitate growths occur on either side of flabellate branches, probably formation of young branches. Ectosome and subectosome. Ectosomal skeleton consists of long brushes of ectosomal auxiliary spicules, congregated around protruding subectosomal extra-axial styles. Density of brushes on surface directly related to distribution of extra-axial megascleres, so that between extra-axial spicules ectosomal skeleton simply membraneous. Subectosornal skeleton consists of radially arranged individual extra-axial styles, with bases embedded in axial skeleton and points protruding through surface. Extra-axial megascleres closely compacted in most places on ectosorne, producing prominently hispid-villose surface. Mesohyl in peripheral skeleton much denser than in axial core, with heavy granular brown pigmented spongin. Choanosome. Axial skeleton contains small condensed core of light spongin fibres (25-75 pm diameter) running longitudinally through branches, forming elongate meshes (up to 230x60 pm), cored by one or few choanosomal styles. Styles in core run longitudinally through branches, but towards edge of fibre-core choanosomal styles become plumose, radiating outwards towards surface. Bases of extra-axial styles embedded in outer edge of axial fibres. Echinating megascleres absent. Tracts of plumose axial spicules terminate about 400 pm below surface, producing prominent zone of transition between axial and ectosomal regions. Megascleres (refer to Table 18 for dimensions). Choanosornal axial styles thin, slightly curved centrally or near their points, with evenly rounded bases and fusiform points. Subectosomal extra-axial styles relatively long and thick, usually slightly curved at centre but occasionally straight, with evenly rounded bases and hastate or stepped tips. Ectosomal auxiliary spicules vary from relatively long and thick styles to raphidiform anisoxeas, usually straight, with evenly rounded or hastate bases and fusifom points. Echinating megascleres absent. Microscleres absent, although raphidiform ectosomal megascleres may be present. Remarks This species is very close to the New Zealand species C . cuneiformis, but differs significantly in the dimensions of the megascleres (Table 18). In growth form it somewhat resembles C. expansa and C . erecta, but spicule geometry and dimensions are quite different. Like C. dichotoma (Fig. 74a-f) the differentiation of spicules into three categories in C . montebelloensis appears to be slightly artificial. The geometry of styles in the ectosomal, axial and extra-axial skeletons is more or less identical, whereas these spicules differ notably in size between each region. Etymology This species is named after the Monte Bello Is from which the holotype was collected. The type locality, Flag I., was one of several radioactive fallout areas affected during the testing of early British atomic weapons during the 1950s. Australian Raspailiidae Ceratopsion palmata, sp. nov. (Figs 78, 79; Table 18) Material Examined (All material collected by the author using SCUBA, unless otherwise indicated.) Holotype. NTM 23051: N. of Amphinome Shoals, NWS, W.A., 19" 19 -7-23.3/S., 119" 08.812.2'E., 50 m depth, 19.vii.1987 (U.S.S.R. RV 'Akademik Oparin', stn NWS55, beam trawl). Paratypes. NCI Q66C-1505-A (fragment NTM 23487): N. of Barrow I., Exmouth Gulf region, NWS, W.A., 20" 38 8 IS., 115"28.8/E., 17 m depth, 26.viii.1988, coll. D. Low Choy & NCI (stn NWS99, SCUBA).NTM 2755: NW. of Yampi Sound, NWS, W.A., 15" 27.4'S., 121"49.0rE., 76 m depth, 29.iv.1982 (CSIRO RV 'Sprightly', stn Don. 17, SP4-42, dredge). Other material. Cobourg Peninsula, N.T.: NTM 21355: Coral Bay, Port Essington, 11" 11 3 S., 132' 03 -75IE., 5 - 5 m depth, 16.v.1983 (stn CP60). Wessel Is, N.T.: NTM 23923: Bay N. side of Cumberland Strait, 11' 27.5 S., 131' 28 - 8/E., 20 m depth, 14.xi.1990 (stn WI-6). Northwest Shelf, W.A.: NTM 2644: W. of Eighty Mile Beach, 19' 33.05/S., 119" 05.07'E., 35 m depth, 30.iv.1982, coll. CSIRO RV 'Sprightly', dredge (stn SP4/82-78). - Substrate and Depth Range Shallow coastal rock and coral rubble reefs, with sand and shell grit sediment, 5.5-76 m depth. Geographical Distribution Northwest Shelf, W.A. to the Wessel Is, N.T. (Fig. 78e). Description Shape. Stipitate, palmate-digitate sponges (44-128 mm high, 32-70 mm maximum breadth), with short cylindrical stalks (12-20 mm long, 4-5 mm diameter), producing fused digits forming flattened palmate fronds in holotype (50 mm long, 30 mm wide, up to 10 mm thick) branching in 2 planes, and/or flattened cylindrical digits (11-38 mm long, 6-11 rnm diameter), with cylindrical-digitate margins (13-25 mm long, 6-8 mm diameter), dividing di- or trichotomously, and taper to rounded tips at apex of digits. Colour. Shallow water specimen (NTM 21355) dark orange-brown alive (Munsell 2.5YR 6/8), whereas deeper water material dark grey-brown in life (5YR 412). Internal coloration orange-brown. Oscula. Not observed. Texture and surjCace characteristics. Texture firm and flexible. Surface optically even, microscopically villose and very hispid. Ectosome and subectosome. Ectosomal skeleton with darkly pigmented heavy layer of spongin, up to 300 p m thick, with discrete bundles of ectosomal oxeas/anisoxeas protruding, surrounding extra-axial subectosomal megascleres. Ectosomal spicule bundles usually perpendicular to surface, sometimes lying at oblique angles or tangentially. Subectosomal extra-axial skeleton with 2 components: radial paucispicular bundles of long extra-axial styles, originating approximately halfway from axial skeleton (not embedded in axial fibres) and extending through surface; and plumose (or plumo-reticulate) brushes of choanosomal axial oxeas ascending from axial core close to surface, but not piercing ectosome. Large subdermal cavities present, marking limit to which axial megasclere brushes extend. Mesohyl throughout ectosomal and subectosomal regions contains relatively heavy dark brown granular spongin. Choanosome. Choanosomal axial skeleton with condensed core of spongin fibres forming close-set reticulation; larger primary fibres (90-150 p m diameter) run longitudinally through branches, cored by multispicular tracts of axial oxeas, interconnected by secondary transverse fibres (40-70 p m diameter) cored by uni- or paucispicular tracts of axial oxeas. Meshes produced by fibre anastomoses elongate or oval (30-70 p m diameter), usually containing moderate amounts of granular spongin. Fibres heavily invested with spongin, but only lightly pigmented. Echinating megascleres absent. Choanocyte chambers oval, 40-60 pm diameter. J. N. A. Hooper Fig. 78. Ceratopsion palmata, sp. nov. (holotype NTM 23051): a, choanosomal axial and extra-axial oxeas; b, subectosomal extra-axial styles; c, ectosomal auxiliary oxeas/anisoxeas; d, section through peripheral skeleton; e, known Australian distribution. Megascleres (refer to Table 18 for dimensions). Choanosomal axial and extra-axial megascleres thick oxeas, usually slightly curved at centre, tapering to sharp fusiform points or with stepped points. Subectosomal extra-axial megascleres long, relatively thin styles, with slight curvature near basal end, evenly rounded bases, tapering to sharp points. Ectosomal auxiliary spicules relatively long thin oxeas or anisoxeas, always straight, thicker at midsection than apex, with raphidiform points or asymmetrical hastate bases. Echinating megascleres absent. Microscleres absent. Remarks In growth form C. palmata is closest to C. ramosa, in that both have biplanar palmate4igitate branching. In spiculation the species approaches C. expansa to a certain extent, having choanosomal oxeas, subectosomal styles and ectosomal oxeas, but spicule dimensions differ considerably between the two species (Table 18). The presence of ectosomal spicules, which are relatively long in comparison with choanosomal oxeas, is characteristic of this species, and in this respect C. palmata has affinities with C. cuneiformis. Paratype NTM 2755 and specimen NTM 21355 are slightly different from the other two specimens of C. palmata in spicule dimensions, having marginally larger megascleres in all categories (Table 18), particularly the specimen from the Cobourg Peninsula, in dimensions Australian Raspailiidae Fig. 79. Ceratopsion palmata, sp. nov.: a, holotype (NTM 23051) (scale = 30 mm); b, SEM of peripheral skeleton; c, SEM of skeletal structure. of its extra-axial spicules, but these differences are not considered to be significant since in all other details the specimens are identical. Etymology This species is named for its shape, like the palm of the hand (L., palmatus). Genus Thrinacophora Ridley & Dendy, 1886: 483; 1887: 193; Dendy, 1905: 186 Thrinacophora Ridley, 1885: 572,Ridley (in part); Hallmann, 1916b: 634-7. Type species: Thrinacophora funiformis Ridley & Dendy, 1886: 484 (by monotypy; holotype BMNI-I 1887.5.2.53, from off Bahia, Brazil) (Fig. 81h-i). Diagnosis Cylindrical, arborescent or encrusting and massive growth forms. Surface not prominently hispid but evenly conulose. Axial and extra-axial skeletons well differentiated. Choanosomal skeleton axially condensed, with dense reticulate core occupying large proportion of branch diameter, cored by short, stout oxeas, anisoxeas or occasionally styles. Echinating megascleres absent. Extra-axial skeleton plumose, with uni- or paucispicular tracts radiating towards surface, cored by long, thick subectosomal styles or anisoxeas. Extra-fibre megascleres abundant and identical to ectosomal spicules. Ectosomal skeleton with a special category of styles (sinuous and apically pronged in the type species), forming erect or paratangential brushes around extra-axial spicules. Structural megascleres short oxeas and long styles, some with strongylote or oxeote modifications, and ectosomal styles. Microscleres raphides occurring singly or in bundles (trichodragmata). 1340 3. N. A. Hooper Remarks Hallmann (1916b) restricted Thrinacophora to include only species with a specialised ectosomal skeleton of styles. He redistributed other nominal species (i.e. Thrinacophora, sensu Dendy, 1905) amongst his new genera Dragmaxia, Dragmacidon and Axidragma (=Tragosia Gray) in the Axinellidae, and he included only 11 species in Thrinacophora. Of these species allocated to Thrinacophora by Hallmann, only the following should remain there: T. jkniformis Ridley & Dendy (Fig. 81h-0, T. cervicornis Ridley & Dendy (1887: 194) from the Philippines (Figs 80, 81a-g), T. spinosa Wilson (1902: 400) from Puerto Rico, Raspailia (Syringella) rhaphidophora Hentschel (1912: 371) (which is a synonym of T. cervicornis) from the Arafura Sea, and T. incrustans Kieschnick (1896: 533) from the Moluccas. Both Thiele (1903: 935) and Hallmann (1916b: 636) had doubts concerning the validity of Kieschnick's poorly described encrusting species. No further comment is appropriate until the holotype of the species is located and redescribed, but it is not known if this material still exists (it was not found in PMJ collections; F. Wiedenmayer, personal communication). The presence of special ectosomal megascleres and a typical raspailiid ectosomal skeletal structure, and the presence of axial condensation and axial and extra-axial differentiation indicate that Thrinacophora has obvious affinities with the Raspailiidae, although earlier authors (e.g. Hallmann 1916b) included the group with Axinellidae. It is interesting to note that these characteristics are also represented in a small group of Microcionidae [viz. Clathria (Axociella) canaliculata (Whitelegge), C. (A.) cylindrica (Ridley & Dendy), and C. (A,) thetidis (Hallmann)] that closely resemble species like T. cervicornis, although the ectosomal specialisation in the microcionids is not the same as the specialised raspailiid condition. Thrinacophora is similar in many respects to Ceratopsion, with a specialised ectosomal skeleton, large extra-axial rnegascleres -and a more or less radial arrangement of the extra-axial skeleton, which protrudes a long way through the surface, well developed axial and extra-axial differentiation, and the lack of echinating megascleres. However, Thrinacophora has a dense axial reticulation of short, stout choanosomal spicules, usually oxeas but also including anisoxeas or rarely styles, and this axial region lacks any evidence of spongin fibres. The criss-cross of axial spicules, which occupies most of the diameter of branches, closely resembles the axial skeleton of Reniochalina (also known as Axiamon) and other axinellids. On this basis, the genus is also distinguished from Raspailia (Syringella). Earlier authors (e.g. Ridley & Dendy 1887; Hallmann 1916a) considered that the possession of raphides as microscleres was an important diagnostic feature at the generic level, placing secondary emphasis on such things as ectosomal specialisation (a character that is used here to define the Raspailiidae), megasclere geometry and skeletal architecture. In Thrinacophora at least, it is unlikely that the possession of raphides can define the genus (as supposed by Ridley & Dendy 1887: 193), but it is certainly another character that can be used to differentiate the genus from the allied Ceratopsion. Thrinacophora cervicornis Ridley & Dendy (Figs 80, 81a-g; Table 19) Thrinacophora cewicornis Ridley & Dendy, 1886: 483.-Ridley & Dendy, 1887: 194-195, pl. 36, fig. 1, pl. 40, fig. 4; Hentschel, 1912: 416-417; Hallmann, 1916b: 636. Raspailia (Syringella) rhaphidophora Hentschel, 1912: 371-372, pl. 13, fig. 10, pl. 19, fig. 31. Thrinacophora rhaphidophora.-Hallmann, 1916b: 635. Material Examined Lectotype (here designated). BMNH 1887.5.2.60: Off the Philippines, 11" 37'N., 123' 31 'E., 36 m depth, 17.i.1875 (HMS 'Challenger', stn 208, dredge). Holotype of T. rhapidophora. SMF 1904: NE. of Penambulai, Aru I., Arafura Sea, 6' S, 134' 501E., 8 m depth, 2.iv.1908, coll. H. Merton (stn 10, dredge) (schizotype MNHN LBIM DCL 2373). Other material (all material collected by the author unless otherwise indicated). Indonesia: SMF 1554: Sungi Barkai, Aru I., Arafura Sea, 6'S, 134' 501E., 18 m depth, 10.iv.1908, coll. H. Merton (stn 14, dredge). Arafura Sea, N.T.: NTM 2607: Cootarnundra Shoals, N. of Melville Australian Raspailiidae 1341 I., 10' 49-07/S., 129' 12.0g1E., 31 m depth, 6.v.1982, coll. B. Thom & R. Lockyer (ref. 319). Northwest Shelf: NTM 21182: W. of Bedout I., Amphinome Shoals, NWS, W.A., 19'30-9'S., 118" 48.7'E., 40 m depth, 26.iv.1983 (CSIRO RV 'Soela', stn NWS8, S02-B7, beam trawl). NTM 21291: NW. of Bedout I., Amphinome Shoals, NWS, W.A., 19" 04-3'S., 118'50-5'E., 83 m depth, 27.iv.1983 (stn NWS14, S02-B5). NTM 21873: N. of Bedout I., Amphinome Shoals, NWS, W.A., 19" 05 -7'S., 118"57-4'E., 83 m depth, l.ix.1983, coll. T. Ward (CSIRO RV 'Soela', stn NWS28, S04-133). NTM 2651: W. of 80 Mile Beach, NWS, W.A., 19' 33-5/S., 119'05-7/E., 35 m depth, 4.v.1982 (CSIRO RV 'Sprightly', sm Don. 15, SP4-78, dredge). Scott Reef, W.A.: PIBOC unreg.: 16" 36.7'S., 121" 1l.l1E;, 50m depth, 17.xi.1990, coll. V. Krasochin, U.S.S.R. RV 'Akademik Oparin', dredge (stn 39). Substrate and Depth Range Offshore coastal rock reef, shallow coastal coral rubble reef, and blue-mud substrates, 8-83 m depth. Geographical Distribution Philippines, Arafura Sea and Northwest Shelf, W.A. (Fig. 80g). Description Shape. Stipitate, digitate sponges (110-320 mm high, 74-160 mm maximum breadth), with short cylindrical or flattened stalk (26-35 mm high, 6-10 mm diameter), irregularly branching bifurcated digits, cylindrical or flattened paper-thin (22-220 mm long, 3-7 mm diameter), tapering to sharp or chiselled points at their ends. Colour. Orange-brown to red-brown alive (Munsell 10R 618 to 2.5R 4/6), darkening on exposure to air, fading to yellow-grey to grey-brown in ethanol. Oscula. Not observed. Texture and surface characteristics. Texture of stalk woody, branches firm and flexible. Surface optically even but covered with microconules, each with a single long protruding extra-axial style protruding through surface up to 4 mm. Ectosome and subectosome. Ectosome prominently hispid; surrounding hispidating spicules are discrete brushes of large ectosomal styles or anisoxeas, forming dense clusters perpendicular to surface. Peripheral skeleton with relatively heavy layer of granular spongin, 200-600 pm thick, irregularly dispersed ectosomal spicules occurring singly, together with abundant trichodragmata. Subectosomal extra-axial skeleton with a radial array of very long styles protruding through surface; spicules occur singly, never in bundles, with bases embedded deeply within axial skeleton. Extra-axial skeleton occupies only a small proportion (10-30%) of branch diameter. Choanocyte chambers oval, 40-110 pm diameter. Choanosome. Choanosomal axial skeleton condensed, consisting of irregular reticulation of tightly packed choanosomal oxeas (or anisoxeas), without a visible spongin fibre component; spicules not obviously organised into major tracts, although many run longitudiially through branches. Axial skeleton occupies majority of branch diameter; shape of axial core usually irregularly stellate, with arms of star corresponding to distribution of surface conules and extra-axial megascleres. Mesohyl in axial core, when not obscured by spicule tracts, contains only little spongin. Echinating megascleres absent. Megascleres (refer to Table 19 for dimensions). Choanosomal axial megascleres thick, straight or slightly curved at centre, with hastate points or stepped tips, occasionally sharply pointed; predominantly oxeas, although anisoxeote or stylote modifications may occur. Subectosomal styles very long, relatively thin, straight or slightly curved, with evenly rounded bases and tapering to sharp points. Ectosomal auxiliary megascleres straight styles with tapering hastate bases and fusiform raphidiform points, occasionally anisoxeote with asymmetrical fusiform ends, with thickest section occurring centrally. Echinating megascleres absent. Microscleres raphides which usually occur in bundles (trichodragmata). J. N. A. Hooper Fig. 80. Thrinacophora cervicornis Ridley & Dendy (holotype BMNH 1887.5.2.60): a, choanosomal axial oxeas; b, subectosomal extra-axial style; c, ectosomal style/anisoxea; d, trichodragmata; e, section through peripheral skeleton; f, Raspailia (Syringella) rhaphidophora Hentschel (holotype SMF 1904), choanosomal styles/anisostrongyles; g, known Australian distribution. Remarks Ridley & Dendy (1887) erected T. cervicornis based on two specimens, but only one (designated here the lectotype) could be found in the BMNH. The redescription of this species, presented here, from the Northwest Shelf of W.A. is a new record for Australian territorial waters, although the species is known from Am I., which lies only 300 km from northern Australian reefs, and is undoubtedly a part of the Dampierian zoogeographical province. The Philippine lectotype and Hentschel's (1912) Arafura Sea specimen are virtually identical, with closely corresponding spicule dimensions (Table 19), although trichodragmata in Hentschel's material are significantly larger than in the lectotype, and in sections these structures are not as well aggregated into tight bundles as in the Philippines material. Both Ridley & Dendy (1887) and Hentschel (1912) recorded that extra-axial styles in the lectotype and in the Arafura Sea specimen were 5200 pm long, but re-examination of both specimens showed that this estimate was far too high; the maximum sizes observed were 4543 pm and 4353 pm, respectively. Additional material from the Northwest Shelf had extra-axial styles that were even smaller (3705 p m maximum length). Australian Raspailiidae 1343 Fig. 81. a-g, Thrinacophora cervicornis Ridley & Dendy: a, holotype (BMNH 1887.5.2.60); b, schizotype of Raspailia (Syringella) rhaphidophora Hentschel (schizotype MNHN LBIM DCL 2373), a junior synonym of T. cervicornis; c, specimen (NTM 21872) (scale = 30 mm); d, SEM of skeletal structure; e, SEM of extra-axial skeleton;f , SEM of peripheral axial skeleton; g, ectosomal specialisation (scale = 250 pm). h-i, T. funiformis Ridley & Dendy (type species of the genus Thrinacophora Ridley): h, holotype (BMNH 1887.5.2.53) (scale = 30 mm); i, section through peripheral skeleton (scale = 1 mm). Thrinacophora rhaphidophora is merged here with T. cervicornis. The two nominal species appear to differ only in seemingly minor details: the degree to which branches are flattened laterally (compare Fig. 81b and 81a); the variability in spicule dimensions (refer to Table 19), whereby both axial and extra-axial megascleres of T. rhaphidophora are slightly smaller than those of T. cervicornis; and the styloid modifications to axial megascleres in the holotype of T. rhaphidophora. Nevertheless, in other details of skeletal construction and spiculation the two species are virtually identical and obvious synonyms. It is peculiar that Hentschel (1912) recorded both nominal species from Aru I. but failed to recognise 1344 J. N. A. Hooper Table 19. Comparisons in spicule measurements between specimens of Thrinacophora cervicornis Ridley & Dendy Measurements are given in micrometres, presented as ranges (and means) of lengthx width, and represent 25 spicules measurements per specimen for each spicule category Material BMNH 1887.5.2.60 SMF 1554 NTM various (n = 4) SMF 1904 Choanosomal oxeasl anisoxeas Subectosomal styles Ectosomal styles/ anisoxeas Lectotype 212-325 1930-4543 288-608 x 9-19 x 10-31 ~0.5-5.5 ( 2 7 9 . 3 ~ 1 4 ~ 8 ) (3433.7~21-0) ( 4 7 3 ~ 1 ~ 2 . 8 ) 252-339 x 10-21 (303.5~16.3) 189-317 X 5-25 ( 2 4 8 . 2 ~10.4) (oxeas) 194-282 x 5-19 (246-9~11.8) (styles/oxeas/ anisoxeas) Specimens 17854353 273-563 x 6-28 xO-5-6 (3055.7~16-8) (439.7~3.0) 1110-3705 328-652 X 11-25 x 1.5-5 (2384.1~16.5) (506.7~3.3) (styles) (styles) Holotype of T. rhaphidophora 1435-2040 402-604 x 7-21 x 3-8 (1774.8~14.8) ( 4 9 6 - 3 ~ 5 . 5 ) (styles) (styles1 anisoxeas) Echinating megascleres Trichodragmata Absent 51-96 x 6-10 (66.3~8.3) Absent 72-1 11 X9-13 (92.3~10.8) 53-101 x6-11 (72-4~8.3) (raphides) Absent Absent 61-94 X7-13 (77.1~933) (raphides) their obvious similarities, and even more peculiar that he referred them to two different genera. Important characters such as skeletal construction, ectosomal specialisation, and gross spicule geometry indicate that the two are, at the very least, sibling species. Genus Axechina Hentschel Axechina Hentschel, 1912: 417.-.de Laubenfels, 1936: 102. Type species: Axechina raspailioides Hentschel, 1912: 417-8 (by original designation and monotypy; holotype SMF 991, schizotype MNHN LBIM DCL 2261, from the Arafura Sea) (Fig. 82). Diagnosis Bushy flabellate growth form. Surface rugose, conulose, and hispid. Choanosomal skeleton with 2 components: condensed, tightly reticulate axis cored by oxeas with spined points; plumose peripheral region formed by columns of smooth anisoxeote spicules. Echinating megascleres absent. Extra-axial skeleton radial, with tracts of long, thick, subectosomal styles embedded in plumose section of axis, with tracts ascending to surface and projecting through it. Ectosomal skeleton composed of peculiarly curved or sinuous toxiform styles with spined terminations, grouped around a central extra-axial style in typical raspailiid formation. Structural megascleres include oxeas with spined terminations, anisoxeas, large styles or anisoxeas with rounded bases, and dermal flexuous toxiform styles with spined terminations. Microscleres absent. Remarks The type and only species of Axechina has a typical raspailiid ectosome in which ectosomal spicules occur in brushes surrounding extra-axial megascleres. Although the genus was previously included with the Axinellidae (Hentschel 1912), its affinities, as indicated by its Australian Raspailiidae 1345 ectosomal skeleton, obviously lie with the Raspailiidae. The ornamentation of choanosomal megascleres is also superficially similar to Amphinomia, and axial skeletal architecture is close to Reniochalina (Axinellidae), Ceratopsion and Thrinacophora (Raspailiidae). The genus could probably be reasonably included with Ceratopsion on the basis of its spiculation, or Thrinacophora in having at least part of its axial skeleton consisting of a dense criss-cross reticulation of spicules without obvious fibres, but it fits easily into neither genus. The genus is retained on a tentative basis, differentiated from others by the presence of the following features: an axial reticulate core, a plumose region near the periphery of the axis, an extra-axial radial skeleton and a plumose ectosomal skeleton, the absence of choanosomal fibres or definite axial spicule tracts (cf. Thrinacophora), the presence of basal and apical spination on both choanosomal and ectosomal megascleres, and the absence of echinating megascleres. Axechina raspailioides Hentschel (Fig. 82) Axechina raspailioides Hentschel, 1912: 417418, pl. 14, fig. 5, pl. 21, fig. 55. Material Examined Holotype. SMF 991: SW. of Kola, AN I., Arafura Sea, 5' 32'S., 134' 25'E., 8-10 m depth, 01.iv.1908, coll. H. Merton (stn 9, dredge). Other material. Wessel Is, N.T.: NTM 23903, NCI Q66C 4680-C: N. of Cape Wilberforce, Melville Bay, Gove Peninsula, 11' 52.6'S., 136' 33.3 E., 25 m depth, ll.xi.1990 (stn WI-I). NTM 23950: W. headland, Rimbija I., Cape Wessel, 11' 0.5/S., 136' 43 -8/E., 15 m depth, 16.xi.1990 (stn WI-9). NTM 23967: S. of Truant I., English Company Is, 11" 40.45/S., 136' 50.2'E., 26 m depth, 19.xi.1990 (stn WI-13). Substrate and Depth Range Rock reef, 8-26 m depth. Geographical Distribution Arafura Sea (Fig. 82f). Description Shape. Stipitate, bushy arborescent sponges (58-70 mm high, 45-55 mm maximum breadth), with a short irregularly cylindrical stalk (12-16 mm long, up to 6 mm diameter), and with bushy more-or-less cylindrical branches, prominently conulose or clathrous (12-32 mm long, up to 11 mm diameter). Colour. Live coloration orange to red-orange (Munsell 5YR 6/10), beige-brown in ethanol. Oscula. Pores (up to 1 - 5mm diameter) scattered over branches, between conules. Texture and surface characteristics. Texture firm and flexible. Surface prominently macro- and rnicroconulose, with large megascleres protruding through conules, extending through surface for a larger distance, and between conules exists a skin-like covering stretched across surface. Ectosome and subectosome. Ectosomal skeleton hispid from 3 sources: extra-axial styles extending a long distance through surface; brushes of sinuous styles grouped around these styles; and choanosomal anisoxeas from plumose section of axial skeleton. Spongin relatively heavy, darkly pigmented and granular in dermal region, varying considerably in thickness, from 60 p m thick on surface conules up to 210 p m thick between conules. Extra-axial skeleton with long subectosomal styles embedded in plumose section at periphery of axial skeleton, and spicules ascend to surface and project through it for approximately 1 mm. Subectosomal region occupies only small proportion (10-20%) of branch diameter. Choanosome. Choanosomal skeleton has 2 structural components: condensed axis of tightly reticulate criss-crossed spicule tracts running longitudinally through branches, resembling Reniochalina for example, occupying majority of branch diameter. Stellate J. N. A. Hooper Fig. 82. Axechina raspailioides Hentschel (holotype SMF 991) (type species of the genus Axechina Hentschel): a, choanosomal axial oxeas; b, choanosomal extra-axial anisoxeas; c, subectosomal extra-axial style; d, ectosomal auxiliary styles; e, section through peripheral skeleton;f , known Australian distribution; g, holotype (scale = 30 mm); h, photomicrograph of axial skeleton (scale = 1 mm). shape of axis more-or-less corresponds with distribution of surface conules. Axis cored by oxeas with spined points. Plumose or radial columns of smooth anisoxeote spicules embedded in axis, on its periphery, producing cactiforrn appearance of axis; columns branch but do not anastomose, and tips of anisoxeas barely pierce surface. Mesohyl in axial region contains little spongin, also lacking choanosomal fibres, but entire region obscured by heavy criss-crossed spicules. Choanocyte chambers oval and 30-80 p m diameter. Megascleres. Choanosomal megascleres of 2 distinct types: oxeas in axial core relatively thin, usually symmetrically curved, with smooth shafts and hastate points covered with Australian Raspailiidae 1347 - microspines [I954237 3)-264 x 8 4 1 0 4)-16 pm]; anisoxeas in plumose axial region thick, large, entirely smooth, with hastate points [305-(355 9 ) 4 2 5 x 16-(20.8)-24 pm]. Subectosomal extra-axial styles long, relatively thin, straight or slightly curved near base, with evenly rounded bases and tapering fusiform tips [1044-(1927 8)-2760 x 15(17 2)-20 pm]. Ectosomal auxiliary megascleres flexuous, curved or sinuous, rarely straight styles, with evenly rounded bases and either hastate or tapering-rounded tips. Bases have terminal microspination and points have terminal or subterminal granular microspination [23 14256-0)-305 x 2 4 4 3)-6 pm]. Echinating megascleres absent. Microscleres absent. - - - Remarks This species has only been recorded from 134-136" longitude, on northern Australian coastal reefs. The material cited above is the first record for the species in Australian waters, and it is suspected that this species is endemic to the Aru and Wessel Is. Hentschel (1912) omitted to record that in the type specimen of A. raspailioides the ectosomal auxiliary styles have spines on both the basal and distal ends, similar to the spination on choanosomal megascleres, although the two forms of spicules do not appear to be otherwise related. Genus Echinodictyum Ridley Echinodictyum Ridley, in Ridley & Duncan, 1881: 493.