The BEAGLE Records of the Northern Territory Museum of Arts and Sciences Number 2 December, 1990 Volume 7 The BEAGLE, Records of the Northern Territory Museum of Arts and Sciences (formerly ‘‘Occasional Papers of') ABSTRACTED IN ZOOLOGICAL RECORD AND BIOSIS Editorial Board J.N.A. Hooper H.K. Larson J.J. Coe S. Gabrieli Editor Associate Editor (Sciences) Associate Editor (Humanities) Assistant Editor C. Jack-Hinton A.J. Bruce P.F. Murray U. Dasari Distributions The Beagle is a refereed journal published by the Scientific Publications of the Northern Territory Museum to disseminate the results of research in the following areas: (i) Systematic and other studies of the terrestrial, marine and freshwater flora and fauna of the Northern Territory, tropical Australia, Southeast Asia and Oceania; (ii) Australian Aboriginal, Southeast Asian and Oceanic art, material culture and archaeol¬ ogy; (iii) Northern Territory and Indo-Pacific history, maritime history and archaeology; (iv) Museum conservation, display techniques, taxidermy, publication, public relations, communication and education. The Beagle is published once or twice a year, depending upon the material available and budgetary considerations. All contributions to The Beagle are peer reviewed by two referees, and where possible at least one international referee is selected by the Editorial board. Whilst articles for The Beagle will normally be 10-30 pages in length, shorter communica¬ tions and notes. Book Reviews and review articles may be acceptable also. Longer articles, significant works or substantial revisions, which form integral studies, may be considered for separate publication in the Monograph Series of the Northern Territory Museum of Arts and Sciences. The Monograph Series is distributed with the Beagle. Prospective authors should contact the Editor, Scientific Publications. The Beagle may be obtained by subscription or by exchange. Subscription rates for one year, which include the Beagle and Monograph Series are: Individuals $30.00, Institutions $40.00. Postal charges, etc. In Australia only - $5.00; overseas - $10.00 (surface), $20.00 (airmail). Cheques should be made payable to ‘Northern Territory Museum’. All subscriptions, back numbers and exchange enquiries should be addressed to the Librarian, Northern Territory Museum, P.O.Box 4646, Darwin, NT 0801, Australia. AUTHORS OFFPRINTS Twenty-five offprints are provided free for each published paper. Additional offprints may be ordered when returning proofs. SUBMISSION OF MANUSCRIPTS Contributions, including the original and two copies of each manuscript, a computer floppy disk if available, with copies of art work should be addressed to: The Editor, Scientific Publications Northern Territory Museum of Arts and Sciences, GPO Box 4646, Darwin NT 0801, AUSTRALIA. © Northern Territory Museum of Arts and Sciences, 1990. ISSN 0811-3653 Cover: Paradicapretta brucei gen. et sp. nov. Drawing by A. Hirayama. Government Printer of the Northern Territory The BEAGLE, Records of the Northern Territory Museum of Arts and Sciences Volume 7 Number 2 December, 1990 CONTENTS SCIENCES BRUCE, A.J. - A second species of Bresilia B. plumifera sp. nov., new to the Australian fauna (Crustacea: Decapoda: Bresiliidae). BRUCE, A.J. - Recent additions to the pontoniine shrimp fauna of Australia. AKIRA HIRAYAMA - Two new caprellidean (n. gen.) and known gammaridean amphipods (Crustacea) collected from a sponge in Noumea, New Caledonia. SADLIER, R.A. - A new species of scincid lizard from Western Arnhem Land, Northern Territory. RUSSELL, B.C. - A new species of Nemipterus (Pisces: Nemipteridae) from the south-western Pacific. MURRAY, P. - Primitive marsupial tapirs (Propalorchestes novaculacephalus Murray and P. ponticulus sp. nov.) from the mid-Miocene of North Australia (Mar- supialia: Palorchestidae). MURRAY, P. - Alkwertatherium webbi, a new zygomaturine genus and species from the Late Miocene Alcoota local fauna, Northern Territory (Marsupialia: Dip- rotodontidae). 1 9 21 29 35 39 53 HUMANITIES GUNN, M. - A brief note on the Malagan Curvunavunga from Tabar, New Ireland Province, Papua New Guinea... SPILLETT, P. - Wehiriya - a missing grave and a lost town. BOOK REVIEW - Dragonflies (Naturalists' Handbooks; 7) 95 HANLEY, J.R. SCIENCES The Beagle, Records of the Northern Territory Museum of Arts and Sciences, 1990 7(2):l-8 A SECOND SPECIES OF B RES I LI A, B. PLUMIFERA SP. NOV., NEW TO THE AUSTRALIAN FAUNA (CRUSTACEA: DECAPODA: BRESILIIDAE). A.J. BRUCE Northern Territory Museum of Arts and Sciences GPO Box 4646, Darwin NT 0801 , Australia. ABSTRACT A new species of Bresilia, B. plumifera (Crustacea: Decapoda: Bresiliidae), is described and illustrated, increasing to four the number of species now known in this genus. The single specimen, the smallest in the genus (total length about 6 mm), was collected from the western Tasman Sea, at only 133 m, the most shallow depth from which this deep-sea genus has been recorded. Keywords: Crustacea, Decapoda, Bresiliidae, Bresilia, new species, Tasman Sea. INTRODUCTION The small genus Bresilia was first described by Caiman (1896) from a single specimen collected off Ireland at 1372 m depth. Some ninety years passed before a further specimen of this genus was reported, when Forest and Cals (1977) described B. corstcana, again based on a single specimen from the western Mediterranean Sea, at a depth of 450 m. Recently, B. antipodarum has been recorded from off eastern Tasmania, at 800 m, by Bruce (1990). The discovery of a single specimen of another species of Bresilia in eastern Austra¬ lian waters increases the size of the genus from three to four species, with two in the northern hemisphere and two at nearly the opposite pole of the southern hemisphere. All species are still known only from the single holotype specimens. All are very small shrimps and doubtless are easily overlooked in trawl or dredge catches. Carapace length refers to the postorbital carapace length. Type material is deposited with the Australian Museum, Sydney (AM). SYSTEMATICS Genus Bresilia Caiman Bresilia plumifera sp. nov. (Figs 1-4) Type material. HOLOTYPE - AM P40086: 1 female, Taupo Sea Mount, western Tasman Sea, 33° 14.-2US., 56° 10.68’E, 133 m, sled, 2 May 1989, coll. J.K. Lowry etal. F.R.V. Fran¬ klin, stn. FR0589-7. Description. A small-sized, slenderly built shrimp, of generally subcylindrical body form, lacking parts of antennae, second to fifth pereiopods, parts of pleopods, with distal caudal fan damaged. Carapace slightly compressed, smooth; ros¬ trum well developed, acute, about 0.3 of cara¬ pace length, straight, slightly shorter than proximal segment of antennular peduncle, dorsal carina distinct, shallow, with 4 small acute teeth, first tooth on carapace proximal to posterior orbital margin, second tooth over posterior orbital margin, distal teeth at about 0.4 and 0.8 of rostral length; ventral carina distinct, shallow, ventral border convex, un¬ armed, non-setose; lateral carinae distinct, broadened posteriorly, expanded to form or¬ bital rim; inferior orbital angle acute, anterior margin of branchiostegite deeply convex, pterygostomial angle acutely produced; su¬ praorbital, hepatic, and antennal spines ab¬ sent. Abdomen smooth; third segment strongly produced posterodorsally; sixth segment about 2.2 times length of fifth, 2.0 times lodger than deep, compressed, posterolateral angle produced; acute, posteroventral angle small, subacute; pleura of first three segments ex¬ panded, broadly rounded; fourth posteriorly produced, rounded; fifth angular, with small posteroventral tooth. Caudal fan damaged, 1 A.J. Bruce Fig. 1 . Bresilia plumifera sp. nov., female holotype, Tasman Sea. Scale divisions in 0.5mm. lacking distal telson and distal rami of uro- pods. Uropod with distolateral angle of proto- podite acute. Eye with rounded, pigmented cornea, obliquely hemispherical; stalk subcylindrical, slightly inflated proximally, without dorsal tooth. Antennule with proximal segment of peduncle broad, about 1.3 times longer than wide, stylocerite broad, acute, not reaching distal margin of segment, medial border setose, with small ventral tooth; statocyst obsolete; distal segments and flagella lacking. Antenna with basicerite robust, with small distal dorsolateral tooth, larger ventrolateral tooth and strong ventral medial tooth; ischio- cerite bearing elongate plumose setae, mero- cerite normal; scaphocerites and flagella lack¬ ing. Epistome with long slender, forwardly di¬ rected acute median process. Mandible with feeble corpus; palp well developed, 2-segmented, proximal segment subcylindrical, distal segment suboval, flat¬ tened, with numerous short plumose or serru¬ late spiniform setae; molar process feeble, slender, slightly expanded and compressed distally, with numerous slender acute mar¬ ginal teeth; incisor process broad, anterior margin with distinct knob-like process disto- dorsally, cutting edge with 10 acute teeth, anteriormost tooth largest. Maxillula with slender, curved palp, with 2 unequal, pre¬ terminal setae; upper lacinia broad, with 12 stout, dentate marginal spines, largest ven- trally, with submarginal row of 11 more slen¬ der serrulate spines and proximal row of 6 slender serrulate spines, dorsal and ventral margins with setulosc setae; lower lacinia slender, with 5 long stout serrulate spines distally, setulose setae ventrally. Maxilla with elongate tapering palp, distal half subcylindri¬ cal, styliform, with few short proximomedial setae; basal endite bilobed, distal lobe subcir¬ cular, proximal lobe more elongate, both sparsely setose, with long finely setulose se¬ tae; coxal endite simple, similarly setose, with scaphognathite well developed, about 4.3 times longer than wide, posterior lobe narrow, with elongate posterior marginal setae, ante¬ rior lobe large, broad, 2.0 times longer than broad, rounded. First maxilliped with small simple palp, with about 6 plumose setae dis- tomedialiy; basal endite small, broad, sparsely setose medially, coxal endite similar, smaller, exopod with small, narrow caridean lobe, with sparse stout plumose marginal setae, flagel¬ lum slender, with several short plumose setae distally; epipod bilobed, elongate, narrow. Second maxilliped with endopod stout, dac- 2 Bresilia plumifera sp. nov., from the Tasman Sea tylar segment terminal, about 1.6 times longer than wide, distally rounded, with numerous long serrulate marginal spines, oblique trans¬ verse row of similar submarginal spines; pro¬ pod about 2.0 times dactyl length, 2.5 times longer than wide, sparsely setose; carpus short, unarmed; ischiomerus 2.2 times longer than wide, 1.2 times longer than propod, sparsely setose medially; basis subequal to ischiomeral length, exopod with long slender Fie 2 Bresilia plumifera sp. nov., female holotype. A, anterior carapace, rostrum and eye, right antennae removed, epistome and cpistomal spine indicated by dashed line. B, antennule, proximal segment of peduncle. C, basicerite, ventrolateral aspect. D, first pereiopod. E, same, chela. F, first pleopod. G, fifth to eight thoracic sternites; above, epistomal spine. 3 A.J. Bruce Fig. 3. Brasilia plumifera sp. nov., female holotype. A, mandible. B, maxillula. C, maxilla. D, first maxilliped. t, second maxilliped. F, third maxilliped. 4 Bresilia plumifera sp. nov., from the Tasman Sea flagellum with numerous short plumose setae distally; coxa medially rounded, with single seta, with reduced rounded epipod laterally. Third maxilliped with long slender endopod. exceeding proximal segment of antennular peduncle by length of distal segment; antepe¬ nultimate segment about 9.0 times longer than central width, tapering feebly distally, sparsely setose medially, with single elongate plumose seta at distodorsal angle, penultimate segment 0.3 of proximal segment length, 4.3 times longer than proximal width, feebly ta¬ pering distally, with 3 transverse rows of short serrulate spines along distal half of ventral border; terminal segment slender, tapering, about 9.0 times longer than proximal width, with stout subterminal distal spine, 0.4 of segment length, with 7 transverse rows of serrulate spines on ventral margin; basis dis¬ tinct from ischiomerus, short, with slender plumose setae medially, exopod well devel¬ oped, flagellum (lacking on right) slender, with short plumose setae distally, reaching to about 0.65 of ischiomeral length; coxa robust, with 5 long plumose setae medially, small globular epipod dorsolaterally. Thoracic sternites narrow, first to fourth unarmed, fifth with submedian pair of large divergent triangular teeth, sixth and seventh with slender pairs of acute subparallel teeth, eighth with blunt median process, feebly di¬ lated and flattened distoventrally. First pereiopod present on right side only, well developed, exceeding basicerite by car¬ pus and chela; chela about 0.6 of carapace length, palm oval in section, smooth, about 2.0 times longer than central width, increasing slightly in width distally, distodorsal margin with small acute tooth, proximal ventral mar¬ gin with three transverse rows of short serru¬ late spines with single larger stout simple spine at 0.5 of ventral border length; fingers strongly deflexed; dactyl about 0.75 of palm length, slightly exceeding fixed finger, with slightly upturned blunt tip, about 3.5 times longer than proximal depth, deeply concave laterally, dorsal margin convex, carinate, with 5 small acute denticles, ventromedial margin concave, with 9 small spinules, ventrolateral margin straight, strongly dentate, with about 40 strongly compressed teeth, largest centrally, decreasing strongly in size distally; fixed fin¬ ger about 2.0 times longer than proximal width, deeply concave dorsolaterally, taper¬ ing strongly distally, cutting edge straight, with series of deep fossae laterally for recep¬ tion of dactylar teeth, medially with a row of stout simple spines, about 35, becoming shorter, stouter distally, extending to tip of fixed finger; carpus short, stout, as wide dis¬ tally as long, slightly expanded, excavate dis¬ tally, with small dorsal lobe with single short stout simple spine, rounded lateral lobe, small ventral lobe with strong simple medial and lateral spines, separated by four slender serru¬ late spines; merus about 0.8 of palm length, 3.4 times longer than central depth, unarmed; ischium obliquely articulated with merus, subequal to palm length, 3.5 times longer than distal width, tapering feebly ventrally, distov- entral angle produced, with small acute tooth, ventral border sparsely setose; basis short, stout, without special features, exopod well developed, flagellum (lacking on right) slen¬ der, reaching to proximal merus, with numer¬ ous plumose setae distally; coxa normal, with rounded dorsolateral protuberance (epipod?). Second of fifth pereiopods lacking; well developed exopod on basis of second pereio¬ pod, (lacking on right) shorter than first pereiopod exopod; posterior pereiopods with¬ out exopods. Pleurobranchs present on first four pereio¬ pods only, decreasing strongly in size anteri¬ orly. Pleopods damaged; first pleopod with ba- sipodite 1.75 times longer than wide, with 6 elongate distally plumose setae medially, with inner row of 6 ovigerous setae; exopod lack¬ ing; endopod small, about 0.45 of basipodite length, tapering distally, with about 10-12 plumose marginal setae. Second pleopod lack¬ ing exopod and endopod. Posterior pleopods without special features, all lacking endopods. Measurements (mm). Total length (ap¬ prox.), 6.0 +; carapace and rostrum, 2.5; cara¬ pace, 1.8; first pereiopod chela, 1.2. Associated fauna. One small badly dam¬ aged hippolytid shrimp, unidentifiable at genus level. Etymology. From pluma, Latin, a feather, and fero, Latin, to carry or bear. Systematic Position. Bresilia plumifera is most closely related to the only other Austra¬ lian species of the genus, B. antipodarum Bruce, which it particularly resembles in the presence of the postero-dorsally produced third abdominal segment, a feature not found 5 A.J. Bruce Fig. 4. Bresilia plumifera sp. nov., female holotype. A, mandible, incisor process. B. maxillula, palp. C, third maxilliped, tip of distal segment of endopod. D, first pereiopod, chela, fingers. E, same, detail of cutting edges. in the two other species of the genus. Bresilia plumifera may be distinguished from B. an- tipodarum by the presence of only four dorsal rostral teeth, with the ventral border of the rostrum unarmed, whereas the latter has nine dorsal rostral teeth and a small ventral tooth. The rostrum is also much shorter in B. plumif¬ era, much less than half the carapace length, whereas it is well over half in B. antipodarum. Bresilia antipodarum also has the antennal tooth and pterygoslomial angle more acutely produced than in B. plumifera and the postero- dorsal enlargement of the third abdominal segment is larger and more compressed. In B. plumifera the proximal segment of the anten- nular peduncle is much broader than in B. antipodarum, with the stylocerite stouter, not reaching the level of the distal end of the segment. The mouth parts of the two species are essentially similar, but in B. plumifera. the incisor process of the mandible is much broader and more strongly dentate, the molar process more strongly provided with simple teeth and the distal segment of the palp with serrulate spiniform setae; the upper laciniaof the maxillula is also much more strongly den¬ tate. The first pereiopods of the two species are remarkably similar, but B. antipodarum lacks the ventral palmar spine present in B. plumifera, and the small distodorsal palmar tooth. Also lacking is the distoventral tooth on the ischium. In B. plumifera the dorsal margin of the dactylus is carinate, with five small acute teeth, whereas in B. antipodarum, this surface is rounded and unarmed. The features of B. plumifera that are without parallel in other species of Bresilia so far described, are the medial epistomal process and the long plumose setae on the ischioccrite and the distal antepenultimate segment of the endopod of the third maxilliped. Both Australian species of Bresilia are char¬ acterized by the presence of a distinct flagel¬ lum on the exopod of the first maxilliped, a character in which they also resemble Encan- tada, and of a distinct posterodorsal carina on the third abdominal segment, features not present in the northern hemisphere species. It is possible that they should be considered as generic-ally distinct, but this is best postponed until such time as intact specimens are avail¬ able. DISCUSSION The posterior thoracic sternites of Bresilia plumifera show a great similarity to those of B. 6 Bresilia plumifera sp. nov., from the Tasman Sea corsicana (Forest and Cals 1977: fig. 20). Forest and Cals (1977) noted that the interpre¬ tation of these processes presented difficul¬ ties, and this applies equally to B. plumifera, at only half the size of B. corsicana , and cannot be elucidated further without destroying the unique specimen. The function of the plumose setae on the antennular ischiocerite and the endopod of the third maxilliped, which probably functions as a complex in association with the median epis- tomal process, is not readily apparent. Appar¬ ently unique to B. plumifera , it seems unlikely that these setae would have been overlooked, particularly in the detailed study by Forest and Cals (1977), but it is possible that they could have been lost during capture of the speci¬ mens. This could not have happened in the case of the epistomal process. The four species of the genus Bresilia are known only from the holotype specimens, incomplete in the case of B. antipodarum and B. plumifera. All are small shrimps, with B. atlantica the largest, with a total body length of 27 mm and B. plumifera the smallest, with a length of about 6 mm. Their small size probably contributes to the paucity of records of these species. Their zoogeography is clearly incompletely known (Fig. 5) and their bathymetric range is now known to extend from 1375 m to 133 m. Bresilia antipodarum is also known from the Tasman Sea, off Cape Freycinet, eastern Tasmania (Bruce 1990). Most of the bresiliiform species of the Bre- siliidae are deep-water species and the only other species so far reported from shallow waters is Encantada spinoculata Wicksten (1989) from 55-92 m depth. The four species of Bresilia may be distin¬ guished by the following key: 1. Third abdominal segment posterodorsally produced; first maxilliped endopod with flagellum .2 Third abdominal segment not posterodor¬ sally produced; first maxilliped endopod without flagellum...3 2. First pereiopod with posteroventral angle unarmed; posterodorsal process of third ab¬ dominal segment compressed; rostrum with nine dorsal and one ventral tooth. . B. antipodarum Bruce First pereiopod with posteroventral angle dentate, dactyl dorsally carinate, dentate; posterodorsal process of third abdominal segment not strongly compressed; rostrum with four dorsal teeth, ventrally unarmed . B. plumifera sp. nov. 3. Posterior margin of telson truncate, with 6- 7 pairs of spines. B. atlantica Caiman Posterior margin of telson with median notch, with four pairs of spines and four submedian spinules... . B. corsicana Forest and Cals 7 A.J. Bruce ACKNOWLEDGMENTS I am most grateful to Dr. J.K. Lowry for the opportunity to examine this specimen from the collections of the Australian Museum. REFERENCES Bruce, A.J. 1990. Two deep-sea shrimps new to the Australian fauna - Psathyrocaris hawaiiensis Rathbun (Pasiphaeidae) and Bresilia antipo- darum sp. nov. (Bresiliidae). with remarks in Encantada spinoculata Wicksten (Bresi¬ liidae). Invertebrate Taxonomy 4(4):847-866. Caiman, W.T. 1896. On some Deep-sea Crustacea from the South West of Ireland. Transactions of the Royal Irish Academy 31(l):l-20, pis 1- 2 . Forest, J., and Cals, P. 1977. Une deuxieme espece du genre Bresilia Caiman, B. corsicana sp. nov. Comparaison avec B. atlantica Caiman (Crustacea Decapoda Bresiliidae). Bulletin du Museum National d'Histoire Naturelle (3) 453, mars-avril, Zoologie 316:549-565. Wicksten, M. 1989. Encantada spinoculata , a new genus and species of shrimp from the Galapa¬ gos Islands (Caridea:Bresiliidae). Journal of Crustacean Biology 9(4):667-671 . Accepted 16 August 1990 8 The Beagle, Records of the Northern Territory Museum of Arts and Sciences, 1990 7(2):9-20 RECENT ADDITIONS TO THE PONTONIINE SHRIMP FAUNA OF AUSTRALIA. A.J. BRUCE Northern Territory Museum of Arts and Sciences GPO Box 4646, Darwin NT 0801, Australia. ABSTRACT Recent additions to the pontoniine shrimp fauna of Australia are reviewed and data are provided on seven species not previously known from Australia: Onycocaris spinosa Fujino and Miyake, Periclimenes mahei Bruce, Platypontonia hrevirostris (Miers), Pontonia stylirostris Holthuis, Tuleariocaris holtluiisi Hipeau-Jacquotte, Vir orientalis (Dana) and V. philippinensis Bruce and Svoboda. Recent nomencla- tural amendments are included. The number of species presently known is increased from 136 to 168 and their distributions and zoogeography are discussed. Keywords: Crustacea: Decapoda: Palaemonidae, Australian fauna, recent addi¬ tions, new records, zoogeography, Indo-West Pacific. INTRODUCTION In 1983, Bruce (1983a) provided a review on the occurrence of 136 species of pontoniine shrimp in the seas around Australia, described up to 1980. Since that publication, three of the species, of the genus Anchistioides, have been transferred to the resurrected family Anchis- tioididae Borradaile, and two species, of Gnathophylloides, have been transferred from the family Gnathophyllidae Dana. Two of the species of Periclimenaeus have now been placed in separate monospecific genera. In addition, 21 new species of pontoniine shrimp have been described from Australian waters, including four new monospecific genera. Seven species previously recorded from the Indo-West Pacific have also been found in Australian waters, and a further seven species are reported here for the first time. The pos¬ sible occurrence of Coralliocaris macroph- thaltna (H. Milne-Edwards) has been reported from Heron Island, Queensland (Bruce 1977), but because of its dubious status, it is not considered further in this review. These addi¬ tions increase the total number of species known from Australia to 168. Although a rich and diverse fauna is present in the seas around Australia, comparison with other regions is difficult as comparable studies in many areas have not yet been carried out. Although detailed studies of the Indonesian fauna have been made through the activities of the Siboga and Snellius expeditions (1899- 1900, 1929-1930), these were carried out before the common use of SCUBA equipment. Undoubtedly many of the recently discovered tropical Australian species will be found to also occur in Indonesian waters in due course, probably together with much that is com¬ pletely new. The pontoniine fauna of the Phil¬ ippines is not well known, and that of Papua- New Guinea even less. All these areas, with their extensive reef systems, the primary habi¬ tat for most pontoniine shrimps, may be ex¬ pected to have a pontoniine fauna at least as diverse as that of Australia. The enumeration of species follows the list¬ ing given in Bruce (1983a), after adjustment for the alterations mentioned above. Species with numbers represent new records; species without numbers (cited as -) represent nomen- clatural changes. ANNOTATED SPECIES LIST Genus Apopontonia Bruce, 1976 134. Apopontonia dubia Bruce, 1981: This species was originally described from a single female specimen from 20m depth at Shag Rock, off North Stradbroke Island, Queens¬ land, collected in 1979 in association with a 9 AJ. Bruce sponge, Ircinia cf. echinata (Keller) (Bruce 1981). It has subsequently reported from Heron Island, Queensland (Bruce 1983f); recently from 43 m depth, off New Caledonia. The genus contains only two other species, A. falcirostris Bruce, known only from the type specimen from 73 m depth off Madagascar, and the following species. 135. Apopontonia tridentata Bruce, 1988: A single ovigerous female of this species was collected in 1983 from the Northwest Shelf from 54 m depth at 19° 41.9’S, 117° 57.15E (not 17° 57.15’E, as given in the original de¬ scription (Bruce 1988b)). There have been no further reports of this species, whose host remains unidentified, but is most probably a sponge. Genus Carinopontonia Bruce, 1988 136. Carinopontonia paucipes Bruce, 1988: The only known specimen of this unusual shrimp, unfortunately lacking many of its appendages, was collected from 83 m depth on the Northwest Shelf at 19° 04.4’S, 118° 47.55’E in 1983 (Bruce 1988b). The host of this species is unknown, but it is certainly a commensal species and a coelenterate associa¬ tion seems highly likely. Genus Chernocaris Johnson, 1967 137. Chernocaris placunae Johnson, 1967: Twelve specimens were collected in 1981, from Placuna placenta from 27 m depth in the Arafura Sea at 12° 58.0’S, 132° 10.0’E (Bruce 1983d). This is only the second recorded oc¬ currence of this species, otherwise known only from Singapore. Genus Conchodytes Peters, 1875 - Conchodytes kempi Bruce, 1989e: This species was referred to in the list given by Bruce (1983a) under the name of C. biun- guiculatus Paulson 1875. There is still doubt as to the distribution of Paulson’s species and whether or not it occurs in Australian waters. Conchodytes kempi has been reported from the Andaman Islands and the Philippines, and it seems likely that Australian specimens will belong to this species. Further study of Con¬ chodytes specimens from pinnid hosts will be necessary belore the general distribution of these shrimps can be satisfactorily clarified. 138. Conchodytes maculatus Bruce, 1989: A pair of specimens and a single male speci¬ men were first collected from pearl oysters, Pinctada maxima , from about 40 m depth in the Arafura Sea, at about 16° 33.5’S, 121° 25.75’E in 1985 (Bruce 1989c). There are no other known specimens of this species. Genus Dasella Lebour, 1945 139. Dasella ansoni Bruce, 1990d. A male and ovigerous female were found together in an ascidian host, Phallusia depressiuscula Heller, from 27 m depth in the Arafura Sea, at 12° 58.0’S, 132° 10.0’E collected in 1981. (Bruce 1983b). No further occurrences have been reported. - Dasella brucei Berggren, 1990: A speci¬ men of this species was initially reported by Bruce (1981 a), who recorded a single male in the ascidian Herdmania momus from Heron Island, Queensland, as D. herdmaniae , from 14 m depth. This species has been recently recognized as a distinct species, so far known from type material only, by Berggren (1990). Dasella herdmaniae is known only from southern India and Mozambique. Genus Epipontonia Bruce, 1977 140. Epipontonia anceps Bruce, 1983: First described from four specimens from Heron Island, southern Great Barrier Reef from 18 m depth collected in 1979 from a sponge host, Dysidea sp. (Bruce 1983c). There have been no further records of this species. The genus contains one other species, £. spongicola Bruce, 1977 known only from Kenya. Genus Exopontonia Bruce, 1988 141. Exopontonia malleatrix Bruce, 1988: A single ovigerous female specimen of this species was collected from pulverized coral rubble from intertidal reef flat at Ashmore Reef, Timor Sea in 1987 (Bruce 1988a). There are no other records of the species, whose host is unknown, but which may be an associate of sponges. Genus Gnathophylloides Schmitt, 1933 142. Gnathophylloides mineri Schmitt, 1933: First recorded in Australian waters from the North Solitary Islands, New South Wales, from 12 m depth, in association with the ur¬ chin Tripneustesgratilla (L.) collected in 1988 (Bruce 1989a). Originally described from Caribbean specimens, this species is now also known from Malpelo Island, Colombia; Ha- 10 Additions to Pontiniine shrimp fauna waii, Tonga, Zanzibar and the Seychelle Is¬ lands, and is therefore one of the few Austra¬ lian shrimps that have a circumtropical distri¬ bution. The Australian records also represent the most southerly extent of the distribution of this species. 143. Gnathophylloides robustus Bruce, 1973: This species, known only from the type material, a male and three ovigerous females, was found on the sea-urchin, Centrostephanus tenuispinus H.L. Clark at 3 m depth off Point Moore, Geraldton, Western Australia (Bruce 1973). There have been no subsequent records of this shrimp. Genus Hamopontonia Bruce, 1970 144. Hamopontonia essingtoni Bruce, 1987: Described from 13 specimens from Coral Bay, Port Essington, Cobourg Penin¬ sula, in association with a coral host, Stylo- phora pistil lata (Esper) collected in 1985 from 6 m depth (Bruce 1987a). No further records have been reported. Genus Miopontonia Bruce, 1985 145. Miopontoniayongei Bruce, 1985: This species is known only from a male and an ovigerous female collected from 40 m and 80 m depth on the Northwest Shelf, by the FRV Soela in 1983, from stations at 19" 29.0’S, 118° 52.0’E and 19° 04.3’S, 118° 55.5’E re¬ spectively (Bruce 1985). The host animals were not identified but are considered most likely to be gorgonians or antipatharians. The genus Miopontonia is monospecific and most closely related to the Caribbean genera Couti- erea Nobili and Pseudocoutierea Holthuis. Genus Onycocaridites Bruce, 1987 146. Onycocaridites anomodactylus Bruce, 1987: The type material and only known speci¬ mens of this species, a damaged male and female, were obtained from an unidentified sponge from 60 m depth in the Arafura Sea, at 10° 40’S, 133° 50’E in 1986 (Bruce 1987c). Genus Onycocaris Nobili, 1904 147. Onycocaris spinosa Fujino and Miyake, 1969: Not previously reported from Australian waters, a male and two ovigerous females were collected from a depth of 10 m depth in a blue sponge on the reef slope off Heron Island, Queensland, in 1980 by R. Boer. Originally described from Yoron-jima, Ryukyu Islands, and subsequently from Ok¬ inawa, there have been no other records of this species. Genus Orthopontonia Bruce, 1982 - Orthopontonia ornatus (Bruce, 1969): This species was referred to in the previous report (Bruce 1983a) under the name of Peri- climenaeus ornatus. Orthopontonia is a mono- specific genus and O. ornatus occurs from East Africa to Australia. The species is an associate of sponge hosts. Genus Periclimenaeus Borradaile, 1915 148. Periclimenaeus orontes Bruce, 1987: Known only from a single ovigerous female specimen, collected on Orontes Reef off Port Essington, Cobourg Peninsula, from a sponge host Jaspis stellifera (Carter), from a depth of 3 m depth in 1982 (Bruce 1987a). There have been no subsequent records of this shrimp. 149. Periclimenes agag Kemp, 1922: Bruce (1990d) recently examined specimens of this species from Lizard Island, Queensland, from baited traps set at Coconut Beach, Watson’s Bay and off Eagle Island at 1-17 m depth collected by S. Keable in 1988. There have been no previous records of this species from Australian waters. Originally reported from the Andaman Islands by Kemp (1922), and subsequently from the Red Sea, New Caledo¬ nia and the Marshall Islands, the species is probably a free-living micropredator. 150. Periclimenes alegrias Bruce, 1987: First found at Coral Bay, Port Essington, Cobourg Peninsula, and subsequently at North-West Vernon Island, this species is known from only four specimens, from 2-4 m and 8 m depth in association with crinoid hosts, Stephanometra spicata (Carpenter), Lamprometrapalmata (Muller) and Comanth- ina variabilis (Bell) (Bruce 1987a). There have been no further reports of this species. 151. Periclimenes anacanthus Bruce, 1989: A dozen specimens of this species were first collected from Zostera beds in southern More- ton Bay, Queensland, in 1987 (Bruce 1989b). There have been no subsequent reports of this species, which is probably free-living. 152. Periclimenes andamanensis Kemp, 1922: Shrimps of a taxon resembling this species were reported from shallow-water sea- grass beds throughout Moreton Bay, Queens¬ land, by Wadley (1978), as Periclimenes nr. 11 A.J. Bruce andamanensis. Some specimens also referred to this taxon by Wadley, from 3m depth over Posidonia beds from the South West Arm, Port Hacking, New South Wales, have also been examined, but unfortunately no second pereiopods have been preserved, so the identi¬ fication can not be fully confirmed. 153. Periclimenes darwiniensis Bruce, 1987: First discovered at Weed Reef, Darwin Harbour, Northern Territory, in 1985, this species is now known to be moderately com¬ mon in intertidal pools in Darwin Harbour (Bruce 1987b) and also occurs in the Gulf of Carpentaria. The species is apparently free- living, and presumably a micropredator. 