LATE CAMBRIAN (POST-IDAMEAN) TRILOBITES FROM THE HIGGINS CREEK AREA, WESTERN TASMANIA PETER A. JELL, NIGEL C. HUGHES AND ANTHONY V. BROWN Jell, P.A., Hughes, N.C. and Brown, A.V. 1991 08 01: Late Cambrian (post-ldamean) trilobites from the Higgins Creek area, western Tasmania. Memoirs ofthe Queensland Museum3Q(3): 455-485. Brisbane. ISSN 0079-8835. Trilobitesaredescribedfrom ninelocalitiesonloggingtracksbetweentheHuskissonRiver and Bums Peak, NNEofRenison Bell,westernTasmania. The 18 taxa, 7 in open nomen- clature. indicate anage in the pre-Payntonian Stage (Queensland .scheme)or late Early to early Late Sunwaplan Stage (North American .scheme). Seven new species are described asLotagnostustullahensis,Rhaptagnosfusmji.Cermatopsthalasta,Asiocephaluslatosug- grundus, Olenusapoxysomafus.Chekiungaspisconcavus, and Wujiajiania distorta. This fauna is important as it represents an off-shelf assemblage, possibly with some shelf inhabitantsmixed in,thatiscontemporaneouswithoneofthecarbonateshelfassemblages ofwestern Queensland. LateCambrian, trilobites, Tasmania. PeterA. Jell andNigel C. Hughes, Queensland Museum, PO Box 300, South Brisbane, Queensland 4101, Australia; Anthony V. Brown, Deportment ofResources and Energy, POBox56, RosnyPark, Tasmania 7018, Australia; 10May, 1991. TasmanianLateCambriantrilobiteshavebeen His Trilobita incertae sedis, specimen 3 (Jago, recorded (Jago,1972, 1974, 1978, 1987; andref- 1978. pi.2. fig. 18) is assignable to Cermatops erences in Jago, 1979 and Banks, 1982) from thalasta sp. nov. as it has the same structure of several different parts of the State and from furrowson the pleural areasasdescribed forthat several different horizons. Thispaper presents a species below. Moreover, we concur with Jago furtherdiscovery,in theHuskisson Group,in the that his cranidial fragment (Jago, 1978, pi.2, HigginsCreekareabetweentheHuskisson River fig.17) is conspecific with the pygidium but the and^Burns Peak, 15 km NNEofRenison Bell, of possible assignment to Briscoia is revised. We a fauna of 18 trilobite taxa plus brachiopod and suggestthat both specimens belongto C. thalas- bradoriid elements; much of the fauna has not ta. Lotagnosius occurs in both faunas, with un- been previously described from Australia and availability of features (through poor pres- has affinitieswith faunas from other partsofthe ervation) on Jago's (1978, pi.2, fig.l) specimen world, in particular central China. The nearest preventing specific comparison. Olenus occurs occurrences oftrilobitesare those from the Hus- in both faunas, albeit as entirely different kisson River about 7 km NNE of Renison Bell species; .some importance is attached to its (Jago, 1974)where the index fossil Glypiagnos- genericrange in Europe,but in ChinaO, sinensis tus reticulatus indicates an early Idamcan age. occurs with taxonomic relatives of some ofour ALateCambrianfaunawasdescribedfromthe Higgins Creek fauna in horizons younger than Climie Formation, Dundas Group (Jago, 1978) those yielding the genus in Europe (Lu and Lin, about 12 km south ofRenison Bell. Jago (1978) 1989). Based on this series of observations we favoured apost-ldamean age overtheotherpos- deduce that Jago’s Climie Formation fauna and sibility,latestIdamean.Jago'sPe/rw/'a(?)sp. has our Higgins Creek fauna w'ere approximately a glabella with convex lateral margins and contemporaneous.Jago(inShergoldetal., 1985) glabellar furrows that are almost straight in the reassessedtheageofhisClimie Formationfauna two arms of the chevron and continuous across on the basis ofthe discovery ofHedinaspis in a theaxis;botharefeaturesofWujiajianiadistorta correlative horizon in the Professor Range; on sp. nov. and we suggest that may be the more thatbasisheconsidereditlate LateCambrianbut likely identity of Jago’s specimen. His Cerato- tabulated it (Shergold el al., 1985, chart 7, col. pygidae, gen. ct sp. indet (Jago, 1978, pi.2, figs 53) as medial Late Cambrian with which we 16, 20) are indistinguishable from Procerato- agree. Wc are unaware of this record of pyge gordonensis Jago, 1987, a close rcla- Hedinaspis being illustrated and suggest that it tive/descendant ofwhich is found in our fauna. should be carefullv