-Ridley, 1884: 454; Ridley & Dendy, 1887: 164, Topsent, 1894: 19; Dendy, 1896: 44; 1905: 175; 1916: 129; Thiele, 1899: 15; Kieschnick, 1900: 570; Hentschel, 1911: 385; 1912: 369; Hallmann, 1912: 171-5; Burton, 1931: 348; Burton & Rao, 1932: 347; LBvi, 1965: 19; 1969: 966. Kalykenteron Lendenfeld, 1888: 216.-Hallmann, 1912: 171 (type species K, elegans Lendenfeld, 1888: 216, here designated) (Fig. 101f). Dictyocylindrus.-Carter, 1879b: 297 (in part); Ridley, 1884: 454. Not Dictyocylindrus Bowerbank, 1862: 1108 (type species Spongia hispida Montagu, 1818: 81, by original designation). Kieplileta de Laubenfels, 1954: 116 (type species K. antrodes de Laubenfels, 1954: 116, by monotypy). Type species: Spongia bilamellata, var. B, Larnarck, 1816: 436 (in part) [by original designation and monotypy (Ridley, in Ridley & Duncan 1881: 493)l (Fig. 101c); junior synonym of Spongia mesenterina Lamarck, 1814: 444 (Topsent 1932: 101) (Fig. 101a). Diagnosis Erect, vasiform, flabellate, ramose or massive growth forms. Surface typically rugose, with ridges, conules and other processes. Texture characteristically harsh, brittle and flexible, reflecting the high ratio of silica to spongin in skeleton. Axial and extra-axial skeletons undifferentiated. Choanosomal skeleton not condensed, and architecture typically irregularly reticulate. Spongin fibres usually massive, fully cored by oxeas, and echinated by acanthostyles. Extra-axial skeleton vestigial, consisting of individual subectosomal styles embedded in peripheral fibres, and projecting through ectosome or merely dispersed between fibres. Ectosomal skeleton usually membraneous, sometimes skin-like, typically without specialised spiculation except in 1 species. Ectosomal region contains heavy deposits of collagenous spongin, frequently with dense deposits of pigment granules. Structural megascleres oxeas of 1-2 sizes, together with acanthostyles, long or short subectosomal styles, and short slender ectosomal styles in 1 species. Microscleres absent. Remarks This genus is atypical of Raspailiidae. Its affinities with the family lie mainly in the presence of echinating acanthostyles and subectosomal styles, which appear to be remnants of an extra-axial skeleton. However, arguments for placing the genus in the Raspailiidae are strengthened by the existence of one species (E. nidulus) that has an ectosomal skeleton 1348 J. N. A. Hooper typical of other members of the family. This is interpreted as the retention of an ancestral character, whereas other species are considered here to be highly derived raspailiids that have lost their ectosomal specialisation. Echinodictyum is most similar to the nominal raspailiid genus Clathriodendron, the type species of which also has a predominantly reticulate skeleton, lacks any specialised ectosomal skeleton, and in which the fibres are echinated by microcionid-like acanthostyles (i.e with evenly distributed recurved spines, and with or without an- aspinose 'neck'). Echinodictyum is exclusively reticulate, with only vestigial development of the extra-axial skeleton (occurring as individual extra-axial styles), and choanosomal megascleres are always short, stout oxeas. By comparison, Clathriodendron has more obvious affinities with Raspailia, with at least some differentiation of the axial and extra-axial skeleton and with choanosomal styles or modified styles coring well developed spongin fibres. The comparison should also be made between Echinodictyum and the microcionid genus Echinochalina Thiele. Echinochalina is reticulate, with minimal spongin, with either oxeas, quasi-monactinal or quasi-diactinal megascleres coring fibres, and it has smooth (s.s.) echinating styles, whereas Echinodictyum invariably has acanthostyles. Echinochalina has been included with the Microcionidae (e.g. Hallmann 1912), and more recently with the Desmacididae (or Esperiopsidae) in the Poecilosclerida (e.g. van Soest 1984). Echinodictyum also shows affinities with the Anchinoidae (also known as Phorbasidae; e.g. Phorbas), which is also known to have oxeote structural megascleres. Morphological comparisons between the raspailiid genera Clathriodendron and Echinodictyum, and the microcionid genera Echinoclathria Carter and Echinochalina, provide evidence for the close relationship between these two groups of taxa. The higher systematic placement of Echinodictyurn may be debatable, but the genus itself appears to be a relatively homogeneous and cohesive taxon. The genus Echinodictyum is predominant in the Indo-Pacific. Thirty-five nominal species are presently assigned to it, including one new species described below, in addition to which there are many more nominal species that are obvious junior synonyms of well established taxa. Due to the paucity of morphological features in this group, and because many of those characters show a high level of intra-specific variability (or intergrade or overlap between one species and another), it appears that species must be differentiated mainly or even exclusively on the basis of growth form. There are usually only minute differences between some taxa in acanthostyle geometry, acanthostyle spination, megasclere size, and the presence or absence of long subectosomal styles. Furthermore, little is known about the inter-specific variability of characters within any group of sponges, and it is possible that only few of those nominal species are valid. For example, a study of the variability in spicule dimensions with latitudinal gradients for the widely distributed E. mesenterinum, described below (Table 25), shows that it may be unreliable to place too much emphasis on spicule size in identifying species, or at least in differentiating between morphologically similar species. One answer may lie in the use of chemotaxonomic and genetic tools to differentiate between sibling species, as demonstrated for the Microcionidae (Hooper et al. 1990), but, unfortunately, fresh material of all species is rarely available. A comparative study of the biochemical characteristics of some of the Australian Echinodictyum (i.e. proteins, carotenoid pigments and free amino acid profiles) has been described elsewhere (Hooper et al. 1992). A worldwide revision of all nominal type species is not presently possible, owing to the unavailability of several type specimens, and so only Australian species are evaluated here. Two of these species are excluded from the genus Echinodictyum, and referred here to Echinochalina (family Microcionidae). These are E. spongiosum Dendy (1896: 45; lectotype from Port Phillip Heads, Victoria, NMV G2409), and E. ridleyi Dendy (1896: 44, lectotype from Port Phillip, Victoria, NMV G2409). Similarly, E. gorgonoides Dendy (1916: 129; holotype from Kattiawar, Indian Ocean, BMNH 1920.12.9.38) is also referred here to the family Microcionidae, genus Clathria, having horny spongin fibres cored by plumose tracts of styles bearing anisoxeote modifications, abundant echinating acanthostyles, subectosomal auxiliary styles or anisostyles, and rare toxas. Thirteen species from Australian waters, including one new species, are included in Echinodictyum, and the spicule dimensions of these species are surnrnarised in Table 28. Australian Raspailiidae 1349 In addition to this comparatively diverse Australian fauna, there are another 17 species in the genus that are not represented in Australian territorial waters; there are possibly many more misidentified species that could also be placed in Echinodictyum. Known species are as follows: Dictyocylindrus aceratus Carter (1886b: 67; holotype from the Mergui Archipelago, Indian Ocean, BMNH 1889.6.9.5) (Fig. 83a-c), E. axinelloides Brondsted (1929: 225; type material presently unknown), E. cavernosum Thiele (1899: 15; holotype from Sulawesi, Celebes Sea, NMB 21) (Fig. 83d-f), E. clathratum Dendy (1905: 175; holotype from Sri Lanka, Indian Ocean, BMNH 1907.2.1.68) (Fig. 83g-i), E. dendroides Hechtel (1983: 68) (holotype USNM), E. jlabellatum Topsent (1906: 561; holotype from Djibouti, west Africa, MNHN LBIM DT 1874) (Fig. 83j-I), Acanthella fibelliforme Keller [1889: 394; holotype not known, from the Red Sea (Gvi 1965: 19), 'representative specimen' MNHN LBIM DCL 3381 (Fig. 83m), E. jousseaumi Topsent (1892b: 24; holotype from the Red Sea, MNHN LBIM DT 3478) (Fig. 83n-p), E. lacazei Topsent (18926: 25; schizotypes from St. Nazaire, Mediterranean, BMNH 1910.1.1.2374-5) (Fig. 84a), Dictyocylindrus laciniatus Carter (1879b: 296; holotype from Mauritius, BMNH 1882.6.3.11) (Fig. 84b-d), E. longistylum Thomas (1968: 246; holotype IM), Pandaros lugubris Duchassaing & Michelotti (1864: 89; type material presently unknown), E. macroxiphera G v i (1969: 966; holotype from Vema Seamount, Atlantic, MNHN LBIM DCL 1432) (Fig. 84e-g), E. marleyi Burton (1931: 348; holotype from Bats Cave Rock, South Africa, BMNH 1936.11.26.39) (Fig. 84h-j), Spongia nervosa Lamarck (1814: 450; MNHN holotype missing, locality unknown); sensu Ridley in Ridley & Duncan 1881, specimen from the coast of Arabia, 'John Murray' Expedition, BMNH 1936.3.4.110) (Fig. 84k-m), Pandaros pennatum Duchassaing & Michelotti (1864: 88; type material unknown), and Dictyocylindrus pykei Carter (1879b: 297; lectotype from Mauritius, BMNH 1882.4.6.9) (Fig. 84n-p). Echinodictyum arenosum Dendy (Fig. 85) Echinodictyum arenosum Dendy, 1896: 46.-Ayling et al. 1982: 100. Material Examined Holotype. NMV G2289 Port Phillip Heads, Vic., 38" 201S., 144" 401E., depth and date of collection unknown, coll. J.B. Wilson, dredge (schizotype BMNH 1902.10.18.366). Substrate and Depth Range Unknown. Geographical Distribution Port Phillip Heads, Vic. (Fig. 85e). Description Shape. Massive plate-like sponge with rounded margins and flat upper and lower surfaces. Colour. Sandy beige in ethanol. Oscula. Minute pores scattered over surface, up to 3 mm diameter. Texture and sulface characteristics. Texture arenaceous, firm, incompressible, and easily broken. Surface irregularly sculptured with raised meandering ridges and depressions, interconnected by reticulated dermal membrane (some still intact). Ectosome and subectosome. Ectosomal skeleton a thin layer of tangentially dispersed choanosomal megascleres aggregated into multispicular tracts, producing a close-meshed dermal reticulation. Tracts supported by plumose brushes of choanosomal megascleres arising from ascending primary fibres; peripheral tracts perpendicular to surface protrude only a small distance. Peripheral skeleton is a well developed tight reticulation of sand grains bound together by abundant spongin. Choanosome. Choanosomal skeleton irregularly reticulate, with sand grains incorporated into heavy spongin fibres. Each sand grain and each spongin fibre incorporates irregularly dispersed choanosomal oxeas; fibres and grains abundantly echinated by small acanthostyles. 1350 J. N. A. Hooper Fig. 83. a-c, Echinodictyum aceratus (Carter): a, holotype (BMNH 1889.6.9.5) (scale = 30 mm); b, section through peripheral skeleton (scale = 1 mm); c, echinating acanthostyle (scale = 50 pm). d-f, E. cavernosum Thiele: d, holotype (NMB 21) (scale = 30 mm); e, section through axial skeleton (scale = 1 mm); f , echinating acanthostyle (scale = 50 pm); g-i, E. clathratum Dendy: g, holotype (BMNH 1907.2.1.68) (scale = 30 mm); h, section through peripheral skeleton (scale = 1 mm); i, echinating acanthostyle (scale = 50 pm). j-I, E. flabellatum Topsent: j, holotype (MNHN LBIM DT3496) (scale = 30 mm); k, section through peripheral skeleton (scale = 1 mm); 1, echinating acanthostyle (scale = 50 pm). m, E. flabelliforme (Keller) (specimen MNHN LBIM DCL 338), megascleres (scale = 100 pm). n-p, E, jousseaumi Topsent: n, holotype (MNHN LBIM DT 3478) (scale = 30 mm); o, section through skeleton (scale = 1 mm); p, echinating acanthostyle (scale = 50 pm). Australian Raspailiidae 1351 Fig. 84. a , Echinodic~um lacazei Topsent (schizotype BMNH 1910.1.1.2374), section through peripheral skeleton (scale = 500 pm). b-d,E. laciniatus (Carter): b, holotype (BMNH 1882.6.3.11) (scale = 30 mm); c, section through peripheral skeleton (scale = 1 mm); d , echinating acanthostyle (scale = 50 pm). e-g, E. macroxiphera L6vi: e, holotype (MNHN LBIM DCL 1432) (scale = 30 mm); f , section through skeleton (scale = lmm); g, echinating acanthostyle (scale = 50 pm). h-j, E. marleyi Burton: h, holotype (BMNH 1932.3.26.39) (scale = 30 mm); i, section through peripheral skeleton (scale = 1 mm); j, echinating acanthostyle (scale = 50 pm). k-m, E. nervosa (Lamarck) sensu Ridley: k, specimen from the coast of Arabia (BMNH 1936.3.4.110) (scale = 30 mm); I, echinating acanthostyle (scale = 50 pm); m, section through skeleton (scale = 1 mm). n-p, E. pykei (Carter): n, lectotype (BMNH 1882.4.6.9) (scale = 30 mm); o, section through skeleton (scale = 1 mm); p, echinating acanthostyle (scale = 50 pm). J. N. A. Hooper Echinodictyum arenosum Dendy: a, choanosomal oxeas; b, echinating acanthostyles; c, section through skeleton; d, fibre characteristics (hatched areas are detritus); e, known Australian distibution; f, holotype (NMV G2289) (scale = 30 mm); g, echinating acanthostyles (scale = 50 pm). Fig. 85. Thin whispy multispicular tracts of oxeas are scattered throughout choanosome, occurring irregularly between sand grains; tracts lightly echinated by acanthostyles and fully enclosed in spongin fibres; fibres 35-70 pm diameter. Choanocyte chambers relatively large, oval, 110-260 p m in diameter, lined with fine layer of slightly granular spongin. Megascleres. Choanosomal megascleres, in fibres and on ectosome, are slender oxeas, straight or slightly curved at centre, with abruptly pointed hastate tips; slight but not definite tendency for these spicules to be asymmetrical, with one tip is very slightly swollen [I724185 -4)-198 x 343.9)-5 - 4 pm]. Subectosomal auxiliary megascleres absent. Ectosomal megascleres absent. Echinating acanthostyles small, thin, slightly curved centrally, entirely spined. Spines sharply pointed and erect, standing more-or-less perpendicular to shaft [52-(73 -2)-91 x 3(5 .5)-9 pm]. Microscleres absent. Australian Raspailiidae 1353 Remarks The original description of this species by Dendy (1896) was in error in describing megascleres of E. arenosum as tylostrongyla or tylota. Had this actually been the case, the species would have been referred to the microcionid genus Echinochalina. However, re-examination of the holotype showed that these spicules were clearly hastate oxeas. In addition, measurements of spicules given by Dendy also differ slightly from those actually observed, and the statement that echinating acanthostyles were scarce is incorrect: these spicules were found to be very abundant. Echinodictyum arenosum is atypical of the genus, and of Raspailiidae in general, in being predominantly arenaceous, much like the desmacidid species of Psammascus and Desmapsamma that are so abundant in southern Australian waters. The geometry of the two types of spicules gives the only clue as to the affinities of this species. Scanning electron microscopy of the skeleton and spicules of this species was not possible due to the highly arenaceous consistency, rendering sections too brittle to retain their form on stubbs. Echinodictyum asperum Ridley & Dendy (Figs 86, 87, 110c; Table 20) Echinodictyum asperum Ridley & Dendy, 1886: 477.-Ridley & Dendy, 1887: 165, pl. 32, fig. 2; Whitelegge, 1897: 328-329; Topsent, 1897: 446, pl. 20, fig. 23; Burton & Rao, 1932: 348; L6vi, 1961: 524, fig. 15; Desqueyroux-Faundez, 1981: 757, table 11; Hooper, 1984: 55. Material Examined Holotype. BMNH 1887.5.2.17: Papiete Harbour, Tahiti, 40 m depth, HMS 'Challenger'. Other material (all material collected by the author using SCUBA,unless otherwise indicated). Ellice Is: AM G1662: Lagoon of Funafuti Atoll, 8" 301S., 179' 05/E., 26 m depth, 1897, coll. C. Hedley (dredge). Indonesia: MNHN LBIM DT 1760: Ambon, Indonesia, no other details known. Philippines: MNHN LBIM DCL 706: Lucap Bay, Lingayen Gulf, Philippines, depth and date of collection unknown, collector unknown. Papua New Guinea: NCI Q66C 4381-C: 'Rasch Pass', Madang, 5' 09.47'S., 145' 49.g1E., 25 m depth, 5.xi.1990, coll. NCI. Darwin Region, N.T.: NTM 2971: EPMFR, 12" 24-5'S., 130" 48-O'E., 10-12 m depth, 13.ix.1982 (stn EP9). NTM 22061: same locality, 12' 25 .OIS., 130' 48.4/E., 6-10 m depth, 10.v.1984 (stn EP14). NTM 22641: same locality, 9-12 m depth, 3.iv.1986 (stn EP28). NTM 2445: Lee Point region, Shoal Bay, 12" 17'S., 13Oe54/E., 30 m depth, coll. N.T. Fisheries (stn D5, shot 8-Don. lA, trawl). NTM 21971, 1982: W. side of Weed Reef, 12" 29.2' S., 130' 47.1 E., 12 m depth, 11.v.1984 (stn WR2, SCUBA).NTM 2870: vicinity of pearl oyster beds, Channel I., Middle Arm, 12" 32.7/S., 130" 52.5'E., 12-13 m depth, 20.viii.1982, coll. P.N. Alderslade (stn C13, SCUBA).Wessel Is, N.T.: NTM 23908: N. of Cape Wilberforce, Melville Bay, Gove Peninsula, 11" 52.6/S., 136" 33.3'E., 25 m depth, ll.xi.1990 (stn WI-1). NCI Q66C 4721-W: NE. tip of Wigram I., English Company Is, 11" 44.4'S., 136' 37.S1E., 16 m depth, 12.xi.1990, coll. NCI. Northwest Shelf, W.A.: NTM 21210: N. of Bedout I., 19' 29-4'S., 118" 52. l1E., 39 m depth, 26.iv.1983 (CSIRO RV 'Soela' S02/83, stn B8-NWS9, beam trawl). NTM 21853: W. of Port Hedland, 19" 28.giS., 118" 52.3'E., 38 m depth, 31.viii.1983, coll. T. Ward (CSIRO RV 'Soela' S04/83, stn 127-NWS27, trawl). NMV unregistered (J.H. 20) (fragment NTM 21486): Bonaparte Archipelago, 14' 06-08'S., 124" 29-31 E., 93-95 m depth, 26.iii.1981, coll. C.C. Lu (RV 'Hai Kung', stn 70032604Don. 60, trawl). Exmouth Gulf Region, W.A.: NTM 23398: 2 km W. of Hermite I., Monte Bello I. Small I. Group, 20' 27.1 IS., 115' 34 2'E., 6 m depth, 29.viii.1988, coll. D. Low Choy & NCI (sm NWS74pLC47). NCI Q66C-1294-W. (duplicate NTM 23466): 500 m NW. of jetty, Leannonth, Exmouth Gulf, 22' 12 - 8 S., 114" 06.1 E., 2 m depth, 16.viii.1988, coll. NCI (stn NWS88). Substrate and Depth Range Restricted to hard substrates (rock and dead coral) on subtidal and shallow coastal reefs, 6-94 m depth. Geographical Distribution Widely dispersed across the Indo-Pacific, from Papiete, Tahiti (Ridley & Dendy 1886, 1887), Ellice I. (Whitelegge 1897), Ambon, Indonesia (Topsent 1897; Desqueyroux-Faundez J. N. A. Hooper Fig. 86. Echinodictyum asperum Ridley & Dendy (specimen NTM 20445): a, choanosornal axial oxeas; b, subectosomal extra-axial style; c, echinating acanthostyle; d, section through peripheral skeleton; e, known Australian distribution. 1981), Lucap Bay, Philippines (Lkvi 1961), North-west Australia (Fig. 86e) (Hooper 1984, and present study), to Tuticorin, India (Burton & Rao 1932). The species has not yet been recorded from eastern Australia. Description Shape. Honeycomb-like, clathrous reticulation of flattened or cylindrical branches, forming spherical, elongate-lobate, or distinctly arborescent and digitate mass, up to 35 cm maximum height, 4 - 5 cm maximum branch diameter, with or without a basal stalk, and with a slightly expanded and flattened basal attachment. Branches 2-8 mm diameter, forming a loose and irregular reticulation but producing relatively close meshes. Colour. Live coloration of shallow-water material light brown to grey-brown (Munsell 7.5YR 714); deeper water material light brown or beige (2 - 5Y 8 / 4 4 ) (Fig. 110c). Fixed specimens retain live coloration. Oscula. Oscula not differentiated from cavernous openings produced by branch anastomoses. Texture and surface characteristics. Texture firm, compressible, difficult to tear. Surface optically uneven but smooth, microscopically hispid, prominently cavernous, with peripheral fibres producing microscopic dermal projections. Ectosome and subectosome. Ectosome lacks special spiculation; dermal skeleton consists of an interconnected network of small fibre bundles ('trabeculae' of authors) up to 2 mm diameter, producing secondary reticulation (similar to gross morphological structure, but on a microscopic scale). No membraneous dermal skeleton observed. Reticulation of peripheral fibre bundles composed of individual multispicular spongin fibres, 150-220 pm diameter, fully cored by oxeas and heavily echinated by acanthostyles. Minute surface reticulation produced by protruding acanthostyles. Subectosomal styles (remnants of an extra-axial skeleton) rare and confined to choanosomal mesohyl matrix. Australian Raspailiidae Fig. 87. Echinodictyum asperurn Ridley & Dendy: a, holotype (BMNH 1887.5.2.17) (scale = 30 mm); b, bulbous specimen (NTM 20870) (scale = 30 mm); c, lobate specimen (NTM 23398) (scale = 30 mm); d, SEM of skeletal structure; e, axial skeleton; f, echinating acanthostyles; g, spine morphology. Choanosome. Choanosomal skeletal architecture lacks axial condensation or any axial and extra-axial differentiation. Choanosomal fibres irregularly reticulate, composed of individual fibres (cf. periphery where they become bundled more closely together), forming ovoid, cavernous meshes up to 450 p m diameter, containing minimal quantities of loose spongin and very few interfibril spicules. Megascleres (refer to Table 20 for dimensions). Choanosomal axial (coring) oxeas slightly curved at centre, rarely straight, relatively short and thin, with slightly fusiform tips which may be sharply pointed or mamillifom. Subectosomal extra-axial (auxiliary) styles rare, only found between fibres, usually straight, fusiform, with slightly subtylote bases. 1356 J. N. A. Hooper Table 20. Comparisons in spicule measurements between specimens of Echinodictyum asperum Ridley & Dendy Measurements are given in micrometres, presented as ranges (and means) of lengthxwidth. For material described in this work these measurements were obtained by sampling 25 spicules per specimen for each spicule category Material Choanosomal oxeas Subectosomal styles BMNH 1887. 5.2.17 (Tahiti) 162-565 X3-13 (326.0~8.2) Holotype 339-548 x 3-8 (425.3~4.5) AM G1662 (Ellice I.) MNHN LBIM DT 1760 (Ambon) MNHN LBIM DCL 706 (Philippines) NTM various (North Australia) (n = 12) Specimens 193-618 X3-13 (399.5~7.0) 199-515 x 3-8 (348-4~5.8) 244-466 x 3-7 (373~6~4.5) 157474 ~2-6.5 (346.2~3.7) Ectosomal megascleres Echinating acanthostyles Absent 99-147 x 5-8 (124.8~6.4) Absent Absent Absent Absent 97-112 x 4-7 (105.2~6-0) 104-1 21 x 6-12 (114.3~8-7) 95-141 X 7-1 1 (118.0~9-3) 93-135 x6-11 (113.3~7.4) Ectosomal megascleres absent. Echinating acanthostyles evenly tapering club-shaped, with slightly rounded or fusiform points and subtylote bases, and with relatively even spination except for a slightly aspinose area proximal to base. Spines slender, sharply pointed and recuwed. Microscleres absent. Remarks This species closely resembles figures and descriptions of earlier authors, and the identity of northern Australian material has been confirmed by re-examination of type material from Tahiti, indicating that the species is relatively conservative morphologically across a wide Indo-Pacific distribution. A number of Echinoclathria-like (honeycombed reticulate) Echinodictyum species, such as E. clathratum (Fig. 83g-i), E. cavernosum (Fig. 83d-f), E. costiferum (Fig. 97), E. lacunosum (Fig. 99) and others, have been recorded from the Indo-Pacific. The confirmation of a unique identity, and the distinguishing features of all those taxa are poorly understood and unlikely to be provided by traditional systematic techniques. In relation to other raspailiid genera, species of Echinodictyum are very simple, with minimal variability in spicule diversity or spicule geometry. Similarly, in many species the range of spicule sizes often overlap, illustrating the difficulty in amving at unequivocal species identifications on the basis of morphological features. Chemotaxonomic tools may prove very useful for differentiating between species of this genus. Nevertheless, the combination of growth form and the specific dimensions of spicules serve to define E. asperum and to distinguish it from other species (Fig. 87). Echinodictyum austrinus, sp. nov. (Figs 88, 89, 11Od) Material Examined Holotype. SAM TS4022 (fragment NTM 21602): SW. of bay, 8 krn SW. of Althorpe I., off S. Yorke Peninsula, Investigator Strait, SA, 35' 05'S., 137"45'E.. depth unknown, 13.iii.1986, coll. H. Chemoff. Australian Raspailiidae Fig. 88. Echinodictyum austrinus, sp. nov. (holotype SAM TS4022): a, choanosomal axial oxeas; b, subectosomal extra-axial style; c, echinating acanthostyle; d, section through peripheral skeleton; e, known Australian distribution. Paraiype. NCI Q66C-2627-V (fragment NTM 23755): Outer edge of reef, W. of Mullalo-Hilaxy's Boat Harbour, Perth region, W.A., 31" 49.9 S., 115"42.25 E., 7 m depth, 12.iii.1989, coll. NCI. Substrate and Depth Range Rock reef, from 7 m depth. Geographical Distribution Yorke Peninsula, S.A. and Perth region, W.A. (Fig. 88e). Description Shape. Thick vasiform sponge, 184 mm long, 130 mm maximum width at rim of vase, with thick walls, up to 18 mm thick at rim, and short stalk, 48 mm long, 22 mm maximum diameter, with enlarged basal attachment. Colour. Live coloration dark red (Munsell 2 - 5R 5/12) (Fig. 1104, dark chocolate-brown in ethanol. Oscula. Large oscula (0.8-1 5 rnrn diameter) situated around rim of cup, forming close-set bundles. Small inhalant pores scattered over both interior and exterior surfaces of vase. Texture and surface characteristics. Texture of vase firm, compressible, slightly flexible; stalk more rigid. Surface optically even on both interior and exterior walls of vase; microscopically rugose, with small projecting peripheral fibres. 1358 3. N. A. Hooper Ectosome and subectosome. Ectosome lacks specialised spiculation, but regularly-spaced bundles of spicules, 230-370 pm apart, protrude through surface for 100-220 pm; bundles composed exclusively of acanthostyles, standing perpendicular to peripheral fibres, and spacing of surface brushes corresponds exactly with distribution of peripheral fibres below. Between surface brushes is stretched a skin-like membrane, with spicule or fibrous components. Extra-axial skeleton vestigial, consisting of rare subectosomal styles dispersed between fibre meshes. Choanosome. Peripheral fibre skeleton regularly reticulate as described above (Fig. 89b), whereas choanosomal skeleton much more irregular, and fibres may become slightly sinuous in places (Fig. 886). Fibres (75-105 pm diameter) only lightly invested with spongin type A, without marked differentiation of primary or secondary elements, cored for 70-95% of their diameter with choanosomal oxeas. Fibre meshes near surface relatively small, compact and oval or elongate, 130-480 pm diameter, whereas closer to axis fibres much larger and irregular in shape, 530-1150 pm maximum diameter. Fibres in choanosomal skeleton only relatively lightly echinated by acanthostyles; exterior fibres in peripheral skeleton have dense perpendicular bundles of acanthostyles. Mesohyl matrix relatively heavily invested with light brown type B spongin, with abundant megascleres scattered within meshes, and oval or elongate choanocyte chambers 80-155 pm diameter. Megascleres. Choanosomal oxeas slender, short, with slight or moderate central curvature, tapering to hastate tips, and invariably with stepped points [holotype: 112-(170 2)-211 x 3(6 5)-9 pm; paratype: 1494189 7)-248 x 447.5)-10 pm]. Subectosomal extra-axial styles rare, slender, straight or slightly curved towards base, with evenly rounded bases and abruptly pointed, hastate tips [holotype: 2194338- 5)463 x 2 4 4 - 3)-6 pm; paratype: 2484288 - 0)-386 x 2 4 3 9)-5 pml. Ectosomal megascleres absent. Echinating acanthostyles short, slender, straight, with subtylote bases and slightly swollen tips. Spines fairly evenly distributed over spicules, although they may be more dense on bases and points. Spines conical, robust, slightly recurved, with sharply pointed tips Bolotype: 61470 2)-82 x 4 4 6 - 6)-8 pm; paratype: 70-(82.8)-92 x 4-(6.2)-8 pm]. Microscleres absent. - - - Remarks This species is unusual amongst Echinodictyum in being red (Fig. 1106) (not purple or black like most species), and having a sculptured ectosomal skeleton composed of perpendicular bundles of acanthostyles. This latter feature is analogous to the paratangential ectosomal features of the haplosclerid sponge Cribrochalina Schmidt (Niphatidae), although in most other respects the two species are unrelated. The present species also shows some similarities in its ectosomal features with E. clathrioides, although perpendicular acanthostyles on peripheral fibres of that species do not form brushes but merely produce a dense erect palisade of spicules. Echinodictyum clathrioides, which is also a vasifonn sponge, is thin-walled with a very smooth surface, and spines on its echinating acanthostyles are spatulate and serrated, whereas spines of E. austrinus are sharply pointed, like most other species. Echinating acanthostyles of this species are also similar in geometry to those of E. carlinoides, E. clathrioides, E. costiferum, and E. mesenterinum in having slightly swollen and more-heavily spined points. The species name refers to its southern Australian distribution. Echinodictyum cancellaturn (Lamarck) (Figs 90, 91, 110e; Table 21) Spongia cancellata Lamarck, 1814: 456. Echinodictytum cancellaturn.