154. Periclimenes denticulatus Nobili, 1906: A single male specimen from Lizard Island, Queensland, lacking both second pereipods, has been referred to this species. The specimen was caught in a trap at 37 m depth (Bruce, 1991a). The species was origi¬ nally described from Gatavake. Gambier Is¬ lands, and subsequently reported only from the Tuamotu and Marshall Islands and the northern South China Sea. - Periclimenes gonioporae Bruce, 1989d: This species was included as a nomen nudum in Bruce (1983a). A description of this species has been recently published, thereby validat¬ ing the use of this name (Bruce 1989d). Peri¬ climenes gonioporae is an associate of Lo- bophyllia and scleractinian corals ( Goni- opora, Galaxea, Porites and Montipora). 155. Periclimenes franklini Bruce, 1990b: A deep-water species, collected from about 300 m. at 17° 21’S, 146° 48.52’E in the Coral Sea in 1986. This species, whose host has not been identified, is thought to be probably a gorgonian associate (Bruce 1990b). 156. Periclimenes laccadivensis (Alcock and Anderson, 1894): A single female speci¬ men has recently been caught at 720 m depth off Cape Freycinet, Tasmania (Bruce 1991a), and represents one of the few deep-water and southerly components of the Australian pon- toniine fauna. Previously recorded from the Laccadive Islands, South China Sea, and off Hawaii, it is the most southerly occurrence of a pontoniine shrimp with an otherwise Indo- West Pacific distribution. The species is possi¬ bly associated with gorgonians. 157. Periclimenes sp. nov. a - Bruce, 1991a: A single specimen of a new species has been recorded (currently in press) from a trap at 23 m depth from Chinaman’s Ridge, Watson’s Bay, Lizard Island, Queensland (Bruce 1991a). This species is known only from a single occurrence, and is thought to be of free- living habits. 158. Periclimenes mahei Bruce, 1969: Not previously reported from Australian seas, numerous specimens including ovigerous females were collected from a colony of Pocil- lopora damicornis from Point Quobba, West¬ ern Australia, in 1980. This species has been previously recorded only from the Seychelle Islands, Comoro Islands and Zanzibar. 159. Periclimenes sp. nov. b - Bruce, 1991a: This species (in press) known only from the holotype specimen collected in a trap at 3.5 m depth at Blue Lagoon, Lizard Island, Queen- land (Bruce 1991a). It is probably a commen¬ sal species, possibly associated with echino- derrns (echinoids?). 160. Periclimenes venustus Bruce, 1990d: Numerous examples of this species were col¬ lected from Port Essington, Cobourg Penin¬ sula in 1985, mainly from unidentified anemo¬ nes. but with two lots from Heliofungia actini- formis, mostly from unrecorded depths, but some from 2.5 -3m depth (Bruce 1991b). Except for its colour pattern, this species is closely similar to P. holthuisi Bruce, and it is likely that some of the specimens previously identified as P. holthuisi , particularly where the original colour patterns were not recorded, will prove to belong to P. venustus on re¬ examination. Periclimenes venustus is also known from Scott Reef, Western Australia and the Philippines (Bruce 1989e). 161. Periclimenesyaldwyni Holthuis, 1959: Known only at present from one male and three ovigerous females collected in 1983 from 16-28m depth in Mercury Passage, off Maria Island, Tasmania (Bruce and Kropp 1984). This species was first described from New Zealand material (as Brachycarpus audouini Bate 1888), and is one of the small number of carideans occurring in both south¬ ern Australian and New Zealand waters. Genus Periclimenoides Bruce, 1990c - Periclimenoides odontodactylus (Fujino and Miyake, 1968): This species was referred to in the species list given by Bruce (1983a) under the name of Periclimenaeus odontodac¬ tylus. The genus is monospecific, and P. odon¬ todactylus has been reported from Japan, 12 Additions to Pontiniine shrimp fauna Hong Kong and Australian waters only, in association with sponge hosts, (Bruce 1990c). Genus Platypontonia Bruce, 1968 162. Platypontonia brevirostris (Miers, 1884): This species has not been recorded previously from Australian waters. A pair of specimens were collected by R. Boer from a specimen of Lopha cristigalli from 20 m depth on the reef slope of Heron Island, Queensland in 1980. This species was first described from specimens from the Seychelle Islands, col¬ lected by the H.M.S. Alert Expedition (1882), with further material subsequently also col¬ lected from the Seychelle Islands. There have been no reports of the species from other localities. Genus Pontonia Costa, 1844 163. Pontonia stylirostris Holthius, 1952: This species has not been previously recorded from Australian waters. A single incomplete male specimen, with the characteristic ros¬ trum and minor second pereiopod, was found amongst coral rubble from 42 m depth off Townsville, Queensland, at 18° 42’S,. 147° 1,0’E by R.A. Birtles in 1979. This species has only been previously reported from Indonesia, between Misool and New Guinea, at 32 m depth, and from Tanganyika and Oman. Its associations have not been identified, but are likely to be ascidians. Genus Tuleariocaris Hipeau-Jacquotte, 1965 164. Tuleariocaris holthuisi Hipeau-Jac¬ quotte 1965: Not previously recorded from Australian waters, a pair of specimens were found on Diadema setosun Leske, at Myora Light, North Stradbroke Island, at 0.5 m depth in September 1967. Genus Typton Costa, 1844 165. Typton dimorphus Bruce, 1986: A male and ovigerous female, collected from coral rubble, although not necessarily a natural pair, were found in a sample from 5 m depth on Ashmore Reef, Timor Sea, at 12" 15.0 S, 123° 00.0’E in 1984 (Bruce 1986). No other speci¬ mens are known. Like other Typton species, T . dimorphus is likely also to be a sponge associ¬ ate. 166. Typton nanus Bruce, 1988: Typton nanus is known only from the female holotype specimen, collected in 1985, from 40-46 m depth on the Northwest Shelf (Bruce 1987c). Its host remains unidentified, but is presuma¬ bly a sponge. This shrimp is one of the small¬ est pontoniine shrimps known, with a postor¬ bital carapace length of only about 1.0 mm. Genus Vir Holthius, 1952 167. Vir orientalis (Dana, 1852): A single example of this species, the first recorded from Australian waters, was collected by N. Coleman in 1979 from 10 m depth from Os¬ prey Reef, Coral Sea, from an unidentified coral host. The species was first described from the Sulu Sea and was subsequently re¬ ported from Kenya, Zanzibar, Seychelle Is¬ lands, Andaman Islands, Mariana Islands and the South China Sea. 168. Vir philippinensis Bruce and Svoboda, 1984: This species has not been recorded pre¬ viously from Australian waters. Two speci¬ mens were collected from Pleurogyra sinuosa from 8 m depth by J..E.N. Veron, from Fitzroy Reef, Queensland, in 1980. A further speci¬ men was collected by T. Fromm, from Euphyl- lia sp., off Lizard Island, Queensland, in 1987. The species was first described from speci¬ mens from Cebu, Philippine Islands, and it has since been reported from Okinawa, Ryukyu Islands. DISCUSSION At present, 168 species of pontoniine shrimp, of 46 different genera, are known to occur in Australian waters. Four of these gen¬ era, all monospecific, are not yet known to occur outside Australian seas. Of the remain¬ ing 42 genera, 17 are also monospecific. As some 308 species have been reported from the Indo-Wcst Pacific region as a whole, about half (54.2%) may be found in Australian wa¬ ters, and 86% of the known genera are also represented in the Australian fauna. Of the 168 Australian species, 141 occur on the eastern seaboard, with 49 on the northern and western, and only 6 on the southern coasts. The high figure for the eastern coasts is proba¬ bly a reflection of the presence of the Great Barrier Reef combined with the increased sci¬ entific collecting effort that it has attracted over many years. Only one species, Anchistus custos, an associate of pinnid molluscs, occurs in all three regions. Twenty-four species (14.3%) have so far been found to occur in both eastern and north-western regions. Con- 13 A.J. Bruce Fig. 1. Subdivision of Indo-Pacific zoogeographic zones, relating to areas described in Table I. A, Red Sea. B, East Africa. C, Southern Arabia, Arabian Gulf, Gulf of Oman. I). Madagascar, Comoro, Mascarene and Seychelle Islands. E, Western Indian peninsula, Maidive, Laccadive and Chagos Islands. F, Sri Lanka, Bay of Bengal, Andaman and Nicobar Islands, Burma. G, Malaya, South China Sea, Taiwan. Philippines. H. Indonesia. I, Australia (e, eastern: nw, north-western: s, southern). J, east China, Japan, Ryukyu Islands, Korea. K, Marshall, Caroline and Marianas Islands. L, Papua New Guinea. M, New Caledonia, Fiji, Vanuatu, Tonga, Kiribati, Tuvulu, Samoan and Phoenix Islands. N, Tokelau, Cook, Line, Society, Austral, Tuamotu and Marquesas Islands. O, Hawaiian Islands. P, Eastern Pacific region. versely, 115 species (168%), including 13 “endemics” (7.7%), are known from the east¬ ern coasts only and 23 species, including 18 “endemics”, occur only on the north-western coasts. Many of the Australian species belong to taxa with very wide Indo-West Pacific distri¬ butions. Thirty two species are also known from the Red Sea, while 80 (48.2%) species are known from the east coast of Africa, with 82 (49.4%) from the combined Red Sea-East African region. The central East African re¬ gion, principally Kenya and Tanzania, has a pontoniine shrimp fauna of 134 species (Bruce 1976), of which 80 species (60%) have so far been found to occur around Australia. The pontoniine fauna of the Pacific Ocean has not yet been adequately examined. The faunas of the extensive reef systems and deeper waters have not yet been generally studied in detail, except in the region of New Caledonia. Diver¬ sity appears to decrease in an easterly direc¬ tion and only 19 species found in Australian waters occur also in the eastern parts of the Indo-West Pacific region (Fig. 1, zones N,0,) but several Indo-West Pacific species have ranges that extend to the eastern American seabord. Of the species presently known to do so, six occur in Australian waters ( Allopon- tonia iaini, Fennera chacei, Gnalhophylloi- des mineri, Harpiliopsis depressa , H. spini- gera and Periclimenes soror). These species, which are all commensal associates of other marine invertebrates, also occur in East Afri¬ can waters and therefore have ranges that extend across the whole Indo-Pacific region, but only G. mineri has so far been found in the Atlantic region, although Palacmonella rotu- mana has now extended its range through the Suez Canal into the eastern Mediterranean Sea (Holthuis and Gottlieb 1958). No free-living Indo-West Pacific species have yet been re¬ ported from the Eastern Pacific region. As the pontoniine shrimp faunas of Indone¬ sia and New Caledonia have been studied in more detail than many of the adjacent regions, they may be conveniently compared with the fauna of Australia, although all must still be considered incompletely known. Such are the vagaries of field collections that even such an abundant and widely distributed species as Periclimenes spiniferus has not yet been for¬ mally recorded from New Caledonia, although it is inconceivable that it is absent from the 14 Additions to Pontiniine shrimp fauna coral reefs there. The almost equally common Harpiliopsis beaupresii has also apparently not been collected. As further collecting is carried out, it is expected that the overlap between the faunas will increase, although all regions may be expected to have some indige¬ nous species. At present, 204 species are known from the combined faunas of Indone¬ sia, Australia and New Caledonia, about two- thirds of the number of species occurring in the whole Indo-West Pacific region, with 87 species from Indonesia (Holthuis 1952; Bruce 1983e; Fransen 1989) and 67 from New Cale¬ donia (Bruce 1990a, 1991b). Eighty-seven of the 168 species found in Australia (49%) are also found in Indonesia or adjacent parts of Papua-New Guinea. The Siboga-Snellius col¬ lections in Indonesian waters and more recent collections in New Caledonian waters have indicated that a number of pontoniine shrimp occur in depths of over 100 m. These species are particularly poorly represented in the Australian fauna due to lack of deep-water sampling. At present only Periclimenes al- cocki , P.franklini , P. hertwigi, P. laccadiven- sis and Mesopontonia gorgonicola are known from these depths, 2.8% of the Australian fauna, with P. laccadivensis reaching the greatest depth at about 720 m. In Indonesia, only four species occur in 100 m or over, representing 3.4 % of the pontoniine fauna but nine species, 15%, of the New Caledonian fauna have been found in these depths. Of the 33 species that are presently known only from Australian waters, it is difficult to consider that many of these are truly endemic, that is, with a distribution limited to Austra¬ lian waters. Due to the small size and cryptic behaviour of most of species, it is probable that many have so far been merely overlooked in other parts of the Indo-West Pacific region. This applies particularly to the recently de¬ scribed taxa from the northern and eastern coasts of Australia, which seem highly likely to be components of the general Indo-West Pacific marine fauna. Species recently de¬ scribed from the Great Barrier Reef have al¬ ready been found as far away as the Ryukyu Islands ( Periclimenes magnificus : Bruce 1979; Nomura et al. 1988). Species from the south are more likely to be truly endemic and possibly members of an earlier marine Gond- wana fauna. Species such as Periclimenes aesopius, the first pontoniine shrimp to be described from Australian waters (Bate 1863), is still only known from St Vincent Gulf, South Australia. It seems probable that it is restricted to this region, and may be a true relict. Another candidate as an endemic spe¬ cies is Pontonia mi nut a, first described by Baker (1907) from a single specimen from an unrecorded host, with a further occurrence at Meroo Point, New South Wales, indicating a rather less restricted distribution than P. aesopius. Periclimenes aesopius is a conspicu¬ ous associate of anemones. Pontonia minuta is undoubtedly a commensal species but proba¬ bly not a true Pontonia s. str., and its host remains to be identified; polychaete worms have been suggested. Once the host has been identified, the distribution of P. minuta , which is probably dependent on that of its host and closely related species, may be rapidly clari¬ fied. Table I The Indo-West Pacific distributions of the Australian pontoniine shrimp fauna (+ = species present; O - species presently known only from Australian waters; ? = uncertain record) SPECIES LOCALITY abcdefgh I nw e s 1. Allopontonia iaini 2. Anapontonia denticauda 3. Anchisius australis 4. Anchisius custoides 5. Anchisius custos 6. Anchisius demani 7. A nc hi si us gra vi eri 8. Anchisius miersi 9. Anchisius pectinis 10. Apopontonia dubia 11. Apopontonia falcirostris 12. Apopontonia tridentata 13. Carinopontonia paucipes 14. Chernocaris placunae 15. Conchodytes kempi 16. Conchodytes maculatus 17. Conchodytes meleagrinae O O O J K L M N O P 15 A.J. Bruce SPECIES 18 . 19 . 20 . 21 . 22 . 23 . 24 . 25 . 26 . 27 . 28 . 29 . 30 . 31 . 32 . 33 . 34 . 35 . 36 . 37 . 38 . 39 . 40 . 41 . 42 . 43 . 44 . 45 . 46 . 47 . 48 . 49 . 50 . 51 . 52 . 53 . 54 . 55 . 56 . 57 . 58 . 59 . 60 . 61 . 62 . 63 . 64 . 65 . 66 . 67 . 68 . 69 . 70 . 71 . 72 . 73 . 74 . 75 . 76 . 77 . 78 . 79 . 80 . 81 . 82 . 83 . 84 . 85 . 86 . 87 . 88 . 89 . 90 . 91 . 92 . 93 . Conchodytes monodactylus Conchodytes nipponensis Conchodytes tridacnae Coralliocaris brevirostris Coralliocaris graminea Coralliocaris superba Coralliocaris venusta Coralliocaris viridis Dasella ansom Dasella brucei Dasycaris ceratops Dasycaris Zanzibarica Epipontonia anceps Exopontonia malleatrix Fennera chacei Gnathophylloides mineri Gnathophylloides robustus Hamodactyloides incampletus Hamodactylus aqabai Hamodactylus boschmai Hamodactylus noumeae Hamopontoma corallicola Hamopontonia essingtoni HarpiUopsis beaupresii Harpiliopsis depressa HarpiUopsis spinigera Ischnopontonia lophos Jocaste japonica Jocaste lucina Mesopotamia gorgoniophila Miopontonia yongei Onycocaridella monodoa Onycocaridella prima Onycocaridites anomodactylus . Onycocans amakusensis Onycocaris oligodentata Onycocaris quadratophthalma . Onycocaris spinosa Orthopontonia ornata Palaemonella pottsi Palaemonella rotumana -t Palaemonella spinulata Paranchistus armatus Paranchistus pycnodontae Parapontonia nudirostris Paratypton siebenrocki + Periclimenaeus arabicus Periclimenaeus ardeae Periclimenaeus bidentatus + Periclimenaeus diplosomatis Periclimenaeus djiboutensis + Periclimenaeus gorgonidarum . Periclimenaeus hecate Periclimenaeus orbitospinatus . Periclimenaeus orontes P ericlimenaeus pachydentatus . Periclimenaeus rastrifer Periclimenaeus rhodope Periclimenaeus tridentatus Periclimenaeus tuomotae Periclimenes aesopius Periclimenes qffinis Periclimenes agag + Periclimenes alcocki Periclimenes alegrias Periclimenes amboinensis P ericlimenes amymone Periclimenes anacanthus P ericlimenes andamanensis Periclimenes attenuatus P ericlimenes brevicarpalis + Periclimenes brocketti Periclimenes cf. catmani Periclimenes carinidactylus Periclimenes ceratophthalmus . Periclimenes colemani O O O O O O + O o o o o o . . o + o 16 Additions to Pontiniine shrimp fauna SPECIES LOCALITY ABCDEFGH I JKLMNOP nw e s 94. Periclimenes commensalis 95. Periclimenes consobrinus 96. Periclimenes cornutus 97. Periclimenes cristimanus 98. Periclimenes darwiniensis 99. Periclimenes denticulatus 100. Periclimenes diversipes + 101. Periclimenes elegans + 102. Periclimenes franklini 103. Periclimenes galene 104. Periclimenes gonioporae 105. Periclimenesgranulimanus 106. Periclimenes hertwigi 107. Periclimenes hollhuisi + 108. Periclimenes imperator + 109. Periclimenes incertus 110. Periclimenes indicus 111. Periclimenes inornatus 112. Periclimenes investigatoris 113. Periclimenes kempi + 114. Periclimenes kororensis 115. Periclimenes laccadivensis 116. Periclimenes sp. nov. a 117. Periclimenes lanipes 118. Periclimenes longirostris + 119. Periclimenes lutescens + 120. Periclimenes madreporae 121. Periclimenes magnificus 122. Periclimenes mahei 123. Periclimenes nilandensis 124. Periclimenes ornaiellus 125. Periclimenes ornatus + 126. Periclimenes pectiniferus 127. Periclimenes platycheles 128. Periclimenes sp. nov. b 129. Periclimenes psamathe 130. Periclimenes ruber 131. Periclimenes seychellensis + 132. Periclimenes soror + 133. Periclimenes spiniferus 134. Periclimenes tenuipes + 135. Periclimenes tenuis 136. Periclimenes toloensis 137. Periclimenes venustus 138. Periclimenes yaldwyni 139. Periclimenes zanzibaricus 140. Periclimenoides odontodactylus .141. Philarius gerlachei + 142. Philarius imperialis + 143. Philarius lifuensis 144 Platycaris latirostris 145. Platypontonia brevirostris 146. Phopontonia furtiva 147. Pontonia urdeae 148. Pontonia katoi 149. Pontonia minuta 150. Pontonia okai 151. Pontonia sibogae 152. Pontonia stylirostris 153. Pontonides sp. + 154. Pontoniopsis comanthi + 155. Propontonia pellucida 156. Stegopontonia commensalis . 157. Thaumastocaris streptopus + 158. Tuleariocaris holthuisi 159. Typton anomalus 160. Typton australis 161. Typton bawii 162. Typton dentatus 163. Typton dimorphus 164. Typton nanus 165. Typton wasini 166. Vir orientalis 167. Vir philippinensis 168. Zenopontonia noverca 17 AJ. Bruce Table 2. The pontoniine shrimp faunae of Indonesia (I), Australia (A) and New Caledonia (NC).(+ = species present; O = not yet known outside designated area; X = species occurring in depths of more than 100 m) SPECIES LOCALITY SPECIES LOCALITY I. A. NC I. A. NC 1. Allopontonia iaini Bruce, 1972 2. Altopontonia disparostris Bruce, 1990 3. Anapontonia denticauda Bruce, 1966 4. Anchistus australis Bruce, 1977 5. A custoides Bruce, 1977 6. A. custos (ForskAJ, 1775) 7. A. demani Kemp. 1922 8. A. gravieri Kemp, 1922 9. A. miersi (De Man, 1888) 10. A. pectinis Kemp. 1925 11. Apopontonia dubia Bruce, 1981 12. A. falcirostris Bruce, 1977 13. A. tridentata Bruce, 1988 14. Araiopontonia odontorhyncha Fujino and Miyake. 1970 15. Carinopontonia paucipes Bruce, 1988 16. Chernocaris placunae Johnson, 1967 17. Conchodytes kempt Bruce, 1989 18. C. muculatus Bruce, 1989 19. C. meleagrinae Peters, 1852 20. C. monodactylus Holthuis, 1952 21. C. nipponensis (De Haan, 1844) 22. C. tridacnae Peters, 1852 23. Coralliocarix brevirostris Borradaile, 1898 24. C. graminea (Dana, 1852) 25. C. superba (Dana, 1852) 26. C. venusta Kemp. 1922 27. C. virtdis Bruce, 1974 28. Dasella ansoni Bruce, 1983 29. D. bracet Berggrcn. 1990 30. Dasycaris c era top 5 Holthuis, 1952 31. D. symbiotes Kemp. 1922 32. D. zanzibarica Bruce, 1973 33. Epipontonia anceps Bruce, 1983 34. Exopontonia malleatrix Bruce, 1987 35. Fennera chacei Holthuis, 1951 36. Gnathophylloides mineri Schmitt, 1933 37. G. robustus Bruce 1973 38. Hamodactyhides incompietus (Holthuis, 1953) 39. Hamodactylus aqabai Bruce and Svoboda, 1983 40. H. bosehmai Holthuis, 1952 41 . H. noumeae Bruce. 1970 42. Hamopontonia corallicola, Bruce, 1970 43. H. essingtoni Bruce. 1986 44. Harpiliopsis beaupresii (Audouin, 1825) 45. H. depressa (Stimpson, I860) 46. H. spinigera (Ortmann, 1890) 47. Ischnopontania lophos (Barnard, 1962) 48. Isopontonia platycheles Bruce. 1982 49. Jocaste japonica (Ortmann. 1890) 50. J. lucina (Nobili, 1901) 51. Mesopontonia gorgoniophtla Bruce, 1967 52. M. gracilicarpus Bruce, 1990 53. M'iopontonia yongei Bruce, 1985 54. Onycocaridella monodoa Fujino and Miyake, 1969 55. O. phma Bruce. 1981 56. O. stenolepis Holthuis, 1952 57. Onycocaridites anomodactylus Bruce 1987 58. Onycocaris amakusensis Fujino and Miyake, 1969 59. O. longirostris Bruce. 1980 60. O. oligdentata Fujino and Miyake 1969 61. O. quadratophthaIma (Balss, 1921) 62. O. spinosa Fujino and Miyake, 1969 63. Orthopontonia ornata (Bruce, 1969) 64. Palaemonella lata Kemp. 1922 65. P. dolichodactylus Bruce, 1990 66. P.pottsi (Borradaile, 1915) 67. P rotumana (Borradaile, 1898) 68. P. spmulata Yokoya, 1956 69. P. tenuipex Dana, 1852 70. Par anchistus armatus (H.Milne-Edwards. 1837) 71. P. pycnodontae Bruce, 1978 72. P. serenei Bruce, 1983 73. Parapontonia nudirostris Bruce, 1968 74. Paratypton siebenrocki Balss, 1914 75. Periclimenaeus arabicus (Caiman, 1939) 76. P. ardeae Bruce, 1970 77. P. arthrodactylus Holthuis, 1952 78. P. bidentatus Bruce, 1970 O O O + + O o o o o o o o X o o x O 79. P. diplosomatis Bruce, 1980 80. P. djiboutensis Bruce, 1970 81. P. gorgonidarum (Balss, 1913) 82. P. Hecate (Nobili, 1904) 83. P. holthuisi Bruce, 1969 84. P. minutus Holthuis, 1952 85. P. nobilii Bruce, 1974 86. P orbitospinatus Bruce, 1969 87. P orontes Bruce, 1986 88. P pachydentatus Bruce, 1969 89. P. rastrifer Bruce. 1980 90. P. rhodope (Nobili, 1904) 91. P. spongicola Holthuis, 1952 92. P. tridentatux (Miers, 1884) 93. P. truncatus (Rathbun, 1906) 94. P. tuamotae Bruce, 1969 95. Periclimenes aesopius (Bate, 1863) 96. P. affinis (Zchnlncr, 1894) 97. P agag Kemp, 1922 98. P. alcocki Kemp, 1922 99. P. alegrias Bruce, 1986 100. P. amboinensis (De Man. 1888) 101. P. amymone De Man, 1902 102. P anacanthus Bruce, 1989 103. P andamanensis Kemp, 1922 104. P attenuatus Bruce, 1971 105. P. hrevicarpalis (Schenkel, 1902) 106. P. brocketti Borradaile, 1915 107. P. brocki (De Man, 1888) 108. P. calmani Tallcrsall. 1921 109. P. carinidactylus Bruce, 1969 110. P. ceratophthalmus Borradaile, 1915 111. /’, colemani Bruce, 1975 112. P. commensal is Borradaile, 1915 113. P. consobrinus (De Man, 1902) 1 14. P. cornutus Borradaile, 1915 115. P. cristimanus Bruce, 1965 1 16. P. dansiniensis Bruce, 1987 117. P. denticulatus Nobili, 1906 118. P. digitalis Kemp. 1922 119. P. diversipes Kemp, 1922 120. P elegans (Paulson. 1875) 121. P.franklini Bruce. 1990 122. P.fujinoi Bruce, 1990 123. P. galene Holthuis, 1952 124. P. gonioporae Bruce, 1990 125. P grandis (Stimpson, 1860) 126. P granulimanus Bruce, 1978 127. P. hertwigi Balss. 1913 128. P holthuisi Bruce, 1969 129. P. irnperator Bruce, 1967 130. P. incertus Borradaile, 1915 131. P. indicus (Kemp, 1915) 132. P inornatus Kemp, 1922 133. P. investigaioris Kemp, 1922 134. P. ischiospinosus Bruce, 1990 135. P. jugal is Holthuis, 1952 136. P kempt Bruce. 1969 137. P. kororensis Bruce, 1977 138. P. laccadivensis (Alcock and Anderson, 1894) 139. P. sp. nov. a - Bruce, 1991a 140. P. lanipes Kemp, 1922 141. P. latipollex Kemp, 1922 142. P. longirostris (Borradaile. 1915) 143. P lutescens auct. 144. P. madreporae Bruce, 1969 145. P. magnificus Bruce, 1979 146. P. mahei Bruce. 1969 147. P. nilandensis Borradaile, 1915 148. P. novaecaledoniae Bruce, 1968 149. P. ornate I lus Bruce, 1979 150. P, ornatus Bruce, 1969 151. /'. pan-ispinatus Bruce, 1990 152. P. parvus Borradaile, 1898 153. P. pectinlferus Holthuis. 1952 154. P. platycheles Holthuis, 1952 155. P. sp. nov. b - Bruce, 1991b 156. P psamaihe (De Man, 1902) 157. P. richeri Bruce, 1990 O O O O O O O + + O o o + o o xO x x xO X O X O xO O xO 18 Additions to Pontiniine shrimp fauna SPECIES LOCALITY SPECIES I. A. NC LOCALITY I. A. NC 158. P. ruber Bruce, 1982 159. /*. seychellensis , Borradaile. 1915 160. /*. sibogae Holthuis, 1952 161. P. soror Nobili, 1904 162. P. spiniferus De Man, 1902 163. P. tenuipes Borradaile, 1898 164. P. tenuirostris Bruce, 1990 165. P. tenuis Bruce, 1969 166. P. toloensis Bruce, 1969 167. P. uniunguiculatus Bruce, 1990 168. P vaubani Bruce, 1990 169. P. venustus Bruce, 1990 170. P. yaldwyni Holthuis. 1959 171. P. zanzibaricus Bruce. 1969 172. Periclimenoides odontodactylus (Fujino and Miyake, 1968) 173. Phiiarius gerlachei (Nobili. 1905) 174. P. imperialist (Kubo, 1940) 175. P. lifuensis (Borradaile. 1898) 176. Platyvaris latirostris Holthuis, 1952 177. Platypontonia brevirostris ( Miers, 1884) 178. P. hyotis Hipcau-Jacquotte. 1971 179. Pliopnntonia furtiva Bruce, 1973 180. Pontonia ardcae Bruce. 1981 O O O xO xO 181. Pontonia ascidicola Borradaile, 1898 182. P. katoi Kubo, 1940 183. P. minuta Baker, 1907 184. P. monnioti Bruce. 1990 185. P. okai Kemp, 1922 186. P. slbogae Bruce, 1972 187. P. styiirostris Holthuis, 1952 188. Pontonides unnger Caiman, 1939 189. Pontonides sp. 190. Pontoninpsis comanthi Borradaile, 1915 191. Propontnnia pellucida Bruce, 1969 192. Stegopontonia commensaiis Nobili, 1906 193. Thaumastocaris streptopus Kemp, 1922 194. Tuleariocaris holthuist Hipcau-Jacquotte, 1965 195. Typton anomalus (Bruce, 1979) 196. T. australis Bruce. 1973 197. T. bauii Bruce, 1972 198. T. dentatus Fujino and Miyake, 1969 199. T. dimorphus Bruce, 1986 200. T. nanus Bruce. 1987 201. T. wasini Bruce, 1977 202. Vir nrientalis (Dana. 1852) 203. V. philippinensis Brute and Svoboda, 1984 204. Zenopontonia noverca (Kemp, 1922) x O REFERENCES Berggen, M. 1990. Dasella herdmaniae (Lebour) (Decapoda: Natantia: Pontoniiae) from Mozambique and establishment of a new spe¬ cies, D. brucei. Journal of Crustacean Biology 10(3): 554-559. Bruce, A.J. 1968. Notes on some Indo-Pacific Pontoniinae, XII. The re-examination of Pontonial brevirostris Miers. 1844, with the designation of a new genus Platypontonia (Decapoda Natantia, Pontoniinae). Crusta- ceatta 15:289-297. Bruce, A.J. 1973. Gnathophylloides robustus sp. nov., a new commensal gnathophyllid shrimp from Western Australia, with the designation of a new genus Levicaris (Decapoda, Caridea). Crustaceana 24( 1): 17-23. Bruce, A.J. 1976. A synopsis of the pontoniinid shrimp fauna of central East Africa. Journal of the Marine Biological Association of India 16(2) (1974):462-490. Bruce, A.J. 1977. The possible identity of Coral- liocaris macrophthalma (H. Milne Edwards, 1837) (Decapoda Natantia, Pontoniinae). Crustaceana 32(2):203-305. Bruce, A.J. 1981a. Notes on some Indo-Pacific Pontoniiae, XXXVII. Additional information on Dasella herdmaniae (Lebour) (Decapoda, Natantia). Crustaceana 40(l):50-56. Bruce, A.J. 1981b. Notes on some Indo-Pacific Pontoniinae, XXXVIII. Apopontonia dubia sp. nov., from a southern Queensland sponge host. Crustaceana 41(3):225-232. Bruce, A.J. 1982. Notes on some Indo-Pacific Pontoniinae, XLI. Orthopontonia, a new ge¬ nus proposed for Periclimenaeus ornatus Bruce. Crustaceana 43(2): 163-176. Bruce, A.J. 1983a. The pontoniine shrimp fauna of Australia. Australian Museum Sydney Memoirs XVIILI95-218. Bruce, A.J. 1983b. A second species of the genus Dasella Lebour, D. ansoni sp. nov., from the Arafura Sea. The Beagle, Occasional Papers of the Northern Territory Museum of Arts and Sciences 1 (3):21 -29. Bruce, A.J. 1983c. Epipontonia anceps sp. nov., a sponge associated pontoniine shrimp from Heron Island. Queensland (Crustacea: Deca¬ poda: Palaemonidae). Records of the Austra¬ lian Museum 35(1): 19-28. Bruce, A.J. 1983d. Additions to the marine fauna of the Northern Territory. 1. Decapod Crusta¬ cea: Caridea and Stenopodidea. The Beagle, Occasional Papers of the Northern Territory Museum of Arts and Sciences 1(5):41 -49. Bruce, A.J. 1983e. Expedition Rumphius II (1975). Crustacds Ddcapodes (l"' partie: Natantia Pontoniinae). Bulletin du Museum National d'Histoire Naturelle Paris (4)5 (A) (3):871- 902. Bruce, A.J. 1983f. Further information on Apopon¬ tonia dubia Bruce (Decapoda, Pontoniinae). Crustaceana 45 (2):210-213. Bruce, A.J. 1985. Notes on some Indo-Pacific Pontoniinae, XL! I. Miopontonia yongei gen. nov., sp. nov., from the Australian Northwest Shelf. Crustaceana 48(2): 167-178. Bruce, A.J. 1986. Notes on some Indo-Pacific Pontoniinae, XLII1. A new species of Typton from Ashmore Reef, Timor Sea (Decapoda, Palaemonidae). Crustaceana 50(3):278-286. 19 A.J. Bruce Bruce, A.J. 1987a. Three new species of commen¬ sal shrimps from Port Essington, Arnhem Land, Northern Australia (Crustacea: Deca- poda: Palaemonidae). The Beagle, Occasional Papers of the Northern Territory Museum of Arts and Sciences 3(1): 143-166. Bruce, A.J. 1987b. Notes on some Indo-Pacific Pontoniinae, XLIV. Periclimenes darwinien- sis sp. nov., from the Northern Territory, Australia (Decapoda, Caridea). Crustaceana 52(l):29-39. Bruce, A.J. 1987c. Onycocaridites anomodactylus, new genus, new species, (Decapoda: Palaemo¬ nidae), a commensal shrimp from the Arafura Sea. Journal of Crustacean Biology 7(4):771 - 799. Bruce, A.J. 1988a. Exopontonia malleatrix, new genus, new species, a palaemonid shrimp from Ashmore Reef, Timor Sea. Journal of Crusta¬ cean Biology 8(1):122-130. Bruce, A.J. 1988b. Two new palaemonid shrimps (Crustacea: Decapoda) from the Australian Northwest Shelf. Journal of Natural History 22:1263-1276. Bruce, A.J. 1988c. Typton nanus sp, nov., a new commensal shrimp (Crustacea: Decapoda: Palaemonidae) from the Australian North- West Shelf. The Beagle, Records of the North¬ ern Territory Museum of Arts and Sciences 4:49-56. Bruce, A.J. 1989a. A note on Gnathophylloides mineri Schmitt (Crustacea: Decapoda: Pa¬ laemonidae), including its first occurrence in Australian waters. The Beagle, Records of the Northern Territory Museum of Arts and Sci¬ ences (1988) 5:97-100. Bruce, A.J. 1989b. A new palaemonid shrimp from the Zostera- beds of Moreton Bay, Queensland (Decapoda: Palaemonidae). The Beagle, Rec¬ ords of the Northern Territory Museum of Arts and Sciences (1988) 5:105-114. Bruce, A.J. 1989c. Notes on some Indo-Pacific Pontoniinae, XLV. Conchodytes maculatus sp. nov., a new bivalve associate from the Australian Northwest Shelf. Crustaceana 56:182-192. Bruce, A.J. 1989d. Periclimenes gonioporae sp. nov. (Crustacea: Decapoda: Palaemonidae), a new coelenterate-associated shrimp. The Beagle, Records of the Northern Territory Museum of Arts and Sciences 6( 1): 149-156. Bruce, A.J. 1989 e. A report on some coral reef shrimps from the Philippine Islands. Asian Marine Biology 6:173-192. Bruce, A.J. 1990a. Crustacea Decapoda: Deep-sea Palaemonoid shrimps from New Caledonian waters. In: A. Crosnier (ed.) R6sultats des Campagnes MUSORSTOM, 6. Memoirs du Museum National d'Histoire Naturelle (A) 145:145-219. Bruce, A.J. 1990b. Periclimenes franklini sp. nov., a new deep-sea shrimp from the Coral Sea (Crustacea: Decapoda: Palaemonidae). The Beagle, Records of the Northern Territory Museum of Arts and Sciences 7(l):55-64. Bruce, A.J. 1990c. Additions to the Marine Shrimp Fauna of Hong Kong. In: Proceedings of the Second International Marine Workshop in Hong Kong, 1986, 611-648. Bruce, A.J. 1990d. A new coelenterate-associated shrimp from Port Essington, Arnhem Land Australia. Indo-Malayan Zoology 6:223-243. Bruce, A.J. 1991a. Two new species of Periclime¬ nes (Crustacea: Decapoda: Palaemonidae) from Lizard Island, Queensland, with remarks on some related taxa. Records of the Austra¬ lian Museum : in press. Bruce, A.J. 1991b. Shallow-water shrimps from New Caledonia (Crustacea: Decapoda: Pa¬ laemonidae. In: B. Richer de Forges (ed.) Le benthos des fonds meubles dans les logons de Nouvelle-CaUdonie, Volume 1. fitudes et Theses: Paris, ORSTOM. in press. Bruce, A.J., and Kropp, D.A. 1984. A redescription of Periclimenes yaldwyni Holthuis ( Bra- chycarpus audouini Bate, 1888, Crustacea, Decapoda, Palaemonidae) and its occurrence in Australian waters. Pacific Science 38(3): 189-198. Fransen, C.H. J.M. 1989. Notes on caridean shrimps collected during the Snellius-II Expedition. 1. Associates of Anthozoa. Netherlands Journal of Sea Research 23(2): 131-147. Holthuis, L.B. 1952. The Decapoda of the Siboga Expedition, XI. The Palaemonidae collected by the Siboga and Snellius Expeditions with remarks in other species. II. Subfamily Pontoniinae. Siboga Expeditie Monographic 39 a 10:1-263, tab. 1. Holthuis, L.B. and Gottlieb, E. 1958. An annotated list of the Decapoda of the Mediterranean coast of Israel, with an appendix listing the Deca¬ poda of the eastern Mediterranean. Bulletin of the Research Council of Israel 7B( 1-2): 1 -126. Nomura, K., Kamezaki, K., Hamano, T. and Misaki, M. 1988. Guide Book of Marine Ani¬ mals of Okinawa. 