scrutinized to determine 456 MEMOIRS OFTHE QUEENSLAND MUSEUM FIG. 1. Locality map forfossil localities mentioned in text. Area is situated between the Huskisson River and Burns Peak NNEofRenison Bell. whether it may be the similar Asiocephalus or post-Idamean age with possible close proximity We not. Similarly the record of Hedinaspis from ofthe ages ofboth faunas. accept Jago’s (in western New South Wales (Webby et al., 1988) Shergold et al., 1985) assessment of its age. could represent Asiocephalus-, distinctive fea- Geological setting and age of the Higgins tures are unavailable on specimens illustrated. Creekfaunaindicatethatitisburiedinaturbidite The fauna from the Singing Creek Formation sequence probablydeposited on orat the base of of the Denison Range, southwestern Tasmania the continental slope. Some ofthe taxa found in (Jago,1987) has in common with faunas descri- only one of the three faunas (Higgins Creek, bed below MicragnostuSy Pseudagnostus, Aph- Climie Formation, Denison Range) may have elaspis, Proceratopyge, but this does not in- been shelf benthos, having been incorporated dicate contemporaneity in itself. However, note intotheoffshelfdepositsby slumporflow atthe that Jago (1987) compared his Leiostegiacean shelfmargin. Those taxa common to the faunas gen. et sp. indet. to families that are normally may have been off shelf benthos or inhabitants We post-Idamean in China and described a species ofthewatercolumn. suggestthatthesethree of Pseudoyuepingia, that genus being post- faunas were essentially contemporaneous, post- Idamean in China and interpreted as such in Idamean in age; discussion of this age is western New South Wales(Webby etal., 1988). presented below. Moreover, Jago (in Shergold et al., 1985) at- tributedapost-Idamean topre-Payntonianageto LOCALITYAND GEOLOGICALSETTING thisfauna. Comparisonofelementsofthisfauna with the one described below also suggest a During mapping for the Corinna 1:50,000 UPPER CAMBRIANTRILOBITES FROM TASMANIA 457 Geological AtlasMapSheet(Turneretal. 1991), The Higgins Creek sequence containing the a small area containing a sequence of inter- fossils belongs to a belt of rocks formed during bedded pebble to granule conglomerate, with the latest Idamean, which is now thrust against dominantly siliceous clasts and matrix, graded Precambrian successions to the west and has sandstone and lithicwacke and siltstone was en- Middle(?) Cambrian volcano-sedimentary suc- countered along the poorly accessible cessions thrust against it on the east. Due to southeastern margin of the map sheet in the discontinuous structures to the north and south Higgins Creek area (Fig.l). Tn March 1990, of the Higgins Creek area and the unresolved during a reconnaissance traverse into this area, structural complexity ofthe rocks to the east, it nine localitieswith fossiliferouslithicwacke and is not possibletogive a structural relationshipof siltstonewerefound incuttingsalonganewroad this east facing sequence w'ith the Middle(?) west from the Murchison Highway towards the Cambrian volcano-sedmentary sequences. Huskisson River; the area is about 15 km NNE Theimplicationofstructuralcomplexityofthe NW of Renison Bell and 12 km ofTuliah, both HigginsCreekarea, the discontinuousstructures towns beingon the Murchison Highway. both to the north and south ofthe Higgins Creek Lithologically, the sequence is dominated by area, and the lack ofknowledge as to the bound- interbeddcd siltstone and well bedded lithic- ary relationship of this area with the sequences wacke,with lensesandchannelinfillingsofcon- containing, the as yet undescribed, late Mid- glomerate. The conglomerate units are of both dle(?)Cambrian faunaswithin thevolcaniclastic clast and matrix supported varieties. The open sequencestothecastofBurns Peak(Corbettand frameworkconglomerate unitshave dominantly McNeil, 1986; Corbett and Solomon, 1988: chert clasts whereas the closed framework units 101), and the tectonic significance of the new are dominated by rounded quartzite clasts. The fauna, will be discussed wdthin the forthcoming thicker sand grade beds and conglomerate units Explanatory Notesto theCorinna 1:50,000Map have scour bases and, in