-Ridley 1884: 457; Hentschel 1912: 370; Topsent 1933: 7. Echinodictyurn pulchrum Brondsted, 1934: 18. Australian Raspailiidae Fig. 89. Echinodictyum austrinus, sp. nov.: a, holotype (SAM TS4022) (scale = 30 mm); b, transverse section of the surface, showing the three dimensional ectosomal structure (scale = 1 mm); C, SEM of skeletal structure; d, SEM of fibre characteristics; e, SEM of echinating acanthostyles; f , spine morphology. Material Examined Holotype. MNHN: LBIM DT 619: unspecified locality in 'Australian Seas', Peron and Lesueur collection (schizotype BMNH 1953.11.9.62). Other material (all material collected by the author using SCUBA, unless otherwise indicated). Indonesia: SMF 1700, 1701 (fragment MNHN LBIM DCL 2319): Pulu Bambu Bay, Aru I., 6'S, 134" 301E., 10 m depth, 3.iv.1908, coll. H. Merton (stn 11, dredge). Great Barrier Reef, Qld: BMNH 1896.2.28.8: Warrior Reef, Torres Strait, 9" 31's.. 143'37'E., beach debris, coll. R.W. Coppinger, HMS 'Alert'. QM GL800: E of Murdock I., Howick Group, 14' 36/ S., 145" 03/E., 14 m depth, 18.ix.1979, coll. Queensland Fisheries Service (stn AQS IB/19-2, trawl). Gulf of Carpentaria, Qld: NTM 23991: SW. of Ken Reef, 12' 12 8IS., 141' 05.8-08.5/E., 42 m depth, 4.xii.1990, coll. J.N.A. Hooper, U.S.S.R. RV 'Akademik Oparin', sled (stn JH-90-044). Darwin Region, N.T.: - 1360 J. N. A. Hooper NTM 2495, 502,503, 514: Fannie Bay Beach, 12" 25IS., 130' 501E., stom debris, 9.ii.1982 (stn EP7, by hand). NTM 2904: EPMFR, 12" 25. O'S., 130' 48 -4'E.. 10 m depth, 31.viii.1982 (stn EP8). NTM 2936: same locality, 12' 24.5'S., 130" 48. OIE., 10-12 m depth, 13.ix.1982 (stn EP9). NTM 21059: same locality, 12" 24-7'E., 130" 48'E., 13 m depth, 9.xi.1982 (stn EP11). NTM 21095: same locality, 12" 25 -OIS., 130" 48 -4/E., 6-7 m depth, 22.xii.1982 (stn EP12). NTM 22253: same locality, 12" 24.5 S., 130" 48. OIE., 10 m depth, 12.iv.1985, coll. J.R. Hanley and C. Hood (stn EP22). NTM 22401: same locality, 8 m depth, 29.vii.1985 (stn EP23). NTM 22683: same locality, 9-12 m depth, 3.iv.1986 (stn EP28). NTM 23202: same locality, 7 m depth, 18.ix.1987, coll. N. Smit (stn EP33). NCI Q66C-0582-W. (fragment NTM 23752): same locality, 6 m depth, 17.viii.1987, coll. NCI. NTM 2781: Angler's Reef, Lee Point, 12' 18.7IS., 130' 52 .OIE., 5-6 m depth, 19.vii.1982, coll. P.N. Alderslade (stn LP3). NTM 21946, 1951: Stephen's Rock, Weed Reef, 12" 29-2IS., 13O047-llE., 12m depth, 27.iv.1984 (stn WR1, SCUBA).NTM 22019: W. side of Weed Reef, 12" 29.2l S., 130" 47. l1E., 5 m depth, 11.v.1984 (stn WR2). NTM 22166: 'Bommies', Weed Reef, 12" 29 2'S., 130" 37.6/E., 8-10 m depth, 5.x.1984 (stn WR5). NTM 22658, 2659: Nightcliff Reef, 12" 22.4IS., 130"49.6IE., 8 m depth, 5.vi.1986, coll. L. Vail (sm NR1, SCUBA). NTM 2838: Channel I., Middle Arm, 12' 33 -8IS., 130" 51 -4IE., 20 m depth, 18.vii.1982, coll. S. Chidgey (Channel I. EIS (FN A19), Don. 24, hooker). NTM 2867: vicinity of pearl oyster lease, Channel I., Middle Arm, 12' 32.7IS., 130' 52.5/E., 12-13 m depth, 20.viii.1982, coll. P.N. Alderslade (stn C13, SCUBA).NTM 21079: Fish Reef, Bynoe Harbour, 12' 26 2IS., 130" 26.2IE., 10-12m depth, 24.xi.1982 (stn FR1, SCUBA).Cobourg Peninsula, N.T.: NTM 2154: Sandy I. No. 2, 11' O5.6'S., 132' lglE., 8-9 m depth, 22.x.1981, coll. P.N. Alderslade (stn CP29, SCUBA). Wessel Is, N.T.: NCI Q66C 4829-P: off Low Point, Marchinbar I., SE. of Cape Wessel, 11' 01. a's., 136" 46.2'E., 20 m depth, 17.xi.1990, coll. NCI. Northwest Shelf, W.A.: WAM 154-82 (fragment NTM 21707): Off Port Hedland, 20" 13'S., 118'28/E., 18 m depth, 5.viii.1982, coll. J. Fromont (CSIRO RV 'Soela' S04182, stn 13-Don. 169, trawl). WAM 141-82 (fragment NTM 21723): same locality, 20' 12/S., 118" 2g1E., 16 m depth, 25.vii.1982 (stn 01-Don. 167). WAM 139-82 (fragment NTM 21718): same locality, 20' 12IS., 118" 25 'E., 14 m depth, 3.viii.1982 (stn 09-Don. 168). NTM 21158: W. of Port Hedland, 19" 29 6 / S., 118' 52.2'E., 38 m depth, 26.iv.1983 (CSIRO RV 'Soela' S02183, stn B8-NWS6, epibenthic sled). NTM 21175: same locality, 19' 30. glS., 118' 48 -7'E., 40 m depth, 26.iv.1983 (stn B7-NWS8, beam trawl). NTM 21189: same locality, 19' 29-4'E., 118" 52. llE., 39 m depth, 26.iv.1983 (stn B8-NWS9). NTM 21233: same locality, 19' 28.5'S., 118'55.3'E., 40m depth, 26.iv.1983 (stn B9-NWS10). NTM 21277: same locality, 19" 05.3/S., 118" 53.S1E., 82m depth, 27.iv.1983 (stn B4-NWS13). NTM 21477: same locality, 20'001S., 116" 30.5'E., 54 m depth, 4.xii.1982, coll. T. Ward (CSIRO RV 'Soela' S06182, stn 119-Don. 66, Frank & Bryce trawl). NTM 21879: same locality, 18' 05 -7/S., 118" 57.4IE., 83 m depth, l.ix.1983, coll. T. Ward (CSIRO RV 'Soela' S04/83, stn 133-NWS28, beam trawl). NTM 21891: same locality, 19" 02.2'S., 118"04.1 IE., 84 m depth, l.ix.1983 (stn 136-NWS29). NTM 21005: NW. of Dampier Archipelago, 19" 54'S., 115' 39'E., 108 m depth, 18.ix.1982, coll. L. Bullard (stn Don. 25, Taiwanese Pair-Trawl). NTM 21017: same locality, 20" lo's., 117" 3g1E., 35 m depth, 20.x.1982, coll. J. Blake (CSIRO RV 'Soela', stn 48-Don. 28, trawl). NTM 21035: same locality, 20' 201S., 117" 28'E., 26 m depth, 22.x.1982 (stn 64-Don. 35). NTM 2643: W. of 80 Mile Beach, 19" 33-5/S., 119'05.7'E., 35 m depth, 4.v.1982 (CSIRO RV 'Sprightly' SP4f82, stn 78-Don. 15, dredge). NTM 22461: NW. of Amphinome Shoals, 19" 12/S., 118" 36.5'E., 76-80 m depth, l.vi.1985, coll. B.C. Russell (sm 8515 haul 6-NWS36, Taiwanese Pair-Trawl). NTM 22477: same locality, 19' 07/S., 118" 2g1E., 82-86 m depth, 2.vi.1985 (stn 8518 haul 4 NWS39). NTM 23046: N. of Amphinome Shoals, 19" 19 7-23.3 IS., 119" 08.8-12.2/E., 50 m depth, 19.vii.1987 (U.S.S.R. RV 'Akademik Oparin', stn NWS55, beam trawl). NTM 2696: W. of Buccaneer Archipelago, 16" 20. 0' S., 120' 10 1 E., 35 m depth, 28.iv.1982 (CSIRO RV 'Sprightly' SP4182, stn 39-Don.19, dredge). NTM 22353, 2359: NW. of Lacepede I., 16" 34-OIS., 121" 27-llE., 4 U 6 m depth, 17.iv.1985, coll. B.C. Russell (stn 85-2-NWS35, Taiwanese Pair-Trawl). NTM 22327, 2277: same locality, 16' 29-33 S., 121" 27-29 E., 38-40 m depth, 17.iv.1985 (stn 85-4-NWS34). NTM 2710: N. of Adele I., Collier Bay, 15' 58.3'S., 122" 39.7'E., 59 m depth, 21.iv.1982 (CSIRO RV 'Sprightly' SP4/82, stn 40-Don. 20). NTM 22974: Due W. of Camarvon, 24" 55.6-56.5' S., 112" 50.8-53.5'E., 80-85 m depth, 19.vii.1987 (U.S.S.R. RV 'Akademik Oparin', stn SB6, beam trawl). - Substrate and Depth Range Usually associated with soft substrates, attached to hard objects embedded in the sediment, or less frequently found on rock and dead coral reefs; exclusively subtidal, not found above 5 m and extending to 108 m depth. Australian Raspailiidae 1361 Geographical Distribution Southern Indonesia (Hentschel 1912) to north-east Australia (Torres Strait, Ridley 1884; present study) and north-west Australia (south to Shark BayICamarvon regions; present study). The 'Alert' material from Torres Strait in the BMNI-I,and the more recent QM material from the Great Barrier Reef are the only records of the species in the Pacific Ocean (Fig. 90e). Echinodictyum cancellatum (Lamarck) (specimen NTM 21079): a, choanosomal axial oxeas; b, subectosomal extra-axial style; c, echinating acanthostyle; d, section through peripheral skeleton (scale = 1 mm); e, known Australian distribution. Fig. 90. Description Shape. Erect, arborescent, reticulate branching sponges, ranging from 10-58cm high, up to 36cm wide, and 1.5-3.5 cm lateral thickness; with long or short cylindrical woody stalk arising from an elongated peduncle, with or without rhizomous projections; stem usually with several longitudinal major branches, interconnected by horizontal side branches forming more-or-less regular rectangular reticulation, usually planar or occasionally multiplanar. Branch anastomoses produce rectangular or circular meshes 5-10 rnm diameter, with branches varying from 3 to 10 mm wide. Branches slightly flattened laterally. Colour. Live coloration varies only slightly, ranging from evenly light mauve-blue (Munsell 5RP 4/8), blue-purple mottled with brown (2-5B 4/6 and 5RP 4/6) to simply light brown (10R 518) (Fig. 110e); pigmentation stable in ethanol. Oscula. Not seen. Texture and surface characteristics. Surface optically even, but microscopically convoluted into ridges surmounted by small conules. Surface projections and ridges made hispid by protruding acanthostyles from peripheral skeleton, extending only short distances from surface. Corrugations especially prominent on tips of branches, produced by relatively open reticulation of peripheral spongin fibres. Texture firm, almost harsh, stiff and barely flexible. J. N. A. Hooper Fig. 91. Echinodictyurn cancellaturn (Lamarck): a , holotype (MNHN LBIM DT 619) (scale = 30 mm); b, specimen (NTM 22359) (scale = 30 mm); c, SEM of skeletal structure; d, fibre characteristics; e, echinating acanthostyle; f , spine morphology. Ectosome and subectosome. Ectosomal region lacks specialised spiculation, but with aspiculose membraneous spongin 'skin' stretched over peripheral reticulate spongin fibres; membraneous covering disintegrates upon preservation. Subectosomal styles rare, vestigial (very thin and small), and interpreted as remnants of an extra-axial skeleton. Choanosome. Choanosomal skeletal architecture simply reticulate, not axially condensed, and with undifferentiated axial and extra-axial skeletons. Choanosomal spongin fibres form irregular meshes, fully cored by multispicular tracts of oxeas and echinated by acanthostyles. Fibres lightly invested with spongin, up to 250 p m diameter. Fibre anastomoses form circular or ovoid meshes, up to 800 p m diameter, with very light mesohyl spongin dispersed between fibres, confined mostly to fibre junctions, and usually containing small purple pigment granules, 2 4 p m diameter. Small oxeas and subectosomal extra-axial styles dispersed between fibre meshes. Australian Raspailiidae 1363 Megascleres (refer to Table 21 for dimensions). Choanosomal oxeas, coring fibres and occurring between fibres, straight or slightly curved centrally, symmetrical, almost fusiform, with mamillifom or sharply pointed tips. Subectosomal extra-axial styles uncommon, small, thin, straight or only slightly curved, with subtylote bases and fusiform, often hair-like points. Ectosomal megascleres absent. Echinating acanthostyles large, relatively thick, with rounded tips and slightly subtylote bases; spines dispersed relatively evenly over spicules; spines robust, strongly recurved, with slightly splayed spatulate points. Microscleres absent. Table 21. Comparisons in spicule measurements between specimens of Echinodictyum cancellotum (Lamarck) Measurements are given in micrometres, presented as ranges (and means) of lengthxwidth. For material described in this work these measurements were obtained by sampling 25 spicules per specimen for each spicule category Material Choanosomal oxeas Subectosomal styles MNHN LBIM DT 619 ('Australia') 147-402 x 6-22 (259.4~14.3) Gotype 65-105 x 2-5 (85-7~3.5) SMF 1700, 1701 (Indonesia) BMNH and QM various (Qld) (n = 2) NTM various (N.T.) (n = 23) 162-366 X5-16 (237.2~10.0) 124-332 x4-17 (212.4~12.1) 115445 x 2-19 (221.9~8.7) 138-378 x 3-20 (217.4~9.9) Specimens 55-112 x2 4 (80.8~3.3) 53-118 x 3-6 (83.4~4.0) 55-170 x24.5 (106.4~3.0) 58-198 x 2-5 (111.9~3.0) NTM various (W.A.) (n = 22) Ectosomal megascleres Echinating acanthostyles Absent 108-188 Absent Absent Absent Absent 92-146 X8-13 (116.4~10.4) 94-133 x8-12 ( 1 1 5 . 4 ~10.1) 108-193 x 6-13 (144.6~9.9) 118-198 x8-13 (149.7~10.5) Remarks This species is well characterised by its external morphology, which superficially resembles that of Clathria coppingeri Ridley (Microcionidae). However, like most other Echinodictyum species, there are few other features that distinguish it. The reticulate branching growth form of E. cancellatum should also be compared with E. nervosum from the western Indian Ocean (Larnarck 1814; Ridley, in Ridley & Duncan 1881; Burton & Rao 1932) (Fig. 84k). Although E. nervosum is poorly known since the type material is missing from the MNHN Lamarck collection, the species is taken in the sense of Ridley, in Ridley & Duncan (1881). It differs from E. cancellatum in having more robust branches with fewer anastomoses, smaller and more irregularly spined acanthostyles, and larger oxeas. Echinodictyum carlinoides (Lamarck) (Fig. 92; Table 22) Spongia carlinoides Lamarck, 1814: 449.-Lamarck, 1816: 377. Echinodictyum glomeratum Ridley, 1884: 456-7.-Hentschel, 1912: 369; Topsent, 1932: 111-12, pl. 6, fig. 4. Material Examined Holotype. MNHN LBIM DT 636: Unknown locality, suspected to be Australian, Turgot collection. Other material. Holotype of E. glomeratum, BMNH 1881.10.21.283: Thursday I., Torres Strait, 1364 J. N. A. Hooper Qld, 10" 35'S., 142" 14'E., 8-10 rn depth, 14.vi.1881, coll. R.W. Coppinger (HMS 'Alert', bottom sand). Indonesia: SMF 1.523 (fragment MNHN LBIM DCL 2302): Meriri and Leer, Am I., Arafura Sea, Indonesia, 6-10 m depth, 31.iii.1908, coll. H. Merton (stn 8, dredge). SMF 1528: E side of Aru I., Arafura Sea, Indonesia, unknown depth, 01.ii.1908, coll. H. Merton (dredge). SMF 1668 (fragment MNHN LBIM DCL 227.5): Straits of Dobo, Aru I., Arafura Sea, Indonesia, 40 rn depth, 20.ii.1908, coll. H. Merton (stn 4, dredge). Substrate and Depth Range Sand and limestone substrates, 6-40 m depth. Geographical Distribution Restricted to northern Australian-southem Indonesian waters (Arafura Sea) (Fig. 92d). Description Shape. Globular, arborescent, bushy growth form, with short stem, enlarged basal attachment, tapering digitate processes at apex of sponge, resembling staghorn-coral in some material (45-80 mm high, 25-60 rnm wide, with branches up to 5 mm long, 5-12 mm diameter). Colour. Live coloration unknown, but in preserved and dry state it ranges from even yellow-brown to grey-brown with purple hue. Oscula. Oscula and inhalant pores not easily differentiated from open-reticulate, pock-marked surface. In dried and preserved specimens, pores 1-3 mm diameter, distributed evenly over surface. Texture and suiface characteristics. Surface uneven, with small lobate and/or flabellate branches, which in turn bear small prominent, sharply pointed conules. Surface abundantly pocked with small openings, appearing insubstantial. Texture very firm barely flexible in dry and preserved states. Ectosome and subectosome. Ectosome membraneous, lacking any specialised skeleton; some tangential spicules scattered over surface, and some primary acending spicule tracts poke through surface forming pointed conules, but most of ectosomal skeleton is even and consists of relatively small spongin fibres, 40-125 p m diameter, forming oval meshes, 150-530 p m diameter. Echinating acanthostyles on peripheral fibres as equally abundant as echinating spicules on choanosomal tracts. Subectosomal region represented by remnants of extra-axial skeleton (scattered long styles). Choanosome. Choanosomal skeleton regularly or irregularly reticulate, without any differentiation of axial and extra-axial components. Choanosomal fibres contain very little spongin, fully cored with oxeas, moderately heavily echinated by acanthostyles. Fibre diameter 190-270 pm, and fibre anastomoses produce oval to eliptical meshes 170-640 p m diameter. Choanocyte chambers oval, 60-110 p m diameter, bounded by very light collagenous mesohyl. Few spicules occur between fibres. Megascleres (refer to Table 22 for dimensions). Choanosomal oxeas vary considerably in size and curvature. Oxeas typically long, thin, slightly curved at centre, and each end curves in a different direction, so that more severely curved examples are almost sigrnoid. Spicules taper to sharp fusiform points. Extra-axial styles relatively long, thin, usually slightly curved, tapering to sharp points, and with evenly rounded bases. Ectosomal megascleres absent. Echinating acanthostyles short, thin or thick, with slightly swollen subtylote bases, and rounded and/or slightly swollen points. Both base and pointed ends prominently spined, whereas there is often a spine-free area just below basal swelling. Spines spatulate, disk-shaped, lacking serrated edges or secondary spines. Microscleres absent. Australian Raspailiidae Fig. 92. Echinodictyum carlinoides (Lamarck): a, choanosomal axial oxeas; b, subectosomal extra-axial style; c, echinating acanthostyles; d, known Australian distribution; e, holotype (MNHN LBIM DT 636) (scale = 30 mm); f , holotype of E. glomeratum Ridley (BMNH 1881.10.21.283) (scale = 30 mm); g , photomicrograph of echinating acanthostyle (scale = 50 pm); h, photomicrograph of skeletal structure (scale = 1 mm). Remarks The species is distinctive in its globular arborescent growth form, with staghom-like digits at the apex of branches, in having relatively long, straight or setaceous extra-axial styles, some flexuous or even sigmoid axial oxeas, and the presence of swollen bulbous tips on acanthostyles. In the latter feature E. carlinoides resembles E. austrinus (Figs 88c, 89e), E. clathrioides (Figs 93c-k, 94e) and, to a lesser extent, E. costiferum (Fig. 97c,g) and E. mesenterinum (Figs 100c, 102d-e). Spicule dimensions observed in material examined were found to differ slightly from those measurements actually published, particularly the smaller sizes of extra-axial styles (recorded as 2 mm by Ridley 1884, for example). However, most of the styles observed were broken, and it is possible that they attain a larger size than observed here. Echinodictyum carlinoides is a poorly known Indo-Australian species, known only from material redescribed above. 1366 J. N. A. Hooper Table 22. Comparisons in spicule measurements between specimens of Echinodictyum carlinoides (Lamarck) Measurements are given in micrometres, presented as ranges (and means) of lengthxwidth. For material described in this work these measurements were obtained by sampling 25 spicules per specimen for each spicule category Material MNHN LBIM DT 636 (unknown locality) BMNH 1881. 10.21.283 (Torres Strait) SMF various (n = 3) (Arafura Sea) Choanosomal oxeas Subectosomal styles Holotype 520-1434 x 6-13 (896.2~9.2) Specimens 772-1565 ~5.5-16 ( 1 0 8 0 . 4 ~10.1) Ectosomal megascleres Echinating acanthostyles Absent 88-118 x 5-1 1 (101.2~7.4) Absent 82-110 3-1 1 (100.4~8.0) Absent 68-127 ~5-12 (107.4~9.4) Echinodicbum clathrioides Hentschel (Figs 93, 94, 110f, 1log; Table 23) Echinodictyum clathrioides Hentschel, 1911: 389-390. Material Examined Syntypes (4 specimens, not seen). ZMH 4454: NW. of Middle Bluff, Shark Bay, W.A., 25" 49/S., 113" 27'E., 7-8 m depth, 21.ix.1905, coll. W. Michaelsen & R. Hartmeyer, (Hamburg Expedition, stn 11, dredge); E of bore, between Eagle Bluff and Baba Head, Freycinet Estuary, Shark Bay, W.A., 26" 06'S., 113" 35/E., 7-11 m depth, 6.ix.1905, coll. W. Michaelsen and R. Hartmeyer (Hamburg Expedition, sm 10, dredge). Other material (all material collected by the author using SCUBA, unless otherwise indicated). Southwest Coast, W.A.: WAM 666-81(1) (fragment NTM 21720): Cockbum Sound, 32' lllS., 115" 43'E., depth unknown, 18.x.1958, collector unknown. WAM 648-81(1) (fragment NTM 21704): W. of City Beach, Perth region, 31" 56.g1S., 115'42. llE., 16.5 m depth, 12.x.1976, coll. L. Marsh (FRV 'Hinders', stn 16, trawl). NCI Q66C-2922-R (fragment NTM 23756): Bussellton Jetty, Geographe Bay, 33" 37.8'S., 115" 20.2'E., 8 m depth, 29.ii.1989, coll. NCI. Houtman-Abrolhos Is, W.A.: NCI Q66C-4230-0: NW. side of Goss Passage, Wallabi Group, 28" 29 - OIS., 113" 36.6'E., 20 m depth, 13.ix.1990, coll. NCI. Q66C-4669-Q: 1 km N. of Gun I., 2 km off Sandy I., Pelsart Group, 28' 52 -51IS., 113' 51 -07/E., 5 m depth, 19.ix.1990, coll. NCI. Shark Bay Region, W.A.: WAM 637-81(1) (fragment NTM 21709): Unspecified locality, Shark Bay, 8.vii.1962, coll. unknown (FRV 'Peron', stn B16). NTM 22955: Sunday I., Dirk Hartog. I., Shark Bay, 26" 07.5'S., 113" 14-O'E., 8-9 m depth, 15.vii.1987 (U.S.S.R. RV 'Akademik Oparin', stn SB5). NTM 22969: W. of Carnarvon, 24" 55.6-56.5IS., 112" 50.8-53.5IE., 80-85 m depth, 14.vii.1987 (U.S.S.R. RV 'Akademik Oparin', stn SB6, beam trawl). Exmouth Gulf Region, W.A.: NCI Q66C-1293-V (fragment NTM 23465): 500 m NW. of Learmonth Jetty, Exmouth Gulf, 22" 12- a's., 114' 06.1 E., 2 m depth, 16.viii.1989, coll. NCI. NCI Q66C-1499-U (fragment NTM 23375): 1 . 5 km from shore, N. of Barrow I., 20' 38. 8'S., 115' 28.7/E., 22 m depth, 26.viii.1988, coll. D. Low Choy & NCI (stn W S 6 8 ) . Northwest Shelf Region, W.A.: NTM 23026: N. of Amphinome Shoals, 19" 19-7-23.3'S., 119" 08.8-12.2'E., 50m depth, 19.vii.1987 (U.S.S.R. RV 'Akademik Oparin', stn NWS55, beam trawl). NTM 22426: NW. of Amphinome Shoals, 19" 12' S., 118' 36-5'E., 76-80 m depth, 01.vi.1985, coll. B.C. Russell & AFZ (stn BCR 8515, haul 6, NWS36, trawl). Substrate and Depth Range Sand, silt and rocky substrates; 2-85 m depth. Australian Raspailiidae Geographical Distribution South-west and north-west coasts of Australia (Fig. 93m). Description -----+ape. Regularly folded or convoluted vases (80-365 mm high, 70-275 mm wide), half vases or fans, with symmetrical or crumpled, thin or thick walls (3-11 mm), with regular or convoluted margins, and usually with basal stalk 50-96 mm long, 15-35 mm diameter. Colour. Live coloration grey to grey-black (Munsell 2 5Y 312-2 -5B 412) (Figs 110f, 110g) or grey-beige in shallow water material, whereas deeper water specimens grey-brown alive. Dried and preserved specimens light yellow or yellow-brown, with reddish or violet tinge. Oscula. Numerous small oscula, 1-3 rnm diameter, evenly dispersed over at least 1 surface of vases or fans, whereas inhalant pores not easily discernible from tightly reticulate fibre construction at surface. 1368 J. N. A. Hooper Texture and su$ace characteristics. Texture coarse and difficult to tear. Surface of most material regular and smooth on both interior and exterior faces of vases or fans. However, in type material, according to the original description, and in two other specimens (WAM 648-81(1) and NTM Z2426), exterior surface slightly uneven and 3-dimensional, although not as pronounced as in E. mesenterinum for example, and interior surface much smoother and more regular, with very few digitate projections. Microscopic net-like appearance of surface (produced by close-set cavities and pores) enhanced by even surface, whereas in most other species of Echinodictyum this trait is masked by paratangential appearance of surface. Ectosome and subectosome. Ectosome reticulate, lacking any specialised spicule skeleton, but with dense layer of erect acanthostyles protruding through ectosome, lying just below dermal membrane, producing dense paratangential surface. Echinating spicules much more prevalent on exterior fibres than on choanosomal fibres. Subectosomal extra-axial skeleton reduced to individual styles dispersed throughout choanosomal and subdermal regions, echinating fibres in sparse bundles. Choanosorne. Choanosomal skeleton regularly or irregularly reticulate, sometimes meandering, producing exceptionally wide and elongate meshes, 260-1400 pm long, 120850 pm wide. Fibres typically broad, 70-240 pm diameter, with very little spongin, fully cored by oxeas. Choanosomal fibres sparsely echinated by acanthostyles, unlike ectosomal region, and echinating spicules stand perpendicular to fibres. Mesohyl very light, and when intact, choanocyte chambers can be seen between fibre meshes, up to 230 pm diameter. Megascleres (refer to Table 23 for dimensions). Choanosomal axial oxeas symmetrical, of 1 size category, relatively slender, gently curved at centre, tapering to abruptly sharp points, slightly rounded tips, or occasionally strongylote. Extra-axial subectosomal styles long and slender, usually straight, with evenly rounded or tapering bases, and with sharply pointed or bluntly rounded tips; few are anisoxeote. Ectosomal megascleres absent. Echinating acanthostyles short, mostly thick, conical, straight, with slightly subtylote regions on both base and tips. Spines relatively evenly distributed over base, shaft and tip, although usually with an aspinose area just below base. Density and size of spines on spicules varies considerably between specimens (Fig. 93c-k). Spines spatulate, with serrated margins. Microscleres absent. Remarks A prominent feature of the skeleton in this species, which serves to distinguish it from other Echinodictyum, is the concentration of acanthostyles standing erect on the exterior surface of peripheral fibres. The preponderance of acanthostyles in this region produces a dense ectosomal skeleton, and this also produces the optically even surface that is characteristic for this species (but atypical for the genus). In having acanthostyles perpendicular to the surface of the sponge, E. clathrioides shows similarities to the South Australian species E. austrinus, although in the latter species these spicules form discrete brushes. Another interesting feature discovered in E. clathrioides (by scanning electron microscopy), which differentiates it from most other Echinodictyum is the presence of spatulate spines on acanthostyles, where each spatulum is serrated (Fig. 94e): other Echinodictyum have simply sharply pointed spines. However, the value of this character in the systematics of the whole genus is restricted at the moment, since this feature has not yet been described for all species. A comparison between specimens from four localities along the W.A. coast (Table 23) shows that the species is relatively conservative in its spicule dimensions, except for specimens from the Northwest Shelf, which appear to have relatively larger acanthostyles. However, these two specimens are from deeper water (50-85 m depth), whereas other material examined was from shallower depths accessible by SCUBA. The division of choanosomal oxeas into two size categories by Hentschel (1911) was unnecessary and misleading: the range of spicule size in E. clathrioides is no greater Australian Raspailiidae Fig. 94. Echinodic@um clathrioides Hentschel: a, specimen (WAM 648-81(1), fragment NTM 21704) (scale = 30 mm); b, specimen (NTM 22955) (scale = 30 mm); c, SEM of skeletal structure; d, SEM of fibre characteristics; e, SEM of echinating acanthostyles; f , spine morphology. than in any other Echinodictyum species. Similarly, the widths of both extra-axial styles and acanthostyles recorded by Hentschel differ from those observed in present material, but these differences are trivial and in all other details all specimens correspond with Hentschel's species. As noted above, this species is similar to E. austrinus, E. carlinoides, E. costiferum and E. mesenterinum in having swollen tips on acanthostyles. It is also similar to E. mesenterinum in shape (see below), with comparable spicule dimensions, but surface features are quite different: E. mesenterinum has a very rugose paratangential surface composed of protruding fibre endings, whereas E. clathrioides has a very smooth surface striated with a microscopic net-like reticulation. Hentschel (1911) also compares the species with E. costiferum (see below) and E. jabellatum Topsent (Fig. 83j-I), although spicule dimensions and surface features differ among these species. 