8. Crustacea (Macrura and Anomura). Southern Press: Okinawa. Accepted 20 August 1990 20 The Beagle, Records of the Northern Territory Museum of Arts and Sciences, 1990 7 ( 2 ): 21-29 TWO NEW CAPRELLIDEAN (N. GEN.) AND KNOWN GAMMARIDEAN AMPHIPODS (CRUSTACEA) COLLECTED FROM A SPONGE IN NOUMEA, NEW CALEDONIA. AKIRA H1RAYAMA Biological Laboratory, Department of Liberal Arts, Asia University, 5-24-10 Sakai, Musashino-shi, Tokyo 180, JAPAN. ABSTRACT Colomastix lunalilo J.L. Barnard and a new genus of caprellid amphipod were collected from a sponge, Siphonochalina sp., in Noumea, New Caledonia. Colo¬ mastix lunalilo is redescribed on the basis of mature specimens because only small and immature specimens have been previously observed. The new caprellid amphipod belongs to the subfamily Protellinae McCain and is clearly distinguished from other genera of this subfamily by a combination of unilobate abdomen with biarticulate abdominal appendages, of which the distal article is remarkably reduced, and uniarticulate pereopods 3-4. Keywords: Crustacea, Amphipoda, Colomastix lunalilo, Paradicaprella brucei gen. et sp. nov., New Caledonia. INTRODUCTION Some amphipod specimens were collected by A.J. Bruce from a sponge, Siphonochalina sp., in Noumea, New Caledonia, in 1978, and consisted of two species, Colomastix lunalilo J.L. Barnard, 1979 (Barnard 1970) and a new caprellid species. Colomastix lunalilo has been recorded from coral, algae and other hosts in the tropical zone of the Indian and Pacific Oceans (Barnard 1970; Ledoyer 1978, 1979, 1982; Myers 1985). However, the previ¬ ous specimens, including the type specimen, were all immature (shorter than 2.0 mm). On the other hand, the present specimens of this species are mature (10.2 mm - 4.0 mm), and morphological variations, which I consider due to growth, are observed between the pre¬ vious and the present specimens. Therefore, I have redescribed C. lunalilo here on the basis of the present mature male and female speci¬ mens. The new caprellid species clearly belongs to the subfamily Protellinae McCain (McCain 1970) with the fallowing morphological char¬ acters: 1) mandible with molar process and palp; 2) pereonites 3-4 present but reduced; 3) abdomen of single reduced segment. How¬ ever, the abdomen of the male in the caprellid species is unique in this subfamily, with the distal segment of biarticulate appendages remarkably reduced and with an unilobe (maybe completely fused). I erect a new genus on the basis of these characters and discuss the relationships between the new genus and six other closely related genera, Deutella Mayer, 1890, Luconacia Mayer, 1903. Monoliropus Mayer, 1903, Pseudoprotella Mayer, 1890, Triantella Mayer, 1903 and Triliropus Mayer, 1903. All the specimens are deposited in the col¬ lection of the Northern Territory Museum of Arts and Sciences, Darwin (NTM). SYSTEMATICS Order Amphipoda Suborder Gammaridea Family Colomastigidae Colomastix lunalilo J.L. Barnard (Figs 1-3) Colomastix lunalilo J.L. Barnard, 1970:96, 100, figs 51-52; Barnard 1971:55, figs 24-25; 21 Akira Hirayama Fig. 1. Colomastix hinalilo J.L. Barnard. Male (no. 1, 10.2 mm). Ledoyer 1978: 233, fig. 15; Ledoyer 1979:26, fig. 9; Ledoyer 1982:156-158, fig. 54: Myers 1985:56, fig. 57. Material examined. NTM Cr.006944: Hot Maitae, Noumea, New Caledonia. Stn. 203 (N.C. Amph. 1); male, nos. 1-3 (10.2 mm, 4.4 mm. 4.2 mm); female, nos. 4-6 (5.2 mm, 4.4 mm, 4.0 mm); collected from a sponge, Si- phonochalina sp., by A.J. Bruce on 3 October 1978. A part of no. 1 (male) and no. 4 (female) are mounted on glass slides in gum-chloral medium. Description of male (no. 1, 10.2 mm). Body. Head broadly even and slightly concave dorsally; rostrum broad, central projection small, acute; anterocephalic lobe rounded, with eyes; anteroventral projection extending to midlength of peduncular article 1 of an¬ tenna 1, triangular, slender. At least pereonites 1-3 slightly ridged dorsally. Pleonal epimera 1-3 rounded. Gills present on pereonites 2-6. Antennae. Subequal in length. Antenna 1: peduncular articles 1-3 in length ratios 8:9:4, acutely produced distoventrally; accessory flagellum absent; flagellum 3-articulate, proximal article projecting remarkably as hood-like process far beyond following ar¬ ticles. Antenna 2: peduncular articles 3-5 projecting acutely on both lateral ends of ventral side, length ratios almost 3:4:3; flagel¬ lum 2- or 3-articulate, both lateral projections reaching to ends of following articles. Mouthparts. Upper and lower lips un¬ known (maybe absent). Maxilla 1: inner plate rather small, pubescent; outer plate broad, provided with 5 small conical teeth and one inner-distal cusp; palp 1-articulate, with 4 apical setae. Maxilla 2: both inner and outer plates coalesced at base, setose. Mandibles similar, consisting of broad and quadrate plate and comb-like process, proximal tooth broad¬ est, others slender. Maxilliped: inner plates completely fused, small; outer plates rectan¬ gular, with 2 small setae distally, distal end projecting; palp 4-articulate. penultimate ar¬ ticle medially swollen, pubescent, with pair of setae at midlength. Gnathopods. Gnathopod 1: coxa produced anteroventrally, lacking nipple-like anterior point; length ratios of segments from basis to propod almost 8:5:5:7:3; dactyl unknown (maybe absent), replaced by 6 finely pinnate setae. Gnathopod 2: coxa rectangular, with blunt nipple-like anterior point; anterior mar¬ gin of basis tuberous; carpus provided with anterior lobe medially setose; palm oblique, defined by pair of cusps, with 2 teeth and distal triangular process; dactyl falcate, with blunt projection near base of grasping margin. Pereopods 3-7. Homopodous except for coxae and basis; coxae 3-7 lacking nipple-like anterior point; basis of pereopods 3-4 slender, of pereopods 5-7 with thin hind plate; propod with 2 or 3 spines, without locking spines. Pleopods. Peduncle stout, twice as long as rami, with 2 coupling spines on pleopod 1 and 1 spine on pleopods 2-3; rami broad, short, 4- or 5-articulate; swimming setae long. Uropods. Uropod I extending slightly be¬ yond uropod 2: rami equal in length, finely 22 Amphipods from New Caledoniia 1.0mm A,B,C,D,l,K,M,hl,0, O.Smm J,L,S,T,U 0.05mm S1 0,5mm C 1 ,C2 ,D 1,D2,H, 0.3mm C3,D3,I1,J1 0.2mm E 0.1mm F,G Fig. 2. Colomastix lunalilo J.L. Barnard. Male (no. 1. 10.2 mm) and female (no. 4, 5.2 mm: J, J1 and L). A Head. B Anteroventral projec.ion of head. C, Cl, C2, C3. Antenna 1, peduncular articles 1-2 and flagellum D D1 D2 D3 Antenna 2, peduncular articles 1-2 and flagellum. E, Maxilla 1. F, Maxilla G, Mandible. H, HI, Maxilliped in'dorsal and ventral views. 1.11, Male gnathopod 1 and distal part of propod. J, ,11, Female gnathopod 1 and distal part of propod. K, Male gnathopod 2. L, Female gnathopod 2. M, Pereopod 3. N, Pereopod 4. (), Pcreopod 5. P, Pereopod 6. Q, Pereopod 7. R. Pleon. S, SI, Pleopod 1 and coupling spines. T, U, Coupling spine of pleopods 2-3. V, Urosome. W, Uropod 1. X, Uropod 2. Y, Uropod 3 and telson. 23 Akira Hirayama serrate marginally. Uropod 2 similar to uro- pod 1, peduncle shorter. Uropod 3 distinctly extending beyond uropod 1; peduncle and outer ramus slightly shorter than half the length of inner ramus. Telson. Triangular, slender with 2 opposing peniciliate setae at midlength, apex with api- comedial tooth and 2 setae. Description of female (no. 4, 5.2 mm). Gnathopod 1: length ratios of segments from basis to propod almost 6:10:8:7:12; propod with 4 pinnate setae. Gnathopod 2: length ratios of carpus, propod and dactyl 7:12:4; propod gently rounded, setose posteriorly, palm undefined; dactyl slender. Remarks. Colomasti.x lunalilo has been recorded from the Hawaiian Islands (Barnard 1979, 1971), Madagascar (Ledoyer 1979), Mauritius (Ledoyer 1978) and Viti Levu, Fiji (Myers 1985), in the tropical zone of the In¬ dian and Pacific Oceans, and is easily distin¬ guished lrom the other colomastigids by the following morphological characters: 1) in antenna 1, the first article of flagellum extends remarkably beyond the following articles; 2) in gnathopod 2 of the male, the dactyl is shorter than half the length of the propod; 3) the inner plates of the maxilliped are com¬ pletely fused; 4) the outer ramus of uropod 3 is shorter than half the length of the inner ramus; 5) the telson is triangular and slender. These speedic characters suggest that the present specimens are C. lunalilo. However, all the specimens collected from the above localities and previously described are small (less than z * u mm) and may be immature (see <the gnathopod 2), whereas the present specimens are large (10.2 mm to 4.0 mm) and mature. Therefore, although allometric variations of this species are not well known, the following morphological variations, which 1 estimate are due to growth, are observed between the present and previous material: 1, eyes large (Barnard 1970; Myers 1985) (rather than medium or small in the new material); 2, coxae 1-5 (Barnard 1970) or at least coxae 1-2 (Myers 1985) have nipple-like anterior points (absent from coxae 1 and 3-5 in the new material); 3, in gnathopod 2 of the male, the palm is defined by large (Barnard 1970) or small (Myers 1985) cusps (in the new mate¬ rial, this is defined by a pair of small cusps, in addition to two teeth), the dactyl is slightly geniculate near the apex (not geniculate in the new material); 4, in gnathopod 1, the propod and carpus are subequal in length (Barnard 1970; Ledoyer 1982; Myers 1985), (in the new material, the propod of the male is as long as the carpus and 0.75 times that of the female; 5, in uropod 3, the outer ramus is about 0.25 times as long as the inner ramus (Barnard 1970; Ledoyer 1982; Myers 1985) (in the new material, it is nearly equal to half the length of the inner ramus); 6, the head is smooth and rounded dorsally (Barnard 1970; Ledoyer 1982; Myers 1985) (in the new material, it is broadly even and slightly concave dorsally); 7, the telson lacks peniciliate setae (Barnard 1970; Ledoyer 1982; Myers 1985), the apex is truncate and serrate (Barnard 1970) (in the present material, it is provided with opposing peniciliate setae midmarginally, and the apex has an apicomedial tooth). 24 Amphipods from New Caledoniia Fig. 4. Paradicaprella brucei gen. et sp. nov. A, Holotype, male, 10.4mm. B. Paratype no 3, female, 9.6 mm. Suborder Caprellidea Family Caprellidae Subfamily Protellinae Paradicaprella gen. nov. Diagnosis. Antenna 2 lacking swimming setae; flagellum 2-articulate. Mandibular palp 3-articulate, not reduced, setae formula for ultimate article 1 - x (plumose setae) -1; molar process developed. Inner lobe of maxilliped medium but smaller than outer lobe; both lobes less setose. Pereopod 5 6-articulate, inserted near posterodistal end of pereonite 5. Abdo¬ men of male with single lobe and pair of biarticulate appendages, terminal article of appendages reduced; in female, without ap¬ pendages. Gills present on pereonites 3-4. Type species. Paradicaprella brucei sp. nov. Etymology. A compound word is derived from “ Paradi -” = paradise, an allusion to New Caledonia, and caprella. The gender is feminine. Remarks. The new genus Paradicaprella is closely related to six other genera in the sub¬ family Protellinae: Deutella Mayer, 1890 (Cavedini 1981; Laubitz 1970; Mayer 1903; McCain 1968), Luconacia Mayer, 1903 (Laubitz 1972; Mayer 1903; McCain 1968), Monoliropus Mayer, 1903 (Arimoto 1976; Griffiths 1973; Mayer 1903; Sivaprakasam 1967), Pseudoprotella Mayer. 1890 (Sars 1895; Chevreux and Fage 1925; Schellenberg 1942; Vassilenko 1974), Triantella Mayer, 1903 (Mayer 1903; Schellenberg 1931) and Triliropus Mayer, 1903 (Arimoto 1976; Mayer 1903). This group of genera has the following characters: pereopod 5 6-articulate, inserted near posterodistal end of pereonite 5; pereronites 6-7 free; 3-articulate mandibular palp not reduced; abdomen of male with one pair of appendages (McCain 1970). However, Paradicaprella distinctly differs from these six genera in the abdomen of the male which has one lobe and the ultimate article of the male’s abdominal appendages is greatly re¬ duced. Paired lobes are present in these six genera, and perhaps in the whole subfamily. In Monoliropus and Luconacia, the ultimate ar¬ ticle of the abdominal appendages is not re¬ duced as observed in the new genus; the other four genera have uniarticulate appendages. Further, Paradicaprella, Triliropus and Monoliropus have unisegmented pereopods 3-4, and these pereopods of the other genera 25 Akira Hirayama are biarticulate although the ultimate article of their pereopods 3-4 is remarkably reduced. Paradicaprella brucei sp. nov. (Figs 4-5) Type material. HOLOTYPE - NTM Cr.006946: Male, 10.4 mm, taken from a sponge, Siphonochalina sp., in Ilot Maitae, Noumea, New Caledonia; 3 October 1978; coll. A.J. Bruce. PARATYPES - NTM Cr.006947: Two males (no. 1,8.3 mm; no. 2, 8.1 mm) and one female (no. 3, 9.6 mm), collected with the holotype. The holotype and a part of the paratype no. 3 (female) are mounted on glass slides in gum-chloral me- Fig. 5. Paradicaprella brucei gen. et sp. nov., male holotype (10.4 mm) and female paratype no. 3 (9.6 mm: L and R). A, Upper lip. B, Lower lip. C, Maxilla 1. D, Maxilla 2. E, Et, One mandible. F, the other mandible. G, Gl, Maxilliped and palp. H, Antenna 1. I, II, Antenna 2 and flagellum. J. Male gnathopod 1. K. Male gnathopod 2. L, Female gnathopod 2 (Para. 3). M, Pcreopod 3. N, Pereopod 5. O, Pereopod 6. P, Pereopod 7. Q, Male plcon and penis in ventral view. Ql, Male pleon in dorsal view. R, Female pleon in ventral view (para. 3). 26 Amphipods from New Caledoniia dium. Collection number: Stn. 203 (NC Amph. 2 ). Description of the male holotype (10.4 mm). Body. Head armed with triangular proc¬ ess dorsally. Length ratios of pcreonites 1-7 9:33:42:39:32:18:8. Pereonite 2 armed with pair of tubecles dorsomedially. Abdominal appendages not lobate, biarticulate, terminal segment with apical seta; abdominal lobe unilobate. Antennae. Antenna 1: length ratios of peduncular articles 5:11:5; flagellum as long as peduncle, with aesthetascs. Antenna 2: length ratios of peduncular articles 4-5 and flagellum 20:25:8; flagellum 2-articulate. Mouthparts. Maxilla 1; outer plate armed with 7 tooth-like spines; ultimate article of palp with 4 spines and 3 setae apically. Max¬ illa 2: inner plate with 5 (right) and 7 (left) setae; outer plate with 6 (right) and 7 (left) setae. Mandibles similar; incisor with 5 teeth; lacinia mobilis ornamented with 2 teeth; with 2 accessory blades in one mandible and 3 in the other; palp 3-articulate, middle article with 4 setae, setal formula for ultimate article 1-7 (plumose) - 1. Maxilliped: inner plate medium in size, with 3 apical setae; outer plate with 2 apical setae; palp rather slender, 4-articulate, penultimate article geniculate, pubescent on grasping margin. Gnathopods. Gnathopod 1 : propod as long as carpus, triangular, palm defined by spine and serrate margin. Gnathopod 2: carpus coalescent with propod; propod swollen, flat¬ tened or slightly concave on half of dorsal margin extending to well developed distal projection; palm gently curved, almost reach¬ ing base, defined by small protrusion with pal¬ mar spine, distal triangular process well devel¬ oped, following poison tooth; dactyl reaching palmar protrusion when closed, provided with numerous pits on grasping margin. Pereopods. Pereopods 3-4 1 -articulate, with 4 setae distally. Pereopods 5-7 ho- mopodous, becoming longer; propod with palmar protrusion armed with pair of spines, palm spinose; dactyl reaching palmar protru¬ sion when closed. Description of the female paratype no. 3 (9.6 mm). Small dorsal projection on head. Gnathopods 1-2 similar to those of male ex¬ cept distal dorsodistal process small. Abdo¬ men without appendages. ACKNOWLEDGMENTS I wish to thank Dr. A.J. Bruce of the North¬ ern Territory Museum for giving me the op¬ portunity to examine the present material. Thanks are also due to Prof. Y. Suzuki and Mr. T. Nakano of Asia University, Tokyo, for providing me with working space and facili¬ ties. REFERENCES Arimoto, I. 1976. Taxonomic studies of caprellids (Crustacea, Amphipoda, Caprellidae) found in the Japanese and adjacent waters. Publica¬ tions of the Seto Marine Biological Labora¬ tory Special Publication Series 3:1-229. Barnard, J.L. 1970. Sublittoral Gammaridea (Amphipoda) of the Hawaiian Islands. Smith¬ sonian Contributions to Zoology 34:1-286. Barnard, J.L. 1971. Keys to the Hawaiian marine Gammaridea, 0-30 metres. Smithsonian Con¬ tributions to Zoology 58:1-135. Cavedini, P. 1981. Contributo alia conscenja dei caprellidi del Mediterraneo (Crustacea, Amphipoda). Bolletino del Museo civico di Storia naturale di Verona 8:493-531. Chevreux, E. and Fage, L. 1925. Faune de France. 9: Amphipodes. Lechevalier, Paris: 1-488. Griffiths, C.L. 1973. The Amphipoda of southern Africa. Part 1. The Gammaridea and Caprel- lidea of southern Mozambique. Annals of the South Africa Museum 60:265-306. Laubitz, D.R. 1970. Studies on the Caprellidae (Crustacea, Amphipoda) of the American North Pacific. Publications in Biological Oceanography 1:1 -89. Laubitz, D.R. 1972. The Caprellidae (Crustacea, Amphipoda) of Atlantic and Arctic Canada. Publications in Biological Oceanography 4:1-82. Ledoyer, M. 1978. Amphipodes gammariens (Crustacea) des biotopes cavitaires organo- genes recifaux de L’lle Maurice (Ocean Indien). Mauritius Institute Bulletin 3:197- 332. Ledoyer, M. 1979. Les gammariens de la pente externe du Grand Recif du Tulear (Madagas¬ car). (Crustacea, Amphipoda). Memorie del Museo civico di Storia naturale di Verona (2° series). Sezione Scienze della Vita 2:1-150. Ledoyer, M. 1982. Faune du Madagascar. 59 (I). Crustaces amphipodes gammariens. Families des Acanthonotozomatidae a Gammaridae. Centre National de la Recherche Scientifique,Paris: 1-598. Mayer, P. 1903. Die Caprellidae der Siboga- Expedition. Siboga-Expeditie, Monographic 34:1-160, pis 1-10. 27 Akira Hirayama McCain, J.C. 1968. The Caprellidae (Crustacea: Amphipoda) of the western North Atlantic. U.S. National Museum Bulletin 278:1-147. McCain, J.C. 1970. Familial taxa within the Ca- prellidea (Crustacea: Amphipoda). Proceed¬ ings of the Biological Society of Washington 82:837-842. Myers, A.A. 1985. Shallow-water, coral reef and mangrove Amphipoda (Gammaridea) of Fiji. Records of the Australian Museum Supple¬ ment 5:1-143. Sars, G.O. 1895. Amphipoda: An account of the Crustacea of Norway with short descriptions and figures of all the species. Volume 1: 1- 711, pis 1-240, 8 supplementary plates. Alb. Cammermeyers: Christiana and Copenhagen. Schellenberg, A. 1931. Gammariden und Caprel- liden des Magellagebietes, Siidgeorgiens und der Westantarktis. Zoological Results Swedish Antarctic Expedition 1901-1903 2:1-290, pi. 1 . Schellenberg, A. 1942. Krebstiere oder Crustacea. IV. Flohkrebse oder Amphipoda. Biologic der Tierwelt Deutschlands 40:1 -252. Sivaprakasam, T.E. 1967. Notes on some am- phipods from the south east coast of India. Journal of the Marine Biological Association of India 9:372-383. Vassilenko, S.V. 1974. Caprellids of the sea of the USSR and adjacent waters. Akademija Nauk USSR 107:1-288 [In Russian], Accepted 19 June 1990 r 28 The Beagle, Records of the Northern Territory Museum of Arts and Sciences, 1990 7(2):29-33 A NEW SPECIES OF SCINCID LIZARD FROM WESTERN ARNHEM LAND, NORTHERN TERRITORY. ROSS A. SADLIER The Australian Museum PO Box A285, Sydney South NSW 2000, Australia. ABSTRACT A new species of Egernia is described from the sandstone escarpment on the western edge of the Arnhem Land plateau. It is most similar in external morphol¬ ogy to the east Australian species Egernia frerei, but can be differentiated from this by significant differences in colouration and also by several scalation characters. Keywords: Scincidae, Egernia , Australia, taxonomy, new species. INTRODUCTION A large species of Egernia from the East Alligator Rivers region, bordering the western edge of Arnhem Land, Northern Territory, has been known since the fauna was initially listed by Cogger (1974) as Egernia cf. frerei. Fur¬ ther specimens have since been collected in that region, and the species has been figured by various authors (Cogger 1975 and subse¬ quent editions; Swanson 1979, 1987; Wilson and Knowles 1988), but its status has never been investigated in detail. Cogger (1975 and subsequent editions) figured this species as E. frerei and included its distribution in the Northern Territory under the description of E. frerei. Swanson (1987) cited the photograph (pi. 106) of this species as Egernia richardi, but provided no explanation for doing so. Presumably, the basis for this action lies in the type locality for Tropidolepisma richardi Pe¬ ters (1869) being “Alligator River ... North Australia.” Reference to the description of T. richardi provided two characters which indi¬ cate that it could not be conspecific with the species described here: a flattened out (oblate) body and tail; and 30 mid-body scale rows. Further, Mr Glen Shea (pers. comm.) has examined the type specimen of T. richardi Peters, and he concluded it is conspecific with Egernia carinata Smith (1939) from south¬ west Western Australia, and that the cited type locality for T. richardi is most likely errone¬ ous. Wilson and Knowles (1988) figure this species (Fig. 472) as Egernia sp. “a member of the E. frerei group,” but failed to give any diagnostic characters that would separate it from any other species of Egernia. In the species account of E. frerei they assign it to the E. major group. Apparent inconsisten¬ cies in group allocation and failure to define either the “£. frerei group” or “£. major group” have done little towards justifying their inclusion in the text. Similarly, the un¬ diagnosed or unreferenced presentation of the Arnhem Land Egernia as an unassigned species has done little to clarify its status. In this paper the Arnhem Land taxon pre¬ viously allied to Egernia frerei is described as a new species. MATERIALS AND METHODS Measurements were made to the nearest millimetre and, other than maximum snout- vent length (SVL), are here expressed as a percentage of SVL. Scalation characters discussed follow Tay¬ lor (1935: 11) for the headshields; ‘midbody scale rows’ are the number or longitudinal scale rows around the body counted at a point midway between the axilla and groin; ‘par¬ avertebral’ scales are the number of scales in a paravertebral row posterior to the parietals to a point opposite the anus; fourth toe lamel¬ lae are the number of scales on the underside of the toe from its apex with the third digit to the nail. Specimens of E. frerei (n = 175) held in the Australian Museum, Museum of Victoria, 29 Ross A. Sadlier Fig. 1. Holotype (NTM R0809) of Egernia arnhemensis Queensland Museum, South Australian Mu¬ seum and American Museum of Natural His¬ tory were examined for comparison with known specimens of the new species described below. Abbreviations used in the text are as follows: AM, Australian Museum, Sydney; NTM, Northern Territory Museum. Egernia arnhemensis sp. nov. (Figs 1-3) Type material. HOLOTYPE - NTM R0809: 6.1 km SW of Oenpelli, NT, 12°22’S, 133°02'E, collected by G. Gow and B. Jukes, 21 May 1975; adult (Fig. 1). PARATYPES - NTM R1190: 3 km SW of Oenpelli, NT, 12°21’S, 133°02’E; NTM R8032-34: R8181 Nourlangie Rock NT, 12°52'S, 132°50’E; N I’M R8387-88: Little Nourlangie Rock, NT, I2°52’S, 132°47’E; AM R38384: Koongarra, Mt. Brockman Range, NT, 12°33’S. 132°56’E; AM R100018: SE corner of Jabiluka outlier, Magela Creek, NT, 12°33’S, 132°56’E. Diagnosis. A large species of Egernia (maximum SVL 208 mm), distinguished from all other members of the genus by the follow¬ ing combination of scalation and colour char¬ acteristics: dorsal scales moderately smooth from Oenpelli, NT. with several longitudinal striations; enlarged subocular scale row complete, separating upper labials from contact with lower eyelid; nuchals usually 2 or 3 either side; midbody scale rows 44-48; paravertebral scale rows 59- 65; colouration dark. Details of holotype. Snout-vent length 192mm; tail length about 260mm, distal 42% reproduced; forelimb to snout length 68mm; axilla to groin length 120mm; hindlimb length 68mm. Prefrontals narrowly separated; supra- ciliaries 8/7, 1st and 2nd enlarged, posteri- ormost present as 2 (1 upper and 1 lower) unenlarged subequal scales (lower similiar in size to adjacent penultimate supraciliary); nuchals 3/3; primary temporal single; enlarged upper secondary temporals 1/1, each with 1 adjacent smaller temporal bordering the parie¬ tal; upper labials 7/8; enlarged auricular lob¬ ules 5/5. Midbody scale rows 44; paravertebral scale rows 59; lamellae beneath the fourth toe 21/24 Description. Maximum SVL 208 mm; lore- limb to snout length 35-38.9% of SVL (mean=36.1, N=9); axilla to groin length 57.9 - 62.5% of SVL (mean=60.3, N=9); hindlimb length 30.2 - 36.8% of SVL (mean=34.3, N=9); tail length approximately 135-140% of SVL (N=2). 30 New species of scincid lizard Rostral in broad contact with frontonasal; prefrontals moderately to narrowly separated (44.4%), to narrowly to broadly contacting; supraciliaries 7(25%), 8(40%), 9(20%) or 10(15%); anterior supraciliaries variably en¬ larged with anteriormost 2(35%), or 3 larger; posterior supraciliaries variably enlarged or fragmented, with the posteriormost either larger(50%), or present as 2 (1 upper and 1 lower), unenlarged subequal scales (the lower similar in size to the adjacent penultimate supraciliary), and the penultimate occasion¬ ally similarly fragmented, or both the penulti¬ mate and last supraciliaries with a single large scale above (interdigitating between 3rd and 4th supraocular); nuchals 1-3 either side (mode 2,75%), anteriormost pair occasionally separated by presence of large distinct inter- nuchal; primary temporal single, large and distinct; enlarged upper secondary temporals I either side, with 1 (90%) - 2 adjacent smaller temporals bordering the parietals; lower sec¬ ondary temporal single large and distinct; nasals widely separated, each with a promi¬ nent postnarial groove; upper labials usually 8(65%), or 7(30%), rarely 9(5%); subocular scale row complete, separating upper labials from lower eyelid; ear opening large with 4(20%), 5(40%), 6(30%), or 7( 10%), enlarged auricular lobules anteriorly. Body scales with 3-4 faint striations, in¬ creasing in number towards the nape as scales broaden; midbody scale rows 44-48 (mean=45.3, sd=1.5, N=10); paravertebral scale rows 59-65 (mean=61.4, sd= 1.9, N= 10). Lamellae beneath fourth toe 20-24 (mean=22, sd= 1.1, N=9), basal portion broad (with a median division), distal portion com¬ pressed (undivided). Colour and pattern. Dorsal surface grey to light brown, each scale with a mid to dark brown longitudinal streak medially, head and nape without dark streaking. Lateral surface grey-brown with dark streaking (similar to dorsal surface), becoming lighter ventrally; upper lateral surface between ear opening and forelimbs black (with occasional white fleck¬ ing) interrupting continuity of the grey-brown dorsal and lateral colouration in this region. This dark colouration tends to break up pos¬ teriorly of forelimb to merge with dark longi¬ tudinal streaks of the upper to mid lateral surface respectively; facial surface grey- brown with marked dark flecking posterior to orbit. Venter anterior to forelimbs strongly marked with dark flecks, tending to form se¬ ries of roughly transverse bars. Soles of feet and lamellae mid-dark brown. In life venter with a yellow flush posterior to forelimb. Mr Keith Martin (pers. comm.) records the colouration of a juvenile individual (approx. 150 mm total length), as uniform dark brown- black marked with bands of yellow spots dor- sally. Etymology. The name “ arnhemensis ” has been chosen for the species to emphasis the restricted distribution of the species to a habi¬ tat associated with the Arnhem Land plateau. Habits. Mr Keith Martin has collected and observed a number of specimens during the Fig. 2. Paratype (AM R100018) of Egernia arnhemensis from the Jabiluka outlier, Magela Creek, NT. 31 Ross A. Sadlier Fig. 3. Map showing the distribution of Egernia arnhemensis (specimen records indicated by closed stars, sight records indicated by open star) in the Alligator Rivers Region, NT. course of field work in the Mt Brockman Nourlangie region. In this region (K.Martin, pers. comm.) E. arnhemensis occurs in thickly vegetated, wet, rocky gorges with numerous deep crevices. The sites at which specimens were collected were “closed forest” and “rocky crevice” habitats. Individuals were only observed to be active in the vicinity of deep crevices in the late afternoon and most specimens were collected in small mammal traps set in the very late evening and checked at first light. This pattern was also observed 32 New species of scincid lizard for the specimen collected by Ms Anne Kerle on the Jabiluka outlier. Mr Grant Husband observed E. arnhemensis at El Sharana, some 95 km south of the Mt Brockman region. Comparison with other species. Egernia arnhemensis is unlikely to be confused with any other member of the genus apart from E. frerei, which is similar in size and superficial morphology. It is not the purpose of this paper to document variation in E. frerei in Eastern Australia and New Guinea, but rather to point to the distinct differences in features of scala- tion and colouration between E. arnhemensis and E. frerei. Those characters which readily distinguish E. arnhemensis from E. frerei include the following: more numerous midbody scale rows (44-48 vs. 30-36); more numerous paravertebral scales (59-65 vs. 44- 56); and a lack of differentiation in colour between the dorsal and lateral surfaces in contrast to a variety of colour patterns in E. frerei , most of which feature not only dorsal and lateral differentiation, but also a distinct laterodorsal zone (if in E. frerei the dorsal and lateral zones are poorly differentiated, then there is generally an accompanying lack of the dark flecking that is a feature of the colour pattern of E. arnhemensis). ACKNOWLEDGMENTS I thank the following individuals and insti¬ tutions for the loan of material: Mr P. Horner, Dr M. King and Mr G. Gow (former curator), Northern Territory Museum of Arts and Sci¬ ences; Ms J. Covacevich, Mr G. Ingram and Mr G. Czechura, Queensland Museum; Mr J. Coventry, Museum of Victoria; Ms A. Ed¬ wards, South Australian Museum; Dr R. Zwiefel, American Museum of Natural His¬ tory. I would also like to thank Mr K. Martin for permission to use his observations on E. arnhemensis-, Mr F. Parker and Dr R. Zwiefel for their observations on E. frerei-, and Mr G. Shea for use of his transcription of Peters’ (1869) description of T. richardi, and com¬ ments on the status of the type specimen of this species. Dr A. Greer provided the photo for Figure 2; Mr T. Farr the photo for Figure 1; and P. Koshland the artwork for Figure 3. I thank Ms K. Havansky and Ms J. Nancarrow for typing the various drafts of this manu¬ script. REFERENCES Cogger, H.G. 1974. Amphibians and Reptiles. In: Alligator Rivers Region Environmental Fact Finding Study, Wildlife. 1-16. C.S.I.R.O. Division of Wildlife Research Report: Can¬ berra. Cogger, H.G. 1975. Reptiles and Amphibians of Australia. A.H. and A.W. Reed: Sydney. First Edition. Peters, W. 1869. Mittheilung iiber neue Saurier (Chaundoemus multicarinatus, Tropidole- pisma Richardi und Gymnodactylus Ste- udneri) und Batrachier (Cyclorhamphus fas- ciatus und Hyla gracilenta). Monatsbericht der Koniglich Preussischen Akademie der Wissenschaften zu Berlin 1869:786-790. Smith, H.M. 1939. A new Australian lizard with a note on Hemiergis. Zoological Series of Field Museum of Natural History 24(3): 1 1-14. Swanson, S. 1979. Some Rock-dwelling reptiles of the Arnhem Land Escarpment. Northern Ter¬ ritory Naturalist 1(2): 14-18. Swanson, S. 1987. Lizards of Australia. Angus and Robertson: North Ryde, N.S.W. Taylor, E.H. 1935. A taxonomic study of the Cos¬ mopolitan Scincoid Lizards of the Genus Eumeces with an account of the Distribution and Relationships of its Species. Bulletin of the University of Kansas 36(14): 1-642. Wilson, S. and Knowles, D. 1988. Australia's Rep¬ tiles. Collins: Australia. Accepted 6 September 1990 33 The Beagle, Records of the Northern Territory Museum of Arts and Sciences, 1990 7(2):35-38 A NEW SPECIES OF NEMIPTERUS (PISCES: NEMIPTERIDAE) FROM THE SOUTH-WESTERN PACIFIC BARRY C. RUSSELL Northern Territory Museum of Arts and Sciences, GPO Box 4646, Darwin NT 0801, Australia. ABSTRACT A new species of Nemipterus from Fiji and Vanuatu is described and figured. The new species appears most closely related to N. aurifilum (Ogilby) and N. bathybius Snyder, but differs principally in colour pattern: N. aurifilum has a single yellow stripe from beneath the origin of the lateral line to the caudal peduncle; N. bathybius has two yellow stripes, the first from beneath the origin of the lateral line to the peduncle and the second from behind the pectoral fin to the caudal fin base; whereas the new species has a single indistinct pale yellow stripe from above the pectoral fin base to the peduncle. Keywords: Nemipterus, new species, Fiji, Vanuatu. INTRODUCTION In the course of revisionary studies of the Indo-West Pacific fish genus Nemipterus, presently being undertaken by the author, numerous specimens were received for identi¬ fication. Amongst the material obtained were several specimens of an undescribed species of Nemipterus sent from Fiji by Dr A.D. Lewis. Additional specimens from Vanuatu (New Hebrides) were found also in the Australian Museum. These specimens are described herein as a new species of Nemipterus. Methods of counting and measuring speci¬ mens follows Russell (1986). Lengths of specimens are standard length (SL) unless otherwise indicated, and all measurements are expressed in millimetres rounded to the near¬ est 0.1 mm. In the description, measurements and counts for the holotype are given first; those for the paratypes, where different from the holotype, are enclosed in parentheses. Specimens are deposited in the Australian Museum. Sydney (AMS) and Northern Terri¬ tory Museum, Darwin (NTM). SYSTEMATICS Family Nemipteridae Nemipterus vitietisis sp. nov. (Fig. 1, Tables 1-2) Type material. HOLOTYPE - NTM S. 11744-001: 170 mm SL, male, Vatutotolu Reef, VitiLevu, Fiji, 17°10’S, 177°38’E,hook and line, 273 m depth, coll. P. Mead, 16 March 1984. PARATYPES - AMS 1.12012: 140.9 mm SL, Efati L, Vila, Vanuatu, 17°45'S, 168°18'E. AMS 1.12013: 161.7 mm SL, Vila, Vanuatu, 17°45’S, 168°18’E, 27.3 m depth. AMS 1.26907-001: 2 specimens, 142.5-159.0 mm SL, Suva market, Viti Levu, Fiji, coll. A.D. Lewis, May 1984. NTM S.l 1140-001: 2 specimens, 190.5-218.5 mm SL, Suva market, Viti Levu, Fiji, coll. A.D. Lewis, 17 February 1984. Diagnosis. Pectoral-fin rays ii,14-16; gill rakers 12-14; body moderately elongate, depth 3.2-3.8 in SL; head length about equal to body depth, depth 1.0-1.1 in head length; diameter of eye equal to snout length, 1.0 in snout; lower margin of eye tangent to a line from tip of snout to upper pectoral fin base; suborbital shallow, its lower edge emarginate, least depth 2.0-3.0 in eye diameter; imaginary line ex¬ tended upward from posterior edge of subor¬ bital reaching dorsal profile about 3-7 scale rows before origin of dorsal fin; pectoral and pelvic fins moderately long, both reaching to or beyond level of vent; upper lobe of caudal fin with long, trailing filament. Description. Dorsal-fin rays X,9; anal-fin rays 111,7; pectoral-fin rays ii, 14 (left) or 15 (right) (ii, 14-16); lateral-line scales 46 (46- 48); transverse scale rows 3/11 (3/10-11); gill rakers 14 (12 in New Hebrides paratype only; gills removed in Fiji paratypes). 