JOURNAL OF THE ROYAL SOCIETY OF WESTERN AUSTRALIA VOLUME 48 PART 1 PUBLISHED 12TH MARCH, 1965 REGISTERED AT THE G.P.O., PERTH FOR TRANSMISSION BY POST AS A PERIODICAL THE President Fast President Vice-Presidents Joint Hon. Secretaries Hon. Treasurer . Hon. Librarian Hon. Editors ROYAL SOCIETY OP WESTERN AUSTRALIA COUNCIL 1964-1965 W. R. Wallace, Dip.For. C. F. H. Jenkins, M.A. J. H. Lord, B.Sc. D. Merrilees, B.Sc. Margaret E. Redman, B.Sc. A. B. Hatch, M.Sc., Dip.For. R. D. Royce, B.Sc. (Agric.). Ariadna Neumann, B.A. R. W. George, B.Sc., Ph.D. (1964). J. E. Glover, B.Sc., Ph.D. (1965). B. E. Balme, B.Sc. (Hons.). J. S. Beard, M.A., B.Sc., Ph.D. A. S. George, B.A. J. G. Kay, B.Sc. P. E. Playford, B.Sc., Ph.D. R. T. Prider, B.Sc., Ph.D., M.Aust.I.M.M., F.G.S. W. D. L. Ride, M.A., D.Phil. D. L. Serventy, B.Sc, (Hons.), Ph.D. Journal of the Royal Society of Western Australia Vol. 48 Part 1 1. — The Physiography, Vegetation and Vertebrate Fauna of the Wallabi Group, Houtman Abrolhos By G. M. StoiT* Manuscript Accepted — 18th August, 1964. The physiography and vegetation are des- cribed. All known vascular plants are listed (in- cluding those from North Island). Brief accounts are given of the reptiles, land-birds and mammals. The biogeography of the islands is discussed. Introduction In order to procure live wallabies (Macropus eugenii) for experiments and to observe them in the field, the Zoology Department organised three short trips to the Wallabi Islands in 1959 (April 16-22, June 21-25 and September 2-12) and one in 1960 (April 22-27). The writer had the good fortune of taking part in all of them and was able to make extensive collections of vascular plants and vertebrate animals. The physiography of the Wallabi Group has been described by Dakin (1919) and still more fully by Teichert (1947). The following account is therefore brief — just enough for describing plant habitats and for a later discussion of the biogeography of the islands. The vegetation will be dealt with at greater length, for scarcely any- thing was previously known of it. In the ac- counts of the fauna, the writer has drawn freely on the observations of the naturalists who pre- ceded him in these islands. Physiography The Houtman Abrolhos lie about 40 miles off the west coast of Western Australia between latitudes 28 and 29. The sea-bed west of the mainland descends in the first 5-10 miles to a depth of 20 fathoms, but one must go a fur- ther 40-45 miles before reaching the 40-fathom line, westwards from which the sea-bed dips steeply to the edge of the continental shelf. The Abrolhos are thus situated towards the western margin of a gently sloping platform. Each of the island clusters (Wallabi, Easter and Pelsart Groups) is located on a plateau-like hump on this platform. Although soundings are limited, it seems that North Island rises from the same submarine plateau as the Wallabi Group. ♦Zoology Department, University of Western Australia. Present address: Western Australian Museum, Perth. East and West Wallabi, with areas of 900 and 1,500 acres and dunes up to 50 feet high, are by far the largest and highest of the Abrol- hos Islands. They and their nearest neigh- bours emerge from a broad rock-flat, composed mainly of old coral reefs which were planed in an earlier cycle of erosion to a level of 1-2 feet above present sea-level. Subsequent erosion has lowered much of the fiat, though considerable areas in the lee of West Wallabi still lie above low water. On much of West Wallabi and almost all of East Wallabi the old reefs are overlain by beds of younger limestone, 3-10 feet thick. Their composition varies: in some areas, shells or coral fragments predominate: in others, there is little or no macroscopic structure. But what- ever their composition, the beds are remarkably uniform in appearance at the surface. Every- where the fairly level exposures are cracked, so that the resultant slabs form, as it were, a gigantic crazy pavement. There is sufficient soil between the slabs to support a moderately varied assemblage of shrubs. Subterranean caverns are common, and several sink-holes have been enlarged or cleaned out to form wells that provide a modest supply of fresh water. This pavement limestone occupies a large un- broken portion of the western half of East Wallabi. Its occurrence on West Wallabi is dis- continuous. The largest area is in the central south of the island, extending from low cliffs iii Rocky Bay to the inner foot of the west coast dunes. The next largest occurrence is on the ridge immediately inland from the north- east coast between Blowfish and Slaughter Points. Towards the northern end of this ridge the beds attain their highest elevation (c.lO feet). All the islets in the vicinity of the two main islands (Pigeon, Little Pigeon, Seagull, Tattler, Pelican, etc.) are composed entirely of this limestone. UniformJy about six feet high and surrounded by vertical or undercut cliffs, they are confusingly similar in appearance. In contrast, the low islets at the eastern margin of the Wallabi Group (Beacon, Long, Dick’s, etc.) consist largely of coral boulders and 1 ^FISH POINf PELICAN ISLAND BLOWFISH POINT SHAG BAY PELICAN POINT Eagle Hill PEOItllT BAY HOmYABOXt Horseyord » n '~-i • Horse Well SEAGULL^ ISLAND^ SLAHGHTEP POINT TATTLEP BAY ff WTLER ISLAND c:>MANGROVE ISLAND 0 ^^PIGEON ISLAND LITTLE PIGEON ISLAND EAGLE POINT \ L MILES 0 J l_ N Fig. 1. — Wallabi Islands. Houtman Abrolhos. Western Australia. shingle, much of it too recently thrown up from neighbouring lagoons to have become consoli- dated- The only considerable part of the islands in which the old reef flat is not overlain by pave- ment limestone is the northern end of West Wallabi, west of the limestone ridge. Low-lying- areas (less than two feet above sea-level), where the reef flat may appear at the surface, carry sheets of water after winter rains. Elsewhere the flat is covered with up to five feet of shell grit, interrupted here and there by narrow banks of guano-rich soil, usually striking N.E.- S.W. Evidently the whole of this area was a part of Shag Bay when sea-level was slightly higher than now. There is another and smaller break in the pavement limestone at the salt-lake, a little inland from the south-east coast of West Wallabi. At high tide the lake is connected with the sea by a small creek. On East and West Wallabi (and North Island), alone of the Abrolhos, the pavement limestone may be overlain by dunes. Along the north and west coasts of East Wallabi and the central west coast of West Wallabi the dunes of calcareous sand are unconsolidated, and in some places they are completely blown out. Be- tween these dunes and the sea there are sandy beaches. In the eastern half of East Wallabi and the south-west corner of West Wallabi the dunes are consolidated, their orientation gen- erally being N.N.W/S.S.E. and their bulk con- sisting of aeolianite, as can be seen in exposures at coastal cliffs. Where covered with sand the underlying presence of aeolianite is indicated by the greater density and variety of the vegeta- tion compared to that on sandy dunes. Vegetation The dominant plant above beaches is the coarse tussocky grass, Spmifex longifolius. Other species are scarce and are repi*esented by scattered individuals or an occasional stand of Atriplex cinerea, Salsola kali, Senecio lautus and Poa caespitosa. Shrubs are more common in the sandy dunes, e.g., Olearia axillaris, Acantlio- carpus preissi and occasionally Myoporum insu- lare: but much of the ground is bare. Seeing that the distinction between sand and shell grit is merely one of particle size, the vegetation of the two are surprisingly different. The plant cover is much more continuous on 2 Sandy Beaches and Dunes Consolidated Dunes Pavement Limestone OutQr edge of Reef Flats Fig 2 — Wallabi Islands, showing physiographic divisions. The unshaded areas mainly consist of shell grit. The broken line in the middle of West Wallabi encloses land less than two feet above sea-level; it is an old rock-flat with little or no coveiHng of shell grit: to its west is a small ridge of pavement limestone that was inadvertently left unhatched on this map. the shell grit. A low saltbush is dominant, Atripiex paludosa, which tends to be replaced in low-lying areas by the shrubby samphire, Arthrocnemum halocnemoides. Other species include Threlkeldia diffusa, Senecio lautus, Frankenia pauciflora and stunted shrubs of Olearia axillaris. The whole of this country on West Wallabi is riddled with the burrows of mutton-birds iPuffinus pacificus). On East Wallabi, where beds of shell grit are much less extensive, the saltbush and samphire is replaced by thickets of Diplolaena dampieri and Alyxia huxifolia. In the narrow ecotone between the shell grit and pavement limestone the vegetation becomes sparse and virtually restricted to Frankenia. Despite the scarcity of soil, the pavement limestone supports a fairly rich vegetation. The dominant species are shrubs from 3-8 feet tall: Pittosporum phillyreoides, Diplolaena dampieri, Grevillea argyrodendron Spyridiu7u globulosuvi. Exocarpus aphylla, Capparis spinosa, Pimelea microcephala, Sarcostemina australe, Olearia axillaris and Beyeria viscosa. Also commonly present are Dianella revoluta, Acanthocarpus preissii and Hibhertia suhvaginata. All plants of the pavement limestone may be found on consolidated dunes; and on East Wallabi, where the aeolianite is usually close to the surface, several additional shrubs appear: Dodonaea inaequifolia, Trichinium divaricatum, Lasiopetalum angustifolium, Leucopogon insu- laris, Acacia bivenosa, Mirbelia ramulosa and Bossiaea rufa. In one valley there is a thicket of Eucalyptus oraria, about 6 chains long and 1-2 chains wide. Though their stems are up to six inches in diameter, they have been so bent over by the prevailing southerlies that the 3 tops of the trees are only 3-8 feet above the ground. Possibly some of the larger trees have been removed, for in 1843 Gilbert found a tree whose trunk was 18 inches through. Otherwise this thicket seems to be much as it was 120 years ago. In contrast to East Wallabi, the flora of the consolidated dunes on West Wallabi is im- poverished. Large areas are completely over- run by Nitraria schoberi; and the introduced succulent, Cryophytum crystallinum, monopo- lises otherwise bare ground. Annual rainfall in the Abrolhos is probably about 12-15 inches, most of it falling from May to September. During my visit in April 1959, the condition of the vegetation was gener- ally poor. The first substantial rain fell (in my absence) on May 29. Some showers were received on June 17, 2-3 inches on June 19, and a considerable amount on June 20 (the day before my second visit). By June 22 the vegetation had recovered: Pimelea was shooting profusely: Carpobrotus was clearly filling out; and Diplolaena had lost its grey half -dead appearance. Seedlings of annual species were abundant, but most of them were too young for identification. By early September the annuals had matured and were flowering. They v/ere especially plentiful in mutton-bird rookeries and on offshore islets. One of the anomalies of the vegetation of the islands is its greater richness on East Wallabi than on the considerably larger West Wallabi. The latter, though generally similar physio- graphically to East Wallabi, differs in certain other respects. First, West Wallabi has (or had) large deposits of guano. Second, it alone is colonised by burrowing petrels. Finally, but not so certainly, it has a denser population of wallabies. It has already been mentioned that the con- solidated dunes of West Wallabi are dominated by Nitraria, whereas those of East Wallabi sup- port a good variety of shrubs and low trees, including several species unknown from West Wallabi. This difference can only be due to the ubiquitous presence of guano in the south- western dunes of West Wallabi. As Gillham (1963) has shown, few species of plants tolerate a high concentration of guano. Extraction of the guano between 1884 and World War I involved the vegetation in large- scale disturbances. Before removing the guano, which mostly lodges among boulders of lime- stone, all plants were stripped from an area, the stones laid to one side, and the soil shovelled with hoes and swept with heavy brooms into heaps (Helms 1902). Apart from the remains of a jetty, tramline and horseyard, and some obviously unnatural hollows and heaps of stones in the dunes, there was little evidence in 1959 of this once profitable industry. All the old excavations were completely revegetated but whether any plant species were permanently affected by this activity could not be ascertained. Likewise unknown is the effect of stock on the islands, e.g. the flock of goats formerly de- pastured on East Wallabi. The presence of burrowing petrels on West Wallabi has had a twofold effect: (1) the suppression of sclerophyllous shrubs (with the partial exception of Frankenia and Olearia) and their replacement by halophytic shrubs: (2) the encouragement of coprophilous annuals such as Urtica urens, Chenopodium murale, Stellaria media etc. The dominance of saltbush and samphire on the beds of shell grit inland from Shag Bay seemed at first sight to be due to soil salinity. But on East Wallabi similar beds be- tween the limestone pavement and the beach at the southern tip of the island were found to carry a scrub of Diplolaena and Alyxia. The conclusion was inescapable that the petrels, whose burrows occurred throughout the West Wallabi shell grit, were in some way responsible for the dominance of the halophytes. Apart from their ‘hedging’ of shrubs (details of which are given later) it is not easy to dis- cern the effect of wallabies on the vegetation. If one compares the vegetation of the Wallabi Islands with that of North Island (on which wallabies no longer occur), the only obvious differences are the greater abundance of Myoporum on North Island, and the greater height and density of the annuals, especially grasses and composites, several species of which have not yet been recorded from the Wallabi Islands. As for differences between the two Wallabi Islands, the only one that seems attri- butable to wallabies is the relative abundance of Eremophila glabra in the eastern dunes of East Wallabi, where the animals are apparently not plentiful. This highly palatable under- shrub was found on West Wallabi only in a small area east of Horse Well. If a portion of the old air-strip on East Wallabi were fenced off, the effect of the numerous wallabies in that area would soon be revealed. Annotated List of Plants In a brief account of the vegetation of North Island (Storr 1960) only perennial plants were mentioned. The opportunity has been taken here to include ail the species that were col- lected or observed on that island. An asterisk before a specific name signifies that the plant has been introduced, and one after the name of an island signifies that the record is based on material identified by Mr. R. D. Royce. Many of the herbaceous species were only seen in winter and spring; though recorded herein as “annual”, some of them could in fact be perennial in the strict sense of the word. POTAMOGETONACEAE Cymadocea antarctica (Labill.) Endl. and Posidonia australis Hook. f. Both these sea- weeds were found cast up on the coast of West Wallabi*, and the first was also found on North I*. SCHEUCHZERIACEAE Triglochin mucronata R. Br. Small annual. North I.* (banks of salt-lake). West Wallabi^ (damp flats). T. trichophora Nees. Small annual. North I.*, East Wallabi*. r. muelleri Buch. Small annual. East V/allabi* GRAMINEAE Setaria dielsii Herm. Annual. North I.*, V/est Wallabi*, Pigeon I.*. 4 Spinifex longifolius R. Br. Perennial. Coastal sand: North I.*, East Wallabi*, West Wallabi*. *Phalaris minor Retz. Annual. North I.* (near fishermen’s camp), West Wallabi (mutton-bird rookeries) . Stipa elegayitissima. Labill. Perennial. East Wallabi (consolidated dunes), West Wallabi*, Pigeon I. S. variabilis Hughes. Perennial. North I., East Wallabi* (pavement), West Wallabi* (rare), Tattler I. S. crinita Gaud. North I.*. Sporobolus virginicus (L. ) Kunth. Perennial. North I.* (damp soil south of salt-lake), West Wallabi* (shingly sea-sprayed beach on south coast) . Agrostis avenacea Gmel. East Wallabi*, West Wallabi*. Polypogon monspeliensis (L.) Desf. Annual West Wallabi*. P. tenellus R. Br. Annual. East Wallabi*, West Wallabi*. *Aveyia fatua L. Annual. North I.*. Danthonia caespitosa Gaud. Perennial. North I.*, East Wallabi*. ^Koeleria phleoides Pers. Annual. North I.*. *Viilpia myuros (L.) Gmel. Annual. North I.*, East Wallabi*. W”est Wallabi* (mutton-bird rookeries). Eragrostis dielsii Pilger. North I.*. Poa caespitosa Forst. Perennial. East Wal- labi* (sand dunes). ^Bromiis m.adritensis L. Annual. North I.*. B. arenarius. Labill. Annual. North I.*, East Wallabi* (pavement), West Wallabi* (north- east coast and mutton-bird rookeries). *S. molliformis Lloyd. Annual. North I.*. "^Cynodon dactylon Rich. Perennial. North I.*, (around fishermen’s huts). "^Hordeum leporinum Link. Annual. North I.*. CYPERACEAE Scirpus antarcticus L. Small annual. East Wallabi*, West Wallabi*. JUNCACEAE Juncus hufonius L. Small annual. North I.*, East Wallabi (stony gully in eastern dunes). LILIACEAE Anguillaria dioica R. Br. Annual. North I.*, East Wallabi*. West Wallabi*. Bulbine semibarhata (R. Br.) Haw. Annual. North I.* (shallow soil over limestone). East Wallabi* (consolidated dune). West Wallabi* I pavement and shell grit). Thysanctus patersoni R. Br. Perennial twiner. North I.*, East Wallabi*, West Wallabi* (pave- ment) , Pigeon I. Dianella revoluta R. Br. Perennial. East Wal- labi* (consolidated dune), West Wallabi and Pigeon I. (pavement). Acanthocarpus preissii Lehm. Low sprawling shrub. Pavement, consolidated dune, shallow sand over limestone: North I.*, East Wallabi, West Wallabi, Pigeon I. ORCHIDACEAE Microtis unifolia (Forst.) Reichb. Annual. East Wallabi* (pavement). URTICACEAE *Urtica urens L. Annual. West Wallabi* (mutton-bird rookeries). Parietaria debilis G. Forst. Small annual. North I.*, East Wallabi* (consolidated dunes), West Wallabi* (mutton-bird rookeries). PROTEACEAE Grevillea argyrophylla Meissn. Shrub. East Wallabi* (consolidated dunes), West Wallabi* and Pigeon I. (pavement). SANTALACEAE Exocarpus spartea R. Br. Shrub. North I.* (sand dunes). E. aphylla R. Br. Shrub. East Wallabi* (con- solidated dunes), West Wallabi* and Tattler I. (pavement). CHENOPODIACEAE Rhagodia baccata (Labill.) Moq. Succulent shrub. Shell grit, consolidated dunes, pavement: North I.*, East Wallabi*, West Wallabi*, Tattler 1., Pigeon I. Chenopodivm nitrariaceum (F. v. M.) Benth. Low succulent shrub. East Wallabi*. Ch. carinatum R. Br. Annual. East Wallabi* (gull rookery at Fish Point). *Ch. murale L. Annual. North I.* (? sp.). West Wallabi* (mutton-bird rookeries and guano), Pelican I. Ch. plantaginellum (F. v. M.) Aellen. Annual. East Wallabi* (gull rookery at Fish Point). Atriplex paludosa R. Br. Low succulent shrub. Abundant on shell beds, shell grit, shingles: North I.*, West Wallabi*, Pelican I., Mangrove 1.. Tattler I., Pigeon I., Long I. A. cinerea Poir. Low succulent shrub. North I.* and East Wallabi (fore-dune). Enchylaena tomentosa R. Br. Low succulent shrub. North I.*, Pelican I., Mangrove I., Pigeon I. Threlkeldia diffusa R. Br. Small ascending succulent shrub. Coastal sand, shingles and rock, and inland on shell grit: North I.*, East Wallabi*, West Wallabi*, Tattler I., Pigeon I. Arthrocnemum arbuscula R. Br. Succulent shrub. North I.* (muddy floor of sink-hole, and banks of salt-lake), East Wallabi (coastal shell beds). West Wallabi* (shell grit and especially banks of guano). Pigeon I. A. halocnemoides Nees. Succulent shrub. Shallow shell grit over limestone: North I.*, East Wallabi, West Wallabi*. Salicornia australis Banks &: Sol. Small flaccid succulent shrub. North I.* (damp mud in sink- hole, and low-lying flat south of salt-lake), West Wallabi (edge of tidal creek), Tattler I. (beneath mangroves). Suaeda australis (R. Br.) Moq. Small flaccid succulent shrub. North I.* (with Salicornia and Sporobolus in damp soil south of salt-lake). Tattler I. (beneath mangroves). Salsola kali L. Annual. North I.* (leeward slopes of east dunes), West Wallabi (beach). 5 AMARANTHACEAE Trichinium obovatum Gaud. Small shrub. East Wallabi* (pavement). T. divaricatum Gaud. Shrub. East Wallabi^' (consolidated dunes). T. eriotrichum (W. V. Pitzg.) C. A. Gardn. Climbing shrub. East Wallabi* (consolidated dunes). West Wallabi* (pavement). Ptilotus villosifiorus P. v. M. Annual. West Wallabi*. AIZOACEAE *Cryophytum crystallinuvi (L.) N. E. Br. Prostrate succulent annual. North I.* (banks of salt-lake). West Wallabi (guano workings). Pelican I. (abundant), Tattler I., Pigeon I. Carpobrctus aequilaterus (Haw.) N. E. Br. Prostrate succulent perennial. North I.*, East Wallabi (consolidated dunes), West Wallabi (pavement, especially near coast). Tattler I, Pigeon I. White flowers as well as red on North and Tattler Islands. PORTULACACEAE Calandrinia calyptrata Hook. f. Small succu- lent annual. North I.*. CARYOPHYLLACEAE *Stellaria media (L. ) Vill. Small annual. West Wallabi* (mutton-bird rookeries) . *Sagina apetala L. Small annual. East Wal- labi*. ’^Spergularia rubra (L.) J. & C. Presl. Small annual. North I.*. East Wallabi*, West Wallabi*. Pigeon I. RANUNCULACEAE Ranunculus parviflorus L. Small annual. East Wallabi*. CAPPARIDACEAE Capparis spinosa L. Trailing shrub. Pavement and consolidated dune: East Wallabi*. West Wallabi*, Tattler I., Pigeon I. CRUCIFERAE Hymenolobus procumbens (L.) Nuttall. Small annual. North I.*, East Wallabi*, West Wallabi*. Cakile maritima Scop. Annual. North I.* (beaches and blown-out dunes), East Wallabi. CRASSULACEAE Crassula colorata (Nees) Ostenf. Small an- nual. North I.*, East Wallabi* (sand dunes), West Wallabi*. PITTOSPORACEAE Pittosporum phillyreoides DC. Tall shrub. Pavement and consolidated dunes: East Wal- labi*, West Wallabi*, Tattler I., Pigeon I. CUNONIACEAE Aphanopetalum clematideum (Drumm. & Harv.) C. A. Gardn. Shrub. East Wallabi*. LEGUMINOSAE Acacia bivenosa DC. Shrub. East Wallabi* (consolidated dunes). Mirbelia ramulosa (Benth.) C. A. Gardn. Shrub. East Wallabi* (consolidated dunes). Bossiaea rufa R. Br. var. foliosa Benth. Shrub. East Wallabi* (consolidated dunes). GERANIACEAE Erodium cygnorum Nees. Annual. North I.*. East Wallabi*, Pigeon I. (? sp.). *E. cicutarium (L.) L’Her. Annual. East Wal- labi* (consolidated dunes). West Wallabi* (mut- ton-bird rookeries). Pelargonium littorale Hugel. East Wallabi*. OXALIDACEAE Oxalis corniculata L. Small annual. East Wallabi* (consolidated dunes ) , West Wallabi* . ZYGOPHYLLACEAE Zygophyllum apiculatum F. v M. Small an- nual. North I.*, East Wallabi*, West Wallabi*. Nitraria schoberi L. Succulent shrub, some- times tall. Coastal: North I.*, East Wallabi, West Wallabi, Pelican I., Mangrove I., Tattler I., Pigeon I., Long I., Beacon I. RUTACEAE Diplolaena dampieri Desf. Shrub. East Wal- labi* (pavement, consolidated dunes, shell grit). West Wallabi* and Pigeon I. (pavement). EUPHORBIACEAE Phyllanthus calycinus Labill. Small shrub. East Wallabi* (consolidated dunes). West Wal- labi. Euphorbia drummondii Boiss. Small annual. East Wallabi*. E. clutioides (Porst.) C. A. Gardn. Annual. North I.* (inner slope of western dunes). Beyeria viscosa (Labill.) Miq. Shrub. East Wallabi* (pavement and consolidated dunes). West Wallabi* (pavement). STACKHOUSIACEAE Stackhousia viminea Sm. Perennial herb. North I.*, East Wallabi*, West Wallabi*. SAPINDACEAE Dodcnaea aptera Miq. Shrub. North I.*, East Wallabi* (consolidated dunes and pavement). West Wallabi* (pavement). D. inaequifolia Turcz. Shrub. East Wallabi* and West Wallabi’" (pavement). RHAMNACEAE Spyridium globulosum (Labill.) Benth. Shrub, sometimes tall. Shallow sand over limestone, consolidated dunes and pavement: North I.*, East Wallabi. West Wallabi*. MALVACEAE Lavatera plebeja Sims. Tall perennial herb. Pelican I. (single plant). "^Malva parviftora L. Annual. North I., East Wallabi*, Pigeon I.*. STERCULIACEAE Lasiopetalum angustifolium W. V. Pitzg. Shrub. East Wallabi* (consolidated dunes). DILLENIACEAE Hibbertia subvaginata Steud. Small shrub. East Wallabi*, West Wallabi* (pavement). FRANKENIACEAE Frankenia paucifiora DC. Small shrub. North I.* (shallow soil over limestone, especially at inner foot of western dunes). East Wallabi* (consolidated dunes). West Wallabi* (stony country near coast and shell grit, especially where it contacts pavement), Tattler I.. Pigeon I. 6 THYMELAEACEAE Pimelea microcephala R. Br. Shrub. North I.* (shallow sand over limestone; rare), East Wal- labi* (consolidated dunes), West Wallabi* (pavement), Tattler I., Pigeon I. MYRTACEAE Eucalyptus oraria L. A. S. Johnson. Small tree. East Wallabi* (consolidated dunes). Restricted to a single valley in the eastern dunes between Flag and Eagle Hills. A specimen col- lected by Dr. D. L. Serventy in 1945 was identi- fied by Mr. C. A. Gardner (1949) as E. gracilis, a species otherwise known only from the semi- arid woodlands of the southern interior of the State. In F. gracilis the bark is deciduous, whereas all but the smallest branches of the East Wallabi trees are covered with rough grey bark. The trees were beginning to flower on April 20, 1959. HALORAGACEAE Haloragis trigonocarpa F. v. M. Perennial herb. East Wallabi=^ UMBELLIFERAE Didiscus pilosus Benth. Annual. East Wal- labi*. Hydrocotyle diantha DC. Small annual. East W'allabi* (stony gully in eastern dunes), West Wallabi*. Apium australe Pet. -Thou. Small annual. East Wallabi* (consolidated dunes), West Wal- labi* (mutton-bird rookeries). Daucus glochidiatus (Labill.) Fisch., Mey. & Ave-Lall. Small annual. North*, East Wallabi* (consolidated dunes). West Wallabi*. EPACRIDACEAE Leucopogon insularis A. Cunn. Shrub. East Wallabi* (pavement and consolidated dunes). PRIMULACEAE *Anagallis femina Mill. Annual. West Wal- labi*. PLUMBAGINACEAE Livionium salicorniaceuvi (F. v. M. ) Low spreading shrub. North* (shallow soil over limestone). East Wallabi, (hollows in consoli- dated dunes' West Wallabi* /shell grit and guano banks', Tattler I., Pigeon I. GENTIANACEAE *Erythraea centaurium Pers. Annual. North I.* (shallow soil over limestone). East Wallabi*, West Wallabi* (pavement). APOCYNACEAE Alyxia buxifolia R. Br. Shrub. East Wallabi* (shell grit, pavement, consolidated dune). West Wallabi (limestone near coast). ASCLEPIADACEAE Sarcostemma australe R. Br. Sprawling suc- culent shrub. East Wallabi* (consolidated dunes). West Wallabi and Pigeon I. (pavement). BORRAGINACEAE Cynoglossum- australe R. Br. Annual. East Wallabi, West Wallabi*. Trichodesma zeylanicuvi (L.) R. Br. Tall an- nual. North I.*. VERBENACEAE Avicennia marina (Forsk.) Vierh. Small tree (mangrove). West Wallabi* (a few in salt creek and at foot of nearby islet), Tattler I. (good clump on northern sheltered side of islet). LABIATAE Westringia dampieri R. Br. Shrub. East Wal- labi* (pavement, consolidated dunes and, rarely, sand dunes). West Wallabi (limestone near coast) . SOLANACEAE Solanum nigrum L. Herb. East Wallabi* (con- solidated dunes and pavement, especially in and near wells), West Wallabi*. Nicotiana rotundifolia Lindl. Annual. North 1.*, East Wallabi*, Pigeon I.*. SCROPHULARIACEAE *Dischisma arenarium E. Mey. Small annual. West Wallabi (shell grit). MYOPORACEAE Eremophila glabra (R. Br.) Ostenf. Low. somewhat flaccid shrub. East Wallabi* (con- solidated dunes), West Wallabi (in a restricted area a little east of Horse Well). Mycporum insulare R. Br. Shrub, sometimes tail. North I.* (inner slope of eastern dunes). East Wallabi*. West Wallabi (pavement and sand near coast). Pigeon I., Long I. PLANTAGINACEAE Plantago varia R. Br. Annual. North L* (shallow soil over limestone), East Wallabi*. West Wallabi*. RUBIACEAE Galium sp. Small annual. East Wallabi*. GOODENIACEAE Scaevola crassifolia Labill. Shrub. North I.* (dunes). East Wallabi (sandy and consolidated dunes) . COMPOSITAE Brachycome iberidifolia Benth. Annual. East Wallabi*. B. ciliaris (Labill.) Less. Annual. North I.*, East Wallabi*. Vittadinia triloba (Gaud.) DC. Annual. East Wallabi*. Olearia axillaris (DC) F. v M. Shrub. North I.*, East Wallabi*, West Wallabi, Pigeon I. Gnaphalium luteo-album. L. Annual. North I.' East Wallabi*. Pcdosperma angustifolium Labill. Annual. North I.*, West Wallabi*. Calocephalus aeruoides (F. v. M.) Benth. An- nual. East Wallabi*, West Wallabi*. Gnaphalodes uUginosus A. Gray. Annual. North I.*, East Wallabi*. West Wallabi*. Senecio lautus Soland. Annual (or perennial in favourable situations). Mainly coastal. North I.*. East Wallabi*, West Wallabi, Pelican I.. Pigeon I. S. brachyglossus F. v. M. Annual. North I.*. East Wallabi*. Picris hieracioides L. Annual. East Wallabi*. *Sonchus oleraceus L. Annual. North I.*, East Wallabi. West Wallabi*, Pelican I., Tattler I.. Pigeon I. 7 Reptiles The Wallabi Islands have a surprisingly rich reptile fauna. Of the 18 species* definitely re- corded from the group, all occur on West Wallabi which has an area of only 2.3 square miles. Moreover, several species are represented by abnormally dense populations, viz. Egernia siokesi, Amphibolurus harbatus, Phyllurus viilii and Python spilotus. On the other hand, cer- tain species (notably the geckoes Gehyra varie- gata and Heteronota bynoei) are rare by main- land standards. Nevertheless, reptiles are clearly the dominant vertebrates in the group, and they have evidently profited from the scarcity of land birds (5 resident species). Apart from their ecological interest, Abrolhos reptiles are important taxonomically. for the types of seven species came from these islands. Knowledge of this fauna began in 1840 with the collections of HMS Beagle, two of whose officers are commemorated in Egernia stokesi and Heteronota bynoei. Two years later John Gilbert visited the Abrolhos, but unfortunately, as with the ‘Beagle’ collections, no record was kept of the islands on which his numerous speci- mens were obtained. Gilbert (1843) does men- tion a lizard on Pelsart Island that is un- doubtedly Egernia kingi. He also saw a Green Turtle in the Wallabi Group, and on most islands he found the remains of “Hawksbill Turtles.” The ornithologist A. J. Campbell visited the Abrolhos in December 1889 and recorded five species, only one of which {Python spilotus) explicitly came from the Wallabi Group (Camp- bell 1890). During a long visit in 1894-5, Otto Lipfert (a former preparator in the W.A. Museum) seems only to have collected Rhynchoelaps bertholdi, a single specimen from West Wallabi. The naturalists of the Percy Sladeri Trust Expedition to the Abrolhos in 1913 and 1915 were the first generally to record the precise locality of their specimens. They obtained ten species on the Wallabi Islands, eight of them constituting the first definite record for the group (Alexander 1922). An expedition from Harvard University spent a fortnight on the Wallabi Islands in October 1931. Their large collection of reptiles included six species that were new for the group (Love- ridge 1934). During four visits in 1959-60, my colleagues and I were only able to add another two species. It would seem, then, that the fauna was almost completely known. Yet in the British Museum (according to its catalogues) there are six more species from the Abrolhos: Diplodactylus vittatus Gray, Egernia whitei Lacepede, Tiliqua rugosa (Gray), Sphenomorphus richardsoni (Gray), Hemiergis quadrilineatus (Gray), and Deniscnia coronata (Schlegel), as well as two species of frog: Limnodynastes dorsalis (Gray) and Myobatrachus gouldi (Gray). The speci- men of Sphenomorphus richardsoni, collected by Bynoe, remains unique. All the others were collected by Gilbert. Perhaps most if not all of them actually came from the mainland; though it is possible that they were collected ^Recently an additional species was collected by tbe Aquinas College Expedition, viz. the skink, Rhodona nigriceps (Glauert). ill the imperfectly explored Pelsart and Easter Groups, from which only two species are cer- tainly known: Egernia kingi (Pelsart Island) and E. stokesi (Rat Island). However, more field work must be carried out on the Islands before these problematical species are disposed of. As none of them are localised, they are excluded from the following list. GEKKONIDAE Phyllurus milH Bory. East Wallabi. West Wallabi, Pigeon I. Abundant; especially under slabs of limestone. Commonly seen at night when out feeding on bare ground. Diplodactylus spinigerus Gray. West Wallabi. Moderately plentiful. Several were found walk- ing on the ground in daylight. Gilbert collected the two syntypes in the Abrolhos. Phyllodactylus ocellatus (Gray). West Wal- labi. Uncommon. The two syntypes of Diplodactylus bilineatus Gray (currently re- garded as synonomous with ocellatus) were obtained by Gilbert in the Abrolhos. Phyllodactylus marmoratus (Gray). East Wallabi, West Wallabi, Pelican I., Tattler I. Not common on the larger islands, but abundant under slabs of limestone on offshore islets, where it is the only gecko. My specimen from East Wallabi came from a hollow eucalypt log. Four of the five syntypes were collected by Gilbert in the Abrolhos. Heteronota bynoei Gray. West Wallabi. In contrast to their abundance on the mainland of Western Australia this and the following species are the rarest geckoes in the group. The Harvard Expedition collected one, and I got two under driftwood on a shingly beach. The type was obtained by Gilbert in the Abrol- hos. Gehyra variegata (Dumeril & Bibron). West Wallabi. The two I collected are the only specimens from the Abrolhos apart from one in the British Museum whose collector is un- known. PYGOPODIDAE Lialis burtoni Gray. East Wallabi, West Wallabi. Uncommon. This may be the ‘small grey snake’ seen by Alexander on East Wallabi and tentatively referred to Denisonia coronata. Delma fraseri Gray. West Wallabi. Alexander’s two specimens are the only record for the Abrolhos. AGAMIDAE Amphibolurus barbatus (Cuvier). East Wal- labi, West Wallabi. Abundant, especially in sandy country among clumps of Spinifex longi- folius. These lizards differ considerably from those on the mainland of Western Australia and were described by Loveridge as a distinct race. minimus. They are mainly distinguished on their smaller size, narrower head, and longer tail and hind-legs. SCINCIDAE Egernia stokesi (Gray). East Wallabi, West Wallabi, Tattler I.. Pigeon I. Abundant, especi- ally under slabs of limestone. Two of my East Wallabi series were taken from the hollow stems of dead shrubs. The 12 syntypes were collected in the Abrolhos, four of them by Gil- 8 belt, the others presumably by the 'Beagle’. Abrolhos specimens never attain the size of mainland animals. E. kingi (Gray). East Wallabi, West Wallabi, Pigeon I. Moderately plentiful, favouring Spini- fex longifolius and other coastal habitats. Also seen, but not captured, on Tattler Island, a f acre islet off the east coast of West Wallabi. The Abrolhos population probably merits sub- specific rank. Ctenotus lesueuri (Dumeril & Bibron). East Wallabi. West Wallabi. Plentiful in all habitats. The 14 specimens obtained by the Harvard Ex- pedition were misidentified by Loveridge as Egernia formosa Fry, a species that is restricted to the Kalgocrlie-Laverton region. Rhodona praepedita (Boulenger). East Wal- labi, West Wallabi. Uncommon. My East Wal- labi specimen was collected under leaf litter in the eucalypt thicket. Two specimens obtained a few days earlier on North Island were listed in Storr (1960) under Lygosoma praepeditum Boulenger, a commonly used synonym of lineata. Ablepharus bcutoni (Desjardin). East Wal- labi, West Wallabi. My specimens came from low cliffs of limestone, into the crevices of which they dart when disturbed. They belong to the race plagiocephalus Cocteau. A small skink seen on Tattler Island was possibly of this species. Ablepharus lineocellatus Dumeril & Bibron. East Wallabi, West Wallabi. Moderately plenti- ful in sandy country. This is the species refer- red to by Alexander as a Lygosoma with an ‘orange head and pink throat’. Ablepharus elegans (Gray). West Wallabi. The three specimens of the Harvard Expedi- tion are Che only record for the Abroholos. BOIDAE Pytho7i spilotus Lacepede. East Wallabi, West Wallabi. Common in all habitats. During the day they are usually found coiled up beneath a bush, though an occasional one may be seen moving around. At night they are more active. They probably hibernate in winter: at any rate none was seen during my visit in late June 1959. One of the specimens collected by the Harvard Expedition had just swallowed a young Egernia stokesi. J. Akerstrom, a fisherman living on West Wallabi. has seen young wallabies inside Carpet Snakes. ELAPIDAE Rhynchoelaps bertholdi (Jan). East Wallabi, West Wallabi. Uncommon. Usually found under slabs of limestone. Land Birds In contrast to the diversity of the herpeto- fauna, only five species of bird are certainly resident in the Wallabi Group. Two more species, the kestrel and cuckooshrike, are pos- sibly resident, and the cuckoo could be a regu- lar visitor in spring. Little has been learnt of the land birds since Gilbert’s visit in 1843. Later workers have paid much more attention to the sea birds, which nest in great num.bers on these islands. A paper devoted to marine and littoral species has been prepared for publication in the Emu. TURNICIDAE Turnix varia (Latham). The Painted Quail is common on East and West Wallabi. Also occurred on one of the Pigeon Islands (Alexan- der 1922'. Most of the West Wallabi population occurs in low dunes covered with Spinifex and in the more open parts of the saltbush and samphire flats. They are even found in the dense tangles of Nitraria covering the old guano workings north of Eagle Point; but they shun the large areas of pavement limestone. Much of their food IS obtained by scratching in the surface soil, the resultant scrapes being very charac- teristic of the areas they are abundant in. On loose sand, e.g., just above high-water mark, the scrapes tend to be circular depressions about 5 inches in diameter and 1-2 inches deep. Fur- ther Inland they are less deep and usually crescentic in shape and thus remarkably like the hoof-marks of horses. They are excavated by scratching 2-3 times with one foot, then half-rotating the body and scratching with the other foot. They also visit the fishermen’s camp at Pelican Point to feed on scraps. At night they sleep on the ground, usually in pairs side by side at the foot of a bush. Their eyes shine red in torch-light, and they can often be picked up by hand. The weight of six birds was, 61, 65, 65. 