1370 J. N. A. Hooper Table 23. Comparisons in spicule measurements between specimens of Echinodictyum clathrioides Hentschel Measurements are given in micrometres, presented as ranges (and means) of lengthxwidth. For material described in this work these measurements were obtained by sampling 25 spicules per specimen for each spicule category Material Choanosomal oxeas ZMH 4454A Shark Bay,W.A. 160-592 x5-10 SW coast, W.A. (n = 2) Shark Bay, W.A. (n = 3) Exmouth Gulf W.A. (n = 2) N.W. Shelf W.A. ( n = 2) Ectosomal megascleres Echinating acanthostyles Absent 60- 105 x4 Absent 69-97 x6-11 (85.5~9-0) 68-97 x 8-14 ( 8 6 . 0 ~10.6) 72-97 ~5-11 (84.2~8.9) 94-132 x 7-15 (115.6~11.5) Syntypes 320400 x2 4 69-358 x5-13 (221.3~9.2) 82-604 x 3-17 (245.1~9.1) 132-342 x3-13 (234.4~9.1) 176-353 x4-15 (240.8~10.1) - A Subectosomal styles Specimens 264327 x4-6 (308.3~5.2) 232-378 x3-5.5 (294.6~4.1) 223-298 ~ 2 4 . 5 (273.5~3~4) 222-300 x 3-5 (251-4~3.8) - - Absent Absent Absent ----- From Hentschel (1911: 389). Echinodictyum conulosum Kieschnick (Figs 95, 96, 110h; Table 24) Echinodictyum conulosum Kieschnick, 1900: 570-1. Material Examined Holotype. PMJ POR85: Thursday I., Torres Strait, Queensland, 'shallow' depth, no other details known. Other material (all material collected by the author using SCUBA,unless otherwise indicated). Northwest Shelf, W.A.: NCI Q66C 1619-2 (fragment NTM 23493): W. end of Lewis I., Dampier Archipelago, 20" 36.6' S., 116" 35.7/E., 8 m depth, 01.ix.1988, coll. NCI. NTM 21828: W. of Port Hedland, 19" 26.g1S., 118' 54.2/E., 50 m depth, 30.viii.1983, coll. T. Ward (CSIRO RV 'Soela' S04/83, stn 126, trawl). NTM 21789: same locality, 19" 04.1 IS., 118" 57.g1E., 84 m depth, 29.viii.1983, coll. T. Ward (CSIRO RV 'Soela' S04/83, stn 124, trawl). Darwin Region, N.T.: NTM 2890: Dudley Point Reef, EPMFR, 12' 25 .OIS., 130" 48.4'E., 10 m depth, 31.viii.1982 (stn EP8). NTM 22252: same locality, 20" 24.5'S., 130" 48.01E., lorn depth, 12.iv.1985, coll. C. Hood & J.K. Hanley (stn EP22). Gulf of Carpentaria, Qld (all material collected by author, aboard U.S.S.R. 'Akademik Oparin', dredge): NTM 23978: N. of Mornington I., 15" 39.g1S., 139" 36.2'E., 40 m depth, 2.xii.1990 (stn JH-90-038). NTM 23984: W. of Cape Keerweer, 13" 58.3'S., 141" 04 6'E., 32 m depth, 3.xii.1990 (stn JH-90-041). NTM 23993: W. of Ken Reef, 11" 39.7' S., 140" 10.5'E., 59 m depth, 5.xii.1990 (stn JH-90-045). NTM 24001: SW. of Torres Strait, 11' 12.0' S., 1400 35 .6l E., 53 m depth, 6.xii.1990 (stn JH-90-047). Great Barrier Reef, Qld: QM GL800 (fragment NTM 21507): E of Murdock I., Howick Is Group, 14' 36'S., 145" 03'E., 14 m depth, 18.ix.1979, coll. QFS (stn AQS 1BI19-2). - Substrate and Depth Range Shallow coastal and shallow offshore rock reefs, in mud or areas with high sedimentation; 8-84 m depth. Geographical Distribution North-west and north-east coasts of Australia (Fig. 95e). Australian Raspailiidae Fig. 95. Echinodictyum conulosum Kieschnick (holotype PMJ POR85): a, choanosomal axial oxeas; b, subectosomal extra-axial style; c, echinating acanthostyle; d, section through peripheral skeleton (scale = 1 mm); e, known Australian distribution. Description Shape. Elongate, conical, lobate or digitate, reticulate-clathrous sponges, 75-320 mm long, 35-200 mm wide, with short basal stalk, 13-76 mm long, 4.5-26 mm diameter, and cavernous construction. Colour. Live coloration jet black, or black with purple tinge (Munsell 5RP 3-2-512) (Fig. 110h), usually silt covered in situ in shallow water material. Preserved material varies from black to purple. Deeper water specimens lack pigmentation entirely, being drab beige or beige-brown. Oscula. Oscula undifferentiated from excavated reticulate surface, but with numerous inhalant pores, 90-450 pm diameter, scattered over intact dermal membrane. Texture and surface characteristics. Texture firm and flexible. Surface covered with numerous large conulose (or small digitate) projections, up to 4 mm high, conical in shape, with pointed, bifurcated, furry tips. Remainder of surface excavated by cavities with a skin-like membrane stretched across, and this is still intact in places in preserved material. Overall appearance is reminiscent of an Echinoclathria (Microcionidae) or Acanthella (Axinellidae). Ectosome and subectosome. Ectosome membraneous, without any specialised spiculation, although tips of megascleres (especially extra-axial styles) often protrude through surface; subdermal-surface brushes most common on ends of conules. Embedded in skin-like membrane of all shallow-water material are many cell-like pigment granules, but these 1372 J. N. A. Hooper pigment spots do not occur in deeper water material. Remnants of subectosomal skeleton occur as individuals or groups of styles embedded in (or echinating) spicule tracts, also scattered throughout mesohyl, and often tips poke through membraneous ectosomal skeleton. Choanosome. Choanosomal fibres relatively large, 110-360 p m diameter, differentiated into primary ascending and secondary transverse connecting tracts. Fibres fully cored with choanosomal oxeas, moderately heavily echinated by acanthostyles standing perpendicular to fibres. Fibres have minimal spongin content, and fibre anastomoses form large elongate meshes, 110-750 p m diameter, lining ovoid choanocyte chambers, 40-110 p m diameter. Mesohyl matrix fairly dense, with abundant (purple or brown pigmented) spongin type B and megascleres scattered between fibres. Echinodictyum conulosum Kieschnick: a, holotype (PMJ POR85) (scale = 30 mm); b, specimen (NTM 20890) (scale = 30 mm); c, specimen (NTM 22252); d, SEM of skeletal structure; e, SEM of peripheral skeleton, showing long extra-axial styles; f , SEM of fibre characteristics; g, echinating acanthostyle. Fig. 96. Australian Raspailiidae 1373 Megascleres (refer to Table 24 for dimensions). Choanosomal oxeas variable in length, usually slender, larger forms mostly straight, smaller forms with even central curvature, tapering to sharply pointed or slightly rounded tips. Subectosomal extra-axial styles slender, straight or slightly curved towards tips, with evenly rounded bases. Many anisoxeote forms of styles also occur. Ectosomal megascleres absent. Echinating acanthostyles relatively long, slender, cylindrical, with slightly subtylote bases, tapering to long points. Small spines evenly distributed over shafts, bases and points of spicules. Spines sharply pointed and slightly recuwed. Microscleres absent. Table 24. Comparisons in spicule measurements between specimens of Echinodictyum conulosum Kieschnick Measurements are given in micrometres, presented as ranges (and means) of lengthxwidth. For material described in this work these measurements were obtained by sampling 25 spicules per specimen for each spicule category Material Choanosomal oxeas Subectosomal styles Ectosomal megascleres Echinating acanthostyles PMJ POR85 (Torres Strait) Holotype 354-545 x 3-9 (430-4~5.3) Absent 92-129 ~7-12 (115.3~9.3) NW. Shelf, W.A. NCI and NTM various (n = 3) Darwin, N.T. NTM various (n = 2) GBR, Qld QM GL800 Specimens 328-568 ~2.5-15 (409.4~4.8) 202-672 x2-11 (359.5~4.6) 311-506 ~2.5-7 (413.6~4.3) Absent 85-118 x2-11 (98.3~8.8) 79-1 12 x 6-11 (97.9~8.8) 89-124 x 3-8 (110.8~6.0) Absent Absent Remarks The published description of E. conulosum implies that the holotype is from Ambon, Indonesia (Kieschnick 1900), whereas the species is actually from Torres Strait (Wiedenmayer et al. in press). The published description also contains several other errors: skeletal fibres of E. conulosum are not undifferentiated, as cited, but they can be separated easily into primary and secondary elements, with the larger primary fibres ascending towards the surface, interconnected by smaller secondary elements running transversely, and the maximum diameter of fibres greatly exceeds the dimensions quoted [i.e. 280-750 pm not 200-300 pm]. Kieschnick also omitted to describe the extra-axial styles, which are fairly abundant in this species. Spicule dimensions for the species across its broad, tropical geographical range appear to be relatively consistent, although material from the east coast of Australia has marginally larger acanthostyles than specimens from the west coast (Table 24). In growth form E. conulosum belongs to the group of reticulate Echinodictyum, which includes E. asperum, E. costiferum, E. clathrioides, and E. lacunosum. It appears to be most closely related to E. asperum in coloration and surface features, but spicule dimensions differ between these species; E. conulosum has smaller choanosomal oxeas, larger subectosomal styles, and smaller acanthostyles (compare Figs 86-87 and 95-96). Echinodictyum costiferum Ridley (Fig. 97) Echinodictyum costiferum Ridley, 1884: 455-456, pl. 42, fig. r. Not Spongia costifera Lamarck, 1814: 432.-Topsent, 1932: 63. J. N. A. Hooper Fig. 97. Echinodictyum costiferum Ridley (holotype BMNH 1881.10.21.292): a, choanosomal axial oxeas; b, subectosomal extra-axial styles; c, echinating acanthostyle; d, known Australian distribution; e, holotype (scale = 30 mm); f , section through peripheral skeleton (scale = 1 mm); g, photomicrograph of echinating acanthostyle (scale = 50 pm). Material Examined Holotype. BMNH 1881.10.21.292: Port Molle (now known as Airlie), Whitsunday Is region, Queensland, 20" 15'S., 148" 43'E., depth unknown, May 1881, coll. R.W. Coppinger (HMS 'Alert', dredge). Substrate and Depth Range Coral reef, depth unknown. Geographical Distribution Known only from Whitsunday Is region, Great Barrier Reef, Qld (Fig. 9 7 4 . Description Shape. Vasiform, clathrous, reticulated sponge, 50 mm high, 70 mm maximum breadth of vase, with walls 3-7 mm thick, without stalk or basal attachment. Colour. Live coloration unknown, beige-brown in dry state. Australian Raspailiidae 1375 Oscula. Oscula and inhalant pores are undetectable amongst the excavated reticulated surface of the sponge. Texture and surjice characteristics. Texture harsh and brittle when dry. Inner surface of vase uneven, with scattered small digitate projections 2-5 mm high. Exterior surface of vase with numerous close-set digitate projections, 4-8 mm high, terminating in sharp points, between which are ridges and canals. Both surfaces cavernous and excavated, formed by close-set reticulated fibres. Ectosome and subectosome. Ectosomal skeleton membraneous, without specialised spiculation, with remnants of a skin-like membrane stretched across surface cavities. Subectosomal skeleton vestigial, with only single extra-axial styles dispersed between fibres, not protruding through surface, many embedded in fibres in axial skeleton, standing perpendicular to them, singly or in brushes. Choanosome. Choanosomal skeletal architecture irregularly reticulate, with well differentiated primary ascending fibres, 125-310 pm diameter, extending into peripheral skeleton, forming characteristic pointed digitate projections, and interconnected by secondary transverse fibres, 40-150 pm diameter. Fibres have light spongin A, fully cored by choanosomal oxeas, and lightly echinated by acanthostyles. Fibres form oval meshes, 14C-750 pm diameter, wider in axis than at periphery. Fibre anastomoses enclose small oval choanocyte chambers, 35-90 pm diameter; abundant megascleres between fibres. Megascleres. Choanosomal oxeas short or long, always thick and robust, straight or only slightly curved at centre, tapering to sharply pointed or stepped tips [I724287 1)532 x 6-(10.3)-14 pm]. Subectosomal extra-axial styles relatively short and slender, with evenly rounded bases, tapering to sharp and often sinuous points; larger examples straight or slightly curved near basal end [290-(382.4)-47 1 x 1 - 5 4 3 .4)-6 pm]. Ectosomal megascleres absent. Echinating acanthostyles relatively long and slender, with slightly subtylote bases, tapering to thin and rounded tips, often with slight swelling near apex. Spines evenly dispersed over spicules or slightly more abundant and more recurved on extremities of spicules. Spines recurved, spatulate and serrated 1924102)-113x6-(7 -0)-9 pm]. Microscleres absent. a Remarks Echinodictyum costiferum belongs to the honey-combed reticulate group of species, superficially resembling the microcionid Echinoclathria group of species, and this species is altogether most reminiscent of the West Indies Pandaros acanthifolium (Duchassaing & Michelotti) (family Microcionidae). The present species can be differentiated from other Australian reticulate Echinodictyum in the specific dimensions of its megascleres, the geometry and spination of its (more slender) acanthostyles, but little else. It is particularly close to E. lacunosum in the geometry and size of megascleres, and it is possible that the two species are synonymous, but both are relatively poorly known only from single specimens. The presence of slight swellings on the tips of acanthostyles, bearing small, abundant, and slightly recurved spines, is similar to (but not as well developed as) those found in E. austrinus, E. carlinoides, E. clathrioides and E. mesenterinurn. Ridley (1884) also remarks on the similarities between E. costiferum and the Indian Ocean species E. pykei (Fig. 84n-p) and E. laciniatus (Carter) (Fig. 84b-d). ? Echinodictyum fruticosum Hentschel (Fig. 98) Echinodictyum fruticosum Hentschel, 1911: 390.-Dendy, 1922: 73. Material Examined None. The holotype has not yet been found [ZMH 4455: Entrance to South Passage, Shark Bay, W.A., 26" 07/S., 113" 10IE., 9 m depth, 16.vi.1905, coll. W. Michaelsen & R. Hartmeyer (Hamburg Expedition, sm 23, dredge)]. J. N. A. Hooper Fig. 98. ? Echinodictyum fruticosum Hentschel: a, holotype (ZMH 4455); b, choanosomal axial oxeas; c, subectosomal extra-axial style; d, echinating acanthostyle; e, known Australian distribution ( a 4 redrawn from Hentschel 1911: fig. 53). Substrate and Depth Range Gravel and solitary rocks, 9 m depth. Geographical Distribution Known only from Shark Bay, Western Australia (Fig. 9%). Description Shape. Shrubby growth form, 50 mm high, 40 mm wide, 30 mm thick, with plumose and anastomosing branches on the surface, about 5-8 mm apart. Colour. Exterior surface black due to numerous pigment deposits, whereas choanosome is paler. Oscula. Not observed. Texture and surface characteristics. Surface with wide-meshed reticulation covered with small pointed digits; surface digits produced by terminal fibres between 5-8 mrn apart, across which stretches a skin-like dermal membrane; membrane completely smooth, but ends of protruding fibre endings roughened and hairy. Texture unknown. Ectosome and subectosome. Ectosome membraneous, without specialised spiculation, with fibre endings protruding a relatively long distance through surface. Subectosomal skeleton reduced to individual extra-axial styles dispersed throughout mesohyl. Choanosome. Choanosomal skeleton consists of thick fibres, up to 20 spicule-widths or 4 mm wide, densely packed with oxeas and heavily echinated by acanthostyles; acanthostyles stand vertical to fibres, with bases only slightly covered by spongin. Fibres composed of very light spongin, and fibre meshes elongate. Megascleres. Choanosomal oxeas slender, cylindrical, curved, and abruptly pointed (153-256 x 6-9 pm).. Extra-axial subectosomal styles relatively rare, smooth, slender, slightly curved or straight, tapering slightly towards base (253-306 x 1 pm).. Ectosomal megascleres absent. Echinating acanthostyles slender, evenly tapering from thick base to sharp point; shaft evenly spined, base heavily spined. Spines very small, barely noticeable under low magnification, (and from Hentschel's illustration they appear to be) sharply pointed (83-96 x 6-7 pm).. Microscleres absent. Australian Raspailiidae 1377 Remarks This poorly known species is known only from the holotype from Shark Bay, W.A. Unfortunately, that material is not presently available for loan, and the description and figures provided here are taken from Hentschel (1911). Hentschel notes that E. fruticosum is similar to three other species in its growth form, surface features and in having sparsely spined and slender acanthostyles: E. asperum, E. cavernosum and E. clathratum. He suggests that E. asperum differs from the present species in lacking extra-axial styles, which is incorrect, whereas the other two species have significant differences the dimensions of acanthostyles in particular, It is possible that E. fruticosum is a synonym of E. asperum, since the two species are very similar in growth form, and spicule dimensions are not substantially different, but this synonymy has yet to be confirmed from re-examination of Hentschel's (1911) type material, if it becomes available for loan. Echinodictyum lucunosum Kieschnick (Fig. 99) Echinodictyum lacunosum Kieschnick, 1898: 56.-Kieschnick, 1900: 570, pl. 44, fig. 9. Material Examined Holotype. PMJ POR84: Thursday I., Torres Strait, Qld, 10" 35 I S., 142" 13I E., 'shallow' depth, no other details known. Substrate and Depth Range Unknown. Geographical Distribution Known only from type locality. Torres Strait (Fig. 9 9 4 . Description Shape. Elongate, lobate digitate, irregularly cylindrical, clathrous sponge, now 75 mm long (previously 85 mm long; Kieschnick 1900), 24 mm maximum diameter, without stalk or basal attachment (although reported as having a 'wide basal part'). Colour. Live coloration unknown, yellowish brown in ethanol. Oscula. Several large oscula present on lateral sides of sponge, 3-4 mm diameter, but difficult to distinguish from excavated porous consistency of the sponge. Texture and surface characteristics. Texture tough and flexible. Sponge has honeycombedreticulate structure, excavated by canals and pores; surface covered with numerous pointed or lobate digitate conules, 2-4 5 mm high. An easily detachable skin-like dermal membrane covers most surface excavations, stretched between conules. Ectosome and subectosome. Ectosomal skeleton membraneous, without specialised spiculation, although tips of acanthostyles on peripheral fibres may poke through surface. Subectosomal skeleton vestigial, with only single or groups of extra-axial styles scattered between fibres; brushes of subectosomal spicules may also protrude through surface, particularly at ends of surface conules. Choanosome. Choanosomal skeletal architecture simply irregularly reticulate, without any obvious differentiation into primary or secondary fibre network, although fibre diameter varies considerably (145-490 pm diameter). Fibres with minimal spongin, fully cored with choanosomal oxeas and heavily echinated by acanthostyles;. fibre anastomoses form elongate or oval meshes, 160-580 pm diameter, lining small oval choanocyte chambers, 45-120 p m diameter, and mesohyl matrix only lightly invested with spongin. Moderate quantities of spicules scattered between fibres, usually subectosomal styles. Megascleres. Choanosomal oxeas very variable in size, ranging from robust to whispy forms, usually only slightly curved at centres, sometimes straight, mostly tapering to sharp points or sometimes rounded tips; few anisoxeote forms also observed [187-(300-5)622 x 8 4 12 2)-20 pm] . - J. N. A. Hooper Fig. 99. Echinodictyum lacunosum Kieschnick (holotype PMJ POR84): a, choanosomal axial oxeas; b, subectosomal extra-axial style; c, echinating acanthostyles; d, known Australian distribution; e, holotype (scale = 30 mm); f , section through peripheral skeleton; g, photomicrograph of echinating acanthostyle (scale = 50 pm). Subectosomal extra-axial styles relatively short and slender, straight or slightly curved near basal end, with rounded bases and tapering to sharply pointed tips [371-(403.4)447 x 2 5 4 3 8)-5 pm]. Ectosomal megascleres absent. Echinating acanthostyles moderately long, slender, conical, with slightly subtylote bases, tapering to rounded tips. Spines sparsely dispersed over spicules, heaviest on base and point, lighter on shaft, with an aspinose area just below base. Spines sharply pointed and recurved [96-(114 7)-132 x 7 4 1 0 - 2)-14 pm]. Microscleres absent. - - - Remarks The published description of E. lacunosum implies that the holotype is from Ambon, Indonesia (Kieschnick 1900), whereas the species is actually from Torres Strait (Wiedenmayer et al. in press). This species has an Echinoclathria-like honeycomb-reticulate growth form, similar to E. asperum, E. conulosum and E. costiferum (see above), and it is also very reminiscent of the growth form of the common Indo-Pacific microcionid sponge Clathria Australian Raspailiidae 1379 vulpina (Lamarck) [also known as C. ji-ondifera (Bowerbank); Hooper, unpublished data]. Echinodictyum lacunosum is known only from the holotype, and it has few other features to distinguish it from other species other than its reticulate growth form and specific megasclere dimensions (Fig. 99). Echinodictyum mesenterinum (Lamarck) (Figs 100-103, 110i; Table 25) Spongia mesenterina Lamarck, 1814: 444. ~chinodictyurnmesenterinurn.-Carter, 1882: 114; Ridley, 1884: 185; Topsent, 1932: 101; Hooper, 1984: 55. Spongia bilamellata Lamarck, 1816: 436 (in part, var. 8). Echinodictyum bilamel1atum.-Ridley, in Ridley & Duncan, 1881: 493, pl. 28, figs 1-6; Ridley, 1884: 454; Hentschel, 1911: 385; Hallmann, 1912: 299; 1914a: 267; Dendy & Frederick, 1924: 504; Topsent, 1932: 69, 101, pl. 6, fig. 5; 1933: 23; Burton, 1938: 15, 20; Guiler, 1950: 7. Kalykenteron elegans Lendenfeld, 1888: 216.-Hallmann, 1914a: 267. Echinodictyum e1egans.-Hallmann, 1912: 171, pl. 23, fig. 1, text-fig. 35. Kalykenteron silex Lendenfeld, 1888: 217.-Hallmann, 1914a: 267. Echinodictyum topsenti de Laubenfels, 1936: 63. Thalassodendron @pica.-Whiteleg 1901: 86 (in part); Hallmann, 1912: 171, 203. Not Thalassodendron typica Lendenfeld, 1888: 233. Echinonema vasiplicata Carter, 1882: 114.-Ridley, 1884: 454; Hentschel, 1911: 385. Material Examined Lectotype (here designated). MNHN LBIM DT568: ? Australian Seas, Peron & Lesueur collection. Paralectotype. MNHN LBIM DT3355: 'Baie des Chiens Marins' (Shark Bay), Westem Australia, Peron & Lesueur collection. [Not-MNHN LBIM DT569, erroneous identification (=Cribrochalina, LBIM DT 603: Haliclonidae)]. Lectotype of Spongia bilamellata, var. 8 (here designated)-MNHN Kangaroo I., Great Australian Bight, Peron & Lesueur collection. Paralectotype of Spongia bilamellata, var. 8-MNHN LBIM DT 3384: same locality. Holotype of Echinonema vasiplicata-BMNH 1877.5.21.1854 (NMV sponge archives photo 47/14): Fremantle, W.A., Bowerbank collection, dry. Lectotype of Kalykenteron elegans (here designated)-AM G9129 (NMV sponge archives photo 1/33-34): precise locality unknown, W.A., dry. Holotype of Kalykenteron silex -AM G9130 (NMV sponge archives photo 318-9): precise locality unknown, E Australia, dry. Holotype of Thalassodendron typica -AM 2958 (=B5481, B5492): precise locality unknown, W.A., dry. Paratypes (possible) of T. typica -AM 22141: Precise locality unknown, Tas. (specimen recorded in Burton 1938: 20). AM E958: FIV 'Endeavour' specimen, locality unknown. Other material (all material collected by the author using SCUBA,unless otherwise indicated). Philippines: NCI Q66C 5811-1: N. of Dumaguete, Negros Orientale, 9" 20. OIN., 123" 18-6lE., 18 m depth, 24.iv.1991 (stn JH-91-007). North-east Coast, Qld: BMNH 1881.10.21.265: Port Curtis, Gladstone region, Qld, 23" 55/S., 151" 23'E., beach debris, April 1881, coll. R.W. Coppinger (HMS 'Alert'). AIMS FN T268 (fragment NTM 22730): Davies Reef, GBR, Qld, 18" 501S., 147' 3g1E., 15 m depth, 25.ii.1982, coll. C.R. Wilkinson (stn Don. 214, scma). NCI Q66C-0387-L (fragment NTM 23751): S. comer of Pioneer Bay, Orpheus I. reef slope, GBR, 18'37'S., 146" 29. llE., 10 m depth, 19.ii.1987, coll. NCI. Cobourg Peninsula Region, N.T.: NTM 256: Coral Bay, Port Essington, CPMNP, 11" 11.5IS., 132" 03/E., 4-6 m depth, 17.x.1981 (stn CP20). NTM 289: same locality, 6 m depth, 18.x.1981 (stn CP21). NTM 291: same locality, 11" 09.4' S., 132" 04'E., 5 m depth, 19.x.1981 (sm CP23). NTM 2368, 373: same locality, 11' 10-8/S., 132" 03. llE., 8 m depth, 19.vii.1981 (stn CP11). NTM 21360: same locality, 11" 11.3/S., 132" 03.75/E., 5 m depth, 16.v.1983 (stn CP60). NTM 21387: same locality, 11" 11.3' S., 132' 03 -7IE., 6 m depth, 17.v.1983. NTM 22490: same locality, 1lo09.4lS., 132" 04.01E., 7 m depth, 13.ix.1985 (stn CP71). NTM 23298, 3303, 3305: Leeward of outer coral bamer, Coral Bay, 11" O9.4/S., 132" 04. OIE., 7 m depth, 12.ix.1986 (stn CP70). NTM 21341: Table Head, Port Essington, CPMNP, 11" 13 - 5IS., 132" 10.5'E.. 6 m depth, 12.v.1983 (stn CP46). NTM 23259, 3269, 3273, 3282, 3287, 3290: SW. side of cliff, Table Head, 11" 13.5/S., 132" 10 5'E., 5 m depth, ll.ix.1986 (stn CP69). NTM 22524: Orontes reef, mouth of Port Essington, 11' 03.6's.. 132' 05.4/E., 18 m depth, 16.ix.1985 (stn CP78). NTM 2541: Port Bremer, 11" 08.5IS., 132" 18. SIE., 7 m depth, 1.v.1982 (stn CP33). NTM 2118: Sandy I. No. 2, 11" 05.5/S., 132" 17'E., 7 m depth, 20.x.1981 (stn CP26). NTM 2547: same locality, 11' 05. OIS., 132' 16.5IE., 14 m depth, 2.v.1982 (stn CP34). Darwin Region, N.T.: NTM 2907, 916: EPMFR, 12" 25 -OIS., 130' 48.4/E., 10 m depth, 31.viii.1982 (stn EP8). NTM 2926: J. N. A. Hooper 1380 same locality, 12" 24-5/S., 130" 48.01E., 12 m depth, 13.ix.1982 (stn EP9). NTM 21060: same locality, 12" 24.7/S., 130" 48 .OIE., 13 m depth, 9.xi.1982 (stn EP11). NTM 22254: same locality, 12'24-5'S., 130'48.11E., 10m depth, 12.iv.1985, coll. C. Hood & J.R. Hanley (stn EP22). NTM 22398: same locality, 8 m depth, 29.vii.1985 (stn EP23). NTM 22645: same locality, 12 m depth, 3.iv.1986 (stn EP28). NTM 22185: 'Bommies', Weed Reef, 12' 29.2/S., 130' 37.6/E., 6 m depth, 16.xi.1984 (stn WR6). NTM 2207, 208: Lee Point, 12' 19 2 / S., 130' 53.1 E., intertidal, 14.xi.1981 (stn LP1, by hand). NTM 2432: same locality, 13.xii.1981 (stn LP2). NTM 2780: Channel I., Middle Arm, 12' 33-4'S., 130°52.5/E., 2 m depth, 7.vii.1982 (stn CI2). NTM 21075: Fish Reef, Bynoe Harbour, 12" 26.2/S., 130' 25 -2/E., 12 m depth, 24.ix.1982 (stn FRI). NTM 2245: Indian I., Bynoe Harbour, 12" 35/S., 130" 33/E., 3 m depth, 31.viii.1981, coll. P. Byers (FV 'Skeleton', stn Don. 1, snorkel). NCI Q66C-0521-H, 0531-R (fragments NTM 23078,3086): Parry Shoals, Timor Sea, 11" 11-41IS., 129' 43 01 IE., 18 m depth, 13.viii.1987, coll. A.M. Mussig & NCI (stn AM 87-3). NTM 23135: same locality, 11" 12. 2S1S., 129" 42.71 'E., 16 m depth, 14.viii.1987 (stn AM 87-5). Wessel Is, N.T.: NTM 23905, NCI Q66C 4684-K: N. of Cape Wilberforce, Melville Bay, Gove Peninsula, 11" 52.6/S., 136' 33.3/E., 25 m depth, ll.xi.1990 (stn WI-1). Ashmore Reef, W.A. (all material collected by L. Vail & A.M. MussighNTM 22759: Coral reef slope, boat anchorage, ME. of West I., 12' 14.28/S., 122" 59. llE., 15 m depth, 23.vii.1986 (stn AR7). NTM 23231: same locality, 20 m depth. NTM 22771: same locality, 24.vii.1986, 18 m depth (stn AR9). NTM 22779: Channel into lagoon, 12" 13.4/S., 122" 59.01E., 15 m depth, 26.vii.1986 (stn AR12). NTM 22818: Outer reef slope, 12" 14.6/S., 122' 55 .OIE., 13 m depth, 28.vii.1986 (stn AR15). NTM 23236: same locality, 12' 14 3/S., 123" 56. OIE., 17 m depth, 27.vii.1986 (stn AR14). Scott Reef region, W.A. (all material collected by V. Krasochin, U.S.S.R. RV 'Akademik Oparin', dredge): PIBOC 012-274: near Seranga-patam, 13' 53 8 / S., 123' 52.8' E., 35-40 m depth, 20.xi.1990 (stn 71). PIBOC 012-71: near Scott Reef, 16" 41 .4'S., 121" 09 -6/E., 51-54 m depth, 4.xi.1990 (stn 26). PIBOC 012-112: 16" 32-2'S., 121' 10.9/E., 4 3 4 4 m depth, 4.xi.1990 (stn 28). PIBOC unreg.: 16' 36.7'S., 121' 11-llE., 50m depth, 17.xi.1990 (stn 39). PIBOC unreg.: 16" 05.01S., 121" 14.7/E., 73-75 m depth, 5.xi.1990 (stn 31). PIBOC unreg.: 16" 25-4/S., 121' 10.5/E., 47-49 m depth, 4.xi.1990 (stn 29). Northwest Shelf, W.A.: NMV unregistered (JH 15) (fragment NTM 21489): NW. of Bonaparte Archipelago, 13" 59'-14" OOIS., 124" 31-55'E., 86 m depth, 27.iii.1981, coll. C.C. Lu (RV 'Hai Kung', stn 70032702, trawl). NTM 22291: NW. of Lacepede I., 16'29-33/S., 121' 27-29/E., 38-40 m depth, 17.iv.1985, coll. B.C. Russell & AFZ (stn PT 8 5 4 , pair trawl). NTM 22268, 2275, 2309: NW. of Lacepede I., 16" 29-33/S., 121" 27-29/E., 3 8 4 0 m depth, 17.iv.1985, coll. B.C. Russell (stn I T 85-4 (tray 2)-NWS34, Taiwanese Pair-Trawl). NTM 23392: 50 m from shore, 2 km W. of Hermite I., Monte Bello Is, small group, 20" 27. llS., 115" 34.2/E., 6 m depth, 29.viii.1988, coll. D. Low Choy & NCI (stn NWS74PLC41). NCI Q66C-1507-F (fragment NTM Z3372), Q66C-1499-U (fragment NTM 23753): 2.5 krn from shore, N. of Barrow I., 20" 38 -8/S., 115" 28.8'E,, 22 m depth, 26.viii.1988 (stn NWS68). WAM 156-82 (fragment NTM 21716): Off Port Hedland, 20" 12'S., 118" 25/E., 14 m depth, 3.viii.1982, coll. J. Fromont (CSIRO RV 'Soela' S04182, stn 09-Don. 168, trawl). WAM 158-82 (fragment NTM 21730): same locality, 20" 13'S., 118' 28'E., 18 m depth, 5.viii.1982 (stn 13-Don. 169). NTM 21422: NE. of Port Hedland, 19" OllS., 119" 25/E., 80 m depth, 19.iv.1983, coll. R.S. Williams (stn NWS17, Taiwanese Pair-Trawl). NTM 2671, 673: N. of Port Hedland, 19" 16. OIS., 118' 50. OIE., 70 m depth, 4.v.1982 (CSIRO RV 'Sprightly' SP4182, stn 77-Don. 18, dredge). NTM 21170, 1171, 1172: W. of Port Hedland, 19' 30-9/S., 118" 48.7/E., 40 m depth, 26.iv.1983 (CSIRO RV 'Soela' S02183, stn B7-NWS8, beam trawl). NTM 21188: same locality, 19' 29.4' S., 118" 52.1 E., 39 m depth, 26.iv.1983 (stn B8-NWS9). NTM 21227: same locality, 19" 28 5'S., 118" 55 3/E., 40 m depth, 26.iv.1983 (stn B9-NWS10). NTM 22456: NW. of Amphinome Shoals, 19" 12/S., 118" 36-5/E., 76-80m depth, l.vi.1985, coll. B.C. Russell (stn BCR 8515 (haul 6)-NWS36, Taiwanese Pair-Trawl). NTM 23036: N. of Amphinome Shoals, 19' 19 -7-23.3 IS., 119' 08.8-12 2 / E., 50 m depth, 19.vii.1987 (U.S.S.R. RV 'Akademik Oparin', stn NWS55, beam trawl). NTM 2634: N. of Dampier Archipelago, 19' 30-41 IS., 116" 30-51 'E., 53 m depth, 11-12.v.1982, coll. L.A. Bullard, (stn OP(1,3,5)-Don. 5, BETD 'Yuan Yu' Taiwanese Pair-Trawl). NTM 21019: NW. of Dampier Archipelago, 20' 13/S., 117" 38 ' E., 30 m depth, 20.x.1982, coll. J. Blake (CSIRO RV 'Soela', stn 49(1)-Don. 29, trawl). NTM 2642: W. of 80 Mile Beach, 19'33.5/S., 119" 05.7/E., 35 m depth, 4.v.1982 (CSIRO RV 'Sprightly' SP4182, stn 78-Don. 15, dredge). Midwest Coast, W.A.: NTM 22954: W. of Sunday I., Dirk Hartog I., Shark Bay, 26" 07.5'S., 113" 14.01E., 9 m depth, 13.vii.1987 (U.S.S.R. RV 'Akademik Oparin', stn SB5). NTM 23318, 3319, 3320: Monkey Mia, Shark Bay, 25' 45/S., 113' 44/E., beach debris, June 1988, coll. R. Smoker (camed on dolphin's rostrum). Southwest Coast, W.A.: WAM 653-81(1) (fragment NTM 21725): W. of City Beach, Perth, 3lo56.9/S., 115"42.11E., 11.5 m depth, 12.x.1976, coll. L. Marsh (FRV 'Hinders', trawl). WAM 650-81(1) (fragment NTM 21728): W. of Swanbourne, 31' 58.42/S., 115" 39 - I 1 E . , 18.3 m depth, 12.x.1976, coll. L. Marsh (FRV 'Hinders', trawl). - - - - Australian Raspailiidae Fig. 100. Echinodictyum mesenterinum (Lamarck) (specimen NTM 20118): a, choanosomal axial B oxeas; b, subectosomal extra-axial style; c; echinating acanthostyle; d , section through peripheral skeleton; e, known Australian distribution. 