places, contain rip-up Sheet (in prep.). siltstone clasts in the basal section. Sand grade beds are usuallygraded andconsistentlygive an FAUNA AND AGE east facing(easterly dip). Some sand grade beds contain intraformational soft sedimentdeforma- We describe 18 trilobite taxa, with 7 ofthose tion and small cross-bedded channel infills. in open nomenclature (Table 1). Proceratopyge Siltstone beds are usually laminar but cross- gordonensishasbeen found in Australiain strata bedded units are also present. Sedimentological we consider approximately coeval with the col- features within this succession suggest a sub- lectionsdescribed below butconsideredbyJago marine fan environment ofdeposition. (1987) to be Idamean. Pseudagnostus sp. is The rock sequence is considered to belong to compared with P, idalis Opik, 1967, although the upper part of the Huskisson Group and a that is a dubious comparison; P. idalis is correlate ofthe upper part ofthe Dundas Group restricted to the Idamean in the restofAustralia. (as described by Brown, 1986). Rock sequences Apart from its record in the Climie Formation lithologicallysimilartothosecontainingthefos- (Jago. 1978) mentionedabove, hasbeen sil sites occur to the north and south ofthe area. recorded in the Idamean ofwestern Queensland To the north, the succession contains mudstone, (Opik, 1963), and the Olenus Zones of Europe quartzite, greywacke, and tuffaceous grey- are correlated with the Australian Idamean. wacke-mudstonewithconglomerateandcrystal- Aphelaspis is now' widely known in Australia, vitric tuffunits (Collins et al., 1981). butitsgreatestdevelopmentisin NorthAmerica; To the south, exposureswere available in two i\\QAphelaspis Zone of that continent is corre- locationsduringtheearly to mid 1980s.Thefirst lated with the early half of the Australian of these were in roadside quarries made during IdameanStage. However, therange ofthistaxon the construction of the Lower Pieman Dam is not certain and depends to large extent on Road; however,theseexposureshavesince been betterdefinitionofthisand relatedtaxacurrently re-vegetated. The second location was in the assignedtoseparate butundifferentiatedgenera, Pieman RiverGorgetothe southofthe road, but asdiscussedin thesysycmaticsbelow. Olentella floodingtherivertoform Lake Pieman drowned forexamplerangesintothepost-Idamean.These these localities. Decription of the sequences in taxacouldbe usedtosuggestcorrelationwiththe these areas can be found in Green (1983) and latest Idamean (Shcrgold et al., 1990). Brown (1986). However,therestofthe faunaisnotobviously 458 MEMOIRS OFTHE QUEENSLAND MUSEUM TRILOBITES \ LOCALITIES 1 2 3 4 5 6 7 8 9 Acmarhachisl so. X Micra^ostussx). cf. M. intermedinsPalmer, 1968 X Pseudagnostids X X Lotamostustullahensissp. nov. X Pseudamostussp. X X Pseuda^nostus(Sulcata^nostus) sp. X Neoamostus clavusShereold, 1972 X Rhapta^nostusconver^ens Palmer, 1955 X Rhapiasnostusmiisp. nov. Proceratopysesp. cf. P. sordonensis Jaeo, 1987 X X X X X Cermatops thalasta sp. nov. X X X Asiocephaluslatosussmndus sp. nov. X X Olenusapoxx'somatus sp, nov. X X X X X Chekiangaspisconcavus sp. nov. X Wuiiaiiania distortussp. nov. X X X X Aphelaspissp. X X Conokcphalinidae indet. X Aposolenopleura sp. X TABLE 1. Distribution offauna at the nine collecting localities marked on the locality map (Fig. 1). in accord with this date. Aposolenopleura and that horizon should be correlated with pre- the nearest specific comparison ofLotagnostus Saukiella horizons in North America, i.e. with tullahemisare to be found in the Hiingaiamag- the late Early Sunwaptan (=late Franconian). nifica Faunule which was originally thought to Cermatops occurs in western Queensland in the be oflateTrempealeau age(Rasetti, 1944); sub- post-ldamean Wentsuia iota-Rhaptagnostus sequently, Lochman and Wilson (1958) corre- apsis Zone (Shergold,!980), and the Welsh lated it with the late Franconian (Sunwaptan) species of Hughes and Rushton (1990) is from Ptychaspis-Prosaukia and Saukia Zones. Pal- the contemporaneous Parabolina spinulosa mer (1968) accepted correlation with the late Zone, twozonesyoungerthan the Idamean. The Franconian (late Early Sunwaptan)and erected Chinese