66, 67 and 70 g. The colour of the upper bill was dark bluish grey, the iris vermilion, and the legs deep chrome-yellow. On April 26, 1960, one was flushed from a nest south-east of the ‘Walls’. It was located among shrubs of Frankenia paudflora, which on West Wallabi dominates the narrow zone between shelly flats and limestone ridges (this ecotone constitutes a favourite feeding area). The nest was a saucer shaped depression nearly 4 inches across and I inch deep, lined with fine twigs, and sheltered from above by a Frankenia. The three white eggs were finely dotted with brown and had underlying grey spots. Fur- ther south (and a little inland from Rocky Bay) a nest was found that was protected from above by a sparse dead bush and a few wind- blown inflorescences of Spinifex. The three eggs were somewhat different in colour to the first clutch; they were pale brown, finely flecked with reddish and dark brown. That evening two chicks were found among Spinifex behind the fishermen’s huts at Pelican Point. Hall (1902) found eggs cn October 20, which indicates a long breeding season. The Abrolhos populations constitute an endemic race, scintillans Gould. The differences between it and the mainland race, stirlingi Ivlathews, are discussed by Alexander (1922). COLUMBIDAE Phaps elegans (Temminck). The Brush Bronzewing is common on East and West Wal- labi and Pigeon Island. On West Wallabi they occur in all kinds of vegetation, including (unlike the quail) the scrub growing on limestone. On April 21, 1959, an empty nest was found on top of a Spinifex tussock. The saucer-shaped platform of Spini- fex roots was 5 inches in diameter and .sheltered from above by a few blades of the grass. On 9 Fig. 3. — Location of Brush Bronzewing nest on Tattler Island. Height of cliff about five feet. April 26 of the following year a pigeon was flushed from two eggs in a relatively substantial nest in a dense bush of Myoporum insulare near East Well. On the same day a remarkable nest v;as found on Tattler Island: the two eggs had been laid on a few fine twigs placed in a slight depression in the limestone at the edge of a low cliff, the nest being sheltered from above by the dense foliage of a mangrove (see sketch, Figure 3). Since Alexander found a nest with a fresh egg as late as November, the breeding- season must be spread over many months. Four birds . The samples of April 22 were also tested for content of crude protein. The following values (% dry weight) were kindly supplied by the Government Chemical Laboratories, and for comparison Rottnest estimates for that date are again given in brackets where available. Nitraria schoberi .. . 19.2 (13..5) Capparis spinosa 18.8 Arthrocnemum halocnemoides .. 13.9 ( 6.0) Pimelea microcephala 11.2 Arthrocnemum arbuscula 11.0 ( 0.0) Atriplex paludosa 10.3 ( 9.7) Myoporum insulare 8.3 ( 7.4) Diplolaena dampieri ^ 8.1 ( 5.8) Beyeria viscosa ^ 6.7 Olearia axillaris 6.3 ( 6.7) Sarcostemma australe 5.6 Pittosporuni phillyreoides 4.2 ( 5.0) Grevillea argyrodendron 3.8 Although practically nothing is known of the relative amount of the various plants eaten by the Tammars, some inferences on their diet can be made from the above data. The water content of the vegetation, presumably, as on Rottnest, is lowest in late summer and highest in late winter. The disparity in this respect between summer and winter diets will be even greater than is apparent from Table 1, when it is recalled that annual herbage is available in winter and spring. Generally the water content averages somewhat lower than on Rott- nest. This is especially so in the succulent species, and is a consequence of the fact that the Abrolhos only receive half as much rain at Rottnest. Table 1 Water content (% wet weight) of West Wallabi plants. Estimates for Rottnest in brackets April 22 June 25 Sept. 11 Arthrocnemum arbuscula 69 76 79 Arthrocnemum halocnemoides 77 (79) — (83) — (90) Atriplex paludosa 56 (75) 77 (78) 79 (85) Beyeria viscosa 38 Capparis spinosa 75 76 78 Diplolaena dampieri 47 (56) 67 (68) 67 (68) Grevillea argyrodendron 49 43 58 Myoporum insulare 69 (68) 79 (75) 81 (78) Nitraria schoberi 81 (83) 84 (91) 82 (90) Olearia axillaris 45 (41) 61 (58) 62 (53) Pimelea microcephala 62 78 71 Pittosporum phillyreoides 59 (57) 60 (58) 60 (61) Rhagodia baccata — (75) — (81) 78 (88) Sarcostemma australe 77 88 87 Threlkeldia diffusa 69 (81) — (91) 83 (89) With minor exceptions, protein levels are con- siderably higher on West Wallabi than Rott- nest, due no doubt to the abundance of nitro- gen in the guano-rich soil. Probably as on Rott- nest, protein levels are at their lowest in April, rising like the water content to maxima in late winter. And as for Rottnest Quokkas, the few weeks before the break of season would be most critical for the Tammars. The relatively high level of protein in the vegetation should help tide them over this period. The average excess of protein in West Wallabi plants over Rott- nest plants of the same species was 27% on April 22, at a time when the protein con- tent of the diet of Rottnest Quokkas varies locally between 6 and 9%- (Storr, 1964). The breeding season seems to be much the same as in Rottnest Quokkas, which is under- standable in view of the similar distribution of rainfall. All of four females examined in late April 1959 had a naked joey in the pouch. The largest of these was 23 cm long; two of the others were almost as large, but the smallest was only 2 cm long. In the following Septem- ber many young were seen at heel. According to the fisherman on West Wallabi. the juveniles are preyed on by Carpet Snakes (Python spilotus) and Sea-Eagles ' Haliaeetus leucogaster ) . This predation does not seem to be sufficiently intense to affect their numbers. In the past, human predation was at times much heavier. For example, Stokes and two com- panions shot 76 on West Wallabi in four hours (after which he aptly named the easternmost extremity of the island Slaughter Point). His colleague Surgeon Bynoe, in a curious account of marsupial reproduction, mentions examining the uteri of “between two and three hundred” wallabies (Stokes 1846, p. 156). It is little won- der that Gilbert (1843) found the animals more abundant on East than West Wallabi two years after the visit of the ‘Beagle’. In a recent unpublished study, J. Kelsall (pers. comm.) found the Abrolhos wallabies to differ in several characters from mainland animals. The name, Iwutmanni Gould, is avail- able for the Abrolhos race. MURIDAE Rattus glauerti Thomas. In November 1907. C. P. Conigrave collected a young adult male rat in the sand dunes of East Wallabi. It was later 11 sent to the British Museum, where Oldfield Thomas (1926) made it the type of a new species. Thomas believed it was related to R. fuscipes Waterhouse of the south-western mainland, as did Iredale and Troughton (1934), who regarded it as a subspecies of fuscipes. Tate (1951) dis- sociated glauerti from fuscipes, suggesting rather that it was a recent introduction of a rat of the Malayo-Polynesian concolor-exulans group. In April 1959 the writer collected three adult rats on West Wallabi, one of which was a female with 10 nipples (4 pectoral, 6 abdominal). As the mammary formula in the concolor-exulans group is consistently 4 + 4 (Tate 1936 and 1951), the Abrolhos rat clearly does not belong there. Despite its belated discovery (compared to that of the Tamrnar), there is no reason for doubting that glauerti is truly indigenous, and that it is an insular race of one or another of the southern Australian species of rats. As the following description shows, glauerti is most like R. greyi Gray. The external measurements of the rats (Nos. 1-3) are given in Table 2. All three had Similar pelage: long soft hair with dark bluish grey bases, tipped on the back with reddish brown and below with whitish. The short hair of the hands, feet and muzzle was whitish. The tail was uniformly dark and had 13 or 14 scales per cm. The caudal hairs were black and twice as long as the scales. The skull prepared from specimen No. 1 measured as follows: total length 36.2, zygomatic breadth 17.1, inter-orbital breadth 5.3, cranial breadth 15.8, interparietal 10.2, nasals 11.0 X 3.6, zygomatic plate 3.2, palatal foramina 6.2 X 1.7, bulla 6.8, molar crowns 5.8, m^ 2.9 X 1.8. The rostrum was broad and flat, the braincase rounded and unridged, and the bullae were rounded. On the following June 22 another five adults were caught on West Wallabi: No. 4 ( <$ , weigh- ing 79.5 g). No. 5 (9. 99), No. 6 (5, 105), No. 7 (9, 82.5, and No. 8 (^, 81). On the evening of September 8, 1959, two adults were taken in break-back traps in the sand dunes behind Turtle Bay, East Wallabi: No. 9 (^, 82), and No. 10 (9, 67 — it had lost all but the tip of its tail). Few rats were seen on West Wallabi in September, and the only ones caught were two juvenile females early in the morning of the 12th: No. 11 (45) and No. 12 (47). On April 23-25, 1960, three adult males were caught on West Wallabi: No. 13 (not weighed). No. 14 (70) and No. 15 (81); the testes of the last two were greatly enlarged. Specimens No. 1-12 have been lodged in the Western Australian Museum. As specimens No. 13-15 are no longer available in Western Australia, their measurements are given in Table 2 (I measured No. 13 and J. Kelsall measured No. 14 and 15). Immediately after capture, the pelage of No. 14 was described as fellows: dorsal and lateral hairs greyish brown, tipped reddish brown; guard hairs blackish: hands and feet whitish; underneath white. Most of the rats were caught in sandy country well vegetated with Spinifex longifolius, a few were among saltbush on the flats of shell Table 2 External measurements (mm) of six Rattus glauerti from West Wallabi Field Number Sex Head ~'r Body Tail Pes Ear 1 127 120 26.5 17 2 d 129 113 27 18 3 9 131 123 26 18 13 104 117 26 18 14 cT 134 104 26 20 15 d" 141 107 27 19.5 grit, and only one was collected on pavement limestone (it was disturbed in the daytime from a heap of stones beside East Well). Though an occasional rat may be seen in the daytime, they are really active only at night. With a head-torch their red eyes can be picked up whenever they come out to feed on the bare ground between patches of vegetation. Be- wildered by the light, they are easily run down and captured. They readily bite if given the opportunity. Apparently they spend the day hidden in the bases of shrubs and tussock grasses. They certainly do not burrow. The rats were so numerous on both islands in 1959-60 that I cannot understand why the species was not discovered before 1907. Perhaps their numbers undergo large fluctuations and my visits happened to coincide with a popula- tion peak. With so many fishermen resident in the Wallabi Group during the crayfishing season, the future of this beautiful little rat is very insecure. A few feral cats could soon exterminate it. OTARIIDAE Neophoca cinerea (Peron & Lesueur). The Hair Seal was formerly abundant, even though the Abrolhos are towards the northern limit of its Western Australian range. In his account of the Wallabi Islands, Gilbert (1843) wrote, “all the islands in this group are thickly inhabited by the seal; we would frequently come upon groups of 7, 8, and 9, lying asleep on the sandy beach; they are very easily captured, by walking steadily up to them with a club, when a single blow on the nose brings them down.” In 1889 Campbell observed that they were “principally found in the Easter and Pelsart groups but now getting scarce.” I saw none during my four visits, though one was missed by only three days; it followed fisherman G. Klee in his boat for a short dis- tance past Pelican Point on April 16, 1959. In the following September I found the remains of an adult male among the mangroves off the south-east coast of West Wallabi. It had evi- dently been dead for a month or more. Its skull was brought back and later lodged in the Western Australian Museum. This magnificent seal shows little fear of man, which has resulted in its decimation throughout the greater part of its range. Discussion In his summary of the vertebrate fauna of the Abrolhos, Alexander (1922) contrasted the wealth of the Wallabi Islands with the poverty 12 of the Easter and Pelsart Groups. He con- cluded that the land fauna of the Abrolhos had been derived from the mainland via the Wallabi Islands by transmarine dispersal. He also con- cluded that when the islands were populated with this fauna the opposite mainland must have been more like the extreme South-west than it is today. Neither of these hypotheses is tenable now. Alexander was evidently unaware of the general post-Pleistocene rise in sea-level, and that islands as distant from the mainland as the Abrolhos could nevertheless be only a few millennia in age. The separation of the Wallabi Islands from the mainland was dated by Main (1961) at 11,500 years ago. While some doubt may attach to the precise date of their separa- tion, other evidence confirms that the Wallabi Islands were formerly part of the mainland. Although the rocks underlying them are of marine origin, the dunes of the Wallabi Islands (and North Island) could only have formed when the area was continental. Admittedly, calcareous sand is produced on islands by the weathering of limestone; production, however, is generally outpaced by removal when sea- level is rising or stationary. Even on an island as large as Rottnest, dune and beach sand is continually being lost to the sea (Storr 1963). Moreover the fauna and flora of the Wallabi Islands are distinctly continental in their rich- ness and balance. It is too much to suppose that two species of mammal, two snakes, 16 lizards (in 12 genera and four families) and 114 vascular plants (in 94 genera and 50 families) have crossed Geelvink Channel in sufficient numbers to become established on these small islands. That many of these animals occur on other west coast islands is a reflection of their ability to survive on shrinking land masses rather than ability to cross the sea. A few species, e.g. Delma fraseri, Egernia stokesi and Rattus glauerti have not, to the writer’s knowledge, been found on other islands. Similarly with the flora, while there are many species common to a number of west coast islands, there are some that are not otherwise insular. In the latter category are the orchid Microtis unifolia and the shrubs Grevillea ar-gyrophylla, Tri- chiniuvi eriotrichum, T. divaricatum, Aphano- petalum cleviatideum, Bossiaea rufa, Hibbertia subvaginata and Sarcostemma australe. The greater part of the fauna and flora is widespread in coastal situations, especially those with calcareous rocks and soils. But none of the animals and few of the plants seem especially adapted for transmarine dispersal. Alexander’s second hypothesis was based on the belief that certain vertebrate animals oc- curred considerably further north on the Abrol- hos than on the west coast mainland. Alexander was not explicit as to the nature of the change on the mainland: it has been interpreted as one of “climatic deterioration” (Serventy and Whit- tell 1948, p.50). In the last few years naturalists have paid increasing attention to the mainland north of Perth, and the known range of most of the species cited by Alexander has been advanced. one by one, to the latitude of the Abrolhos and even beyond. Since Alexander’s time the known northern limit of the Brush Bronzewing has been extended from the Moore River successively to the Hill River (Storr and Ford 1959) and Freshwater Point, 25 miles south of Dongara (Ford 1960); the Painted Quail from the Moore River to the lower Murchison (Sedgwick and Morrison 1949) and to Peron Peninsula (Bath- gate, pers. comm.); the Spotless Crake from Perth successively to Yanchep (Serventy and Whittell 1948) and Hutt Lagoon (Ford 1962); and Egernia whitei (sensu lato) from 30 miles north of Perth to Eneabba (on the mainland* and to Bernier Island in Shark Bay (Douglas and Ride 1962). There only remain Denisonia coronata, Rattus fuscipes and Macropus eugenii. The known range of Denisonia coronata on the mainland has only been extended north from Perth to Gingin (Glauert 1957); but the supposed oc- currence of this species in the Abrolhos has never been confirmed. It is also irrelevant that Rattus fuscipes is still known only from the south coast, for the Abrolhos rat has been shown since Alexander’s time to belong to a different species (R. glauerti). The Tammar has not been collected on the mainland any further north than in Alexander’s time. There is, however, some evidence that it exists on the mainland opposite the Abrolhos. In October 1962, I had a fleeting glimpse of a small macropod in Acacia rostellifera scrub, 12 miles north of Balline; it did not appear to be a young Grey Kangaroo, the common species in that area. A few months later, Mr. Tom Pepper, of Tamala Station, told me that a “grey wallaby” lived in York Gum and other thickets on Lynton Station (especially the Port Arthur thicket), but he had not heard of them for twenty years. Recently Dr. W. D. L. Ride (pers. comm.) examined a rough skin of a wallaby that was shot five miles inland from Port Gregory: he believes that it was almost certainly from a Tammar. Alexander did not consider the flora of the Abrolhos. Had he done so, he might just as easily have arrived at the opposite conclusion, that the climate of Geraldton was formerly drier and warmer. Most Abrolhos plants extend well to the north and south of the islands. I know of none that has its northern limit in the islands: whereas at least two species (Capparis spinosa and Sarcostemma australe) are not known from so far south on the main- land coast. The fate of Alexander’s hypothesis is a sober- ing reminder that while it is fairly easy to dis- cover what is present on an island, it is much more difficult to ascertain what is absent from the mainland. Churchill (1960) and various members of the A.N.Z.A.A.3. Quaternary Shoreline Committee have cited evidence, from Australia and New Zealand, that 4-5000 years ago sea-level was 9-10 feet higher than now. At first sight the small islands immediately east of the Wallabi Islands fit in well with the concept of a pre- vious 9-10 foot higher sea-level. Their flat tops, 5-8 feet above present sea-level, are con- sistent with submarine planing. The richness of their fauna and flora, however, is quite in- consistent with any hypothesis that demands their submergence since they were initially severed from the main islands. Pigeon Island, for example, is only 400 yards long and 100 wide and is separated from East Wallabi by half a mile of sea. Yet its vegetation is not noticeably less diversified than any similar area of pavement limestone on East or West Wallabi. Its flora comprises 30 species, most of which are woody plants. It is inconceivable that so many shrubs like Cavvaris, Sarcostemvia and Grevillea could have become established on this island in 2-3000 years. It is possible that these islands were at least 7 feet higher 4,000 years ago and that falling sea-level has kept pace with their loss of eleva- tion. If the islands have been lowered by sur- face erosion, the pavement limestone on the Wallabi Islands should attain elevations of up to 15 feet where it is protected by overlying dunes. No such elevations have been observed, but then exposures are very limited in which the contact can be seen between pavement and dune limestone. At any rate surface erosion is probably a less potent force here than sub- terranean solution and the subsequent collapse of the caverns so produced. Alternatively the islands may have lost elevation through general subsidence of the continental shelf (in which case the age of the islands would be over- estimated). Whereas a 9-10 foot rise in sea-level would drastically reduce the area of East and West Wallabi and completely obliterate all the smal- ler islands, a two-foot rise would hardly effect the latter, margined as they are with vertical cliffs. It has already been conjectured that Shag Bay must formerly have penetrated deeply into the north-western part of West Wallabi. West of East Well the pavement limestone ter- minates abruptly in low cliffs that have every appearance of having formed under coastal conditions. Such a transgression of the sea would have required a rise in sea-level of no more than two feet. Acknowledgments My visits to the Abrolhos were primarily for marsupial studies and were financed by a grant from the C.S.I.R.O. to the Zoology Department of the University of Western Australia. Trans- port to the islands was generously provided by the Golden Gleam Fish Processing Co. and Mr. A. J. Fraser (Director of Fisheries). I am grateful to Mr. R. D. Royce (Chief Botanist, Government Herbarium) for the identification of plant specimens and to the fishermen on West Wallabi and Pigeon Island whose hospital- ity we enjoyed. References A.N.Z.A.A.S. Quaternary Shorelines Committee (1961). — Report to Brisbane Congress. Aust. J. Sci. 24; 121-124. (1964). — Report to the Canberra Congress. Aust. J. Sci. 26: 388-391. Alexander, W. B. (1922). — The vertebrate fauna of Hout- man’s Abrolhos (Abrolhos Islands), Western Australia. J. Linn. Soc. (Zool.) 34: 457-486. Campbell. A. J. (1890). — Notes on the zoology of Hout- man’s Abrolhos. Rep. Aust. Ass. Advance. Sci. 2: 492-496. Churchill. D. M. (I960).— Late Quaternary changes in the vegetation on Rottnest Island. W. Aust. Nat. 7: 160-166. Dakin. W. J. (1919). — The Percy Sladen Trust Expedi- tions to the Abrolhos Islands (Indian Ocean). Report I. J. Linn. Soc. (Zool.) 34: 127-180. Douglas, A. M. and W. D. L. Ride (1962).— Bernier and Dorre Islands. Reptiles. Fauna Bull. No. 2. (Fisheries Department. Western Austra- lia.) Ford, J. (1960). — The relationship between the avifauna of the Abrolhos Islands and the South-West. Emu 60: 284-285. .. (1962). — Northern extension of the ranges of the Spotless and Spotted Crakes. Emu 62: 61-62. Gardner, C. A. (1949). — Eucalyptus from Abrolhos Is- lands. W. Aust. Nat. 2: 47. Gilbert, J. (1843). — In a letter to the editor of the ‘Inquirer’ (Perth, April 19). Gillham, Mary E. (1963). — Association of nesting sea- birds and vegetation types on islands off Cape Leeuwin. South-western Australia W. Aust. Nat. 9: 29-46. Glauert, L. (1957). — “Snakes of Western Australia” (Western Australian Naturalists' Club: Perth.) Helms, R. (1902). — Houtman’s Abrolhos. J. Dept. Agric. W. Aust. 5: 33-55. Iredale, T. and Troughton, E. L. (1934). — A check-list of the mammals recorded from Australia. Mein. Aust. Mus. 6. Loveridge, A. (1934). — Australian reptiles in the Museum of Comparative Zoology, Cambridge, Massa- chusetts. Bull. Mus. Comp. Zool. 77 (6). Main. A. R. (1961). — The occurrence of Macropodidae on islands and its climatic and ecological implications. J. Roy. Soc. W. Aust. 44: 85- 89. Mayr, E. and R. Wolk. (1953). — The South-west Austra- lian races of the Spotted Scrub-Wren, Scri- cornis maculatus. W. Aust. Nat. 4: 66-70. Sedgwick, E. H. and P. C. Morrison (1949). — Observations on the Lower Murchison R.A.O.U. Camp. September 1948. Emu 48: 212-242 Serventy, D. L. and H. M. Whittell ( 1948).— “Birds of Western Australia” (Paterson Brokensha; Perth. ) Stokes, J. L. (1846). — “Discoveries in Australia". II. (Lond: Boone). Storr, G. M. (1960). — The physiography, vegetation and vertebrate fauna of North Island, Houtmar Abrolhos. J. Roy. Soc. W. Aust. 43: 59-62. (1963). — Some factors inducing change in the vegetation of Rottnest Island. W. Aust Nat. 9: 15-22. (1964). — Studies on marsupial nutrition. IV. Diet of the quokka, Setonix brachyurus (Quoy and Gaimard) on Rottnest Island. W'estern Australia. Aust. J. Biol. Sci. 17: 469-481. Storr, G. M. and J. R. Ford (1959). — Northern extension of known range of Brush Bronzewing. W. Aust. Nat. 7: 51 Tate, G. H. H. (1936). — Some Muridae of the Indo- Aus- tralian Region. Bull. Amer. Mus. Nat. Hist. 72: 501-728. (1951). — Result of the Archbold Expeditions. No. 65. The rodents of Australia and New Guinea. Bull. Amer. Mus. Nat. Hist. 97: 183- 430. Teichert, C. (1947). — Contributions to the geology of Houtman’s Abrolhos. Western Australia. Proc. Linn. Soc. N.S.W. 71: 145-196, and pis. 6-16. Thomas, O. (1926). — Two new Australian Muridae. Ann. Mag. Nat. Hist. (9) 18: 308-310. 14 2. — The Ocypode Ghost Crabs of Western Australia (Crustacea, Brachyura) By R. W. George and Mary E. Knott* Manuscript Accepted — 18th August, 1964. Four species of Ocypode are recorded in West- ern Australia. Two. O. ceratophthalma and O. cordimana are widespread Indo-west Pacific species occurring along the northern coasts of Australia; the other two are probably Australian endemics. O. convexa occurring on the west coast and O. fabricii on the northern and upper west coasts of Australia. The first male pleopod is figured and described for each species. A key and short descriptions are provided to enable adults and most sub- adults to be identified. Notes on the behaviour of the two probably endemic species are re- corded. ^ Introduction In recent years, active collecting has produced much crustacean material from Western Aus- tralia; perusal of the previous literature reveals inadequate comparisons of local and circum- Indian Ocean material in many instances. This report deals with the genus Ocypode, pointing out several previous misidentifications in the literature. Full bibliographic treatment has been attempted for O. fabricii and O. convexa but not for O. cordimana and O. ceratophthalma. Catalogued specimens now housed in the Western Australian Museum are indicated by the abbreviation WAM preceding the catalogue number and the measurements given in the text are those of the length of the carapace. Acknowledgments We wish to record our appreciation to Dr. Isabella Gordon of the British Museum (Natural History), who kindly examined specimens of Ocypode identified as O. kuhlii in the collection of that Museum; to Dr. L. Bott of the Sencken- berg Museum and Dr. Daniele Guinot of the Museum National D’Histoire Naturelle who kindly made comparisons of our material with the holotypes of O. nobilii and O. fabricii re- spectively; to Dr. L. B. Holthuis of the Leiden Museum and Dr. J. Yaldwyn of the Australian Museum for photocopies of original descriptions: to Dr. F. H. Talbot of the South African Museum for the loan of specimens identified as O. kuhlii from East Africa: to Dr. A. Crosnier of the Centre D’Oceanographie de Pointe-Noire for his advice; and to the many collectors of material dealt with in this paper, particularly those who made special efforts to secure crabs for us. Mrs. Marina Tyndale-Biscoe and Miss Sue Woods gave valuable technical assistance in the initial and final stages of the work. Mrs A. Neumann’s assistance in translating certain French and German texts is gratefully acknow- ledged. •Western Australian Museum, Beaufort Street, Perth. First Male Pleopods The general shapes of the first male pleopods of the four species of Ocypode examined are similar, consisting of a straight or slightly curved main shaft closely applied to the convex thorax in vivo. The term “upper surface” re- ferred to in the text is that side of the pleopod exposed when the abdomen is pulled back. Descriptions were facilitated by removal of the entire pleopod followed by low power micro- scopic examination. When viewed directly above the upper sur~ face, the pleopod of some species, e.g. O. ceratophthalma (fig. lA). can be easily sub- divided into main shaft, constricted “neck” and expanded “head”. The head when viewed obliquely or laterally is expanded for all species examined. The head of each pleopod curves outward to terminate in a hard chitinous tip- The main shaft of all species is covered with sparse pubescence. Fig. 1. — First male pleopods of Ocypode. Actual pleopod lengths given: A. B. O. ceratophthalma (18.1 mm); C. O. cordimana (14.9 mm); D. O. fabricii (15.8 mm); E. O. convexa (16.9 mm). A, viewed directly above upper surface, pubescence omitted; B C D E. head viewed obliquely from upper outer aspect, m.s. = main shaft, n. = neck, h. = head. v. = vibrissae. Drawn by Miss R. Hunt. 15 The diagnostic features are: the shape of the distal expansion of the lower surface ridge, the arrangement of the vibrissae on the head, and the form and position of the chitinous tip surrounding the distal aperture. The figures are of the head of the left pleopod viewed obliquely from the upper, outer aspect. Measurements of total pleopod lengths are given in the figure legends. Ocypode Weber 1795 Type species. — Cancer ceratophthalmus Pallas, 1772. The diagnostic features of the genus are fully described by Alcock (1900, p. 344). The spell- ings Ocypode and Ocypoda have both been used by previous workers for this genus and here the spelling Ocypode is adopted in view of the recommendations made to the International Commission on Zoological Nomenclature by Hol- thuis (1962, pp. 237, 244 and 245). Representatives of this genus occur on the Tropic and Subtropic coasts of all the major oceans of the world and the largest number of species have evolved in the Indo-west Pacific region. Although we have not attempted a complete review of the genus, it would appear that there are 14 species in the Indo-west Pacific, three in the Atlantic and two in the East Pacific regions. Current research on this genus by the French workers, Drs. Guinot and Cros- nier, will no donbt help to clarify the taxonomy of the group. Major revisions of this genus have previously been given by such workers as Kingsley (1880), Miers (1882) and Ortmann (1894 and 1897); and valuable contributions to the Indo-pacific Ocypode species have been published by Alcock (1900) and Tesch (1918). Most of these workers recorded variations in some characters (par- ticularly length of ocular stylet and shape of outer orbital angle) yet these variable charac- ters have sometimes been used in the construc- tion of their identification keys. This has led to some misidentifications, particularly of sub- adult and juvenile specimens. The diagnoses and the key given here are designed to facilitate the identification of adult and most subadult specimens of the four West- ern Australian species of Ocypode; only tenta- tive identifications are possible for extremely juvenile specimens. Close microscopic examination of these juven- iles revealed certain adult characters which could be relied upon, and some which could not. When examined microscopically (particularly if dry), the stridulating organ is possibly the best character on which to base an identifica- tion. Although it may be extremely difficult to accurately count the components of the juvenile stridulating organ, it is easy to see the actual type of structures of which it is composed. For example, the stridulating organ of juvenile O. ceratophthalma is obviously composed of two different types of structures, whereas in O. fahricii, the stridulating organ is composed of only one type of structure, transverse ridges; and these are obviously different from the ar- rangement of separate tubercles that forms the stridulating organ of O. convexa. In the juveniles, the leg hairbrushes are repre- sented by sparse lines of hair compared with the dense brushes developed in the adult. In the species from Western Australia, the general shape of the orbit and outer orbital angle vary little from juvenile to adult. In other species from other areas, e.g. O. saratan these vary with size, as discussed later. Key to the Western Australian species of Ocypode 1 . 2 . 3. Palm of large cheliped with stridulating organ 2 Palm of large cheliped without stridulating organ (1) Stridulating organ sole- ly or partly composed of transverse ridges .... 3 Stridulating organ sole- ly composed of round- ish tubercles (10-25); lower orbital edge with a deep lateral and a deep median notch (2) Stridulating organ com- posed of round and/or elongated tubercles in the upper part and transverse ridges in the lower: lower orbital edge with no lateral notch Stridulating organ com- posed solely of fine transverse ridges (108- 141); lower orbital edge with a broad, U-shaped lateral notch O. cordimana O. convexa O. ceratophthalma O. fahricii Ocypode cordimana Desmarest (Fig. 2A) Ocypode cordimana; Desmarest 1825, p. 121; Stimpson 1907, p. 110; Boone 1934, p. 191. Ocypode cordimana; Kingsley 1880, p. 185; Haswell 1882, p. 95; Miers 1882, p. 387; Alcock 1900, p. 349; Tesch 1918, p. 35; Sakai 1939, p. 613. Ocypode cordimanus; Barnard 1950, p. 84. Diagnostic features . — No stridulating organ cn palm of large cheliped. Lower orbital edge with a broad key-hole shaped lateral notch and a very slight median notch. Eyes without sty- lets. Outer orbital corners acute, directed for- wards. Anterior upper border of propodus of second and third pereiopod in male with single hairbrush: in female, single hairbrush on second pereiopod only. Inner dorsal margin of merus of large cheliped with curved, toothed flange. Inner margin of carpus of large cheliped with single main tooth and very small subsidiary teeth; outer distal margin of carpus granulate, without distinct teeth. Description of male pleopod (Fig. 1. C, WAM 27-63). — Viewed directly above upper surface, main shaft not expanded to form distinct head. Lower surface of shaft with ridge terminating in an expanded, smooth, oval lobe proximal to head. Tip bilobed, straight-edged, obliquely- directed. Upper and lower margins of outer surface of head each with a fringe of short vibrissae. Material examined Western Australia. — One male (28.3 mm) and one female (25.5 mm) East Montalivet Is. (west of Darwin), A. Whitworth on “DOROTHEA”, 21.x. 1962, WAM 27-63 and 140-63. 16 Fig, 2. — Male Ocypode from Western Australia. A. O. cordiviana WAM 27-63; B. O. ceratophthalma WAM 20-64; C. O. fahricii WAM 4-64; I). O. convexa WAM 24-63. Other localities. — Two immature males (10.2 and 11.6 mm» Ashmore Reef, Timor Sea, J. McIntyre, C.S.T.R.O.. August 1961, W.A.M. 28-63: four juveniles (4.0 to 8.2 mm) Marion Reef, Paget Island, Coral Sea. G. P. Mees, 2o.xl. 1961, WAM 76-63; two males (16.5 and 13.8 mm) south end Heron Island, Queensland, R. W. George. 