4 = Substrate and Depth Range Associated with subtidal and shallow coastal rock, live and dead coral reefs, with a bathymetric distribution (in Australasia) from intertidal-86 m depth. Geographical Distribution Circum-Australian distribution (Fig. 100e): Fremantle region, W.A. (Carter 1882; present study), Abrolhos I. (Dendy & Frederick 1924), Bunbury, Geraldton and Shark Bay regions (Lendenfeld 1888; Hentschel 1911; Hallmann 1912; present study), Northwest Shelf, W.A., oceanic coral atolls of north-westem W.A., around to Cobourg Peninsula, N.T. (Hooper 1984) and Wessel Is, N.T. (present study), inshore reefs and inter-reef regions of central Great Barrier Reef region, Qld (Ridley 1884; Hallmann 1912; present study), Port Jackson, N.S.W. (Whitelegge 1901), Maria I., Tasmania (Burton 1938; Guiler 1950), to Kangaroo I., Great Australian Bight, S.A. (Lamarck 1814; Topsent 1932). Description Shape. Growth form characteristic, aptly termed 'birds nest' by earlier authors. Sponges typically erect, vasiform or cup-shaped, on short stalk or attached to substratum directly on an elongated basal attachment (Fig. 101a-I). Majority of material more-or-less symmetrical, with completely concentric vase walls; occasionally specimens incompletely concentric, convoluted flabelliform, or planar lamellate (particularly material from turbid waters). Size of specimens varied considerably, ranging from small vases less than 80 mm high, 60 mm maximum diameter, up to massive vasiform or flabelliform structures approximately 550 mm high, 420 mm maximum diameter. Walls of vases relatively thick, 10-28 mm wide; short basal stalk, when present, 25-67 mm long, 14-46 mm diameter. Colour. Live coloration consistent, varying only slightly in presence or absence of characteristic purple pigmentation. Subtidal shallow water or clear water specimens vary from evenly pigmented dark purple or blue-purple throughout (Munsell 5RP 3/2), or mottled 1382 J. N. A. Hooper Fig. 101. Echinodictyum mesenterinum (Lamarck): a, lectotype (MNHN LBIM DT568) (type species of the genus Echinodictyum Ridley) (scale = 30 mm); b, paralectotype (MNHN LBIM DT3355) / (MNHN LBIM DT603); (scale = 30 mm); c, lectotype of Spongia bilamellata Lamarck, variety 3 d, paralectotype of S. bilamellata var. /3 (MNHN LBIM DT3384) (scale = 30 mm); e, holotype of Echinonema vasiplicata Carter (BMNH 1877.5.21.1854) (scale = 30 mm); f , lectotype of Kalykenteron elegans Lendenfeld (AM G9129) (type species of the nominal genus Kalykenteron Lendenfeld) (scale = 30 mm); g, holotype of Kalykenteron silex Lendenfeld (AM G9130) (scale = 30 mm); h, holotype of Thalassodendron typica, sensu Whitelegge (AM 2958) (scale = 30 mm); i, flabellate specimen (NTM 22524) (scale = 30 mm); j, specimen, atypical form with even surface (NTM 22291) (scale = 30 mm); k, vasiform specimen (NTM 23392) (scale = 30 mm); 1, detailed view of the rugose surface characteristic of E. mesenterinum (scale = 30 mm). Australian Raspailiidae 1383 with peripheral fibres darkly pigmented and interior fibres unpigmented (beige or light brown; 2.5Y 814) (Fig. 110i), or rarely specimens almost entirely black. Deeper water material or specimens from highly turbid areas lack any pigmentation, being merely beige or brown. Pigmentation stable in ethanol. Oscula. Numerous small oscula scattered over interior and exterior walls of vases. Exhalant pores 2-5 mm diameter, more easily observed on interior walls where they are evenly scattered, whereas on exterior surface of vases pores more irregularly dispersed between uneven surface ridges. Inhalant pores minute, 0 -5-2 mm diameter, seen on intact sections of skin-like dermal membrane. Texture and surface characteristics. Texture very harsh, only slightly compressible or flexible, difficult to tear, which is consistent and very characteristic for the species. Fibre reticulation and entire skeletal construction is cavernous, with irregular projections dominating external surface, although interior and exterior surfaces usually quite different in ornamentation. Surface on exterior face greatly undulating, with pronounced 3-dimensional fibre reticulation and numerous ridges running more-or-less vertically. Exterior surface also interdispersed with low, rounded or pointed conules, with prominent net-like reticulation. Inner surface of vase much less undulating, and superficial fibre reticulation less markedly paratangential and more compressed. Both surfaces optically hispid. Ectosome and subectosome. Where intact, ectosome membranous, without specialised spiculation, with tympanic skin-like membrane stretched between adjoining surface conules. Membrane usually perforated with small inhalant pores. Echinating megascleres on peripheral fibres also pierce surface and partly obscure features of ectosomal skin. Ectosomal fibres contain heavy deposits of darkly pigmented cell-like granules, 5-10 pm diameter. Fibres in peripheral skeleton taper to relatively sharp points at ends, and aggregations of peripheral fibres form prominent surface conules. Long subectosomal extra-axial styles may protrude from tips of surface conules, singly or in sparse brushes, representing remnants of extra-axial skeleton. Choanosome. Choanosomal skeletal architecture irregularly or occasionally regularly reticulate, without obvious axial condensation or any axial and extra-axial differentiation. Fibre anastomoses form cavernous meshes, usually ovoid and occasionally rectangular, 450-1500 pm diameter, with a light mesohyl matrix confined to vicinity of fibre junctions. Choanosome may or may not contain pigment granules. Spongin fibres typically heavy, multispicular, entirely cored by oxeas; no obvious differences between fibre size or the number of coring spicules in fibres between peripheral and axial skeletons. Fibres vary in size from 30-320 pm diameter. Echinating acanthostyles heavily and evenly distributed over fibres in peripheral skeleton and at core. Thinner oxeas may be dispersed between fibres, but their presence or absence and abundance varies considerably within an individual sponge. Megascleres (refer to Table 25 for dimensions). Choanosomal oxeas variable in length, usually slender, straight or slightly curved at centre, typically symmetrical, tapering to hastate points which may be sharply pointed or marnilliform. Several authors divide these spicules into 2 length categories, but this division is artificial and not recognised here. Subectosomal extra-axial styles uncommon or rare in some specimens, short and slender, straight or only slightly curved, with rounded non-tylote bases. Ectosomal megascleres absent. Echinating acanthostyles moderately long and slender (although there is considerable variability in spicule length between specimens), club-shaped, with prominent subtylote swelling at base and with slightly swollen or simply rounded blunt tip. Spines concentrated on extremities, whereas near basal swelling there are fewer (or sometimes no) spines. Spines slightly spatulate, recurved and hook like, but lack serrated edges. Microscleres absent. Remarks This species is now relatively well known from the literature under one or another of its various synonyms, the most common being E. bilamellatum. It is a conspicuous member of the Australasian sponge fauna, with an extensive (possibly circum-Australian) distribution, J. N. A. Hooper 1384 Table 25. Comparisons in spicule measurements between specimens of Echinodictyum mesenterinum (Lamarck), showing variability in dimensions related to geographic locality (and latitude) Measurements are given in micrometres, presented as ranges (and means) of lengthxwidth. For material described in this work these measurements were obtained by sampling 25 spicules per specimen for each spicule category Material Choanosomal oxeas Subectosomal styles Ectosomal megascleres Echinating acanthostyles Holotype of Spongia mesenterina 'Australian Seas' MNHN DT568 112-276 x 3-15 (188.2~7.6) 112-318 x 4-7 (222-8~5.2) Absent Lectotype of Spongia bilamellata Great Aust. Bight, S.A. (36's.) (n = 1) 131-334 -~4-15 (193.4~10.6) 181-398 X 3-7 (265.3~4.8) Absent Specimens South-west Coast, W.A. (31"s.) (n = 2) Mid-west coast W.A. (n = 4) (25-26" S) NW. Shelf W.A. (n = 22) (14-20"s) ME. coast QLD (n = 2) (18-23"s) Ashmore Reef, W.A. (12's.) (n = 6) Darwin region N.T. (n = 17) (12's.) Cobourg Pen. N.T. (11"s.) (n = 22) 164-4 19 x4-11 (234.0~6.9) 151-376 x 1.5-9 (214~5~5.2) 92-595 X1-5-15 (244-4~7.4) 192-508 x 2-12 (286.3~6.9) 188-502 ~3-12 (275.5~7.7) 140-474 X 1-18 (254.7~7.4) 11-78 x 1-16 (250-8~7.0) 224402 x 3-6 (318.6~4.3) 154467 X 2-7 (321.4~4.4) 110-483 X 2-7 (286.0~4.4) 208-523 x 3-7 (384.3~4.8) 292-603 x 3-14 (380.8~5.6) 130-425 x 2-7 (290.5 ~ 4 . 4 ) 105-485 x 3-6 (259.0~4.8) Absent Absent Absent Absent Absent Absent Absent and is particularly abundant in the tropics. It is characterised mainly by its growth form and surface features, with few other unique characters, although such features as pigmentation and acanthostyle geometry may also be useful in differentiating E. mesenterinum from other Echinodictyum species. In growth form, E. mesenterinum should be compared to E. vasiplicatum, E. laciniatus (Fig. 84b-d), E. clathrioides (Figs 92-93) and E. costiferum (Fig. 97). This species has slightly swollen terminations on the points of acanthostyles, like E. austrinus (Figs 88-89), E. carlinoides (Fig. 92), E. clathrioides (Figs 93-94), and E. costiferum (Fig. 97). These spicules are heavily ornamented with small recurved spines, but in E. mesenterinum this feature is much more poorly developed than in other species. Echinodictyum mesenterinum is most similar to E. clathrioides, but the two species can usually be easily differentiated on the basis of their spicule dimensions (Tables 23, 25), the presence of a paratangential surface in E. mesenterinum, and in the geometry of spines on acanthostyles (spatulate and serrated in E. clathrioides, slightly spatulate but with sharp edges in E. mesenterinum). This latter feature is especially valuable in separating atypical growth forms of either species (e.g. Figs 94f, 102f). The use of spicule dimensions as a diagnostic character for this species is unreliable, as spicules vary considerably between specimens. To illustrate this variability, specimens Australian Raspailiidae 1385 Fig. 102. Echinodictyum mesenterinum (Lamarck): a, SEM of skeletal structure; b, SEM of fibre characteristics; c, SEM of echinating megascleres; d, SEM of echinating acanthostyle; e, SEM showing swollen bulb-like termination on the distal end of acanthostyles; f , spine morphology. were grouped by geographic locality (Table 25), and ranges of measurements of spicules were compared between each locality (Fig. 103). Both choanosomal oxeas and echinating acanthostyles were slightly larger in specimens from the tropics than those recorded for specimens from temperate latitudes. Extra-axial styles also varied in length between material, but no latitudinal trends for these spicules were apparent. Echinodictyum nidulus Hentschel (Figs 104, 105, 110j; Table 26) Echinodictyurn nidulus Hentschel, 1911: 387. Material Examined Holotype. ZMH 4453 (not seen): Brown Station, Dirk Hartog I., Shark Bay, W.A., 26" OOIS., 113' 12'E., 2-4.5 m depth, 17.vi.1905, coll. W. Michaelsen & R. Hartmeyer (Hamburg Expedition, stn 28, dredge). J. N. A. Hooper 1386 -5 - ,0° 5 - 200 - 5 o, -3a, { : {, 0 .a 0 . I 100- 0 dl, 11 Fig. 103. Latitudinal gradients in spicule sizes, showing the mean lengths of spicules for specimens = 130 0.05 of Echinodictyum mesenterinum (Lamarck) from various localities. r2 = 66.3% Also shown is a linear regression f = 13r18 plot for each spicule type (9) and p <O'Oo8 the result of ANOVA (F statistic and significance level) for each r2 = 74.6% regression line. A, styles; a, oxeas; f = 2053 W, acanthostyles. p <0,003 r* = I5.7010 , , , 15 .- I 20 / 25 30 35 40 Latitude ("S) Other material (all material collected by the author using SCUBA, unless otherwise indicated). Houtman-Abrolhos I., W.A.: NTM 22882: Pelsart Islets, 28" 47.6/ S., 114" 00-7 / E., 20 m depth, 9.vii.1987 (U.S.S.R. RV 'Akademik Oparin', stn HA1). Exmouth Region, W.A.: NCI Q66C-1292-U (fragment NTM 23464): 500 m NW. of Learmonth Jetty, Learmonth, 22" 12. SIS., 114' 06 1'E., 2 m depth, 16.viii.1989, coll. NCI. Darwin Region, N.T.: NTM 2260: Dudley Point, East Point Marine Fish Reserve, 12' 25.0iS., 130'49- liE., intertidal, 31.viii.1981 (sm EP2, by hand). NTM 2307, 312: same locality, 20.ix.1981 (sm EP5). NTM 2160, 171, 180: same locality, 13.xi.1981 (stn EP6). NTM 22126: same locality, 26.ix.1984 (stn EP16). NTM 22420: same locality, 3 m depth, 14.viii.1985 (stn EP24, snorkel). NTM 22539: same locality, intertidal, 4.x.1985 (stn EP25). NTM 22553: same locality, 12.xii.1986, coll. A.M. Mussig & C. Hood (stn EP27). NTM 23204: same locality, 25.ix.1987, coll. N. Smit (stn EP34). NTM 2393: Vestey's Beach, Bullocky Point, 12' 26.2/S., 130' 49.g1E., intertidal, ll.xii.1981 (stn MB4). - Substrate and Depth Range Sand-algae substrate, 2 4 . 5 m depth (holotype); other material from 'cold-water' Acropora coral reefs (Abrolhos I.), tropical fringing coral and rock reefs (Exmouth Gulf and Darwin), and mud covered rock-reef flats (Darwin). Bathymetric distribution in northern Australia intertidal-5 m depth, whereas 1 specimen from Abrolhos I. was collected from 20 m depth, possibly a consequence of warmer sub-surface waters of the Leeuwin Current. Geographical Distribution West coast of Australia (Fig. 104e). Description Shape. Usually small irregular fans, half vases or cups ['bird's nest' of Hentschel (1911)], but occasionally semi-encrusting with digitate surface projections, particularly specimens from cryptic habitats. Size 60-180 mm high, 40-120 mm wide, with relatively thick walls 7-19 mm, mostly with very short stalk, 15-44 mm long, 15-24 mm diameter, attached to substrate by broad holdfast, up to 26 mm diameter. Holotype significantly smaller than other material, 30 mm high, 40 mm transverse diameter, but similar in shape. Colour. Live coloration jet black or black with purple tinge (Munsell 2 -5B 2 -512 or darker) (Fig. 110~3,immediately recognisable in the field; pigmentation identical in preserved material, slightly soluble in ethanol, and upon contact with skin produces a purple mucous which stains skin. Australian Raspailiidae a b Fig. 104. Echinodictyum nidulus Hentschel (specimen NTM 20310): a, choanosomal axial oxeas; b, ectosomal auxiliary style; c, echinating acanthostyles; d , subectosomal extra-axial style; e, known Australian distribution; f , section through peripheral skeleton (scale = 1 mm). Oscula. Large oscula (2-3 - 5 mm diameter) scattered evenly over surface, marginally more common on upper surface and in indentations on surface. In live specimens oscula surrounded by a membraneous lip which protrudes slightly above surface. Inhalant pores undifferentiated from excavated reticulate surface. Texture and surjface characteristics. Surface composed of many small, wart-like, close-set conules, with shaggy or papillose structure. Between conules, running longitudinally along branches, usually occur deeply excavated drainage canals. Surface of flabelliform and vasiform specimens identical on both faces, and entire surface of all material composed of close-set net-like reticulation, formed by compact spongin fibres in periphery. Texture harsh, specimens firm and flexible. Ectosome and subectosome. Unlike all other species of Echinodictyum, this species has a specialised ectosomal skeleton consisting of thin raphidiform oxeas, forming erect brushes on surface usually surrounding tips of protruding choanosomal and subeciosomal spicules. Ectosomal spicules protrude through surface for short distances (150-350 pm), most prevalent on ends of surface conules. Presence of specialised ectosomal skeleton J. N. A. Hooper Fig. 105. Echinodictyum nidulus Hentschel: a, specimen (NTM20260) (scale = 30 mm); b, specimen (NTM 23204) (scale = 30 mm); c, specialised ectosomal skeleton (scale = 500 pm); d, SEM of skeletal structure; e, SEM of fibre characteristics; f, SEM of echinating acanthostyles; g, SEM of spine morphology. not obvious in all sections since they are relatively sparse by comparison with other Raspailiidae. Surface conules interconnected by membraneous skin-like covering, through which no or few spicules protrude but containing abundant pigment granules. Subectosomal skeleton with large extra-axial styles scattered throughout peripheral skeleton, particularly near surface conules, standing perpendicular to surface and protruding through it for some considerable distance (400-1080 pm). Choanosome. Choanosomal skeleton reticulate, with relatively regular ascending and transverse spongin fibres (90-160 pm diameter), and with some suggestion that ascending fibres slightly larger than transverse connecting elements. Fibres fully cored by oxeas, containing only sparse but heavily pigmented (violet-brown) spongin. Fibres moderately heavily echinated by acanthostyles standing perpendicular to fibres, with only a small portion of bases embedded in spongin. Echinating spicules marginally more prevalent on Australian Raspailiidae 1389 peripheral fibres than in axis. Fibre anastomoses produce both small and large oval meshes (110-440 p m diameter), with peripheral fibres usually smaller than those in axis. Meshes lined with cell-like (purple) pigment granules; choanocyte chambers ovoid (90-240 p m diameter). Few megascleres scattered between fibres, and mesohyl containing moderately light but heavily pigmented spongin type B. Megascleres (refer to Table 26 for dimensions). Choanosomal axial oxeas short, moderately stout, slightly curved at centre, tapering to symmetrical sharp points. Subectosomal extra-axial styles long, slender, straight or only very slightly curved near tip, with evenly rounded base, tapering to long and sharply pointed tips. Ectosomal auxiliary styles small, very slender, usually straight, with tapering or evenly rounded base and long sharp point. Echinating acanthostyles relatively long and slender, straight, with slightly subtylote base and very long sharply pointed tip. Spines slender, erect, sharply pointed, evenly dispersed over spicules. Microscleres absent. Table 26. Comparisons in spicule measurements between specimens of Echinodictyum nidulus Hentschel Measurements are given in micrometres, presented as ranges (and means) of lengthxwidth. For material described in this work these measurements were obtained by sampling 25 spicules per specimen for each spicule category Material Choanosomal oxeas Subectosomal styles Ectosomal styles Echinating acanthostyles ZMH 4453* (Shark Bay) HoutmanAbrolhos I. (n = 1) Exmouth Gulf (n = 1) Darwin Region (n = 12) A Specimens 443-1140 ~7-17 (772-0~11.8) 471-1052 x 6-15 (77404~11.2) 543-1545 x 6-15 (1001.3~11.5) From Hentschel (1911: 387). Remarks This species is immediately recognisable in the field by its low fan-shaped or cup-shaped growth form, warty surface sculpturing, jet black coloration, and the production of a purple mucus upon contact with the skin, similar to that of Iotrochota (family Desmacididae). It differs from all other Echinodictyum in possessing both very long subectosomal extra-axial styles and shorter ectosomal styles, which produce an ectosomal skeleton typical of the family but unique to the genus. Although it has not yet been possible to borrow the holotype from the ZMH, there is no doubt that recent material described above is conspecific with E. nidulus. Hentschel's (1911) brief description of the species is quite explicit concerning peculiarities in its growth form, coloration, presence of a specialised ectosomal skeleton, and spicule geometry and size. Hentschel (1911) also compares this species to E. clathratum Dendy (Fig. 83g-i) and E. pykei (Carter) (Fig. 84n-p), which also have unusually long extra-axial styles, but these species differ in other details of spicule geometry, growth form, and absence of a specialised ectosomal skeleton. In this regard E. nidulus should also be compared with E. longistylum Thomas from India, which has the following spicule dimensions: oxeas 1390 J. N. A. Hooper 283-1703 x 3-16 pm, styles 566-4811 x 4-29 pm, acanthostyles 6 7 - 1 0 0 ~6 pm. Hentschel also suggests that E. nidulus and E. mesenterinum are most closely related, showing similarities in growth form and spicule diversity, but these similarities are superficial. Cup-shaped, flabellate and vasiform specimens are known for both species, and in both it is common for the mesohyl to be permeated by purple pigment granules, but geometry and spination of acanthostyles, size of various megascleres and the presence of an ectosomal skeleton in E. nidulus clearly differentiate the species. Echinodictyum nidulus is also very similar in growth form, surface features and coloration to the South African species E. marleyi Burton (Fig. 84h-j), but spiculation is quite different. Echinodictyum rugosum Ridley & Dendy (Figs 106, 107; Table 27) Echinodictyum rugosum Ridley & Dendy, 1886: 477.-Ridley & Dendy, 1887: 165, pl. 32, fig. 1; Topsent, 1892a: 25; Hentschel, 1912: 370. Material Examined Holotype. BMNH 1887.5.2.65: Arafura Sea, SE. of Am I. and off Cape Wessel, N.T., 8" 56/S., 136" 05'E., 98 m depth, 12.ix.1874 (HMS 'Challenger', stn 190). Other material. Indonesia: SMF 1543 (fragment MNHN LBIM DCL2350): Pulu Bambu, Am I., 6" S, 134' 501E., 10 m depth, 03.iv.1908, coll. H. Merton (stn 11). Darwin, N.T.: NTM 21945: Stephen's Rock, Weed Reef, 12' 29.2'S., 130' 47.1 'E., 12 m depth, 27.iv.1984, coll. J.N.A. Hooper (stn WR1). Northwest Shelf, W.A.: NTM 22473: NW. of Amphinome Shoals, 19'04/S., 118" 30.5'E., 76-84 m depth, 01.vi.1985, coll. B.C. Russell & AFZ (stn BCR 8516-NWS38, trawl). Substrate and Depth Range Dead coral rubble, rock reef, shell-grit and sand substrates, 12-98 m depth. Geographical Distribution Arafura and Timor Seas (Fig. 106d). Description Shape. Flattened palmate, planar fans, with or without palmate-digitate or cylindrical digitate modifications to margins, 142-185 mm long, 109-132 mm wide, with digits 27-73 mm long, 8-12 mm diameter, if present, and with long or short cylindrical basal stalks 18-1 10 mm long, 8-22 mm diameter. Colour. Live coloration ranges from orange-brown (Munsell 5YR 7/10) to brown (5YR 5/4), and preserved specimens beige or light brown. Oscula. Infrequent oscula scattered over surface, between surface conules, 1-3 mm diameter, not raised above surface. Inhalant pores minute, 200-500 p m diameter, evenly scattered over surface. Texture and surjflace characteristics. Texture firm, flexible and tough, surface evenly rugose, with small, apically pointed microconules evenly dispersed over surface, 1-3 mm high, 1-2 5 rnm apart; conules interconnected by ridges with a skin-like membrane stretched between them. Ectosome and subectosome. Ectosomal skeleton membraneous, without specialised spiculation, although points of both choanosomal oxeas and extra-axial styles from peripheral fibres may poke through surface, especially on points of conules. Ectosome with heavy concentrations of spongin type B containing heavy deposits of dark brown pigment granules; pigmentation heaviest in ectosomal and subectosomal regions, extending up to 150 pm below surface. Subectosomal skeleton vestigial, consisting only of rare extra-axial styles scattered between fibres and also echinating some fibres, particularly on surface. Choanosome. Skeletal architecture regularly reticulate, with broad fibres, 48-110 pm diameter, containing very little spongin type A, fully cored with stout oxeas and heavily echinated by long slender acanthostyles. Acanthostyles perpendicular to fibres, with only a small part of their bases embedded in fibres. No division of fibres into primary or Australian Raspailiidae 1391 secondary elements; fibres form straight or curved tracts, thickest at nodes of fibre junctions; fibre anastomoses usually form oval meshes, occasionally elongate, 90-360 pm diameter, enclosing oval choanocyte chambers, 28-75 pm diameter. Spongin type B in mesohyl matrix occurs in moderate quantities, but only few pigment granules present in this region. Few megasceleres scattered between fibres. Megascleres (refer to Table 27 for dimensions). Choanosomal oxeas moderately short, always stout, straight or with only slight central curvature, with abruptly tapering, hastate points. Subectosomal extra-axial styles long, slender or moderately thick, slightly curved towards basal end, with evenly rounded bases, tapering to sharp fusiform points. Ectosomal megascleres absent. Echinating acanthostyles relatively long and slender, straight, with slightly subtylote bases and with long tapering tips, ending in sharp points. Spines evenly dispersed over base, shaft and tip of spicules. Spines slender, conical, sharply pointed, and slightly recurved. Microscleres absent. Fig. 106. Echinodictyurn rugosum Ridley & Dendy (specimen NTM 21945): a, choanosomal axial oxeas; b, subectosomal extra-axial styles; c, echinating acanthostyles; d, known Australian distribution. Remarks This species is unusual, and it can be differentiated from all other Echinodictyum, in having relatively thick oxeas and styles, and a palmate or palmate-digitate growth form with a regularly microconulose surface superficially resembling the axinellid sponge Reniochalina stalagmitis Lendenfeld (previously known as Axiamon folium Hallmann). Skeletal preparations of this species are also immediately recognisable by the regular reticulation of very stout oxeas, the heavily echinated fibres, and the entire skeleton appears to be one almost completely dominated by silica with very little fibre (type A) spongin. 1392 J. N. A. Hooper Extra-axial styles are rare in all known specimens of this species, but the occurrence of stout forms is unusual to the genus. Ridley & Dendy (1886, 1887) overlooked their existence entirely, and Hentschel (1912) suggested that there were two varieties: stout and slender forms. In the holotype only slender forms were found, whereas in the other three Fig. 107. Echinodictyum rugosum Ridley & Dendy: a, holotype (BMNH 1887.5.2.65) (scale = 30 mm); b, digitate specimen (NTMZ2473)(scale = 30 mm); c, flabellate specimen (NTMZ1945)(scale = 30 mm); d, e, SEMs of skeletal structure; f , fibre characteristics; g, SEM of echinating acanthostyle; h, SEM of spine morphology. Australian Raspailiidae 1393 Table 27. Comparisons in spicule measurements between specimens of Echinodictyum rugosum Ridley & Dendy Measurements are given in micrometres, presented as ranges (and means) of lengthxwidth. For material described in this work these measurements were obtained by sampling 25 spicules per specimen for each spicule category Material Choanosomal oxeas BMNH 1887. 