Wujiajiania and Chekiangaspis occur Asiocephalus for trilobites from the same together in the Lotagnostus punctatus Zone of faunule.ThisgenusisrepresentedinourHiggins western Zhejiang; this zone is numbered 14 in Creek fauna. In Kazakhstan, Hedinaspis (Asio^ the scheme ofLu and Lin (1989) withPseudag- cephalus) sulcata is the nominal species of a nostusidalisoccurringin x\\c,Proceratopyg€ fen- zone correlated with the Trisulcatagnostus ghuangensisandErixanumZones(Zones 11 and trisulcus and Eolotagnostus scrobicularis 12 of Lu and Lin’s scheme) those zones being Zones; Shergold et al. (1990) correlated these correlatedwith thethird andsecondlastzonesof zones with the Saukia Zone {Saukiclla junta the Australian Idamean, respectively (Lu and Subzone)of North America even though Lin, 1989, table 8). Therefore, if Lu and Lin’s Saukiella does not appear in the Kazakh se- zone 13 (Pseudoglyptagnostus clavatus-Sino- quenceuntiltwozoneslaterin ihzHarpidioides- proceratopygekiangshanensisZone) is equated PlatypeltoidesZone (Apollonov and Chugaeva, with the last Idamean or an immediately post- 1983);Shergold(pers.comm.June, 1991)would ldameanZone(J.H.Shergold(pers.comm.June, now correlate those Kazakh zones with older 1991)suggestedequivalence with i\\tIrvingella horizons in the rest of the world thus coming Zone) then ihcLotagnostuspunctatusZone (14) closer to the suggestion made here . We suggest is post-ldamean. In so far as the base of the L. UPPER CAMBRIAN TRILOBITES FROM TASMANIA 459 pimctatus Zone of western Zhejiang is aligned cussed above approximate contemporaneity with the base ofthe Maladioidella Zone of the with Jago's (1978. 1987) Climie and Singing North China Provinceand thatZone isshown by Creekfaunasissuggestedastheonlyspeculation Shergold et al. (1990, fig.7)to be equivalent to at a more specific level. The lower fauna in the the post-/m/ige//flZoneofQueensland, we cor- Watties Bore section (Webby et al., 1988) relate the L. punctatus Zone, with the late Early resembles the Tasmanian faunas in containing Sunwaptan (=late Franconian). Therefore, 6 Pseudoyuepingia, Proceratopyge, Pareuloma members of the Higgins Creek fauna suggest (close to Chekiangaspisas discussed by Webby correlation with this North American horizon. etal. (1988, p.914)), ?iX\dHedinaspissp.(maybe Neoagnostus clavus is found in western Asiocephalus as discussed above); however, Queensland in the penultimate and prepenul- these similarities may be due to their occurring timate zonesofthepre-Payntonian. Shergoldet in the same offshelfenvironment where a num- al. (1990) also correlated this horizon with the ber of genera are beginning to appear to have Saukiella junia Subzone of North America but quite long rangesas well as to proximity ofage, that correlation was based on conodont work ifany.The significance ofthefaunais itsfurther which has been revised providing a different revealing the oceanic fauna ofeastern Australia picture (NicollandShergold, 1991). llscontem- in the Late Cambrian and providing more infor- poraneous horizons in the North American se- mation for correlation within Australia and the quence may be near the Early /Late Sunwaptan Asian region. boundary (Ludvigsen and Westrop, 1985). Similarly, Rhaptagnostus convergens is asc- SYSTEMATIC PALAEONTOLOGY ribed an age (Shergold, 1977) in the Saukiella pyrene Subzone ofNevada; thathorizon, imme- The fossils described herein are housed in the diately on top of the Ellipsocephaloides Zone Geological Survey of Tasmania (prefix GST). (Longacre, 1970, text-fig.5) indicates the Early Descriptive terminology follows Moore (1959), to Late Sunwaptanboundary. Thesetwo species where possible, with the notation for glabellar may therefore be added to those that suggest a lobes and furrows following Henningsmoen Sunwaptan age forourfauna even ifthesugges- (1957) (i.e. lobestermed LO, LI, L2, L3 etc. and tion is for a zone younger than suggested by the furrows SO, SI, S2 etc from the occipital for- previous six species. ward). The Higgins Creek fauna, therefore, contains 8 species that suggest correlation with horizons ClassTRILOBITA just below or above the Early/Late Sunwaptan Order MIOMERA Jaekel, 1909 (=Iate