35 B.C. Russell Fig. 1. Nemipterus vitiensis. drawing based on paratype (NTM S.l 1140-001, 190.5 mm SL). Depth 3.2 (3.4-3.8) in SL; head 3.2 (3.2-3.5) in SL; head length about equal to body depth, depth 1.0 (1.0-1.1) in head; snout short, bluntly rounded In profile, 3.2 (3.0-3.3) in head; eye 3.1 (3.0-3.2) in head; diameter of eye equal to snout length, 1.0 in snout; lower margin of eye tangent to line from tip of snout to upper pectoral fin base; interorbital width 1.7 (1.6-1.8) in eye; least depth of suborbital 2.5 (2.0-3.0) in eye; ventral margin of suborbi¬ tal emarginate, its posterior edge rounded; an imaginary line extended upward from poste¬ rior edge reaching dorsal profile about 3-7 scale rows before origin of dorsal fin; naked width of preopercle 2.3 (1.8-2.3) in scaly width; margin of preopercle finely denticu¬ late; caudal peduncle depth 1.3 (1.3-1.4) in peduncle length; dorsal fin length 1.9 in SL; dorsal spines 8-10 longest, 1.8 (1.5-1.9) times length of first dorsal spine; dorsal rays 6-8 longest, 1.3 (1.4-1.6) times length of longest dorsal spine; anal fin length 5.6 (5.6-5.9) in SL; first anal spine 2.2 (1.7-2.2) in second; second anal spine 1.3 (1.2-1.3) in third; pecto¬ ral fins moderately long, reaching to or be¬ yond level of vent, length 1.1 (1.0-1.2) in head; pelvic fins moderately long, reaching almost to or just beyond vent, length 1.5 (1.3- 1.4) in head; pectoral fins length 0.7 (0.7-0.9) in pelvic fins length; caudal fin forked, upper rays produced into long, trailing filament. Maxillary reaching to below anterior mar¬ gin of pupil; jaw teeth small, pointed, in nar¬ row tapering bands in both jaws; upper jaw Table 1. Measurements (mm) of type specimens of Nemipterus vitiensis. Holotype NTM Measurement S. 11744-001 Paratypes AMS 1.12012 AMS 1.12013 AMS 1.26907-001 AMS 1.26907-001 NTM S.l 1140-001 NTM S.l 1140-001 Standard length 170.0 140.9 161.7 142.5 159.0 190.5 218.5 Depth 52.4 39.2 46.0 38.0 45.7 56.3 62.0 Head length 53.4 44.6 50.5 41.7 45.5 54.8 64.0 Snout length 16.9 13.8 15.3 13.0 15.0 17.4 20.3 Eye diameter 17.1 15.0 16.0 13.3 15.0 18.0 20.0 Intcrorbital width 10.2 8.5 9.1 8.0 9.1 11.6 12.5 Suborbital width 6.8 4.8 5.4 5.3 6.1 8.0 10.2 Peduncle depth 17.9 13.8 16.2 14.4 15.0 20.0 21.7 Peduncle length 24.0 19.0 22.1 20.7 20.0 28.3 31.4 Dorsal fin length 91.7 72.2 85.1 74.3 83.7 99.1 112.3 Anal fin length 30.4 23.8 28.9 25.5 28.5 32.4 37.9 Pectoral fin length 48.6 39.5 49.0 35.5 43.8 54.5 63,3 Pelvic fin length 35.9 30.2 35.6 32.4 35.0 41.0 49.8 First dorsal spine length 12.0 10.3 9.7 11.4 12.5 12.4 14.1 Longest dorsal spine length 21.2 18.0 18.8 16.8 16.5 23.1 26.7 Longest dorsal ray length 27.2 20.8 25.2 24.4 27.1 32.2 42.1 First anal spine length 6 0 5.0 6.0 5.6 6.1 6.0 9.0 Second anal spine length 12.9 10.2 11.9 12.9 13.2 14.9 Third anal spine length 17.4 12.8 14.9 14.2 15.5 17.2 19.4 Preopercle scaly width 12.9 9.3 9.5 9.0 10.1 13.0 13.7 Preopercle naked width 5.7 5.1 4.7 4.5 5.6 5.7 7.2 36 New species of Nemipterus with anterior outer row of 4 pair of small, recurved canines. Colour in alcohol: body pale brown, with traces of silvery white on opercle and ventral surface. Faint dusky submarginal band along dorsal fin. Colour in life (from colour transparencies): head and body silvery white, dusky blue on dorsum; indistinct, pale yellow stripe from above pectoral fin to caudal peduncle; broad lemon-yellow band on either side of ventral midline; eye silvery; dorsal fin light dusky, upper margin of fin yellow-edged, with dusky mauve submarginal band; anal fin transparent; caudal fin dusky pink, caudal filament yellow; pelvic fins transparent: pectoral fins pink, with dusky base. Etymology. Named vitiensis after the main Fiji island, Viti Levu, where the holotype and 4 of the 6 paratypes were collected. Remarks. Nemipterus vitiensis appears to be most closely related to N. aurifilum (Ogi- lby, 1910), a species that is restricted to tropi¬ cal and subtropical eastern Australia, and N. bathybius Snyder, 1911, which ranges from southern Japan to northwestern Australia. All three species are characterised by having a long caudal filament and a yellow band on either side of the midventral line. However, the position of stripes on the sides of the body readily distinguishes fresh specimens of N. vitiensis from N. bathybius and N. aurifilum: N. vitiensis has a single, indistinct pale yellow stripe from above the pectoral fin to the caudal peduncle; N. bathybius has two yellow stripes, one from beneath the origin of the lateral line to the peduncle, and the other from behind the pectoral fin to the caudal base; and N. auri¬ filum has a single yellow stripe from beneath the origin of the lateral line to the peduncle. In preserved specimens all traces of striping on the body are lost. Meristic data are not useful in separating species of Nemipterus, and proportional meas¬ urements do not provide complete separation of the taxa. Selected meristic and morphom¬ etric data for N. vitiensis, N. aurifilum and N. bathybius are given in Table 2. Only a nar¬ rower interorbital width (1.6-1.8 in eye diame¬ ter) clearly separates N. vitiensis from N. auri¬ filum (interorbital width 1.3-1.5 in eye diame¬ ter), although vitiensis also tends to have a slightly longer snout (3.0-3.3 in head length, versus 3.3-3.9) and somewhat broader naked Table 2. Selected counts and proportional measurements (mm) of Nemipterus vitiensis, N. aurifilum and N. bathybius. N. vitiensis N. aurifilum N. bathybius Pectoral branched rays 14-16 13-15 13-15 Lateral line scales Transverse scale rows 46-48 46-49 46-49 above lateral line Transverse scale rows 3.5-3.5 3.0-3.5 3-4 below lateral line 10-11 9-10 9-11 Gill rakers (total) 12-14 14-20 13-16 Depth in SL 3.2-3.8 3.2-3.8 2.9-3.6 Head in SL 3.2-3.5 3.1-3.6 2.9-3.5 Depth in head length 1.0-1.1 1.0-1.2 0.9-1.1 Snout in head length 3.0-3.3 3.3-3.9 3.1-3.9 Eye in head length 3.0-3.2 2.6-3.4 2.7-3.6 Eye in snout length 0.9-1.0 0.7-1.0 0.7-1.2 Interorbital width in eye diameter 1.6-1.8 1.3-1.5 1.3-1.9 Suborbital w idth in eye diameter 2.0-3.1 2.2-4.7 1.9-4.7 Peduncle depth in peduncle length 1.3-1.4 1.2-1.6 1.0-1.4 Dorsal fin length in SL 1.9-2.0 1.8-2.1 1.8-2.1 Anal fin length in SL 5.6-5.9 5.4-6.1 5.0-6.0 Pectoral fin length in head length 1.0-1.2 1.0-1.3 1.0-1.4 Pelvic fin length in head length 1.3-1.5 1.3-1.6 1.1-16 Pectoral fin length in pelvic fin length First dorsal spine length 0.7-0.9 0.6-1.0 0.7-1.0 in longest dorsal spine Longest dorsal spine length 1.5-1.9 1.5-2.9 1.3-2.4 in longest dorsal ray First anal spine length in second 1.3-1.6 1.1-1.4 1.0-1.6 anal spine Second anal spine length in third 1.7-2.2 1.7-2.3 1.6-2.4 anal spine 1.2-1.3 1.2-1.4 1.1-1.4 Preopercle scaly width in naked width 1.8-2.3 1.4-1.9 1.5-2.2 preopercle (preopercle scaly width 1.8-2.3 in naked width, versus 1.4-1.9) than aurifilum. In Nemipterus vitiensis and N. bathybius the angle of the posterior edge of the suborbital is more acute, and a line drawn upwards from the posterior edge of the suborbital reaches the dorsal profile before the origin of the dorsal fin; while in N. aurifilum the angle of the posterior margin of the suborbital is distinctly obtuse, and a line drawn upwards from it reaches the dorsal profile at or behind the origin of the dorsal fin. Nemipterus vitiensis may also have a lower gill raker count (12-14) than N. aurifilum (14-20) or N. bathybius ( 13- 16), but only two of the specimens examined had intact gill arches and further material needs to be examined. Nemipterus vitiensis is so far known only from Fiji and Vanuatu (New Hebrides). Speci¬ mens were collected in depths ranging from 27-273 m. ACKNOWLEDGEMENTS I am grateful to Dr A.D. Lewis for collect¬ ing and photographing the specimens from Fiji, and Dr J.R. Paxton who permitted me to examine specimens in the Australian Museum. I thank P. Lastrico (FAO, Rome) for his draw¬ ing, based on a photograph of one of the 37 B.C. Russell paratypes. Dr J.E. Randall provided critical comments on the manuscript. REFERENCES Ogilby, J.D. 1910. On some new fishes from the Queensland coast. Endeavour Series (1):85- 139. Russell, B.C. 1986. Review of the western Indian Ocean species of Nemipterus Swainson 1839, with description of a new species. Senckenber- giana Biologica 67:19-35. Snyder, J.O. 1911. descriptions of new genera and species of fishes from Japan and the Riu Kiu Islands. Proceedings of the United States National Museum 40:525-549. Accepted 17 July 1990 38 The Beagle, Records of the Northern Territory Museum of Arts and Sciences, 1990 7(2):39-51 PRIMITIVE MARSUPIAL TAPIRS {PROPALORCHESTES NOVACULACEPHALUS MURRAY AND P. PONTICULUS SP. NOV.) FROM THE MID-MIOCENE OF NORTH AUSTRALIA (MARSUPIALIA: PALORCHESTIDAE). PETER MURRAY Northern Territory Museum of Arts and Sciences, GPO Box 2109, Alice Springs NT 0871, Australia. ABSTRACT The upper molar dentition of Propalorchestes novaculacephalus demonstrates a transitional state between the bilophodont marsupial tapirs and the selenodont wynyardiids. Although Propalorchestes had developed bilophodont crowns, the metacone and stylar cusp D remained sufficiently differentiated to verify the development of bilophodonty in diprotodontoid (vombatimorphian) marsupials from a selenodont condition, in which the primary buccal cusp is formed by stylar cusp D rather than the metacone. KEYWORDS: Palorchestinae, Wynyardiidae, molar evolution, diprotodontoid sys- tematics, Bullock Creek Local Fauna, Riversleigh “Systems” Fauna. INTRODUCTION A cranial fragment from the Bullock Creek Local Fauna indicated that tapir-like marsupi¬ als (Palorchestinae) were already highly modified forms by mid-Miocene times and that they differed in many significant respects from the palorchestid Ngapakaldia tedfordi Stirton (Murray 1986). Several years had elapsed since the cranium of Propalorchestes was described before any palorchestine denti¬ tions from the Camfield Beds came to light. What was thought to be the first to be recov¬ ered, a right maxillary fragment containing one and a half molars, was actually preceded several years by a dentary collected by Dr Thomas Rich in 1982, from the Horseshoe West locality of the Camfield beds. The den¬ tary and maxilla fragment are well-matched and of the correct proportions to belong to the same species as the edentulous cranium. Cast¬ ing further afield, Dr Mike Archer and Ms Sue Creagh assembled a small assortment of iso¬ lated teeth and dentary fragments from River¬ sleigh which have palorchestid features. The total collection of late Oligocene to mid-Miocene palorchestine material consists of only eight specimens: three dentary frag¬ ments, a maxilla, a cranial fragment and three isolated teeth. Despite the sparseness and frag¬ mentary condition of the sample, it substan¬ tially improves our resolution of the sys- tematics of the marsupial tapirs and moreover, adds a previously unknown transitional ele¬ ment to the interpretation of the phylogeny of bilophodont dentitions within the Vombati- morphia. The “diprotodontoid” (vombatimorphian) affinity as opposed to a macropodoid (Owen 1874: Raven and Gregory 1946) affinity of palorchestids was first noted by Woods (1958). Since then, palorchestid affines pos¬ sessing vombatimorphian character com¬ plexes were diagnosed by Stirton (1967) ( Ngapakaldia , Pitikantia). However, the mo¬ lar crown morphology of these genera showed few synapomorphic features to unite them with members of the genus Palorchestes. Ngapakaldia is usually treated as a plesiomor- phic sister group of the apparently more de¬ rived Palorchestinae (Stirton et al. 1967). However, Archer (1984) correctly observed that there is very little to support an even monophyletic union of the two groups or the state of the morphocline for their dental and cranial characters. 39 P.F. Murray Evidence is presented indicating that the distinctive upper molar morphology of Propalorchestes more closely resembles that of a wynyardiid than any other described, potentially antecedant, vombatimorphian group. Moreover, if the transitional state of the dentition is taken as an indication of the polarity of the morphocline, the Subfamiliae Incertae palorchestids (Ngapakaldia, Pitikan- tia) must represent a more derived state, in¬ deed. if they are even to remain within the Family Palorchestidae. The following abbreviations are used in the text: NTM, Northern Territory Museum; NMV, Museum of Victoria (formerly National Museum of Victoria); AR, Palaeontology Laboratory University of New South Wales; CPC, Commonwealth Palaeontological Col¬ lection. SYSTEMATICS Family Palorchestidae Archer and Bartholomai Subfamily Palorchestinae Tate Genus Propalorchestes Murray Propalorchestes Murray, 1986:195. Type species. Propalorchestes novacu- lacephalus Murray, 1986:195. Additional species. Propalorchestes pon- ticulus n. sp.; Propalorchestes cf. ponticulus. Diagnosis. To the diagnostic features given in Murray (1986) I add the following: dentary shorter, thicker, more massive relative to size of dentition than in Palorchestes painei', inter- coronoid sulcus absolutely and relatively wider and deeper, forming a definite crest extending to below M„ defining a deep sub- alveolar fossa. Horizontal ramus deepens an¬ teriorly in contrast to attenuated depth of ramus in Palorchestes painei. Lower molars absolutely smaller and much smaller relative to the size of the horizontal ramus. Crowns lower, interloph valleys less steeply inclined toward the buccal and lingual sides than in Palorchestes painei. Anterior cingulum thicker, buccal cingulum more clearly defined than in Palorchestes painei. Maxilla relatively shorter, wider and deeper than in Palorchestes spp., zygomatic root of maxilla thicker and projects laterally perpendicular to the tooth row in contrast to a definite posterior slant of the process in Palorchestes spp. Anterior sur¬ face of infraorbital fossa flat; infraorbital fo¬ ramen higher and relatively larger than in any Palorchestes species. Upper molars trapezoi¬ dal in outline shape, metaloph markedly nar¬ rower than protoloph. Buccal cingulum ab¬ sent. Stylarcusp E well developed; stylar cusp D forming primary buccal cusp of metaloph. metacone reduced to a swelling between the metaconule (“hypocone”) and stylar cusp D. Conspicuous postmetacrista slants buccally from the metacone to the base of stylar cusp E. Propalorchestes novaculacephalus Murray (Figs 1-2, 3A-B) Propalorchestes novaculacephalus Mur¬ ray, 1986:195-211, figs 1-6. Type material. HOLOTYPE - NTM P8552-I0: left and part of the right side of neurocranium including most of the zygo¬ matic arch and latex endocast (Murray 1986: figs 1 -6). PARATYPES - NTM P862-27: right maxilla with M 2 ' 3 (Fig. IA-B). NMV P187282: right dentary, containing M, 5 (Fig. 2A-B), Camfield Beds, Northern Territory. AR 9682: isolated right M\ damaged on the buccal side, Riversleigh Queensland (Fig. 3B). Type locality: Camfield Beds “...16 miles Southeast of Camfield Homestead in the north central Northern Territory” (Plane and Gate¬ house 1968). Formation. Camfield beds, light coloured calcareous sandstone, siltstone and limestone with basal ferruginization and silification at the top. Age. Mid-Miocene, determined on the basis of the stage of evolution of diprotodonlid species (Woodburne et al. 1985). Description. Lower molars: M 25 essen¬ tially homomorphic; lower molars rectangu¬ lar, shallowly waisted at the midvalley and equidistant between the trigonid and talonid (Fig. 2A-B). Relatively long and narrow (length to width proportions about 1.5: 1.0). Molar gradient a very slight increase from front to back, (M 3 is 98% of the length of M,). Molars are about 15% shorter than Palorch¬ estes painei and much smaller than in other Palorchestes species. Breached lophids have crescentic, slightly oblique cementum exposures with basinlike expansions on the protoconid and hypoconid. Protoconid large, offset antcrolabially and divided anteriorly from the metaconid by a deep sulcus originating from a fossette devel¬ oped in the lingual margin of the precingulid. 40 Primitive marsupial tapirs Fig. 1 . Maxilla fragment of Proplorchestes novacu- lacephalus, paratype NTM P862-27. A, lateral aspect; B, ventral aspect. Abbreviations; iof, infraorbital foramen; mxs. maxillary sinus; zpr, zygomatic process; nips, maxillopalatine suture. In worn molars this sulcus results in a narrow¬ ing of the lophid by an invaginated crescent of enamel similar to the neomorphic cuspid de¬ scribed in Muramura williamsi (Pledge, 1987). Precingulid extends from labial side of an¬ terior portion of protoconid, ascending and expanding slightly to the midline of the crown then descending abruptly before reaching the lingual side of the crown. The area of maxi¬ mum expansion and height of the precingulid forms a small cuspid (paraconid). Cristid obliqua (midlink) broad, low, complete, situ¬ ated near the longitudinal midline of each molar. Faint, rounded labial cingulid present on all molars; interlophid valley on either side of the midlink broad, shallow and gently tilted to¬ wards the margins of the crown. Anteropos- teriorly deep, rounded postcingulid extending slightly higher up the crown lingually than labially; connected to the middle of the hy- polophid by a stout post-link demarcated on either side by deep, crescentic sulci. (Meas¬ urements approximate; M 3 length 16.6; ante¬ rior width 10.6; posterior width 10.0; M 4 length 16.5, anterior width 10.0; posterior width 11.0; M 5 length 17.1; anterior width 11.0; poste¬ rior width 11.2). Upper molars: M 2 broad, bilophodont, low crowned and trapezoidal in occlusal outline with the protoloph markedly wider than the metaloph (Figs 1A-B, 3A). Labial side of the crown composed of four stylar cusps. Moie¬ ties deeply divided labially by a sulcus extend¬ ing from the midlink down to the cementoe- namel contact of the crown. Protoloph broader and wider than metaloph; precingulum wide and deep anteroposteriorly, merging anterola- bially with a short postparastylar crest termi¬ nating in a small paraconal facet. Stylar cusp C. separated from the paracone by a shallow labial sulcus, forms the largest and buccal-most cusp of the protoloph. The posterior enamel crest of the breached loph curves sharply mesiad to form a short remnant of the centrocrista. The interloph valley is narrow and V-shaped, slightly expanded lin¬ gually behind the protocone to terminate at a well-developed lingual cingulum (postproto¬ crista). Within the interloph valley, slightly labial to the midline of the crown, a small rhomboidal fossa separates two enamel bulges at the base of each loph lingually and the remnant centrocrista labially, forming a double midlink-like structure. The posterior moiety of the crown retains a strong triangular outline due to the promi¬ nence distobuccally of stylar cusp E. The la¬ bial cusp of the metaloph is composed of stylar cusp D which is separated from stylar cusp E by a deep sulcus. The metacone is represented by an enamel thickening about midway along the metaloph, and is connected to stylar cusp E by a long, obliquely oriented postmetacrista. The posterior surface of the metaloph angles anterolingually towards the metaconule (“hypocone”) presenting a smooth, steeply angled face, emarginated by a well-developed postcingulum, situated low on the crown. M 3 is similar to M 2 , with a slightly narrower metaloph and a less obliquely oriented and less prominent, though still remarkably ro¬ bust, postmetacrista. (Measurements; M 2 length 16.8; anterior width, 13.2; posterior width, 12.0; M 3 estimated length, 15.7; poste¬ rior width, 11.6). A possible M J is represented by a system C Riversleigh specimen (AR9682, paratype). This isolated, lightly worn crown (16.0 mm 41 P.F. Murray Slf Fig. 2. Right dentary of Propalorcliestes novaculacephalus , paratype NMV PI 87282. A, lateral aspect; B, dorsal aspect. Abbreviations: inf. masseteric fossa; sas, subalveolar sulcus; smc, submassteric crest; pmf, posterior mental foramen; ics, intercoronoid sulcus; cp, coronoid process; mrt, mid-ramal torus; pft, pit for fleshy temporalis muscle; ptf, pterygoid fossa; pap, postalveolar process; pas. postalveolar shelf; slf, sublingual fossa; tt, transverse torus; gp, genial pit. 1-CP long) shows all of the previously noted fea¬ tures in a slightly more suppressed form than the M 3 of the holotype. Because the M 3 of NTM P862-27 shows a similar reduction of its features, a gradient along the molar row is likely. However, AR9682 could also be inter¬ preted as a slightly more derived M 3 of Propalorchestes novaculacephalus. Dentary: The right horizontal ramus of Propalorchestes novaculacephalus (NMV PI87282) is distinctive in its robustness and strong angularity. Compared to Palorchestes painei. P. novaculacephalus is slightly shorter, but significantly thicker (23.8 mm thick below M 3 4 ), and markedly triangular in section. In contrast to P. painei , the dentary 42 Primitive marsupial tapirs slightly increases in depth towards the sym¬ physis, rather than tapering forward (42.0 mm deep below P 3 ; 40.0 mm below M,). Relative to the length, depth and thickness of the hori¬ zontal ramus, the cheek teeth of P. novacu¬ lacephalus are proportionally much smaller than in P. painei. The 18.0 mm wide furrow¬ like intercoronoid sulcus, which is a conspicu¬ ous feature of palorchestines, is exceptionally broad in P. novaculacephalus and is continued forward to form a wide, sloping, shelf-like crest extending 25.0 mm from the molar alve¬ oli to the mid region of the dentary. In other Palorchestes species a notable, though less distinguished (due to the weakness of the coronoid crest) subaiveolar fossa is likewise present. In P. novaculacephalus , the coronoid crest is prominent and sharp; grading into a definitive mid-ramal torus below M r The lower portion of the ascending ramus is more produced laterally, and more confined dorsoventrally than in other palorchestines. Consequently, the masseteric fossa is deeper and the submasseteric crest ascending to the posterior masseteric eminence is thicker and more prominent. The pterygoid fossa is rela¬ tively deeper and more confined than in P. painei. As in other palorchestines, the digas¬ tric fossa is shallow and indistinct. The postalveolar shelf is significantly shorter than in other palorchestine species. The postalveolar crest is higher, more robust and more laterally deflected. A distinct 10.0 mm diameter fossa for the fleshy temporalis insertion in the posterior end of the intercoro¬ noid furrow in P. novaculacephalus is absent or indistinct in Palorchestes species. Propalorchestes novaculacephalus has large, deep genial pits in the lower mid-poste¬ rior surface of the dentary symphysis, and a thick, though short transverse torus. Palorch¬ estes painei has small genial pits located rela¬ tively higher on the posterior surface of the symphysis. No transverse torus is present. The dentary symphysis of P. novaculacephalus is shorter, more rounded, relatively deeper than in P. painei and apparently slightly less decur- ved. Relative to the posterior margin of the symphysis and in combination with its less horizontal inclination, the diastema was per¬ haps shorter in P. novaculacephalus than in P. painei. The diastemal crest descends towards the incisor alveolus less steeply than in P. painei, but overall, the dentaries of the two species appear to have been very similar. Maxilla: The maxilla (paraty pe NTM P862- 27) containing previously described M 2 ' 3 , corresponds in size and robusticity to the dentary (Fig. 1A-B). The occlusal surfaces and thegotics of the assigned upper and lower dentitions match closely. Although there are no contact points with the cranium (holotype, NTM P8552-10) the maxilla readily corre¬ sponds to the projected missing proportions in relation to its shape and the thicknesses of its broken margins. Compared to the same region in other Palorchestes species, the mediolateral thick¬ ness of the maxilla above M, is much greater in Propalorchestes novaculacephalus due to the formation of a large sinus. This sinus is not present in P. parvus and is either reduced or absent in P. painei. The expanded maxillary antrim creates a conspicuous swelling beneath the infraorbital foramen in P. novaculacepha¬ lus, and a deep groove for the emergent in¬ fraorbital nerve is inscribed on its external surface. Unlike P. parvus, in which the facial plane slants posteriorly, the anterior surface of the zygomatic root abruptly projects at a right angle to the surface of the maxilla. In some respects, P. novaculacephalus more closely resembles P. painei, in which the zygomatic roots and jugals must have presented a broad, flat facial plane extending around the orbits and nearly up to the corresponding lateral expansions of the frontal crests. However, the entire facial plane of P. painei also retreats more than in P. novaculacephalus. The rela¬ tively large diameter of the infraorbital nerve canal and foramen (10.0 mm by 8.0 mm) and its slightly elevated position compared to P. parvus, are also common to P. painei. Judging from the more anterior position of the maxillopalatine suture, which transects the palate between M 2 and M\ the palate must have been proportionally different from P. painei. This could, of course, indicate that the molar series is M 3 ' 4 , but relative to the zygo¬ matic root, the position of the molars corre¬ spond closely to M 2 3 in Palorchestes painei. However, assuming that a more primitive condition was manifest, similar to that of Murarnura williamsi (Pledge, 1987), the posi¬ tion of the molars relative to the maxillary process of the zygoma in P. novaculacephalus might eventually be found to differ substan- 43 P.F. Murray tially from that of Palorchestes species. In any case, I have taken the least speculative course, and await the resolution of the problem from a more complete specimen. The internal surface of the maxilla pre¬ serves the contour of the internal nares, which was relatively more expanded than in the cor¬ responding regions of P. parvus. Palorchestes painei appears to show an intermediate condition between P. novacu- lacephalus and P. parvus. Overall, the maxil¬ lary fragment indicates that the midfacial region in P. novaculacephalus was relatively more laterally expanded and deeper than in any species of Palorchestes and perhaps that the palate was shorter. The large infraorbital foramen and deep, broad, perpendicular facial plane of the zygoma, ostensibly for the attach¬ ment of nasolabialis musculature, strongly suggest the presence of a trunk-like append¬ age in this species. There is however, an indication that the narial aperture may have been much wider transversely and perhaps relatively shorter than in P. painei. In all Palorchestes species, the plane of the zygoma slopes posteriorly about 10 degrees relative to the occlusal line, whereas in Propalorcliestes novaculacepha¬ lus it slopes to about the same extent anteri¬ orly, as in more typical diprotodontoid marsu¬ pials. This indicates that the nasofrontal re¬ gion projected more anteriorly than in Palorchestes painei , which would result in a greater overhang of even highly reduced nasal bones. As no portion of the rostrum of this species has been recovered, I can only con¬ clude that while Propalorcliestes novacu¬ lacephalus probably had a well-developed tapir-like trunk, its proportions and perhaps its extent of mobility differed somewhat from the more evolved Palorchestes species. In conjunction with its cranial morphology (holotype NTM P8552-10), Propalorchestes novaculacephalus had a deep cranium, with a high, narrow braincase and robust, wide zygo¬ matic arches which terminated abruptly in a wide, flat facial platform for the attachment of the facialis muscles of the trunk. The rostrum was probably relatively broader, shorter and less downturned than in Palorchestes painei. The fragmentary remains of this species indi¬ cate that the more primitive palorchestines had fully developed, by mid-Miocene times, a characteristic masticatory complex which is distinctive for the subfamily (Murray 1986). Indeed, no other species within this lineage shows this constellation of characters as con¬ spicuously, relative to the size of its dentition and braincase. Presumably, as palorchestines increased in size, an allometric adjustment of the adductor complex resulted in the modera¬ tion of some of these features in later forms. Propalorcliestes ponticulus n.sp. (Fig. 3C) Type material. HOLOTYPE - AR1802: isolated right M 2 (Fig. 3C), D-site (System A, Archer et al. 1989) Riversleigh, Queensland. Provisionally referred material. AR11302: ?M S (Fig. 3E,F), Wayne’s Wok Locality (System B, Archer etal. 1989) River¬ sleigh Queensland. AR1779: dentary frag¬ ment (Figs. 3D; 4A-C), fully fused symphysis with portion of M 2 , no site designated, River¬ sleigh. Queensland. SGM1008; fragment of dentary, hemimandibles fully fused at the symphysis with P 3 -M, 3 crowns sheared off (Fig. 4D-F), Sticky-Beak locality (System A, Archer et al. 1989), Riversleigh Queensland. Type locality. “D-Site”, northern tip of Carl Creek Limestone exposures, 2.4 km South of Verdon Creek (Tedford, 1966). Formation. Carl Creek Limestone: mottled, vuggy cream and yellow limestone. Age. Late Oligocene or early Miocene (Tedford 1966; Archer et al. 1989). Diagnosis. Smallest palorchestine, with cheek teeth similar in size to Pitikantia ; char¬ acteristic, nearly equal-sized, narrow lophids and anterior displacement of the protoconid; low, centrally divided and labially situated cristid obliqua, weaker than in other palorch¬ estines, deep, tapering postcingulid and hin- dlink. Differs from other palorchestines but resembles Pitikantia dailyi in having a more labially situated cristid obliqua, deeply notched in the interlophid valley and travers¬ ing a narrower declivity between the lophids; differs from Pitikantia dailyi in having a sec¬ ondary swelling on the anterolingual side of the hypolophid, in having a clearly defined preprotocristid immediately mesial to the protoconid and a broad shallow sulcus in the middle of the anterior surface of the protolo- phid of M y Description. Considered to be an M 2 on the basis of its narrow interproximal facet and less expanded protoconid (Fig. 3C); it is also pos- 44 Primitive marsupial tapirs med end cm Fig. 3. Upper and lower molars of Propalorchesles species. A, upper right second molar of Propalorchesles novaculacephalus, paratype NTM P862-27; B, upper right third or fourth molar of Propalorchesles novaculacepha¬ lus. paratype AR9682; C, lower right second molar of Propalorchesles ponticulus sp. nov., holotype AR1802; I), lower left second molar of Propalorchesles cf. ponticulus, SGM 1008; E, occlusal aspect of left lower '.’fifth molar of Propalorchesles cf. ponticulus. ARI1302; F, labial aspect of left lower ?fifth molar of Propalorchesles cf. ponticulus. AR11302. Abbreviations: E. stylar cusp E; D. stylar cusp D; C, stylar cusp C; B, stylar cusp B; par, paracone; ppc, preprotocrista; pro, protocone; Ic, lingual cingulum (postprotocrista); mcl, metaconule (“hypocone"); pmcc, postmetaconulecrista; met, metaconc; prac, postmetacrista; ml, “midlink"; end, entoconid; Is, lingual swelling of the base of the hypolophid; med, metaconid; pc, precingulid; prd, protoconid; fp, fossa behind protoconid; co, cristid obliqua; hyd, hypoconid; pi, postlink; poc, postcingulid. 