162-502 X3-18 ( 2 5 0 . 2 ~11.4) Subectosomal styles Ectosomal megascleres Echinating acanthostyles Holotype 5.2.65 316411 ~2.5-5 (357-3~3.6) Absent 93-128 x 8-1 1 (111-6~9.4) Absent 96122 ~7-12 (109-4~9.5) 89-122 x 6-9 (110.0~8.0) 96-136 x 8-12 (116.6~9.9) Specimens SMF 1543 NTM 21945 NTM 22473 149-378 x 5-15 (217.3~9.7) 123-286 ~2-16 (201.1~10.6) 162-345 ~4-19 ( 2 4 6 . 8 ~12.2) 28 1-462 x 3-10 (359.0~5.9) 312402 ~2.5-8 (361.6~4.9) 292-411 ~2.5-7 (355.2~4.7) Absent Absent known specimens both types of spicules were noted, but their division into two categories is artificial and they are not separated here. The dimensions of spicules do not differ substantially between any of the specimens examined (Table 27), but this species has a very restricted geographical and bathymetric range. Zncertae sedis Genus Tethyspira Topsent Tethyspira Topsent, 1890a: 197.-Topsent, 1893: 33; SarB, 1958: 252; Vacelet, 1961: 39; Pulitzer-Finali, 1983: 606. Type species: Tethea spinosa Bowerbank, 1874: 279 (by monotypy; holotype BMNH 1877.5.21.394, from Fowy Harbour, English Channel) (Figs 4w, 70a-b). Diagnosis Encrusting to massive growth forms. Surface hispid, with microconules or papillae. Choanosomal skeleton with condensed basal layer of spongin lying on substrate, with microcionid-like spongin-fibre nodes, and lacking a separate category of choanosomal megascleres. Extra-axial skeleton plumose, with non-anastomosing, multispicular columns of smooth subectosomal styles. Peculiarly formed acanthose spicules are dispersed around (? echinate) choanosomal fibre nodes and extra-axial plumose spicule tracts, and also erect on basal layer of spongin. Spongin minimal and mostly confined to region surrounding spongin-fibre nodes. Ectosome lacks specialised spiculation, but plumose brushes of extra-axial spicules occur on surface. Structural megascleres consist of 1 category of long, smooth styles, and acanthose spicules have very large perpendicular spines. Microscleres absent. Remarks Tethyspira has uncertain affinities with the Raspailiidae. It shows similarities in skeletal structure and spiculation to Raspaciona, and is also reminiscent of Hymeniacidonidae (e.g. Ulosa, Hymeniacidon) (Fig. 70a), although it possesses a plumose skeletal structure on the surface. In fact van Soest et al. (1990) referred the genus to the (revised) order Halichondrida. Unlike Raspaciona, this genus lacks ectosomal specialisation, and the only 1394 J. N. A. Hooper obvious link with the Raspailiidae (e.g. Pulitzer-Finali 1983) or Euryponidae (e.g. Sara 1958; Vacelet 1961) is the presence of acanthose megascleres, which may or may not be truly echinating. The peculiar spination of those megascleres (Fig. 4w) is apparently characteristic of this monotypic genus (e.g. Vacelet 1961). Australian Species None. Genus Sigmeurypon Topsent Sigmeurypon Topsent, 1928: 59. Type species: Microciona fascispiculiferum Carter, 1880: 44 (by original designation and monotypy). Diagnosis Thinly encrusting growth form. Surface even and hispid. Choanosomal skeleton basally condensed uncored spongin fibres lying on the substrate and a thin mesohyl containing raphides and sigmas. Subectosomal skeleton with long smooth styles embedded in basal spongin, standing perpendicular to substrate and protruding through ectosome. Echinating acanthostyles present and erect on substrate, inter-dispersed with choanosomal megascleres. Ectosomal skeleton with protruding styles from choanosome and bundles of raphides lying tangential to surface. Structural megascleres styles and acanthostyles. Microscleres raphides occumng singly or in bundles (trichodragmata). Remarks There is no surviving material of M. fascispiculiferum, since the holotype was destroyed in the LFM during World War 11 (S. Stone, personal communication). Therefore the type species of Sigmeurypon is only known from Carter's (1880) brief original description. It is possible that raphides recorded by him are raphidiform toxas, which are known to occur in several microcionids (e.g. Simpson 1968). Furthermore, it is also possible that 'sigmas' recorded by Carter (1880) in this species are actually examples of bidentate sigmoid isochelae. These possibilities prompted de Laubenfels (1936: 110) to refer this species to his genus Damoseni in the family Microcionidae, but it is not possible to confirm or refute those ideas without checking type material. Australian Species None. Genus Cantabrina Ferrer-Hernandez Cantabrina Ferrer-Hernandez, 1914a: 453.-Ferrer-Hernandez, 1914b: 22. Type species: Cantabrina erecta Ferrer-Hernandez, 1914a: 453 (by monotypy; holotype Madrid, paratype BMNH 1930.1.2 1.9, schizotype MNHN LBIM DCL174L, from Spain). Diagnosis Erect lobate growth form. Surface conulose and hispid. Choanosomal axial skeleton halichondroid, with disorganised criss-cross of long styles. No fibres present, and mesohyl contains relatively light extra-fibre spongin. Subectosomal extra-axial skeleton more-or-less plumose, with long axial styles protruding through surface only on tips of conules. Extra-fibre spongin relatively heavy near periphery. Echinating megascleres rare. Ectosomal skeleton membraneous, without specialised spiculation. Structural megascleres consist of very long styles of single category, with stepped tips and rounded bases, and smooth echinating styles with rhabdose bases. Microscleres absent. Remarks The affinities of this monotypic genus are unknown. Echinating megascleres, claimed to be present in this species (Ferrer-Hernandez 1914a, 1914b) were not observed in the Australian Raspailiidae 1395 BMNH paratype at all, although a spicule slide made from that material in MNHN did show some rare examples of rhabdostyles (Fig. 108~).The skeletal architecture is that of a halichondrid sponge. It is possible that these rhabdostyles are contaminants, in which case the species should be referred to Halichondriidae. Fig. 108. Cantabrina erecta Ferrer-Hernandez: a, paratype (BMNH 1930.1.21.9) (scale = 30 mm); b, section through peripheral skeleton (scale = 1 mm); c, echinating rhabdostyle (scale = 200 pm); d, structural spicules (scale = 500 pm). Australian species None. Discussion Phylogenetic Relationships Throughout this study comparisons have been made between species of Raspailiidae, which are traditionally assigned to the order Axinellida, subclass Tetractinomorpha, and the family Microcionidae in the order Poecilosclerida. This comparison is not casual, and for various reasons discussed here it is proposed to return Raspailiidae to the Poecilosclerida, to which it was originally assigned by Hentschel (1923). The higher systematic relationships of the Microcionidae are relatively straightforward, as demonstrated by the possession of chelae microscleres, which is an apomorphy for the order Poeciloscleria. By comparison, the higher systematics of raspailiids are less obvious and have been debated by several authors (e.g. Ridley & Dendy 1887; Topsent 1894; Dendy 1905; Vosmaer 1912; Wilson 1921). The presence of similarly formed acanthostyles in both families is an obvious clue to their common ancestry, but this feature is interpreted here as representing the retention of an ancestral character state, and, as such, the group cannot be defined solely on this basis. Bergquist noted that there was now (i.e. 1970: 32) substantial evidence from various sources that indicated closer affinities between the two families, but that evidence was still apparently insufficient to prove any definite relationships between the raspailiids and Poecilosclerida. The assignment of Raspailiidae to the order Axinellida rests solely on supposed homologies in skeletal architecture between representative taxa (i.e. the possession of a differentiated axial and extra-axial skeleton and a condensed axial skeleton). However, it is obvious from an examination of all raspailiid genera that the family demonstrates a wide range of 1396 J. N. A. Hooper Table 28. Comparisons of mean spicule dimensions between Australian species of Echinodictyum Measurements are given in micrometres and presented as means (22 standard errors) of lengths and widths Material Choanosomal oxeas Subectosomal styles E. arenosum Absent E. asperum 373.4240.3 4.621.0 E. austrinus, sp. nov. 338.5&100-7 4.321 . 7 E. cancellatum 93-6d7.3 3.4d-9 E. carlinoides 1012-12353.4 10.223.2 E. clathrioides 281.9229.4 4.1kO.8 E. conulosum 403 - 2287.2 4-8d-6 E. costiferum 382-4264.4 3.421.4 E. fruticosumA 279.5 1.0 E. lacunosum 403 - 4229 - 9 3.821.0 E. mesenterinum 303-22104.6 4.721.4 E. nidulus 849.22281.3 11.523.8 E. rugosum 358-3248-4 4-8d.O A Ectosomal styles Echinating acanthostyles Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent 339.1265.2 3.6k1.2 Absent Median values. Fig. 109. a, Raspailia vestigifera Dendy (specimen NTM Z2534), Coral Bay, Cobourg Peninsula, N.T. (photo J.E.N. Veron); b, Raspailia phakellopsis, sp. nov. (paratype NTM Z2158), Weed Reef, Darwin, N.T. (photo author); c, Raspailia darninensis, sp. nov. (holotype NTM Z2239), East Point, Darwin, N.T. (photo author); d, Raspailia nuda Hentschel (specimen NTM Z2394), East Point, Darwin, N.T. (photo author); e, Ectyoplasia tabula (Lamarck) (specimen NCI Q66C-1341-X), Muiron I., Exmouth Gulf, W.A. (photo NCI); f, Ectyoplasia tabula (Lamarck) (specimen NTM Z2961), Dirk Hartog I., Shark Bay, W.A. (photo author); g, Ectyoplasia vannus, sp. nov. (specimen NCI Q66C-1339-V), Muiron I., Exmouth Gulf, W.A. (photo NCI); h, Trikentrion flabelliforme Hentschel (specimen NCI Q66C-0530-Q), Pany Shoals, Timor Sea, N.T. (photo NCI); i, Trikentrion flabelliforme Hentschel (specimen NTM Z2705), East Point, Darwin, N.T. (photo author);j, Amphinomia sulphurea, gen.sp. nov. (specimen NCI Q66C-0558-V), Parry Shoals, Timor Sea, N.T. (photo NCI). Australian Raspailiidae 1397 J. N. A. Hooper 1398 Table 29. Summary of characters used to judge apomorphy in the construction of the cladogram of relationships between genera of Raspailiidae based on outgroup comparisons with members of the families Microcionidae, Myxillidae and Axinellidae (Fig. 111) List of Characters and Character States Structure of the choanosomal skeleton Dense axial compression of criss-crossed spicules, without axial fibres Axially condensed reticulation of spongin fibres and/or spicule tracts Slightly axially condensed reticulation of spongin fibres and/or spicule tracts Basally condensed skeleton lying on the substrate Reticulation of spongin fibres and/or spicule tracts, without any trace of axial compression Loosely aggregated or plumose axial fibres Differentiation of the axial and extra-axial skeletons Well developed Slight Vestigial, with only the rudiments of an extra-axial skeleton or individual spicules present Absent, without any trace of an extra-axial skeleton Architecture of the axial skeleton Renieroid reticulation of fibres and/or spicules Regular or irregular reticulation of fibres and/or spicules Plumo-reticulate structure, with differentiated primary and secondary tracts Condensed reticulate core with plumose peripheral tracts Plumose, non-anastomosing fibres Reduced hymedesmoid-microcionid skeleton, with a basal layer of spongin lying on the substrate Megascleres in the axial skeleton Entirely smooth choanosomal styles, subtylostyles or anisoxeas Entirely smooth choanosomal rhabdostyles Choanosomal styles or subtylostyles with microspined bases Choanosomal styles or subtylostyles with spines on the basal and distal ends Acanthose choanosomal rhabdostyles Sinuous styles or anisoxeas Choanosomal oxeas Acathostrongyles or acanthotylostrongyles Both spined oxeas and smooth anisoxeas Absent Architecture of the extra-axial skeleton Renieroid reticulate Regularly or irregularly reticulate Plumo-reticulate Plumose, or very slightly plumo-reticulate Radial, with slightly diverging plumose spicules Radial, with only few spicules perpendicular to the axis Megascleres of the extra-axial skeleton Smooth or basally spined subectosomal styles, subtylostyles or anisoxeas Smooth or basally spined subectosomal strongyles or oxeas choanosomal megascleres form extra-axial tracts, and near the periphery setaceous subectosomal styles are embedded in peripheral fibres Absent, but extra-axial tracts are composed of choanosomal megascleres Choanosomal megascleres produce extra-axial tracts but vestigial extra-axial styles are dispersed throughout the mesohyl Structure of the ectosomal skeleton, and presence or absence of ectosomal specialisation Specialialised spiculation consisting of ectosomal styles or anisoxeas forming brushes around protruding megascleres from the extra-axial skeleton Specialialised spiculation consisting of ectosomal oxeas or strongyles forming brushes around protruding megascleres from the extra-axial skeleton Structure typical of the family, but ectosomal spicules are undifferentiated from choanosomal spicules, and these form plumose brushes surrounding the protruding extra-axial megascleres Specialised spiculation of ectosomal megascleres, but these form a continuous or discontinuous palisade perpendicular to the surface and they do not appear to be associated with the protruding extra-axial megascleres Australian Raspailiidae 1399 Table 29. Continued List of Characters and Character States Specialised spiculation of ectosomal megascleres, but these are scattered throughout the ectosome and subectosomal regions, and these may lie tangential, paratangential or perpendicular to the surface Protruding extra-axial megascleres which are surrounded by bundles of raphides dispersed over the ectosome Acanthostyles forming plumose brushes around the protruding extra-axial megascleres Lacks a specialised skeleton, although subdermal megascleres (including choanosomal and subectosomal spicules) protrude through the surface Lacks any sort of specialised spicule skeleton and is simply membraneous Geometry of echinating megascleres Microcionid-like, club-shaped, with rounded or sharp points, subtylote bases, and with evenly or unevenly distributed spines Acanthose, club-shaped, but with strongly curved hooks on the shaft Acanthose, club-shaped or strongylote, with strongly curved hooks on the base and shaft (cladotylote) Acanthose, club-shaped, with large spines lying perpendicular to the shaft Acanthostyles with smooth or acanthose rhabdose bases Acanthostyles with smooth rhabdose bases, and large recurved spines are distributed over the shaft Acanthostyles with bulbous tylote bases, with or without acanthose points and other modifications to the distal portion Club-shaped with clavulate points Absent, although acanthose choanosomal rhabdostyles echinate skeletal tracts Absent, but diactinal acanthorhabds are dispersed throughout the skeleton Sagittal monact-, diact- or tetractinal spicules (acanthoplagiotriaenes) with only one spined ray Sagittal tetract- or pentactinal spicules (rarely with fewer rays) (acanthoplagiotriaenes) with all or most rays spined Absent Geometry of spines on acanthostyles Straight, perpendicular to the shaft, cylindrical, sharply pointed Recurved, conical, sharply pointed Recurved, spatulate, with sharp or even edges Recurved, spatulate, with serrated edges Absent Distribution of echinating megascleres Not confined to any particular region in the skeleton, and are relatively evenly dispersed over skeletal tracts Confined to a particular region of the skeleton, at the junction of axial (or basal) and extra-axial skeletons Confined to a particular region of the skeleton, and form plumose brushes on the ectosome, surrounding protruding extra-axial spicules Confined to a particular region of the skeleton, and are concentrated mainly near the peripheral skeleton Forming a rigid interlocking skeleton Absent Microscleres Absent Raphides (or microxeas), occurring singly or in bundles (trichodragmata) Chelae and/or toxas architectural types, and many of these structures show only vague or no similarities with axinellid forms at all. In fact the reverse is true: many skeletal types in the family clearly show structural relationships with other poecilosclerids. For example, of 26 Raspailia species described here, 58% lacked axinellid skeletal structure, having instead reticulate, plumo-reticulate and plumose skeletons. The raspailiid genus Aulospongus is peculiar in 1400 J. N. A. Hooper Australian Raspailiidae 1401 having unusual tubular, plumose, skeletal fibre bundles, whereas identical fibres are known for the Clathria 'parthena' group of Microcionidae species. Similarly, raspailiids with typical axinellid skeletal structure [e.g. Raspailia (Raspailia) vestigifera] have structural counterparts in the Microcionidae (e.g. Clathria (Axociella) canaliculata, with a compressed axis, plumose-radial extra-axis and isochelae microscleres) (Hooper, unpublished data). The diversity of raspailiid skeletons is further illustrated below in the analysis of genera. Several other homologies between Raspailiidae and Microcionidae are also apparent. Incorporation of detritus into spicule-bearing fibres is a common trait amongst the Poecilosclerida [e.g. Clathria (Clathriopsamma) spp.], and it is also recorded here for several Raspailiidae [e.g. Raspailia (Clathriodendron) spp., Echinodictyum arenosum]. Several raspailiids have completely lost echinating megascleres [e.g. Raspailia (Syringella) spp., Ceratopsion spp.], and this is also a feature of some microcionids [e.g. Clathria (Axociella) spp.]. One raspailiid (Aulospongiella monticularis) incorporates echinating acanthostyles secondarily into fibres, which is also a feature peculiar to the rnicrocionid Clathria 'phorbasiformis' species group (Hooper, unpublished data). Some raspailiids lack the specialised ectosomal structure characteristic of the family (e.g. Raspailia australiensis), and have instead a continuous spicule crust typical of one group of Microcionidae [Clathria (Thalysias)]. Other examples are listed in the systematic section of this study. Thus, morphological comparisons made between raspailiids and other groups throughout this study indicate that Raspailiidae is more closely related to Poecilosclerida, such as Microcionidae, than they are to axinellids, such as Hemiasterellidae and Axinellidae. Preliminary discussions on this relationship have been presented earlier (Hooper 1988b, 1990a), but in the present study the Raspailiidae are formally defined in the order Poecilosclerida. This conclusion is supported by biochemical evidence (general protein electrophoresis, carotenoid pigments, free amino acids) (Hooper 1990a; Hooper et al. 1992). Postulated relationships between genera of Raspailiidae were investigated using a numerical computer method for inferring phylogenies (PAUP, Swofford 1985). This analysis uses the Wagner method, taking the preferred phylogenetic tree as the most parsimonious one, i.e. the one with the fewest number of evolutionary steps. Data used in this analysis were derived from an unordered multistate character set, and apomorphy was judged by outgroup comparisons. These are surnrnarised in Table 29. Outgroups chosen were from both closely and more distantly related taxa in the families Microcionidae (Clathria, Thalysias), Myxillidae (Acarnus) and Axinellidae (Reniochalina). A consensus tree, produced from 26 minimum-length trees, is depicted in Fig. 111. Three major groups of genera emerge, as indicated by the transformation series of skeletal architectural types. These groups (labelled 1 to 3 in Fig. 111) show a range of skeletal structures extending from an axially condensed, axinellid-like construction to a strictly reticulate, myxillid-like condition. Six subgroups are also evident, as indicated by the geometry of megascleres composing their axial, extra-axial and echinating skeletons. These groups are delineated and named as follows: raspailoids (Raspailia, Ectyoplasia and Endectyon); axinelloids (Ceratopsion, Axechina i d Thrinacophora); triaenoids (Trikentrion and Cyamon); hymedesmoid-microcionoids (Aulospongus, Raspaciona, Eurypon, Rhabdeurypon and Hymeraphia); myxilloids (Amphinomia and Echinodictyum); and plocamoids (Lithoplocamia and Plocamione). Synapomorphy for all these groups is the possession of a specialised (raspailiid) ectosomal skeleton, and synplesiomorphy is Fig. 110. a, Amphinornia sulphurea, gen. et sp. nov. (holotype NTM Z1787), Amphinome Shoals, Northwest Shelf, W.A. (photo T. Ward); b, Ceratopsion montebelloensis, sp. nov. (holotype NCI Q66C-1528-A),Monte Bello Is, W.A. (photo NCI); c, Echinodictyum asperum Ridley & Dendy (specimen NTM Z1853), Port Hedland region, Northwest Shelf, W.A. (photo T. Ward); d, Echinodictyum austrinus, sp. nov. (paratype NCI Q66C-2627-V), Perth region, W.A. (photo NCI); e, Echinodictyum cancellaturn (Lamarck) (specimen NCI Q66C-0582-W), East Point, Darwin, N.T. (photo NCI); f, Echinodictyurn clathrioides Hentschel (specimen NCI Q66C-2922-R), Bussellton, Geographe Bay, W.A. (photo NCI); g, Echinodictyum clathrioides Hentschel (specimen NCI Q66C-1293-V), Learmonth, Exmouth Gulf, W.A. (photo NCI); h, Echinodictyum conulosurn Kieschnick (specimen NCI Q66C-1619-Z), Dampier Archipelago, Northwest Shelf, W.A. (photo NCI); i, Echinodictyurn mesenterinum (Larnarck) (specimen NCI Q66C-0387), Orpheus I., Great Barrier Reef, Qld (photo NCI); j, Echinodictyum nidulus Hentschel (specimen NTM Z0312), East Point, Darwin, N.T. (photo author). 1402 J. N. A. Hooper the possession of echinating acanthostyles: not all taxa possess these features, but where absent their obvious affinities to Raspailiidae are indicated by one or more other features. Thus, the demonstrated heterogeneity in architectural types amongst genera of Raspailiidae indicates that skeletal structure cannot be regarded as the primary diagnostic character for the family, although it is useful in differentiating groups of genera amongst the family. Raspailia 5~ 1 6d, 8h la-c 3 bld Ectyoplasia Endectyon Ceratopsion Axechina Thrhacophora Trikentrion 8 ~IOCI , l a - 1l b Ic 8k11 Jf, 3 e , 6c1,8f lob[ lf3e95ds 8a 3a,e/f 7a 2b 3 1 d f 8d ,.............- .4 I 7b98~99a 2c,5fg8g lld,3f Cyamon AU ~ O S ~ O Raspaciona 11b ~ ~ U S Eurypon Tethyspira? Rhabdeurypon Hymeraphia Amphinomia Echinodictyum Lithoplocamia L Plocamione Fig. 111. Cladogram of the hypothesised relationships between genera of Raspailiidae based on computer-generated phylogenetic analysis using parsimony (PAUP), with indication of three major groups and six subgroups. Each number and letter on the cladogram corresponds to a character and character state respectively, listed in Table 29, and represents an evolutionary change from a relatively plesiomorphic to a relatively apomorphic state. The tree is derived from a consensus of 26 possible minimum length trees, based on an unweighted, unordered, multistate character set (consensus information = 0.373). with pleisiomorphy determined by outgroup comparisons (outgroups = Microcionidae, Myxillidae, Axinellidae; consistency index = 0.671; tree length = 73 .0). Biogeography of Species The delineation of the major Australian marine zoogeographic zones used in this study follows the scheme proposed by Bennett & Pope (1953), modified from Whitley (1932), which divides continental Australia into five provinces. These include two tropical zones (Dampierian, Solanderian), two warm temperate zones Windersian, Peronian) and one cool temperate region (Maugean). Of course, actual sponge distributions are unlikely to coincide exactly with the proposed boundaries between each of these provinces, since these boundaries are known to vary considerably between the major phyla. Nevertheless, this scheme does provide a convenient nomenclature on which to base zoogeographical analyses, and it provides a reasonable level of accuracy in delineating adjacent provinces based on both physical and biological factors. However, it is likely that a more general model, such as that discussed by Wilson & Allen (1987: fig. 3.13), provides a more Australian Raspailiidae 1403 realistic interpretation of the major influences on marine faunal distributions. That model loosely corresponds with the five zoogeographic zones described above, but it recognises that there are major areas of overlap between zones, as well as more restricted areas of endemism. More importantly, the scheme also incorporates more recent data on the patterns of oceanic and coastal water circulation. This model notes that for many invertebrate groups the northern tropical fauna is greatly influenced by the general Indc-West Pacific input (both genetic and physical), whereas it suggests that the southern temperate area has a much higher level of endemism, having brought much of its sponge fauna with it during the continent's migration from the Cretaceous austral landmass. For the Porifera this pattern is certainly demonstrated by some groups, such as the preponderance of arenaceous species in the southern provinces (Wiedenmayer 1989), but our knowledge of Australian sponge taxonomy is still so rudimentary, and only a small proportion of the fauna has yet been described (estimated at less than 20%; Hooper, in press), that it is not possible to generalise on Poriferan biogeography with any great degree of confidence. The following analysis is based on a restricted fauna of Raspailiidae described in this study. The literature implies that there is only a low diversity of raspailiid species in Australian waters, but from recent extensive collections along the mid-west and north-west coasts (i.e. Dampierian province, Shark Bay region, W.A. to Cape York, Qld) it is shown that the family is very prevalent in this region (38 species) (Table 30). Of these species, 15 are not found elsewhere, and some of these may be true endemics. Further collections along the north-east coast, especially in the inter-reef regions of the Great Barrier Reef, which have high sediment loads and large tides comparable with conditions of the north-west coast, may show a higher diversity than is presently known for that region. Presently, only 13 species are known to occur in the Solanderian province, of which five are apparently endemic. This relatively low number is despite examination of extensive collections held by NCI and QM from the inter-reef and other Queensland regions. If these collections are accurate representations of the north-east coast raspailiid fauna, then the number of species on the north-west coast is abnormally high. For sponges, this conclusion has other empirical support. Ltvi (1979 and personal communication) remarks that MUSORSTOM collections and the literature clearly demonstrate that the southern Indonesian-northern Australian region has the highest diversity of all Indo-Pacific biogeographical provinces, and is the centre of dispersal for Indo-Pacific sponges. He notes that sponge faunas diminish in diversity away from this centre. By comparison, in temperate waters a relatively low diversity of raspailiids is indicated. The south-eastern coast Peronian province has 10 species, five of which are endemic; the cool temperate Maugean province in southern Victoria, Bass Strait and Tasmania has only seven species, of which three are endemic; and the Flindersian province, extending from western Victoria to the mid-south coast of W.A., also has only seven species, three of which are endemic. This low diversity in temperate waters is probably truly representative of the fauna, rather than merely an artifact of sampling effort by the author. A recent, extensive study of the Maugean benthos provided an estimate of the total number of sponge species for that region at approximately 650 species (Wiedenmayer 1985), which is comparable with the north-west coast of Australia (approximately 800 species: Hooper, unpublished data), but only one raspailiid species was recorded for the southern region (Wiedenmayer 1989). Antarctic-subantarctic waters have an even lower diversity of raspailiids, with only one species recorded in the literature. Thus, similar to most other marine benthic invertebrate groups, the diversity of raspailiid species is lowest in subantarctic waters and increases substantially towards the tropics. However, this trend is not universal for all sponge groups (e.g. Desmacididae, Wiedenmayer 1989). Of the 56 species of Raspailiidae recorded in Australasian waters, only 15 have been described outside this region. This represents only 27% of the fauna, or, in other words, a level of endemism for the family of nearly 73%. However, it is unlikely that this figure is completely accurate as it is undoubtedly biased by the poorer knowledge of the sponge fauna outside the range of the present study. Of these regions in the vicinity of Australia, the raspailiid fauna of New Zealand is the best known, consisting of nine J. N. A. Hooper 1404 Table 30. Zoogeographical distribution of Raspailiidae species in Australian marine provinces (after Bennett & Pope 1957), and their known distribution elsewhere Key to symbols: ++, known only from one province, possibly endemic; +, present in more than one province; -, absent from a particular province Species 1 Zoogeographic region 2 3 4 5 Other known Distribution Genus Raspailia, subgenus Raspailia R. atropurpurea R. echinata R. gracilis R. phakellopsis R. pinnatifida R. tenella R. vestigifera R. wilkinsoni - - - - + ++ ++ - ++ + - ++ - - +? Subgenus Clathriodendron ++ R. arbuscula R. bifurcata R. cacticutis R. dunvinensis R. desmoxyiformis R. keriontria R. melanorhops R. paradoxa - R. compressa R. frondula R. reticulata R. wardi - + - - + + + - - - - ++ - - Subgenus Syringella R. R. R. R. R. australiensis clathrata elegans nuda stelliderma R. irregularis E. frondosa E. tabula E. vannus + ++ - - - - - - - + - - - Genus Endectyon - T.flabelliforme - - Genus Trikentrion - - - Genus Cyamon - New Zealand - ++ Antarctica - - Indonesia, Thailand East coast of India ? - Indonesia - Indonesia Genus Eurypon Genus Amphinomia E. graphidiophora - A. sulphurea - - Genus Ceratopsion C. dichotoma C. axifera C. montebelloensis C. palmata - Subgenus Hymeraphiopsis Genus Ectyoplasia ? E. elyakovi E. fruticosa aruensis E. thurstoni E. xerampelina C. aruense + - Subgenus Raspaxilla - ++ New Zealand - - ++ - - - - Indonesia Australian Raspailiidae 1405 Table 30. Continued Species Zoogeographic region 1 - T. cervicornis A. raspailioides E. arenosum E. asperum 2 3 4 5 Other known Distribution - - + Genus Thrinacoahora - - Indonesia, Philippines Genus Axechina - - Indonesia Genus Echinodictyum - ++ - E. austrinus E. cancellatum E. carlinoides E. clathrioides E. clathrioides E, costjferum E. fruticosum E. lacunosum E. mesenterinum E. nidulus E. rugosum Indo-Pacific, from Tahiti to Gulf of Manaar Indonesia Indonesia Philippines Indonesia 1. Dampierian province, Geraldton, W.A. to Cape York, Qld. 2. Solanderian province, Cape York, Qld. to Coffs Harbour, N.S.W. 3. Peronian province, Coffs Harbour, N.S.W. to shallow coastal regions of northern Vic., and deeper waters off NE. Tasmania. 