23. v. 1961, WAM 74-63: one male (14.1 mm) Urunga, near Coff’s Harbour, New South Wales, Mrs. R. W. George, 1.x. 1959, WAM 75-63. Distribution. — “Mauritius, east coast of Africa to Red Sea, Indo-pacific to Japan” (Barnard), East Australia (Haswell); now recorded from Western Australia. The type locality is He de France (Mauritius) — see Boone 1934, p. 191. Ocypode ceratophthalma (Pallas) ( Fig. 2B ) ^ Cancer ceratophthahnus Pallas 1772, p. 83, pi. 5, figs. 7 and 8. Ocypoda ceratophthalma; Miers 1882, p. 379; Miors 1886, p. 238; Alcock 1900, p. 345; Tesch 1918, p. 36; Balss 1935. p. 140; Sakai 1939, p. 614; Gillett and Mc- Neill 1962, pi. 118. Ocypode ceratophthalma; Edmondson 1946, p. 310; Tweedie 1950, p. 321. (feeding behaviour). Ocypode ceratophthalmus; Barnard 1950, p. 86. Ocypode kuhlii; Miers 1882, p. 384 (part) and 1884, p. 237 (part). Diagnostic features. — Stridulating organ on palm of large cheliped consists of round and/or elongated granules in the upper part and fine transverse ridges in the lower, margined distally by sparse pubescence in females and a dense hairbrush in males. Lower orbital margin with- out median or lateral notches. In our series, specimens over 24 mm with distinct ocular sty- lets, length of the stylet usually in proportion with length of carapace; specimens less than 24 mm with rudimentary or no ocular stylet. At equivalent size, males have longer ocular stylets. Outei- orbital corners right-angled and outwardly directed. Anterior surface of pro- podus of second and third pereiopods of male with triple hairbrush; females with single hair- brush and few scattered hairs. Inner dorsal margin of merus of large cheliped almost straight, toothed; without flange. Inner margin of carpus of large cheliped with single main tooth (a number of subsidiary teeth develop in large specimens ) : outer distal margin of carpus denticulate. Descriptioji of male pleopod (Fig. 1. A, B, WAM 20-64). — Viewed directly above upper surface, main shaft constricted to form neck and expanded distally to form head. Lower surface of shaft with ridge terminating in small, smooth, acute projection proximal to head. Tip bilobed, almost straight-edged, longitudinally- directed. Outer margin of head with double fringe of long vibrissae. Material examined Western Australia. — Ten males (26.1-38.4 mm), thirteen females (24.3-36.5 mm) and six juveniles (6.7-19.0 mm) from fifteen separate localities between East Montalivet Is. and Dorre I., Shark Bay. WAM 66-55. 69-55, 33-63, 34-63, 78-63, 79-63, 81-63, 83-63, 85-63, 86-63, 88-63, 90-63, 91-63, 93-63, 94-63, 95-63, 98-63. 136-63, 143-63, 19-64. 20-64. Other localities. — Five males (26.9-35.2 mm), three females (29.8-33.0 mm) and two juveniles (8.2 and 12.8 mm) from the following localities; Penang; Christmas I. and Cocos Is., Indian Ocean; Heron I., Queensland; Port Macquarie and Forster, New South Wales. WAM 80-63, 82-63, 84-63. 87-63, 89-63, 92-63, 96-63. 17 Coviinents. — Miers (1882, p. 385) recorded "a series of rather smaller specimens” of O. kuhlii from Thursday Island. However, re-examina- tion of three of his specimens by Dr. Gordon at the British Museum and of another small speci- men by us from the Australian Museum, showed that they were all juvenile O. ceratophthalma. Further, Dr. Gordon examined the remaining material, recorded by Miers (1882, p. 385) as O. kuhlii (except the female from Japan which could not be located), and states (pers. comm.) '‘Although Miers seems to have adopted de Man’s characters for O. “kuhlW in his paper, the fact remains that not one of his specimens of supposed “kuhlii” agrees with de Man’s descrip- tion of the two specimens in the Leiden Museum. Until someone re-examines and figures O. “/cw/i/zz” holotype from Java, I cannot be sure that any material so-called in our collection belongs to that species.” The Shark Bay record of O. kuhlii by Miers 1882 is in fact O. convexa ( discussed later ) . Distribution. — “Mauritius, east coast of Africa to Red Sea, Indo-pacific” (Barnard). The type locality is India. Ocypode fabricii Milne Edwards (Fig. 20 Ocypoda fabricii Milne-Edwards 1837, p. 47 and 1853. p. 142. Ocypoda aegyptiaca; Balss 1935, p. 140. Non O. fabricii, Kingsley 1880. p. 182. Diagnostic features. — Stridulating organ on palm of large cheliped composed of 108-141 extremely fine transverse ridges, accompanied distally by a dense patch of pubescence in males and scattered hairs in females. Lower orbital edge with a broad median notch and a U-shaped lateral notch. Stylets on eyes very short. Outer orbital corners very acute, usually directed outwards. Anterior surface of propodus of second pereiopod with triple hair- brush (largest male 37.9 mm), double hairbrush (other males) or single hairbrush (females). Distal half of anterior surface of propodus of third pereiopod with single hairbrush (largest male) or sparse or absent (remainder of mate- rial). Inner dorsal margin of merus of large cheliped straight with regular denticles along proximal half, distal half with clusters of irre- gular teeth. Inner margin of carpus of large cheliped irregularly toothed with two (subadult) to nine teeth (female 29.8 mm). Descriptioii of male pleopod (Fig. 1. D, WAM 4_64. — Viewed directly above upper surface, main shaft expanded to form head but not markedly constricted to form neck. Lower sur- face with ridge terminating in a very low, smooth, bulbous expansion proximal to head. Tip trilobed, spooned, longitudinally-directed; distal aperture opens between the two proximal lobes. Outer margin of upper surface of head with two tufts of long vibrissae. Material examined. — One male (17.6 mm) East Montalivet Is., A. Whitworth on “DORO- THEA", 21.x. 1962, WAM 102-63; one female (16.6 mm) Derby, Driver Stuart. March 1945, WAM 277-45; two males (19.4 and 19.8 mm) Yampi Sound. G. A. Robinson. March 1959, WAM 103-63: tv/o juveniles (9.8 and 11.2 mm) Yampi Sound. G. A. Robinson, July/August 1960, WAM 105-63; one female (30.3 mm) Adele I.. W. Goode on “DOROTHEA”. 19.x. 1962. WAM 100-63; one female (recently moulted) (32.2mm) Crab Creek, Broome, M. C. MacDonald, 12.i.l962, WAM 138-63: one female (29.8 mm) probably Broome. E. J. Stuart. 1917. WAM 9391; one juvenile (11.0 mm) Legendre I., Dampier Archi- pelago, H. Williams and M. C. MacDonald. 10.vi.l962. WAM 97-63; two males (13.9 mm) three females (16.5 to 18.5 mm) one juvenile (8.5 mm) Dolphin I., Dampier Archipelago, R. D. Royce on “DAVENA”, 29.V.1960. WAM 30-63: one female (18.4 mm) four males (12.7 to 24.5) and one (abdomen damaged) (11.7 mm) be- tween east and west Lewis Is., Dampier Archi- pelago, B. R. Wilson on “DAVENA”. 12.vi.l960. WAM 104-63; one female (23.5 mm) Shark Bay. Mrs. J. Watson, May 1961, WAM 101-63; one male (29.4 mm) and one female (31.3 mm) Den- ham. Shark Bay. B. R. Wilson. 29.xii.1959, WAM 145-63 (donated to Paris Museum); one male (37.9 mm) Denham’s Hummock east side of Shark Bay, R. Slack-Smith, 9.1.1963, WAM 144-63; two males (32.8, 34.7 mm) and one female (33.1 mm) Little Lagoon, I 2 mile N. Den- ham, Shark Bay, D. Bathgate, 12.iii,1964, WAM 4-64, 18-64: one male (21.4 mm) unknown locality 9390: one ovigerous female (30.8 mm) unknown locality WAM 99-63 donated to Senck- enbevg Museum. Comments. — Milne Edwards (1837) gave only a very general description of this species, unac- companied by illustrations; and this species has remained in relative obscurity after Miers (1882) and Ortmann (1897) included O. fabricii with O. ceratophthalma. Their reasons for this procedure were based on an erroneous descrip- tion given by Kingsley (1880) whose specimens from Australia and Natal are not O. fabricii since they had tubercles in place of fine striae on the stridulating ridge. In view of the previous obscurity of O. fabricii. we initially believed that our material belonged to an unknown species so we are indebted to Dr. Crosnier for suggesting correctly, that our West- ern Australian material of this species might be O. fabricii. He has recently been studying Ocypode and other genera from Madagascar for which he examined closely the type of O. fabricii. On his suggestion, we sent several of our specimens to Dr. Daniele Guinot, Museum National D’Histoire Naturelle, Paris, for direct comparison with the holotype of O. fabricii and she reports that our specimens agree with the holotype from “Oceanie”. The precise locality of the holotype is not known but since our mate- rial and two specimens of this species which Dr. Guinot has seen from “west of Darwin", are all Australian in origin, it is likely that the holo- type was collected in the western part of Oceania, if not Australia. No specimens of this species were found in the collections of the British Museum or the Zoological Survey of India when one of us (R.W.G.) examined them in 1963; there were however some specimens from Northern Aus- tralia incorrectly identified as O. kuhlii in the Singapore Museum. On the basis of the number of striae in the stridulating organ O. fabricii closely resembles O. nobilii de Man, 1902. The stridulating ridge of the O. nobilii male holotyps (18 mm c.l.) is formed of approximately 115 transverse striae. A personal report by Nobili to de Man (1902, p. 481) records a large male (21 mm) with 120 transverse striae. An adult female (WAM 99- 63) and small males and females (WAM 30-63) of O fabricii were sent to Dr. Bott of the Senckenberg Museum for comparison with the holotype of O. nobilii. He regarded the holotype male of O. nobilii as an adult which clearly dif- fered from the specimens of O. fabricii sent for comparison. He pointed out that the most striking difference between the species, apart from the much greater size of O. fabricii, was the shape of the merus of the walking legs. The ratio of the length to breadth of the merus of pereiopod 4 in O. nobilii is 2.6 whereas in O. fabricii this ratio is 3.25. Another distinguish- ing feature is the shape of the lateral notch on the lower margin of the orbit; in O. nobilii it is shallow and obtuse but in O. fabricii it is deep and U-shaped. Balss (1935, p. 140) identified an adult speci- men from Shark Bay as O. aegyptiaca Ger- staecker 1856 and he probably used Ortmann’s (1897) key for his identification. Indeed all our specimens of O. fabricii key out closest to that species using Ortmann’s key. In addition, Balss regarded his Shark Bay specimen as conspecific with O. aegyptiaca specimens from the Red Sea; in our opinion, his Red Sea specimens were not adult, but subadult animals. Miers (1882, p. 382) noted that the shape of the orbital corners and the length of the ocular stylet of O. aegyp- tiaca vary with size and this observation was confirmed when one of us (R.W.G.) collected a series of the common Red Sea ghost crab from East Aden Protectorate in March 1963. This collection w'as made during a short term cray- fish assignment with the F.A.O. United Nations (FAO/UN 1963). Holthuis (1958) pointed out that Forskal’s (1775) name of O. saratan for the Red Sea ghost crab is valid and antedates O. aegyptiaca Gerstaecker (1856). The comparison of our East Aden series of O. saratan with local mate- rial is represented diagrammatically in Figure 3; there is little obvious difference between the subadult O. saratan (fig. 3. B, C) and adult O. fabricii from Shark Bay (fig. 3. D, E). The two species can be readily distinguished at most stages of growth by the 100 or more fine striae in the stridulating ridge of O. fabricii compared with 60-100 striae on the striciulating organ of O. saratan. The notched form of the outer orbital angles of O. saratan as figured by Laurie (1915, p. 467) is represented in four of our nine adults from East Aden and all our adults have the stylet recurved posteriorly, not anteriorly as shown by Laurie. It is also noteworthy that the juveniles of O. saratan from East Aden would be identi- fied as O. kuhlii using the key produced by Miers (1882). It is not unlikely that juveniles of this species and other species (see comments on O. ceratophthalmai have in the past been desig- nated O. kuhlii. Fig. 3. — Growth Changes of Ocypode. A-C; O. saratan^ East Aden Protectorate WAM 142-63. A male 42.2 mm. B female 22.5 mm. C female 18.6 mm; 1>-F: O. fabricii. Western Australia. D WAM 144-63 male 37.9 mm. E WAM 145-63 female 31.3 mm, F WAM 104-63 male 18.3 mm. Measurements are those of the carapace length. Distribution. — North and west coasts of Aus- tralia from East Montalivet Is. (west of Dar- win) to Shark Bay. The type locality is “Oceanie”. Ocypode convexa Quoy and Gaimard (Fig. 2D) Ocypode convexus Quoy and Gaimard 1824, p. 52.5. Ocypoda convexus; Kingsley 1880, p. 185. Ocypoda kuhlii; Miers 1882, p. 384 (part). Ocypode pygoides Ortmann 1894, p. 766 and 1897, p. 362; Montgomery 1931, p. 451 (sic pygioides); Ride 1962, p. 17. Ocypoda pygoides; Balss 1935, p. 140. non O. convexa; Stimpson 1858, p. 100 and 1907, p. 109 -- O. stimpsoni Ortmann 1897, p. 364. non O. convexa Nobill 1900. p. 518 = O. nobilii de Man 1902, p. 478. Diagnostic features. — Stridulating organ of 10 to 25 roundish tubercles. Lower orbital edge with two deep notches, one central and one lateral. Eyes without stylets. Outer orbital corner acute, directed forwards. Propodus of second and third pereiopods with single hair brush on upper anterior surface. Inner dorsal margin of merus of large cheliped with curved toothed flange. Inner margin of carpus of at least one cheliped with distinct bifid tooth. Description of male pleopod (Fig. 1. E, WAM 24-63). — Viewed directly above upper surface, main shaft not expanded to form distinct head. Lower surface of shaft with ridge terminating in a well-developed, rounded, pubescent lobe proxi- mal to head. Tip bilobed, straight-edged verti- cally-directed. Upper and lower outer margins of head each with fringe of short vibrissae. Material examined. — Twenty nine males (20.1- 42.5 mm), twenty-six females (18.7-38.4 mm) and twenty-seven juveniles (7.5-20.0 mm) from twenty-one West Australian localities from Bar- row I., and between North West Cape and Yal- lingup. WAM 7662, 7726, 9492, 9613, 9980, 10032, 10455/8, 128/9-46, 24/26-63, 49/73-63, 77-63, 106-63, 107-63, 137-63, 13-64. Comments . — There is little doubt that Ort- mann’s suspicion (1897, p. 362) that his name O. vyQOides is synonymous with O. convexa is true. Quoy and Gaimard’s original figure and short description of O. convexa agree with the material examined hei’e. Kingsley (1880, p. 185) and Miers (1882, p. 385) noted the features of the bifid tooth on the carpus of the cheliped and the hairbrush on the second and third pere- iopods; both these features are clearly seen in Quoy and Gaimard's figure (in the figure, the hairbrush on right second pereiopod has not been illustrated). Miers thought O. convexa might be referred to either O. kuhlii or O. cordi- mana but O. cordimana lacks a stridulating organ on the palm of the cheliped and accord- ing to de Man (1881, p. 252), O. kuhlii lacks the propodal hairbrushes. In 1960. Dr. Gordon re- examined the male from Shark Bay which Miers (1882) identified as O. kuhlii and concluded that it was an adult O. convexa. Distribution . — West coast of Western Austra- lia, from North West Cape to Yallingup and also Barrow I. The record by Balss (1935) from Bar- row Island is confirmed. The type locality is Dirk Hartogs Island, Shark Bay. General Observations and Behaviour The ghost crabs of Western Australia do not as a rule leave their sand burrows during the day; at night they forage along the strand lines of the beaches, retreating to their burrows or to the sea if disturbed. The burrows of all West- ern Australian species are near-ventical with simple openings; the excavated sand is carried away from the burrow mouth and dispersed. At Montalivet Is. three species, O. cordimana, O. ceratovhthalma and O. fabricii were dug from burrows on one beach and the burrows were not obviously different. However, it is likely that cn closer examination, differences in burrow position, shape or construction might be dis- covered. By comparison, the appearance of O. saratan in the East Aden Protectorate is the daylight feature of the beaches: their burrows are well marked by large volcano-shaped piles of sand alongside the burrow mouths. The crab carries out an “armful” of sand, climbs to the top of the pile and deposits a load there each time. During the winter of 1961, Mr. J. Brouwer studied a number of burrows of juvenile O. con- vexa at City Beach, near Perth, and he found that the entrances were all well above high water mark and 45-60 feet from the edge of the sea. The burrows were more or less vertical but all entrances were south of the bottom of the burrow, presumably to prevent sunlight from penetrating the burrow. Mr. Brouwer also recorded that these juveniles survived immersion in freshwater for 12 hours and on release were able to dig burrows again. At Bernier Is., Shark Bay, Ride (1962) reported adult O. convexa (as O. pygoides) foraging up to one-third of a mile inland following heavy rain. The ability of O. convexa to withstand immersion in freshwater and to actively forage in rain-soaked vegetation may also apply to ether species: perhaps de Man’s (1888, p. 108) record of O. cordimana in freshwater at Sulli- van I. and Mier’s (1886, p. 239) record of Ocypode sp. in freshwater at Fiji are similar examples. The following observations on O. fabricii by Mr. D. G. Bathgate at the Little Lagoon, 1^ miles north of Denham, Shark Bay, on the 12.iii.64 are also worthy of recording here. The official maximum shade temperature at Denham on that day was 114 deg. P. at 1300 hours. Mr. Bath- gate’s observations on the behaviour of the crabs from 0930 to 1630 hours were almost certainly in response to the very hot, dry conditions when the temperature and water loss of the crabs would be unusually high. The normal burrows along the beach at Little Lagoon were about 3 feet above the high water mark at 2 feet in depth but remarkably, they were not occupied by crabs during the above daylight observations: about 30 burrows were investigated by Mr. Bathgate and found to be empty. The crabs were actively engaged in digging temporary burrows closer to the water line. The burrows were only about 1 foot from the water line: the completed burrow ran obliquely a distance of about 6 inches, where it reached the water table. The most remarkable aspect of the crab’s behaviour was the fairly regular and very rapid visits to the water in the lagoon to completely submerge for about 2-3 minutes. They returned more slowly to the temporary burrow; there to continue digging, first at the burrow mouth and then deeper in the burrow. About five visits to the burrow were required to excavate the burrow to 6 inch depth; subsequent visits to the same burrow were continued until a total of about 20 visits were made (approximately 2h hours’ duration) after which the crab may com- mence digging at a new burrow site. On his return to the area in the cool of the night. Mr. Bathgate noted that the temporary burrows were not being used and were only faintly recognisable, as the rising tide had al- most erased them. The crabs at this time were behaving normally, foraging near the stand line and retreating, when disturbed, to their perma- nent burrows or to the waters of the lagoon. References Alcock, A. (1900). — Materials for a carcinological fauna of India. J. Asiat. Soc. Beng. 69: 343-350. Balss, H. (1935). — Brachyura of the Hamburg Museum Expedition to South West Australia, 1905. J. Roy. Soc. W. Aust. 21: 113-151. Barnard, K. H. (1950). — Descriptive catalogue of the South African decapod Crustacea. Ann. S. Afr. Mus. 38: 1-837. Boone, L. (1934). — Scientific results of the world cruises of the yacht Alva. Bull. Vanderbilt Oceanogr. (Mar.) Mus. 5: 1-210. Desmarest, G. A. (1825). — “Considerations generales sur la classe des Crustaces . . .“. (Paris). Edmondson, C. H. (1946). — “Reef and Shore Fauna of Hawaii*’ (Bernice P. Bishop Museum: Hon- olulu.) FAO/UN. (1963). — Report to the Government of Aden on the Crawfish Resources of Eastern Aden Protectorate. Based on the work of R. W. George. Rep. FAO/EPTA, (1696): 1-23. Forskal, P. (1775). — “Descriptiones Animalium”. (Hauniae.) 20 Gerstaecker, A. (1856). — Carcinologische Beitrage. Arch. Naturgesch. 22: 132-137. Gillett, K. and McNeill, F. (1962). — “The Great Barrier Reef and adjacent isles”. (Coral Press: Sydney.) Haswell, W. A. (1882). — “Catalogue of the Australian Stalk- and Sessile-Eyed Crustacea”. (Aus- tralian Museum: Sydney.) Holthuis, L. B. (1958). — Contributions to the knowledge of the Red Sea. No. 9. Bull. Minst. Agric. Israel Div. Fish. Haifa. No. 17: 41-54. (1962). — Forty-seven genera of Decapoda (Crustacea); Proposed Additions to the Official List. Z.N. (S.) 1499. Bull. Zool. Norn. 19: 232-253. Kingsley, J. S. (1880).— Carcinological Notes, No. 3— Revision of the genus Ocypoda. Proc. Acad. Nat. Sci. Philad. 1880: 179-186. Laurie, R. D. (1915). — Reports on the marine biology of Sudanese Red Sea. 21. On the Brachyura. J. Linn. Soc. (Zool.) 31: 407-475. Man, J. G. de (1881). — Carcinological Studies in the Leyden Museum No. 2. Notes Leyden Mus. 3: 245-254. (1888). — Report on the podophthalmous Crustacea of the Mergui Archipelago. J. Linn. Soc. (Zool.) 22: 1-312. (1902). — Die von Herrn Prof. Kukenthal im Indischen Archipel gesammelten Dekapoden und Stomatopoden. Ahh. Senchenb. Naturf. Ges. 25: 478-483. Miers, E. J. (1882). — On the Species of Ocypoda in the Collection of the British Museum. Ann. Mag. Nat. Hist. (5) 10: 376-388. (1884).— Crustacea. In “Report of the Zool- ogical Collections made in the Indo-Pac- ific Ocean during the Voyage of H.M.S. Alert. 1881-2”, Pt. 1. Pp. 178-322, 513-575. (Taylor and Francis: London.) (1886). — Report on the Brachyura collected by H.M.S. Challenger during the years 1873-76. Rep. Sci. Res. “Challenger” Voyage, 1873-76 Zool. 17.- (49). Milne Edwards, H. (1837). — “Histoire Naturelle de Crus- taces” 2 (Roret: Paris.) (1852). — Observations sur la classification naturelle des crustaces. Ann. Sci. Nat. (3) 18: 109-166. Montgomery, S. K. (1931). — Report on the Crustacea Brachyura of the Percy Sladen Trust Expe- dition to the Abrolhos Islands. J. Linn. Soc. (Zool.) 37: 405-465. Ortmann, A. E. (1894). — Die Decapoden-Krebse des Strassburger Museums. Zool. Jh. (Abt. I) 7: 761-772. (1897). — Carcinologische Studien. Zool. Jh. (Abt. I) 10: 258-372. Pallas, J. (1772). — “Spicelegia Zool”. Part 9. Quoy, J. R. C. and Gaimard, P. (1824). — “Voyage autour du Monde . . . sur les corvettes I’Uranie et la Physicienne. Zoologie”. (Paris.) Ride, W. D. L. (1962). — In Bernier and Dorre Islands. W. Aust. Fish. Dept., Fauna Bull. No. 2: 1-131. Sakai, T. (1939). — “Studies on the Crabs of Japan. IV. Brachygnatha, Brachyrhyncha”. (Yokendo: Tokyo.) Stimpson, W. (1858). — ;Prodromus descriptionis animal- ium evertebratorum. Part 5. Crustacea Ocy- podoidea. Proc. Acad. Nat. Sci. Philad. 1858: 93-110. (1907). — Report on the Crustacea (Brach- yura and Anomura) collected by the North Pacific Exploring Expedition, 1853-56. Smithson. Misc. Coll. 49: 1-240. Tesch, J. J. (1918). — The Decapod Brachyura of the Siboga Expedition 1. Hymenosomidae, Ret- roplumidae. Ocypodidae, Grapsidae and Gec- arcinidae. Siboga Exped. 39c: 1-148. Twcedie, M. W. F. (1950). — Notes on grapsoid crabs from the Raffles Museum. 2. On the habits of three Ocypodid crabs. Bull. Raffles Mus. No. 23: 321-24. Weber, F. (1795).— “Nomenclator Entomologicus Secun- dum Entomologiciam Systematicum III. Fabricii Adjectis Speciebus Recens Detectis et Varietabilis”. (not seen.) 21 3. — Two species of the extinct genus Sthenurus Owen (Marsupialia, Macro- podidae) from south-eastern Australia, including Sthenurus gilli sp. nov. by D. Merrilees* Manuscript Accepted 18th Axigust, 1964. A new species of Sthenurus, smaller than any so far known, is described from deposits near Strathdownie, Western Victoria, of presumed Pleistocene age. A variant of the same species occurs in Haystall Cave, Naracoorte, South Aus- tralia, and it may have ranged into Western A second (larger) species of Sthenurus, re- sembling S. occidentalis Glauert, occurs in the Strathdownie deposit. The Haystall Cave de- posit also contains a second species of Sthenurus resembling S. occidentalis. The taxonomic rela- tionships of the larger Strathdownie species, the larger Haystall Cave species, S. occidentalis and S. oreas are to be considered later in a separate paper. Introduction During an investigation of occurrences of the genus Sthenurus in Western Australia, I was able to borrow an extensive series of specimens *C/o Western Australian Museum. Perth, Western Australia. from the National Museum of Victoria, for comparative purposes. Among them was a large sample from a site near Strathdownie in western Victoria, representing two species. One of these resembled the Mammoth Cave sample from which Glauert <1910 a and b) described S. occidentalis; the other did not fit any pub- lished description. Data on these two species from Strathdownie were assembled, and con- clusions from these data are reproduced below. Another series of specimens loaned by the South Australian Museum included a large sample of Sthenurus from Haystall Cave, Nara- coorte, South Australia. Two species were pre- sent in this sample, probably the same two species as at Strathdownie. Data on the two species from Haystall Cave are also presented below. Pigs. 2. 3 . — Sthenurus gilli sp. nov. from Strathdownie, Victoria. 2. — Upper cheek teeth in different aspects, P 21729 d (P‘) originally separate from P 21615 (molars in maxilla). Note subdued masseteric process, un- damaged in P 21663. 3. — Left mandibular ramus, buccal aspect. 23 Copies of the raw data have been lodged in the libraries of the National Museum of Victoria (Russell St., Melbourne C.I., Victoria), of the South Australian Museum (North Terrace, Adelaide, S.A.) and of the Western Australian Museum (Beaufort St., Perth, Western Aus- tralia) . The Strathdownie Deposit Gill (1957) describes the Strathdownie deposit as containing a rich and varied assemblage of marsupials, with one monotreme. He states that it occurs in a limestone ridge standing some 20 ft. above the level of a plain extending from Casterton to Mt. Gambier, one of a series of such limestone ridges. The limestone is de- scribed as a beach or shallow water sediment containing marine molluscs. It is believed to have been subjected to cave formation during the latter part of the Pleistocene, and the fossil remains were found in red “cave earths” re- vealed by quarrying. The specimens described below were presented to the National Museum of Victoria in 1956 and 1957 by Messrs. C. Austin, W. Brooker and C. B. Basse. The Sthenurus sample from Strathdownie This consists of 186 specimens in an excellent state of preservation, from isolated teeth to maxillae and mandibular rami bearing com- plete sets of teeth. 65 of these specimens show teeth generally rather larger than those of Sihenurus occidentalis from Mammoth Cave, and 121 show teeth conspicuously smaller. It is concluded below that the two groups within the Strathdownie sample represent two species of Sthenurus. At least 9 individuals of the larger species are represented, and at least 23 of the smaller. For statistical purposes, care has been taken that each individual animal has been repre- sented only once for each dimension. For example, when considering the dimension "length” of lower permanent premolar, speci- mens from the left-hand side were first assembled. Each right-hand lower permanent premolar was then compared with each left- hand permanent premolar in general form and in state and pattern of wear. Any right pre- molar which could be construed as coming from the opposite side of the same animal as any of the left premolars was then rejected from the statistical treatment, and only those right pre- molars judged to derive from additional animals were accepted. Tables 1 and 2 record distri- bution of specimens between right and left sides, and the tables of raw data to be lodged in the libraries of the Western Australian Museum and the National Museum of Victoria record the side from which each specimen measured derives. I have examined 7 skulls or ankylosed mandibles of Sthenurus showing the cheek teeth of both sides, in none of which was there any sign of differential form or wear between the two sides. Thus Tables 1 and 2 record maximum num- bers of individuals for each dimension, but without the bias which might result from the double representation of some individuals and single representation of others if every available specimen had been included. For ease of comparison, data have been tabulated in the same form as Marcus (1962) used for his Sthenurus andersoni from Bingara, N.S.W. Dental enatomical terms are as used by Ride (1961); tooth designations are those of Thomas, used for the reasons advocated by him (Thomas, 1922) and by Ride (1964). Methods of measurement are detailed with the records of raw data mentioned above. Sthenurus Owen A diagnosis of the genus Sthenurus, with some comments, has been published recently by Bartholomai (1963). It should be noted that the premolar designation used by Bartholomai differs from that of Thomas (1922) used herein. Sthenurus gilli* sp. nov. Diagnosis. — M2, 3, 4 narrower than any known species of Sthenurus (i.e. than S. atlas Owen, S. andersoni Marcus, S. vales (De Vis) S. notahilis Bartholomai, S. antiquus, Bartholomai, S. creas (De Vis) or S. occidentalis Glauert). Lower molars with trigonid basin having an inclined facet to the forward face with several (usually 31 small depressions or furrows in it. Upper molars with wide anterior shelf divided by longitudinal ridglet into larger buccal and smaller lingual ncrtions. Hclotype. — National Museum of Victoria speci- men P21609, ankylosed right and left mandibu- lar rami, both lacking coroncid and condylar regions. Full juvenile dentition preserved, except hindmost molar on left. Left permanent premolar excavated, showing crown fully formed, and portion of roots. See Fig. 1. Type locality. — Shire Quarry, Section 22, Parish of Kaladbro, Strathdownie, western Victoria. Paratypes . — Same locality as holotype. (a) Used in statistics, Table 1: — Nat. Mus. Viet, specimens P21587, 21598, 21607-21609, 21613, 21614, 21618, 21621. 21624,. 21642, 21643, 21645-21648, 21654, 21655, 21657, 21659, 21662, 21666, 21677, 21681, 21724c, 21729c, 21732p (mandibular) and P21611, 21615-21617, 21626, 21635, 21637, 21649, 21671, 21674, 21683, 21701, 21724h, 217241, 21725c, 21729d (maxil- lary). (b) Not used in statistics: — Nat. Mus. Viet, specimens P21595, 21597, 21612, 21619, 21630, 21636, 21652, 21688, 21704. 21706, 21732c, 21732e-g, 21732i-n (mandibular) and P21627, 21634, 21650, 21653, 21663, 21676, 21703, 21724a, 21724d, 21724f, 21725d, 21728a, 21729a (maxillary). Some isolated molar teeth, both upper and lower, were not included in the statistics be- cause their position in the tooth-row was not certain. (c) Premaxillary specimens, believed to be referable to S. gilli, but not certainly associated with any maxillary specimen: — Nat Mus. Viet, specimens P21700, 21730b, 21730d, 21730f, 21730h-i, 21731, 21732r-s, 21733a, 21733c-d. *Named after E. D. Gill, Curator of Fossils, National Museum of Victoria, in recognition of his stimulating contributions to our understanding of Australian Quarternary events. 24 Table 1 Dental Data on Sthenurus gilli from Strathdownie, Victoria. Upper Dimension Kxamiii(‘d* X umber of Speeimens Right Side Obsem'd Jtange Sami)le Mean Sample Standard D(*viation Samj)le Coetlieient of Variation mm. mm. mm. P® Length ') h-4 10-0 9 • 82 0-25 2-5 Width 5 S-2- 0-h 9- 12 ()-()] ()•() J)L‘ Length ') 8-() - 8-‘) 8 -SO 0-12 1-4 Widtli 5 8- 9- 9-7 9 • 2() 0 - 28 :)•() Length 12 15-2 18-;l 15-98 0 - 92 5-7 Width 12 9-9-12-5 11) -95 0-87 8-0 M‘ r.engfli 10 9:1 10:1 9 • 79 0-:l() :M WiP, Liuigth 7-8 5 7 7 -(5- 8-4 7-97 0-2(5 3-2 Widtli . . 7-3 5 7 7-3- 7-9 7-48 0-20 2-7 P. Length 13-3 9 7 13-3- 15-0 14-3(5 0-49 3-4 Width 8-2 9 (5 8-0- 9-3 8 -(50 0-41 4-8 Hasin Width 3-8 (5 8 3-8- 5-5 4 - 45 0-41 9-2 Ml lauigth 8-5 9 (5 8-4- 9-3 8-81 0 - 28 3-2 Width . . 8-0 9 (5 7-8- 8-2 8-01 0-12 1-5 M, Length 9-(5 8 4 9 • 5 5) - 9 9 - (57 0-13 1-3 Width . 8-3 8 4 8 - 2 - 8-8 8 - 53 0-20 2-4 M 3 lauigth 9-7 4 4 9-7 10-7 10-13 0-3(5 3-(> Width 8-8 4 4 8 - 8 - 9-9 9-19 0-33 3 -(5 .Al, lyength ... .. . in 2 2 9-5- 9-9 9-70 0-18 1-9 Width alvt'ohis 2 9-1 9-5 9 - 33 0-17 -9 * Details of measiirin*' procedures, with estimates of a<-curacy and consistency, lod«jed with data on individual s])ecimens in Xational Museum of Victoria and Western Australian Museum. All dimensions maximal, widths in molars across protoloph or protolophid, depth 1, perpendicular to lonu axis of tooth, basin width 1’, across posterior centra! basin. Comparison of S. gilli with other species See Figs. 1 (holotype, showing lower dentition), 2 (upper dentition) and 3 (buccal aspect of mandibular ramus), and Table 1. Distinguishable from S. andersoni Marcus by procumbency of lower incisor (much less pro- cumbent in S. gilli) and by width of Pa, P.i and M 2 - 1 : in S. gillu and P 4 wider but M 2 , 3, 4 narrower than in S. andersoni. Trigonid basin projecting relatively further forward, and cheek tooth row (P;i, DP 4 , Ml-,, inclusive in juveniles) much longer, in 5, andersoni than in S. gilli. Upper permanent premolars not known in S. andersoni, but upper molars stated by Marcus (1962) to lack forelink. Distinguishable from S. occidentalis Glauert and from the larger Strathdownie species not only by dental dimensions, but also by lack of a prominent descending masseteric (zygomatic) process, and by lack of molar ornamentation. In S. occidentalis and in the larger species at Strathdownie, the descending process is very marked, projecting downward further than the ccclusal surface of the upper cheek tooth row, whereas in S. gilli it is so subdued and smoothly I’ounded as hardly to justify the term “process.” In S. occidentalis and in the larger Strathdownie species, there is much more ornamentation on both upper and lower molars than in S. gilli. 1 have been able to make direct comparison of S. gilli specimens with only one specimen of S. oreas (De Vis), viz. Queensland Museum specimen P3814, figured by Bartholomai (1963- Fig. 5). This specimen, portion of a maxilla, differs markedly from any specimen of S. gilli not only in the greater lengths and widths of molars in S. oreas, but also in their showing much more ornamentation. Prom Bartholomai’s (1963) description of his revised S. oreas, it would appear that the mandi- bular ramus in S. oreas closely resembles that of S. gilli in form, except that the masseteric crest and masseteric foramen would appear to be a little lower in S. gilli (see Fig. 3) ; however, in almost all quantitative respects (except width of lower permanent premolar), both bone and teeth appear to be markpdly smaller in man- dibles of S. gilli than S. oreas. The small size of the molar teeth in 5. gilli suffices to distinguish this species from any other species of Sthenurus. However, two aspects of tooth morphology also appear to be distinctive of S. gilli, those noted in the diagnosis above, one for upper and one for lower molars. The larger species of Sthenurus at Strathdownie Specii7iens . — (a) Used in statistics, Table 2. — Nat. Mus. Viet, specimens P21586, 21633, 21638, 21640. 21644. 21656, 21724a, 21724b, 21725b, 21729b. 21732a-b. 21732d (mandibular) and P21629. 21641. 21660, 21673, 21687. 21705. 21720b. 21724g, 21725a, 21728c. 21729e (maxil- lary). (b) Not used in statistics; — Nat. Mus. Viet, specimens P21732 o (mandi- bular), P21672, 21721h, 21728b. 21749 (maxillary) and P21678. 