Franconian)boundary (3 or4zones post- Superfamily AGNOSTOIDEA McCoy, 1849 Idamean) and 4 that suggest latest Idamean but Family AGNOSTIDAE McCoy, 1849 much less strongly. In discussions of Wujia- jianiadistortadj\^Asiocephaluslatosuggrundus Lotagnostus Whitehouse, 1936 we point out that they are probably ancestral to members of the respective genera found else- TypeSpecies where in the world. Therefore, they may be Agnosias trisectus Salter, 1864 from the Late interpreted as indicatingaslightly older horizon Cambrian ofBritain and Sweden. inTasmania.We initiallyconsideredthatamong the 9 localities, which could not be placed in Lotagnostus tullahensis sp. nov. stratigraphicsequence,somemay havebeen late (Fig. 3G-I) Idameanandothersseveral zonesposi-ldamean. However,careful examination ofthecollections Etymology show the reported associations to be valid with Near the town ofTullah, western Tasmania. some of the unexpected co-occurrences on one piece of rock. Material Wc suggest that the age of the fauna may be Holotype GST14375 and paratypes post-ldameanwithintheEarlySunwaptanStage, GST14373 and 14374 from Loc. 9. with which its constituents have been compared above; nozonalcorrelation maybesubstantiated Diagnosis with any other part of the world, not even with Cephalon with well-impressed preglabellar Queensland or other parts of Australia. As dis- median furrow reaching border furrow, 460 MEMOIRS OFTHE QUEENSLAND MUSEUM scrobiculate cheek areas; anterior glabellar lobe Dresbachian of Alabama, by original designa- large, subquadrate with slight anteromedial tion. projection and rounded anterolaterally; S2well- impressed, chevron-shaped over axis; glabellar Acmarhachis ? sp. nodejustbehindS2;basallobeslarge,triangular, (Fig. 2A,B) extending well forward. Pygidium with wide, straight-sidedaxisfinishingwell forw'ardofbor- Material der furrow; first axial ring of two lateral lobes; GST14355 from Loc. 1 second axial ring with high-rounded median node descending forward to articulating furrow Discussion and dividing first ring; cheek areas weakly This specimen is too poorly preserved for scrobiculate; border flat but not wide, with pair taxonomictreatmentbut it is assigned ioAcmar- ofmarginal spines posterior ofwidest point. hachisonthebasisofglabellarshapewithbroad- ly rounded anterior and long anterior lobe, on Remarks lackofpreglabcllarmedian furrowandon shape Generic assignment is made using the diag- of the pygidial axis which reaches to the nosisofLudvigsenetal.(1989)wherein theonly posterior border furrow. This may be slim point of possible disagreement could be the in- evidence but Shcrgold (1980) showed that the dication that the pygidial axis extends close to genus ranges higher into the Late Cambrian in the posterior border furrow. In this Tasmanian Australia and the featuresoutlined are sufficient species the axis does not appearto extend as far for tentative assignment ofthis poor specimen. posteriorly (Fig. 31) as in cogeners. Variation in this feature is evident in otherspecies. Micragnostus Howell, 1935 The Tasmanian cephalon is virtually indistin- guishablefromthatofL. americanus(cf.Rasetti, Type Species 1944, pi. 36, fig. 1; Ludvigsen et al. 1989, pi. 1, Agnostus clavus Lake, 1906 from the Trem- fig.25-upper right). Anterior glabellar shape adocian ofWales, by original designation. and strongly dividedposteriorglabellarlobe are distinctive. However,thepygidiumofthatNorth Micragnostus sp. cf. M. intermedins American species has awide border constricted (Palmer, 1968) acrolobcs, axis reaching close to border furrow (Fig. 2G-I, K(right only) and differently shaped first and second pygidia! axial rings and tubercle thereon. In this last fea- Material ture, the furrow between the first and second Three cranidia GST14359-14361 and one pygidial axial rings is usually transverse and pygidium, GST14366 all from Loc. 1. continuousinpreviouslydescribedspeciesofthe genus but in L. americanusas in ourTasmanian Discussion species (Ludvigsen et al. 1989, pl.l, fig.l5; These cephala are tiny and the pygidium is Rasetti, 1944,pl.36,fig.2)thetubercleinterrupts only a little larger. It is, therefore, difficult to that furrow', extending over