4. Maugean province, Bass Strait and shallow waters of Tasmania. 5. Flindersian province, western Vic. to Geraldton, W.A. species, of which only two also occur in Australian waters (both surprisingly in the tropical Australian fauna). Only four species of Raspailiidae have been described so far for New Caledonian waters, but three of these are indicated as being endemic to that region ( U v i 1967; U v i & LCvi 1983). At least two more undescribed raspailiids are known to occur there (Hooper, unpublished data). By comparison, the sponge fauna of Papua New Guinea is poorly documented, with only about 30 species described from the entire region (Bowerbank 1877; Kelly Borges & Bergquist 1988). No Raspailiidae have yet been reported but Echinodictyum asperum has recently been found near Madang (NTM collection). From the few publications on the Indonesian sponge fauna and from the large collections of unpublished material from that region in the ZMA and MNHN (Siboga, Snellius II and MUSORSTOM expeditions), approximately 30 species of raspailiids are indicated for the archipelago (R. van Soest and C. Lkvi, personal communication), of which only about five are endemic, but ten of these also occur in northern Australian waters (Table 30). The Indian Ocean Christmas I. fauna is relatively well known, given the small size of its reefs, consisting of 33 species (Kirkpatrick 1900, 1910), but none of these are raspailiids. Further north, to the western side of the Malay Peninsula extending into the Andaman Sea and southern Bay of Bengal we know of only 45 species of sponges (Carter 1886b; Thomas 1977), two of which are raspailiids, and one of these is endemic. Recent unpublished collections from this region by the author confirm this surprisingly low level of diversity for the Andaman Sea. Levels of apparent endemism within the Indo-West Pacific raspailiid fauna are indicated in Fig. 112, but this analysis is certainly preliminary, and species diversity in the Indonesian archipelago in particular is expected to be much richer than presently known. From an analysis of the number of species common to Australian waters and also occurring in neighbouring regions, the affinities of the tropical raspailiid fauna lie mainly with the Indo-Malay region, as predicted by the biogeographic model presented by Wilson & Allen (1987, fig. 3.13), whereas species found in more temperate waters are not generally found elsewhere. The number of species common to both the Indonesian and northern J. N. A. Hooper 1406 & F G . ~ ~ ~'y5' Subantarctic f 3 # 12O0 I , I 165' Fig. 112. Regional variation in the levels of apparent endemism for the family Raspailiidae within the Indo-West Pacific. Open blocks represent the number of raspailiid species known to occur in each region; solid blocks represent the number of raspailiid species which are indicated as being endemic (a presumption based mostly on records from the literature). Australian regions (i.e. ten) will probably increase as our knowledge of the Indonesian fauna increases, whereas the level of endernism indicated for southern Australian waters may be a more accurate representation of the fauna. Only one species (Echinodictyum asperum) is widely distributed throughout the Indo-Pacific, from Tahiti in the east, throughout the Indo-Malay Archipelago, to the Gulf of Manaar in the west. Acknowledgments Completion of various aspects of this study would not have been possible without the generous assistance of a number of people and organisations. I thank Patricia Bergquist (University of Auckland) for her encouragement and assistance at the commencement of this project, and Felix Wiedenmayer (NM Basel) for giving me access to his various notes, unpublished manuscripts, photographs and collections of type fragments from various eastern European Museums (lodged with the NMV sponge archives). I am most grateful to Claude U v i (MNHN Paris) for providing a post-doctoral fellowship at the MNHN and giving me access to its type collections, including all the Lamarck material, and Shirley Stone (BMNH London) for providing access to and assistance with the vast BMNH collections. Similarly, I acknowledge financial assistance from the Sir Winston Churchill Australian Raspailiidae 1407 Memorial Trust and an Australian Biological Resources Study grant, which provided access to European museum collections. For their help with the taxonomic detective work, in tracking down type material, providing additional material from localities which would otherwise be inaccessible to me, or in getting me to otherwise inaccessible localities, and for other information cited in the text I thank the following people: Joe Baker (AIMS, Townsville), Lester Cannon (QM, Brisbane), Frank Climo (NMNZ, Wellington), Ruth Desqueyroux-Faundez (MHN, Geneva, and LMJG, collections), George Elyakov (PIBOC, Vladivostok), Manfred Grasshoff (SMF, Senckenberg), the late Takomura Hoshino (MMBS, Hiroshima), Frank von Knorre (PM, Jena), Dieter Kiihlmann (ZM, Berlin), Vladimir Krasochin (PIBOC, Vladivostok), Claude U v i (MNHN, Paris), C.C. Lu (NMV, Melbourne), A.K. Mandal (IM, Calcutta), Loisette Marsh (WAM, Perth), Peter Murphy and associates (NCI, Townsville), Shane Parker and Wolfgang Zeidler (SAM, Adelaide), Urs Rahm (NM, Basel), Frank Rowe, Bill Rudman and Penny Barents (AM, Sydney), Klaus Ruetzler (USNM, Washington), Rob van Soest (ZM, Amsterdam), Shirley Stone (BMNH, London), B.R. Stuckenberg (NM, Natal), Jean Vacelet (SME, Marseille), J.E.N. (Charlie) Veron (AIMS Townsville), Trevor Ward (CSIRO, Sydney), Clive Wilkinson (AIMS, Townsville), the Masters, crew and scientific staff accompanying the research expeditions of the CSIRO RV 'Soela', RV 'Sprightly', U.S.S.R. RV 'Akademik Oparin', Cootamundra Shoals Survey team (UK), FV 'Rachel' (Darwin) and FV 'Skeleton' (Darwin). I am grateful to Jodie Baxter, Cathie Johnston, Anne-Marie Mussig and Rex Williams for their technical assistance at various times, and I particularly thank Daniel Low Choy for his assistance in the field, tireless preparation of material for histology and photographic assistance. References Anonymous, 1985. 'International Code of Zoological Nomenclature.' International Trust for Zoological Nomenclature. (H. Charlesworth & Co.: Huddersfield.) 3rd edn, 338 pp. Ayling, A.L., S. Stone & B.J. Smith, 1982. Catalogue of types of sponge species from southern Australia described by Arthur Dendy. Reports of the National Museum of Victoria 1, 87-109. Bennett, I. & E. Pope, 1953. Intertidal zonation of the exposed rocky shores of Victoria, together with a rearrangement of the biogeographical provinces of temperate Australian shores. Australian Journal of Marine and Freshwater Research 4, 105-59. Bergquist, P.R., 1961. Demospongiae (Porifera) of the Chatham Is and Chatham Rise, collected by the Chatham Is 1954 Expedition. New Zealand Department of Scientific and Industrial Research Bulletin 139. Biological Results of the Chatham Is 1954 Expedition (5), 169-206. Bergquist, P.R., 1970. The marine fauna of New Zealand: Porifera, Demospongiae, Part 2. (Axinellida and Halichondrida). New Zealand Department of Scientific and Industrial Research Bulletin.New Zealand Oceanographic Institute Memoir 197, 1-85, pls 1-20. Bergquist, P.R., 1978. 'Sponges.' (Hutchinson: London.) 268 pp. Bergquist, P.R., 1980. A revision of the supraspecific classification of the orders Dictyoceratida, Dendroceratida and Verongida (class Demospongiae). New Zealand Journal of Zoology 7, 443-503. Bergquist, P.R. & J. Fromont, 1988. The marine fauna of New Zealand: Porifera, Demospongiae, Part 4 (Poecilosclerida). New Zealand Oceanographic Institute Memoir (96), 1-197, pls 1-57. Boury-Esnault, N. & M. van Beveren, 1982. Les DBmosponges du plateau continental de KerguelenHeard. Territoire des Ten-es Australes et Antarctiques Fran~aises.Cornit6 National Francaise des Recherches Antarctiques (52), 1-132, pls 1-21. Boury-Esnault, N. & M.T. Lopes, 1985. Les DBmosponges littorales de 1'Archipel des A~ores.Annales de I'lnstitut Oc6anographique, Paris 61(2), 149-225, pls 1-5. Bowerbank, J.S., 1862. On the anatomy and physiology of the Spongiadae. Part 111: On the generic characters, the specific characters and the method of examination. Philosophical Transactions of the Royal Society, London 152, 1087-135, pls 72-74. Bowerbank, J.S., 1864. 'A Monograph of the British Spongiadae.' (Ray Society: London.) Vol. 1, 290 pp., 37 pls. Bowerbank, J.S., 1866. 'A Monograph of the British Spongiadae.' (Ray Society: London.) Vol. 2, 388 pp. 1408 J. N. A. Hooper Bowerbank, J.S., 1873. Report on a collection of sponges found at Ceylon by E.W.H. Holdworth Esq. Proceedings of the Zoological Society of London (1873), 25-31, pls 5-7. Bowerbank, J.S., 1874. 'A Monograph of the British Spongiadae.' (Ray Society: London.) Vol. 3, 360 pp. Bowerbank, J.S., 1877. Description of five new species of sponges discovered by A.B. Meyer on the Philippine Is and New Guinea. Proceedings of the Zoological Society of London 1877, 456-64. Brondsted, H.V., 1924. Papers from Dr Th. Mortensen's Pacific Expedition 1914-16. 23. Sponges from New Zealand. Part 1. Videnskabelige Meddelelser fra Dansk naturhistorisk Forening 77, 435-83. Brondsted, H.V., 1929. Neue Schwamme aus Amoy an der FormosaStrasse. Zoologischer Anzeiger, Leipzig 81, 224-7. Brondsted, H.V., 1934. Resultats Scientifiques du voyage aux Indies Orientales NeBrlandaises. Sponges. Brusseles Institut royal des sciences naturefles de Belgique. Memoires 2(15), 3-26. Burton, M., 1928. Report on some deep-sea sponges from the Indian Museum collected by the R.I.M.S. 'Investigator'. Part 11. Tetraxonida (concluded) and Euceratosa. Records of the Indian Museum, Calcutta 30(1), 109-38, pls 1-2. Burton, M., 1930. Additions to the sponge fauna of Plymouth. Journal of the Marine Biological Association of Plymouth 16, 489507, Burton, M., 1931. On a collection of marine sponges mostly from the Natal coast. Annals of the Natal Museum 6(3), 337-58, pl. 23. Burton, M., 1932. Sponges. Vol. 6, pp. 237-392, pls 48-57 in: 'Discovery Reports'. (Cambridge University Press: Cambridge.) Burton, M., 1934~.Sponges. Vol. 4(14), pp. 513-621, pls 1-2 in: 'Scientific Reports of the Great Barrier Reef Expedition 1928-29'. British Museum (Natural History): London.) Burton, M., 1934b. Sponges. Vol. 3(2), pp. 1-58, pls 1-8 in: 'Further Zoological Results of the Swedish Antarctic Expedition 1901-1903'. Stockholm: Norstedt & S6ner.) Burton, M., 1935. The family Plocamiidae with descriptions of four new genera of sponges. Annals and Magazine of Natural History (15) 87, 399-404. Burton, M., 1938. Supplement to the littoral fauna of Krusadai I. in the Gulf of Manaar. Porifera. Bulletin of the Madras Government Museum ( n . ~ . )Natural , History Section 1(2), 1-58, pls 1-9. Burton, M., 1956. The sponges of west Africa. No. 4 pp. 111-47 in: 'Atlantide Report. Scientific Results of the Danish Expedition to the Coasts of Tropical West Africa 1945-46'. Danish Science Press: Copenhagen.) Burton, M., 1959. Sponges. Vol. 10(5), pp. 151-281 in: 'Scientific Reports of the John Murray Expedition 1933-34'. (British Museum (Natural History): London.) Burton, M. & H.S. Rao, 1932. Report on the shallow-water marine sponges in the collection of the Indian Museum. Records of the Indian Museum 34(3), 299-356, pl. 18. Cabioch, L., 1968a. Contribution B la connaissance de la fame des Spongiaires de la Manche occidentale. Demosponges de la region de Roscoff. (Travaux de la Station Biologique de Roscoff. Nouvelle SBrie 18.) Cahiers de Biologie Marine 9, 211-46. Cabioch, L., 19686. Contribution B la connaissance des peuplements benthiques de la Manche Occidentale. (Travaux de la Station Biologique de Roscoff. Nouvelle Sene 18). Cahiers de Biologie Marine 9, 493-720. Capon, R.J., J.K. MacLeod & P.J. Scammells, 1986. The Trikentrins: novel indoles from the sponge Trikentrionjabelliforme. Tetrahedron 42(23), 6545-50. Carter, H.J., 1876. Descriptions and figures of deep-sea sponges and their spicules, from the Atlantic Ocean, dredged up on board H.M.S. 'Porcupine', chiefly in 1862 (concluded). Annals and Magazine of Natural History (4) 18, 226-40, 307-24, 388-410, 458-73, pls 12-16. Carter, H.J., 1879a. On Holostrella, a fossil sponge of the Carboniferous Era, and on Hemiastrella, a new genus of recent sponges. Annals and Magazine of Natural History (5) 3, 141-50, pl. 21. Carter, H.J., 1879b. Contributions to our knowledge of the Spongida. Annals and Magazine of Natural History (5) 3, 284-304, 343-60, pls 25-27. Carter, H.J., 1880. Report on specimens dredged up from the Gulf of Manaar and presented to the Liverpool Free Museum by Capt. W.H. Cawne Warren. Annals and Magazine of Natural History (5) 6, 35-61, 129-56, pls 4-6. Carter, H.J., 1882. New sponges, observations on old ones, and a proposed new group. Annals and Magazine of Natural History (5) 10, 106-25. Carter, H.J., 1885. Descriptions of sponges from the neighbourhood of Port Phillip Heads, South Australia, continued. Annals and Magazine of Natural History (5) 16, 347-68. Australian Raspailiidae 1409 Carter, H.J., 1886~.Supplement to the descriptions of Mr. J. Bracebridge Wilson's Australian Sponges. Annals and Magazine of Natural History (5) 18, 271-90, 369-79, 445-66, pl. 10. Carter, H.J., 18866. Report on the marine sponges, chiefly from King I., in the Mergui Archipelago, collected for the Trustees of the Indian Museum, Calcutta, by Dr John Anderson, F.R.S., Superintendent of the Museum. Journal of the Linnean Society of London, Zoology 21, 61-84, pls 5-7. Chiffings, A.W., 1983. Western Australian marine environments. Australian Marine Sciences Association Bulletin (81), 14-19. de Laubenfels, M.W. (see Laubenfels, M.W. de). Dallwitz, M.J. & T.A. Paine, 1986. User's guide to the DELTA System. A general system for processing taxonomic descriptions (Third Edn). Commonwealth Scientific and Industrial Research Organisation Australia, Division of Entomology Technical Report (13), 1-106. Dendy, A., 1887. The sponge-fauna of Madras. A report on a collection of sponges obtained in the neighborhood of Madras by Edgar Thurston, Esq. Annals and Magazine of Natural History (5) 20, 153-64, pls 9-12. Dendy, A., 1889. Report on a second collection of sponges from the Gulf of Manaar. Annals and Magazine of Natural History (6) 3, 73-99, pls 3-5. Dendy, A,, 1896. Catalogue of non-calcareous sponges collected by J. Bracebridge Wilson, Esq., M.A., in the neighbourhood of Port Phillip Heads. Part 2. Proceedings of the Royal Society of Victoria 2(8), 14-51. Dendy, A., 1897. Catalogue of non-calcareous sponges collected by J. Bracebridge Wilson, Esq., M.A., in the neighbourhood of Port Phillip Heads. Part 3. Proceedings of the Royal Society of Victoria 2(9), 230-59. Dendy, A., 1905. Report on the sponges collected by Professor Herdman, at Ceylon, in 1902. Vol. 3(18), pp. 57-246, pls 1-16 in: 'Report to the Government of Ceylon on the Pearl Oyster Fisheries of the Gulf of Manaar'. Ed. W.A. Herdman. (Royal Society: London.) Dendy, A., 1916. Report on the non-calcareous sponges collected by Mr James Hornell at Okhamandal in Kattiawar in 1905-1906. Vol. 17, pp. 96-146, pls 1 4 in: 'Report to the Government of Baroda on the Marine Zoology of Okhamandal (2)'. Dendy, A., 1922. Report on the Sigmatotetraxonidacollected by H.M.S. 'Sealark' in the Indian Ocean. Transactions of the Linnean Society of London, Zoology 18, 1-164, pls 1-18. in: 'Reports of the Percy Sladen Trust Expedition to the Indian Ocean in 1905', Volume 7. Dendy, A., 1924. Porifera. Part I. Non-Antarctic sponges. Vol. 6(3), pp. 269-392, pls 1-15 in: 'British Antarctic ('Terra Nova') Expedition, 1910. Natural history report'. (British Museum (Natural History), Zoology: London.) Dendy, A. & L.M. Frederick, 1924. On a collection of sponges from the Abrolhos Is, Western Australia. Journal of the Linnean Society of London, Zoology 35, 477-519, pls 25-26, Desqueyroux-Faundez, R., 1981. Revision de la collection d'eponges d'Amboine (Moluques, Indonksie) constituee par Bedot and Pictet et consewee au Museum d'histoire naturelle de GenBve. Revue Suisse de Zoologie 88(3), 723-64, figs 1-132. Dickinson, M.G., 1945. Sponges of the Gulf of California. Vol. 11(1), pp. 1-57, pls 1-97 in: 'Reports on the Collections Obtained by Allan Hancock Pacific Expeditions of Velero 111 off the coast of Mexico, Central America, South America, and Galapagos Is in 1932-40'. Duchassaing, De Fonbressin, P. & G. Michelotti, 1864. Spongiaires de la mer CaraYbe. Naturk. Verh. Holland Maatsch. Wetensch Haarlem (ser. 2 ) 21(3), 1-124, pls 1-25. Ehlers, E, 1870. 'Die Esper'schen Spongien'. Zoologischen Sammlung der Konig. Universitgt Erlangen. Programm zum Eintritt in der Senat der Koniglichen Friedrich-Alexanders-Universittit in Erlangen: Erlangen, 36 pp. Esper, E.J.C., 1794. 'Die Pflanzenthiere in Abilungen nach der Natur mit Fabbenerleuchtet nebst Beschreibungen.' (Zweyter Theil: Niirnberg.) 303 pp., 109 pls. Ferrer Hernhdez, F. (see Hernhdez, F. Ferrer). Gray, J.E., 1867. Notes on the arrangement of sponges, with description of some new genera. Proceedings of the Zoological Society of London 1867, 492-558, pls 27-28. Guiler, E.R., 1950. Notes on Tasmanian marine sponges. Records of the Queen Victoria Museum, Launceston 3(2), 5-14. Hallmann, E.F., 1912. Report on the sponges obtained by the F.I.S. 'Endeavour' on the coasts of New South Wales, South Australia, Queensland, and Tasmania, 1909-10. Part 1. Zoological Results of the Fishing Experiments carried out by the F.I.S. 'Endeavour' 1909-10 (2), 117-300, pls 21-36. 1410 J. N. A. Hooper Hallmann, E.F., 1914a. A revision of the monaxonid species described as new in Lendenfeld's 'Catalogue of the Sponges in the Australian Museum. Part 1. Proceedings of the Linnean Society of New South Wales 29, 263-315, pls 15-24. Hallmann, E.E, 1914b. A revision of the monaxonid species described as new in Lendenfeld's 'Catalogue of the Sponges in the Australian Museum. Part 1. Proceedings of the Linnean Society of New South Wales 29, 398-446 pls. 15-24. Hallmann, E.F., 1916a. A revision of the genera with Microscleres included, or provisionally included, in the Family Axinellidae, with descriptions of some Australian species. Part ii. (Porifera). Proceedings of the Linnean Society of New South Wales 41(3), 495-552, pls 29-38. Hallmann, E.F., 1916b. A revision of the genera with Microscleres included, or provisionally included, in the Family Axinellidae, with descriptions of some Australian species. Part iii. (Porifera). Proceedings of the Linnean Society of New South Wales 41(4), 634-75, pls 29, 33, 3 8 4 . Hallmann, E.F., 1917. On the genera Echinaxia and Rhabdosigma [Porifera]. Proceedings of the Linnean Society of New South Wales 42(2), 391-405, pls 21-22. Hanitsch, R., 1889. Second report on the Porifera of the L.M.B.C. District. Proceedings of the Biological Society of Liverpool 3, 155-73, pls 5-7. Hanley, J.R., 1988. Invertebrate fauna of marine habitats in Darwin Harbour. Pp. 135-52 in: 'Darwin Harbour'. Eds H.K. Larson, M.G. Michie & J.R. Hanley. Australian National University North Australian Research Unit Mangrove Monograph Number 4. Hartman, W.D., 1982. Porifera. Pp. 640-66 in: 'Synopsis and Classification of Living Organisms'. Volume 1. Ed. S.P. Parker. (McGraw-Hill: New York.) Hechtel, G., 1983. New species of marine Demospongiae from Brazil. Iheringia Shies Zoologica, Porto Alegre 63, 59-89. Hentschel, E., 1911. Tetraxonida. 2. Teil. Vol. 3(10), pp. 279-393 in: 'Die Fauna Siidwest-Australiens'. Eds W. Michaelsen & R. Hartmeyer. (G. Fischer: Jena.) Hentschel, E., 1912. Kiesel- und Hornschw8mme der Aru und Kei-Inseln. Abhandlungn Senckenbergiana naturforschende Gessellschaji 1912, 2 9 5 4 8 . Hentschel, E., 1914. Monaxone Kieselschw8mme und Hornschwamme der Deutschen SiidpolarExpedition 1901-1903. Deutsche Sudpolar-Expedition 1901-1903 15, (Zoologie 7) 37-141. Hentschel, E., 1923. Erste Unterabteilung der Metazoa. Parazoa. Einziger Stamm und einzige Klasse der ersten Unterabteilung: Porifera = Schwiimme. Handbuch der Zoologie 1, 307-417. Hernandez, F. Ferrer, 1914a. Estudios sobre Espongiarios. Boletin de la Real Sociedad Espaiiola de Historia Natural, Madrid 14, 452-4. Hernindez, F. Ferrer, 1914b. Esponjas del Cantfibrico. Parte Segunda. 111: Myxospongida; IV: Tetraxonida; V: Triaxonida. Trabajos Museo Nacional de Ciencias Naturales Madrid, Serie Zoolbgica (17), 1-46. Hooper, J.N.A., 1984. A new genus and two new species of Haplosclerid sponges (Porifera :Demospongiae) from the Timor Sea, North-west Australia. Proceedings of the Royal Society of Victoria 96(2), 55-60. Hooper, J.N.A., 1988a. Structural features of the benthic community of East Point Reef Fish Reserve. A comparative study between oceanic, near-shore and inshore reefs of north-west Australia. Pp. 214-25 in: 'Darwin Harbour'. Eds H.K. Larson, M.G. Michie & J.R. Hanley. Australian National University North Australia Research Unit Mangrove Monograph, Number 4. Hooper, J.N.A., 19886. Character stability, systematics and affinities between Microcionidae (Poecilosclerida) and Axinellida (Porifera :Demospongiae). A revision of the Australasian Microcionidae'. Unpublished Doctor of Philosophy Thesis, University of Queensland, Department of Zoology, Volumes 1 and 2, 1208 pp., 141 figs, 43 pls. Hooper, J.N.A., 1990a. Character stability systematics and affinities between Microcionidae (Poecilosclerida) and Axinellida. 4.28494 In: 'Proceedings of the 3rd International Conference on the Biology of Sponges, Woods Hole, U.S.A., November 17-23 1985'. Smithsonian Contributions to Zoology. Hooper, J.N.A., 1990b. A new species of Rhabderemia Topsent (Porifera: Demospongiae) from the Great Bamer Reef. The Beagle, Records of the Northern Territory Museum of Arts and Sciences 7(1), 65-78. Hooper, J.N.A., in press. Phylum Porifera. Sponges. In: 'A Coral Reef Handbook'. Eds Patricia Mather and Isobel Bennett. (Surrey Beatty & Sons: Sydney). Third Edn. Hooper, J.N.A., unpublished data. Revision of the family Microcionidae (Porifera :Demospongiae), with description of Australian species. Invertebrate Taxonomy Hooper, J.N.A., R.J. Capon, C.P. Keenan & D.L. Parry, 1990. Biochemical and morphometric differentiation of two sympatric sibling species of Clathria (Porifera :Demospongiae) from northern Australia. Invertebrate Taxonomy 4(1), 123-48. Australian Raspailiidae 1411 Hooper, J.N.A., R.J. Capon, C.P. Keenan, D.L. Parry & N. Smit, 1992. Chemotaxonomy of marine sponges: families Microcionidae, Raspailiidae, and Axinellidae, and their relationships with other families in the orders Poecilosclerida and Axinellida (Porifera :Demospongiae). Invertebrate Taxonomy 6, in press. Hoshino, T., 1975. The sponges of the Anan coast. Zoological Magazine 84(1), 30-8, pls 1 4 . Johnston, G., 1842. 'History of British Sponges and Lithophytes.' (Edinburgh, London, Dublin.) 264 PPKeller, C., 1889. Die Spongienfauna des rothen Meeres. I. Halfte. Zeitschrift fur Wissenschaftliche Zoologie 48, 311-406, pls 20-25. Keller, C., 1891. Die Spongienfauna des rothen Meeres. 11. Halfte. Zeitschrift fur Wissenschafliche Zoologie 52, 294368, pls 16-20. Kelly Borges, M. & P.R. Bergquist, 1988. Sponges from Motupore I., Papua New Guinea. Indo-Malayan Zoology 5, 121-59. Kieschnick, O., 1896. Silicispongiae von Temate nach den Sammlungen von H e m Prof. Dr. W. Kiikenthal. Zoologischer Anzeiger 19, 526-34. Kieschnick, O., 1898. Die Kieselschwamme von Amboina. Inaugural Dissertation, Jena, 66 pp. Kieschnick, O., 1900. Kieselschwtimme von Amboina. Vol. 8, pp. 545-82, pls 44-45 in: 'Zoologische Forschungoreisen in Australien und den Malayischen Archipel ausgefuhrt in den Jahren 1891-1893'. Number 515. Ed. R. Semon. Denkschriften Medizinisch Naturwissenschaftliche Gesellschaft, Jena. Kirk, H.B., 1911. Sponges collected at the Kermadec Is by Mr W.R.B. Oliver. Transactions of the Royal Society of New Zealand 43, 574-81. Kirkpatrick, R., 1900. On the sponges of Christmas I. Proceedings of the Zoological Society of London 1900, 127-41. Kirkpatrick, R., 1910. On a remarkable pharetronid sponge from Christmas I. Proceedings of the Royal Society of London 83, 124-33. Kirkpatrick, R., 1903. Descriptions of South African sponges. Part III. Cape of Good Hope, Department of Agriculture Bulletin, No. 16. Marine Investigations in South Africa 2, 233-64, pls 5-6. Koltun, V.M., 1964. [Sponges of the Antarctic. 1. Tetraxonida and Comucospongida.] Issledovaniya Fauny Morei [Biological Results of the Soviet Antarctic Expedition (1955-1958)l 2, 6-131, pls 1-15 [Russian]. Koltun, V.M., 1976. Porifera-Part 1: Antarctic sponges. Report of the B.A.N.Z. Antarctic Research Expedition 1929-1931, Series B (Zoology and Botany) 9(4), 153-98. Lamarck, J.B.P. de Monet, 1813 (1814). Sur les polypiers empltts. Annales du Mustum &Histoire naturelle, Paris 20, 294312, 370-86, 432-58. Lamarck, J.B.P. de Monet, 1814 (1815). Suite des polypiers emptitks. Mtmoirs du Mustum &Histoire naturelle, Paris 1, 69-80, 162-8, 331-40. Lamarck, J.B.P. de Monet, 1816. 'Histoire Naturelle des Animaux sans Verttbres.' Verditre: Paris, Vol. 2, 568 pp. Laubenfels, M.W. de, 1930. The sponges of California. Stanford University Bulletin 5(98), 24-9. Laubenfels, M.W. de, 1936. A discussion of the sponge fauna of the Dry Tortugas in particular, and the West Indies in general, with material for a revision of the families and orders of the Porifera. Carnegie Institute of Washington Publication Number 467. Papers of the Tortugas Laboratory 30, 1-225, p l ~1-22. Laubenfels, M.W. de, 1950. The Porifera of the Bermuda Archipelago. Transactions of the Zoological Society of London 27, 1-154, pls 1-2. Laubenfels, M.W. de, 1954. The sponges of the west-central Pacific. Oregon State Monographs, Zoology 7, 1-306, p l ~1-12. Lendenfeld, R. von, 1887a. On the systematic position and classification of sponges. Proceedings of the Zoological Society of London (for 1886), 558-662. Lendenfeld, R. von, 1887b. Die Chalineen des australischen Gebietes. Zoologische Jahrbiicher, Jena 2, 723-828, p l ~18-27. Lendenfeld, R. von, 1888. 'Descriptive Catalogue of the Sponges in the Australian Museum, Sydney.' (Taylor & Francis: London.) 260 pp., 12 pls. Lendenfeld, R. von, 1890. Experimentelle Untersuchgen iiber die Physiologie der Spongien. Biologisches Centralblatt 10, 71-81, 102-10. Ltvi, C., 1958. Rtsultats scientifiques des campagnes de la 'Calypso'. Fascicule III. V. Campagne 1951-1952 en Mer Rouge. 