21730g, 21733b, 21733e (premaxil- lary). Some isolated lov/er molars were not included in the statistics because their position in the tooth row was not certain. Comparison of larger Strathdownie species with other species See Figs. 4 (showing upper dentition) and 5 (showing lower dentition) and Table 2 (of dental dimensions). Closely resembles S. occidentalis Glauert and S oreas (De Vis) in form. In lengths P^ and widths M 1;^ considerably exceeds Mammoth Cave sample of S. occidentalis (direct com- parison — D.M.). In lengths P;| and widths M1;2 exceeds Queensland sample of S. oreas, according to data published by Bartholomai (1963) and to my direct comparison with one Queensland specimen (F3814 — see above). De- spite the quantitative differences observed be- tween the large Strathdownie sample and those from Mammoth Cave and from (Queensland, I believe the animals concerned were related, and propose to investigate the relationship further. In the meantime, it appears desirable to leave open the matter of the taxonomic status of the larger Sthenurus at Strathdownie. Distinguishable from 5. antiquus Bartholomai on the form of the lower permanent premolar and the spacing of the lower molars, according to my direct comparison with Queensland Museum specimens F2931 and F2932, figured in association by Bartholomai (1963 — Fig. 9). In the larger Strathdov/nie specimens, each lower molar oyerlaps the base of the preceding tooth in the cheek tooth row to a greater extent than in 5. antiquus. The lingual crest on the lower permanent premolar F2932 of 5. antiquus in- clines inward to a much greater extent, and the buccal crest is relatively much lower, extends less far forward, and is more clearly separated on the rear face of the tooth from the lingual crest, than in any P.i from the larger Strath- downie species. Upper incisors of Sthenurus at Strathdownie No specimens were available in the Strath- downie sample which associated upper molars or premolars referable to S. gilli with upper in- cisors. However, available I^ and I- specimens from Strathdownie fell into two distinct groups, larger and smaller: it would appear justifiable to associate the smaller incisors with S. gilli rather than with the larger species occurring at Strathdownie. Unfortunately, no specimen showed an U associated with a smaller I^ or I'; but it was possible to divide the I'* speci- mens available into two kinds. One of these groups could be associated with I^ and and clearly derived from the larger species at Strath- downie: furthermore, in form this group re- sembled the I‘‘ teeth of S. occidentalis. In the second of the I'* groups, the individual teeth were more stiap-like than F in S. occidentalis, but were less rugose on the lingual aspect, and bore a less pronounced “fold” in the enamel of the lingual aspect of the antero-occlusal corner of the tooth. Since the larger Strathdownie species resembled S. occidentalis in many parti- culars. it would appear reasonable to assign the more strap-like upper third incisors to S. gilli rather than to the larger species. It is on these grounds that the allotment of incisor teeth to the two different species in the Strathdownie deposit has been made. See Fig. 6. S. Aust. Mus. specimen P 13687, attributable to Sthenurus gilli (see below) does show and F. but unfortunately not 1~, in association with cheek teeth distinctive of this species. See Fig 7. It confirms the division made among the unattached Strathdownie upper incisors. The Haystall Cave Deposit According to labels supplied with the speci- mens from Haystall Cave, all the specimens on loan to me appear to have been recovered from a red to yellow sandy deposit at depths up to 2 ft. 6 in. One left mandibular ramus (P 13682) from the top 6 in. appears to derive from the same animal as right ramus P 13830a from 2 ft. 6 in.: thus it is probable that all the speci- mens are approximately contemporaneous. They are probably of late Quaternary age. The sample was collected and presented to the South Australian Museum in 1963 and 1964 by the Cave Exploration Group (South Austra- lia). The Sthenurus sample from Haystall Cave This consists of 52 well-preserved specimens, most of them juveniles, 24 forming a homo- geneous group with larger teeth; and 28 forming a similarly homogeneous group with smaller 26 Table Z Dental data o?i the larger species of Sthenurus from Strathdownie, Victoria. Upper l>ini<‘iisioii Kxaiiiint“(l* N iimbi'r of SjH'cinuMis Left Kijiht Oh/StTved Kariue Sample .M<*an Samiih' Standard l)(“viatiou mm. mm. mm. )»3 LeiiKUi .) 11 12-0-1 2(1 12-00 Width 0 l()-2 10 •« 10-4.5 0-20 OF* (1 10-.') 11-7) 1 1 ■ 00 0-71 Width •> (1 1 1 -2 12-2 1 1 - 7.5 0-77 F' Lenjitli ■> 2 17-1 18-H IS -22 0 - 0;5 Width .> • ) 121 -12-9 12-OC. 0-7:5 M- Lenijtli •> (1 12-8- 12-2 12-00 0-2S Width •> 0 12-9 12-1 12-00 ()-14 M'‘ L(‘iiKtii '•> 1 2 • 7 - 1 4 • <) 14-lS 0 • 22 Widtli •> J2-t)-14'2 1 2 - S2 0 • 2S Lontjtli 1 4 • 2- 1 .5 • 14-72 0-00 Width 12-()-12-K 12-70 0-14 M* ]/t‘imtii •) 120-I4-0 1 2 - 02 0-S.5 Widtli •» 12-7 14-2 1 2 - 22 0-S4 Lower ^J>fpth {) 2 I2-0 12-2 12-07 0-0;5 IviMitith o 9-0 -10- I 9 - 90 0-27 Width •> 7-7 - S-0 7 - S7 0-10 DF„ Length 9-7 10- 1 9-90 0-2S Widtli 9-0 9-7 9 - 2.5 0-49 P. J.enjith lO-O 17-t) 10-79 0-52 Width • > h 9-S 10-7 10-29 0-27 M, Len^tli ] 4 11-') 12-4 11-74 0-2S Width 4 101 110 10- 2S 0-27 Ms la-iifith 1 12-7 12-7 1 2 ■ 70 Widtli 1 1 11-4 11 -9 1 1 - 0:5 0-2.5 M. Lcniith 1 1 12-212-4 12-20 0-14 Width 1 12 0 12-2 12-10 0-14 M„ la'ii^th 11 12-SO Width 1 1) 1 1 • .50 * All diimMisioiiis niaxinial. widths ill iiioiars across jirotoloph or protolophid, depth 1, pcriicmlicular to long of Variation 1 1 ) () • .') Ii-O 5 • 7 ’2 • '2 T 1 2 :{ 2-,S 41 I •() <» • :i .')• I 2-7 2-0 2-9 'v2 :}-2 2t) ;{-2 • r> 0 1 • 1 } -2 tootii. teeth. At least 12 individuals of the larger species are represented, and at least 9 of the smaller. From these groups, the specimens listed below were selected for statistical purposes in the same way as frcm the twQ groups at Strath- o'ownie: — (a) Larger species. S. Aust. Mus. specimens P 13703, 13831 g-h (maxillary) and P 13674- 13676. 13678, 13680-13681. 13696, 13711, 13713- 13714, 13830a. 13831 c-d (mandibular). (b) Smaller species. S. Aust. Mus. specimens P 13687 (complete upper dentition of both sides except right F, left F'-M, P 13690 a-b, 13691- 13694 (maxillary) and P 13688, 13702, 13704, 13706, 13708, 13710, 13717, 13830 b, 13831 a-b (mandibular) . Measurements on these specimens are sum- marized in Tables 3 and 4. The smaller species of Sthenurus from Haystall Cave Direct comparisons of specimens of the smaller Strathdownie with the smaller Haystall Cave species show that the two are closely similar. On morphological grounds, one would equate them. Compare Figs. 1, 2 and 3 with Fig. 7 below. However, mean dimensions in the Haystall Cave sample exceed those in the Strathdownie sample in 28 cases out of 30, and there are even 2 cases (length DP-' and width Mh in which ranges in the two samples do not overlap. See Tables 1 and 3. Such a situation, in which strong morpholo- gical resemblances between two samples are coupled with quantitative distinctions, has been discussed by Ride (1964). He has proposed numerical tests for the recognition of sub- species in this situation. On Ride's tests, either length of DP^ or width of M‘* could conceivably serve to differentiate as subspecies the Haystall Cave and Strathdownie samples of the smaller Sthenurus. But none of the differences revealed by comparison of Tables 1 and 3 below can, in my opinion, be construed as differentiating full species. The Haystall Cave sample appears to me to be merely a geographical or temporal variant of Sthenurus gilli as defined above. 27 Figs. 4, 5 . — Sthenurus sp., the larger species from Strathdownie, Victoria. 4. — Upper cheek teeth, P-< excavated. Note prominent masseteric process. 5. — Lower cheek teeth. P.j excavated. 28 The larger species of Sthenurus from Haystall Cave Like the smaller species, the larger species at Haystall Cave and at Strathdownie are closely similar in form, but show quantitative differences. Compare Figs. 4 and 5 with Fig. 8, and Table 2 with Table 4. Note that the Hay- stall Cave sample of lower teeth iz the more numerous of the two. In a substantial minority of cases (13 out of 29) mean dimensions (or single measurements where only one specimen occurs) in the less numerous sample fall outside the observed range in the more numerous sample, and in several cases (e.g. length P‘, depth Ii, width Pa) there is no overlap in range. Despite the strong resemblances between the two samples, some doubt therefore remains about the grade of their relationship. Since in any case, the question of relationships of both samples with the Darling Downs (Queensland) sample of Sthenurus areas and with the Mam- moth Cave (W.A.) sample of S. occidentalis are to be taken up later, the question of con- speciffcity of the Haystall Cave and Strath- downie samples is not further pursued here. Provisionally, the two samples appear to me to be geographical or temporal variants of one species. Possible occurrences of the Strathdownie and Haystall Cave species of Sthenurus in Western Australia A small lower permanent premolar (P4) of Sthenurus was described and figured by Lun- delius (1963) from Madura Cave on the Nuliar- bor Plain. This specimen is now lodged in the Chicago Natural History Museum, but I have examined a plaster cast of it (W. Aust. Mus. specimen 63.6.1). Its length is 14.2 mm., its width 8.1 mm. and its basin width is 3.5 mm., if measurements made from the cast can be taken as accurate. (“Basin width” is the maximum distance separating buccal from lin- gual crests on the hinder portion of P4). The Madura Cave tooth resembles P^ in both Haystall Cave and Strathdownie samples of S, gilli in form, and in length falls within the range shown in both samples. In width it falls within the Strathdownie but not quite within the Haystall Cave ranges, and its basin width is smaller than any in these two samples. Only provisionally, therefore, can it be referred to S. gilli. Since his original discovery at Madura Cave, Lundelius has participated in further excava- tions there (Lundelius E. L. 1964 pers. comm.) Fig. 6.— Comparison of Sthenurus upper incisors from Strathdownie, Victoria. Left, P 21731 P21733 a and P 21730 h attributed to S. gilli. Right, P 21678 and P 21730 g attributed to the larger species. 29 Table 3 Dental data on the smaller species of Sthenurus from Haystall Cave, NaracoortCy South Australia. Upper NuinlK'r 1 )imenssion Kxaniiuc(l * of Specimen!? Left (tight ()t)served Range Samjth* .Mean Sam])l(‘ Standard Deviation Sami)le (•oettieient . of Variation mm. mm. mm. ]' Length 0 e. 5-9 J3 Lengtli 0 1210 Lengtli 1 h-S lO-.') 1015 0-29 2-8 Width 1 8-4- 8-90 0-:57 4-2 m*’ Length . .. 1 9-:}- 9-<) 9-4:5 0 - 1 5 1-6 Width ... 1 :i 9-4 10-2 9-80 0-:54 :5-4 Length .) 4 ir)-2-ld-4 15-82 0-45 2-8 Width r 4 1 1 -0- 12-2 1 L - 48 0-71 6-2 M' Length ... 0 4 lOO-IO-d 10-2:5 0-27 2-6 Width ... 0 4 9-9- 10-(i 10-25 0-29 2-8 M* J.engtli ,. :l 4 lU-4 11-7 10-90 0-54 5-0 width 1 4 10 -2-10 -9 10- 50 0-25 2-4 Length 1 2 1 1 • :M 1 • 9 1 1 -t)0 0-:50 2-6 WidtJi 1 10 -0 11 -0 10-77 0-21 2-0 L(“iigth 1 lOO-llO 10-80 0-28 2-6 Widtli 1 10-1-10-4 10-25 o:57 :5-6 Lower J 1 Dt'pth 0 4 9-711-2 10:55 0 • 6t) 6-:3 I’a Lengtli 0 5 8-5- 9-4 9-10 0-:59 4:5 Width 0 5 71- 7-9 7 • 56 0:30 :5-9 m\ Length 0 5 8-1- 8-6 8-26 0 19 2-;3 Width 0 5 7-:5- 8-:5 7-64 0-42 5 • 5 I\ J.ength 0 9 1 :5-7 16-2 14-97 0 • 78 5-2 Width 0 9 8-:5- 9-9 9-01 0 • 48 5-:i Hasin Width 0 7 4-1- 5-4 4-69 0 • 52 11-0 Mi Length 0 8 8-5- 9-5 9-1:5 o-:54 :5-7 Width .. 0 9 7-9- 9-0 8-:57 0-41 4-9 M., Length 0 6 9:5-10-8 9 - 9:5 0 - 5:5 5-3 Widtli 0 6 8-5- 9-7 8 - 9:5 0-49 5-5 M., Length ... 0 4 9-7 11-1 10-:55 0-58 5 • t5 Width 0 4 9-1 10-4 9 • 60 0-59 6-2 M, I.ength ,. 0 2 9-5 10-4 9 - 95 0-64 6-3 Width 0 2 9-4 10-4 9-90 0-71 7-1 * All diiiu'iisions maximal, widths iu molars across protoloidi or protoloidiid. depth 1, jXTpendicular to long axis of tooth. but Without finding any more specimens of Sthenurus. It is unfortunate that a single premolar does not necessarily provide more than a general guide to the probable size of the molars in Sthenurus. For example, S. oreas (De Vis) as revised by Bartholomai (1963) has lower permanent premolars which are shorter than those of 5. occidentalis Glauert, together with lower molars which are longer. Both premolars and molars are similar in form. If. in these samples, only premolars had been known, one might have predicted from them that the molars of S. areas would have been shorter than those of S. occidentalis. By analogy, it is possible that the molars of the Madura Cave species of Sthenurus may have differed quite markedly from those of S. gilli. Tentatively, however, it may be suggested that Sthenurus gilli ranged into Western Aus- tralia. As remarked above, the larger species of Sthenurus at Haystall Cave and at Strathdownie resemble 5. occidentalis from Mammoth Cave, W.A. in form and in size of the lower permanent premolars, but not closely in size of the molars. These samples could conceivably represent geographical variants of a wide-ranging species which once included Western Australia in its range. Acknowledgments I am indebted to Mr. E. D. Gill (Curator of Fossils, National Museum of Victoria), to Dr. D. W. P. Corbett (Curator of Fossils and 30 S, Aust. Mu9 P \%Q1 cm 1 Figs. 7, 8 . — Sthenurus from Haystall Cave, South Australia. 7. — Variant of Sthenurus gilli sp. nov. Skull with left ?■' and DP» removed to expose P*. 8. — The larger species of Sthenurus at Haystall Cave. Right mandibular ramus with P- and DPi displaced to the side, exposing Pi. 31 Table 4 Dental data on the larger sjjecies of Sthenurus from Haystall Cave, Naracoorte, South Australia. Upper Dimi'iision Examined* Number of Specimens Left liight Ohs(‘rved Range Sample .Mean Samjjle Stamlard Deviation Sample C’oeltieient of Variation mm. mm. mm. 1” Length 0 1 10 -SO Width 0 1 10- 10 J)P‘ Length 0 1 U ■()() Width 0 1 10 (>() Length (1 2 10-7 Ki-S 1() • 75 0- 10 0 -() Width 0 2 120 12-5 12-25 0-:l5 2-0 Length 0 1 12-10 Width 0 1 1 1-50 W I.ength 0 U-;i() Width 0 l:l-(iO yP Length 1 0 U-30 Width 0 Ll-SO *M‘ I.ength 0 LMO Width 1 0 12-70 Lower J, Dejith 4 1 10-9-11 (5 11-20 0-79 7-0 Length 8 1 . 9-4-10-4 9-90 0-42 4-2 Width 8 1 8 • 7 • 9-7 8 - 98 0 -:i 8 4-2 in\ hength 8 9-:J~10-5 10-05 0 • 43 4-3 Width 8 1 8-7- 9-9 9 - 29 0-35 3-7 1\ Length 8 1 5 - 5-1 7 - 5 10-45 0-54 3-3 Width 9 2 9110-0 9 - 85 0;18 3-8 M, Length 7 2 1 1 ■ () 1 2 - ;5 11 -90 0-57 4-8 Width 7 2 9-8-10-9 1 0 - 32 0-32 3-0 3L Length 1 :i 12-9 14-0 13-70 0-54 3-9 Width 2 11-5 11-9 11 -70 0-20 1 -7 M 3 Length 0 14-2 14-5 14-37 0-10 1-1 Width 2 0 12-5 12-7 12-00 0-14 l-I .M, Length 2 0 13-0 13-9 1 3 - 45 0 - 59 4-4 Width 2 0 11 -7-12-5 12-10 0-57 4-7 * All dimensions maximal, width in molars across protolo])h or protolophiti. (h'ptli J, l)erpendiciilar to lontf axis of tooth. Minerals, South Australian Museum), to Mr. A. Bartholomai (Curator of Geology, Queensland Museum), to Mr. H. O. Fletcher (Curator of Fossils, Australian Museum, Sydney), to Mr. F. L. Sutherland (Geologist, Queen Victoria Museum, Launceston, Tasmania), to the Directors of these Museums and to Mr. J. A. Mahoney (University of Sydney) for the loan of the samples described above and of a wide range of comparative specimens. Messrs. Gill and Bartholomai, and Dr. B. Daily (University of South Australia) also very kindly provided information on the specimens. I am grateful also to Dr. W. D. L. Ride and to the staff of the Western Australian Museum for constructive criticism of this and ancillary work, and for access to the modern and fossil marsu- pial collections and to records of the Western Australian Museum, and to Mr. W. B. Sewell for his care with the photographs reproduced herein. References Bartholomai, A. (1963). — Revision of the extinct Macro- podid genus Sthenurus Owen in Queensland. Mem. Qd. Mus. 14: 51-76. Gill, E. D. (1957). — The ancestors of our marsupials. Australian Amateur Mineralogist 3: No. 3. Glauert, L. (1910a — January). — The Mammoth Cave. Rec. W. Aust. Mus. 1: 11-36. ( 1910b — March ) . — Sthenurus occidentalis (Glauert). Bull. Geol. Surv. W. Aust. 36: 53-69. Lundelius, E. L. (1963). — Vertebrate remains from the Nullarbor Caves, Western Australia. J. Roy. Soc. W. Aust. 46: 75-80. Marcus, L. F. (1962). — A new species of Sthenurus (Mar- supialia, Macropodidae) from the Pleistocene of New South Wales. Rec. Aust. Mus. 25: 299-304. Ride, W. D. L. (1961). — The cheek-teeth of Hypsiprym- nodon moschatus Ramsay 1876 (Macropodi- dae: Marsupialia) . J. Roy. Soc. W. Aust. 44: 53-60. (1964). — A review of Australian fossil mar- supials. J. Roy. Soc. W. Aust. 47: 97-131. Thomas, O. (1922). — On mammals collected by C. Keys- ser in . . . New Guinea. Ann. Mag. Nat. Hist. 9: 669-676. 32 INSTRUCTIONS TO AUTHORS Papers may be submitted to the Society in accordance with Rules and Regulations 38 to 41 inclusive (see below). They should be addressed to The Honorary Secretary, Royal Society of Western Australia, Western Australian Museum, Perth, Authors are solely responsible for the factual accuracy and for any opinion expressed in their papers. They are particularly requested to verify references. Alterations to MSS. submitted to the printer will be allowed only under excep- tional circumstances, and no changes will be permitted after galley-proof stage. In the preparation of MSS. authors are required to follow the C.S.I.R.O., Guide to Authors (C.S.I.R.O., Melbourne, 1953), except that papers longer than 10,000 words (30 foolscap pages of pica type, with 6 inch lines 40 to the page) will not normally be accepted. Authors may be required to meet half the cost of preparation of the blocks of diagrams and illustrations. Authors shall receive a total of 30 reprints free of charge. Further reprints may be ordered at cost, provided that such orders are submitted with the MS. RULES AND REGULATIONS 38. Every paper intended to be read before the Society or to be published in the Society’s Journal must be sent to the Secretaries at least seven days before the date of the next ensuing Council meeting, to be laid before the Council. It will be the duty of the Council to decide whether such contribution shall be accepted, and if so, whether it shall be read in full, in abstract, or taken as read. All papers accepted for publication must be read or otherwise communicated at an ordinary meeting prior to publication. 39. A Publications Committee, appointed by the Council, shall recommend to the Council whether a paper presented to the Society shall be published in the Society’s Journal. The Publications Committee may obtain an opinion from any person it may select on the suitability of any paper for publication. 40. Publication in the Society’s Journal shall only be available to (a) Ordinary Members, Cb) Honorary Members, (c) Non-members resident outside Western Australia, who ^must communicate the paper through an Ordinary or Honorary Member. No paper shall be accepted from a Non-member resident in Western Australia. 41. The original copy of every paper accepted for publication by the Society, with its illustrations, shall become the property of the Society, unless stipulation is made to the contrary, and authors shall not be at liberty to publish their communicated papers elsewhere prior to their appearance in the publications of the Society unless permission for so doing is given by the Society, or unless the Society fails to publish the paper in the Journal of the year in which it is read or otherwise communicated, or of the succeeding year. Journal of the Royal Society of Western Australia Volume 48 1965 Part 1 Contents 1. — The Physiography, Vegetation and Vertebrate Fauna of the Wallabi Group, Houtman Abrolhos. By G. M. Storr. 2. — The Ocypode Ghost Crabs of Western Australia (Crustacea, Brachyura). By R. W. George and Mary E. Knott. 3. — Two species of the extinct genus Sthenurus Owen (Marsupialia, Macro- podidae) from south-eastern Australia, including Sthenurus gilli sp. nov. By D. Merrilees. Editor: R. W. George Assistant Editor: R. D. Royce The Royal Society of Western Australia, Western Australian Museum, Perth 91629/12/64—570 ALEX. B. DAVIES, Government Printer. Wcsterf’ Australia JOURNAL OF THE ROYAL SOCIETY OF WESTERN AUSTRALIA VOLUME 48 PART 2 PUBLISHED 30TH JUNE, 1965 REGISTERED AT THE G.P.O., PERTH FOR TRANSMISSION BY POST AS A PERIODICAL THE ROYAL SOCIETY OF WESTERN AUSTRALIA COUNCIL 1964-1965 President Past President Vice-Presidents Joint Hon, Secretaries Hon. Treasurer Hon. Librarian Hon. Editors W. R. Wallace, Dip.For. C. F. H. Jenkins, M.A. J. H. Lord, B.Sc. D. Merrilees, B.Sc. Margaret E. Redman, B.Sc. A. B. Hatch, M.Sc., Dip.For. R. D. Royce, B.Sc, (Agric.). Ariadna Neumann, B.A. R. W. George, B.Sc., Ph.D. (1964). J. E. Glover, B.Sc., Ph.D. (1965). B. E. Balme, B.Sc. (Hons.). J. S. Beard, M.A., B.Sc., Ph.D. A. S. George, B.A. J. G. Kay, B.Sc. P. E. Playford, B.Sc., Ph.D. R, T. Prider, B.Sc., Ph.D., M.Aust.I.M.M., F.G.S. W. D. L. Ride, M.A., D.Phil. D. L. Serventy, B.Sc. (Hons.), Ph.D. Journal of the Royal Society of Western Australia Vol. 48 Part 2 4. — A History of the Royal Society of Western Australia, and its role in the Community. Presidential Address, 1964. By C. F. H. Jenkins, M.A.* Delivered — 20th July, 1964. Ladies and gentlemen; I was greatly honoured when this society elected me as President for the year 1945-46 taut I regard it as a much greater honour to have the privilege of holding office for a second term, particularly in the Society’s Jubilee Year. I joined the Royal Society in 1929 and have served on the council since 1932. During that period I have been associated with every office excepting that of treasurer and so have been very closely connected with the Society’s activi- ties for over 30 years. This experience has encouraged me to investigate the earlier history of the organisation and to present these findings together with a number of general comments as my Presidential Address for this anniversary year. If any other excuse is necessary for presenting what may be regarded as largely parochial material to a gathering such as this, I might point out that Royal Society presidents in all the other states have given comparable ad- dresses, as did Sir Howard Florey in his 1963 Anniversary Address to the Royal Society of London. The Royal Society of London is of course, the most celebrated scientific society in the world and received its charter from King Charles II on 22nd April, 1663. The formation of this society marked the development of modern science in Britain and provided a pat- tern to be followed by such empire countries as South Africa, Canada, New Zealand and Australia, all of whom modelled their societies on the parent body. Modern science came to Australia with James Cook and Joseph Banks in 1770, but it was not until 1821 that the first scientific society was formed when Sir Thomas Brisbane, Governor of N.S.W. founded the Philosophical Society of Australia. This survived for little more than 12 months, but was followed in 1850 by the Australian Philosophical Society, for the en- couragement of Arts. Science, Commerce and ♦Department of Agriculture, Jarrah Road, South Perth. Agriculture. In 1885 this became the Philoso- phical Society of N.S.W. and in 1866 it assumed the title of the Royal Society of N.S.W. The Royal Society of Van Diemans Land for Horticulture. Botany and Advancement of Science was formed in 1843, later of course to become the Royal Society of Tasmania. The Royal Society of Victoria was established in 1854, the Royal Society of South Australia in 1877 and that of Queensland in 1884 (Elkin. A. P. 1962, AJ.S. 25: 2). A Royal Society was formed in Canberra in 1930 and at first assumed the name of the Royal Society of Australia. The various state Royal Societies were opposed to this title however and in 1955 it was changed to the Royal Society of Canberra. Royal Society of Western Australia The foundations of the Royal Society of W.A. were laid as far back as 1891 when under the presidency of Sir John Forrest (later Lord Forrest) the Western Australian Natural History Society was founded. This society did not pub- lish a journal and failed to survive beyond 1895. Then in 1897 the Mueller Botanical Society was established with Sir John Forrest president (With E. J. Bickford Chairman) and His Excel- lency the Governor, Lt. Col. Sir Gerard Smith K.C.M.G. Patron. The remaining members of Council were as follows: Vice Presidents: Hon. E. H. Wittenoom, M.L.C. G. Leak, M.L.A. Hon. Secretary: F. J. Skewes. Council: Dr. F. Tratman. Dr. F. Wilkinson. A. W. Milligan. J. S. Battye. In 1903 the scope of the Society was broadened and the name was changed to “Western Aus- tralian Natural History Society, with which is incorporated the Mueller Botanical Society.” The following comprised the first Council: Patron: His Excellency the Governor, Admiral Sir Frederick Bedford, G.C.B. President: C. R. P. Andrews. Vice Presidents: Dr. P. Tratman. B. H, Woodward. A. Purdie. Editors: C. R. P. Andrews. P. Crossman. Treasurer and Librarian: A. Purdie. Secretary: P. S. Dobbie. Council: A. F. Crossman. E. W. Hursthouse. A. Gibb Maitland. A. W. Milligan. Justice R. MacMillan. H. C. Prinsep. In 1909 the Societies’ work broadened still further and the name was changed accordingly to the “Western Australian Natural History and Science Society.” The first officer bearers were as follows: Patron: His Excellency the Governor Sir Gerald Strickland K.C.M.G. President : Dr. F. Tratman. Vice Presidents: A. Gibb Maitland. Dr. A. Morrison. Treasurer: T. Thorpe. Editors: H. M. Giles. C. P. Conigrave. Secretary: C. P. Conigrave. Librarian: C. P. Conigrave. Council: E. S. Simpson. W. Catton Grasby. T. N. Lee. P. G. Brown. H. M. Giles. It is recorded that in 1907 several meetings lapsed for want of a quorum and a committee consisting of Messrs. B. H. Woodward, E. A. Mann, J. B. Allan and Dr. J. B. Cleland, was appointed to investigate the possibility of form- ing a Royal Society or alternatively affiliating wuth the Royal Society of South Australia. No immediate action was taken however, and the title of Royal Society was not assumed until 1914. The letter conferring this honour was presented to Council by Professor Dakin on December 2nd, 1913 and published in the “West Australian” on December 13th. The first office bearers for the new Society were as follows: Patron: His Majesty King George V. Vice Patron: His Excellency the Governor, Major General Sir Harry Barron K.C.M.G. President : Professor W. J. Dakin. Vice Presidents: A. Gibb Maitland. Professor A. D. Ross. Hon Secretaries: M. A. Browning. W. B. Alexander. Hon. Treasurer: J. J. East. Hon. Librarian: A. O. Watkins. Hon. Editors: A. Gibb Maitland. M. A. Browning. W. B. Alexander. Council: H. B. Curlewis. W. J. Hancock. A. Montgomery. E. S. Simpson. B. H. Woodward. Functions and Aims of The Royal Society The main functions and aims of the Society are summarised in paragraph 2 of the 1936 Constitution which reads as follows: “To pro- mote and to assist in the advancement of Science in all its branches.” This gives the Society the widest possible scope and places no restriction upon membership, providing correct election procedures are followed. In addition to producing an annual journal, the Society has conducted regular meetings for the presentation of scientific papers, discussions and popular talks. A perusal of the records show that the Society has concerned itself with many matters of public interest and has offered ad- vice and criticism where these have been con- sidered necessary. In addition to subjects such as Geology and Biology which constitute the major contributions to the Society, papers and reports have been presented on such matters of topical interest as the following: Spiritual Healing (Vol. X, p xxiv> Water Divining (Vol. VI, p 112) Housing (Vol. V, p 62) Decimal Currency in Australia (Vol. IV, p 16) Town Planning (Vol. V, p 39) as well as other subjects of public concern or popular interest. Special Committees. The Society has on many occasions appointed working committees to investigate problems of public interest or major importance. Some idea 34 of the range of subjects covered will be gained from the following list: Vermin Committee (Vol. VII, p 112) Salinity Committee (Vol. XIII, p xiii) Flora & Fauna Conservation Committee (Vol. XIII, p xiv) Excursion Committee and number (N) in samples < which have been partitioned into generations) . i. isolepis i. gularis m. maculatus X. K. M. X. K. M. X. K. M. December 1 22 22 2 29 31 30-2 January * 0 27-35 30-7 \ 29 29 February March 7 24 34 30-1 April 2 38-39 38-5 May 8 28-40 33 • 5 29 30 49 41 -3 June 5 33-45 37-8 July 17 31 52 42-4 35 36 55 47-8 Augtist 32 34-54 43 • 5 20 34-59 44 4 Sei)teinber 29 40-65 55 - 1 2 48 48 1 52 52 October 9 44-60 52-3 48 58 50-4 Xovember 31 43-65 55 8 18 42 65 53 • 1 December 21 47-50 54 • 9 0 49 56 52-4 January 50 46 66 54-9 10 48 59 52-4 February 11 55 65 00 0 March 0 62-70 00-9 25 50-63 50 • 3 April 0 58 66 02 0 May 1 61 01 53 ful. In the few places where maculatus and isolepis are sympatric, as at North West Cape and east of Southern Cross, no differences are evident in habitat preferences. Amphibolurus fordi occupies an area in which scanty rains are received in summer as well as winter. It is confined to isolated patches of red sand clothed with small spinifex and scat- tered mallee. a habitat it shares with the south- ernmost populations of i. gularis. Although its geographical range is small. /e??7,oraZzs occurs in a variety of habitats. North- east of Giralia it is restricted to the almost bare tops of red sand dunes, whereas the much larger i. rubens monopolises the spinifex-covered interdunes. A little further west, beyond the range of rubens, femoralis is sympatric with i. isolepis and m. badius. It has been found with both of these on heavy loam vegetated with spinifex. It has also been found with isolepis on stony, clayey soils, and with bndius on red dunes clothed with proteaceous shrubbery. Phylogeny The common feature in the colour pattern of the various members of this species-group are the pale lateral and dorsolateral stripes, which are otherwise rare in Amphibolurus. However a pale dorsolateral stripe occurs ia Physignathus, Diporiphora and Tympanocryptis. Strongly keeled dorsal scales are also shared by members of these genera with the maculatus group. In their dorsal colour pattern these genera tend to combine pale longitudinal stripes with dark transverse bands, as in Tympanocryp- tis lineata and Diporiphora spp. Among the maculatus group this kind of pattern is best preserved in maculatus itself, where the break in the transverse bars may represent a former vertebral stripe now submerged in the dorsal ground coloration. Despite their bright colours and bold pat- terns, all members of the species-group are inconspicuous against their natural background. The general coloration matches well with the substrate. All but three taxa occur on reddish soils, and all except these three are dorsally reddish. The pattern too is responsive to the environment. The transverse bars of maculatus are useful in habitats where the vegetation is high and dense enough to throw numerous shadows. But on the bare sunlit ground of arid habitats, barring would undoubtedly be dis- advantageous. Furthermore, in arid habitats the dorsolateral stripe has undergone some modification. In femoralis and z. rubens it has almost disappeared; while in z. gularis and, to a lesser extent, z. isolepis it tends to break up into ocelli similar to those on the back. While it is not unlikely that maculatus is the most primitive member of the group in colour pattern, it is not at all certain that it is primitive in other characters. With respect to number of pores and extent of ventral black in males, femcralis, fordi, maculatus and isolepis form a series. Whether the series has been formed by reduction, addition or both, there seems to be no way of ascertaining. Regarding pores, I favour the hypothesis of reduction. Successive loss of preanal pores from near the midline would reduce the strong arch of isolepis to the slight bow of maculatus and fordi and the straight alignment of femoralis. Such a sequence demands the loss of pores la common event in the Agamidae) rather than the evolution in isolepis of a unique alignment of preanal pores. The extent of ventral black is roughly cor- related with body size. I. citrinus is clearly derived from z. gularis; it is considerably smaller than its parent race and has less ventral black. Conversely z. rubens is much larger than its presumed parent, z. isolepis, and has much more ventral black. In the series m. griseus, m. maculatus and m. badius, decreasing size is ac- companied by decreasing ventral black. Within a taxon the ventral black (first attained by males when their snout-vent length is between 40 and 50 mm) continues to spread with growth, even after breeding. Hence change in the num- ber of pores and relative extent of ventral black could be consequent of phylogenetic change in body-size. Body-size is not consistently influenced by non-biotic factors such as latitude. Whereas maciuatus increases in size from north to south, the opposite obtains in the sequence isolepis- gularis-citrinus. On the other hand the prior presence in a habitat of another member of the group could well determine whether an in- vader would advantageously increase or decrease its body size. With a snout-vent length of up to 67 mm. griseus is by far the largest race of maculatus, which is otherwise a smaller species than isolepis. East of Southern Cross griseus coexists with citrinus (SVL up to 57). the small- e:^t race of isolepis. In the sandhill country east of Exmouth Gulf, rubens (SVL up to 83 and m.uch the largest race of isolepis) coexists with femoralis (SVL up to 57). Over most of the State the maculatus group is represented by only one species — isolepis in the north-east, and maculatus in the south- west. It is probable that the greater part of the evolution of the taxa has been allopatric, and that their morphological divergence has been acquired during their adaptation to the physical environment. More or less recently, certain taxa have expanded their range and have locally become sympatric with ether mem- bers of the group. The taxa geographically intermediate between maculatus and isolepis, viz. femoralis and fordi, are especially exposed to competition. It is possible that each of these has independantly undergone selection for sexual precocity, and that their reduction in pores and ventral black are concomitants of this precocity. References Glauert, L. (1961). — "A Handbook of the Lizards of Western Australia.’* (W.A. Naturalists’ Club: Perth). Gray. J. E. (1845). — “Catalogue of the Specimens of Lizards in the Collection of the British Museum (Brit. Mus.: Lond.). Loveridge, A. (1934). — Australian Reptiles in the Museum of Comparative Zoology, Cambridge, Massa- chusetts. Bull. Mus. Comp. Zool. 77 (6). Lucas, A. H. S.. and Frost, C. (1896). — Reptilia m “Re- port on the Work of the Horn Scientific Ex- pedition to Central Australia.” (Dulau: Lond.) . 54 6. — The Field Oecurrence of the Mt. Padbury Meteorite By W. H. Cleverly* Manuscript accepted — 16th February, 1965 Fragmentation of the Mt. Padbury meteorite, a newly discovered mesosiderite. probably oc- curred at the penultimate stage of atmospheric flight, distributing fragments over a very small impact ellipse, the larger fragments being em- bedded several inches in the soil. The con- centration of nearly 99% of the material in a circle whose area is only 0.4 that of the whole ellipse is notable and suggests that the line of flight was nearly vertical with only some minor fragments scattered further out when fragmentation occurred. Further fragmenta- tion probably occurred on impact, and subse- quently the material suffered considerable weathering and redistribution of the resulting fragments by secondary processes. This may indicate that the fall occurred many years ago — possibly centuries. but the unstable mineral lawrencite in the meteorite could have produced the same effect in a very short time. Location The find was made on Mt. Padbury station, the homestead of which is situated 68 miles north-north-west of Meekatharra. The site of find was 9 miles from the homestead on a bear- ing slightly south of east, at a point on 118® 15' east longitude and 25® 42' south latitude. The mountain from which the station takes its name is closely visible to the ncrth-north-east. * Department of Geology, School of Mines of Western Australia, Kalgoorlie. The country about the site is, perhaps, best described as “mulga plain”. It carries stunted mulga and is very poor pastoral country (Pig. 1). There is a stony soil cover and a tough fer- ruginous hardpan occasionally outcropping, but more often at depths up to one foot. Occasion- ally this hardpan is exposed at the surface or occurs as detached floaters. History of the find and its recovery This new meteorite find was made on 12th March, 1964 by Mr. W. C. Martin, senior part- ner in the Mt. Padbury Pastoral Company, dur- ing the course of mustering sheep. Despite protests from a companion that the material was valueless, Mr. Martin took some pieces from one of the “outcrops” back to the station home- stead. There his partner, Mr. A. H. Bell, attempted to “dolly” one of the pieces for gold and also applied an oxy-acetylene flame to the material, causing a white metal to run from it. Still puzzled, Mr. Bell forwarded pieces to the School of Mines in Kalgoorlie for identification. There they came to the attention of the writer and were identified as a mesosiderite, a stony- iron meteorite of rare type. Mr. Bell described the occurrence in general terms as possibly several hundredweights of material, occurring in four or five small outcrops Fig. 1. Mass 1. — The largest piece of the Mt. Padbury meteorite (195 lb.) in situ, with de- tached fragments in foreground. Mr. W. C. Martin, the finder (on left) and his partner, Mr. A. H. Bell, who called attention to the material. Note the stony, almost “gibber-plain” surface and the poor mulga vegetation. 