45 P.F. Murray sible that the more prominent preprotocristid is a gradient-related feature, as no complete Propalorchestes novaculacephalus cheek¬ tooth series is known. Compared to P. novacu¬ lacephalus, the roots are small relative to the crown and less divergent from one another anteroposteriorly. Judging from the narrower interloph valley, and depth of the breached lophids, the lophids were relatively high and steep. The crown base lacks the distinctive narrow swellings above the cementoenamel junction in P. no¬ vaculacephalus. Although the protoconid is offset anterobuccally as in other palorchesti- nes, it is not conspicuously hypertrophied. This feature may also relate to the molar gra¬ dient, but extrapolating from the condition in P. novaculacephalus , the condition is adduced to be a characteristic of the entire molar series. The cristid obliqua is divided in the mid¬ valley rather than continuous as in P. novacu¬ lacephalus. It is situated in the buccal third of the interlophid valley rather than traversing the longitudinal midline of the tooth. This feature is also present on the M 2 of Pitikantia dailyi. The anterior surface of the protolophid is concave due to a shallow sulcus commenc¬ ing immediately above the precingulum. This concavity is accentuated by a distinct but low, broad and steep preprotocristid (forelink) and a low, broad swelling emanating from the base of the metalophid on the lingual side. The anterior surface of the hypolophid is analogous to the protolophid. The labially situ¬ ated cristid obliqua delineates a shallow con¬ cavity in the middle third of the tooth, accen¬ tuated lingually by a distinct swelling of the enamel on the anterior surface of the lingual third of the hypolophid. The lingual swelling, resembling a faint mid-link, was a low bulge in the enamel that only becomes an obvious fea¬ ture of the hypolophid structure after advanced attrition of the crown. The labial interlophid valley is more dis¬ tinctively concave behind the protolophid than in P. novaculacephalus. This indicates that the posterior surface of the protolophid was more hollowed-out and probably more posteriorly recurved on the labial protolophid apex than in P. novaculacephalus. Like P. novaculacepha¬ lus, the interloph surface was less cambered labially and lingually than in other palorch- estines. A short, weak lingual cingulum is present. (Measurements: length, 14.5; ante¬ rior width 9.2; posterior width, 10.5). Etymology. Latin, ponticulus “little bridge” in reference to the weaker, shorter, partially divided midlink and alluding to its apparent phyletic position relative to other palorchestids. Propalorchestes cf. ponticulus (Figs 3D-F, 4) Material. AR11302: a perfectly preserved lower molar crown of a palaorchestine marsu¬ pial from the Wayne’s Wok Locality, River- sleigh, Queensland (Fig. 3E,F). According to Archer et al. (1989) Wayne’s Wok is a system B (provisionally early-Miocene) equivalent. Remarks. The specimen may represent an encrypted M s in which the enamel surface was only just completed and was perhaps slightly dehydrated before fossilization. However, the specimen preserves, in a conservative way, some of the key features of Propalorchestes ponticulus and is therefore assigned provi¬ sionally to that species. These characters are: protolophid offset as in typical palorchestines, but not greatly enlarged at the base, well de¬ veloped preprotocristid, which however, be¬ comes faint as it reaches the precingulid; a shallow sulcus on the anterior surface of the protolophid; a somewhat hollow posterior surface of the protolophid immediately below the protolophid; a swelling at the base of the anterior surface of the hypolophid; crown base not swollen above the cementoenamel junc¬ tion; divided cristid obliqua confined to the labial side of the interloph valley, though stronger, higher and not situated as labiad as in the holotype of Propalorchestes ponticulus (AR1802). The crown is slightly longer (0.3mm) and relatively broader anteriorly (0.8mm) than the holotype, although the slight front to back molar size increase in palorchestines could easily account for this. The primary differ¬ ences are that the protolophid is slightly wider than the hypolophid (characteristic of the M 5 of Palorchestes species), and that the floor of the interloph valley is more cambered on ei¬ ther side of the midlink. (Measurements: length 15.1, anterior width, 10.0, posterior width, 9.6). Though slightly smaller and otherwise less derived than other palorchestine species, the lophids of P. ponticulus are high, sulcate an¬ teriorly and recurved posteriorly as in the later palorchestines. Because all of the P. novacu- 46 Primitive marsupial tapirs lacephalus specimens are heavily worn, it can only be inferred that they were similarly high- crowned but larger, thicker, with a wider inter- loph valley and a relatively larger protoconid. The provisionally referred specimen AR1779 is a dentary fragment with both hemimandibles fully fused at the symphysis. The anterior portion of the protolophid and the lingual sides of both lophids of M, are broken and the lophids are severly worn. Based on the size of the crown base and roots, the River- sleigh M, is slightly smaller than the M, of the Camfield P. novaeulacephalus (Figs 3D, 4A- C). Fig. 4. Dentary fragments of Propalorchestes cf. ponticulus. These specimens share a variety of features with Pitikantia, Ngapakaldia and Propalorchestes. A, anterior portion of the dentary of AR1779, lateral aspect: B, dorsal aspect of ARI779; C, ventral aspect of ARI779: D, lateral aspect of SGM 1008. Abbreviations: dc, diastemal crest; gc, genial crest; gp, genial pit; slf. sublingual fossa; ia. incisor alveolus; pmf, posterior mental foramen. The profile of the dentaries more closely resemble those of the genera Ngapakaldia and Pitikantia than the genus Palorchestes; like the former genera, there is an apparently greater front-to-back molar gradient and the incisor implantation is more like that of Ngapakaldia than Palorchestes. Like Propalorchestes and Palorchestes, the dentary symphysis is fully ankyloscd, with a long, deep sublingual fossa, but the genial pits are located on the ventral surface rather than posterior to the symphysis. Because A-C (AR1779) retains a lower second molar that resembles the holotype of P. ponticulus, and because the dentary morphology of D-E (SGM 1008) closely resembles AR1779, I have provision¬ ally included them in the species P. ponticulus. These fragments support a close relationship of Pitikantia- like forms to the palorchestines and may represent a stage of evolution close to their divergence. 47 P.F. Murray The remnants of the crown reveal a nar¬ rower, steeper interloph valley than in P. no- vaculacephalus and the presence of an enamel bulge on the anterior surface of the base of the entoconid is only just apparent. The midlink is short, though well-developed and situated immediately labiad of the midline of the tooth. The dentary fragment is more informative. The portions in common with P. novacu- lacephalus indicate that AR1779 was not as deep or thick tranversely through each hem- imandible. The symphysis is shallower posteriorly and less steeply inclined. The genial pits are lo¬ cated on the ventral surface of the symphysis rather than comparatively high up on the pos¬ terior surface. As in later palorchestines, the diastemal crests are prolonged anteriorly to form a deep, U-shaped sublingual trough. The provisionally referred specimen SGM1008 is a more complete dentary frag¬ ment, but lacking cheek tooth crown details (Fig. 4 D-F). It is therefore aligned with the formerly described specimen on the basis of its overall morphological similarity and the comparision of the sizes of the cheek tooth alveoli. Although the P, crown is damaged on the lingual side, the occiusal and lateral profile of the tooth, combined with its size and pro¬ portions, the steepness and bredth of the ante¬ rior cuspid and the indication of a sulcus just lingual to the preprotocristid all favour an incipiently palorchestine morphology. The proportions of the crown bases of the molars and their alveoli indicate that the molar gradient was very slight, as in all palorchesti¬ nes. The lateral surface of the anterior portion of the horizontal ramus is flat, moderately deep with a broad, rounded inferior border. The gently upturned symphysis is long, and narrows gradually towards the incisor alveoli. The incisor alveoli are large, vertically ori¬ ented, oval cavities separated throughout their considerable depth by a thin bony septum. The implantation angle of the incisors was shal¬ low, but due to a lack of symphyseal deflec¬ tion, typical of Palorchestes species, SGM1008 had only moderate procumbency of the incisors, comparable to that of Pitkantia dailyi (Stirton, 1967). As in all known palorch¬ estines, the diastema is exceptionally long and forms a narrow trough extending to the base of the incisor crowns. In contrast to P. novaculacephalus, the genial pits are located on the ventral surface of the symphysis as in AR1779. These are sepa¬ rated by a prominent genial crest that extends anteriorly within the sublingual sulcus to be¬ low the P v A remnant of this crest is visible in AR1779, immediately anterior to the level of P,, and as the posterior portion of the sym¬ physis is severely weathered, the genial crest must have been equally well developed. Although the fossil material is fragmentary, I am confident that the presence of these two palorchestine species in the lower Oligocene and mid-Miocene of Australia will eventually be confirmed by more complete specimens. Propalorchestes ponticulus is a less robust form, that could be ancestral to P. novacu¬ lacephalus as the specimens appear to be confined to system B and A faunas at River- sleigh. Experience with other genera from Camfield, Riversleigh and Wipajiri (Wakaleo spp., Neohelos tirarensis) has shown that such species distinctions as I have just made be¬ come blurred when larger samples are ob¬ tained. It is therefore with the same reserva¬ tion that Murray and Megirian (1990) place upon the distinction of W. vanderleueri from W. otdfieldi that I designate P. ponticulus as distinct from P. novaculacephalus. A closer relationship of SGM 1008 to the genus Piti- kantia cannot be ruled out. However, that Riv¬ ersleigh specimen differs from Pitkantia dai¬ lyi in being more robust, in possessing a pos¬ terior mental foramen, in having an apparently more strongly ankylosed symphysis and in showing some differences in the cheektooth crown proportions and morphology. CONCLUSIONS Tedford et al. (1977) first noted that wynyardiid upper molars show a transitional stage in the development of hilophodont mo¬ lars in which the labial side of the lophs are composed of stylar cusps B, C and D, with the metacone and paracone taking up a median position within gradually evolving transverse crests. This evidence supported the observa¬ tions of Winge (1941), Ride (1971) and Archer (1976) that a selenodont pattern resembling phascolarctids and peramelids, gave rise to the highly developed, albeit superficially uncom¬ plicated bilophodonty characteristic of, for example, the nototherine and zygomaturine Diprotodontidae. The traditional descriptive nomenclature (Stirton et al. 1967) applied to the Diprotodontidae assumed that the hy- 48 Primitive marsupial tapirs DIPROTODONTIDAE I PALORCHESTIDAE _ WYNYARDIIDAE Fig. 5. Phenogram depicting a hypothesis of the evolution of bilophodont “diprotodontoids" from a selenobilopho- dont ancestor partially based on the morphology of the second upper molar. A. Neohelos (Zygomaturinae); B, Pyramios (Diprotodontinac); C, Ngapakaldia (Subfamily Incertae); D, Propalorchestes (Palorchestinae); E. Mura- mura (Wynyardidiidae, Subfamily Incertae). I, (symplesiomorphic) expression of “posterobuccal triangle"; 2, (synapomorphic) loss of “posterobuccal triangle”; 3, (synapomorphic) suppression of the metaconc; 4, (synapomor- pliic) widening of intcrloph valley and transverse expansion of lophs. The postulated connection between Propalorch¬ estes and wynyardiids is deduced from phenetic resemblances; possible synapomorphies include the robust, elogated postmetacrista and the distinct swelling in the middle of the protoloph that could have been a precursor to the double midlink complex (?centrocrista + postparacrista) in Propalorchestes. These resemblances are more compelling than with other sclenodont vombatimorphians or phascolarctomorphians. poconc and the metacone formed the respec¬ tive lingual and buccal cusps of the metaloph. Tedford et al. (1977: fig. 3A-D) illustrated a series of second upper molars of a wynyardiid, two vombatoids and a phasco- larctid from the Pinpa Local Fauna. This series clearly demonstrated the cusp homologies from a selenodont phascolarctid in which the hypocone was not present, to an incipiently bilophodont wynyardiid, that unambiguously demonstrated the intraloph position of the metacone, absence of the hypocone and labial position of stylar cusp D, to form the metaloph of the molar, with the metaconule as the proxy “hypocone”. Although wynyardiids such as Muramura williamsi Pledge (1987) and Namilamdeta snideri (Rich and Archer, 1979) are incipi¬ ently bilophodont, they retain a predominately selenodont pattern. The possibility of para- phyletic bilophodonty could not be ruled out, particularly with repect to the diprotodontids, among which no clear representation of the conspicuous posterobuccal triangle of cusps (D, E and metacone) remains. Propalorchestes novaculacephalus is fully bilophodont but retains the “posterobuccal triangle” of the wynyardiids and demonstrates the reduction of the metacone and hypertro¬ phy of the metaconule to form the metaloph 49 P.F. Murray (Fig. 5). Moreover, a hypso-bilophodont pat¬ tern, in which the postmetacrista and discrete expression of the metaconule and stylar cusp D, obscured by the development of the loph, had developed in parallel among the Palorch- estinae. Among species of the more derived genus Palorchesies, a fully bilophodont pat¬ tern is present, and only by tracing the specific details of its transformation can the homolo¬ gous structures relating to the obfuscated se- lenodont heritage of the lineage be under¬ stood. It is by close analogy to this subtle transi¬ tion, that I consider the characteristic bilopho- donty of the Diprotodontidae to have occurred. By association with synapomorphic features in the basicranium (the large epitympanic fenestra) and the dentition (P 3 morphology) we can assume a close relationship between the fully bilophodont genus Ngapakaldia and the Palorchestinae. Because Ngapakaldia has further suppressed its selenodont heritage, it represents a state, too derived to be directly ancestral (e.g. absence of a postmetacrista), but immediately post-transitional, to the dip¬ rotodontidae (Zygomaturinae, Diprotodonti- nae). SUMMARY The genus Propalorchestes was initially diagnosed on the basis of a cranial fragment representing a primitive palorchestine. Denti¬ tions assigned to this genus are likewise ple- siomorphic, and show the transition between selenodont wynyardiid or wynyardiid-like dentitions and the fully evolved bilophodont dentitions that characterize members of the family Diprotodontidae. Derived palorches- tids paralleled the diprotodontids by also in¬ corporating stylar cusp D into the metaloph while suppressing the metacone; therefore little, if any, indication of this state is evident in the later Tertiary genus Palorchesies. Palorchestine palorchestids were present in the late Oligocene as evinced by Propalorch¬ estes ponticulus. In retaining a remnant meta¬ cone and a large postmetacrisla, Propalorch¬ estes novaculacephalus from the mid-Mio¬ cene (post-Wipajiri) Camfield beds appears be less derived than the Riversleigh system C specimen. The distinctive features of the up¬ per molar crowns observed for P. novacu¬ lacephalus from Camfield are obviously pres¬ ent, though suppressed in the Riversleigh example. Although this could be attributed to changes along the molar gradient, the possibil¬ ity of a chronological difference has not been eliminated. ACKNOWLEDGEMENTS I thank Thomas Rich, Mike Archer and Sue Creagh for providing a substantial portion of this interesting material. Thanks to Dirk Megirian for preparing the Bullock Creek maxillary fragment of Propalorchestes and thanks also to Tim Flannery for constructive comments on the manuscript. REFERENCES Archer, M. 1976. Phascolarctid origins and the potential of the selenodont molar in the evolu¬ tion of diprotodont marsupials. Memoirs of the Queensland Museum 17:367-371. Archer, M. 1984. The Australian marsupial radia¬ tion. In: M. Archer and G. Clayton (eds.) Vertebrate Zoogeography and evolution in Australasia. 633-808. Hesperian Press: West¬ ern Australia. Archer, M. and Bartholomai, A. 1978. Tertiary marsupials of Australia; a synoptic review. Alcheringa 2:1-19. Archer, M., Godthelp, H., Hand, S., and Megirian, D. 1989. Fossil mammals of Riversleigh, northwestern Queensland: preliminary over¬ view of biostratigraphy, correlation and envi¬ ronmental change. Australian Zoologist 25(2):29-65. Murray, P. 1986. Propalorchestes novaculacepha- lus gen. et sp. nov., a new palorchestid (Dipro- todontoidea: Marsupialia) from the middle Miocene Camfield Beds, Northern Territory, Australia. The Beagle. Occasional Papers of the Northern Territory Museum of Arts and Sciences 3( 1): 195-211. Murray, P. and Megirian, D. 1990. Further obser¬ vations on the morphology of Wakaleo van- derleueri (Marsupial ia:Thylacoleonidae) from the mid-Miocene Camfield beds. North¬ ern Territory. The Beagle, Records of the Northern Territory Museum of Arts and Sci¬ ences 7( 1 ):91 -102. Owen, R. 1874. On the fossil mammals of Australia part IX. Family Macropodidae; Genera Ma¬ cropus. Pachysiagon. Leptosiagon, Procopto- donand Palorchesies. 2 Volumes. Philosophi¬ cal Transactions of the Royal Society: Lon¬ don. Plane, M. and Gatehouse, C. 1968. A new verte¬ brate fauna from the Tertiary of Northern 50 Primitive marsupial tapirs Australia. Australian Journal of Science 30:271-273. Pledge, N. 1987. Muramura williamsi, a new genus and species of ?Wynyardiid (Marsupialia: Vombatoidea) from the Middle Miocene Etadunna Formation of South Australia. In: M. Archer (ed.) Possums and Oppossums: studies in evolution. 393-400. Surrey Beatty and Sons and the Royal Society of New South Wales: Sydney. Raven, H. and Gregory, W. 1946. Adaptive branch¬ ing of the kangaroo family in relation to habitat. American Museum Novitates 1309:1- 33. Rich, T. and Archer, M. 1979. Namilamadeta snid- eri, a new diprotodontan (Marsupialia Vomba¬ toidea) from the medial Miocene of South Australia. Alcheringa 3:197-208. Ride, W. 1971. On the fossil evidence of the evolu¬ tion of the macropodidae. Australia Zoologist 16:6-16. Stirton, R. 1967. The Diprotodontidae from the Ngapakaldi Fauna, South Australia. In: R. Stirton, M. Woodburne and M. Plane (eds.) Tertiary Diprotodontidae from Australia and New Guinea. Bureau of Mineral Resources Geology and Geophysics Bulletin 85:3-35. Stirton, R., Woodburne, M. and Plane, M. 1967. The phylogeny of Tertiary Diprotodontidae and its significance in correlation. In: R. Stirton, M. Woodburne and M. Plane (eds.) Tertiary Diprotodontidae from Australia and New Guinea. Bureau of Mineral Reseources Geology and Geophysics Bulletin 85:151-160. Tate, G. 1948. Results of the Archbold Expeditions No.59. Studies on the anatomy and phylogeny of the Macropodidae (Marsupialia). Bulletin of the American Museum of Natural History 91:233-352. Tedford, R., 1966. Fossil mammal remains from the Tertiary Carl Creek Limestone, north¬ western Queensland. Bureau of Mineral Re¬ sources Geology and Geophysics Bulletin 92:217-237. Tedford, R., Archer, M., Bartholomai, A. Plane, M., Pledge, N., Rich, T., Rich, P. and Wells, R. 1977. The discovery of Miocene vertebrates. Lake Frome Area, South Australia. Bureau of Mineral Resources Journal of Australian Geology and Geophysics 2:53-57. Winge, H. 1941. Interrelationships of the mam¬ malian genera. C.A. Reitzels Foriag: Copen¬ hagen 1. Woodburne, M., Tedford, R., Archer, M., Turn- bull, W., Plane, M. and Lundelius, E., 1985. Biochronology of the continental mammal record of Australia and New Guinea. Special Publication of the South Australian Depart¬ ment of Mines and Energy 5:347-363. Woods, J. 1958. The extinct marsupial genus Palorchestes Owen. Memoirs of the Queens¬ land Museum 13:177-193. Accepted 31 August 1990 51 The Beagle, Records of the Northern Territory Museum of Arts and Sciences, 1990 7(2):53-80 ALKWERTATHERIUM WEBBI, A NEW ZYGOMATURINE GENUS AND SPECIES FROM THE LATE MIOCENE ALCOOTA LOCAL FAUNA, NORTHERN TERRITORY (MARSUPIALIA: DIPROTODONTIDAE). PETER MURRAY Northern Territory Museum of Arts and Sciences, GPO Box 2109, Alice Springs NT 0871, Australia. ABSTRACT Alkwertatherium wehbi , gen. et sp. nov., possesses characteristics structurally intermediate between the diprotodontine genus Pyramios and the zygomaturine genus Plaisiodon of the family Diprotodontidae. In its expression of a clearly differentiated and well-developed parastyle on the P\ Alkwertatherium is aligned with the Subfamily Zygomaturinae. In its absence of a hypocone on the P 3 and its markedly constricted diastemal palate, Alkwertatherium resembles the members of the Subfamily Diprotodontinae. Alkwertatherium was derived from diprotodonti- nes at a structural grade similar to that of Pyramios. It may have shared ancestry with the primitive zygomaturines Plaisiodon and Nimbadon. Pyramios, Alkwer¬ tatherium and Plaisiodon exemplify a gradual structural succession in cheek tooth and cranial morphology from the Diprotodontinae to the Zygomaturinae. Keywords: Late Miocene marsupials, Alcoota Local fauna, Diprotodontidae, zygo¬ maturinae, zygomaturine origins, Diprotodontinae, nototheriines, Pyramios, Plaisiodon. INTRODUCTION Primarily a description of a new taxon, this study also attempts to address the character states of the Diprotodontidae in relation to Alkwertatherium, which is compared with other Cheltenhamian-Mitchellian equivalent forms in considerable detail, incorporating a brief review of the mid- to late Miocene dipro- todontid genera. At least three large genera and one smaller diprotodontid genus are present in the Alcoota Local Fauna. The molar dentitions of the three largest genera: Pyramios alcootense Woodburne, Plaisiodon centralis Woodburne and Alkwertatherium webbi, gen. et sp. nov., overlap in size and morphology. Isolated molars of these genera are difficult to dis¬ criminate without the associated third premo¬ lar. It is possible, judging from the size distri¬ butions, that the tables of measurements pro¬ vided by Woodburne (1967a) may contain an admixture of these forms. Sampling of the Alcoota site between 1985 and 1989 indicates that Pyramios and Alkwertatherium are un¬ common and therefore the inclusion of the odd molar of these species would not seriously influence the statistical definition of Plaisio¬ don. The upper third premolars of three of the four Alcoota genera are readily distinguish¬ able. The P 3 of Pyramios alcootense is character¬ ized in having a weakly expressed or absent parastyle, absence of a hypocone and presence of a faint posterobuccal cingulum. Its broadly trilobate occlusal profile, large, conical proto¬ cone and equally large triangular, undivided parametacone with a short, steeply descending postparametacrista are associated with dipro¬ todontine (Diprotodontinae) diprotodontids. The large size of the P 3 relative to the molars is more in proportion with the condition in zy¬ gomaturines. In Plio- Pleistocene diproto- dontines, (eg. Euryzygoma, Nototherium and Diprotodon) the P 3 is reduced relative to the molars. Plaisiodon centralis permanent upper pre¬ molars are elongated, with a prominent, pos¬ teriorly curved parastyle and a long, blade- 53 P.F. Murray like postparametacrista. The posterobuccal cingulum is absent. The hypocone is at least incipiently present, sometimes budding off the posterior side of the protocone. In some individuals the parametacone is weakly differ¬ entiated into two conjoined cusps, but the conspicuous vertical buccal sulcus that ac¬ companies such a division in the zygomaturines Kolopsis torus Woodburne and Zygomaturus species is less distinct. Kolopsis torus has a broad, low molariform, 5-cusped P 3 with a divided parametacone, well-devel¬ oped anterobuccal and posterobuccal cingulae and a low, but distinct mesostyle. While these three genera cannot be mis¬ taken for any other diprotodontid. Alkwertath- erium webbi upper third premolars have a large parastyle combined with a Pyramios- like absence of the hypocone; steep, triangular undivided parametacone and large conical protocone. A weak posterobuccal cingulum is present. An indistinct mesostyle is confluent at its apex with the parametacone, as in Pyr- amios. The premolar is similar to that of Pyr- amios, but larger relative to the molars, and distinctive in its possession of a massive, strongly differentiated parastyle. The apparent absence of unambiguous de- scriminati ve features in the upper molar denti¬ tions and the diprotodontine-zygomaturine continuity in the premolar morphology result¬ ing from the discovery of Alkwertatherium raises some issues in diprotodontid sys- tematics. On the basis of cranial characters, Alkwertatherium is very similar to Plaisiodon, but otherwise expresses affinities with Pyr- amios. Abbreviations of museum accession num¬ bers: NTM, Northern Territory Museum; NMV, Museum of Victoria (formerly National Museum of Victoria); CPC, Commonwealth Palaeontological Collection (BMR); SGM. Spencer and Gillen Museum (NTM Alice Springs); VCMP. University of California Museum of Paleontology. SYSTEMATICS Family Diprotodontidae Gill Subfamily Zygomaturinae Stirton, Woodburne and Plane Genus Alkwertatherium gen. nov. Type species. Alkwertatherium webbi sp. nov. Diagnosis. Cranium narrower (relatively longer in proportion to width) than Pyramios, Plaisiodon, Kolopis or Neohelos; dorsal profile forming a distinct vertex, with weak frontal and low sagittal crests resem¬ bling the type Plaisiodon centralis (CPC 6784). Constricted and elongated postsquamosal lamina, similar to, but more extreme than in Neohelos or Plaisiodon ; oc¬ ciput lower and narrower than any similar¬ sized diprotodontid. Rostrum narrower than in Pyramios, Plaisiodon and Neohelos, con¬ stricted at the level of the infraorbital fora¬ men; diastemal palate similar to, though longer and more narrowly constricted imme¬ diately anterior to P 1 than in Pyramios; later¬ ally expanded premaxillary palate, as in Pyr¬ amios, but with a shallow, wide, as opposed to a narrow, deep, interincisival fossa. Nasals elevated slightly and expanded anteriorly as in Plaisiodon and Neohelos, not downturned and pointed as in Pyramios. Zygomatic arch long, as deep anteriorly as posteriorly, contrasting with the short, anteriorly narrowing zygomatic arches of Pyramios. Squamosal-jugal suture long, squamosal process extends to beneath the orbit as in Pyramios; contour of arch rela¬ tively flat as in Plaisiodon and in contrast to the strongly-bowed zy-gomatics of Pyramios; masseteric processes broad and short as in Neohelos, in contrast to long, robust processes in Pyramios; deep, well-defined nasolabialis fossa similar to Plaisiodon and Neohelos, but differing from Pyramios in which the fossa is poorly developed. Trilobate, posteriorly broad P 3 composed of a large, slightly buccally off¬ set parastyle, pyramidal, undivided parameta¬ cone and conical protocone; hypocone absent, weak posterobuccal cingulum present, differ¬ ing from Pyramios in having a well-developed parastyle as large or larger than that of Plaisio¬ don. Upper molars more similar to Pyramios than to Plaisiodon with gradual progression in molar width from M 2 " 4 ; wide interproximal contact between P 3 and M 2 , but differing from Pyramios in having a more distinct, thicker lingual cingulum, wider mouth of the lingual side of the interloph sulcus, more concave posterior surface of the hypoloph, more dis¬ tinct “midlink", large metastyle on M' and much narrower hypoloph on M 5 . A conspicu¬ ous obliquity of the lophs of M 2 ' 3 resembles the condition of later diprotodontines, but Plaisiodon among the zygomaturines, has similarly obliquity of the lophs. Dentary hori- 54 A new zygomaturine genus Fig. 1. Alkwertalherium webbi gen. et sp. nov. (holotype, SGM 888). A, lateral aspect of cranium; B, ventral aspect of cranium; C, dorsal aspect of cranium. Abbreviations: LC, lambdoid crest; OC, occipital condyle; GF, glenoid fossa; MP, masseteric process; In. incisors; BOC, basioccipital: EG, entoglenoid swelling; PG, postglenoid process; NLF, nasolabial fossa; MPP, mastoid/paroccipital process; MPS, maxillo-palatine suture; FC, frontal crest; NFS, nasofrontal suture; SC, sagittal crest; SS, squamosal sulcus; SJS, squamosal-jugal suture. 55 P.F. Murray Fig. 2. Plaisiodon centralis Woodbume, 1967 (SGM 871) for comparison with Alkwertatherium webbi. A, lateral aspect of cranium; B, ventral aspect of cranium; C, dorsal aspect of cranium. Abbreviations not included in Fig. 1: FM, foramen magnum; PP. paroccipital process; PT. pterygoid; HTS, breached hypotympanic sinus; PER, periotic; PBS, paroccipital-basioccipital suture; FLP. foramen lacerum posterior: TW, vestigial squamosal tympanic wing; IPF, interpterygoid fossa; PLF, posterolateral palatine foramina; PAL, palatine; IF, incisive toramina; SQ, squamosal; JU, jugal. 56 A new zygomaturine genus zontal ramus deep but relatively thin in sec¬ tion, inferior border straight and proportion¬ ally similar to Plaisiodon; differing from the short, deep, thick-sectioned ramus with con¬ vex “rocker” inferior border profile of Pyr- amios ; straight, strong, high diastema! crests expanded into wide, flattened surfaces anteri¬ orly and elongated diastema resembling Plaisiodon proportionally but more similar to Pyramios in its expression; dentaries unankylosed at the symphysis as in Plaisiodon and in contrast to Pyramios ; symphysis ex¬ tremely long and horizontal, extending to posterior M 2 , as opposed to Pyramios in which the symphysis is short and steeply inclined. Sublingual fossa long, narrow and horizontal, similar to Plaisiodon, much longer and more horizontal than in Pyramios; lower incisor crowns wide, thick, and spatulate, resembling those of Pyramios and the New Guinea Plio¬ cene zygomaturine, Kolopsoides. Etymology. Alkwerta (= Alcoota) means “native shield” in Eastern Arrente or Iliaura language. Alkwerta is the accepted Arrente orthography of Alcoota. The area around Al¬ coota Homestead was named after a rocky landmark resembling a shield. As this object is similar in shape to the parastyle of the upper third premolar, reference is made to the promi¬ nence of the cusp which appears to stand up in front of the tooth like a shield. The generic name Alkwertatheriiim also emphasizes the uniqueness of the Alcoota Local Fauna by designating a form with the locality name. Alkwertatherium wehbi n.sp (Figs 1-13) “Plaisiodon" webhi Murray, 1989:12 (nomen nudem). Type material. HOLOTYPE - SGM 888: nearly complete, slightly crushed and asym¬ metrically distorted cranium missing the inci¬ sors, paroccipital processes and portions of the basicranium. PARATYPES - SGM 883: associated dentaries missing third premolars and posterior portion of the ascending rami. SGM P892: right dentary fragment with with complete cheek dentition (P,-M s ). Comparative material. Plaisiodon cen¬ tralis - SGM 887: maxillary palate; SGM 881: right dentary; SGM 772: right dentary frag¬ ment; SGM 885: left maxillary fragment with complete cheek dentition; SGM 884: cast of holotype (CPC 6784); SGM 871: cranium. Pyramios alcootense - SGM 872: palate with complete right cheek dentition. Neohelos ti- rarensis - SGM 891: right and left dentaries from cranium of NTM P8695-38. Kolopsis torus - SGM 889: cranium. Supplementary specimens referred to are noted in the text and figures. Diagnosis. As for the genus. Type Locality. Alcoota, Flinders Quarry, “...Waite Formation, 4 miles [6.4 km] south¬ west of Alcoota Station, 2.1 miles [3.4 km] southwest of junction of Waite and Ongeva Creeks, Northern Territory, Australia” (Woodburne 1967b). Age. The Alcoota Local Fauna is probably Cheltenhamian Stage equivalent or slightly older (Woodburne et al. 1985). Kolopsis sp. and Zygomaturus gilli are reported from Beau¬ maris, Victoria which is considered to be Cheltenhamian Stage (Stirton et al. 1967). A Kolopsis species ( Kolopsis rotundus Plane) is described from the radiometrically dated Oti- banda Formation of New Guinea, at between 7.6-5.7 million years B.P. (Everenden et al. 1964; Stirton et al. 1967). That age has subse¬ quently been adjusted to between 3.5 and 2.5 myBP (Page and McDougall 1972; Hock and Holm 1986). Stirton et al. (1967) consider the Awe Fauna to be a relict of mid-Tertiary Australian faunas and therefore not amenable to an overall stage of evolution correlation. On the basis of the presence of Zygomaturus at Beaumaris, the Alcoota Local Fauna is provi¬ sionally considered to be older than the Beau¬ maris Local Fauna. Description and comparison. General comparison: The cranium of Alkwertatherium webhi (Fig. 1) is more similar in overall ap¬ pearance to Plaisiodon centralis (Fig. 2) and some morphs of Neohelos tirarensis than it is to either Pyramios alcootense or Kolopsis torus. The Alkwertatherium webbi cranium is smaller, more elongated and laterally com¬ pressed, more gracile and more prognathic than any known example of Plaisiodon cen¬ tralis or Pyramios alcootense (Fig. 3). The upper incisor arcade is broader and more procumbent (Figs 4-5; Tables 1-2); the ros¬ trum is narrower and more elongated. The palate is narrower and the cheektooth row is straighter, narrower and shorter (Fig 6). The gracile, elongated Alkwertatherium webbi cranium is proportionally dissimilar to the type Pyramios alcootense (CPC 6749), (Fig. 6), which has a broad, relatively short 57 P.F. Murray Table I. Measurements (mm) of A, upper incisors, B, lower incisor, including implantation angles, and C, dentary dimensions. A. UPPER INCISORS Alveolar length A. wehhi P. centralis implantation angle A. wehhi P. centralis I 3 14.5 20.6 145° 130° I 2 10.6 10.8 145° 115° I.' 15.0 16.5 175° 150° B. LOWER INCISOR Thickness Width Implantation angle (frontal, from horizontal) A. webbi 11.0 29.0 20° P. centralis 12.0 12.9 55° P. alcootense 11.0 28.0 40° C. DENTARY MEASUREMENT P. centralis A. webbi P. alcootense Length I, to P, 80.0 85.0 67.5 Length tooth row 148.5 117.0 112.2 Depth at M, 70.0 55.5 60.0 Depth MTMj 74.5 68.3 80.5 Depth at P, 53.5 60.0 65.0 Length ramus 315.0 276.0 270.0 Length symphysis 121.0 117.0 88.0 I, above inferior border 52.0 23.0 50.0 cranium, long robust masseteric processes and lacks the postorbital and postsquamosal con¬ strictions conspicuous in A. webbi and to a lesser extent P. centralis. However in certain specific features, principally those relating to the trophic complex, Pyramios alcootense shows some striking similarities to Alkwer- tatherium. These are seen in the narrowly constricted diastemal palate, strongly devel¬ oped diastemal crests, flared premaxillary palate, broad, spatulate lower incisors and length of cheek tooth rows. There are no spe¬ cific similarities to Kolopsis torus expressed in the cranium and dentaries of Alkwertath- erium. Cranium: The lateral profile of the cranium of Alkwertatherium webbi is long and rela¬ tively low, but with a definite vertex, unlike that of the holotype of Plaisiodon centralis (CPC 6784) (Fig. 1). A subsequently recov¬ ered specimen of P. centralis (SGM 871) also differs from the holotype in this respect (Fig. 2). It is possible that the holotype represents an extreme condition for the genus Plaisio¬ don. The basicranial axis of Alkwertatherium is however, less upward-flexed relative to the occlusal plane than either the type specimen or SGM 871 Plaisiodon centralis. Alkwertat¬ herium webbi' s cranium is approximately the same length and has the same state of tooth wear as the P. centralis type, which has the advantage of neutralizing the considerable allometric distortions encountered in this family in relation to size, growth stage, and sexual dimorphism (Table 2). The rostrum of Alkwertatherium webbi is narrower, more uniformly tubular and about the same length as P. centralis, CPC 6784, but differs in being constricted bilaterally at the level of the infraorbital foramen, and the na¬ sals are narrowest, about 30 mm across at this point, in relation to the arcing dorsal margin of the maxilla. Each nasal termination of A. webbi is gently rounded and widest there (46 mm) in contrast to P. centralis in which the widest point of the combined nasal bones is immedi¬ ately anterior to the jugal eminence. Pyramios nasals are narrowest distally and combine to form a pyriform, decurved, overhanging nasal process. Table 2. Measurements (mm) of the A, upper, and B, lower cheek dentitions of Alkwertatherium wehhi compared with Pyramios alcootense and Plaisiodon centralis. A. UPPER CHEEK DENTITION A Ikwertatheri um L888 R888 Pyramios L872 R872 Plaisiodon L887 