5. Spongiaires de Mer Rouge recueillis par la 'Calypso' (1951-1952). Annales de I'lnstitut Ocianographique, Monaco 34, 3-46. Ltvi, C., 1960. Les Dtmosponges des cates de France. I. Les Clathriidae. Cahiers de Biologie Marine 1, 47-87. .... 1412 J. N. A. Hooper Levi, C., 1961. Spongiaires des Iles Philippines, principalment recoltkes au voisinage de Zamboanga. Philippines Journal of Science 88(4), 509-33. Levi, C., 1965. Spongiaires recoltes par l'6xpedition Israelienne dans le sud de la Mer Rouge en 1962. Bulletin of the Sea Fisheries Research Station, Israel. Israel South Red Sea Expedition, 1962, Report (13). 3-27. L6vi, C., 1967. DBmosponges r6coltCes en Nouvelle-Caledonie par la Mission Singer-Polignac. Exptdition Frangaise sur les rtcifs coralliens de la Nouvelle-Calddonie. Editions de la Foundation Singer-Polignac 2, 13-26, pls 1-3. Lkvi, C., 1969. Spongiaires du Vema Seamount (Atlantique Sud). Bulletin du Musbum National &Histoire Naturelle 41, 952-73, pls 1-5. Levi, C., 1973. SystCmatique de la classe des Demospongiaria (D6mosponges). Pp 577-631 in: 'Trait6 de Zoologie. Anatomie, Systematique, Biologie. 111. Spongiaires'. (SCr.ed. P-P. Grass&) Eds P. Brien, C. LCvi, M. Sarh, 0. Tuzet & J. Vacelet. (Masson et Cie: Paris.) LCvi, C., 1979. The demosponge fauna from the New Caledonian area. Proceedings of the International Symposium on Marine Biogeography and Evolution in the Southern Hemisphere. New Zealand Oceanographic Institute Special Volume, 307-15. Uvi, C. & P. L k i , 1983. Demosponges bathyales rkcolt6es par le N/O 'Vauban' au sud de la Nouvelle-Caledonie. Bulletin du Musdum National &Histoire Naturelle (4e sdrie) 5(A, no. 4), 931-97, pls 1-8. ~ e s dans 1'Atlantique Oriental par le 'President-TheodoreLCvi, C. & J. Vacelet, 1958. ~ ~ o n recoltees Tissier' (1955-1956). Revue des Travaux. Institut Scientifrques et Technique des P8ches Maritimes, Paris 22, 225-46. Lundbeck, W., 1910. 3. Porifera. (Part 111). Desmacidonidae (Pars.). Vol. 6, pp. 1-124, pls 1-11 in: 'The Danish Ingolf-Expedition'. (Bianco Luno: Copenhagen.) Michie, M., 1987. Distribution of Foraminifera in a macrotidal tropical estuary: Port Darwin, Northern Tenitory of Australia. Australian Journal of Marine and Freshwater Research 38, 249-59. Montagu, G., 1818. An essay on sponges, with descriptions of all the species that have been discovered on the coast of Great Britain. Memoirs of the Wernerian Natural History Society, Edinburgh 2(1), 67-122, p l ~3-16. Nardo, D., 1833. 'Auszug aus einem neuen System der Spongiarien, wornach bereits die Aufstellung in der UniversitMs-Sammlung zu Padua gemacht ist' Isis, oder Encyclopadische Zeitung, von oken: Jena, coll. 519-23. Nardo, D., 1847. 'Prospetto della Faune marina volgare del Veneto estuario con cenni sulle principali specie commestibili dell'Adriatico, ecc.'. Estratto dall'opera: Venezia e la sue lagune, 45 pp. Norman, A.M., 1878. On the genus Haliphysema, with descriptions of several forms apparently allied to it. Annals and Magazine of Natural History (5) 1, 26484, pl. 16. Pick, EK., 1905. Die Ganung Raspailia. Archiv fur Naturgeschichte 1905, 1-48, pls 1-4. Pope, E.C., 1967. The shore reefs of Darwin. Australian Natural History (March 1967), 278-84. Pulitzer-Finali, G., 1977. Report on a collection of sponges from the Bay of Naples. 111. Hadromerida, Axinellida, Poecilosclerida, Halichondrida, Haplosclerida. Bolletino dei Musei e degli Istituti Biologici dell'Universith di Genova 45, 7-89. Pulitzer-Finali, G., 1983. A collection of Mediterranean Demospongiae (Porifera) with, in appendix, a list of the Demospongiae hitherto recorded from the Mediterranean Sea. Annali del Museo Civico di Storia Naturale di Genova 84, 445-621. Ridley, S.O., 1884. Spongiida. 4. 366-482 (Australia), 582-630 (Western Indian Ocean), pls 39-43, 53-54 in: 'Report on the Zoological Collections made in the Indo-Pacific Ocean during the Voyage of H.M.S. 'Alert' 1881-2'. (British Museum (Natural History): London.) Ridley, S.O., 1885. The Monaxonida. Reports on the 'Challenger', Narrative 1(2), 569-73. Ridley, S.O. & A. Dendy, 1886. Preliminary report on the Monaxonida collected by the H.M.S. 'Challenger'. Annals and Magazine of Natural History (5) 18, 325-51, 470-93. Ridley, S.O. & A. Dendy, 1887. Report on the Monaxonida collected by H.M.S. 'Challenger' during the years 1873-76. Vol. 20, pp. 1-275, pls 1-51 in: 'Report on the Scientific Results of the Voyage of H.M.S. 'Challenger' during the Years 1873-76'. (Her Majesty's Stationary Office: London, Edinburgh, Dublin.) Ridley, S.O. & P.M. Duncan, 1881. On the genus Plocamia, Schmidt, and on some other sponges of the order Echinonemata. With descriptions of two additional new species of Dirrhopalum. Journal of the Linnean Society, London 15, 476-97, pls 28-29. Sarh, M., 1958. Studio sui Poriferi di una grotta di marea de Golfo di Napoli. Archivio Zoologica Italiano, Napoli 43, 203-80, pls 1-2. Australian Raspailiidae 1413 Schmidt, E.O., 1862. 'Die Spongien des Adriatischen Meeres.' (Wilhelm Engelmann: Leipzig.) 88 pp., 6 pls. Schmidt, E.O., 1866. 'Zweites Supplement der Spongien des Adriatischen Meeres. Enthaltend die Vergleichung der Adriatischen und Britischen Spongiengattungen.' (Wilhelm Engelmann: Leipzig.) 24 pp., 1 pl. Schmidt, E.O., 1868. 'Die Spongien der KUst von Algier. Mit Nachtriigen zu den Spongien des Adriatischen Meeres. (Drittes Supplement).' (Wilhelm Engelmann: Leipzig.) 44 pp., 5 pls. Schmidt, E.O., 1870. 'Grundzuge einer Spongien-Fauna des Atlantischen Gebietes.'. (Wilhelm Engelmann: Leipzig) 88 pp., 6 pls. Semeniuk, V., 1985. Mangrove environments of Port Darwin, Northern Temtory: the physical framework and habitats. Journal of the Royal Society of Western Australia 67(3/4), 81-97. Shaw, M.E., 1927. On a collection of sponges from Maria I., Tasmania. Proceedings of the Zoological Society of London 18, 419-39, pl. 1. Sim, C-J. & G.J. Bakus, 1986. Marine sponges of Santa Catalina I., California. Occasional Papers of the Allan Hancock Foundation. New Series (5), 1-23. Simpson, T.L., 1968. The structure and function of sponge cells: new criteria for the taxonomy of Poecilosclerid sponges (Demospongiae). Bulletin of the Peabody Museum of Natural History 25, 1-141, p l ~1-17. Simpson, T.L., 1984. 'The Cell Biology of Sponges.' (Springer-Verlag: Berlin.) Soest, R.W.M. van, 1984. Marine sponges from Curapao and other Caribbean localities. Part 111. Poecilosclerida. Studies on the Fauna of Curagao and other Caribbean Is (199), 1-167, pls 1-10. Soest, R.W.M. van, 1987. Phylogenetic exercises with monophyletic groups of sponges. Vol. G13, pp. 227-241 in: 'Taxonomy of Porifera'. Eds J. Vacelet & N. Boury-Esnault. (NATO AS1 Series, Springer-Verlag: Berlin.) Soest, R.W.M. van & S. Weinberg, 1980. A note on the sponges and octocorals from Sherkin I. and Lough Ine, Co. Cork. The Irish Naturalists' Journal 20(1), 1-15. Soest, R.W.M. van, M.C. Diaz & S.A. Pomponi, 1990. Phylogenetic classification of the halichondrids (Porifera, Demospongiae). Beaufortia 40(2), 15-62. Sollas, W.J., 1879a. On Plocamia plena, a new species of echinonematous sponge. Annals and Magazine of Natural History (5) 4, 4453, pls 6-7. Sollas, W.J., 1879b. On Plectronella papillosa, a new genus and species of echinonematous sponge. Annals and Magazine of Natural History (5) 3, 17-27. Sollas, W.J., 1885. A classification of the sponges. Scientijic Proceedings of the Royal Dublin Society (2) 5, 112. Monaxonellida (suborder Stephens, J., 1916. Preliminary notice of some Irish sponges.-The Sigmatomonaxonellida) obtained by the Fisheries Branch of the Department of Agriculture and Technical Instruction, Ireland. Annals and Magazine of Natural History 17, 232-43. Strand, E., 1924. Miscellanea nomenclatorica zoologica et palaeontologica. ArchivJiir Naturgeschichte, Berlin 92A(b), 31-6. Swofford, D.L., 1985. 'PAW. Phylogenetic Analysis Using Parsimony. Version 2.4.' (Illinoise Natural History Survey: Champaign.) Thiele, J., 1898. Studien uber pazifische Spongien. I. Heft. Zoologica 24, 1-72, pls 1-8. Thiele, J., 1899. Studien uber pazifische Spongien. 11. Heft. Zoologica 24, 1-33, pls 1-5. Thiele, J., 1903. Kieselschwiimme von Ternate. 11. Abhandlungen der Senckenbergischen Naturforschenden Gesellschaft 25, 933-68, pl. 28. Thiele, J., 1905. Die Kiesel- und Hornschwamme der Sarnmlung Plate. Zoologische Jahrbucher, Jena Supplement 6 [Fauna chilensis (3)], 407-96, pls 27-33. Thomas, P.A., 1968. Studies on Indian sponges--I. Two new species of silicious sponges belonging to the genera Echinodictyum Ridley and Rhadberemia [sic.] Topsent (class: Demospongiae Sollas, order: Poecilosclerida Topsent). Journal of the Marine Biological Association of India 10(2), 245-9, p l ~1-2. Thomas, P.A., 1976. Endectyon lamellosa n.sp. (Demospongiae :Poecilosclerida, Raspaliidae) from the Indian Seas and a revised key to the Indian species of Endectyon Topsent. Journal of the Marine Biological Association of India 18(1), 169-72, pl. 1. Thomas, P.A., 1977. Studies on Indian sponges. VIII. Four new records of silicious sponges Echinochalina glabra (Ridley and Dendy), Higginsia mixta (Hentschel), Geodia lindgreni (Lendenfeld) and Pachamphilla dendyi Hentschel from the Indian Ocean. Journal of the Marine Biological Association of India 19(1), 115-22. Topsent, E., 1889. Quelques Spongiaires du Banc de Campgche et de la Pointe-a-Pitre. Mdmoires de la Socidtd Zoologique de France 2, 3C-52. 1414 J. N. A. Hooper Topsent, E., 1890~.Eponges de la Manche. Mdmoires de la Socidtd Zoologique de France 3, 195-205. Topsent, E., 1890b. h d e s de Spongiaires. Revue de Biologie du Nord de France 2, 289-98. Topsent, E., 1892~. Contribution 1'6tude des Spongiaires de 1'Atlantique Nord. Rdsultats des Campagnes Scienti'ues Accomplies sur son Yacht par Albert Ier Prince Souverain de Monaco 2, 1-165, p l ~1-11. Topsent, E., 1892b. Eponges de la Mer Rouge. Mkmoires de la Socikte Zoologique de France 5, 21-9, PI. 1. Topsent, E., 1893. Nouvelle skrie de diagnoses d'Eponges de Roscoff et de Banyuls. Archives de Zoologie Expdrimentale et Gdnkrale (3), Notes et Revue 1, 3343. Topsent, E., 1894. Une rkfonne dans la classificaiton des Halichondrina. Memoires de la Socidtd Zoologique de France 7, 5-26. Topsent, E., 1897. Spongiaires de la Baie d'hboine. Voyage de MM. M. Bedot et C. Pictet dans 1'Archipel Malais. Revue Suisse de Zoologie 4, 421-87. Topsent, E., 1904. Spongiaires des A~ores. Rksultats des Campagnes Scientifrques Accomplies sur son Yacht par Albert ler Prince Souverain de Monaco 25, 1-280, pls 1-18. Topsent, E., 1906. Eponges recueillies par M. Ch. Gravier dans la Mer Rouge. Bulletin du MusCum National &Histoire Naturelle 1906, 557-70. Topsent, E., 1907. Poeciloscl6rides nouvelles recueillies par le 'Fran~ais' dans 1'Antarctique. Bulletin du Muskum National &Histoire Naturelle 1907, 69-76. Topsent, E., 1913. Spongiaires de 1'Expkdition Antarctique Nationale Ecossaise. Transactions of the Royal Society of Edinburgh 49(3, 9), 579-643, pls 1-6. Topsent, E., 1920. Spongiaires du Muske Zoologique de Strasbourg. Monaxonides. Bulletin de I'lnstitut Ocdano,graphique (Monaco) (381). 1-36. Topsent, E., 1925. Etude des Spongiaires du Golfe de Naples. Archives de Zoologie Exptrimentale et Gdndrale 63(5), 623-725, pl. 8. Topsent, E., 1927. Diagnoses d'EPonges nouvelles recueillies par le Prince Albert Ier de Monaco. Bulletin de I'lnstitut Ocdanographique (Monaco) 502, 1-19. Topsent, E., 1928. Spongiaires de 1'Atlantique et de la M6diterrante provenant des croisihres du Prince Albert Ier de Monaco. Rksultats des Campagnes Scientifiques Accomplies sur son Yacht par Albert Ier Prince Souverain de Monaco 74, 1-376, pls 1-11. Topsent, E., 1930. Eponges de Lamarck conserv6es au Mustum de Paris. Archives du Musdum National d'Histoir$ Naturale (6) 5, 1-56, pls 1 4 . Topsent, E., 1932. Eponges de Lamarck conservtes au Mustum de Paris. Deuxieme partie (I). Archives du Mustum National d'Histoire Naturelle (6) 8, 61-124, pls 1-6. Topsent, E., 1933. EPonges de Lamarck conserv6es au Mustum de Paris. Fin (I). Archives du Muskum National &Histoire Naturelle (6) 10, 1-60, pls 1-3. Topsent, E., 1936. Eponges observtes dans les parages de Monaco. (DeuxiEme Partie). Bulletin de I'lnstitut Ockanographique (Monaco) (686), 1-70. Uriz, M-J., 1988. Deep-water sponges from the continental shelf and slope off Namibia (South-west Africa): classes Hexactinellida and Demospongia. Monographica Zoologica Marinas 3, 9-157, pls 143. Vacelet, J., 1961. Spongiaires (demosponges) de la region de Bonifacio (Corse). Recueil des Travaux de la Station Marine d'Endourne 36(22), 2145. Vacelet, J., 1969. Eponges de la roche du large et de 1'Ctage bathyal de Mtditerranke. Mkmoires du Mustum National &Histoire Naturelle (A) 59(2-3), 145-219, pls 1-4. van Soest, R,W.M (see Soest, R.W.M. van). Vosmaer, G.C.J., 1912. On the distinction between the genera Axinella, Phakellia, Acanthella a.0. Zoologische Jahrbucher, Jena Supplement 15, 307-22, pls 15-16. Vosmaer, G.C.J., 1935. 'The Sponges of the Bay of Naples. Porifera Incalcaria with Analyses of Genera and Studies in the Variations of Species.' Eds C.S. Vosmaer-Roe11 & M. Burton. (Martinus NijhoR The Hague.) Vol. 2, pp. 457-828. Whitelegge, T., 1889. List of the marine and freshwater invertebrate fauna of Port Jackson and the neighbourhood. Journal of the Royal Society of New South Wales 23(2), 163-323. Whitelegge, T., 1897. The sponges of Funafuti. Memoirs of the Australian Museum 3, 323-32, pl. 18. Whitelegge, T., 1901. Report on sponges from the coastal beaches of New South Wales. Records of the Australian Museum 4(2), 1-70 [55-1181, pls 10-15. Whitelegge, T., 1902. Notes on Lendenfeld's Types described in the 'Catalogue of Sponges' in the Australian Museum. Records of the Australian Museum 4(7), 27488. Whitelegge, T., 1907. Sponges. Part 1.-Addenda. Part 2. Monaxonida continued. In: 'Scientific Results of the Trawling Expedition of H.M.C.S. 'Thetis' Off the Coast of New South Wales in February and March, 1898'. Memoirs of the Australian Museum 4(10), 487-515, pls 4546. Australian Raspailiidae 1415 Whitley, G., 1932. Marine zoogeographical regions of Australia. Australian Naturalist 8, 166-7. Wiedenmayer, F., 1977. 'Shallow-water Sponges of the Western Bahamas.' (Birkhauser: Basel.) Experimentia Supplementum 28, 1-287, pls 1-42. Wiedenmayer, F., 1985. Perspectives in taxonomy of Australian marine sponges. Pp. 72-3 in: 'Abstracts of the 3rd International Conference on the Biology of Sponges, Woods Hole, USA, November 17-23 1985'. Smithsonian Contributions to Zoology. Wiedenmayer, E, 1989. Demospongiae (Porifera) from northern Bass Strait, southern Australia. Memoirs of the Museum of Victoria 50(1), 1-242, pls 1-38. Wiedenmayer, F., J.N.A. Hooper & A.A. Racek, in press. Porifera. In: 'Zoological Catalogue of Australia, Volume 9'. Ed. D.W. Walton. (Australian Government Printing Service: Canberra.) Wilson, B.R. & G.R. Allen, 1987. 3. Major components and distribution of marine fauna. In: 'Fauna of Australia. Volume lA, General Articles'. Eds G.R. Dyne & D.W. Walton. (Australian Govemment Printing Service: Canberra.) Wilson, H.V., 1902. The sponges collected in Porto Rico in 1899 by the U.S. Fish Commission Steamer 'Fish Hawk'. Bulletin of the United States Fish Commission for 1900 2, 375-411. Wilson, H.V., 1921. The genus Raspailia and the independent variability of diagnostic features. Journal of the Elisha Mitchell Scientific Society, University of North Carolina at Chapel Hill 37, 54-60. J . N . A. Hooper 1416 Systematic Index (Names in bold type refer to species described from the Australasian region.) Abila Grays ........................................... 1195 Abilana Strand ........................................ 1195 Acantheurypon Topsent .............................1313 acerata..Dendy. Axinella ........................... 1267 aceratus (Carter). Echinodictyum...............1349 aculeata (Johnston). Raspaciona ................1309 agminata Hallmann. Raspailia .......... 1198. 1222 Amphinomia. gen. nov.............................. 1321 anchoratum var. ramosa Lendenfeld Echinonema ............................... 1222 antrodes de Laubenfels. Kieplitela............. 1347 arbuscula (Lendenfeld),Raspailiu .. 1195. 1197. 1198. 1219. 1222 arenosum Dendy, Echinodictyum ............. 1349 argon Dickinson. Cyamon.........................1304 aruense Hentschel, Cyamon..................... 1305 asperum Ridley & Dendy, Echinodictyum .................................. 1353 atlanticus (Lbvi & Vacelet). Lithoplocamia................................... 1320 atropurpurea (Carter), Raspailiu ..... 1197. 1200 Aulospongiella Burton .............................. 1307 Aulospongus Norman ............................... 1307 australiensis Ridley, Raspailiu ......... 1220. 1256 austrinus, sp nov., Echinodictyum ........... 1356 Axechina Hentschel.................................. 1344 axifera (Hentschel), Ceratopsion.............. 1330 Axinectya Hallmann ........................ 1195. 1199 axinelloides Brondsted. Echinodicty um ....... 1349 Axoplocamia Burton ................................. 1319 . Basiectyon Vacelet ................................. 1284 bifurcata Ridley. Raspailiu ..... 1197. 1219. 1228 bilamellata. in part. ~amarck.Spongia ...... 1347. 1379 cacticutis (Carter). Raspailiu ........... 1197. 1230 cactoides (Burton & Rao). Eurypon ........... 1314 calva Sarh. Raspaciona ............................. 1311 cancellaturn (Lamarck). Echinodictyum .................................. 1358 Cantabrina Ferrer.Hernandez .................... 1394 carlinoides (Larnarck). Echinodictyum ......1363 carteri (Duncan). Plocamione .................... 1319 catalina Sim & Bakus. Cyamon................. 1304 cavernosum Thiele. Echinodictyum.............1349 clathrata Ridley. Raspailiu ..... 1220. 1259. 1260 clopetaria (Schmidt). Plocamione.............. 1319 Ceratopsis Thiele ..................................... 1327 Ceratopsion Strand .................................. 1327 cervicornis (Burton). Raspailia ........ 1195. 1199. cladofagellata Carter. Axinella ..................1213 clathratum Dendy. Echinodictyum .............. 1349 Clathriodendron Lendenfeld .... 1195. 1199. 1222 clathrioides Hentschel. Echinodictyum ...... 1366 clathrioides Lbvi. Aulospongus .................. 1307 clavata Thiele. Ceratopsion....................... 1328 clavatum (Bowerbank). Eurypon ................ 1314 compressa Bergquist. ~as$ilia ...... 1197. 1198. 1245 conulosum Kieschnick. Echinodictyum ..... 1370 costiferum Ridley. Echinodictyum............. 1373 cuneiformis Bergquist. Ceratopsion............ 1328 Cyamon Gray ................................ 1298. 1304 . daminensis. sp nov.. Raspailia ....... 1197. 1232 delaubenfelsi Burton. Endectyon ................ 1285 demonstrans Topsent. Endectyon ................ 1285 dendroides Hechtel. Echinodictyum ............ 1349 desmoxyiformis. sp nov., Raspailia .. 1197. 1236 dichotoma (Whitelegge). Ceratopsion....... 1220. 1328 Dictyocylindrus. in part. Bowerbank........... 1195 Dictyocylindrus Carter .............................. 1347 dirrhopalina Topsent. Plocamione .............. 1319 Dirrhopalum Ridley ................................. 1319 dolichosclera LCvi & Lbvi. Lithoplocamia ................................... 1321 . echinata Whitelegge. Raspailiu ................ 1203 Echinaxia Hallmann ........................ 1195. 1198 Echinodictyum Ridley ...................... 1198. 1347 Ectyoplasia Topsent................................. 1273 elegans Lendenfeld. Kalykenteron ..... 1347. 1379 elegans (Lendenfeld). Raspailia ................. 1262 elyakovi, sp nov.. Endectyon .................... 1285 Endectyon Topsent ................................... 1284 Epicles Gray .......................................... 1313 erecta Ferrer.Hernandez. Cantabrina.......... 1394 erecta Thiele. Ceratopsion........................ 1328 Eurypon Gray ........................................ 1313 Euryponidae Topsent................................ 1193 expansa (Thiele). Ceratopsion.......... 1327. 1328 . falcifera Topsent. Raspailia ..... 1195. 1200. 1220 fascispiculiferum (Carter). Sigmeurypon...... 1394 ferox (Duchassaing & Michelotti). Ectyoplasia............................. 1273. 1327 fabellaturn Topsent. Echinodictyum............1349 fabelliforme (Keller). Echinodictyum..........1349 jkbelliforme Carter. Trikentrion..............1298 1258. 1267 faccida Bergquist. Raspailia ... 1187. 1198. 1199 cewicornis Ridley & Dendy. fagelliformis Ridley & Dendy. Thrinacophora........................... 1340 Raspailia ................................ 1220. 1259 chalinoides var. glutinosa Carter. folium Thiele. Raspailia ................... 1198. 1199 Axinella ........................................... 1213 freyerii Schmidt. Raspailia ...............1195. 1220 Australian Raspailiidae 1417 page frondosa (Lendenfeld).Ectyoplasia .. 1273. 1327 frondula (Whiteleggge). ~ a s ~ a i l... i a1195. 1199. 1248 fruticosa (Dendy). Endectyon ........... 1285. 1290 fruticosa aruensis (Hentschel). Endectyon ........................................ 1290 fruticosum Hentschel. Echinodictyum ....... 1375 funiformis Ridley & Dendy. Thrinacophora ................................... 1339 gardineri (Dendy). Aulospongus ................ 1308 glorneratum Ridley. Echinodictyum ............ 1363 gracilis (Lendenfeld) Raspailia ...... 1197. 1205 graphidiophora Hentschel. Eurypon ......... 1315 gymnazusa (Schmidt). Plocamione ............. 1319 . hamata (Schmidt). Endectyon ........... 1284. 1285 helium Dickinson. Trikentrion.................... 1298 Hemectyon Topsent................................. 1284 Hemectyonilla Burton ............................... 1307 Heterectya Hallmann ................................ 1307 hirsuta Thiele. Raspailia ........................... 1220 horsuta Thiele. Raspalia .................. 1198. 1199 hispida (Montagu). Raspailia ...ll95. 1197. 1215 humilis (Topsent). Raspailia ...................... 1220 Hymeraphia Bowerbank ........................... 1313 Hymeraphiopsis. subg. nov........................ 1270 hystrix (Duncan). Plocamione .................... 1319 inaequalis Dendy. Raspailia .... 1187. 1198. 1199 incrustans Kieschnick. Thrinacophora........ 1340 involutum (Kirkpabick). ~ u l o s ~ & u ........ s 1307. 1327 irregularis Hentschel. Raspailia ....... 1197. 1270 jousseaumi Topsent. Echinodictyum ............ 1349 Kalykenteron Lendenfeld .......................... 1347 keriontria. sp nov.. Raspailia ................... 1237 Kieplitela de Laubenfels ........................... 1347 koltuni Sim & Bakus. Cyamon .................. 1304 . lacazei Topsent. Echinodictyum ................. 1349 laciniatus (Carter). Echinodictyum ............. 1349 lacunosum Kieschnick. Echinodictyum..... 1377 laeve Carter. Trikentrion........................... 1298 lamellosa Thomas. Endectyon .................... 1285 lithistoides Dendy. Lithoplocamia .............. 1320 Lithoplocamia Dendy ............................... 1320 longistylum Thomas. Echinodictyum ........... 1349 lugubris (Duchassaing & Michelotti). Echinodictyum ................................... 1349 macroxiphera LCvi. Echinodictyum ............ 1349 mariana Ridley & Dendy. Axinectya ......... 1195. 1220. 1327 marleyi Burton. Echinodictyum .................. 1349 melanorhops. sp nov.. Raspailia .............. 1240 . Mesapos Gray ........................................ 1313 mesenterinum (Lamarck). Echinodictyum ......................... 1347. 1379 microxephora (Kirkpahick). Ceratopsion.... 1328 rniniaceum..Burton. Eurypon ..................... 1270 minor Pulitzer.Fiali. Ceratopsion.............. 1328 Monectyon LBvi & Vacelet ........................ 1320 montebelloensis. sp nov.. Ceratopsion...... 1335 rnonticularis (Ridley & Dendy). Aulospongus ...................................... 1307 muricata (Esper). Trikentrion.................... 1298 . neon de Laubenfels. Cyamon ..................... 1304 nervosa (Lamarck). Echinodictyum ............. 1349 nidulus Hentschel. Echinodictyum ............ 1385 nigra Lendenfeld. Clathriodendron............ 1222 nuda Hentschel. Raspailia ..... 1197. 1220. 1264 ornata (Dendy). Plocamione ..................... 1319 pachysclera (LCvi & LCvi). Plocamione ...... 1319 palmata. sp nov.. Ceratopsion.................. 1337 papillosa Sollas. Plectronella .................... 1298 paradoxa Hentschel. Raspailia ....... 1197. 1235. 1244 Parasyringella Topsent .................... 1195. 1200 pennatum (Duchassaing & Michelotti). Echinodictyum ............................ 1349 perforata. Lendenfeld. Antherochalina........ 1273 phakellina Topsent. Raspaxilla ................... 1195 phakellopsis. sp nov.. Raspailia ............... 1207 pilosella (Topsent). Eurypon ............. 1313. 1315 pilosus (Vacelet). Endectyon ............. 1284. 1285 pinnatifida (Carter). Raspailia ........ 1197. 1211. 1219 Plectronella Sollas .................................. 1298 Plocamia Schmidt ................................. 1319 Plocamione Topsent................................. 1319 polyplumosa (LCvi). Eurypon .................... 1314 pro@& Ridley & Dendy. Raspailia .......... 1220 Proraspailia LCvi .................................... 1314 Protoraspailia Burton & Rao .................... 1314 pulchrum Brondsted. Echinodictyum ........... 1358 pykei (Carter). Echinodictyum .................... 1349 . . quinqueradiata (Carter). Cyamon ............... 1304 ramosa Lendenfeld. Echinonema ............... 1222 ramosa Thiele. Ceratopsion ...................... 1328 Raspaciona Topsent................................. 1309 Raspailia Nardo .............................. 1195. 1200 Raspailiidae Hentschel .............................. 1193 raspailioides Hentschel, Axechina ... 1326. 1345 Raspailopsis Burton ........................ 1195. 1199 Raspaxilla Topsent................. 1195. 1199. 1245 Raspeloplocarnia Burton ........................... 1319 reticula&, sp nov., Raspailia ................... 1250 Rhabdeurypon Vacelet .............................. 1317 . J. N. A. Hooper Rhaphidectyon Topsent............................. 1307 rhaphidophora Hentschel. Raspailia .. 1220. 1340 rigida Ridley & Dendy. Raspailia ..... 1220. 1259 robusta Sarh. Raspaciona ..........................13 11 rubra var. digitata Lendenfeld. Halichondria ..................................... 1222 rugosum Ridley & Dendy. Echinodictyum ........................... 1390 setacea Carter. Axinella ............................ 1211 Sigmeurypon Topsent ............................... 1394 silex Lendenfeld. Kalykenteron.................. 1379 spinosa (Bowerbank). Tethyspira ............... 1393 spinosa Wilson. Thrinacophora................. 1340 spinosum (Topsent). Aulospongus ......1307. 1327 spinosum Vacelet. Rhabdeurypon ............... 1317 s t e l l i d e m (Carter), Raspailia ........ 1197. 1267 stellifera Bowerbank. Hymeraphia ............. 1313 sulphurea, sp nov., Amphinomia .... 1321. 1322 supraturnescens Topsent. Axinella .............. 1259 Syringella of authors ....................... 1195. 1256 Syringella Schmidt. Raspailia .................... 1195 . tabula (Lamarck). Ectyoplasia ........ 1277. 1327 teissieri Cabioch. Endectyon ...................... 1285 tenax (Schmidt). Endectyon ....................... 1284 tedella (Lendenfeld).Raspailia ................. 1213 tenuis (Ridley & Dendy). Endectyon ..........1285 Tethyspira Topsent ................................. 1393 Thrinacophora Ridley .............................. 1339 thurstoni (Dendy). Endectyon ......... 1285. 1292 topsenti de Laubenfels. Echinodictyum....... 1379 topsenti Dendy. Raspailia ................. 1198. 1199 Tricheurypon Topsent............................... 1314 Trikentrion Ehlers .................................. 1298 tubulatus (Bowerbank). Aulospongus ..........1307 typica Nardo. Raspailia ................... 1195. 1197 typica..Whitelegge. Thalassodendron.......... 1379 Valedictyum de Laubenfels...... 1195. 1200. 1219 vannus. sp nov.. Ectyoplasia 1280 vasiplicata Carter. Echinonema .................. 1379 vesti@fera Dendy. Raspailia ............ 1195. 1215 vickersii Bowerbank. Cyamon.................... 1304 villosa (Thiele). Aulospongus ........... 1307. 1327 viminalis Schmidt. Raspailia ... 1188. 1189. 1195 viridis (Topsent). Eurypon ......................... 1314 . . .................... wardi. sp nov.. Raspailia ......................... 1252 wilkinsoni. sp nov. Raspailia .......... 1198. 1220 . xerampelina (Lamarck). Endectyon ..1185. 1294 Manuscript received 7 May 1990; accepted 2 November 1990

RELATED PAPERS

RELATED TOPICS

Log In

Revision of the Family Raspailiidae (Porifera: Demospongiae), with description of Australian Species.

Revision of the Family Raspailiidae (Porifera: Demospongiae), with description of Australian Species.

Related Papers

RELATED PAPERS

RELATED TOPICS