55 Fig. 2 ’. — Side elevations of three major fragments showing weathering effects. The upper surfaces of all are original dimpled surfaces. Top left: spheroidal core after loss of up to 2 in. by exfoliation from the underside. Bottom right: upward wedging cracks sub-divide the mass into individuals which weather spheroidally. Top right: cracking and weathering advanced, a well-defined spheroid in the middle. The two large specimens are one foot v;ide. rising above the natural cement and spread over an area about thirty yards square. Knowledge of the manner of occurrence of meteorites has frequently been lost in the past through thoughtless collecting, and this occurrence ap- peared from Mr. Bell’s description to be un- spoiled and to provide an ideal opportunity for accurate recording of details of distribution of the pieces. In response to a request from the writer, the finders removed no further material and it was subsequently possible to record and collect from the outcrops with due care. Each outcrop was found to consist of a very badly weathered and disintegrating boulder of meteorite, embedded in the soil. Distribution and state of the material The larger pieces of meteorite were embedded to about the depth of the ferruginous hardpan, but no worthwhile material was recovered from below ground level. Each large mass showed mor^ or less snheroidal weathering like boulders of d'lerite. The material in the ground was completely decomposed and consisted of de- tached exfoliated slabs or scales of “onion-skin” type; the term “iron shale” is sometimes applied to decomposed material of this nature. It was collected with the intention of determining the extent and weight of each mass rather than with any hope that it might be of value for petrographic study. Each outcrop could be lifted off its weathered underpaid. From the underside of some masses cracks extended up- ward dividing them into spheroidally weather- ing parts (Fig. 2). Mass 3 was so frail as to fall apart at a touch. In contrast, the upper surfaces of Masses 2 and 3 were dimpled and appear to be original surfaces; this dimpling on Mass 2 is shown in Fig. 3. Portion of the lateral surface of Mass 1 also showed this struc- ture. This survival is noteworthy because lower parts embedded in the ground were surrounded by up to 4" of iron shale. That thickness might owe something to swelling consequent upon oxidation and hydration. Of Mass 4, initially about 22 kg. in weight, the largest surviving fragment weighed only 0.7 kg., and the remainder was divided into hundreds of small fragments. The great bulk of the curving or irregular fragments were quite useless as specimen material but the rare rounded or ovoid pieces from this or other sources were of relatively high specific gravity and fresh internally. Such rounded pieces are the innermost cores of the speroidally weathered masses. 56 Fig. 3. — Mass 2 disintegrated by secondary agencies and drifted apart. The largest pieces show the dimpled upper surface. There are quite a number of fragments in the foreground but the small pebbles are a soil constituent not meteoritic. Hammer length, 1 foot. Following detachment of fragments as a re- sult of weathering, there had been some second- ary distribution of material by soil creep, rainwash or sheet floods, resulting in small eluvial or alluvial trails of fragments. There is also a possibility that rare, unusually wide floods of the Murchison river might have covered this point. Fragments which could be referred on the basis of proximity and slope to a particular parent mass were mapped and recorded as the eluvium of that mass (Figs. 4 and 5). Thera remain some minor fragments of no evident parentage and these probably belong to at least three categories. Some groups of frag- ments could have resulted from disintegration of other original masses of small dimensions. “Mass” 7 (Fig. 5) is a group of 15 such frag- ments, the largest 0.3 kg. in weight, envisaged as parts of a hypothetical mass of about 1.2 kg. A fragment weighing 0.35 kg. found between Masses 1 and 6 was larger than any fragment in the eluvial trail of Mass 1 but had the ex- foliated form and was quite isolated. Distribu- tion of such fragments by human agency is a possibility: the site is close to an area which was thsroughly prospected for manganese. Finally, some pieces may be original and un- related to distribution effects of weathering. A piece in this category is an ovoid one weighing 0.6 kg. and occurring 20' in direction 170° from Mass 5 without accompanying debris. Frag- ments not known to be related to a parent mass are listed in the table of weights as “aberrant”. Reasons for the primary distribution of mass When a meteorite is found as a number of related fragments, there are at least four pos- sible reasons: (a) Fragmentation pre-dating entry to the earth’s atmosphere. Fig'. 4. — Eluvial spread from Mass 1 looking westerly towards the source. Fragments have been placed upon white sample bags. No attempt has been made to mark all the very numerous fragments near the source. 57 N i FEET Fig. 5. — Distribution of the fragments of the Mt. Padbury meteorite. Larger masses are numbered and their eluvium is included within the broken lines. The very numerous fragments of Masses 1 and 2 are not shown individually. The circle is of 120 feet diameter, the ellipse 130 feet x 275 feet. (b> Fragmentation consequent upon heat- ing and the stresses of atmospheric flight. (c) Fragmentation on impact. (d) Disintegration by weathering and dis- tribution by surface agencies. The first impression in the field was that the major masses formed a triangle of about 90 feet side and their distribution might be due to fragmentation on impact. When further pieces were flagged it was evident that the distribution was roughly elliptical. This is the character- istic distribution of a shower of meteorite frag- ments. If a group of fragments is visualized travel- ling earthward and gradually dispersing in flight then a cone of paths is formed. Gener- ally, when the axis of the cone is not vertical, the earth’s surface forms an elliptical impact area. In theory, the special case of a vertical axis and a circular impact area is also a possi- bility but such a special case will be a rarity. Ellipses of meteorite distribution are known varying from rough patterns to near-perfect ones. The plot of the distribution (Fig. 5) illustrates the pattern recognised in the field. After allowing for the reconcentration of eluvial frag- ments with their parent masses, 94.4% of the weight of the collected material is represented by Masses 1 to 4 at the corners of a triangle and 98.7% lay within a circle of 120 feet dia- meter: the whole of the material can be included within an ellipse measuring 130' x 275', an area of 0.6 acre. The stony soil and tough ferruginous hardpan would combine with the somewhat brittle nature of the meteorite itself to effect breakage on impact, and fractures will have been initiated by atmospheric flight stresses. However, the grouping of the fragments within a small area is the only real evidence for breakage on impact rather than before impact. The ellipti- cal distribution favours arrival of the meteorite as a number of pieces and another point in favour is that the larger masses (which are numbered in order of decreasing weight) were embedded according to their weights. Mass 1 was embedded more than 8", Masses 2 and 3 more than 6", Mass 4 several inches, Mass 5 for 2" and Mass 6 not at all. A depth of 2" to 4" is to be added for the three largest masses, being the thickness of the thoroughly weathered shells. These depth figures are too consistent to be fortuitous and, in any case, there appears to be no reason why fragments scattered on impact should be embedded to such depths. The very small size of the ellipse suggests fragmen- tation at a very late stage in flight: more com- monly, ellipses are measurable in miles. It is a possibility that the dimpled surfaces present on the three biggest masses are parts of the original surface because at least two or three of the dimples coincide with large silicate grains and they could therefore be an ablation effect. They might represent parts of a pos- terior surface, somewhat protected during oriented atmospheric flight. However, the state of the material is such that it is impossible to attempt a reconstruction and this idea there- fore remains as speculation. 58 Table 1 Weights (kilograms) of Mt. Padbury Meteorite Mass 1 . .. Main mass W'eathered . kimberleyensis, is also set well in from the lateral margin of the skull. The fossae for the vomerine and palatine tusks are similarly placed relative to the choana. The conjoined lower jaw rami form a bulbous symphysial shelf which fits within the margins of the upper jaw. The ventral borders of the rami are sharply keeled posteriorly. The retro- articular processes are similar in relative size to those of D. kimberleyensis and the outlines of the glenoid facet and the adductor fossa cor- respond closely. These characters of the lower jaws also indicate a close relationship of the African and Australian genera. A number of characters of the skull and lower jaw distinguish R. capensis from D. kimberley- ensis. R. capensis has a longer and narrower antorbital region of the skull. The external naris is set further back from the snout tip. The anterior part of the postparietal is indented by the posterior margin of the supratemporal. The supraorbital groove is less well defined. The left side of the antorbital fragment of R. capensis appears to be broken off just anterior to the forward edge of the interpterygoid vacuity. The portion of the palate anterior to the vacuity is relatively much longer than it is in D. kimber- leyensis. As restored, the interpterygoid vacu- ities of R. capensis are proportionately much smaller than those of D. kimberleyensis. In R. capensis the orbit lies above and lateral to the anterior part of the interpterygoid vacuity; in D. kimberleyensis the orbit is above and lateral to the center of the vacuity. Ihe lower jaw of R. capensis is deeper and more massive than that of D. kimberleyensis. The posterior edge of the retroarticular process is notched whereas it is pointed in D. kimber- leyensis. Romer (1947, p. 190, fig. 34) referred Micropo- saurus casei (Haughton, 1925, pp. 253-256, figs. 17-18) to R. capensis and made a composite re- storation of the two skulls. He considered the common horizon and the similarity of the skull shape and size to be indicative of identity and the lateral orbit and naris of R. cape?isis and the posteriorly placed quadrates of M. casei to be the results of differential distortion. By including the interotic fragment of R. capensis, a restoration of the dorsal surface which differs materially from Romer’s composite has been produced. The broad cultriform process and the extensive exoccipital-pterygoid suture suggest that R. capensis. unlike Microposaurus caseU is unrelated to the trematosaurs. M. casei displays such typical trematosaur characters as a greatly extended postorbital region, a very long parasphenoid-pterygoid suture, a narrow cultriform process and a double anterior palatal vacuity. The parasphenoid probably extensively underplates the exoccipitals as in other trematosaurs. "Trematosaurus” kannemeyeri (Broom, 1909, pp. 270-271) has tentatively been referred to R. capensis by Kitching (1957, p. 81). Broom did not illustrate the specimen but von Huene (1920. fig. 8) contributed an interpretation restored from the measurements cited by Broom. The skull is very narrow in the orbital region. This fact alone suggests that '‘T.” kannemeyeri should be retained in the Ti'ematosauridae. The generic designation, however, is probably not correct. Plaster and rubber replicas of the holotype skulls of P. erici and P. wimani, generously donated by the Paleontological Institute, Uni- versity of Uppsala, have augmented the excel- lent review of the genus by Nilsson (1946) and have made direct comparisons with the Australian genus possible. Nilsson’s careful reconstructions have been of great value in this work. Peltostega has long been a problematic genus. It has been variously classified in the Tremato- sauridae, the Metoposauridae, the Sclero- cephalidae and in its own family. Peltostegidae. Wiman (1916, pp. 210-216, text figs. 1-3, pi. XV, figs. 1-3, pi. XVI, fig. 1) first described the type and only specimen of P. erici. It was recovered from the Triassic Posidonomya beds of Mt. Anderson, Spitzbergen with Aphaneramma rostratum and Lyrocephalus euri by the third Swedish Spitzbei’gen expedition in 1915. The specimen is the posterior half of a large skull (33-34cm. in median length as restored by Nilsson. 1946). The postorbital part of the skull roof, the entire occipital surface and part of the palate are preserved. The original restora- tions of the missing portions of the skull by Wiman are too foreshoi’tened in the antorbital region. The battered condition of the palate (Wiman, pi. XV, fig. 2) led Romer (1947, p. 189) to the conclusion that the structure identi- fied by Wiman as the cultriform process of the parasphenoid was actually the ventral surface of the overlying sphenethmoid. He suggested that the cultriform process had been a narrow structure as in trematosaurids. Nilsson, how- ever, identified the structure as a cultriform process and it is probable that this interpreta- tion is correct. Nilsson (pp. 4-34) described a second species, P. wimani, from the same horizon in Spitz- bergen. The holotype, another occipital por- tion. was found at Mt. Wallenberg. A referred skull roof fragment was found at Mt. Congress. The holotype of P. wimani is even larger than that of P. erici and differes from it in its com- paratively delicate ornament of the dorsal sur- face, proportionately greater exoccipital con- dyles and in details of the bones of the skull roof. Many characters relate Peltostega to Delta- saurus. Both genera have broad triangular skulls and lateral orbits facing outward as well as upward near the middle of the lateral mar- gins. The parietal foramen of P. erici is close behind a line between the posterior borders of the orbits as in Z>. pustulatus. Angular otic notches are present in D, kimherleyensis and in both species of Peltostega. The frontal bone is excluded from the orbital border in P. erici and in both species of Deltasaurus. The dermal sculpture of Peltostega corres- ponds closely to that of D. kimherleyensis. The resemblance of sculpture to that of D. pustula- tus is less obvious, but as noted above, the sculp- ture of this species is reasonably regarded as a modification of that of D. kimherleyensis. The skull roof bones of the Peltostega specimens are uniformly pitted over most of their surfaces. These pits are small and equidimensional as in D. kimherleyensis but are somewhat more dis- tinct as the concentric ridges of sculpture have a higher relief than in the Australian form. As in D. kimherleyensis there is a basic pattern of radiating ridges which carries small rounded nodes at points of bifurcation and at points of junction with the concentric ridges. The peri- pheries of certain bones of P. wimani, notably the anterior edges of the left squamosal and left quadratojugal, have exceptional develop- ment of the radiating ridges and the pits are poorly developed or lacking due to imperfect development of the concentric ridges; similar unpitted areas also occur on the type skull of D. kimherleyensis as noted in the descriptive section. Common characters of the posterior part of the palatal surface include broad cultriform pro- cesses and broad corpi of the parasphenoid bones and broad posterior margins of the interpterygoid vacuities. The occipital surfaces of P. erici and D. kim- herleyensis are somewhat similar. The supra- occipital fenestrae are round and are larger than the foramina magni. The cheeks slope down sharply from the flat skull table. The descending processes of the squamosal are sep- arated from the ascending processes of the pterygoids by long, narrow palatoquadrate fis- sures. Peltostega has a longer and narrower antor- bital region (as reconstructed by Nilsson), and the posterior margin of the skull table is more strongly excavated. The lateral line system is not as conspicuously developed in Peltostega; the supraorbital groove is not continuous medial to the orbit of P. erici; and the supratemporal groove of this species is shorter: P. wimani seemingly lacks an occipital groove while that of P. erici is discontinuous. Both species of Peltostega have a posterior bifurcation of the jugal groove not present in D. kimherleyensis. The palatal shagreen of P. erici is apparently quite rudimentary although this may be the re- sult of incomplete preservation. The skull has a few small denticles on the pterygoid bone and a row of larger teeth on the posterior part of the left ectopterygoid; these latter are probably also part of the shagreen as they are too irre- gularly orientated and too medial to be part of a palatal tooth row. The basicranial regions of Peltostega erici and D. kimherleyensis show minor differences in construction. The parasphenoid-pterygoid suture is more lateral in D. kimherleyensis. The quadrate ramus of the pterygoid is broader and shorter in P. erici. The Australian species lacks the large rounded excavation of the posterior margin of the palate between the exoccipital and quadrate condyles. P. erici has a small projection of the pterygoid on the posterior margin of the interpterygoid vacuity; this is not present on the outline of the pterygoid which is well-preserved on U.C.M.P. no. 61063. The only notable difference of the occipital surface is the construction of the subotic process of the exoccipital. This is a broad, flat, obliquely orientated structure in D. kimherley- ensis. It is a short vertical bar just ventro- lateral to the condyle in P. erici and P. wimani. The incompleteness of the Peltostega and Rhytidosteus material has made it impossible to determine if either is more closely related to Deltasaurus than the other. Deltasaurus and Rhytidosteus share a great intensification of the shagreen of the palate, but this may also have been true of Peltostega. The dermal sculpture is basically similar in the three genera; that of Rhytidosteus is somewhat coarser than that of Deltasaurus and that of Peltostega is somewhat finer. Peltostega is perhaps the most primitive genus in one respect: it lacks a suture between the ex- occipital and the pterygoid on the palate. Deltasaurus has a short exoccipital-pterygoid suture and Rhytidosteus has an extensive one. If the Rhytidosteidae follow the trend found in the Capitosauridae and Brachyopidae des- cribed by Watson (1919, 1951, 1956), this sutural contract would be progressively en- larged through evolution. 86 Comparisons of the Rhytidosteidae with other families The Rhytidosteidae superficially resemble several genera of the Trematosauridae, par- ticularly Lyrocephalus, in the triangular shape of the skull, in the relatively small interptery- goid vacuities and in the positions of the orbits near the levels of the sagittal midpoints. Nilsson Q946) distinguished the Tremato- sauridae from Peltostega by the elongation of the skull between the orbits and the parietal foramen, by the deeper occiput, by the exceed- ingly high narrow cultriform process and by the great length of the corpus of the para- sphenoid, which reaches back to underplate the exoccipital condyles. These characters distin- guish the Trematosauridae from the Rhytido- steidae. The double anterior palatal vacuities of trematosaurids is an additional difference. Nilsson noted the following characters of the Capitosaurids which distinguish this family from Peltostega: the otic notch is deeper and narrower; the orbits are placed in the posterior portion of the skull roof; the cultriform pro- cess is not wide and it often has a keeled ventral surface; the pteroccipital foramen is not as well defined; and there is a high sharp oblique ridge on the ascending process of the pterygoid (lacking, however, in Paracycloto- saurus). These are also valid differences of this family from the Rhytidosteidae. Caoitosaurids and Rhytidosteids share vertical occiputs and differ from the tremato- saurids and metoposaurids in this respect. The retroarticular process is similar in shape in the two families. Some of the rhytidosteid genera parrallel characters found in certain capito- saurids. Deltasaurus has a single anterior palatal vacuity, similar in shape to that of ParatosauTus and Cyclotosaurus. There is a palatoquadrate fissure between the squamosal and the pterygoid on the occiputs of Delta- saurus, Peltostega and Parotosaurus. Delta- saurus and Rhytidosteus have an exoccipital- pterygoid suture as do Cyclotosaurus and Para- cyclotosaurus. Nilsson remarked on the superficial similar- ities of Peltostega and the metoposaurids in the lateral positions of the orbits and the broad cultriform processes. The metoposaurid orbit, however, does not face dorsolaterally as is the case in Peltostega and the cultriform process is narrow between the vomers in metoposaurids. The metoposaurid cultriform process passes broadly and evenly into the corpus of the para- sphenoid; the corpus expands suddenly in Peltostega. Nilsson distinguished the metoposaurids from Peltostega by the anterior positions of the orbits, by the enlarged postorbital regions, by the enlarged paraquadrate foramina and by the re- duced posttemporal fossae of this family. These distinctions are also valid for the other rhyti- dosteid genera. Additional metoposaurid characters not found in the rhytidosteid genera are occiputs which slope down and back from the skull roof and complex (and poorly understood) relations of the pterygoids and squamosals on the cheek portions of the occiputs. The Brachyopidae share only general labyrin- thodont characters with the Rhytidosteidae. The Rhytidosteidae are a conservative family and retain many primitive rhinesuchid- lydekkerinid conditions. The palatal shagreen, the locations of the orbits near the centre of the skull, the broad cultriform processes, the short parasphenoid-pterygoid sutures and the locations of the parietal foramina close behind the orbits distinguish the Rhytidosteidae from contemporary early Triassic families and also suggest derivation of this family from the Rhinesuchidae, Lydekkerinidae or Uranocentro- dontidae. The undivided anterior palatal vacuity of Deltasaurus kirriberleyensis and the palatoquadrate fissure of both D. kiviber- leyensis and Peltostega erici are probably also conservative features. All late Permian rhine- suchoids in which the snout region is preserved have a single anterior palatal vacuity. Lydek- kerina huxleyi (Watson, 1919, fig. 4) and Limnoiketes palundinatans (Pai’rington. 1948, fig. 6B), show palatoquadrate fissures. The specialized characters of the skull readily distinguish the Rhytidosteidae from the Rhine- suchidae and the Lydekkerinidae and necessi- tate a superfamial separation. The rhyti- dcsteids have narrower snouts and broader occiputs. The orbits lie near the lateral edge of the skull roof. The sculpture pattern is modified. The quadrate condyles have moved forward relative to the exoccipital condyles. Deltasaurus, Peltostega and Rhytidosteus have smaller and more angular otic notches than those found in members of the late Permian families. The palatal exposures of the cultri- form processes of the parasphenoids of rhine- suchids and lydekkerinids terminate in sharp angles between the vomers. In most genera of these families this termination lies behind the level of the anterior borders of the inter- pterygoid vacuities. The rhytidosteid cultri- form processes reach far forward between the vomers and have broad conical terminations. Comparison of the Rhytidosteidae with problematic Triassic genera Laidleria Laidleria gracilis, recently described by Kitching (1957, pp. 67-82, figs. 16-19), consists of the palatal surface of the skull, the lower jaws and the articulated partial skeleton, and is preserved on a sandstone slab in the Albany Museum, Grahamstown, South Africa. It was found in the Karroo of the eastern Cape Province. The exact stratigraphic provenance is uncertain but it is probably from the Cynognathus zone as it was found just below the Stormberg rocks and as the matrix re- sembles Cynognathus zone matrix of other localities. Certain features of the skull and lower jaw are suggestive of the Rhytidosteidae: the skull is triangular and the occiput is very broad; the orbit lies near the lateral border of the skull; the postorbital region is short; the cultriform process is broad; and the symphysis is expanded into a broad plate. Many details of skull construction prevent an inclusion of Laidleria in the Rhytidosteidae. 87 The cultriform process of the parasphenoid. although broad, tapei’s to a sharp point anteriorly in contrast to the broad and conical terminations found in Deltasaurus and Rhyti- dosteus. The snout is slender and pointed. In Kitching’s interpretation, an otic foramen replaces the otic notch (a condition paralleled by Tertrema of the Trematosauridae and by Cyclotosaurus and Paracyclotosaurus of the Capitosauridae) . The frontal forms part of the orbital border; palatal shagreen is either absent or is greatly reduced (a small pitted area on each pterygoid may represent shagreen); thex’e is a palatal tooth row (definitely absent in Deltasaurus)', and the choana lies close to the lateral edge of the skull. The most profound differences of Laidleria from the Rhytidosteidae occur on the occipital surface. The Laidleria occiput is very low and is of uniform height from corner to corner. The descending process of the squamosal has sutural contact with the ascending process of the pterygoid. The exoccipital apparently lacks the processes lamellosus and basalis (a condi- tion also found in Indobrachyops panchetensis (von Huene and Sahni, 1958) and the foramen magnum cannot be distinguished from the supraoccipital and basioccipital foramina. Three characters of Laidleria are unique among the Labyrinthodontia: the dentary teeth are very large, few in number and widely spaced; the contact of the pterygoid and ectcpterygoid in the palate is peculiar — a rounded posterior lapet of the ectopterygoid underlies the anterior extent of the palatal ramus of the pterygoid; the tabular has two sutures with the exoccipital — the usual suture on the paraoccipital bar and an additional suture medial to the posttemporal fossa. Useful characters for systematic purposes are perhaps present on the skull roof of Laidleria gracilis which is still embedded in matrix. The nature of the lateral line system, the position of the parietal foramen and most of the sutural relations of this surface are unknown. Kitching (fig. 19B) shows sculpture on the skull peripheries and on the labial surfaces of the lower jaws. The illustration does not depict this sculpture clearly, and comparisons with the sculpture of the rhytidosteid genera cannot be made. The lower jaw symphysis covers the region of the anterior palatal vacuity and the shape of this structure is not known. Laidleria superficially resembles Lyrocep- halus in skull shape but the cultriform process of the parasphenoid is too broad for a trema- tosaur and the parasphenoid lacks the posterior extension which is characteristic of the Trema- tosauridae. The numerous peculiarities of skull construction exclude this African genus from the other Permian and Triassic families. Laidleria is best placed in its own family, Laidleriidae as Kitching has suggested. How- ever, there is greater resemblance to the Rhyti- dosteidae than to other families and it may be provisionally included in the Rhytidosteoidea. Lastiscopus Wilson (1948, pp. 359-361, pi. 59) described the only known specimen of Lastiscopus dis- jurictus, a poorly preserved skull from the late Triassic Dockum Formation of Texas, and pro- visionally placed it in a new family, Lasti- scopidae. Little could be said of its construction. As Wilson noted (p. 360), the high vault of the skull and the large interpterygoid vacuity suggest the Trematosauridae. He also stated: “The extreme posterior edge of the para- sphenoid is missing but there is good evidence in the broken edges of the bone that it at one time covered the ventral surface of the exoc- cipitals.” This condition would also tend to relate Lastiscopus to the Trematosauridae. Two characters of the skull suggest the Rhyti- dosteidae rather than the Trematosauridae. These are the extreme lateral position of the orbit and the short distance between the orbits and the parietal foramen. It is to be hoped that more evidence establish- ing the systematic position of this animal will be found. It is the only non-metoposaurid labyrinthodont of the North American late Triassic and if it proves to be either a trema- tosaurid or a rhytidosteid, it will be the latest known occurrence of its family. For the present the family, Lastiscopidae should be placed as incertae sedis among the superfamilies of the Triassic Temnospondyli. Age relationships The entire collection of fossil vertebrates from the Blina Shale may be considered as a faunal unit through the occurrence of Deltasaurus kim- herleyensis. Skull and lower jaw fragments of this animal were found at all the Erskine Range localities although these are spaced over two square miles of area and more than 70 vertical feet of section. Four jaws of D. kimberleyensis were collected at the B.M.R. Dry Corner locality and one at U.C.M.P. locality V 6044. There is no evidence of evolutionary change or faunal replacement among the collections from the various localities. The uniform lithology of the Blina Shale exposures in the Erskine Range is consonant with the uniformity of the faunal content. There are no intraformational disconformities within the measured section at V 6040 and there is no other evidence to indicate a major interruption of the depositional cycle. No lateral or vertical changes in the sedimentary facies suggest the presence of more than one depositional environ- ment. The microfloras from the Blina and Kockatea Shales establish a general time equivalence for these units within the lower part of the Scythian Stage (Balme, 1963). The presence of species of Deltasaurxis in both formations supports the correlation. Dickins and McTavish (1963) place the portion of Kockatea Shale which contains Deltasaurus pustulatus in the Otoceratan zone. Minor differences in bone sculpture and skull proportions are the only features which distin- guish D. kimberleyensis from the Kockatea species and it is not likely that the stratigraphic position of the Blina fauna is far removed from the Otoceratan zone. 88 Although detailed considerations of the age relationships of the Blina fauna are deferred to a subsequent paper certain conclusions should be stated here. The similarity between Deltasaurus, Rhytidos- teus and Peltostega indicates that the Blina Shale, the Cynognathus zone and the Posido- nomya beds are close in time to one another. These correlations are supported by the occur- rence in the Blina Shale of a brachyopid which is close to Batrachosuchus of the Cynognathus zone and of trematosaurids which are perhaps congeneric with Aphaneramma and Tertrema of the Posidonomya beds. The Cynognathus zone, in turn, may be equated with the Bunter Forma- tion of Germany, and with Zone VI of the Cis- Uralian region of the U.S.S.R. through the oc- currence in these three units of Parotosaurus and Trematosaurus. The Posidonomya beds con- tain ammonites and other invertebrates as well as fish and amphibians and the ammonites place the level in the upper part of the Scythian Stage (Spath, 1930 and Frebold, 1939). In all proba- bility, therefore, the time range represented by the Posidonomya beds, the Blina Shale, the Cynognathus zone, the Bunter Formation and Zone VI is contained by the Scythian Stage. Watson (1942 and 1957) and Romer (1947) considered the Lystrosaurus zone to be the earliest Triassic level in the Beaufort Forma- tion of South Africa. The temnospondyls of this zone, however, are rhinesuchcids, a group which probably contains the ancestors of the younger capitosauroids (Watson, 1962) and rhytidos- teoids. The cranial structure of the rhinesu- choids is notably more primitive than that found in the younger superfamilies, indicating an appreciable time lapse from the Lystrosaurus zone to the Cynognathus zone and the Blina Shale. The position of the Blina fauna in the lower part of the Scythian stage and of Delta- saurus pustulatus in the basal Otoceratan zone suggests that Cynognathus zone is the earliest Triassic level in the Beaufort Formation and that the Lystrosaurus zone is of late Permian age. Acknowledgments The field program in northwestern Austra- lia was sponsored by the Museum of Paleon- tology, University of California and the Western Australian Museum. Dr. R. A. Stirton, Director of the Museum of Paleontology, arranged finan- cial support and guided the preliminary plan- ning. Dr. W. D. L. Ride, Director of the Western Australian Museum, supervised’the procurement and transport of vehicles and supplies and later arranged the shipment of the fossil collections to the University of California. Each institu- tion provided the use of a vehicle and trailer. The Museum of Paleontology defrayed costs of food and equipment and provided funds for travel. The Western Australian Museum pro- vided salaries for two members of the party. The field party is indebted to British Petro- leum Kwinana Ltd. and in particular to Mr. L. R. Gascoine for the provision of fuel and all service costs at BP garages for the vehicles and to Western Australian Petroleum Pty. Ltd. for the use of a storage building in Derby. It is a pleasure to acknowledge the kind hos- pitality of Mr. and Mrs. P. Slater of Derby, Mr. and Mrs. W. W. Henwood of Calwynyardah Sta- tion, Mr. W. W. Henwood Jr. of Blina Station, and Mr. and Mrs. K. C. Rose of Liveringa Station. The success of the field program is largely a reflection of the experience and leadership of Dr. C. L. Camp. Mr. Duncan Merrilees, Dr. Ken- neth G. McKenzie and Dr. W. D. L. Ride aided greatly in the collection of fossil vertebrates. I wish to thank: Dr. W. D. L. Ride, Director of the Western Australian Museum and Mr. J. M. Rayner, Director and Dr. N. H. Fisher, Chief Geologist of the Bureau of Mineral Resources, Geology and Geophysics for the loan of fossil collections, and the British Museum of Natural History and the Uppsala University Paleontology Institute for providing replicas of type speci- mens. I appreciate the many helpful suggestions made by Drs. Joseph T. Gregory, Ruben A. Stirton, Robert C. Stebbins and Samuel P. W^elles who critically read this manuscript. Mrs. Mary McDonald accomplished the difficult pre- paration of much of the collection. Drawings of the fossil specimens are the wcrk of Mr. Owen J. Poe. I am particularly grateful for the advice, as- sistance and encouragement offered by Dr. Joseph T. Gregory under whose direction this research was conducted. References Balme, B. E. (1963). — Plant Microfossils from the Lower Triassic of Western Australia. Palaeontology 6(1): 12-40. Broom. R. (1909). — The fossil fishes of the Upper Karoo beds of South Africa. Ann. S. Afr. Mus. 7; 251-269. Brunnschweiler, R. O. (1954). — Mesozoic stratigraphy and history of the Canning Desert and Fitz- roy Valley, Western Australia. J. Geol. Soc. Aust. 1: 35-54. Bystrov, A. P. (1935). — Morphologische Untersuchungen der Dekknochen des Schadels der Wirbeltiere. I. Mitteilung. Schadel der Stegocephalen. Acta Zool., Stock. 