R887 P 3 L 18.7 19.2 18.0 24.0 23.3 W 17.7 17.6 - 17.1 17.6 17.7 M 2 L 23.3 22.5 _ 20.0 23.5 23.5 AW 19.2 18.8 - 18.2 20.0 20.0 PW 18.1 18.2 - 18.0 20.0 21.0 M 5 L 27.1 25.2 _ 25.2 27.1 26.5 AW 21.5 - - 21.0 23.0 22.5 PW - 19.7 21.2 19.0 21.6 21.1 M 4 L 28.7 27.7 26.0 26.7 32.0 30.1 AW 22.5 24.1 23.9 22.5 25.5 25.5 PW 18.3 - 20.5 21.0 22.2 - M 3 L 27.7 _ _ 28.5 _ 32.0 AW 20.1 - - 24.9 - 27.2 PW 17.0 — “ 21.0 — 19.8 B. LOWER CHEEK DENTITION Alkwertatherium Pyramios Plaisiodon R892 L883 R883 L89I R891 R88I P,L 13.8 16.5 15.5 19.5 W 9.5 - - 11.0 10.6 13.5 m 2 l 21.5 19.9 18.3 20.6 20.3 24.0 AW 14.5 14.0 - 14.3 15.2 15.2 PW 15.0 15.1 13.7 15.0 14.4 17.0 m 3 l 26.2 23.9 22.1 - - 30.5 AW 19.0 17.3 15.5 17.0 18.5 19.0 PW - - 14.8 - “ 19.8 m 4 l 28.2 25.5 25.8 28.0 27.5 35.0 AW 20.2 18.7 - 20.5 20.0 23.0 PW 18.3 - 18.3 18.8 19.0 23.0 M, L 26.2 27.5 27.4 30.0 30.0 33.2 AW 19.5 20.5 20.3 2 1.4 21.0 24.9 PW 16.6 17.8 18.2 19.2 19.0 21.5 58 A new zygomaturine genus Fig. 3. Dimensions and comparison of the ventral aspect of Alkwertatherium webbi with Plaisiodon centralis. A, Plaisiodon centralis (SGM 871); B, Alkwertatherium webbi (SGM 888). Key to abbreviations (see Figs 1-2), plus: GIC, groove for internal carotid artery; FO, foramen ovale; CF, condylar foramen. Scaled to equivalent condyloba- sal length. In lateral view, the profile of the diastema is slightly more elongated and less curved down¬ wards along with the profile of the incisors than in P. centralis (Fig. 5). As in P. centralis, the premaxillary-maxillary suture is nearly vertical. The dorsal profile of the nasals is gently concave. Though slightly crushed, a natural-looking depression located about 30 mm behind the end of the nasals indicates that they were weakly elevated. A small premaxil¬ lary-nasal notch appears to have been present. The narial aperture is trapezoidal, the nasals lying almost horizontal to the frontal plane. Posteriorly, the nasal bones are gabled result¬ ing in a low, oval elevation situated immedi¬ ately anterior to the frontal depression. Alk¬ wertatherium and P. centralis are similar in this respect, and strong resemblances to Neohelos are also present in the dorsolateral contours of the rostrum and particularly in the basic shape of the nasal bones. Alkwertatherium differs substantially from Plaisiodon and Pyramios in the width of the base of the rostrum at the nasofrontal contact. In A. webbi, the width of this region, 50 mm, is little more than half the width of the similar¬ sized Plaisiodon , CPC 6784, and considerably less than half the width of that of Plaisiodon, SGM 871. Immediately behind the rostrum, Alkwertatherium resumes its proportional similarities with CPC 6784, in which the inter¬ orbital width is only slightly greater than that of the new genus. The frontal region of Alkwertatherium re¬ tains the narrow aspect of the base of the rostrum as it ascends into the frontal crests. The frontal depression is shallow and flat compared to the deeply clefted fossae in CPC 59 P.F. Murray 6784 and SGM 871, Plaisiodon centralis. In Pyramios alcootense the frontal crests are large, rounded swellings in marked contrast to the long narrow, crisply delineated and straight crests of Plaisiodon and Alkwertath¬ erium. As in P. centralis , the frontal crests converge in a low sagittal crest and the flat lateral wall of the braincase descends steeply towards a long, narrow squamosal sulcus. In Pyramios the frontal crests do not converge into a sagittal crest, remaining as separate temporalis crests for the length of the neuro- cranium. The basic shape of the neurocranium of Alkwertatherium differs little from that of P. centralis, except for its greater length and narrowness posterior to the squamosal root, and the extent to which it is elevated above the tooth row. The latter difference however, amounts to only about ten degrees. A larger discrepancy in cranial shape between Alkwertatherium webbi and Plaisio¬ don centralis is the nasofrontal angle, which, as measured at the nasofrontal suture, regis¬ ters a difference of about twenty degrees (Fig. 6). As in P. centralis the parietals of Alkwer¬ tatherium are long and narrow and the squamosal lamina is broad. The postsquamosal arch in Alkwertatherium is considerably longer and more deeply concave than in comparable P. centralis specimens. The zygomatic arch of Alkwertatherium webbi is deep and long. The contour of the undistorted left zygomatic arch is flatter than in P. centralis, closely resembling the condi¬ tion in Neohelos tirarensis (Fig. 6). The squamosal process is long and sustains a con¬ siderable width, extending anteriorly to lie beneath the orbit. The squamosojugal suture also contributes to the posterior orbital margin in Pyramios. Both CPC 6784 and SGM 871 have much shorter squamosal processes that terminate 30 mm or more posterior to the distal orbital margin. Consequently, in Alk¬ wertatherium, the zygomatic arch is at least as deep anteriorly, immediately posterior to the orbit, as it is immediately anterior to the gle¬ noid fossa. In P. centralis and Pyramios alcootense, the zygomatic arch is narrower anteriorly than it is posteriorly. Although the squamosal process does not extend as far anteriorly in Neohelos as in Alkwertatherium, the anterior depth of the arch is approximately the same as the posterior depth, and the squamosal process is relatively longer than in either Plaisiodon or Pyramios. The masseteric process in A. webbi is relatively delicate and short, not extending more than a few millimetres below the oc¬ clusal line. The processes are missing from the holotype of P. centralis but its broken surface indicates that a much more robust process was present. The extent to which the masseteric process was developed in P. centralis is dem¬ onstrated by SGM 871. The masseteric proc¬ esses of Pyramios are also much longer and more robust than in Alkwertatherium. The nasolabial fossa is deeper, more inferiorly directed and more sharply delineated in Alk¬ wertatherium than in Plaisiodon or Pyramios, more closely resembling that of Neohelos ti¬ rarensis. Palate: The diastemal palate of Alkwertath¬ erium webbi is markedly constricted immedi¬ ately anterior to the P J and attains a minimum width of only about 28 mm at 15 mm anterior to the tooth. Anterior to the P\ at 25 mm, the palate begins to flare laterally to attain a maximum width of 52 mm immediately poste¬ rior to the I 3 alveoli. In P. centralis, the dias¬ tema is moderately narrow, relatively more constricted than in most Neohelos tirarensis and Kolopsis torus crania in which the maxi¬ mum interincisival width is approximately the same as the interdiastemal minimum (Figs 5- 6: Tables 1-2). In conjunction with the anteri¬ orly expanded premaxillary palate, the incisor alveoli open more laterally than ventrally and the implantation angles of the incisor roots are closer to horizontal. A more graphic expres¬ sion of this complex is seen in the position of the I 3 alveolus immediately in line with I 2 in Plaisiodon centralis, Neohelos tirarensis and Kolopsis torus in contrast to the laterally off¬ set I 3 alveoli in Alkwertatherium. In the mor¬ phology of the diastemal and premaxillary palate, Alkwertatherium is a bit different from zygomaturines, more closely resembling the notothere Pyramios alcootense. However, in Pyramios, the interincisival fossa is a deep, spoon-shaped depression. In Alkwertath¬ erium, the interincisive region is a shallow, wide concave surface. Although no upper incisors have been iden¬ tified for the species, the implantation angles of the teeth indicate that the crowns were more procumbent than in P. centralis, Neohelos tirarensis or Kolopsis torus. The diastemal palate is also relatively longer than in P. cen- 60 A new zygomaturine genus Fig. 4. Comparison of the anterior palatal and dentary diastemal regions in Alkwertatherium webbi and Plaisiodon centralis. A, ventral aspect of anterior palatal region of P. centralis (SGM 871); B, ventral aspect of symphysial region of the dentary of P. centralis ; C, ventral aspect of anterior palatal region of A. webbi (SGM 888); D, ventral aspect of the symphysial region of the dentary of A. webbi. Not to scale. 61 P.F. Murray LAMBDOID CREST Fig. S. Restorations of Alkwertatherium webbi gen. et. sp. nov. Opposite, restorative sketch of lateral aspect of skull, scale 50mm; A-C, diagrammatic restorations of cranium. 62 A new zygomaturine genus tralis. The incisor implantation in Pyramios is more similar to that of Alkwertatherium, though also less procumbent. The maxillary palate of Alkwertatherium is slightly narrower than in P. centralis, made less visually apparent perhaps, by the nar¬ rower and shorter cheek tooth rows (Figs 1-3, 6). The palate is otherwise similar to that of Plaisiodon, particularly in the shape of the cheektooth arcade and moderate extent of posterior divergence. Pyramios alcootense has a relatively broader, shorter, more domed palate with more posteriorly divergent tooth rows (Fig. 6). Cranial base: The internal nares and in¬ terpterygoid fossa appear to be shallower in Alkwertatherium than in P. centralis, in which an exceptionally deep, cavernous excavation is a conspicuous feature of SGM 871. The cranial base is poorly preserved \n Alkwertath¬ erium. Sufficient surface contour is present to define the extremely wide, obliquely oriented glenoid fossae, ventrolaterally directed pterygoid fossae and low, wide, but appar¬ ently thin, delicate postglenoid processes. SGM 871, Plaisiodon centralis, preserves many basicranial features of the genus and the majority of those portions visible in Alkwer¬ tatherium appear to correspond closely to it. In Pyramios, exceptionally wide glenoid fos¬ sae are oriented at right angles to the axis of the cranium. There is, moreover a highly dis¬ tinctive lateral profile of the glenoid notch which is narrowly V-shaped, as opposed to the broadly U-shaped profile of Alkwertatherium. An interesting similarity between Pyramios and Alkwertatherium is the shared feature of thin postglenoid processes. The occipital region of Alkwertatherium is conspicuously narrower than in Plaisiodon, Pyramios and Kolopsis, but is similar in width though lower than in some Neohelos speci¬ mens (eg. NTM P8695-38). The occipital condyles, which are the most posterior struc¬ tures of the cranium, are comparatively small, transversely elongated and narrow, in contrast to the large, widely separated occipital condyles of Pyramios. The condyles are sepa¬ rated ventrally by a deep, V-shaped inter- Table 3. Cranial measurements (mm) of Alkwertatherium webhi (888) com¬ pared with Plaisodon centralis (884, 871) and Pyramios alcootense (872). Cranial measurements SGM888 872 884 871 Length-11 to occipital condyle 415.0 - - 520.0 Length-lambdoid crest to outer I 1 - - - 525.0 Length-zygomatic arch 204.0 - 195.0 243.0 Length-temporal is fossa 122.0 - 125.0 155.0 Length-diastcmal palate 73.0 - 52.0 78.0 Length-upper cheek tooth row 122.5 120.0 - 149.0 Length-interptcrygoid fossa - - - 138.0 Length masseteric process 65.2 - 104.5 Length-lambdoid crest to nasals 395.0 - - - Hcighl-basioccipital-sagittalcres 70.0 - 122.0 121.0 Depth-vertex to pterygoid 145.0 - - 175.0 Depth-posterior zygomatic arch 50.0 - 52.4 84.5 Depth-anterior zygomatic arch 50.0 - - 56.6 Distance-molars above masseteric pr <5.0 - - 35.0 Depth-maxilla above P' 57.0 - 54.0 - Depth from M 4 5 to top of frontal 110.0 - 104.0 115.0 Width-occiput 98.0 - - 130.0 Width-across mid point of orbits 95.0 — 110.0 140.0 Width-frontal above orbits 58.0 - 88.0 127.0 Width-cranium at frontal converg. 60.0 - 61.0 60.0 Width-PMX across narial aperture 69.0 - 60.0 92.0 Width-zygoma to sagittal crest 120.0 - 90.0 167.0 Width-anterior to lambdoid crest 102.0 - - 190.0 Width-premaxillac between P 41.0 - 26.0 30.0 Width-palate anterior to P' 27.8 39.0 37.7 54.5 Width-maxillae labial sides P’ 67.0 72.0 75.5 91.4 Width-palate lingual M 14 52.0 62.0 - 64.0 Width-maxillae labial sides M 5 87.0 100.0+ - 111.0 Width-interptcrygoid fossa 47.0 - - 54.0 63 P.F. Murray condylar notch. The foramen magnum is trans¬ versely wide and distinctly oval in shape. Immediately above the foramen magnum, the Table 4. Summary of characters of Alkwertatherium webbi gen. ct sp. nov. and key to comparisons shown in Figures 6, 8, II, 13. Summary of characteristics of Alkwertatherium webbi 1.0 Dorsum cranii 1.1 high steep frontal 1.2 constricted prcorbitally 1.3 constricted postsquamosally 1.4 upward basicranial flexion 1.5 high, thin narrow frontal crests 1.6 anteriorly expanded nasals 1.7 trapezoidal narial aperture 1.8 low. narrow occiput 1.9 wide, oval foramen magnum 2.0 Zygomaticofacial 2.1 long, deep squamosal process of zygomatic arch 2.2 zygomatic arch flattened in lateral aspect 2.3 zygomatic arch deep anteriorly 2.4 short, delicate masseteric processes 2.5 deep, well-defined nasolabial fossa 2.6 high, distinct sinuous facial crest 2.7 long, well defined buccinator sulcus 2.8 long squamosal sulcus 2.9 shallowly arcing diastemal profile 3.0 Palate 3.1 flat, narrow and long 3.2 nearly straight arcades, slight posterior divergence 3.3 diastemal palate markedly constricted 3.4 flaring premaxillary palate.-shallow IC fossa 3.5 P lateral to P. alveoli open anterolaterallly 4.0 Basis cranii 4.1 Narrow, shallow interpterygoid fossa 4.2 small, shallow pterygoid fossae 4.3 wide oblique glenoid fossae 4.4 deep, V-shaped intercondylar notch 5.0 Upper third (permanent) premolar 5.1 large parastylc 5.2 hypocone absent 5.3 undivided parametacone 5.4 buccal cingulum present 5.5 high clearly expressed mesostyle 5.6 w idc, deep posterior fossa, distinctively worn 6.0 Upper molars 6.1 weak parastylc, strong postparaconal crest 6.2 strong, short, rounded lingual cingulum 6.3 small protostyle M J 6.4 M 5 in line with M 4 lingually 6.5 P J widely contacts M 2 interproximally 6.6 mctalophs sharply reduced M 4 5 6.7 metastyles low on PMC, strong on M 4 6.8 labial cingulae short, thick, rising confined to MV 7.0 Dentary 7.1 body deep, tapers anteriorly 7.2 diastemal crests, long high, anteriorly expanded 7.3 near horizontal symphysis, terminates behind M 2 7.4 short, shallow digastric fossa 7.5 genial pits in ventral border of symphysis 8.0 Lower incisors 8.1 wide, thick crowns 8.2 nearly horizontal implantation 8.3 enamel on all surfaces 8.4 thick cementum base raised above enamel 9.0 Lower third prcmolar 9.1 small relative to molars 9.2 narrow, oval crowm 9.3 weak buccal cingulum present 9.4 large central cuspid, short posterior moiety 10.0 Lower molars 10.1 paralophid crest weak, protolophid face steep 10.2 protolophids M 2J high relative to hypolophids 10.3 midvalleys V-shaped 10.4 lingual offset of P 3 -M 2 10.5 weak, short precingulid 10.6 reduced size of M s proportionally small occipital surface was deeply excavated for the attachment of nuchal musculature. Upper cheek dentition: The P 3 is a short, broadly trilobate tooth composed of a large, high, transversely oval-sectioned parastyle with a slight buccal offset of the apex, a high, pyramidal, undivided parametacone and a robust, conical but slightly lower protocone (Figs 7-9; Table 4). The hypocone is absent. The apex of the parastyle is set closer to the parametacone than in Plaisiodon, but closely resembles it proportionally. The buccal groove separating the parastyle from the par¬ ametacone is deeper, longer and more cleft- like than in any other zygomaturine genus. The anterolingual basin is deeper, more con¬ fined and forms a more definitive horizontal shelf than in Neohelos or Plaisiodon. The anterolingual cingulum is short but distinct, commencing from the mid-lingual base of the parastyle to ascend the anterolingual aspect of the protocone, forming near its base, a minute protostyle. A distinct secondary basin, slightly elevated above the parastylar basin, lies im¬ mediately anterior to the base of the proto¬ cone. This preprotoconal basin is incipient in Neohelos, distinct but small in Kolopsis and variably expressed in Plaisiodon, in which its presence is associated with a reduction of the parastylar basin. A protostyle is present in some Neohelos and Kolopsis specimens, but appears to be absent in Plaisiodon. The proto¬ cone is large, though slightly smaller than the parametacone and higher, relative to the par¬ ametacone than in Neohelos, Plaisiodon and Kolopsis. In section, the protocone of Alkwertath¬ erium approximates a rounded trapezoidal shape compared to the distinctly rounded sec¬ tion of the cusp in Plaisiodon and Pyramios. A short, thick postprotocrista ascends the cusp near vertically, becoming indistinct immedi¬ ately short of the lingual portion of the postcin¬ gulum. The postcingulum is well-developed lingually, becoming indistinct interproxi¬ mally and distobuccally. The posterior sur¬ faces of the parametacone and protocone are steep and delineate the sides of a deep, oval distal basin, exaggerated to some extent per¬ haps, by wear. Although there is ample room on the posterolingual shelf for a hypocone, not so much as a thickening of the cingulum can be found to betray a phenotypic expression of the structure. 64 A new zygomaturine genus Fig. 6. Character distribution comparison of the cranium of Alkweriaiherium webbi with crania of the larger mid- to late Miocene diprotodontid genera. A, Plaisiodon centralis, (holotypc CPC 6784, restored from cast SGM 884); B, Neohelos tirarensis (NTM P8695-38); C, Alkwertatherium webbi, (SGM 888) restored; D, Pyramios alcootense (holotype cranium CPC 6749, partially restored) (after Woodburne 1967); E, dorsal aspect of snout of P. centralis (from cast SGM 884); F, A. webbi ; G, N. tirarensis; H, P. alcootense (after Woodburne 1967); I, ventral aspect of P. centralis SGM 871 (restored); J, A. webbi (restored); K, N. tirarensis; L, P. alcootense, (restored) (after Woodburne 1967). Numerical references to characters are given in Table 4. Corresponding numbers indicate a strong similarity; brackets indicate a variable condition or less distinct, though similar features. Abbreviations: FRC, frontal crest; FAC, facial crest: BF, buccinator fossa; MP, masseteric process; ICF, Interincisive fossa; DC, diastemal constriction; NLF, nasolabial fossa. 65 P.F. Murray c D 5 Fig. 7. Upper cheek dentition of Alkwertatherium webbi. (A-C) comparison of P’-M J of Plaisiodon centralis with (D) Alkwertatherium webbi. A-C. range of morphological variation in cusp morphology of Plaisiodon centralis ; C, "hypoconeless variant” which has a thickening in the lingual cingulum; in B, the parastyle is exceptionally worn. Note the narrow labiad interproximal contact of the P’ with M ! in Plaisiodon ; E, upper cheek tooth row of Alkwertatherium webbi; F, lower cheek tooth row of Alkwertatherium webbi. Abbreviations: PS, parastyle; PMC, parametacone; PRC, protocone; HYC, hypoconc; PPC, postparaconal crest; MTS, metastyle; BC, buccal cingulum; ML, midlink; PAC, paracone; MS, mesostyle; MEC, metacone; MED, metaconid; END, entoconid; ACD, anterior cingulid; PRD, protoconid; HYD, hypoconid; BCD, buccal cingulid; PCD. postcingulid; MLD, mctalophid or cristid obliqua. 66 A new zygomaturine genus Apparently “hypoconeless” variants of Neohelos usually show a thickening of the cingulum in the position of the hypocone. The posterior portion of the tooth is more similar to that of Pyramios, both structurally and thegoti- cally, except that the posterolingual portion of the postcingulum is more generous in Alkwer- tatherium. The low, interproximally indistinct postcingulum commences on the distobuccal corner of the parametacone after which it ascends the posterolabial margin of the cusp in the form of a steep, thick, slightly irregular postparametacrista, narrowing gradually as it approaches the apex of the cusp. The par¬ ametacone is steep and triangular in section with the anterior face presenting a transversely flattened, apically tapering surface. A distinct cuspule is present immediately lateral to and below the apex of the parametacone. This mesostyle (possibly homologous with stylar cusp C of primitive vombatiform molars) is continuous with the posterobuccal cingulum, which in Alkwertatherium defines a distinct, oval fossette on the distobuccal flank of the parametacone. The P 3 of SGM 872, Pyramios alcootense, is similar to Alkwertatherium in the shape of the parametacone and the degree of development of the postparametacrista. A posterobuccal cingulum is present but does not establish a fossette. The mesostyle is larger but less dis¬ tinctly defined. There is a faint postparastylar sulcus separating a small parastyle from the parametacone and defining the anterior extent of the mesostyle. None of the Pyramios third premolars show a tendency to develop more than an incipient parastyle, although the ante¬ rior crown base is more generous than in other nototheriines, being more comparable to that of zygontaturines, and a faint buccal crease, delineating the base of the parastyle, is present on some specimens. The large size of the P 3 and the incipient parastyle development gives the impression that Pyramios is structurally intermediate to nototheriines (Diprotodonti- nae) and zygomaturincs. The upper molars of Alkwertatherium, in comparison with similar-sized CPC 6748, Plaisiodon centralis, are smaller relative to the splanchnocranium, particularly M 4 ' 5 , which do not abruptly widen behind the third molar as in P. centralis. This marked transi¬ tion in the molar gradient in P. centralis is a distinctive and consistent feature of the spe¬ cies, and indeed among zygomaturines in general. There are two other consistent differ¬ ences in the morphology of the arcade: in Alkwertatherium the interproximal contact between the P 3 and M 2 is wide and centered, and the M 5 is aligned with the M 4 on the lingual side. In Plaisiodon centralis the P 3 is offset buccally leaving a large V-shaped gap be¬ tween the teeth on the lingual side, and the M 5 is offset lingually creating a step-like contour of the occlusal profile on the lingual side. The primary differences in molar morphology between Alkwertatherium and P. centralis are in the degree of development in the buccal and lingual cingulae and crests related to the pos¬ terior margins of the lophs. The second molar in Alkwertatherium is distinctive in having a wider mid-valley bu¬ ccally and in possessing a well-developed buccal cingulum, continuing anteriorly into a short postparaconal crest. The parastyle is poorly developed on all molars. The molar crown morphology is otherwise very similar in the two species and the descrimination of isolated molars is complicated by a high de¬ gree of morphological variability. The M 2 of Alkwertatherium has a weakly developed parastyle, but a distinct, short, postparastylar crest. A short postparaconal crest ascends the base of the paracone and is continuous with a thick buccal cingulum which spans the mouth of the interloph valley to the base of the hypocone, where it is transected by a faint sulcus. The cingulum bulges slightly as a presumptive mesostyle. The protoloph is slightly wider and more massive than the metaloph. A low, short, midlink is present and a second, faint crest transects the interloph valley just lingual to the longitudinal midline of the crown. The postcingulum is wide and rounded. Buccally it ascends the metacone as a distinct postmetac¬ rista, at the base of which is developed a low, indistinct metastyle. Lingually, the postcin¬ gulum terminates near the base of the hy¬ pocone and is separated by the lingual side of the metaloph from a short lingual cingulum that flares out around the mouth of the inter¬ loph sulcus. At its anterior extreme, a small protostyle is present. The anterior cingulum is wide lingually and slightly thickened on the midline. The M 3 is larger than the M 2 with a distinctly wider lingual interloph sulcus and correspond- 67 P.F. Murray 2 2 K b 0 2 W M2 LXPW^ M4 LXPW • • M5 LXAW M5 LXPW ■ • ■ ■ WIDTH MM Fig. 8. Scatter diagram of the dimensions of upper cheek teeth comparing Alkwerlalherium (triangles) with Plaisio- don (dots) and Pyramids (squares); data from Woodburne (1967a) with the addition of new specimens listed under Comparative Material of A. webbi. Note the relatively clear separation of genera with the premolar compared to the large degree of overlap with the molar dimensions. Alkwerlalherium webbi aligns with Pyramios alcootense. Scale is in millimetres. X axis = width. Y axis = length; regressions were not calculated due to to small sample size. 68 A new zygomaturine genus ingly longer lingual cingulum. The protostyle is reduced to a low bulge. The “midlink” (which does not appear to be homologous with a crista obliqua), is a teardrop-shaped enamel boss developed on the base of the paraconal side of the protoloph. The interloph valley is sinuous, deep and narrow. Both moieties are approximately equal-sized. The postmetac¬ rista is indistinct. The postcingulum is re¬ duced lingually but remains wide and low on the posterobuccal corner of the metaloph, where a low metastyle is present. The M 4 is larger than M 3 and is distin¬ guished by its narrower metaloph, wide pro¬ toloph and buccally expanded paracone, wider midvalley, less well-developed postcingulum and less distinct though broader “midlink”. The M 5 closely resembles M J in size and shape, but the metaloph is relatively narrower, the postcingulum is confined to the posterobuccal side of the crown, the midlink is reduced to a faint swelling and the buccal cingulum is absent. Although the molars are indeed distinctive, there are no individual characters of the crown morphology that can invariably separate them from Plaisiodon centralis or from Pyramios alcootense. Except for their smaller size, slightly greater apparent obliquity of the lophs of M 2 " 3 and the difference in the molar gradi¬ ent, the degree of variability in Plaisiodon crown morphology, and indeed, even the dif¬ ferences between the right and left sides of the Alkwertatherium cheek dentition, render the distinction of isolated teeth somewhat dubi¬ ous, given that the size range of Alkwertath¬ erium is unknown. Worn molars, particularly M 2 , have a distinctive thegotic character which will be discussed in conjunction with the lowers. The molar gradient, dimensions, and the low disposition of the metasyle on M 2 of Alkwertatherium more closely resembles Pyr¬ amios, but the proportional widths of protol¬ oph to metaloph for each molar, the occlusal contours, particularly M 4-5 , with a marked re¬ duction of the width of the metaloph, are more similar to Plaisiodon. Dentary: The associated dentary of Alk¬ wertatherium superficially resembles that of Palorchestes in having an extremely long, kangaroo-like, horizontally disposed sym¬ physis, containing broad, spatulate, procum¬ bent lower incisors (Figs 10-11). The molars are typical of diprotodontids and most closely resemble those of the nototheriine, Pyramios alcootense. Although the dentary was found in association with the cranium of Alkwertath¬ erium, which was isolated from other fossil concentrations, direct association of fossils in the Alcoota assemblage is no guarantee that the remains represent a single individual. Fig. 9. Comparisons of the upper cheek dentitions of larger mid- to late Miocene diprotodontid genera with Alkwerthatherium webbi. A, Neohelos tirarensis (NTM P8690) "hypoconeless variant”; B, Pyramios al¬ cootense (SGM | R]872); C, Alkwertatherium webbi (SGM [RJ888); D, Plaisiodon centralis (SGM (RJ887); E, Plaisiodon centralis (SGM 885); F, lingual occlusal profile of Neohelos tirarenesis (NTM P8690) compared with G, Pyramios alcootense (SGM (R]872): H, Alkwer¬ tatherium webbi (SGM 888); I, Plaisiodon centralis (SGM 1RJ887); .1, Plaisodon centralis (SGM 885). Abbreviations; MS, mcsostyle; PS, parastyle; PR. protocone: MT, mctastyle: PC; postcingulum; HY, hypocone; BC, buccal cingulum: ML, midlink; AC. anterior cingulum; LC, lingual cingulum; ME, metacone; PA, paracone. Numerical references to the characters are listed in Table 4. The lingual profiles of the molars show the step-like offset of M' in Plaisiodon, in comparison to the other genera. There is a tendency towards this condition in some Neohelos and some Pyr¬ amios (F, G); but the trait is apparently not present in Alkwertatherium webbi. 69 P.F. Murray However, there are occasionally isolated pockets in which partially articulated material is recovered. Alkwertatherium occurred in one of these situations. The dentary lay slightly above and at the posterior end of the cranium with the dentitions of both facing upwards. As the excavation around the dentary progressed, the cranial base of SGM P888 was partially exposed immediately beneath the posterior end of the inferior border of the dentary, the anterior ends of the jaws and cranium pointing away from one another, as though the dentary had been dislodged backward from the skull. To the extent that the somewhat distorted lower tooth rows can be physically occluded with the uppers, they match closely in terms of occlusal surface angles of the similarly worn lophs and lophids, the distance between them, and in terms of the individual variations of each tooth in the degree of wear, pattern of attrition and thegotic correspondence. The dimensions of the individual molar teeth are also within the expected tolerances for the same individual which ranges normally between 0.85 and 1.0. The length of the right uppers from front to back are: 23.3,27.1.28.7, Fig. 10. Dentary of Alkwertatherium wehbi gen. ct sp. nov. (SGM 883, paratype). A, occlusal aspect; B, lateral aspect. Note the extreme procumbency of the lower incisors, anteriorly tapering horizontal ramus, high long diastemal crests and their bony expansions, long deep, sublingual fossa and short, weak transverse torus penetrated by genial pits. The symphysis, which is unfused, extends to the posterior moiety of M,. Abbreviations: SLF, sublingual fossa; DC, diastemal crest; AF, anteriorly flattened surfaces of diastemal crests; GP, genial pit; TT, transverse torus: PS, postalveolar shelf; PA, postalveolar process; MF, mandibular foramen; PME, posterior masseteric eminence; FM, mental foramen. 70 A new zygomaturine genus Fig. 11. Comparison of the dentaries of larger mid- io late Miocene diprotodontids with Alkwertatherium webbi. A, Plaisiodon centralis (SGM 881); B, medial aspect (left) and occlusal aspect (right) of Neohelos tirarensis (CPC 22530); C, Alkwertatherium webbi (paratype SGM 883); 1), Pyramios alcootense, (SGM 891). Numerical character definitions are given in Table 4. Abbreviations: DC, diastemal crest; DF, digastric fossa; DS, dentary symphysis; GP, genial pit. 71 P.F. Murray 27.7. The right lowers are 18.3, 22.1, 25.8, 27.4; resulting in a proportional congruence of 0.87. The distance between the lophs and lophids for each occluding pair are as closely matched as can be determined by simple measurements. The congruence of the lower incisors with the upper incisor sockets appears to be close. Presumably the low angulation and buccal offset of the empty upper incisor sockets is disposed to the accomodation of the broad, procumbent lower incisors. The dentary of Alkwertatherium differs from that of the nototheriine Pyramios al- cootense in having a nearly straight inferior border of the horizontal ramus, an anteriorly tapering and more slender as opposed to a more massive, anteriorly deep, posteriorly tapering body; much shorter and vertically narrower, more horizontal diastemal portion, and a much longer and more horizontally dis¬ posed symphysis that extends to behind M, as opposed to behind the P r In Pyramios al- cootense, the dorsal contour of the symphysis (sublingual sulcus) is nearly vertical in SGM P891. Ventrally, in Pyramios , a distinct men¬ tal eminence is developed immediately below the incisor alveoli. This is also evident in the paratype UCMP 69784. The dentary is similar to, though much smaller than any mature specimen of Plaisiodon centralis, differing principally from that genus in the more procumbent incisor alveolus and relatively longer diastemal portion of the jaw. The sym- physial outline is more similar to Plaisiodon than it is to Pyramios. The horizontal ramus is relatively deeper in proportion to length than in Plaisiodon, but it is also proportionally less thick in section below the molar alveoli. The diastemal crests of Alkwertatherium are much thicker, higher and less sinuous than in P. centralis. Anteri¬ orly, the diastemal crests become laterally expanded, flat surfaces paralleling the sub- horizontally implanted incisor crown bases. The diastemal crests of Pyramios are similar, though more arched dorsally, much shorter and less expanded anteriorly than in Alkwer¬ tatherium. Internally, the digastric sulcus is slightly deeper than in either Pyramios or Plaisiodon and extends further anteriorly than in the latter but not as far anteriorly as in the former genus. The position of the genial pits is also intermediate between Pyramios, in which they are on the posterior surface of the sym¬ physis and Plaisiodon, in which they are situ¬ ated on the ventral surface. In Alkwertath¬ erium, the genial pits penetrate the ventral surface of the symphysis, but extend onto the posterior surface as well. Lower dentition: The lower incisors are extremely wide, thick, spatuiate horizontally implanted teeth that repose with the lateral margin of the crown at about ten degrees from horizontal (Figs 5, 10-11; Table 1). Plaisio¬ don centralis incisors are only moderately spatuiate, with the crowns oriented about 55 degrees from horizontal in frontal aspect, and lying at about 35 degrees from horizontal in lateral aspect, being considerably less procumbent than in Alkwertatherium. The lower incisor implantation of Pyramios is nearly horizontal, as in Alkwertatherium, but the symphysial angle is substantially more steeply inclined and extends far deeper below the alveoli. The frontal implantation angle of the incisor crowns in Pyramios is acute. The internal angle of the paired incisors of Pyr¬ amios is V-shaped at between 80 and 85 de¬ grees. In Alkwertatherium the internal angle (dorsal surfaces) of the paired incisors pres¬ ents an obtuse surface of approximately 120 degrees. The enamel extends entirely around the crown of the lower incisors of Alkwertath¬ erium, as in Pyramios, in contrast to its being confined to a narrow strips on either side of the dorsal surface of the crown as in the zygomaturine Plaisiodon centralis. The se¬ verely worn tips of the crowns of Alkwertath¬ erium lower incisors precludes a description of their original shape, but there can be little doubt that they were very similar to those of Pyramios alcootense. The transverse width of the crown is at least 28 mm and about 10 mm thick medially, near the base of the crown. The crowns are thicker medially, tapering towards the external margins. The cementum is raised above the enamel on the dorsal surfaces form¬ ing a continuation with the broad, bony dias¬ temal platform. The P, of the paratype dentary SGM P883 was broken off, then placed in a small box which was subsequently mislaid during trans¬ portation of the specimens from the field. It was exceptionally small relative to the molars. Fortunately it is well preserved on SGM P892, a right dentary fragment of Alkwertatherium (Figs 12-13). The P 3 has a small, oval crown 72 A new zygomaturine genus P3 LXW M2 LXAW • • 8 . T 2 2 X 0 z H M3 LXAW ■ ■I • • • • 6 • ▲ M3 LXPW (m M4 LXAW ■I ▲ T A t— i—i— i —i—i—i—r M4 LXPW .V ■ A i- 1 - 1 - 1 - 1 -r • • M5 LXAW • ■ «■ t— i—i—i—i—i—i—r M5 LXPW -i—i—r WTCDTM 1VI1VI Fig. 12. Scatter diagram of the dimensions of lower cheek teeth of Alkwertatherium webbi (triangles), Ptaisiodon centralis (dots) and Pyramios alcootense (squares). As with the upper dentitions, Alkwertatherium is more closely aligned with Pyramios alcootense. Data from Woodburne (1967a) and new specimens given in materials list. Regressions were not calculated due to small sample of Alkwertatherium. 73 P.F. Murray composed of a large central cuspid and a median cristid which extends back to the inter- proximal contact with M, and is conjoined by a weak buccal and a more distinct lingual cingulum. The buccal cingulum trends obliquely towards the anterior edge of the posterior root, then faintly ascends the flank of the central cuspid to define a shallow, U- shaped fossa. The posterolingual cingulum surrounds a kidney-shaped basin. Structurally the P 3 crown is more similar to that of a zygomaturine (e.g. Plaisiodon centralis) and differs from that of Pyramios in being smaller, relatively narrower, shorter posterior to the central cuspid, in possessing a buccal cin¬ gulum and in lacking a clearly defined trans¬ verse crest from the central cuspid. The lower molars are similar to those of Pyramios alcootense (Fig. 13). Alkwertath¬ erium has a distinct but low paralophid crest and a steep, high protolophid on the M,. This high protolophid relates thegotically to the correspondingly low protoloph of the upper and the deeply worn posterior fossa of the P, and anterior cingulum of M r With the excep¬ tion of M,, the protolophids are wider than the hypolophids. The M, is relatively small, and the gradient is that of a gradual front to back increase in size. The lophids of M I3 are more obliquely oriented than the other molars. The protolophid of M 3 is also much higher than the hypolophid. The interlophid valleys of the first two molars are more sinuous and steeply V-shaped than in the last two. A short, low precingulum is present on the lingual side of the anterior base of the protolophid. Its ex¬ pression diminishes gradually from front to back. A low cristid obliqua is present on all molars, extending from the hypoconid down the anterior face of the hypolophid to the base of the posterior face of the protolophid in the approximate longitudinal midline of the tooth. Weak cingulae occur in the buccal and lingual mouths of the interloph sulcus. The postcin¬ gulum is thick, relatively deep and rounded. It is elevated slightly near the longitudinal midline of the tooth, commencing on the labial side proximally and swinging gradually to¬ ward the lingual side distally. Etymology. The species designation “ webbi ” is made in gratitude to Thomas and Wendy Webb of Alcoota Station, Northern Territory, for their valued support in palaeon¬ tological research at Alcoota. Fig. 13. Comparison of the lower cheek dentitions of the larger mid- to late Miocene diprotodontid genera with Alkwertatherium webbi. A, Kolopsis torus (SGM 893); B, Alkwertatherium webbi (paratype SGM 892); C, Pyr¬ amios alcootense (SGM 891); D. Plaisiodon centralis (SGM 881); E, labial aspect of dentition of Kolopsis torus (SGM 893); F, Alkwertatherium webbi (SGM 892); G, Pyramios alcootense (SGM 891); H, Plaisiodon cen¬ tralis (SGM 881). Numerical references to character definitions are given in Table 4. Abbreviations: END, entoconid; MED. metaconid; HYD, hypoconid; PRD, protoconid; CO. cristid obliqua; PLC, paralophid crest; BC; posterobuccal cingulid; AC, anterobuccal cingulid; MV, midvalley (U-shaped in Kolopsis torus, as consid¬ ered typical of zygomaturines); HLD. hypolophid; PLD, protolophid. Analysis of characters. The morphological description of Alkwertatherium webbi broadly samples comparable morphological features, of which only a few are .specifically useful in diprotodontid systematics (Tables 3,5; Figs 6, 8, 11, 13). Many important features (suture configurations, upper incisor morphology etc.) could not be considered due to the poor 74 A new zygomaturine genus state of preservation of the specimen. Though incomplete, this constellation of characters is adequate to formulate an hypothesis of the approximate systematic position of Alkwer- tatherium. Based primarily upon the work of Stirton et al. (1967) and subsequently Archer and Bartholomai (1978), Aplin and Archer (1987), the Diprotodontidae can be divided into two subfamilies, the Zygomaturinae and the No- totheriinae or Diprotodontinae. Recent revi¬ sions have sunk the Nototheriinae. Consensus purports that the two groups of diprotodon- tids, the Zygomaturinae, with a well-devel¬ oped parastyle on the upper third premolar and the Diprotodontinae which either lack or have small parastyles, are monophyletic. It is as¬ sumed that the large zygomaturine parastyle is a synapomorphic character. However, in the case of the Diprotodontinae, the absence or small size of the parastyle is a symplesiomor- phic state. Diprotodontines are also descriminated from zygomaturines on the basis of tighter, more V-shaped interlophid valleys; “cleaner”, less complex internal loph faces; high, steep protolophid of the M,; suppression of the paralophid crest; weak anterior cin- gulid; strong postparaconal crest on the upper molars; reduced size of the upper and lower premolars relative to the size of the molars, and a greater obliquity of the molar lophs and lophids. In derived forms ( Diprotodon ), the lophs (-ids) of the molars are hypsodont. This character complex may succeed in distinguishing the later Tertiary Diprotodonti¬ nae from the Zygomaturinae, but Pyramios alcootense variably expresses zygomaturine- like midlinks, complex interloph shapes and cingulum configurations, as well as possess¬ ing an upper third premolar that is consistently as large relative to M\ as in for example, the zygomaturine, Neohelos. Pyramios alcootense aligns with diproto¬ dontines in having a steep, high protolophid on the M„ a suppressed, practically nonexis¬ tent, paralophid crest; in the molar gradient of the uppers, in which there is a simple front-to- back enlargement and in which the metaloph is not as reduced relative to the protoloph as in zygomaturines, and of course, the upper third premolar, which lacks a well-developed par¬ astyle. In cranial characters, Pyramios resembles Euryzygoma in its diastemal con¬ striction, deep, broad cranium and short, pos¬ teriorly divergent molar arcades. However, the I 3 ’s in Euryzygoma and Diprotodon are in line with the I 2 ’s and the zygomatic processes are flat-sided. The lower incisor morphology, relatively wide though V- shaped interlophid valleys, weak cristid obliquae and comparatively low Table 5. Distribution of characters in representatives of four diprotodontoid families. 1, wynyardiid ( Muramura williamsiV, 2,"palorchcstid" (Ngapakaldta) and 3, palorchcstinc ( Propalorchestes ) examples represent the outgroup; 4, diprotodontinc (Pyramiosalcootense); 5, zygomaturine (Alkwer- tatherium webbi ); 6 , zygomaturine ( Neohelos tirarensis); 7, zygomaturine (Plaisiodon centralis ); 8, zygomaturine (Kolopsis torus). Symbols: o=abscnt; +=present; ±=more or less present; -=not observed; ?=unsure of homology; AP=apomorph, PL=plesiomorphy. 1 2 3 4 5 6 7 8 CHARACTER DISTRIBUTION AND STATE 1.1 o o o o 4 o 4 0 long, high steep frontal (AP) 1.2 0 0 0 0 4 o o 0 constricted preorbitally (AP) 1.3 4 o o o 4 4 + o postsquamosal constriction (PL) 1.4 o ± 0 0 + o 4 o upward BC flexion (AP) 1.5 4 4 0 0 + 0 4 0 thin frontal crests (PL) 1.6 o 0 0 o 4 4 4 4- nasals expand anteriorly (AP) 1.7 0 o 0 0 4 4 + o trapezoid narial aperture (AP) 1.8 + o 0 0 + + 0 0 low, narrow occiput (PL) 1.9 0 4 + + 4 + 4 4 wide, oval foramen magnum (PL) 2.1 0 0 4 4 4 0 o 0 long squamosal process (AP) 2.2 o 4 4 0 4 4 4 4- flattend zygomatic arch (PL) 2.3 o 0 0 o 4 0 o 0 zygomatic deeper anteriorly (AP) 2.4 4 + 0 o 4 4 o 4- short masseteric processes (PL) 2.5 4 4 0 0 4 4 4 4- deep nasolabial fossa (PL) 2.6 o 0 o 4 + o 0 0 distinct facial crest (AP) 2.7 o 0 o 4 + o 0 o distinct buccinator fossa (AP) 2.8 + 4 4 0 4 + 4 0 long squamosal sulcus (PL) 2.9 0 4 7 + 4 0 o o shallow diastemal profile (AP) 3.1 0 0 4 0 + 0 + 0 long, narrow palate (AP) 3.2 + 4 + 0 ± 4 + 4- arcades parallel, straight (PL) 3.3 o o 4 + 4- o 4- o constricted diastema (AP) 3.4 0 o 4 + + 0 0 0 flared premaxillary palate (AP) 3.5 0 o + + + o o o 1 J lateral to V (AP) 4.1 + 0 o o + 0 0 0 narrow, shallow IP fossa (PL) 4.2 4 4 + 4 + 4- 4 small pterygoid fossae (PL) 4.3 0 0 0 0 + 4 4- 4 oblique glenoid fossae (AP) 4.4 0 o + 0 4 0 o 0 deep, V-shaped condylar notch AP) 5.1 7 0 o 0 4 4 4- 4 large parastyle on P' (AP) 5.2 + + + 4 + 0 o o hypoconc absent (PL) 5.3 + 4 4 4 4 + 4- 0 undivided parametacone (PL) 5.4 o o 0 + + 4 o 4 buccal cingulum present (PL) 5.5 ? ± 4 + 4 o o o high, large mesostylc (PL) 5.6 0 0 4 4 + o 0 0 functional posterior basin (AP) 6.1 4 4 4 + + 0 ± o weak parastyle, strong PPC (PL) 6.2 o 0 0 ± 4 0 ± 0 thick, short, rounded LC (AP) 6.3 0 o 0 0 4 0 ± o protostyle on M : (AP) 6.4 4 4 4 ± + o o 0 M 4 -M 5 in line Iingually (PL) 6.5 + + 4 4 + 0 0 0 P 1 widely contacts M ! (PL) 6.6 0 0 o 0 ± + 4 - 4 Metaloph reduction M 4 5 (AP) 6.7 o 0 0 o + o o o large metastyle low on M 4 (AP) 6.8 o 0 o + 4 0 4 - 0 short, thick buccal cingulum (AP) 7.1 + 4 + 0 + + 4 - 4 deep ramus tapers anteriorly (PL) 7.2 o o 4 + + 0 4- o strong diastemal crests (AP) 7.3 0 0 + 0 + 0 ± 0 long, horizontal symphysis (PL) 7.4 0 0 4 + + o 0 o short, weak digastric fossa (AP) 7.5 + + 4 4 ± + 4- 4 genial pits ventral surface (PL) 8.1 0 o + + + o o 0 spatulatc I, crown (AP) 8.2 o o 4 4 + 0 0 0 horizontally implanted 1, (AP) 8.3 o o 4 + 4 o 0 o enamel all sufaccs 1 ( crown (AP) 8.4 0 0 0 + 4 o 0 o ccmentum raised above enamel (AP) 9.1 0 4 0 0 + o o o P, small relative to molars (PL) 9.2 0 + 0 0 4 o o o narrow, oval crown of P, (PL) 9.3 0 0 0 0 4 + + 4 buccal cingulum present on P, (AP) 9.4 0 o 4 0 + 4 o 4 short posterior moiety P, (PL) 10.1 0 0 4 + 4 0 0 o steep protolophid, weak crest (AP) 10.2 0 o + + + 0 o 0 protolophids higher (AP) 10.3 + 4 + + + 0 0 0 V-shaped mid valleys (PL) 10.4 0 0 0 4 + 0 0 0 lingual offset P, (AP) 10.5 0 0 + + + 0 o o weak, short precingulid (AP) 10.6 0 0 0 0 ± + 4- 4 reduced M ? (AP) 75 P.F. Murray lophs and lophids of Pyramios are not sugges¬ tive of a close relationship to Euryzygoma or any other later Tertiary notothere (Woodbume 1967a, 1967b). Having noted these discrepencies in combi¬ nation with the geologically late occurrence of Pyramios , Woodburne (1967a, 1967b) characterized Pyramios as a primitive nototh¬ ere residing near the base of subsequent dipro- todontine radiations. Although Woodburne (1967a) does not make the point explicit, the odontological similarity of Pyramios with certain Alcoota zygomaturines is difficult to ignore, and some of the ambiguities in differentiating Pyramios from the zygomatur¬ ines are implicit in his conclusions. Among the zygomaturines, some Plaisio- don centralis specimens have strong postpara- conal crests, as in Alkwertatherium. The molar parastyles of both Alkwertatherium and Plaisiodon are relatively small; and the inter- loph valleys are proportionally similar in these genera. The thick, short labial cingulae that arise along the anterior ends of the midvalleys are also present in some Plaisiodon specimens and are especially prominent in Alkwertath¬ erium. Assuming monophyly for the Diproto- dontidae, Pyramios and Alkwertatherium could represent remote descendants of the two respective subfamilies (Fig. 14). An alterna¬ tive hypothesis is that Pyramios, Alkwerta¬ therium , Plaisiodon and perhaps Kolopsoides represent an intermediate clade of diprotodon- tids into which a degree of parallel evolution (i.e. independent development of the parastyle and hypocone) must be introduced. The reality is, given the degree of morphological variabil¬ ity (ie. absence of hypocone in some Neohelos specimens) and difficulty in establishing clear complementary distributions in character complexes in these genera (correlation of the presence or absence of a feature with another), combined with the extraordinarily long period of time since their inferred divergence from a parent clade or their respective parent clades, that few “pristine” character states of a pre¬ cisely comparable nature are present for com¬ parison. The cladogram (Fig. 14) is based primarily upon dental morphology, particularly the upper third premolar. The family Diprototo- dontidae is distinguished from the Palorches- tidae on the basis of the synapomorphic loss, incorporation or suppression of stylar cusps C and D into the lophs (Fig. 14 [ 1 ]). The absence of the parastyle on the P 3 is plesiomorphic. A symplesiomorphic diprotodontid character is the presence of a large epitympanic fenestra, which decreases in size in the derived diproto- dontids, but remains more highly expressed in the zygomaturines than in the nototheres. In Pyramios the epitympanic fenestra is small and simple (Woodburne 1967b) whereas in Neohelos, the fenestra is usually large, though sometimes subdivided and narrowed trans¬ versely. The ventral squamosal tympanic wing is virtually absent in plesiomorphic diproto- dontids so that the tympanic cavity is a mere hollowing of the posterior wall of a bulla-like hypotympanic swelling that also forms the entoglenoid eminence. In Pyramios, a secon¬ dary ventral tympanic process is derived from the mastoid (Woodburne 1967b). The polarity of this morphological complex appears to be a progressive reduction of the alisphenoid tym¬ panic process, present in Propalorchestes (Murray 1986) to a short, entirely squamosal tympanic process (e.g. Vombatidae). The diprotodontid conditon is accompanied by a hypertrophy of a hypotympanic sinus immedi¬ ately anterior to the tympanic cavity and the tympanic process is further reduced. In Pyr¬ amios, the reduced anterior processes are re¬ placed by a mastoid tympanic wing. In Dipro- todon the tympanic cavity is partially floored by fusion and hypertrophy of other surround¬ ing elements. In all known zygomaturines, except perhaps, Zygomaturus, in which secon¬ dary bony fusions are evident, the tympanic cavity is open ventrally. The derived diprotodontids show a more distinct front to back molar gradient than in palorchestids and primitive diprotodontids (Fig. 14 [2]), (Rich etal. 1978). Plesiomorphi- cally a strong paracristid (paralophid crest of Stirton et al. 1967) is present. The synapomorphic basis for clade stem [3] is inferred to be the reduction of the paralo¬ phid crest on M 2 . There is a plesiomorphic retention of the small parastyle, high mes- ostylar cusp and simple, undivided parameta- cone on the P 3 . Pyramios alcootense retained or developed a P 3 of large size relative to the molars [4] with a small, though in some individuals (SGM P872) an incipiently zygomaturine-like, par¬ astyle. Symplesiomorphically the upper mo- 76 A new zygomaturine genus Fig. 14. Hypothesis of phylogenetic branching in Miocene diprotodontids based primarily upon a cladistic assessment of odontological characters (Table 5), depicted with minimum circuitry. The approximate Lyellian equivalency of the included genera is related to pan-continental stages (Woodhurne et at. (1985). The discussion contains a more detailed interpretation. [1] loss, incorporation or suppression of stylar cusps C and D in relation to the Iophs; [2] distinct front to back molar gradient; [3] reduction of paralophid crest; (4) P' large relative to molars; [5] reduction of size of permanent premolars relative to molars; |6] large parastyle, separated from the parametacone by a deep cleft; [7] development of a hypocone on the P 5 ; |8) "hook-like” parastyle on P' (Hand et a /., in preparation); [9] basic proportional similarity of the P' in Nimbadnn and Neolielos (narrow posterior fovea of P\ etc); [10] mesoslyle retracted towards cingulum; [11] parametacone in kolopsis divided into two distinct cusps. lars develop midlinks to the extent of Plaisio¬ don and Alkwertatherium. The paralophid crest is weak or absent and the protolophid of M 2 is high and steep. Character complex [5 ] sets the other nototh- eriines apart from Pyramios : synapomorphic reduction of the size of the permanent premo¬ lars relative to the molars and tendency to¬ wards higher, steeper lophs and lophids. The development of a large parastyle on the P 3 [6] unites all zygomaturine diprotodontids. Plesiomorphically it lacks a hypocone, is broad posteriorly and retains a high stylar cusp on the labial side of the apex of the parameta¬ cone. A marked narrowness of the metaloph relative to the protoloph on M 4 and especially M 5 is also characteristic. Subsequent zygomaturines are united by the development of a hypocone on the P 3 [7J. Plesiomorphically the genus Kolopsoides re¬ tains a poorly differentiated parastyle, though interpreted by Plane (1967) to be a derived feature. Autapomorphically, the hypocone is as large or even larger than the protocone in Kolopsoides. Kolopsoides and Plaisiodon appear to be united by a lengthening of the parametacone in which however, the paracone and metacone remain essentially undifferenti¬ ated, the mesostylar cusp remains high and poorly defined labially, and a posterobuccal cingulum is absent in these two genera. Phen- etic similarity in the long, nearly horizontal dentary symphysis of Plaisiodon, Kolopsoi¬ des and Alkwertatherium combined with the large, deep, almost fully enamelled lower incisors in (young only) Kolopsoides (Plane 1967) suggest a minor clade grouping, but evidence is thin. 77 P.F. Murray Nimbadon and Plaisiodon are tenuously united by a “hook-like” shape of the par- astyle of P 3 (Hand et al. personal communica¬ tion) [8]. The shape and proportions of the cusps the P 3 are also similar in Neohelos and Nimbadon [9]. However, Neohelos usually has a prominent buccal cingulum with a small but fully differentiated mesostyle located well below the apex of the parametacone [10]. In Kolopsis the parametacone is divided into two distinct cusps [11] and an anterobuccal in addition to a posterobuccal cingulum, some¬ times with a defined mesostyle, is usually present. DISCUSSION The similarities between Alkwertatherium and Pyramios on the one hand and Alkwertath¬ erium and Plaisiodon on the other, are not easy to reconcile because of discontinuities in some of the character complexes and the con¬ tinuously varying nature of others. My initial impression of Alkwertatherium was that jt might belong to the genus Plaisiodon, (Mur¬ ray 1989), impetuously resulting in a nomen nudem status for the name “ Plaisiodon" webbi. Although the type Pyramios descrip¬ tion seemed more remote relative to Alkwer¬ tatherium than that of Plaisiodon , there are also both general and specific similarities to Pyramios. In order to include Alkwertatherium within one or the other genus, either generic diagnosis would have to be revised to accomo¬ date the specimen. If the differences among Plaisiodon, Pyramios and Alkwertatherium were to be sunk to the specific level, the subfamilial diagnoses as well as the generic diagnoses would require revision due to the higher level systematic emphasis placed upon the presence of a large, differentiated par- astyle on the P 3 of Alkwertatherium and all other zygomaturines. The least disruptive option is to differentiate the new form at the generic level. Alkwertatherium shows divisions of certain character complexes that are perceived as interrelated characters in other Alcoota gen¬ era. For example, the protolophid of M, is steep in Alkwertatherium but the paralophid crest is more differentiated than in Pyramios, suggesting that the two characters are geneti¬ cally independent. The long, horizontal sym- physial profile in Alkwertatherium is more similar to that of Plaisiodon than to the nearly vertical profile of Pyramios, yet the procum¬ bent incisor implantation and morphology of the incisors suggest a closer relationship between Pyramios and Alkwertatherium. Conversely, the deep, spatulate lower incisor crowns of the zygomaturine Kolopsoides are more similar to those of Alkwertatherium and Pyramios than to those of other zygomatur¬ ines. The constricted diastemal palate in Alkwer¬ tatherium is very similar to the condition in Pyramios and other nototheres. It is one of a few consistent apomorphic features that could unite the diprotodontinae (nototheres) and, were it not for the prominent parasty le-bearing Alkwertatherium, a monophyletic grouping might be supported with this character. Plaisiodon centralis also has a more extreme constriction of the diastemal palate than in other zygomaturines, except perhaps for Zy- gomaturus. The condition varies continuously from Pyramios to Alkwertatherium to Plaisio¬ don (Fig. 6). A number of other features (Table 5) align the cranium of Alkwertatherium to the zygomaturines, especially to that of Plaisio¬ don. The zygomaturines Nimbadon, Neohelos and Kolopsis appear to form a distinct sub- clade, largely through a retention of ple- siomorphic zygomaturine features. I surmise that they separated from the Alkwertatherium- Plaisiodon-Kolopsoides subclade at the point of the expression of the hypocone on the P r In order to express this phenomenon on a cla- dogram, I would be compelled to postulate a series of dichotomies for which there is no concrete evidence. A similar series of dichoto¬ mies would be entailed to resolve the relation¬ ship between Pyramios and Alkwertatherium. 1 have not expressed these palaeontologically unsupported branchings on the cladogram. It has not been directly demonstrated in the fossil record that the zygomaturines arose from a parastyle-less ancestor. The condition is inferred from the plesiomorphic alignment of palorchestids with the diprotodontids. The earliest recognizeable diprotodontids (e.g. Ngapakaldia) lack the prominent parastyle of zygomaturines (Murray in press; Murray and Wells, in press). The preferred hypothesis is that zygomatur¬ ines arose from a notothere-like form in which the parastyle was reduced, as in primitive palorchestids. It is therefore more compelling to view Pyramios as a structural intermediate 78 A new zygomaturine genus between the diprotodontines and zygomatur- ines, than to place the genus in a position ancestral to later diprotodontines. The mor¬ phology of Alkwertatherium is resolved as being largely symplesiomorphic with Pyr¬ amios, but in possession of the zygomaturine apomorphy of a large, differentiated parastyle. As yet, there are no recognizable annectent forms for either Alkwertatherium or Pyramios in any of the pre-Cheltenhamian local faunas. The Bullock Creek Neohelos tirarensis ap¬ pears to be closely related to Kolopis torus (Stirton et al. 1967). Nimbadon spp could represent basal zygomaturines with some specific affinity to Plaisiodon and Neohelos (Hand et al., in preparation) but there is no firm evidence to support this observation. It is not likely however, that the apomorphic char¬ acter complexes in Alkwertatherium and Pyr¬ amios evolved entirely within the intervening period between the mid- and late Miocene. It is reasonable to assume that the ancestors of both of these distinctive genera will eventually be recognized in one of the early to mid-Miocene faunas. The presence of Alkwertatherium, Plaisiodon, Pyramios and other unique non- diprotodontid forms such as the primitive macropodid Hadronomas puckridgi, sets the Alcoota Local Fauna apart from the mid- Miocene and early to mid-Pliocene faunas. It appears to represent an assemblage of ple- siomorphic relicts, the ancestors of some of which must antedate the Bullock Creek Local Fauna and which must have evolved in rela¬ tive isolation. Of the diprotodontids, only Kol- opsis is suggestive of later annectents ( Zygo- maturus). I suspect therefore, that a signifi¬ cant portion of the Alcoota Local Fauna may represent a relict of early to mid-Miocene radiations analogous to the New Guinea Plio¬ cene local faunas. Although there is no obvi¬ ous geographic feature that would explain such an isolation, it may be that the isolation of drainage patterns and associated habitats oc¬ curred earliest in the centre, severely limiting population movements by mid-Miocene times. CONCLUSIONS Alkwertherium webbi gen. et sp. nov. estab¬ lishes a close relationship between diproto¬ dontines (= nototheres) and zygomaturines, if it is accepted that Pyramios is a diprotodontine rather than a primitive zygomaturine. The Alcoota form, Plaisiodon centralis, appears to be more closely related to Alkwertatherium than to any other known zygomaturine. It is postulated that there were two minor zygoma¬ turine lineages, one represented by Nimbadon, Neohelos and Kolopsis, and the other repre¬ sented by Alkwertatherium, Plaisiodon and Kolopsoides. The two minor lineages could be united through the ancestor of Nimbadon. This assumes, of course, that I have not described one of the extreme ends of two highly variable generic dines (Plaisiodon and Pyramios) as a new genus. I justify the generic distinction of Alkwertatherium on the basis of its possession of a character that is also em¬ ployed to distinguish the two subfamilies of the Diprotodontidae. The systematic emphasis of this character depends upon the assumption that Alkwertatherium is more derived than Pyramios and more primitive than Plaisiodon. Should it eventually be shown that Alkertath- erium is an extreme variant of either genus, this account would be no less informative, because it will have drawn attention to a de¬ gree of morphological variability sufficient to seriously question current diprotodontid sys- tematics. Meanwhile, I submit that Alkwertatherium accounts for some of the anomalies in the cheek tooth measurements of Pyramios and Plaisiodon provided by Woodburne (1967a). Its morphology, though derived, provides a glimpse of the antecedents of the two subfami¬ lies of diprotodontids and gives some indica¬ tion of the polarity of certain characters as well as documenting structurally, the stages transitional between diprotodontines and zy¬ gomaturines. SUMMARY Alkwertatherium webbi exhibits a unique combination of diprotodontine (= nototh- eriine) and zygomaturine characters in its cranium, dentition and dentary morphology. Although the upper third premolar had devel¬ oped a large parastyle as in the zygomaturine diprotodonts, it retains a weak posterobuccal cingulum, lacks a hypocone and otherwise resembles the P 3 of the notothere, Pyramios alcootense Woodburne. The cranium of Alk¬ wertatherium is basically similar to that of the 79 P.F. Murray zygomaturine Plaisiodon centralis Woo- dburne, differing principally from that genus in having an extremely constricted diastemal palate, expanded interincisival fossa and inci¬ sor angulation more characteristic of nototh- eriines. The dentary resembles some nototh- eriines and especially palorchestines in hav¬ ing a long, narrow sublingual fossa and spatu- late incisors, but is more similar to the zygomaturine Plaisiodon in overall shape and proportions. Alkwertatherium can be ex¬ plained as a relict form which evolved from an ancestral clade structurally intermediate be¬ tween zygomaturines and nototheriines. It is concluded that Alkwertatherium webhi is a late surviving representative of a plesiomor- phic zygomaturine clade ancestral to Plaisio¬ don centralis and Neohelos tirarensis. ACKNOWLEDGEMENTS I thank Rod Wells and Tom Rich for their comments on the manuscript. Thanks also go to the Flinders University Students who helped dig out the extension trench and to Danny Turner who spent the better part of a week excavating the specimen. REFERENCES Aplin, K. and Archer, M. 1987. Recent advances in marsupial systematics with a new syncretic classification. In: M. Archer (ed.) Possums and oppossums; studies in evolution, xv-lxxii. Surrey Beatty and Sons and the Royal Society of New South Wales: Sydney. Archer, M. and Bartholomai, A. 1978. Tertiary mammals of Australia: a synoptic review. Alcheringa 2:1-19. Everenden, J., Savage, D., Curtis, G. and James, G. 1964. Potassium-argon dates and the Ceno- zoic mammalian chronology of North America. American Journal of Science 262:145-198. Hoch, E. and Holm. P. 1986. New K/Ar age deter¬ minations of the Awe Fauna Gangue, Papua New Guinea: Consequences for Papuaus- tralian late Cenozoic stratigraphy. Modern Geology 10:181-195. Murray, P. 1986. Propalorchestes novaculacepha- lus gen. et sp. nov., a new palorchestid (Dipro- todontoidea: Marsupialia from the Miocene Camfield Beds, Northern Territory, Australia. The Beagle, Occasional Papers of the North¬ ern Territory Museum of Arts and Sciences 3(1): 195-211. Murray, P. 1989 (unpublished). "Plaisiodon wehhi" a new zygomaturine species (Marsupi¬ alia: Zygomaturinae) from the late Miocene Alcoota Fauna of Central Australia. Confer¬ ence on Australian Vertebrate Evolution and Systematics, Sydney. (Abstract): 12. Page. R. and Mcdougall, I. 1972. Ages of minerali¬ zation of gold and porphyry copper deposits in the New Guinea Highlands. Economic Geology 67:1034-1048. Plane, M. 1967. Two new diprotodontids from the Pliocene Otibanda Formation, New Guinea. Tertiary Diprotodontidae from Australia and New Guinea. In: Bureau of Mineral Resources Geology and Geophysics Bulletin 85:105-128. Rich, T. Archer, M. andTedford, R. 1978. Raemeo- therium yatkolai, gen. et sp. nov., a primitive diprotodontid from the Medial Miocene of South Australia. Memoirs of the National Museum of Victoria 39:85-91. Stirton, R. 1967. The Diprotodontidae from the Ngapakaldi Fauna, South Australia. In: Terti¬ ary Diprotodontidae from Australia and New Guinea. Bureau of Mineral Resources Geol¬ ogy and Geophysics Bulletin 85:3-35. Woodburne, M. 1967a. The Alcoota Fauna, Cen¬ tral Australia an integrated palaeontological and geological study. Bureau of Mineral Re¬ sources Geology and Geophysics Bulletin 87:1-175. Woodburne, M. 1967b. Three new diprotodontids from the Tertiary of the Northern Territory, Australia. In: Tertiary Diprotodontidae from Australia and New Guinea. Bureau of Mineral Resources Geology and Geophysics Bulletin 85:53-104. Woodburne, M. Tedford, R. Archer, M. Turnbull, W. Plane, M. and Lundelius, E. 1985. Bio¬ chronology of the continental mammal record of Australia and New Guinea. Special Publi¬ cation of the South Australian Department of Mines and Energy. 5:347-363. Accepted 6 September 1990 80 HUMANITIES The Beagle, Records of the Northern Territory Museum of Arts and Sciences, 1990 7(2):83-88 A BRIEF NOTE ON THE MALAGAN CURVUNAVUNGA FROM TABAR, NEW IRELAND PROVINCE, PAPUA NEW GUINEA. MICHAEL GUNN Northern Territory Museum of Arts and Sciences, GPO Box 4646, Darwin NT 0801, Australia. ABSTRACT A type of malagan sculpture previously used for peace-making is described and placed within its current cultural context on Tabar, New Ireland. Keywords: Papua New Guinea, New Ireland, Tabar, malagan, art, peace-maker. INTRODUCTION Malagan traditions of New Ireland first came to notice of the outside world through the work of Richard Parkinson in the late 19th century (Meyer and Parkinson 1895; Parkin¬ son 1907). Parkinson, followed by Kramer (1925), Peekel (1910, 1926-7, 1928, 1929, 1931, 1932) and Walden (Walden and Never- mann 1940) developed the rudiments of the western understanding of malagan: that it is a series of rites accompanied by an outstanding art tradition concerned with honouring and maintaining the memory of the dead. Powder- maker (1931a, 1931b, 1932, 1933), Groves (1934, 1934-5, 1936, 1936-7) and then Biihler (1933) extended descriptive knowledge with detailed studies at the village level, but it was not until Lewis’s work in 1953-4 (Lewis 1969) that the social contextual meaning of malagan was explicated. More recently Brouwer (1980), Heintze (1969), and Klichler (1985, 1987) gave detailed analysis of malagan sym¬ bolism in three locations on mainland New Ireland. Most recently Lincoln (1987) edited a body of work by Bodrogi, Clay, Heintze, Gunn, Lincoln, and Wagner that covered as¬ pects of some of the still active malagan tradi¬ tions in New Ireland. Despite this relatively intensive documen¬ tation, there is no record of those items of malagan sculpture which were used by an¬ tagonistic clans at major peace ceremonies. This paper documents the last of these old Curvunavunga peace-making malagans on Tabar, the traditional home of malagan. The original function of the malagan type is de¬ scribed, followed by a discussion of the subse¬ quent use of the Curvunavunga sculpture found at the Obun village site. DISCUSSION In June 1982, on behalf of the National Museum of Papua New Guinea, the author purchased the last known remaining example of a Curvunavunga malagan on Tabar from Daniel Kariets of Vutigamgam clan ( matabu) at Manggavur village, southern Tatau Island. It is now registered with the National Museum of Papua New Guinea as 82.50.7.1. The four metre horizontal hardwood Curvunavunga malagan sculpture was in a badly weathered condition, but still resting on its support posts on Tirodan clan land in the centre of the old village site of Obun, 2 km east of Manggavur (Plate 1). Obun was last occu¬ pied before the Second World War, then aban¬ doned when the inhabitants moved to Manggavur. Although four people were able to give the author information about the Curvunavunga type malagan, only two locations for this malagan type were recorded and the example under discussion was the only one sighted. Around Mapua and southern Tatau Island (central in the Tabar Group) informants re¬ ferred to the malagan type as Curwunawunga or Curwunawungga. In central and southern Big Tabar it was referred to variously as Surwunawunga, Suruniwunawungga or Suruwiwunawungga. These glosses reflect 83 M. Gunn dialect variants based upon cur- = reproducing or cloning malagan (lit. bed. from which many spring); and vunavunga = ridge-pole of a house. At Tatau village, north Tatau Island, the Curvunavunga type of malagan was re¬ membered as Watirewong by Lomlom, now an old man over 70 years old. Other associations for this name were not known by Lomlom or any other informants. On the basis of data provided in 1982 by informants at Manggavur village and 1983-4 at Wang village, it seems that the Curvun¬ avunga malagan type was used in two phases: initially as a peace malagan, and later as a woman’s malagan used to mark the comple¬ tion of the initiation of young adult women into “big-woman” status. Up until the end of 1884 Tabar was per¬ ceived as hostile to all foreigners and the inhabitants were apparently at war with the people of mainland New Ireland (Romilly 1886:42) as well as with one another. Shortly after the German annexation of New Ireland in 1884, the warships “Elisabeth”, “Albatros”, “Hyane”, and “Marie” made the Imperial German Government's presence known in the area (see Sack and Clark 1979) and the people of Tabar were persuaded to cease fighting. As a peace-making malagan the Curvun¬ avunga assumed a significance above all other malagan types. From oral traditions recorded from Lepan Landavak at Wang village, south¬ ern Big Tabar, I understand that after initial peace negotiations had taken place the neces¬ sary gardens prepared and pigs were located and counted, then the Curvunavunga malagan sculpture was made by sculptors commis¬ sioned by both sides. Upon completion of the sculpture, the malagan ceremony for the Curvunavunga was jointly operated by the two disputing clans. During the ceremony, each clan exchanged a number of lengths of re shell discs in order to “buy the war-dead” of the opposing clan. Both sides then brought out the heads of all those who had been killed and placed them on top of the Curvunavunga. This action was termed “ vavunga kita pi a mu koa ni rivirua ” [this one is to finish for all time the heads of all those killed in the fighting]. After the peace was made the Curvun¬ avunga malagans remained in village centres until they disintegrated or were sold to for¬ eigners. If sold, the support posts would be left in the ground. One immense old fig tree near Datava village on Big Tabar Island still has a pair of posts amongst its roots, firmly planted in the ground. Local people said that these support posts were placed in the ground and consolidated by using strong magic associated with koravar (Zingiber sp.) and lime. This consolidation had the effect of placing a taboo on the posts so that the relics would remain as evidence of the reconciliation for future gen¬ erations. Plate 1. A horizontal Curvunavunga malagan sculpture from Tabar, four metres long, hardwood. 