16: 65-141. Camp, C. L. (1963). — Journey through North-Western Australia and Central Australia in search of Fossil Vertebrates. Western Australian Museum Annual Report, 1S60-61 : 27-32. Case, E. C. (1946). — A census of the determinable genera of the Stegocephalia. Trans. Amer. Phil. Soc. 35: 325-420. Dickins, J. M., McTavish, R. A. and Balme, B. E. (1961). — The Beagle Ridge Bore. Australasian Oil & Gas Journal 7(4): 20-21. (1963). — Lower Triassic marine fossils from the Beagle Ridge (B.M.R. 10) Bore, Perth Basin, Western Australia. J. Geol. Soc. Aust. 10(1): 123-140. Frebold, H. (1939). — Das Festungsprofil auf Spitzbergen. V. Stratigraphie und invertebratenfauna der Alteren Eotrias, Nebst Beschreibung anderer Vorkommen in Spitzbergen. Skr. Svalb. Og. Ishavet 77: 1-58. Guppy, D. J., Lindner. A. W., Rattigan, J. H. and Casey, J. N. (1958). — The Geology of the Fitzroy Basin, Western Australia. Bull. Bur. Miner. Resour. Aust. 36: 1-116. Haughton, S. H. (1925). — Investigations in South Afri- can fossil reptiles and Amphibia. 13. De- scriptive catalogue of the Amphibia of the Karroo System. Ann. S. Afr. Mus. 22: 227- 261. 89 Huene, F. von (1920). — Gonioglyptus, ein alttriassischer Stegocephale aus Indien. Acta. Zool., Stock. 1: 433-464. (1948). — Short review of the lower tetrapods. Roy. Soc. S. Afr. Spec. Publ., Robert Broom Comm. Vol.: 65-106. and Sahni, M. R. (1958). — On Indohrachyops panchetensis gen. et sp. nov. from the Upper Panchets (Lower Trias) of the Raniganj Coalfield. Af on. Pal. Soc. India 2: 1-14. Jaekel, O. (1922). — Neues liber Hemispondyla. Paldont. Z. 5: 1-25. Kuhn, O. (1933). — Labrinthodontin. Fossilium Catalogus, Animalia, Pars 61: 1-14. — (1960). — Amphibia. Supplementem I ad Partes 61 et 84. Fossilium Catalogus I: Ani- malia. Pars 97: 1-164. Kitching, J. W. (1958). — A new small stereospondylous labrinthodont from the Triassic beds of South Africa. Palaeontologia Africana, 5 (1957): 67-82. McKenzie, K. G. (1961). — Vertebrate Localities in the Triassic Blina Shale of the Canning Basin. Western Australia. J. Roy. Soc. W. Aust. 44 (3): 69-76. McWhae, J. R. H.. Playford. P. E., Lindner, A. W., Glenis- ter, B. F. and Balme, B. E. (1958). — The Stratigraphy of Western Australia. J. Geol. Soc. Aust. 4: 1-161. Nilsson, T. (1943). — On the morphology of the lower jaw of Stegocephalia with special reference to Eotriassic stegocephalians from Spitzbergen, I. Descriptive part. K. Svenska VetenskAkad. Handl. (3) 20(9) : 1-46. (1946). — On the genus Peltostega Wiman and the classification of the Triassic Stegocepha- lians. K. Svenska VetenskAkad. Handl. (3) 23(3) : 2-55. Owen. R. (1884). — On a labyrinthodont amphibian (Rhytidosteus capensis) from the Trias of the Orange Free State, Cape of Good Hope, ^uarf. J. Geol. Soc. Lond. 40: 333-339. Farrington, F. R. (1948). — Labyrinthodonts from South Africa. Proc. Zool. Soc. Lond. 118: 426-445. Romer, A. S. (1945). — “Vertebrate paleontology” (Chi- cago, Univ. Chicago Press). (1947). — Review of the Labyrinthodontia. Bull. Mus. Comp. Zool. Harv. 99: 3-352. Save-Sdderbergh, G. (1935). — On the dermal bones of the head in labyrinthodont stegocephalians and primitive Reptilia with special reference to Eotriassic stegocephalians from East Greenland. Medd. Gronland 98(3) : 1-211. Spath, L. F. (1930). — The Eotriassic Invertebrate Fauna of East Greenland. Medd. Gronland 83(1): 5-90. Veevers, J. J. and Wells, A. T. (1961). — The Geology of the Canning Basin, Western Australia. Bull. Bur. Miner. Resour. Aust. 60: 1-323. Watson. D. M. S. (1919). — The structure, evolution and origin of the Amphibia, The “orders” Rachitomi and Stereospondyli. Phil. Trans. (B) 209: 1-72. (1951). — “Palaeontology and modern biology” (New Haven, Yale Univ, Press). (1956), — The brachyopid labyrinthodonts. Bull. Brit. Mus. (Nat. Hist.), Geology 2(8): 318-391. (1957). — The two great breaks in the history of life. Quart. J. Geol. Soc. Lond. 112: 435- 444. (1962). — The evolution of the Labyrintho- donts. Phil. Trans. (B) 245: 219-265. Wiman, C. (1916). — Neue Stegocephalenfunde aus dem Posidonomyaschiefer Spitzbergens. Bull. Geol. Instn. Univ. Upsala 13: 209-222. Wilson, J. A. (1948). — A small amphibian from the Triassic of Howard County, Texas, J. Paleont. 22: 359-361. Woodward, A. S. (1932), — In Zittel, K. A. von, “Text- book of Palaeontology”, Second English edi- tion, translated and edited by C. R. Eastman and revised with additions by Sir Arthur Smith Woodward. (London, Macmillan & Co.) 90 9. — The identity of Mus burtoni Ramsay, 1887 (Rodentia, Muridae, Melomys) from the neighbourhood of Derby, Western Australia by J. A. Mahoney* * Manuscript received I6th February 1965; accepted 27th April 1965 Abstract The identity of Mus burtoni has remained obscure following its erection by Ramsay in 1887. The type skull and mandible of this species have now been found in the Australian Museum. Sydney. Mus burtoni is placed in Melomys Thomas, 1922 where it is predated by Melomys cervinipes (Gould, 1852) and Melomys rufescens (Alston, 1877). It is doubtfully conspecific with the former species and distinct from the latter. Melomys burtoni (Ramsay, 1887) is tentatively accepted as a valid species pending more satisfactory knowledge of the genus. Introduction Mus burtoni was erected by Ramsay (1887b) for a single rodent specimen received from Derby, north Western Australia. The fragmen- tary skull and mandible and the feet are illustrated in the original description and skin (or flesh) dimensions listed. The sex is not stated nor is a catalogue number or depository noted. No further specimens have been recorded for Mus burtoni and subsequent authors have accepted Ramsay’s name for it without comment (Ogilby, 1892 p. 107), indi- cated that its identity is obscure (Longman, 1916 p. 34; Iredale and Troughton, 1934 p. 75; Ellerman, 1941 p. 214), or ignored it (Tate, 1951). The type skull and mandible of Mus burtoni have now been found in the Australian Museum, Sydney. These have been examined and Mus burtoni is placed in Melomys Thomas, 1922. Melomys burtoni (Ramsay, 1887) Holotype: Australian Museum No. S427, fragmentary skull and mandible. The skin of the holotype is not registered under this number and has not been found. The holotype was received from T. H. Bowyer-Bower Esq.,v from Derby, north Western Australia. Ramsay does not clearly indicate that the specimen was collected at Derby consequently its locality is recorded here as the neighbourhood of Derby, Western Australia. The moderate wear on the molars of the holotype suggests that it is a young adult. The holotype skull and mandible are regis- tered in the Australian Museum “S” catalogue as “Rat Skull of”. Prior to registration this specimen belonged in the Museum’s “old collection”. Collector, date of collection, and locality data are not recorded in the catalogue * Department of Geology and Geophysics, University of Sydney. • Thomas Henry Bowyer-Bower accompanied by a taxi- dermist. Walter Burton, left Sydney early in 1886. and collected in north-western Australia, up the Pitzroy River as far as Mount Anderson (see fig. 1). He then collected on Thursday Island, off Cape York, and at Palmerston in Northern Territory, where he died of typhoid fever on December 22, 1886. (auth. Whitten. 1954, Pt. 2. p. 71). for it, nor is any indication given there that it is a type. Mr. G. P. Whitley, recently retired Curator of Fishes, Australian Museum, has examined the catalogue entry for the holotype and believes it to have been made by Mr. E. R. Waite (personal communication). Registration of this specimen, dated August 12, 1893, was made during Dr. E. P. Ramsay’s Curatorship of the Australian Musuem (September 22, 1874 to December 31, 1894). The skull and mandible are accompanied by an unattached label ^pill- box top), shown in figure 2, M. The initials “G. H. B.” at the top of the label are the same as those of G. H. Barrow, the artist who pro- duced the illustrations for Ramsay’s paper on Mus burtoni. The words following “6” at the bottom of the label are difficult to read but may be “times abt.”— Ramsay’s published figures of the skull and lower molars of the holotype of M. burtoni are noted by him as being 6 times natural size. “B Bower” and “Derby” are written in pencil on the right hand side of the label. Two Australian Museum numbers, S427 91 92 and 427, are also included on it. One, at least, of these numbers was apparently written on the label by Waite at the time of registration while the second number may have been added later. Other writing on the label agrees with that of Ramsay (see Whittell, 1954 Pt. 1 Pis. 29, 30 for a sample of Ramsay's hand- writing) . The Australian Museum skull and mandible (S427) agree well with Ramsay’s figures of Mus burtoni if allowance is made for some latitude in their production, and if the illustra- tion of the skull is a mirror image of the original. Ramsay’s figure 2 shows the presence of a left third molar but this is the only tooth now missing from S427. This must have been missing from the skull when Ramsay had it, otherwise Ramsay could not have described it as having a portion of the dentition. One further rodent, the holotype of Hapalotis boweri Ramsay, 1887 from “North West Aus- tralia”, is recorded by Ramsay (1887a) as having been received by him from Bowyer-Bower. The skull of this specimen cannot be found and may not have been sent to Ramsay (see Ramsay, 1887a p. 1154). It is not figured in his paper on H. boweri and does not comply with data on the label accompanying S427. The above details, considered together, justify recognition of S427 as the holotype skull and mandible of Mus burtoni Ramsay. Specimens examined: Holotype. External characters; Ramsay notes that the chief characteristic in this species is its re- markably woolly and soft fur, and uniform colour. His account of the external characters is reiterated below but with the measurements, given by him in inches, converted to millimetres: General colour of a uniform dull ashy-grey or mouse- colour. fur dense, close, thick and soft, of one kind, almost woolly, slightly browner above than on the under surface, which is of a light grey tint; head rather short; ears moderate; tail naked, not quite the length of the body; whiskers black reaching to behind the ears; from snout to eye, 17.8 mm; from snout to ear. 30.5 mm: length of ear 16.5 mm, greatest width 11.4 mm; forearm, 17.8 mm; hand, 11.4 mm; hind foot, 25.4 mm; tail, 104 mm; total length from snout to tip of tail. 226 mm. Ramsay’s figures of the hind and fore feet of the holotype are reproduced in figure 2, K and L. Skull and mandible: Ramsay did not describe the skull and mandible of the holotype. How- ever, he noted that the former is broken and that only the anterior parts of it with a portion of the dentition is left. M-* is missing from the left molar row of S427 and the, posterior portion of the palate is badly damaged: Antero-internal angles of nasals broken but fronts of nasals about level with premaxillae; posterior ends of nasals level with backs of nasal processes of premaxillae. Nasals with maximum width subterminal then tapering to their junction with the frontals. Rostrum short, broad, and deep. Lacrymals small. Postorbital ridges present. Dorso-anterior angles of zygomatic plates rounded, about level with anterior edges of bottom halves of plates (anterior edge of left zygomatic plate damaged). Incisive foramina elliptical with rounded ends and with greater part contained in maxillae; left incisive foramen extending to anterior extremity of M^. Left posterior palatal foramen small, opposite anterior end of M2. Incisors opisthodont. Molars moderately worn, without accessory cusps. Angular processes and coronoid regions of mandibular rami broken. Molars without accessory cusps. Measurements for the holotype are tabulated in table 1. The skull, mandible, and dentition of this individual are illustrated in figure 2, A to J. TABLE 1 Measurements (in millimetres) for holotype skull and mandible (Australian Museum No. S427) of Melomys burtoni (Ramsay). Length from anterior extremity of nasal to posterior extremity of inter- frontal suture Nasal length measured from anterior extremity of nasal to posterior extremity of internasal suture Maximum width across nasals Width across nasals between naso- frontopremaxillary points Width of rostrum at anterior end of incisive foramina ... .... Maximum width of rostrum Interorbital width Height of skull at anterior extremity of Ml Minimum width across zygomatic plate Length of incisive foramen Width across incisive foramina . .. Length of diastema Width of palate between alveoli of antero-internal roots of Mi Ml Length x width M ‘2 Length x width M-i Length x width Mi-3 Length Length of mandibular ramus from tip of incisor to posterior extremity of condyle measured with ventral surface of ramus horizontal Height of condyle above ventral surface of mandibular ramus measured with ventral surface of ramus horizontal .... M^ Length x width M^ Length x width M.j Length x width M ,_3 Length 19.0 approx. 8.8 approx. 3.3 1.9 4.6 5.1 approx. 5.0 7.2 3.3 5.2* 2.0 approx. 7.7 2.9 approx. 3.1 X 1.6 2.2 X 1.6 1.1 X 1.1 5.8 19.4 7.5 approx.* 3.6 X 1.5 2.1 X 1.7 1.4 X 1.3 6.0 Notes: The length and height marked by asterisks were measured on the left side of the skull and mandible; measurements of the other paired structures were taken on the right side of the skull and mandible. Measurements taken by vernier calipers graduated to read to 0.05 mm; tooth measurements do not include roots. Figure 2 (opposite) — Melomys burtoni (Ramsay). A — S427 (Holotype), dorsal view of cranium; B — ventral view of cranium; C — right lateral view of cranium; D — left lateral view of cranium; E — occlusal view of right upper molar row; F — occlusal view of left lower molar row; G — dorsal view of left mandibular ramus; H — lingual view of left mandibular ramus; I — labial view of left mandibular ramus; J — lingual view of right mandibular ramus; K — Holotype, hind foot; L — Holotype, fore foot; M — Australian Museum label accompanying holotype of Melomys burtoni (Ramsay). A — D and G — J x2.6 approx.; E & F x8.0 approx.; H, K & L x2 (reproduced from Ramsay. 1887b PI. 17 figs. 4-5); M xl.3. Discussion Twelve species of Melomys, including M. bur- toni (Ramsay), have been recorded from Australia. Except for the Queensland form, Melomys callopes Pinlayson, 1942, these were erected prior to 1938 when Rtimmler published on the New Guinea Muridae. In that work, Rtimmler (1938 p. 100) reduced the Australian species to two, Melomys cervinipes (Gould, 1852) from Queensland, Northern Territory, and 93 New South Wales', and Melomys lutillus (Thomas, 1913) from Queensland (and New Guinea). Both Ellerman (1941 pp. 230-231 & 1949 pp. 87-88) and Tate (1951 p. 292) concur with Rummler in listing only M. cervinipes and M. lutillus as the Australian species: but, while Ellerman agrees closely with Rummler (1938 pp. 116 & 130) in his allocation of the various Australian populations of Melomys under these two names, Tate’s arrangement is rather differ- ent, and it is clear that further work will be required before boundaries can be placed with any certainty about these taxa. At the time that these authors revised Melomys, M. burtoni was not recognized as a species of that genus and since it is an older name than all available names of Melomys (excluding Pogonomelomys — see under) except M. cervinipes (Gould, 1852) and M. rufescens (Alston, 1877). the identification of its bio- logical status and affinities is nomenclaturally important. M. burtoni is distinct from M. rufescens, which is only known to occur in the Bismarck Archipelago, New Guinea, and the Solomon Islands (Laurie and Hill, 1954). Un- fortunately, so little is known of the morpho- logical ranges of both M. cervinipes and M. lutillus that the relationships of M. burtoni to either of these Australian species cannot be determined at present. It can only doubtfully be placed within M. cervinipes (Gould); there- fore I tentatively accept it here as a valid species of Melomys but recognize that further material and more detailed studies will prob- ably result in its identification with one or more of the extra-Western Australian populations of Melomys. Until now, Melomys was not known to occur in Western Australia. While no species of Pogonomelomys are known to occur in Australia this genus is closely related to Melomys and should be taken into consideration in discussing the status of M. burtoni. Pogonomelomys (type species Melomys mayeri Rothschild and Dollman, 1932) was first introduced, as a subgenus of Meloinys, by Rummler (1936 p. 248) but was later elevated to full generic rank by Tate and Archbold (1941 p. 5). In doing this, they drew particular attention to the prehensile dorsal tip of the tail, provided with tactile skin, in Pogono- melomys; but Harrison (1962 p. 59) has recently noted that the tail tips are partly pre- hensile in Australian Melomys identified by him (1962 p. 57) as M. cervinipes eboreus Thomas and M. lutillus littoralis (Ldnnberg) and the ^Melomys cervinipes (and Melomys) has not been known to extend further south than the Hunter River. New South Wales where Gould reported it in 1852. A number of maxillae and mandibular rami, indis- tinguishable from Melomys cervinipes (Gould), are included in the Quaternary red bone deposit of the Pyramid Cave, Buchan district, eastern Victoria (see Wakefield 1960a & 1960b for an account of the bone deposits in the Buchan District) and, if this material is correctly placed here, this rodent has apparently undergone a recent shrinkage in range in south-eastern Australia. These specimens were identified by me subsequent to publication of Wakefield’s second paper (1960b) and are not in- cluded among the rodent remains recorded therein for the Pyramid Cave deposit. They have now been placed in the palaeontological collections of the National Museum of Victoria and are registered specimens no. P 20673. distinction may not be as useful as Tate and Archbold supposed. Unfortunately, the nature of the skin on the dorsal tip of the tail of M. burtoni is not mentioned by Ramsay so there is no justification in placing it within Pogono- melomys; nevertheless, such characters as it is known to possess do not exclude this possibility, but it should here be noted that of all the species currently placed in Pogonomelomys (Ellerman. 1949; Tate. 1951; Laurie and Hill, 1954) only the name Uromys bruijnii Peters and Doria, 1876 predates M. burtoni, and U. bruijnii and M. burtoni are certainly not con- specific. Acknowledginents I wish to thank Mr. B. J. Marlow, Curator of Mammals, Australian Museum, Sydney for permission to examine the Museum’s rodent collection. A preliminary examination of the label accompanying S427 was carried out by Constable B. J. Fitzgerald, Handwriting Section, Scientific Investigation Bureau, Criminal Investigation Branch, N.S.W. Police Force. Mr. Marlow and Mrs. J. Anderson, Curator of the Macleay Museum, University of Sydney have provided valuable assistance in the elucidation of data on this label. Mr. G. P. Whitley, recently retired Curator of Fishes, Australian Museum has identified the handwriting of the Australian Museum catalogue entry for S427 as that of Mr. E. R. Waite. Mr. N. A. Wakefield, Melbourne Teacher’s College, Victoria has kindly permitted me to examine his Buchan district rodent collection. Mr. R. C. Taylor, Department of Illustration. University of Sydney has provided the photographs for figure 2. The manuscript has been read and criticised by Dr. W. D. L. Ride, Director, Western Australian Museum, Perth. References Alston. E. R. (1877). — On the rodents and marsupials collected by the Rev. G. Brown in Duke-of- York Island, New Britain, and New Ireland. Proc. Zool. Soc. Lond. 1877: 123-127. Ellerman, J. R. (1941). — The Families and Genera of Living Rodents; with a List of Named Forms (1758-1936) by R. W. Hayman and G. W. C. Holt. Vol. 2, Family Muridae; 1-690 (British Museum (Natural History); London). (1949). — The Families and Genera of Living Rodents; with a List of Named Forms (1758-1936) by R. W. Hayman and G. W. C. Holt. Vol. 3. Pt. 1: 1-210 (British Museum (Natural History); London). Gould, J. (1852). — The Mammals of Australia. Pt. 4 (Vol. 3) PI. 14 and Text. Harrison. J. L. (1962).— Mammals of Innisfail. 1. Species and distribution. Aust. J. Zool. (10) 1: 45-83. Iredale, T. and Troughton, E. Le G. (1934). — A check-list of the mammals recorded from Australia. Mem. Aust. Mus. 6: 1-122. Laurie. E. M. O. and Hill. J. E. (1954).— List of Land Mammals of New Guinea, Celebes and Adjacent Islands 1758-1952. 1-175 (British Museum (Natural History); London). Longman, H. A. (1916). — List of Australasian and Austro- Pacific Muridae. Mem. Qd. Mus. 5: 23-45. Ogilby. J. D. (1892). — Catalogue of Australian mammals with introductory notes on general mam- malogy. Aust. Mus. Cat. 16: 1-142. 94 Ramsay, E. P. (1887a). — Description of a new species of HapalotiSy (H. Boweri) from North West Australia. Ptoc. Linn. Soc. N.S.W. (2) 1: 1153-1154. (1887b). — Description of two new species of marsupials (Perameles and Antechinus), and of a new species of Mus (M. burtonijy from the neighbourhood of Derby, N.W.A. Proc. Linn. Soc. N.S.W. (2) 2: 551-553. Rlimmler, H. (1936). — Die Pormen der papuanischen Muridengattung Melomys. Z. Sdugetierk. 11: 247-253. (1938). — Die Systematlk und Verbreitung der Muriden Neuguineas. Mitt. zool. Mus. Berl. 23: 1-297. Tate. G. H. H. (1951). — The rodents of Australia and New Guinea. Results of the Archbold Expeditions. No. 65. Bull. Amer. Mus. Nat. Hist. 97: 183-430 and Archbold, R. (1941). — New rodents and marsupials from New Guinea. Results of the Archbold Expeditions. No. 31. Am. Mus. Novit. No. 1101: 1-9. Thomas, O. (1913). — Some new species of Uromys. Ann. Mag. Nat. Hist. (8) 12: 212-217. (1922). — A subdivision of the genus Uromys. Ann. Mag. Nat. Hist. (9) 9: 260-261. Wakefield, N. (1960a). — Recent mammal bones in the Buchan district — 1. Viet. Nat. 77 (6): 164-178. (1960b). — Recent mammal bones in the Buchan district — 2. Viet. Nat. 77 (8): 227-240. Whittell, H. M. (1954).— The Literature of Australian Birds: A History and a Bibliography of Aus- tralian Ornithology. Pt. 1: 1-116. Pt. 2: 1-786. Pt. 1 index: 787-788. (Paterson Brokensha Pty. Ltd., Perth, Western Australia). 95 INSTRUCTIONS TO AUTHORS Contributions to this Journal should be sent to The Honorary Secretary, Royal Society of Western Australia, Western Australian Museum, Perth. Papers are received only from, or by communication through, Members of the Society. The Council decides whether any contribution will be accepted for publication. All papers accepted must be read either in full or in abstract or be tabled at an ordinary meeting before publication. Papers should be accompanied by a table of contents, on a separate sheet, showing clearly the status of all headings; this will not necessarily be published. Authors should maintain a proper balance between length and substance, and papers longer than 10,000 words would need to be of exceptional importance to be considered for publication. The Abstract (which will probably be read more than any other part of the paper) should not be an expanded title, but should include the main substance of the paper in a con- densed form. Typescripts should be double-spaced on opaque white foolscap paper; the original and one carbon copy should be sent. All tables, and captions for figures, should be on separate sheets. To avoid unnecessary handling of the original illustrations authors arte requested to include additional prints, preferably reduced to the final size required; a choice of either one-column (about 2.8 inches) or two-column (about 5.8 inches) width is available. The preferred positions of figures should be marked on the second typescript copy. In the preparation of references, and for all matters of presentation not otherwise covered in these instructions, authors are required to follow the C.S.I.R.O. Guide to Authors (Melbourne, 1963). Failure to read through this carefully before preparing papers may lead to delay in publication. The use of the various conventional systems of nomenclature recommended in this booklet, and in the supplementary pamphlets referred to in it, is obligatory; for this purpose, palaeontological papers must follow the appropriate recommendations for zoology or botany. All new stratigraphic names must have been previously approved by the Stratigraphic Nomenclature Committee of the Geological Society of Australia, Thirty reprints are supplied free of charge. Further reprints may be ordered at cost, provided that orders are submitted when the galley proofs are returned. Authors are solely responsible for the accuracy of all information in their papers, and for any opinion they express. Journal of the Royal Society of Western Australia Volume 48 1965 Part 3 Contents 8. A new genus of Temnospondyli from the Triassic of Western Australia. By J. W. Cosgriff. 9. The identity of Mus burtoni Ramsay, 1887 (Rodentia, Muridae, Melomys) from the neighbourhood of Derby, Western Australia. By J. A. Mahoney. Editor: A. F. Trendall Assistant Editor: A. S. George The Royal Society of Western Australia, Western Australian Museum, Perth 99036/6/65— 570— t ALEX. B. DAVIES, Government Printer, Western Australia JOURNAL OF THE ROYAL SOCIETY OF WESTERN AUSTRALIA VOLUME 48 PART 4 PUBLISHED 31ST DECEMBER, 1965 REGISTERED AT THE G.P.O., PERTH FOR TRANSMISSION BY POST AS A PERIODICAL THE President Vice-Presidents Past President Joint Hon, Secretaries Hon. Treasurer Hon. Librarian Hon. Editor ROYAL SOCIETY OF WESTERN AUSTRALIA COUNCIL 1965-1966 J. H. Lord, B.Sc. D. Merrilees, B.Sc. R. W. George, B.Sc., Ph.D. W. R. Wallace, Dip.For. Margaret E. Redman, B.Sc., A.L.A.A. A. B. Hatch, M.Sc., Dip.For. R. D. Royce, B.Sc. (Agric.). Ariadna Neumann, B.A. A. F. Trendall, B.Sc., Ph.D., A.R.C.S., F.G.S. A. S. George, B.A. C. F, H. Jenkins, M.A. L. E. Koch, M.Sc. R. J. Little. P. E. Playford, B.Sc., Ph.D. R. T. Prider, B.Sc., Ph.D., M.Aust.I.M.M., F.G.S. L. W. Samuel, B.Sc., Ph.D., P.F.A.C.I., F.R.LC. D. L. Serventy, B.Sc. (Hons.), Ph.D. Journal of the Royal Society of Western Australia Vol. 48 Part 4 10. — The Porcellanidae (Crustacea, Anomura) of Western Australia with descriptions of four new Australian species by Janet Haig* Manuscript received and accepted 16th February, 1965 Abstract Twenty-eight species of Porcellanidae are reported for Western Australia. Three are new species: Pachycheles johnsoni, Porcellana furcillata, and Polyonyx maccullochi. Fifteen previously known species are new records for Western Australia, and seven of those fifteen are new records for Australia. A new species. Pachycheles granti, from Queensland and New South Wales, is also described. Introduction Until now vex’y little attention has been given to Western Australian crabs of family Porcellanidae. Only two papers have been published, dealing specifically with the Western Australian fauna, in which Porcellanidae were treated: these were Rathbun (1924) and Hale (1929). Porcellanids were mentioned inci- dentally in three additional works. Only ten species of the family have been recorded from Western Australia. The incentive for the present study was pro- vided by a large collection of crustaceans dredged off the coast of Western Australia by an expedition sponsored by the Bernice P. Bishop Museum of Honolulu, and donated to the Western Australian Museum (George, 1961). The Porcellanidae from that expedition, together wdth other members of the family in the collections of the Western Australian Museum, were turned over to me for study. I have also been able to borrow numerous Western Austra- lian porcellanids from the Australian Museum; the Porcellanidae collected by the Hamburg Southwest-Australia Expedition , 1905. and housed in the Zoologisches Museum, Hamburg; and a small group of specimens in the U.S. National Museum, comprising a portion of the Porcellanidae collected during Dr. E. Mjdberg’s Swedish scientific expeditions to Australia 1910- 1913 and reported by Rathbun (1924). During a visit to the British Museum (Natural History) I examined a small collection of Western Aus- tralian Porcellanidae housed in that institution. In the synonymy of each species I have tried to include all references to its occurrence any- where in Australia. * Allan Hancock Foundation. University of Southern California. Los Angeles, California, U.S.A. Contribution no. 274 from the Allan Hancock Foundation. Discussion of zoogeographical implications will be included in a future report, which will treat the other anomuran families (hermit crabs excepted) represented in Western Aus- tralia. The following abbreviations are used in the text: W.A.M. — Western Australian Museum; A.M. — Australian Museum; B.M.N.H. — British Museum (Natural History) ; Z.M.H. — Zoologis- ches Museum, Hamburg; U.S.N.M. — United States National Museum. Key to Western Australian genera of Porcellanidae 1. Movable segments of antennal peduncle with free access to orbit 2 Movable segments of antennal peduncle excluded from orbit by a forward projection of basal seg- ment, which meets anterior margin of carapace 4 2. (1) Posterior portion of lateral walls of carapace consists of one or more pieces separated by mem- branous interspaces from anterior portion Pachycheles Lateral walls of carapace entire 3 3. (2) Front a strongly produced, tri- angular rostrum, armed with a row of strong spines; telson of abdomen five-plated Petrocheles Front various. usually rather prominent but never produced into a strongly armed rostrum; telson seven-plated Petrolisthes 4. (1) Carapace broader than long .... 5 Carapace as long as, or longer than, broad 6 5. (4) Dactyl of walking legs with at least two well-developed fixed claws, and often with accessory spinules on lower margin .... Polyonyx Dactyl of walking legs a straight, slender spine with no accessory spinules .... .... .... Raphidopus 6. (4) Carapace markedly longer than broad; dactyl of walking legs very short and stout, with four strong, fixed claws Porcellanella Carapace at most only slightly longer than broad; dactyl of walking legs slender, with one or two fixed claws and a few movable spinules 7 97 7. (6) One cheliped distinctly larger than the other, particularly in males; dactyl of small cheliped twisted out of plane with manus; dactyl of walking legs with a single terminal claw Pisidia Chelipeds subequal or one distinctly larger than the other; dactyl not twisted out of plane with manus (except in occasional large specimens of gravelei); dactyl of walking legs bifurcate or with a single terminal claw .... Porcellana Genus PETROCHELES Miers Diagnosis. Basal antennal segment short, not produced forward to meet anterior margin of carapace; movable segments with free access to orbit. Carapace conversely cordate, broader pos- teriorly than anteiiorly and with a strongly pro- duced, triangular rostrum. Lateral margins of carapace and rostrum armed with a row of strong spines. Chelipeds subequal; chelae long, slender, flattened; carpi armed with strong- spines on margins and on dorsal surface. Telson of abdomen five-plated. Petrocheles australiensis ( Miers) Petrolisthes (Petrocheles) australiensis Miers 1876a, p. 222 (Australia); 1876b, p. 61. Haswell 1882b. p. 147 (Mordialloc). Baker 1905, p. 264, pi. 36, figs 1, la (St. Vincent Gulf; Port MacDonnell). Petrocheles australiensis, Hale 1927a, p. 81, text-fig. 78 (South Australia). Diagnosis. Protogastric region of carapace with a pair of spines. Rostral spines four on either side. Inner margin of carpus of cheliped armed with six or seven strong spines. Material examined. 3 9 , Great Australian Bight, S. of Nullarbor Station, 5.i.l958, R. Hardie, W.A.M. 171-60. “Swims backwards like cray- fish.” Remarks. Largest specimen 18.4 mm in cara- pace length. Distribution. Australia (Victoria and South Australia). Now recorded from Western Aus- tralia. Genus PETROLISTHES Stimpson Diagnosis. Basal antennal segment short, not produced forward to meet anterior margin of carapace; movable segments with free access to orbit. Carapace rounded or subquadrate, usually about as broad as long; armature various. Front triangular or trilobate, often rather prominent, not strongly armed except near base (minutely denticulate in some species). Chelipeds sub- equal; chelae flattened; armature of carpi vari- ous. Telson of abdomen almost invariably seven- plated. Key to Western Australian species of Petrolisthes 1. Supraocular spine present; spines on lateral margins of carapace posterior to epibranchial spine 2 No supraocular spine, nor spines on lateral margins of carapace posterior to epibranchial spine 3 2. (1) Front broad, sinuously triangular; inner orbital angle not distinct, but marked by a large spinule Front narrow, with distinctly marked but unarmed inner orbital angle 3. (1) Epibranchial spine present No epibranchial spine 4. (3) Carapace strongly to faintly rugose, not much longer than broad Carapace smooth, distinctly longer than broad 5. (4) Merus of walking legs unarmed on anterior margin Merus of walking legs armed on anterior margin .... 6. (5) Dorsal surface of chela with rugae and flattened tubercles; inner side of fingers pubescent Dorsal surface of chela with small, upstanding, well-separated granules; no pubescence, or at most only a fine trace, in gape of fingers scabriculus militaris 4 teres 5 ohshimai 6 moluccensis boscii haswelli Petrolisthes scabriculus (Dana) Porcellana scabricula Dana 1852, p. 424 (Sulu Sea); 1855, pi. 26, fig. 13. Petrolisthes scabriculus, Stimpson 1858, p. 227. Haig 1964, p. 358, text-fig. 2. Diagnosis. Carapace transversely rugose. Supraocular spine present. Front broad, triangu- lar, forming a nearly even curve from tip to supraocular spine; inner orbital angle marked by a distinct spinule. Epibranchial region with two spines; midbranchial margin with a few spinules. Inner margin of carpus of chelipeds with four or five broad teeth: outer margin with row of strong spines. Chela with a median longitudinal crest: dorsal surface to outside of this crest covered with hairs, which form a heavy fringe along outer margin. Merus of walking legs armed on anterior margin with a row of spines. Material examined. 19, W. of Flat I. off Onslow (near Long I.), 6-10 fm, 9.vi.l960, B. R. Wilson on “Davena”, W.A.M. 36-62. 1 . Males to 8.7 mm; ncn-ovigerous females to 9.0 mm; ovigerous females to 8.1 mm. Ovigerous females in August at Broome and in December at Shark Bay. Distribution: Recorded only from Gulf of Siam (Haig) and Australia (Western Australia and Queensland). PetroHsthes ohshiinai (Miyake) Porcellana maculata H. Milne Edwards 1837, p. 253 (New Ireland). Petrolisthes maculatus, Miers 1884, p. 558. Lenz 1905, p. 375, pi. 47, fig. 1. Neopetrolistlies ohshimai Miyake 1937, p. 35. text-fig. (Ryukyu Islands); 1942, p. 350, text-fig. 13. McNeill 1953, p. 90 (Hope Island). Petrolisthes ohshimai, Johnson 1960, p. 164. Gordon 1960, p. 166. “Small species of crab.” Saville-Kent 1897, p. 220, pi. 39 (Roebuck Bay or Lacepede Islands, fide Gordon 1960). “Unidentified crab.” McCulloch and McNeill 1923, p. 58 (Great Barrier Reef). Diagnosis. Carapace smooth, hairless, convex, markedly elongate. No supraocular spine. Front very broad, partly covering and reaching well beyond eyes. Epibranchial spine present. Inner margin of carpus of chelipeds with two or three wide-set teeth, proximal one largest; outer margin unarmed except for a short, blunt tooth at distal end. Chela broad, flat, with strongly curved outer margin. Merus of walk- ing legs unarmed on anterior margin. Red spots on carapace and chelipeds, and usually also on abdomen, maxillipeds, eyestalks, and walking legs. The red colour sometimes takes the form of large, irregularly-shaped blotches, 101 and sometimes of very regularly and evenly distributed small spots: compare e.g. the illus- trations in Lenz 1905 and in Miyake 1942. Material examined. 