84 Note on the malagan Curvunavunga infant scarification marks Fig. 1. Reconstructed sketch of Curvunavunga. SUBSEQUENT USE OF THE OBUN CURVUNAVUNGA Those malagans which remained in the vil¬ lage centres became subject to the processes of transfer of ownership which affect all property on Tabar. Traditions of inheritance for the Curvunavunga were quite different to those for normal malagans on Tabar, for the Curvun¬ avunga appear to have been permanent fix¬ tures in village centres. Normal malagan sculpture is ephemeral, for it is made accord¬ ing to inherited prescription which changes owners at a malagan ceremony held in grave¬ yards. An example of the sculpture is made and displayed at the ceremony as part of the change of ownership of normal malagans, and is rarely reused (see Gunn 1987:78-83). Mala¬ gan sculpture is today either thrown into the vunotung “cold house” part of the graveyard after use, or else wrapped in pandanus spathe and kept in the roof of the men’s house to be sold to a passing foreigner. According to oral traditions recorded from Daniel Kariets at Manggavur, the Curvun¬ avunga was owned by the first born woman of each generation and passed from one woman owner to the next (Appendix 1). The first-born girl of the next generation was confined for a period in a separate building inside the men’s house area in the graveyard. She resided with 10 to 20 other children, both male and female, who were undergoing a period of seclusion (,davan) prior to the end of puberty, a process which made them into potential community leaders. Confinement of young women has been reported elsewhere in the literature from other parts of New Ireland. Kramer (1925:27) noted a very similar system in the Madak language area, where the confinement house was called eandavan. Parkinson (1907:272) reported the practice from south-western New Ireland, and Bell’s report (1936) on dafal from the Tanga Group indicates the geographic spread of this custom. Reports of the confine¬ ment of young men as well as women on Tabar was recorded from two unrelated informants and suggests a Tabar variation on the practices found elsewhere. When a girl who was to inherit the Curvun¬ avunga malagan was about to “come out”, she was escorted to the centre of the village and placed on a platform underneath the big malagan. She would then be given a piece of taro to hold in her hand and her father’s clas- sificatory sisters would come and eat Irom the taro. Those who ate the taro would then drape the initiate with re (strings of shell discs). After this section of the ceremony was com¬ pleted, the shell lengths would then be draped over the malagan above her. Following the initiation, women of the fa¬ ther’s clan would “play” with the initiate’s mother’s brothers. A similar “play” between clans during initiation is still carried out today in the beriber cycle of ceremonies for a first¬ born child. In one such “play” witnessed by the author in Monun village, north Simberi Island, a large group of singing women from the child’s clan climbed over the entrance into the men’s house enclosure and graveyard where the men were solemnly feasting, picked 85 M. Gunn up handfuls of food and rubbed it into the faces and hair of the men from the child’s father’s clan. This “play” culminated with the women flogging the men with twigs and then throwing them into the sea. Beriber women’s ceremo¬ nies today often have an attached malagan component, where there is a need to hand a malagan onto the next generation before the original owner dies. Ownership of the rights to produce malagan sculpture can only be trans¬ ferred to the next generation through public display of the sculpture. If the owner dies without passing on his rights to the next gen¬ eration, then the malagans he owns die with him. Consequently, when an owner ap¬ proaches death, every opportunity is taken to display and transfer malagan sculpture. Op¬ portunities include beriber ceremonies, a per¬ son returning home from a long absence and malagan ceremonies held by distant kinsmen. According to the oral traditions received by the current owner of the Curvunavunga, the images on the malagan found at Obun (Fig. 1) are of two women lying opposed, their genita¬ lia joined by the crab erimang. In this context erimang is one of a group of ancestral con¬ cepts called tadar (or variously tandar, tandaro, tendar) on Tabar. Tadar are spirit images linked to clans and strongly associated with land ownership. Tadar concepts will be discussed in a later paper. The author recorded a further example of Erimang on Tabar as Erimang Porpor Merik, a tadar crab belong¬ ing to Betet clan. This tadar is located close to the sea in a creek to the north near Maragat village on the north west part of Tatau Island. Porpor merik literally means “red flower” and is a reference to the red decoration or hair on the crab’s chest. It is not known whether Betet or any related clan had a connection with the malagan, the clans, or with the land in¬ volved with the Curvunavunga under discus¬ sion. An infant (sex now indeterminate) lies on the breast of the woman to the left, but due to the deteriorated condition of the carv ing it was not possible to determine whether a similar infant lay on the other woman’s breast. An¬ other example of Curvunavunga which was burnt in earlier times was described by an informant from Wang village, southern Big Tabar Island, as featuring a woman giving birth. According to a Daniel Kariets the concen¬ tric circles surrounding the women’s navels on the Curvunavunga are indicative of the scarification traditions of earlier times, when each woman had her own personal design. Veivem side rails found top and bottom on the malagan also occur on many current types of malagan sculpture, and often are said to indi¬ cate the spears previously used to support the dead in the bo, the chair on which corpses were previously displayed. Although the antecedents of the Curvun¬ avunga malagan traditions are not known, smaller malagan types are still used today to terminate clan separations and disputes. In the one example of a dispute termination recorded at Pekinberiu village by the author, both sides came together to “work a malagan” as part of a larger ceremony which both of the clans operated. The conciliatory ceremony was termed vedega “looking at one another”, and for this ceremony kupkup ci malaga “fountain of malagan” sculptures was used. Kupkup ci malaga are a quick-to-work type of malagan and are generally small figure sculptures, al¬ though other styles can also have this classifi¬ cation. Two malagan sculptures, together with re shell disc strings and pigs, were exchanged between the two big-men. The final action of vedega culminated with the two big-men each holding one end of a piece of pandanus, a ritual master minguc then cut the pandanus in two with a knife, signifying that “the argu¬ ment was broken”. ACKNOWLEDGEMENTS I would like to thank the people of Tabar for their work in documenting their traditions; and Nick Nani, Fr Bernie Miller, and Barry Craig for their help in securing the old mala¬ gan for the National Museum. The author’s fieldwork on Tabar in 1982 and 1983-4 was funded by the NT Museum in affiliation with the National Museum of Papua New Guinea. REFERENCES Bell, F.L.S. 1936. Dafal .Journal of the Polynesian Society 45(3):83-98. Brouwer, E.C. 1980. A Malagan to cover the grave - funerary ceremonies in Mandak. Unpub¬ lished PhD thesis. University of Queensland: St.Lucia. Bidder, A. 1933. Die Totenfeste in Nord Neu Irland. Verhandlungen der Schweizerischen Naturforschende Gesellschaft 114:243-270. 86 Note on the malagan Curvunavunga Groves, W. 1934. Fishing rites at Tabar. Oceania 4(4):432-457. Groves, W. 1934-5. Tabar today: a study of a Melanesian community in contact with alien non-primitive cultural forces. Oceania 5(2):224-240; 5(3):346-360; 6(2): 147-157. Groves, W. 1936. Secret beliefs and practices in New Ireland. Oceania 7(2):220-246. Groves, W. 1936-7. Settlement of disputes in Tabar. Oceania 7:501-520. Gunn, M. 1984. Tabar Malagan, an outline of the emic taxonomy. In: Philip J.C. Dark (ed.) Developments in the Arts of the Pacific. Pa¬ cific Arts Association Occasional Papers (1 ):81 -92. Gunn M. 1987. The Transfer of Malagan Owner¬ ship on Tabar. In: L. Lincoln (ed.) An Assem¬ blage of Spirits: Idea and Image in New Ireland, pp. 74-83. George Braziller: New York. Heintze, D. 1969. Ikonographische Studien zur Malanggan-kunst Neuirlands. Unter- suchungen an ausgewahlten Vogeldarstel- lungen. Inaugural dissertation (unpublished). Tubingen Schwitalla Himmelsthur. Kramer, A. 1925. Die Malanggane von Tombara. Georg Muller: MUnchen. Kuchler, S. 1985. Malanggan: exchange and re¬ gional integration in northern New Ireland. Unpublished PhD thesis. London School of Economics: London. Kuchler, S. 1987. Malangan: art and memory in a Melanesian society. Man (N.S.) 22:238-255 Lewis, P.H. 1969. The social context of art in northern New Ireland. Fieldiana Anthropol¬ ogy Volume 58. Field Museum of Natural History: Chicago. Lincoln, L. (ed.) 1987. An Assemblage of Spirits: Idea and Image in New Ireland. George Bra¬ ziller (in association with The Minneapolis Institute of Arts): New York. Meyer, A.B. and Parkinson, R. 1895. Schnitzerein und Masken vom Bismarck Archipel und Neu Guinea. Publicationen aus dem Koniglichen APPENDIX 1 Although the mechanism for the establish¬ ment of the original ownership of the Obun Curvunavunga is not clear, the malagan is understood by the current owner to have been passed to the first-born daughter for four gen¬ erations. However, the pattern of ownership gives more exceptions than the norm. Maso, the first owner in the second generation, had no children, so she passed it to her sister Melul. Melul had three children, two boys and a third Ethnographischen Museum zu Dresden. Vol¬ ume 10. Parkinson, R.H.R. 1907. Dreissig Jahre in der Siidsee. Strecker and Schroder: Stuttgart. Peekel, P.G. 1910. Religion und Zauberei aufdem Mittleren Neu-Mecklenburg, Bismarck Archi¬ pel, Sudsee. Aschendorff: Munster, and An- thropos Bibliothek: Vienna. Peekel, P.G. 1926-7. Die Ahnenbilder von Nord- Neu-Mecklenburg. Eine kritische und positive studie. Anthropos 21:806-824; 22:16-44. Peekel, P.G. 1928. Lang-Manu, die Schlussfeier eines Malagan-(Ahnen-) Festes auf Nord-Neu- Mecklenbburg. In: W. Koppers (ed.) Fest¬ schrift Publication d‘hommage offerte au P.W.Schmidt. 542-555. Wien. Peekel, P.G. 1929. Das Zweigeschlechterwesen. Anthropos 24:1005-1072. Peekel, P.G. 1931. Religiose Tanze auf New Irland (Neu Mecklenburg). Anthropos 26:513-532. Peekel, P.G. 1932. Uli und Ulifeier, oder, vom Mondkultus auf Neu Mecklenburg. Archivfiir Anthropologie N.F. 23:41-75. Powdermaker, H. 1931a. Report on field-work in New Ireland. Oceania 1:355-365. Powdermaker, IT 1931(b). Mortuary rites in New Ireland. Oceania 2:26-43. Powdermaker, H. 1932. Feast in New Ireland; the social functions of eating. American Anthropologist 43:236-247. Powdermaker, H. 1933 (1979). Life in Lesu. Wil¬ liams and Norgate: London. Romilly, H. 1886. The Western Pacific and New Guinea. Notes on the natives, Christian and cannibal, with some account'of the old labour trade. John Murray: London. Sack, P. and Clark, D. (eds and translators) 1979. German New Guinea: the annual reports. ANU Press: Canberra. Walden, E. and Nevermann, H. 1940. Totenfeiern und Malagane von Nord-Neumecklenburg - nach aufzeichnungcn von E. Walden. Zeitschrift fur Ethnologie 72:11-38. Accepted 12 September 1990 born girl, who unfortunately was born deaf mute and did not marry. So the Curvunavunga was passed out of the clan to Lagase, the first born son, to pass to his first daughter when she came of age. But Lagase fathered two sons and no daughters, so the pattern was again inter¬ rupted when he was forced to pass it out of the clan to his eldest son, Daniel Kariets. Kariets had intended to pass the ownership of this Curvunavunga to his eldest daughter, Matse in Katobi clan, but decided it would be better kept in the National Museum. 87 M. Gunn A — O [Tirodan clan] O A A A A O O Fig. 2 Ownership history of the Obun Curvunavunga 88 The Beagle, Records of the Northern Territory Museum of Arts and Sciences, 1990 7(2):89-93 WEHIRIYA - A MISSING GRAVE AND A LOST TOWN PETER SPILLETT (DAENG MAKKULLE) Northern Territory Museum of Arts and Sciences, GPO Box 4646, Darwin NT 0801, Australia. ABSTRACT The discovery and identification of the grave of a Buginese princess near Dili, East Timor, indicates a closer nexus between South Sulawesi and Timor than previously thought, and a Bugis/Makassan settlement predating Portuguese occupation. Keywords: Bugis, Dili, graves, Hera, history, historic site, Islam, Makassans, Ti¬ mor. INTRODUCTION This contribution is part of my ongoing research into the Makassan voyages to Austra¬ lia since the 16th century, and the expansion of Islam to the Kingdoms of Gowa and Tallo in South Sulawesi and Waihale in Timor. Whilst working in Makassar, Zainuddin Daeng Pabetta, a descendant of the Kings of Tallo, recounted traditional legends relating to the expeditions to Timor by the Kings of Tallo in the 16th, 17th and 18th centuries at the behest of the Solorese in a bid to oust the Portuguese. DISCUSSION In April 1989,1 took the opportunity to visit East Timor to carry out a survey of archival and library holdings in Dili as well as to con¬ duct interviews with people who might have knowledge of Makassan contact with the area. Following from a radio interview on the purpose of my visit to East Timor, I received a phone call from Ronny Sjamsuddin, a Makas¬ san reporter with Antara Newsagency. He had something to show me. Would I go and talk with him? It turned out to be a small publica¬ tion produced by the Department of Education & Culture which contained illustrations of historic buildings and objects in East Timor together with brief texts. My attention was drawn to a dark photo¬ graph of a carved stone on which were some barely discernible characters. The caption suggested that from the shape of the stone, it was possibly part of an Islamic grave but the decoration was more in keeping with classic Hindu, not East Timor culture. The inscrip¬ tion could not be identified. We discussed the characters and came to the conclusion that there were two distinct scripts and two languages. The first were Arabic numerals which we deciphered as the year 1052 of the Hijrah. The other appeared to be either Lontara; Makassar or Lontara; Bugis. (Arabic is written from right to left, and Lon¬ tara is written from left to right). Apparently the stone had been found by a resident of Dili in the jungle. I made a number of enquiries as to whom the finder of the ‘Nisan’ (Muslim gravestone) was, or where it was deposited, but without success. On a later visit to Makassar I consulted with Drs Muttalib and Drs Bahru of the Department for the Protection and Conservation of His¬ torical and Archaeological Remains and they confirmed the Arabic date as A.D. 1632. The Lontara’ is Lontara’ Bugis which reads - WEHIRIYA. “WE" is an honorific indicat¬ ing a Buginese princess and “Hiriya" her personal name. The Christian year is based on the calculation for year 1052 Hijrah plus 580. This was an exciting find. It was not what I was searching for but could be an indicator. I sent the news to Sjamsuddin in Dili and asked him to check further but unfortunately I had no further contact with him. In September 1989, I returned to Dili and made further enquiries regarding the identity of the finder of the grave stone and it’s loca- 89 P. Spillett Fig. 1. Headstone from the grave of Wehiriya. tion. It was only when I contacted the Director of the East Timor Museum. Drs Paulus Susilo, that I at last had some results. He told me that he now knew who the finder of the gravestone was and where it was lo¬ cated. When I told him of the translation of the inscription he immediately became interested in it and agreed to my request to interview the finder and to visit the site where it was found, since he had not previously visited the site. It was important that we endeavoured to locate the grave site as this could be an indica¬ tor to a cemetery and possibly a village of considerable age. We met with the Camat and the Kepala Desa, Francisco Da Silva, the finder of the stone. Da Silva explained that an area of land was being cleared to make way for development of a new Kampung, when the stone was knocked over. He had removed it to a safe place since he suspected that it was of some importance although he knew nothing about it. We were driven to the place where the stone was kept and a quick inspection showed that the inscription shown in the photograph was identical (Fig. 1). This object was quite defi¬ nitely an Islamic grave post made of locally quarried stone. From the Camat’s office we drove to the small town of Hera, some 20 km from Dili by road (but no more than 3 km by sea). From the road we walked inland a few hundred metres, passing by two very old Beringin trees which probably indicate the presence of a former Kampung. We came to a place in an over¬ grown area admidst scrub land and gathered at a small low mound covered with grass and stones, possibly indicators of other graves. I questioned Da Silva as to the orientation of the head stone because being an Islamic bur¬ ial, as I suspected, the grave would be oriented north-south with the head to the north, and with the body lying on its right side, facing towards Mecca (in this case to the west). He 90 Wehiriya -a missing grave could not remember clearly but indicated an east-west orientation. I explained that it is normal in Islamic burial for there to be one headstone for a male person and a headstone and a footstone, which is usually smaller, for a female person. He could not remember ever seeing a foot- stone. It was rather disappointing and incon¬ clusive. We discussed the situation and searched the area around for signs of other graves, but apart from the slightly elevated ground and the small rocks which were spread around there was nothing else. I then decided to take some photographs of the site and the headstone. The grass was beaten flat and, so that I could get a better view, I asked the men to pull aside a small thick bush atop the mound. Everyone was astounded to find, in situ, below the bush, a small carved stone (Fig. 2). The footstone was small, made of the same material as the headstone, but with no inscrip¬ tion or decoration. But what was most impor¬ tant was that on matching the headstone with the foot-stone, the grave was found to have a north-south orientation, thus indicating it to be an Islamic burial. As a result of these findings further research is now being conducted on the toponymy of the area and its local history, in an endeavour to determine the history of what may quite well be the earlier village of Hera. The two Beringin trees are ‘keramat’, that is, sacred to the local people, and are situated on the eastern side of a small silted-up river about 100 m away and the grave of Wehiriya is approximately 200 m to the east. The name of the grave site is called Makirate, the place of the dead, and the name of the kampung nearest the site is Manuroni. The name of the suku or tribe of this area is Mantarlidu, a part of the Kingdom of Hera which also holds authority over the island of Atauro. From an unidentified, mid-17th century Dutch map of Timor (Fig. 3), which shows a distorted geography of the eastern part, there Fig. 2. Gravesitc of Wehiriya, showing footstone. 91 P. Spillett Fig. 3. Map of Timor, showing (top) a distorted geography of the eastern part, and (bottom) corresponding place names in contemporary usage. 92 Wehiriya -a missing grave is shown a river named “Macassars River”, which is approximately where one would expect to find Hera. CONCLUSION Further research should reveal the history of the site; but what will it be? The town and kingdom of Hera appear to be older than the city of Cioli. Are we searching for the ancient town of Hera or Kampung Makassar where the King of Tallo died? What was a Buginese princess doing in East Timor before the advent of the Portuguese in Dili, being an Islamic princess when Islam was only accepted by the Kingdom of Bone in 1611? Hera . Hira .Hiriya.?? ACKNOWLEDGEMENTS Acknowledgements are made to the follow¬ ing people for their advice, oral history, maps: Zainuddin Dg. Pabetta, Arifuddin Hamid, Ronny Sjamsuddin, Drs. Muttalib, Drs. Bahur, Drs. Susilo, Fransisco Da Silva, Eugenio Sarmento and Kevin Sherlock. Accepted: 20 July 1990 93 The Beagle. Records of the Northern Territory Museum of Arts and Sciences, 1990 7 ( 2 ): 95-96 BOOK REVIEW Dragonflies (Naturalists’ Handbooks; 7) Peter L. Millar Cambridge University Press 1987, ISBN 0521 317657 paperback, pp. 84; RRP. $21.50 This volume is the seventh in a series de¬ signed to aid people without a university edu¬ cation in biology who may be interested in studying local natural history. The editors’ preface states “ The books in this series of¬ fer... the information and ideas needed to plan an investigation and the practical guidance needed to carry it out. They draw attention to regions on the frontiers of current knowledge where amateur studies have much to offer.” The book is divided into 6 chapters, each of which offers information under various sub¬ headings. While other dragonfly faunas are mentioned the emphasis is on the British fauna. The Introduction contains four sections Evolution; The aesthetic appeal and economic importance of dragonflies; The British fauna; Dragonflies are very good animals for field work. Chapter two is titled Eggs and Larvae and provides a good summary of the types of eggs and larvae produced by dragonflies. Other topics discussed in this chapter include habitat specificity, different modes of feeding, loco¬ motion, territorial behaviour and respiration. The third chapter entitled simply “The adult” is by far the largest in the book. There are 21 subheadings covering a wide range of topics which are logically arranged beginning with Emergence, then progressing through such topics as Flight, Thermoregulation, Vi¬ sion, Feeding behaviour, to Reproduction, Courtship, and Oviposition. This chapter is particularly well written. In concise, easy to understand text each of the topics is examined. The reader quickly be¬ comes aware of the diversity of dragonfly biology and of the complexity of dragonfly behaviour. It also becomes apparent that drag¬ onflies are one of the better known groups of insects. At every opportunity the author highlights where there are significant gaps in our knowl¬ edge of dragonflies and suggests research which could be undertaken by the reader. The fourth chapter provides keys to larvae and adults of British species. The key to larvae was prepared by Graham Vick and the key to adults by David Chelmick. Any key to species usually contains a lot of specialist terminology and a poor understand¬ ing of the terminology often leads to errors in identification. Both of the keys presented here are profusely illustrated. Definitions of many terms are provided and important structures are clearly labelled on the illustrations. Chapter five is a brief examination of the status of the British dragonfly fauna which notes that 11 of 39 breeding species are cur¬ rently under threat. Drainage of wetlands for land reclamation and pollution from agricul¬ tural, industrial or sewage disposal are respon¬ sible. Attempts to preserve habitats or create new ones are documented, and descriptions of ideal dragonfly habitats are provided. The dragonfly recording scheme, which seeks to update records of dragonfly species in Britain, is described at the end of the chapter. Chapter six describes some techniques for studying dragonflies and includes some useful tips on experimental design and how to write up the findings of research so that it may be published. Several journals which publish short papers on insect biology are listed and there is also a list of useful addresses, includ¬ ing several dragonfly societies, entomological 95 Book Review suppliers and the headquarters of the dragon¬ fly recording scheme. An excellent bibliography of important works on dragonflies is provided under the heading Further Reading. There are two appendices, the first is a checklist of the British species and the second documents the months of the year when adults of the British species appear. The final part of the book is a small index. The volume is well illustrated throughout with line drawings and there are four good colour plates in the centre of the book. As the subject of this work is the dragonflies of Britain, it could be said that it is of limited use outside the British isles or western Europe. While it is certainly true that most of the species mentioned are more or less endemic to that region, the general description of dragon¬ fly biology provided would be useful any¬ where. The volume is a very handy size for field studies and is printed on good quality paper. One minor fault with the paperback version is the use of two staples to bind the pages to¬ gether; one of the staples fell out not long after I began reading. In conclusion, I enjoyed reading this book which I found informative and stimulating. It is well written and illustrated. Perhaps the most endearing quality of the book is the obvious fascination of the author for all things dragonfly. 1 have no hesitation in recommend¬ ing this volume to anyone who is interested in dragonflies or general natural history. J. RUSSELL HANLEY Northern Territory Museum of Arts and Sciences GPO Box 4646 Darwin, NT 0801, Australia 96 -ST IS .^ GUIDE TO AUTHORS Authors are advised to follow the layout and style in the most recent issue of The Beagle. Three copies of typewritten manuscripts should be submitted. Manuscripts should be in English, double-spaced throughout and have a margin of at least 4cm on the left-hand side. Text should be on one side of good quality A4 bond paper. If avail¬ able, a computer file of the manuscript (on 5" or 3 " MS/IBM DOS floppy disk) should be submit¬ ted together with the printed version. Where ap¬ propriate, articles should conform to the sequence: Title, Abstract, Keywords, Introduction, Materi¬ als and Methods, Text, Discussion, Acknowledge¬ ments, References. The Title should be concise and informative. An abridged title (not exceeding 50 letter spaces) should be included for use as a running head. The Abstract should not exceed 150 words, and should state concisely the scope of the work and give the principal findings. Keywords, to facilitate information retrieval, of up to 15 in number should be chosen to outline the main subjects covered. The Introduction, including a review of litera¬ ture, should not exceed what is necessary to indi¬ cate the reason for the work and the essential back¬ ground. Abbreviations used throughout the text may be explained at the end of the introductory material, or placed separately in the Materials and Methods section. Footnotes are to be avoided, wherever possible, except in papers dealing with historical subjects. The International System of units should be used. In the descriptive text numbers from one to nine should be spelt out and figures used for numbers over nine. For associated groups, figures should be used consistently, e.g. 5 to 10. not five to 10. Systematic papers must conform with the Inter¬ national Code of Zoological Nomenclature and, wherever possible, with their recommendations. Synonymies should be given in the short form (taxon author, date:page) and the full reference cited at the end of the paper. Full citations of taxa used in the text (i.e. taxon author, date) must also be included in the references, whereas the short citation (i.e. taxon author) need not be included. Subsequent citations of taxa given in synonymies should be separated from bibliographical details by a dash (-). TABLES Tables should be numbered with arabic numer¬ als and accompanied by a title. Horizontal rules are inserted only above and below column head¬ ings and at the foot of the table. Footnotes on tables should be kept to a minimum and be re¬ served for specific items in columns. All other explanatory material should be incorporated with the title. ILLUSTRATIONS Line drawings, maps, graphs and photographs are generally regarded as “figures” and are to be numbered consecutively for interspersion through the text. Drawings must be on drafting film or good quality board with appropriate lettering in¬ serted. Black and white photographs must be sharp, of high contrast on glossy paper, and mounted on board. The author’s name, title of paper and figure number must be indicated on the reverse side of all illustrations. Captions or leg¬ ends should be typed together on pages at the end of the text. Colour illustrations may be accepted for larger papers, such as substantial revisions, and only where colour contributes significantly to the de¬ scriptive or scientific merit of the paper. CITATIONS AND REFERENCES Citations of sources within the body of the text should include the author, year of publication and page reference (where appropriate), e.g. Roth (1896); (Roth 1896); (Roth 1896, 1898); (Roth 1896; Smith 1915). Note that commas are not used to separate bibliographical citations (e.g. Roth 1896), whereas they should be included for taxo¬ nomic citations; for the original citation (e.g. Dasyurus Geoffroy, 1796), but not for subsequent citations (e.g. Dasyurus - Gould 1842). References should be arranged alphabetically and chronologically at the end of the paper. Titles of all references must be given in full and wher¬ ever possible citations given in BIOSIS format. Where an author has published more than one work referred to in the same year, the references should be appended with the letter (a), (b), etc. The following examples show the style to be followed: Brake, B., McNeish, J. and Simmons, D. 1979. Art of the Pacific. Oxford University Press: Wel¬ lington. Cogger, H.G. 1981. A Biogeographic Study of the Arnhem Land Herpetofauna. In: Proceedings of the Melbourne Herpetological Symposium: 148-155. Zoological Board of Victoria: Mel¬ bourne. Kaeppler, A.L. 1963. Ceremonial Masks: a Mela¬ nesian art style. Journal of the Polynesian Society 72(2): 118-138. Roth, H.L. 1896. The natives of Sarawak and British North Borneo. 2 volumes. Truslove and Hanson: London [Textual reference: Roth 1896 (II):22-26[. Skwarko, S.K. 1968. Mesozoic. In: Geology of the Katherine-Darwin Region, Northern Territory. Bureau of Mineral Resources. Ge¬ ology and Geophysics Bulletin 82:105-116. The BEAGLE Volume 7 Number 2 December, 1990 CONTENTS SCIENCES BRUCE, A.J. - A second species of Bresilia B. plumifera sp. nov., new to the Australian fauna (Crustacea: Decapoda: Bresiliidae). 1 BRUCE, A.J. - Recent additions to the pontoniine shrimp fauna of Australia. 9 AKIRA HIRAYAMA - Two new caprellidean (n. gen.) and known gammaridean amphipods (Crustacea) collected from a sponge in Noumea, New Caledonia. 21 SADLIER, R.A. - A new species of scincid lizard from Western Arnhem Land, Northern Territory.. 29 RUSSELL, B.C. - A new species of Nemipterus (Pisces: Nemipteridae) from the south-western Pacific. 35 MURRAY, P. — Primitive marsupial tapirs ( Propalorchestes novaculacephalus Mur¬ ray and P. ponticulus sp. nov.) from the mid-Miocene of North Australia (Marsupi- alia: Palorchestidae)... 39 MURRAY, P. - Alkwertatherium webbi, a new zygomaturine genus and species from the Late Miocene Alcoota local fauna, Northern Territory (Marsupialia: Diproto- dontidae). .'. . 53 HUMANITIES GUNN, M. - A brief note on the Malagan Curvunavunga from Tabar, New Ireland Province, Papua New Guinea. 83 SPILLETT, P. — Wehiriya - a missing grave and a lost town. 89 BOOK REVIEW HANLEY, J.R. - Dragonflies (Naturalists' Handbooks; 7). 95 SPS2