19, Roebuck Bay, 1932, Mrs. B. Grey, B.M.N.H. 1 , 19, N.E. side Rosemary I., Dampier Archipel., low tide on hypostome of large anemone, 27. viii. 1961. B. R. Wilson and G. W. Kendrick, W.A.M. 32-62. Remarks. Johnson (1960) transferred Miyake’s species to genus Petrolisthes and reported a specimen from Christmas Island, eastern Indian Ocean, which he considered a westward exten- sion of range. However, Dr. J. Forest (in litt.) informs me that Porcellana maculata H. Milne Edwards is identical with Miyake’s species. There are two dried type specimens in the Paris Museum, labelled "Porcellana maculata, Edw./ MM. Quoy et Gaimard./Nlle Irlande”; one of them, a male 8.5 by 7 mm. was chosen as lectotype by Dr. Forest. On the basis of this record, and subsequent ones by Miers and Lenz, the range of the species may be extended to the western Indian Ocean. It is a commensal with giant sea anemones of genus Stoichactis. Even though H. Milne Edwards’ name antedates Miyake’s, the latter must be used for the species because of Article 23b of the International Code of Zoological Nomenclature, which requires that a name unused as a senior synonym for more than 50 years is to be dis- carded. The name maculatus as applied to this species did not appear in print between its use by Lenz (1905) and by Jacquotte (1964). The red colour in the Rosemary Island speci- mens is in small, evenly distributed spots as in Lenz’s illustration, but there are large areas of unspotted ivory white: most of the frontal region, the fingers, distal and outer part of the palm, proximal and distal ends of carpus of the chelipeds, most of the merus of the walking legs, and the entire carpus, propodus, and dactyl of those appendages. Male 11.2 mm; female 10.0 mm. Distribution. Indian Ocean; Zanzibar, Mozam- bique. Madagascar, Gulf of Mannar, Christmas Island. Pacific Ocean: Ryukyu Islands, Palau Islands, Bismarck Archipelago, Marshall Islands, Fiji Islands. Australia (Western Australia and Queensland) . Genus PACHYCHELES Stimpson Diagnosis. Basal antennal segment short, not produced forward to meet anterior margin of carapace: movable segments with free access to orbit. Carapace rounded or subquadrate, usually about as broad as long in males, a little broader than long in females; unarmed except for spine, present in some species, at outer orbital angle. Front not prominent, usually rounded or transverse in dorsal view, trilobate in frontal view. Epimera (lateral walls of carapace) incomplete, the posterior (sub- branchial) portion consisting of one or more pieces separated by membranous interspaces from anterior portion. Chelipeds large, thick, and robust, one almost always distinctly larger than the other. Telson of abdomen composed of either seven or five plates (five in all Indo- West Pacific species). Male lacking pleopods in many species. Key to Australian species of Pachycheles 1. Carapace covered with scattered setae; pleopods present in males pisoides Carapace devoid of hairs; no pleopods in males 2 2. (1) Carpus and chela with longitu- dinal rows of strong, well- separa ted tubercles. these tubercles covered with short hairs granti Chelipeds devoid of hairs .... 3 3. (2) Anterior regions of carapace smooth; chelae smooth, tuber- culate, or with four longitudinal ridges sculptus Anterior regions of carapace roughened; carpus and chela with large, scalloped-edged tubercles, latter often with free edges . .. johnsoni Pachycheles sculptus (H. Milne Edwards) Porcellana sculpta H. Milne Edwards 1837, p. 253 ( Java) . Porcellana pisum H. Milne Edwards 1837. p. 254 (China). Porcellana pulchella Haswell 1882a, p. 758 (Holborn Island; Port Molle); 1882b, p. 148. Pachycheles pulchellus, Miers 1884, p. 273, pi. 30. fig. A (Port Molle; Albany Island; Thursday Island; Prince of Wales Channel). Henderson 1888, p. 114 (Arafura Sea south of Papua; Flinders Passage). Ortmann 1894. pp. 29, 30 (Thursday Island). Pachycheles sculptus, Ortmann 1894, p. 29. Haig 1964, p. 368. Pachycheles pisum, Rathbun 1924, p. 30 (Cape Jaubert). Ward 1928, p. 245 (Capricorn and Bunker Groups). ^Pachycheles sculptus. Ward 1928, p. 245 (Capricorn and Bunker Groups). Diagnosis. Front broad, transverse in dorsal view and trilobate in frontal view: anterior regions of carapace smooth: carapace and cheli- peds devoid of hairs. Chelipeds, especially carpus and chela, extremely variable: Chela entirely smooth, or with incipient smooth, longitudinal ridges, or with surface somewhat pitted, or with four smooth, narrow, longitudinal ridges, lat- ter often crossed by slightly oblique grooves to form rows of close-set tubercles: the wide inter- spaces between these rows smooth, pitted, or with small, wide-set tubercles. Chelae similar in ornamentation, or minor more strongly ridged and tuberculate than major. Carpus completely smooth, or with low, somewhat overlapping- tubercles, or with irregular rows of strong, squamate tubercles, some of them transversely elcngate. No pleopods in males. Material examined. 1 ^ , 23 miles S.W. of Trough- ton I., 25 fm. W.A.M. 336-62. 1 ^, 2$, Pender Bay, 1932. Mrs. B. Grey, B.M.N.H. 10 S , 11 9, Broome. June 1932, A.M. P.10222. 1 ^ , off Roe- buck Bay, 5-9 fm. June-Oct. 1929, A. A. Living- stone, A.M. P.14124. 2S, entrance to Roebuck Bay, 9 fm, 15.viii.l929, A. A. Livingstone, A.M. P.14100. 1^, 19, near entrance to Roebuck Bay, 5-8 fm. 26. ix. 1929, A. A. Livingstone, A.M. P.14081. 1^, Roebuck Bay, 1932, Mrs. B. Grey. B.M.N.H. 2$, 10-20 miles W. of Lagrange Bay, 12-25 fm, W.A.M. 347-62. 2 9 , 40 miles W. of Cape Jaubert, 23 fm. W.A.M. 326-62 & 329-62. 1 ^,- 19, 42 miles W.S.W. of Cape Jaubert, 70 ft, 26.V.1911, E. Mjoberg. U.S.N.M. 56400. 1 ^ . 42 miles W.S.W. of Cape Jaubert, 70 ft, 30. v. 1911. E. Mjoberg, U.S.N.M. 56401. 1 9, 42 miles W.S.W. of Cape Jaubert, 48 ft, 16.vii.l911, E. Mjoberg, U.S.N.M. 56399. 1 ^ , off Ninety Mile Beach between Cape Jaubert and Wallal, 5 fm. Sept. 1929, A. A. Liv- ingstone, A.M. P.14106. 3^, 3 9, 60 miles W. x N. of Bedout I., 25 fm, W.A.M. 342-62 & 343-62, 19, N.E. side Rosemary I., Dampier Archipel., under stones at low tide, W.A.M. 358-62. 11 <5, 99, 3-4 miles off E. end Delambre I., Dampier Archipel., 10 fm, W.A.M. 125-60 & 126-60. 1^, 19,2 miles W. of Legendre I., Dampier Archipel., 23 fm, W.A.M. 128-60. 1^, Malus I., Dampier Archipel., 10 fm, W.A.M. 180-60. 1 9 , W. ap- proaches to Mermaid Strait, Dampier Archipel., 20 fm, W.A.M. 117-60. IS, 19, Exmouth Gulf or Shark Bay, trawled, W.A.M. 76-62. 1 juv., off Shark Bay, 121 m, “Umitaka Maru”, W.A.M. 356-60. 19, Point Gregory, N.W. side Peron Peninsula, Shark Bay, under stones at low tide, W.A.M. 196-60. 3S, 29, N.E. side Peron Flats, Shark Bay, from sponges and trawl trash, W.A.M. 354-62 Port Hedland, 1905, Gale for Hamburg S.W. Australia Exped., Z.M.H 11512. Remarks. The specimen from the Palau Islands referred to Porcellana nitida by Miyake (1942, p. 359, text-figs. 20-22; 1943, p. 129, text- fig. 50) cannot belong to this species because it has a strong tooth on the lateral margin of the carapace behind the epibranchial angle; in none of the specimens examined, nor in the types, is there an indication of even incipient spinulation or crenulation on the carapace margins. With Miyake’s record removed from the synonymy the species becomes restricted to Australia, where it is perhaps an endemic. Miers’ variety rotundifrons is based on adults of P. nitida, which was described by Haswell from juvenile material. The differences in the form of the front mentioned by Miers can be attri- buted to growth. Males to 7.4 mm; ovigerous females 7.2 and 7.3 mm, off Cape Jaubert in October. Distribution. Australia (Northern Territory and Queensland). Now recorded from Western Australia. Porcellana furcillata, sp. nov. (Fig. 2) Description. Carapace a little longer than broad; nearly smooth, with protogastric regions lightly indicated and other regions scarcely marked. Front broad, horizontal, not deflexed, produced beyond eyes; median lobe broad, its surface concave, apex tipped with two or three strong spinules, a larger spine on either side near tip; separated by broad, U-shaped notches from narrow lateral lobes which have their tips acute and curved inward, inner margins armed with several spinules. Outer orbital angle produced into a strong spine; a strong spine or tooth on hepatic margin; a small but distinct epibranchial spinule. Lateral margins posterior to cervical groove with three or four spinules. First movable antennal segment with a small anterodistal spinule: second and third unarmed; flagellum long, slender, without hairs. Chelipeds smooth, devoid of hairs. Merus with a strong inner lobe, its margin crenulate and bearing two teeth; outer margin with two strong spines; lower surface with a strong spine at inner distal angle. Carpus with five to seven shallow, wide-set teeth on inner margin, these teeth varying in size and some of them occasionally much reduced or obsolescent; outer margin with three strong spines including the one at distal end. Chelae elongate, one some- what larger than other, lying at a very oblique angle to plane of carapace; outer margin with a sharp crest bearing a row of minute granules, these becoming larger on distal portion of manus and on pollex; just to inside of this margin, proximal half of chela with a row of five to eight strong spines. Fingers crossed at tips, which are not notched; outer margin of dactyl smooth, or with a row of fine spinules similar Figure 2. —Porcellana furcillata. A, dorsal view of holotype; B. basal segment of antennuie; C, propodus and dactyl of right first walking leg; D, right first walking leg. Scales for A and D 5 mm; for B, 0.5 mm; for C, 2 mm. no to those of outer margin of pollex; inner side of fingers at their bases with a short tuft of fine hair. In larger chela, fingers generally somewhat gaping, with a tooth on cutting edge of pollex, and shorter in comparison to length of manus than in smaller chela, which has fingers meeting along entire length of straight cutting edges. Walking legs long and slender, with long, scattered setae. Anterior margin of merus armed with several spines. Carpus with an anterodistal spine, more strongly devoloped on first pair of legs. Propodus with a slender, movable spinule about midway along posterior margin and three similar spinules at posterodis- tal end. Dactyl deeply cleft into two strong, subequal fixed claws; lower margin with a single small, movable spinule. Telson of abdomen seven-plated. Holotype. Ovigerous female (5.1 by 4.6 mm). 60 miles W. x N. of Bedout Island, 25 fm, 12.x. 1962, R. W. George on “Dorothea”, W.A.M. 344-62. Paratypes. QUEENSLAND: IS, 19, Albany Passage, Cape York, Aug. -Oct. 1907, C. Hedley and A. R. McCulloch, A.M. P.14764. WESTERN AUSTRALIA: 1 S , Roebuck Bay, 9 fm, Aug. 1929, A. A. Livingstone, A.M. P.14765. 1 9 , off Roe- buck Bay, 5-9 fm. June-Oct. 1929, A. A. Living- stone, A.M. P.14091. 1 ^ , 72 miles W. x N. of Bedout I., 25 fm, 12.x. 1962, R. W. George on “Dorothea”, W.A.M. 357-62. 1 ^,2 miles W. of Legendre I, Dampier Archipelago, 23 fm. 9.vi. 1960, B. R. Wilson on “Davena”, W.A.M. 128-60. Remarks. Porcellana furcillata is perhaps allied to P. quadrilobata Miers 1884, a species which has been collected in Australian waters but is not yet recorded from Western Australia. In P. quadrilobata the median frontal lobe has a distinct notch, and the carpus of the cheli- peds is unarmed along the outer margin. P. furcillata bears a superficial resemblance to Pisidia spinuligera (Dana), from which it is easily distinguished by the deeply bifurcate dactyl of the walking legs. Males 3.5 by 3.0 mm to 4.0 by 3.6 mm; non- ovigerous female 3.6 by 3.2 mm; ovigerous females 4.9 by 4.5 mm and 5.1 by 4.6 mm. Ovigerous females between August and October at Cape York and in October off Bedout I. Distribution. Known only from the localities listed above, in Western Australia and Queens- land. Genus PORCELLANELLA White Diagnosis. Basal antennal segment strongly produced forward and broadly in contact with anterior margin of carapace: movable segments far removed from orbit. Carapace considerably longer than broad, smooth, sides subparallel; without spines except for projection of outer orbital angle. Front horizontal, not at all de- flexed, produced well beyond eyes, strongly tri- dentate. Chelipeds subequal, smooth, unarmed. Chelae elongate, lying nearly vertical to plane of carapace, with a crest of hair on inner sur- face. Walking legs short, smooth, their anterior margins unarmed. Dactyls very short; multi- unguiculate, with four strong, fixed spines. Telson of abdomen seven-plated. Porcellanella triloba White Porcellanella triloba White 1852. p. 394, pi. 5, figs. 2, 2a {off Cape Capricorn). Johnson 1964, p. 100 (Bowen). Porcellana triloba, Haswell i882h, p. 149. Diagnosis. Lateral frontal lobes about three- fourths length of median lobe. Outer orbital angle acute or spiniform. Inner margin of merus of chelipeds with a strongly developed lobe. Most proximal spine of dactyl of walking legs much smaller than other three. Material examined. 1 ^ , 19, 36 miles S.W. of Adele I., 40 fm, 17.X.1962, R. W. George on “Dorothea”, W.A.M. 333-62. 1^, between Broome and Wallal on Ninety Mile Beach, c. 7 fm, 1931, Capt. R. Bourne, A.M. P.10020. 15, Roebourne, 21. ix. 1959, C. Lambert, W.A.M. 161- 60. 25, 29, mainland at Antenni Mia, Point Samson, or Rosemary I., Dampier Archipel., on a pennatulid, 26-30. viii. 1961, B. R. Wilson, W.A.M. 47-62. 15, 1 9, S.E. of Rosemary I., Dampier Archipel., 3-5 fm, 26.viii.1961, G. W. Kendrick and B. R. Wilson, W.A.M. 24-62. 15, N.W. of Heirisson Prong, Shark Bay, 11-124 m, 13.ix.l905, St. 16, Hamburg S.W. Australia Exped., Z.M.H. 11676. 1 9, Freycinet Reach, W. of Middle Flat to N. point of Heirisson Prong, Shark Bay, 11-16 m, 12.ix.l905, St. 14, Hamburg S.W. Australia Exped., Z.M.H. 11665. Remarks. Recently Johnson (1964) showed that Porcellanella picta Stimpson, which many authors have considered synonymous with P. triloba, may be separated on the basis of a num- ber of constant characters. The material from Western Australia all agrees with P. triloba in Johnson’s restricted sense. Johnson assigned earlier records to one or the other of the two species as far as was possible on the basis of reexamined material and published descriptions and illustrations, and concluded that the only certain localities for P. triloba are N.E. Australia and Palau Islands. The Palau Islands record, however, was based on a third species of Por- cellanella recently described by Sankarankutty (1963, p. 273) as p. haigae. P. triloba should thus be considered a strictly Australian form, at least pending the re-examination of the speci- mens on which several published records were based. Males to 9.5 by 6.8 mm; non-ovigerous females to 9.0 by 6.7 mm; ovigerous females to 8.4 by 6.3 mm. Ovigerous females in August at Dampier Archipelago or vicinity, and in October off Adele I. Distribution. Definitely known only from Australia (Queensland, at Bowen and off Cape Capricorn). Now recorded from Western Aus- tralia. Genus POLYONYX Stimpson Diagnosis, Basal antennal segment strongly produced forward and broadly in contact with anterior margin of carapace; movable segments far removed from orbit. Carapace usually broader than long, particularly in females: strongly convex front to back. Front deflexed, trilobate or transverse, usually appearing trans- verse in dorsal view. Chelipeds large, differing in size and form. Dactyl of walking legs with at least two well-developed fixed claws and often with spinules on lower margin. Telson of abdomen seven-plated. Pleopods present in males of all but a few species. Key to Western Australian species of Polyonyx 1. Lateral margin of carapace armed with spines Lateral margin of carapace un- armed 2. (1) Ventral claw of dactyl of walking legs not nearly at a right angle to axis of dactyl Ventral claw of dactyl of walking legs nearly at a right angle to axis of dactyl 3. (1) Claws of dactyl of walking legs subequal or at least with dorsal claw well developed in comparison to ventral claw Ventral claw of dactyl of walking legs much larger than dorsal claw 4. (3) Ventral surface of chelae with a pronounced longitudinal crest in proximal half; a large lobe on inner margin of merus Ventral surface of chelae without such a crest; meral lobe vestigial or absent 5. (4) Dactyl of walking legs with a single rather large, stout spinule in addition to the two main claws; male with pleopods Dactyl of walking legs with two small, slender spinules in addition to the two main claws; male pleopods lacking 6. (3) Front transverse; merus of chelipeds with a vestigial lobe; propodus of walking legs with 12 or more ventral spinules Front with median lobe produced; merus of chelipeds with a well- developed lobe; propodus of walk- ing legs with four or five ventral spinules 2 3 suluensis telestophilus 4 6 triunguiculatus 5 obesulus hiunguiculatus transversus maccullochi Polyonyx suluensis (Dana) PoTcelana suluensis Dana 1852, p. 414 (Sulu Sea); 1855, pi. 26, fig. 4. Rathbun 1924, p. 30, pi. 1, figs. 15-16 (Cape Jaubert). Polyonyx denticulatus Paulson 1875a, p. 32, pi. 1, fig. 11 (nomen nudum); 1875b, p. 89, pi. 11, fig. 6 (Red Sea). Johnson 1958, pp. 98, 100. text-fig. 1. Polyonyx hexagonalis Zehntner 1894, p. 187, pi, 8, figs. 18, 18a (Amboina). Polyonyx suluensis, Haig 1964, p. 373, text-fig. 3. Diagnosis. Carapace roughly hexagonal, as broad as, or slightly broader than, long; surface with scattered long hairs. Lateral margins armed with spines. Front very broad, trilobate, median lobe prominent, acute. Merus of chelipeds with a prominent lobe on inner margin, armed with several spines. Carpus with about four strong spines on inner margin. Major cheliped nude or hairy; minor chela densely covered with long hairs. Propodus of walking legs with one ventral spinule in addition to three at distal end. Dactyl with two large, subequal fixed claws; no accessory movable spinules on lower margin. Material examined. 19, 45 miles W.S.W. of Cape Jaubert, 72-80 ft, May-July 1911. E. Mjdberg, U.S.N.M. 56435. 1 9 , off Ninety Mile Beach between Cape Jaubert and Wallai, 5 fm, Sept. 1929, A. A. Livingstone, A.M. P.14089. Remarks. Both specimens were ovigerous; the carapace length of one of them was 3.2 mm. In view of its abundance throughout most of its range, it is surprising that this species has been so seldom collected in Western Australia. Distribution. Indian Ocean (Red Sea; Sey- chelles; Saya de Malha; Cargados Carajos); Japan southward to Philippine Islands and East Indian Archipelago. In Australia reported only from Western Australia. Polyonyx telestophilus Johnson Polyonyx telestophilus Johnson 1958, pp. 98. 103, text- fig. 2 (Singapore). Haig 1964, p. 376, text-fig. 4. Diagnosis. Carapace roughly hexagonal, a little broader than long; nearly or completely devoid of hairs. Lateral margins armed with spines. Front very broad, trilobate, median lobe prominent, acute. Merus of chelipeds with a prominent inner lobe, minutely denticulate or crenulate and sometimes bearing a single, larger tooth. Carpus with distal portion of inner margin produced as a strongly convex lobe; less produced proximal portion crenulate or with one to three teeth. Chelipeds slightly hairy or devoid of hairs. Propodus of walk- ing legs with one ventral spinule in addition to three at distal end. Dactyl with two large, fixed claws, the ventral one distinctly larger and longer than the dorsal and directed at almost a right angle to axis of dactyl; no acces- sory movable spinules on lower margin. Material examined. 1 9 , between Malus and Gidley Is., Dampier Archipelago, 10 fm, l.vi.l960, Royce on “Davena”, W.A.M. 137-60. Remarks. The single Western Australian specimen, an ovigerous female, had a carapace length of 4.4 mm and breadth of 5.6 mm. John- son (1958) found the species to be strictly commensal with alcyonarian corals of genus Telesto at Singapore. Distribution. Gulf of Siam; Singapore; Amboina. Now recorded from Australia. Polyonyx biunguiculatus (Dana) Porcellana hiunguiculata Dana 1852, p. 411 (type locality not stated); 1855, pi. 26, figs. la-d. Polyonyx hiunguiculatus, Stimpson 1858, p. 229. Johnson 1958, pp. 100, 105, text-fig. 3. Haig 1964, p. 377. 7Porcellana hiunguiculata, Haswell 1882b, p. 147 (Holborn Island). IPolyonyx hiunguiculatus, Ortmann 1894, p. 30 (Thursday Island). Polyonyx tuherculosus var., Rathbun 1924, p. 31. pi. 1, fig. 17 (Cape Jaubert). Diagnosis. Carapace subrectangular, broader Chan long especially in females: surface devoid of hairs. Lateral margins unarmed. Front trilobate, median lobe subrectangular. Merus of chelipeds with a vestigial inner lobe. Inner margin of carpus straight, unarmed. No crest on proximal half of ventral surface of chela. Dorsal surface of chelipeds without hairs. Pro- podus of walking legs with one ventral spinule in addition to three at distal end. Dactyl with two large, subeoual fixed claws; two small, mov- able spinules on lower margin. No pleopods in males. Material examined. 1 9 , King Sound, B.M.N.H. 1 9 , 45 miles W.S.W. of Cape Jaubert, 72 ft, 7.vii.l911, E. Mjoberg, U.S.N.M. 56433. 112 16,2 miles W. of Legendre I., Dampier Archipel., 23 fm, 9.vi.l960, B. R. Wilson on “Davena”, W.A.M. 128-60. 1 0 , Gidley I., Dampier Archipel., 10 fm, l.vi.l960, R. D. Royce on *‘Davena'’, W.A.M. 129-60. 16, 39, W. approaches to Mer- maid Strait, Dampier Archipel., 20 fm, 27.V.1960, R. D. Royce on “Davena”, W.A.M. 117-60. 1^, Malus I., Dampier Archipel., 10 fm, 31.V.1960, R. D. Royce on “Davena”, W.A.M. 180-60. 26, Exmouth Gulf or Shark Bay, trawled, winter 1960, R. McKay on “Peron”, W.A.M. 76-62. Remarks. The U.S.N.M. specimen is part of the material reported by Rathbun (1924) as “Polyonyx tuberculosus de Man, var.”. The Queensland records of Haswell and Ortmann need to be verified in the light of Johnson’s (1958) revision of Polyonyx, in which he showed that P. hiunguiculatus has frequently been con- fused with other species; but specimens I have examined from several Queensland localities are definitely referable to P. hiunguiculatus. Ground colour pale yellow, thickly speckled with orange-red; the colour especially concen- trated on fingers of chelae. Broad bands on walking legs. Males to 6.9 by 8.4 mm; ovigerous females to 4.6 by 6.9 mm. Ovigerous females in May at Dampier Archipelago and in July at Cape Jaubert. Distribution. Indian Ocean (Eritrea; Sey- chelles: ? Ceylon'; western Pacific Ocean from Formosa Strait southward to East Indian Archi- pelago. Australia (Western Australia and Queensland) . Polyonyx obesulus Miers Porcellana obesula White 1847, p. 130 {nomen nudum: ? Madgica-Sima Islands). Adams in Adams and White 1848, p. iii. Polyonyx obesulus, Miers 1884, p. 272. pi. 29, fig. D (Port Darwin: Prince of Wales Channel; Port Denison; West Island). Henderson 1888, p. 115 (Flinders Passage). Johnson 1958. pp. 99, 108, text-fig. 4. Haig 1964, p. 378. 7PoTcellana (Polyonyx) tuberculosa de Man 1888, p. 424. pi. 19, fig. 1 (Amboina). Polyonyx parvidens Nobili 1905, p. 161 (Gulf of Iran). Diagnosis. Carapace subovate, broader than long, particularly in females; surface devoid of hairs. Lateral margins unarmed. Front trilo- bate, median lobe rounded or subrectangular. Merus of chelipeds with a vestigial inner lobe. Inner margin of carpus straight, unarmed. No crest on proximal half of ventral surface of chela. Dorsal surface of chelipeds without hairs. Propodus of walking legs with one ventral spinule in addition to three at distal end. Dactyl with two large, subequal fix.ed claws; a single rather large, stout spinule on lower margin. Material examined. 16, 23 miles S.W. of Troughton I., 25 fm, 22.x. 1962, R. W. George on “Dorothea”, W.A.M. 336-62. 16, Broome, W.A.M. 9457/8. 16, Broome, June 1932, A.M. P.10265. 1 9 , near entrance to Roebuck Bay, 5-8 fm, 26. ix. 1929, A. A. Livingstone, A.M. P.14453. 15, 2 miles S.W. of Peak I., 10 fm, 18.vi.l960, B. R. Wilson on “Davena”, W.A.M. 136-60. 19,9 miles S. W. of Geraldton, 20 fm in large sponge, 30.x. 1956, R. W. George, W.A.M. 43-62. Remarks. The type locality of P. obesulus, ac- cording to White (1847) and as indicated in re- cords at the British Museum (Natural History), was the Madgica-Sima Group, which is in the Ryukyu Archipelago (see Haig 1964, p. 380). But Adams in Adams and White (1848, p. iii) stated that what were evidently the type specimens were dredged in the Sulu Sea. The latter locality seems more likely in view of the fact that the species is not otherwise repox’ted as far north as the Ryukyu Islands. Males to 5.0 by 7.1 mm; ovigerous female 5.7 by 7.7 mm. Ovigerous female in October at Geraldton. Distribution. Indian Ocean including Gulf of Iran; western Pacific Ocean, from Philippine Islands (perhaps from Ryukyu Islands?) south- ward to East Indian Archipelago. Australia (Northern Territory and Queensland). Now re- corded from Western Australia. Polyonyx triunguiculatus Zehntner Polyonyx triunguiculatus Zehntner 1894, p. 185 (Amboina). Johnson 1958, pp. 99, 110. Polyonyx acutifrons de Man 1896, p. 384 (Atjeh,. Sumatra); 1898, pi. 32, figs. 49, 49a-d. Diagnosis. Carapace subovate or subrectangu- lar, broader than long, especially in females; surface devoid of hairs. Lateral margins unarmed. Front trilobate, median lobe prom- inent, narrow, acute. Merus of chelipeds with a well-developed lobe on inner margin. Inner margin of carpus straight, unarmed. Chela with a pronounced longitudinal crest on proximal half of ventral surface. Dorsal surface of chelipeds without hairs. Propodus of walk- ing legs with one ventral spinule in addition to three at distal end. Dactyl with two large, subequal fixed claws; one or two small movable spinules on lower margin. Material examined. 1 cheliped, Broome, June 1932, A.M. P.14454. 1 9 , 40 miles W. of Cape Jaubert, 23 fm on sponge, 13.x. 1962. R. W. George on “Dorothea”, W.A.M. 328-62. 29, 3-4 miles off N.E. end Delambre I., Dampier Archipel.. 10 fm, 5.vi.l960, B. R. Wilson on “Davena”, W.A.M. 126-60. 15, 2 mi. S.W. of Peak I., 10 fm, 18.vi.l960, B. R. Wilson on “Davena”, W.A.M. 136-60. 15, Exmouth Gulf or Shark Bay, trawled, winter 1960, R. McKay on “Peron”, W.A.M. 76-62. 3 5. 19, Shark Bay, 8.vii.l962, “Peron”, W.A.M. 334-62. Remarks. Colour of preserved specimens pale orange, with darker orange areas on carapace and chelipeds. On the dorsal surface of the carapace there is a large patch of the darker shade on either side, leaving a pale median stripe with parallel sides which joins pale areas on frontal and posterior parts of the carapace. On the chelipeds there is irregularly-shaped light and dark mottling. Males to 7.0 by 8.9 mm; non-ovigerous female 3.7 by 4.7 mm; ovigerous females to 6.9 by 9.3 mm. Ovigerous females at Dampier Archipelago in June, Shark Bay in July, and Cape Jaubert in October. Distribution. Throughout the Indian Ocean including the Red Sea; in the western Pacific reported only from Singapore, Amboina, and Sumatra. Now recorded from Australia. Polyonyx transversus (Haswell) Porcellana transversa Haswell 1882a, p. 759 (Bowen)* 1882b. p. 150. Polyonyx transversus, Baker 1905, p. 262, pi. 36, figs. 2, 2a (off Newland Head). Hale 1927a, p. 83, text-fig. 8o! McNeill and Ward 1930, p. 363, pi. 59, fig. 3 (Botany Bay; 113 Port Curtis; Western Port). Pope 1946, p. 91, text-fig. (Pittwater). Dakin 1952', p. 352 (Gunnamatta Bay). Johnson 1958, pp. 99, 115. Diagnosis, Carapace subovate or subrectangu- lar, broader than long particularly in females: surface devoid of hairs except for fringe on frontal margin. Lateral margins unarmed. Front transverse, with median lobe very slightly developed. Merus of chelipeds with a vestigial lobe. Inner margin of carpus developed into a prominent, convex lobe, broadest distally and unarmed. Dorsal surface of chelae and inner margin of carpus thickly hairy. Propodus of walking legs with more than 12 spinules along lower margin. Dactyl with two fixed claws, vent- ral much larger than dorsal one; two stout spines on lower margin. Material examined. 1 ^ , 19, Cockburn Sound, 5.iv.l959, Stn. 146, W.A.M. 113-60. 4 9, S. end Careening Bay, Garden I., Cockburn Sound, from Chaetopterus tubes. I.iii.l959, Mar. Group Nat. Club. W.A.M. 119-60. US, 119 Careening Bay. Garden I., from Chaetopterus tubes on intertidal flats. 26.xi.1961. B. R. Wilson and marine group. W.A.M. 37-62. Remarks. The type specimens were taken from the siphons of a bivalve mollusc. Aspergillum. Johnson (1958) stated that there is no definite information as to the habitat of the species in the more southern records. However, McNeill and Ward (1930, pp. 363 and 364) gave rather detailed information about the habitat of speci- mens in the collections of the Australian Museum. Of specimens from Botany Bay they stated: “Both specimens were taken from “U”- shaped worm tubes, occupied by a species of the polychaet Chaetopterus, where they were found reclining in the inflated basal portion of their sanctuaries, at a depth of about fourteen inches frcm the surface of the tidal flats exposed at low tide.” Of the Port Curtis specimens collected by M. Ward: “He remarked that they [worm tubes 1 were found in the soft mud at the ex- treme lew tide line, or close thereto in shallow drains and pools. In each worm tube examined a male and female crab were present. The tubes were not more than one foot deep in the mud, and, owing to the fragile nature of their struc- ture. great difficulty was experienced in digging them out.” They concluded: “The fact that so very few specimens have been recorded. . . . com- bined with the evidence already to hand, sug- gest that the creature. ... is invariably a com- mensal. This would account for its apparent rarity in collections.” Further information on the Polyonyx-Chaetopterus association is pro- vided by Pope (1946). The ground colour of preserved specimens is pale orange, with mottlings of darker orange on the carapace and chelipeds. Males to 6.4 by 8.3 mm; non-ovigerous females to 7.6 by 10.5 mm: ovigerous females to 8.0 by 11.0 mm. All three lots of specimens examined included egg- bearing females. Distribution. Apparently an Australian endemic; reported from Queensland, New South Wales. Victoria, and South Australia. Now recorded from Western Australia. Polyonyx maccullochi, sp. nov. (Fig. 3) Polyonyx obesulus (?), Grant and McCulloch 1906, p. 41 (Port Curtis). Description. Carapace subovate, only slightly broader than long (c. 1.2 times in both males and females); surface smooth, hairless except for fringe on frontal margin; sides sparsely hairy. Front narrow, with a strong median lobe forming a slightly acute angle and extending well beyond the obtuse lateral lobes; in dorsal view appearing transverse or slightly convex. All segments of antennal penduncle smooth; flagellum long, slender, with a few vestigial hairs. Ventral surface of outer maxillipeds not hairy. Chelipeds rather sparsely hairy; no hairs on upper surface, some on lower surface of merus and chela and in gape of fingers, and a sparse fringe on outer margins of chelae. Upper surface of chelipeds smooth, sometimes lightly punctate but without rugosities or tubercles. Merus with a well-developed, convex lobe on inner margin. Carpus with inner margin pro- duced into a prominent lobe, unarmed and evenly convex throughout its length. Males: Major chela swollen, without crest on surface; fingers strongly curved outward ; fixed finger with a strong rounded tooth on cutting edge at base, movable finger with a tooth at base and another about midway along cutting edge; outer margins of fingers smooth. Minor chela rather slender, without crest, less swollen than major chela; fingers long and slender, not out- curved, their inner margins minutely crenulate; outer margin of fixed finger wth a row of small, sharp tubercles, movable finger with a few similar tubercles on outer surface near tip. Females: Major chela with fingers longer and much less out-turned than in males, on the whole resembling minor chela; teeth on cutting edges of fingers not well developed. Minor chela as in males. Walking legs with a fringe of long, plumose hairs on margins. Merus un- armed on lower margin, that of third walking leg about twice as long as broad. Propodus armed on lower margin with four or five stout movable spinules including the pair at distal end, that of third walking leg about twice as long as broad. Dactyl very narrow in com- parison to propodus: one or two minute spinules on its lower margin proximal to bifid tip. Male with a pair of pleopods. Holotype male (4.2 by 5.1 mm). Entrance to Roebuck Bay, 9 fm, 15.viii.l929, A. A. Living- stone, A.M. P.14117. Paratypes. QUEENSLAND: 19, Port Curtis, 7 fm, F. E. Grant, A.M. G.5754. WESTERN AUSTRALIA: 1^, 19, Broome, June 1932, A.M. P.10268. 19, entrance to Roebuck Bay. 9 fm. 15.viii.l929, A. A. Livingstone, A.M. P.14123. IS, 19, Denham, Shark Bay, 4-22. ix. 1905, St. 65, Hamburg S.W. Australia Exped., Z.M.H. 11733. 19 (juv.), c. 2h miles S.W. of Denham, Shark Bay, 3 m, 10.vi.l905, St. 7, Hamburg S.W. Austra- lia Exped., Z.M.H. 11651. 1 <5 (juv.), Brown Sta- tion, Dirk Hartog I., Shark Bay, 2a-42 m. 18.vi.l905, St. 28, Hamburg S.W. Australia Ex- ped., Z.M.H. 11698. 1^, 19, entrance to Useless Inlet, Shark Bay. -4-2 m, 13. ix. 1905, St. 18. 114 Figure 3 . — Polyonyx maccullochi. A, dorsal view of holotype; B, anterior view of frontal region, paratype; C, propodus and dactyl of left third walking leg, para- type; D, left third walking leg, paratype. Scale for A = 5 mm; for B, 3 mm; for C, 1 mm; for D, 2 mm. Hamburg S.W. Australia Exped., Z.M.H. 11679. 23, Shark Bay, 8.vii.l962, “Peron”, W.A.M. 334- 62. Remarks. Polyonyx maccullochi belongs to a well-defined group of Polyonyx species (desig- nated the “P. sinensis group” by Johnson 1958) in which the lateral margins of the carapace, and the inner margins of the merus and carpus of the chelipeds, are unarmed; the chelipeds are hairy; and the dorsal claw of the dactyl of the walking legs is much smaller, than the ventral claw. The two species of this group now re- ported from Western Australia, P. maccullochi and P. transversus, can be readily distinguished from each other by the characters given in the key and diagnoses. Of the other Indo- West Pacific members of group sinensis treated by Johnson, only P. cometes Walker and P. utinomii Miyake agree with P. maccullochi in having a well-developed lobe on the inner margin of the merus of the chelipeds. In P. cometes the dorsal surface of the chelipeds is densely hairy, whereas in P. maccullochi this surface is devoid of hairs. In P. utinomii the median fron- tal lobe is weakly produced, and the propodus of the walking legs bears only three spinules on its lower margin. The specimen from Queensland listed above is the one questionably referred by Grant and McCulloch (1906) to Polyonyx obesulus. Substrates, where mentioned on the labels with the material examined, were sand or sand and mud. There was no indication of com- mensalism, but it is likely that members of this species occur at times in' association with Chaetopterus or other organisms as do most species of group sinensis. Males to 5.1 by 6.2 mm; females to 5.1 by 6.4 mm. Distribution. Known only from the localities listed above, in Western Australia and Queens- land. Genus RAPHIDOPUS Stimpson Diagnosis. Basal antennal segment strongly produced forward and broadly in contact with anterior margin of carapace; movable segments far removed from orbit. Carapace subovate, broader than long. Front transverse and tri- dentate, not prominent nor greatly defiexed. Chelipeds subequal. Dactyl of walking legs a straight, slender spine, with an acute tip; no supplementary spinules. Telson of abdomen seven-plated. Raphidopus ciliatus Stimpson Raphidopus ciliatus Stimpson 1858, pp. 228, 241 (Hong Kong); 1907, p. 185, pi. 22, fig. 5. Rhaphidopus ciliatus. Grant and McCulloch 1906, p. 42 (Port Curtis). Diagnosis. Sides of carapace, inner and outer margins of chelipeds, upper and lower surface of chelae, and walking legs very heavily setose. Lateral margins of carapace and outer margin of carpus of chelipeds armed with spines. Upper surface of carpus with a median longitudinal row of tubercles or spinules. Material examined. 1 $ , Roebuck Bay, shore on sand flat between tides, 8.viii.l929, A. A. Liv- ingstone, A.M. P.13737. Remarks. The single specimen had a carapace length of 6.0 mm and breadth of 7.7 mm. Distribution. Japan; Hong Kong. Australia at Port Curtis, Queensland. Now recorded from Western Australia. Acknowledgments I am most grateful to Dr. R. W. George of the Western Australian Museum for placing his in- stitution’s ccllection of Porcellanidae at my dis- posal and for his patience in answering numer- ous questions concerning it. I am indebted to Mr. P. A. McNeill and Dr. J. C. Yaldwyn for the loan of specimens housed in the Australian Museum, and in particular to Mr. McNeill for providing information on and photographs of the types of several porcellanid species: and to Dr. A. Panning and Dr. F. A. Chace, Jr., who sent additional material from their respective institutions, the Zoologisches Museum, Hamburg, and the U.S. National Museum. I also wish to thank Dr. Isabella Gordon, through whose kind- ness I was able to examine type and other ma- terial during a visit to the British Museum (Natural History). My sincere thanks go also to several persons who assisted this investigation in other ways. Dr. J. Forest of the Museum d’Histoire Naturelle, Paris, answered questions concerning type speci- mens in his institution. Dr. L. B. Holthuis of the Rijksmuseum van Natuurlijke Historie, Lei- den, provided valuable advice on certain prob- lems of nomenclature. An exchange of opinions and information with Dr. D. S. Johnson of the University of Malaya, Singapore, helped to clarify several points. I also v/ish to thank the administration of the Allan Hancock Foundation for providing the working space in which this investigation was carried out. List of Genera and Species Genus Petrocheles Miers australiensis (Miers) Genus Petrolisthes Stimpson scabriculus (Dana) militaris (Heller) moluccensis (de Man) boscii (Audouin) hasivelli Miex*s teres Melin ohshimai (Miyake) Genus Pachycheles Stimpson sculptus (H. Milne Edwards) johnsoni sp. nov. granti sp. nov. pisoides (Heller) Genus Pisidia Leach spinuligera (Dana) cf. spinuUfrons (Miers) dispar (Stimpson) Genus Porcellana Lamarck habei Miyake gravelei Sankolli ornata Stimpson nitida Haswell furcillata sp. nov. Genus Porcellanella White triloba White Genus Polyonyx Stimpson suluensis (Dana) telestophilus Johnson biunguiculatus (Dana ) obesulus Miers triunguiculatus Zehntner transversus (Haswell) maccullochi sp. nov. Genus Raphidopus Stimpson ciliatus Stimpson References Adams, A. and White, A. (1848). — Crustacea. In “The Zoology of H.M.S. Samarang . . . during the years 1843-1846.” (London.) Audouin, V. (1826). — Explication sommaire des planches de crustaces de I’figypte et de la Syrie, publiees par Jules-Cesar Savigny Description de r^Jgypte, ou Recueil des observations et des recherches qui ont ete faites en figypte pendant I’Expedition de I’armee frangaise, Hist. Nat. 1 (4): 77-98. (Paris.) Baker, W. H. (1905). — Notes on South Australian decapod Crustacea. Part III. Trans. Roy. Soc. S. Aust. 29: 252-269. Borradaile, L. A. (1900).— On the Stoma topoda and Macrura brought by Dr. Willey from the South Seas. In “Zoological Results Based on Material from New Britain, New Guinea. Loyalty Islands and Elsewhere, Collected During the Years 1895, 1896 and 1897, by Arthur Willey ... Part IV”: 395-428. (Cambridge.) Dakin, W. J. with Bennett, Isobel and Pope, Elizabeth (1952). — “Australian Seashores. A Guide for the Beach-lover, the Naturalist, the Shore Fisherman, and the Student.” (Angus & Robertson: Sydney.) Dana, J. D. (1852-53). — Crustacea. United States Explor- ing Expedition during the years 1838, 1839, 1840, 1841, 1842 ... 13: (viii) 1-685 (1852), 686-1618 (1853). (Philadelphia.) (1855).— Crustacea. United States Exploring Expedition during the years 1838, 1839, 1840, 1841. 1842 ... 14, Atlas: 1-27, pis. 1-96. (Philadelphia.) Edmondson, C. H. (1925).— Marine zoology of tropical Central Pacific. Crustacea. Bull. Bishop Mus., Honolulu 27: 3-62. George, R. W. (1961). — Narrative of the personnel from the Western Australian Museum who took part in the Bernice P. Bishop Museum of Hawaii Expedition to Western Australia (May-June, 1960). Annu. Rep. Mus. W. Aust. 1959-60: 24-25. Gordon, Isabella (1931).— Brachyura from the coasts of China. J. Linn. Soc. (Zool.) 37: 525-558. (1960). — Additional note on the porcellanid sea-anemone association. Crustaceana 1: 166-167. 116 Grant, F. E. and McCulloch. A. R. (1906).— On a collec- tion of Crustacea from the Port Curtis district, Queensland. Proc. Linn. Soc. N.S.W. 31 : 2-53. Gravely, F. H. (1927). — The littoral fauna of Krusadai Island in the Gulf of Manaar. Orders Decapoda (except Paguridea) and Stoma- topoda. Bull. Madras Govt. Mus. (n.s.) 1 (1): 135-155. Haig, Janet (1960). — The Porcellanidae (Crustacea Anomura) of the eastern Pacific. Allan Hancock Pacif. Exped. 24; (viii) 1-440. — (1964). — Papers from Dr. Th. Mortensen's Pacific Expedition 1914-1916. 81. Porcellanid crabs from the Indo-West Pacific, Part I. Vidensk. Medd. Dansk Naturh. Foren. Kbh. 126: 355-386. Hale, H. M. (1927a). — “The Crustaceans of South Aus- tralia.” Part I. (Govt. Printer; Adelaide.) ( 1927b). — The fauna of Kangaroo Island, South Australia. No. 1. The Crustacea. Trans. Roy. Soc. S. Aust. 51; 307-321. (1929). — Notes on the fauna of Dirk Hartog Island, Western Australia. No. 4. Crustacea. Trans. Roy. Soc. S. Aust. 53: 67-70. Haswell, W. A. (1882a). — Description of some new species of Australian Decapoda. Proc. Linn. Soc. N.S.W. 6: 750-763. (1882b). — “Catalogue of the Australian Stalk- and Sessile-eyed Crustacea.” (Australian Mus.; Sydney.) Heller, C. (1862). — Neue Crustaceen, gesammelt wahrend der Weltumseglung der k. k. Fregatte Novara. Zweiter vorlaufiger Bericht. Verh. Zool.-Bot. Ges. Wien 12; 519-528. (1865). — Crustaceen. Re'se der oesterreichis- chen Fregatte ‘Novarra’ um die Erde, in den Jahren 1857, 1858, 1859 . . . Zoologischer Theil 2 (3) (1); 1-2B0. (Vienna.) Henderson. J. R. (1888). — Report on the Anomura col- lected by H.M.S. Challenger during the years 1873-76. Rpt. Zool. Challenger Exped. 27: (xi) 1-221 (London, Edinburgh & Dublin.) (1893). — A contribution to Indian carcino- logy. Trans. Linn. Soc. Lond. (Zool.) (2) 5: 325-458. Jacquotte, R. (1964). — Notes de faunistique et de biologie marines de Madagascar I. Sur I’association de quelques crustaces avec des cnidaires recifaux dans la region de Tulear (sud-ouest de Madagascar). Bxill. Rec. Trav. Sta. Mar. Endoume 32 (48): 175-178. Johnson, D. S. (1958). — The Indo-West Pacific species of the genus Polyonyx (Crustacea, Decapoda, Porcellanidae). Ann. Zool., Agra 2: 95-118. (1960). — On a porcelain crab, Petrolisthes ohshimai (Miyake), from Christmas Island. Indian Ocean, with a note on the genus Neopetrolisthes Miyake. Crustaceana 1: 164- 165. (1964). — On the species Porcellanella picta S t i m p s o n (Decapoda, Porcellanidae). Crustaceana 7; 98-102. Kingsley, J. S. (1880). — On a collection of Crustacea from Virginia, North Carolina, and Florida, with a revision of the genera of Crangonidae and Palaemonidae. Prec. Acad. Nat. Sci. Philad. 31: 383-427. Lenz, H. (1905). — Wissenschaftliche Ergebnisse der Reisen in Madagaskar und Ostafrica in den Jahren 1889-95 von Dr. A. Voeltzkow. Ostafrikanische Dekapoden und Stoma- topoden. Abh. Senckenb. Naturf. Ges. 27; 341-392. McCulloch, A. R. (1913). — Studies in Australian Crustacea. No. 3. Rec. Aust. Mus. 9; 321-353. and McNeill, F. A. (1923). — Notes on Aus- tralian Decapoda. Rec. Aust. Mus. 14; 49-59. McNeill, F. A. (1953). — Carcinological notes No. 2. Rec. Aust. Mus. 23; 89-96. and Ward, M. (1930). — Carcinological notes. No. 1. Rec. Aust. Mus. 17: 357-383. Man. J. G. de (1888). — Bericht liber die von Herrn Dr. J. Brock im indischen Archipel gesammelten Decapoden und Stomatopoden. Arch. Naturgesch. 53 (1): 215-600. (1896). — Bericht liber die ... zu Atjeh, an den westlichen Klisten von Malakka, Borneo und Celebes sowie in der Java-See gesam- melten Decapoden und Stomatopoden. Dritter Theil. Zool. Jb. Syst. 9: 339-386. (1898).— Bericht liber die ... zu Atjeh, an den westlichen Klisten von Malakka, Borneo und Celebes sowie in der Java-See gesam- melten Decapoden und Stomatopoden. Sechster (Schluss-) Theil. Zool. Jb. Syst. 10; 677-708. Melin, G. (1939). — Paguriden und Galatheiden von Prof. Dr. Sixten Bocks Expedition nach den Bonin- Inseln 1914. K. Svenska VetenskAkad. Handl. (3) 18 (2) ; 1-119. Miers, E. J. (1876a). — Descriptions of some new species of Crustacea, chiefly from New Zealand. Ann. Mag. Nat. Hist. (4) 17; 218-229. — (1876b). — “Catalogue of the Stalk- and Sessile-eyed Crustacea of New Zealand.” (Colonial Mus. and Geol. Surv. Dept., N.Z.) (1884). — Crustacea. In “Report on the Zoo- logical Collections Made in the Indo-Pacific Ocean during the voyage of H.M.S. ‘Alert’ 1881-2”: 178-322, 513-575. (Brit. Mus. (Nat. Hist.): London.) Milne Edwards, H. (1837). — “Histoire Naturelle des Crustaces, Comprenant I’Anatomie, la Physiologic et la Classification de ces Anim- aux.” Vol. 2. (Paris.) Miyake, S. (1937). — A new crab-shaped anomuran living commensally with a gigantic sea-anemone (Neopetrolisthes ohshimai gen. et sp. nov.). Zool. Mag., Tokyo 49: 34-36. (1942). — Studies on the decapod crustaceans of Micronesia. III. Porcellanidae. Palao Trop. Biol. Stud. 2; 329-379. (1943). — Studies on the crab-shaped Anomura of Nippon and adjacent waters. J. Dep. Agric. Kyushu Univ. 7 ; 49-158. (1961). — Three new species of the Anomura from Japan (Crustacea, Decapoda). J. Fac. Agric. Kyushu Univ. 11: 237-247. Nobili, G. (1905). — Decapodes nouveaux des cotes d’Arabie et du Golfe Persique (diagnoses preliminaires). Bull. Mus. Hist. Nat., Paris 11: 158-164. (1906a). — Mission J. Bonnier et Ch. Perez (Golfe Persique, 1901). Crustaces decapodes et stomatopodes. Bull. Sci. Fr. Belg. 40: 13-159. (1906b). — Faune carcinologique de la Mer Rouge. Decapodes et stomatopodes. Ann. Sci. Nat. Zool. (9) 4: 1-347. Ortmann, A. (1892). — Die Decapoden-Krebse des Strass- burger Museums . . . IV. Theil. Die Abthei- lungen Galatheidea und Paguridea. Zool. Jb. Syst. 6; 241-326. (1894). — Zoologische Forschungsreisen in Australien und dem Malayischen Archipel . . . ausgefuhrt in den Jahren 1891-1893 von Dr. Richard Semon . . . Fiinfter Band: Sys- tematik und Thiergeographie. I. Lieferung. Crustaceen. Denkschr. Med.-Naturw. Ges. Jena 8; 3-80. Paulson, O. M. (1875a). — [Carcinological observations.] Mem. Soc. Nat. Kieff. 4 (1); 27-32. (1875b). — “Studies on Crustacea of the Red Sea.” Part I. (English translation, 1961; Israel Program for Scientific Translations.) Pope, Elizabeth C. (1946). — Chaetopterus — a strange and beautiful worm. Aust. Mus. Mag. 9: 87-92. Rathbun, Mary J. (1924). — Results of Dr. E. Mjdberg’s Swedish scientific expeditions to Australia 1910-1913. 37. Brachyura, Albuneidae and Porcellanidae. Ark. Zool. 16 (23): 1-33. Sankarankutty. C. (1963). — On three species of porcel- lanids (Crustacea-Anomura) from the Gulf of Mannar. J. Mar. Biol. Ass. India 5: 273-279. Sankolli, K. N. (1963). — On a new species of porcellanid crab (Decapoda. Anomura) from India. J. Mar. Biol. As5. India 5: 280-283. Saville-Kent, W. (1897). — “The Naturalist in Australia.” (Chapman & Hall: London.) Stead, D. G. (1898). — Notes on the habits of some of the Australian malacostracous Crustacea. Zoologist (4) 2: 202-212. 117 Stimpson, W. (1858). — Prodromus descriptionis animalium evertebratorum . . . Pars VII. Crustacea Anomura. Proc. Acad. Nat. Sci. Philad. 10; 225-252. (1907). — Report on the Crustacea (Brachyura and Anomura) collected by the North Pacific Exploring Expedition, 1853-1856. Smithson. Misc. Coll. 49: 1-240. Walker, A. O. (1887). — Notes on a collection of Crustacea from Singapore. J. Linn. Soc. (Zool.) 20: 107-117. Ward, M. (1928). — The Crustacea of the Capricorn and Bunker Groups, Queensland. Aust. Zool. 5: 241-246. White, A. (1847). — “List of the Specimens of Crustacea in the Collection of the British Museum.” (Brit. Mus. (Nat. Hist.): London.) (1852). — Descriptions of some apparently new species of Annulosa, (collected by Mr. Macgillivray during the voyage of H.M.S. Rattlesnake). In “Narrative of the Voyage of H.M.S. Rattlesnake . . . During the years- 1846-1850 . . .” Vol. 2, Appendix 6. (London.) Whitelegge, T. (1889). — List of the marine and fresh- water invertebrate fauna of Port Jackson and the neighbourhood. J. Roy. Soc. N.S.W. 23: 163-323. (1897). — The atoll of Funafuti, Ellice Group: its zoology, botany, ethnology, and general structure based on collections made by Mr Charles Hedley . . . VI. The Crustacea. Mem. Aust. Mus. 3: 127-151. Zehntner, L. (1894).— Voyage de MM. M. Bedot et C. Pictet dans TArchipel Malais. Crustaces de I’Archipel Malais. Rev. Suisse Zool. 2; 135-214. 118 11. — A new species of Nectria (Asteroidea, Goniasteridae) from Western Australia By S. A. Sheperd* and E. P. Hodgkin** Manuscript received 16th February, 1965; accepted 27th April, 1965. Abstract A new species of starfish belonging to the genus Nectria is described and figured. All specimens collected so far are from the lower west coast, between 29° and 34° South latitude in shallow water down to 55 feet. Nectria wilsoni sp. nov. (Fig. 1) Diagnosis Disk large, arms thick at base and slightly tapering. The plates on the aboral surface of the disk form elevated paxilliform tabulae, as is typical of the genus, and are covered with relatively large protuberant hemispherical granules. The peripheral granules form a radiating fringe. Distally the aboral tabulae are low and have a large central granule surrounded by a ring of small granules. Papulae are rare or absent from the distal part of each arm. The species differs from other members of the genus in the size and appearance of the granules of the aboral tabulae, both on the arms distally and on the disk. The granules are coarser and larger and the tabulae some- what more elevated and closer together than in N. ocellata (Perrier), to which it otherwise bears some resemblance. Description of holotype Rays 5. R/r is 77/26 mm R 3r. Br is 30 mm at the base, 15 mm at the middle and 12 mm near the tip of the ray. Aboral plates of disk and base of rays form large paxilliform tabulae, elevated about 3-4 mm and flat on top. The shaft of each tabula hourglass-shaped and about 2-3 mm across at narrowest point widening to 4,-6 mm across on the crown. Crowns of tabulae subequal, more or less rounded in outline and crowded so that they frequently touch adjacent crowns. Tabulae covered with swollen hemispherical protuberant granules of from 0.4 to 1.0 mm across; the peripheral granules similar in shape and character, though sometimes longer. Each disk tabula has about 16 to 24 peripheral granules and 10 to 20 inner granules among which there may be considerable disparity in size. On the rays, the tabulae are smaller and lower and on the distal two thirds of ray are crowded, * c/o The South Australian Museum, Adelaide. South Australia. ** Department of Zoology, University of Western Australia. Nedlands. Western Australia. low, and difficult to distinguish. These distal plates each have a large central granule, pro- tuberant and hemispherical, about 1 mm across and surrounded by a ring of 8-12 small hemispherical granules. Papulae on the disk in groups of 6-14 between the connecting ossicles of the aboral plates and in decreasing numbers to about half way along the ray, rare or wanting on the distal half, absent between the marginal plates. Madreporite low, about 2.5 mm across and hidden under the tabulae midway between centrum and margin. Marginal plates distinct, about 21-24 in each series on each side of ray; proximally higher than long and covered with a more or less uniform coat of granules, but on the distal half squarish and a central granule on each plate is enlarged and dominant (fig. 1, B). Oral intermediate plates covered with coarse rounded granules close enough to obscure the outlines of the plates. Ambulacral plates with 2 or 3 (sometimes 4 proximally) furrow spines about 2 mm long and more or less rounded and blunt. Subambulacral spines 2, similar to furrow spines but shorter and thicker, occasionally a second series of granuliform subambulacral spines. Two pedicellariae were detected behind the furrow series each with 4 or 5 slender slightly curved spines bending inwards against one another (fig. 1, E). They were not seen else- where. Oral plates with 6 or 7 stout prismatic or quadrate marginal spines, the innermost stoutest, flattened where they are contiguous. On the surface of each plate there are 3 or 4 stout prismatic spines with rounded edges and distally on each plate a group of 6 to 8 small granules. Colour All specimens have been a strong red when fresh, varying from deep orange to magenta on both surfaces. The skin of the papular areas and between the granules of the oral sur- face may be pale, almost white. Material Examined The Holotype (W.A.M. no. 3-65) is labelled “Sorrento Beach near Perth, 2 fath., 27/1/63” and was collected by B. R. Wilson. Nine other specimens have been examined and are designated paratypes, eight of these are lodged in the Western Australian Museum and the ninth in the South Australian Museum. 119 Figure \.—Necria wilsoni sp. nov. A-E: Holotype. W.A.M. No. 3>65. P: Paratype. W.A.M. No. 1-65. G: Paratype, S.A.M. No. K 613. 120 W.AM. 18-59. R 63 mm Eagle Bay, Cape Naturaliste, 30 ft., rock. W.A.M. 19-59. R 75 mm Dunsborough. 30 ft. W.A.M. 2-62. R 85 mm Hamelin Bay near Cape Leeuwin, on jetty piles. W.A.M. 3-62. R 110 mm Off Beagle Island (29° 50' S.), probably in a cray- fish pot. W.A.M. 1-65. R 54 mm Dunsborough, at 55 ft. on rock. W.A.M. 2-65. R 90 mm Sorrento reef, near Fremantle, on sea grass. W.A.M. 4-65. R 90 mm Sorrento reef. W.AM. 5-65. R 83 mm Sorrento beach, 18 ft., on sea grass, near limestone. S.A.M. K 613. R 69 mm Hall Bank near Fremantle. With the exception of W.A.M. 3-62, all were collected by B. R. Wilson. These specimens show some variability in the coarseness of the granulation, which is parti- cularly noticeable on the aboral tabulae. Two of the .smaller paratypes (W.A.M. 1-65 and S.A.M. K613) have smaller and less protuberant granules 20 mm). This recogni- tion of recruits as distinct from adults does offer increased opportunities in analysing genetic and age composition of population through a series of years. Mortality and longevity The heavy field mortality precluded establish- ment of detailed survivorship curves. For example in November 1961 on Lighthouse swamp 312 juveniles were individually marked. The following November a sample of 151 frogs yielded 22 adults of which only 4 were of the series of 312 juveniles marked the previous year. Recoveries were much higher in the early years. These have not been plotted because there is so much difference between the early and later years; a fact reflected in the change in numbers of adults and juveniles in the population (Tables 2, 3 and 4). The longest recorded period for an individual survival at each locality was; Lighthouse swamp, 2 years, marked as adult November 1960 re- captured November 1962; Bagdad soak, 2J years, marked as juvenile November 1961, recaptured March 1964; Negri soak, 2h years, marked as juvenile November 1960, recaptui'ed March 1963. Population size Population estimates. These have been made by the method in Main (1961), except in 1964 when simple recapture data were used to cal- culate population estimates by means of a Lin- coln index. The results are shown in Table 1. TABLE 1 Estimates of size of post-breeding papulatioiis of C. insignifera at three swamps on Rottnest Island. Your Lighthouse liuKthui Negri 19.55 1,021 1959 261 I960 •■m 189 1961 580 165 1962 702 817 196;i 1964 1.:J41 Errors in method of making estimates. In dry seasons disturbed frogs retreat to secure hiding places and so are not caught. Also, in dry seasons there is not a good admixture of marked and unmarked animals and therefore there is a tendency to get spurious population estimates. November 1962 was a dry period, while the November 1961 sampling period was overcast and with dew each night. During the November 1964 sampling period light rain showers were common. The census in 1955 (Main, 1961) followed an exceptionally wet season and somewhat similar wet conditions prevailed during 1964. There is every reason to believe that the high estimates of these years truly represent the population size. Age structure The specimens have been categorised into adult and juvenile using the snout-vent lengths determined in Figure 1. Data are presented in Tables 2, 3 and 4. 123 40 30 20 10 30 20 10 N = 49 I 1 ' I 28.1 Nov. 1959 Mac 1960 June 1960 Nov 1960 Length in mnn. Figure 1. — Histograms of snout vent measurements of two samples of Crinia insignifera collected from Lighthouse soak, Rottnest Island. Upper. Sample collected June, 1960. Lower. Sample collected November, 1960. 124 Annual changes within localities All November samples from each locality show a pattern characterised by greater adult frequency in the earlier years and greater juvenile frequency in later years. These changes are most marked in the Lighthouse swamp and Negri soak populations in which the differences are statistically highly significant (Lighthouse swamp p < 0.001, Negri soak p > 0.001, < 0.01), at least pronounced at Bagdad soak. These dif- ferences reflect changes in adult survival and re- cruitment, particularly as each of these contri- butes to the population structure: survival was good in the early years and recruitment poor, later the adult survival was poor and recruit- ment relatively good. While the foregoing is true for the series of observations, comparison of each observation with the preceding one shows that at Lighthouse swamp statistical signiflc- ance in the proportion of juveniles to adults is reached between November 1960 to November 1961, and again between November 1962 and 1963. At Negri soak statistical signiflcance in the proportion of juveniles to adults is reached between November 1960 to November 1961, p almost 0.02 and November 1962 p < 0.01. Tests are not possible on the full Bagdad soak series. Inter-population coinparisons Comparisons are only possible for the three years 1960 to 1962. The differences between November 1960 samples from Lighthouse and Negri are highly significant (p< 0.001). In 1961 the differences between Lighthouse swamp and Bagdad soak are highly significant (p < 0.001, but the differences between Light- house swamp and Negri soak are not significant. In 1962 Lighthouse swamp and Negri soak are not different but each is significantly different from Bagdad soak (p < 0.001). In 1963 no sample was taken from Bagdad soak and the Negri soak population was catas- trophically reduced by salt water flooding. In 1964 the Negri soak population was still very small. The Bagdad soak population was not estimated, but juveniles were common, while the Lighthouse swamp population was very large and dominated by juveniles. Genetic composition of populations Lighthouse swamp. The data are presented in Table 2. A statistical analysis of these results shows: Total population. A change in frequency of morphs between November 1959 and Nov- ember 1960 (p < 0.05 > 0.02) with no statis- tically significant change subsequently. Recruitment. There is no statistically signific- ant difference between recruitment from year to year. Differential survival of frogs. The age cate- gories developed suggest that the March sample is comparable with the adults of the following November. Comparison shows November 1959 adults to differ significantly from the March 1959 sample (p almost 0.001, 10.22) but other samples are not significantly different. Bagdad soak. The data are presented in Table 3. A statistical analysis of these results shows; Total population. There is no statistically significant difference in morph frequencies in the samples over the years. Recruitment. There is no difference in morph frequencies among recruits from year to year. Differential survival of frogs. There is a significant difference in frequency of morphs present between March 1961 and adults November 1961 (chi square 4.268 p < 0.05, > 0.02). Negri soak. The data are presented in Table 4. A statistical analysis of these results shows: Total population. There is no significant differ- ence in morph frequencies present from year to year. TABLE % Lighthouse Soak; composition of population samples hy phenotype and age. Year and ^lonth Composition hy J^henotypc Composition by Age and Plienotype Composition by Age lUdged Lyrate Adult Juvenile Adult Juvenile Ridged Lyrate Ridged l.yrato JIareh 49 50 November 54 25 38 n 10 14 49 30 1900— March 41 24 November 90 81 03 59 27 22 122 49 1901— March 22 24 November 200 178 45 27 101 151 72 312 1902— March :34 30 November 85 00 12 in 73 5() 22 129 190:3— March :30 30 November 127 98 ii 3 124 95 0 219 1904— March i:3 7 November 375 270 11 9 304 201 20 625 125 TABLE 3 Bagdad Soak; coviposition of population samples by phenotype and age. Composition by Phenotype Composition by Age and Plienotype Composition by Age Year and Month Kidffed Lyrate Adult Juvenile Adult Juvenile Ridged Lyrate Ridged Lyrate 1960— March 44 48 Xovcmber 1961— 33 32 33 30 .... 2 63 2 March 61 61 November 1962— 51 32 34 16 17 16 50 33 March ... 45 38 November 196:i— 12 5 7 3 5 2 10 7 March November 2 1 1964— March .... 22 12 November 16 9 2 i 14 8 3 22 TABLE 4 Negri Soak; composition of population samples by phenotype and age. Composition by Phenotype Composition by Age and Phenotype Composition by -Age Year and Month Uidged Lyrate Adult Juvenile Adult Juvenile Ridged Lyrate Ridged Lyrate 1960— 3larch .... November 24 21 n 7 13 14 18 27 1961 — .March 35 16 November 1962— 54 34 14 5 40 30 19 70 March .... 13 5 November 1963 — 48 31 3 3 45 28 6 73 Jlarch .... November L7 3 2 2 1964-- -March .. Novem)>er 2 1 1 1 i 2 •> Recruitment. There is no difference in morph frequencies among the recruits in the years 1960 to 1962. Differential survival of frogs. There is no statistical difference in morphs present between March 1961 and November 1961 adults. Intra-population comparisons Total populations. All populations are statis- tically similar from year to year with respect to frequency of morphs present. Recruitment. All populations are similar with respect to frequency of morphs present among recruits. Differential survival of frogs. The November 1961 samples from all populations show a decline in the lyrate morph compared with the previous March. Only in the case of Bagdad soak does this reach statistical significance. However, when the March and November samples from the three localities are pooled, the large sample does show the change to be significant almost at the p .02 level (chi square 4.80). This suggests that the decline in frequency of lyrate animals is real and not an artefact of sampling. Discussion Earlier (Main, 1961) I attempted to explain the differences in morph frequency between the pre- and post-breeding samples in 1959 as follows; “The advantage enjoyed by the ridged morph may be due to : (a) More rapid development so that, as larvae, ridged animals may take advantage of more ephe- meral waters in dry years. (b) The larvae of the morph may be able to tolerate the warmer temperatures which are found in warm dry seasons.” The earlier information was interpreted in the absence of a distinction between the contribu- tions of recruitment and adult survival to the total morph frequency. Ability to distinguish juveniles of the recent breeding from older animals in a post-breeding sample now allows another interpretation which can be considered after reviewing the results presented above. Classification of the November samples into recruits and adults has exposed the threefold nature of the problem viz.: — Constancy of phenotypes in recruitment: changes in survival among frogs of one pheno- type in two seasons (1959; 1961); domination of total population by juveniles towards the end of the study period. It is of interest to see whether any of the observed changes in the populations can be related in any way to the different serai stages present at each locality and/or the different seasonal weather pattern over the years of the study. Ecological succession Of the three localities described earlier Negi*i soak is clearly earliest in the succession, no rot- ting litter has accumulated and the water is still alkaline. On the other hand Bagdad soak is nearest the climax, especially with respect to surrounding sedges. The data suggest that serai stage only has an effect on population structure insofar as at Bagdad soak the decline in adult frequency is delayed compared with the other two sites. Seasonal weather pattern The seasonal weather for the years 1959 to 1964 has been summarised in Table 5. The seasons have varied from warm dry (1959) to cold dry (1960) mild dry (1961) to mild wet (1963). Over this range of environmental con- ditions there has been no statistically signifi- cant change in the proportion of phenotypes present among the recruits. TABLE 5 Perth weather. May to September, years 1959 to 1964 inclusive. Year lYmperature Average Deviation from Mean (°F.) Rain-days Deviation from .Mean (days) Sunshine Deviation from Mean (hours) Rainfall Deviation from Mean (inches) 1950 + 2-2 — 9 -h 66 — 10 I960 — 1-9 — 4 + 81 — 6 1961 4- 1-1 — 12 + 80 — 4 1962 u i-i + 3 — 4 196:3 + 1-1 + 25 — 155 + 4-5 1964 -h 0-24 + 9 + :3-() -f 3-28 In the warm dry 1959 season there was a significantly poorer survival of the lyrate ani- mals at Lighthouse (the only place then under observation). Again in the mild dry 1961 season the data are suggestive that lyrate sur- vival is poorer in all localities. When data of all localities are pooled the decline is significant almost at the p -- .02 level. It thus appears that lyrate adults are less able to stand warm winter conditions than ridged animals. Conclusions Over the period of the observations the seasonal rainfall pattern has gone from dry to wet and during the same period the proportion of adults in the populations has also declined. These changes are statistically most marked at Negri soak and Lighthouse swamp. The ex- planations of these observations appear to lie in a combination of, habitat preferences of adult frog, serai stage and minor local topographic conditions. Crmia insignifera is a frog which can only persist in swamps or marshes that are dry during the summer. During winter time they enter water to breed, but at other times leave the water and occupy adjacent marshy, but not flooded parts such as the Gahnia zone surrounding Bagdad soak. It so happens that the Bagdad soak is the only one in which flood- ing of the adjacent sedges does not take place in wet years. Here, so to speak, the adults are never flooded out into the inhospitable and unsuitable grassland and it is hex’e that the survival of adults appears best. In the other localities adults are driven into unsuitable habi- tats by flooding in wet years and this is re- flected in the declining proportion of adults in population samples as the years become wetter. Acknowledgments The work was done with the facilities offered by the Rottnest Biological Research Station while the author was in receipt of a Research Grant from the University of Western Australia. References Hodgkin, E. P. and Sheard, K. (1959). — Rottnest Island: the Biological Station and recent scientific research. J. Roy. Soc. W. Aust. 42: 65-95. Main, A. R. (1961 ). — Crinia insignifera Moore (Aiiura Leptodactylidae) on Rottnest Island. J. Roy. Soc. W. Aust. 44: 10-13. 127 13. — Notes on the composition and structure of the Duketon Meteorite by M. J. Frost* Manuscript received and accepted 16th March, 1965 Abstract The Duketon Meteorite, whose discovery in 1948 was previously reported by the author, is a medium octahedrite in the sense of Lovering and others (1957). Schreibersite. and possibly cohenite, are present, but neither troilite nor lawrencite was observed. Analyses of two samples gave: — Fe (by difference) 91.9, 92.1; Ni 7.36, 7.14; Co 0.51, 0.48; P 0.244, 0.239; S 0.031, 0.029. A preliminary note on the Duketon Meteorite with a history of the find, a description of the meteorite and photographs has already been published (Frost, 1958). The Duketon Meteorite was found in 1948 resting on the surface about 10 miles north of Duketon, Western Australia. This is approx- imately long. 122° 22'E, lat. 27° 30'S giving the seven figure coordinate number (CN) — 1224, — 275. Its original weight was probably about 119 kg. (263 lb.). The surface, contrary to previous opinion, is not weathered and the very dark brown (Munsell 10YR2/2) striated crust is perfectly preserved. Grinding and etching of a section indicates that the structure is, using the definition of Lovering and others (1957), that of a medium cctahedrite. The average width of the kamacite plates, calculated from measurements after determining the orientation of the etched surface with the aid of table I of Belaiew (1923), is 1.0 mm. On the surface studied no troilite was observed but schreibersite, and possibly * Department of Geology. University of Canterbury, Christchurch, N.Z. cohenite, occur as grains and plates up to 0.02 X 0.3 mm. in cross-section, mainly within the plessite. The plessite is mostly of the fine- grained perthitic variety. There is no evidence of the presence of lawrencite. TABLE I Analyses (columns a and bJ of two samples of Duketon Meteorite a b c Fe 91.9 92.1 0.1 Ni 7.36 7.14 0.04 Co 0.51 0.48 0.01 p 0.244 0.239 0.005 S 0.031 0.029 0.005 Notes: Analytical results in weight per cent. Iron by difference. Column c: estimated standard error of results for a and b. Analyst: P. R. Hentschel. Analyses of two fragments from the interior of the meteorite (Table I) indicate no unusual features. These analyses were made possible by a Research Grant from the University of Canterbury. References Belaiew, N. T. (1923). — The inner structure of the crystal grain as revealed by meteorites and Widmanstatten figures. J. Inst. Met. 29: 379-406. Frost, M. J. (1958). — A preliminary note on the Duketon Meteorite. J. Roy Soc. W. Aust. 41: 55-56. Lovering, J. F., Nichiporuk, W., Chodus, A. and Brown H. (1957). — The distribution of gallium, germanium, cobalt, chromium and copper in iron and stony-iron meteorites in relation to nickel content and structure. Geochim. et Cosmoch. Acta 11: 263-278. 128 INSTRUCTIONS TO AUTHORS Contributions to this Journal should be sent to The Honorary Secretary, Royal Society of Western Australia, Western Australian Museum, Perth. Papers are received only from, or by communication through, Members of the Society. The Council decides whether any contribution will be accepted for publication. All papers accepted must be read either in full or in abstract or be tabled at an ordinary meeting before publication. Papers should be accompanied by a table of contents, on a separate sheet, showing clearly the status of all headings; this will not necessarily be published. Authors should maintain a proper balance between length and substance, and papers longer than 10,000 words would need to be of exceptional importance to be considered for publication. The Abstract (which will probably be read more than any other part of the paper) should not be an expanded title, but should include the main substance of the paper in a con- densed form. Typescripts should be double-spaced on opaque white foolscap paper; the original and one carbon copy should be sent. All tables, and captions for figures, should be on separate sheets. To avoid unnecessary handling of the original illustrations authors are requested to include additional prints, preferably reduced to the final size required; a choice of either one-column (about 2.8 inches) or two-column (about 5.8 inches) width is available. The preferred positions of figures should be marked on the second typescript copy. In the preparation of references, and for all matters of presentation not otherwise covered in these instructions, authors are required to follow the C.S.I.R.O. Guide to Authors (Melbourne, 1963). Failure to read through this carefully before preparing papers may lead to delay in publication. The use of the various conventional systems of nomenclature recommended in this booklet, and in the supplementary pamphlets referred to in it, is obligatory; for this purpose, palaeontological papers must follow the appropriate recommendations for zoology or botany. All new stratigraphic names must have been previously approved by the Stratigraphic Nomenclature Committee of the Geological Society of Australia. Thirty reprints are supplied free of charge. Further reprints may be ordered at cost, provided that orders are submitted when the galley proofs are returned. Authors are solely responsible for the accuracy of all information in their papers, and for any opinion they express. Journal of the Royal Society of Western Australia Volume 48 1965 Part 4 Contents 10. — The Porcellanidae (Crustacea, Anomura) of Western Australia, with descriptions of four new Australian species. By Janet Haig. 11. — A new species of Nectria (Asteroidea, Gonioasteridae) from Western Australia. By S. A. Sheperd and E. P. Hodgkin. 12. — Further studies of the polymorphic species Crinia insignifera Moore (Anura, Leptodactylidae) on Rottnest Island. By A. R. Main. 13. — Notes on the composition and structure of the Duketon Meteorite. By M. J. Frost. Editor: A. F. Trendall Assistant Editor: A. S. George The Royal Society of Western Australia, Western Australian Museum, Perth 99037'6'C5— 570 ALEX. B. DAVIES, Govornmcnt Printer, Western Australia