the length of 2 seg- comparethemwith materialofthe long-ranging ments. Nevertheless,shapeofthe lateral partsof M. intermedius (assigned to Micragnostus by thefirstringandshapeofthetubercledistinguish Fortey (1980:21)) from Alaska (Palmer, 1968), these2species.Lotagnostussp.ofPalmer(1968, Mexico (Robison and Pantoja-Alor, 1968) and pl.l2, figs 3,4) has a pygidium almost identical Newfoundland (Ludvigsen et a!., 1989). How- with the Tasmanian (axial ring is longer, axis ever, the pygidium is distinctive with its axis narrower, border narrower) but the associated finishingwell forwardofthe borderfurrow, and cephalon is distinct (nonscrobiculate, wide bor- markedly narrow border for the genus or for der furrow, poorly divided posterior glabellar agnostoids ofthis age. The latter feature in par- lobe, posteriorly placed glabellar node), ticular,isdistinctiveoiM. intermediusamongits contemporaries,butissharedwith Tremadocian Acmarhachis Resser, 1938 forms described by Fortey (1980, pl.l). The cranidiaarelessdistinctivebut comparisonwith Type Species the smaller eranidia from Mexico (Robison and Acmarhachistypicalis Resser, 1938, from the Pantoja-Alor, 1968, pi.97, figs 4,5) shows the anteriorly tapering glabella with well rounded , UPPER CAMBRIANTRILOBITES FROM TASMANIA 461 Fig.2. Allfrom Loc. 1 exceptLfromLoc,5. AyB^Acmarhachis?sp.internalmouldandlatexcast, respectively, ofcompletespecimenGST14355,x8. C,EffacedpseudagnostidGSTI4356,x7. D-F, K{\cfl),Rhaptagnostus mjisp. nov. D, internalmouldofsmallpygidium GST14357, xl4. E,F, latex castand internalmouldofsmall cephalon GST14358, xl4. K (left), internal mould of small cephalon GST14365, xl4. G-I, K(right), Micragnostussp.cf.M.intermedius (Palmer, 1968).G-I, internalmouldsofsmallcephalaGSTl4359-14361 xl5, xl5, and xll, respectively. K(right), internal mould of pygidium GST14366, xl4. J, Pseudagnostid indet.internalmouldofcompletedeformedspecimen GST14362, x6. L, Pseudagnoslidindet. GST14363, x5. anterior, narrow border, and short but distinct Type Species basal lobes to be shared by both setsofmaterial. Neoagnostus aspidoides Kobayashi, 1955 Identification with this American species is from British Columbia, by original designation. necessarily tentative because cranidia of com- parable size to the larger Amer-ican specimens Neoagnostusclavus (Shergold, 1972) are not available from Tasmania as yet and be- (Fig. 3M) cause the Tasmanian specimens are all internal moulds and therefore may prove to have dif- Material ferent external morphology. GST14379, cephalon with first thoracic seg- Family DIPLAGNOSTIDAEWhitehouse, ment articulated from Loc. 8. 1936 Subfamily PSEUDAGNOSTINAE Discussion Whitehouse, 1936 This cephalon resembles the western Queenslandparatypc figuredbyShergold(1977, Neoagnostus Kobayashi, 1955 pi.16, fig.14), particularly in outline, glabellar lobesandnodearrangement,andwaistedglabel- 462 MEMOIRS OFTHE QUEENSLAND MUSEUM UPPER CAMBRIANTRILOBITES FROM TASMANIA 463 la. Only the preglabellar median furrow on the nosius idalis Opik, 1967 but we are reluctant to Tasmanian specimen could be used to distin- suggest such an assignment on the available guish it from the Queensland species but that is material. Further discussion ofthese specimens not sufficient in view of the other close seems fruitless because the features of the similaritiesexhibited.ThisTasmanianspecimen glabella are uncertain. has been tectonically compressed in the transverse direction so the glabella, border fur- Pseudagnostus (Sulcatagnostus) sp. row (laterally) and basal lobes appear narrower (Fig. 3J,K) than in the undistorted Queensland specimens butthe specificidentity is not in doubt,although Material a pygidium will be necessary to make certain GST14376, 14377 from Loc. 7. identification. Discussion PseudagnostusJaekel, 1909 The internal mould of a whole specimen is laterallycompressedanddamagedin someparts. Type Species Likewise thecranidium (Fig.3K) isstrongly dis- Agnosias cyclopyge Tullberg, 1880 from the torted. Therefore, features of the subgenus are Late Cambrian ofSweden. the scrobiculate cranidium. long straight-sided glabella with poorly impressed transverse fur- Pseudagnostus sp. row,well impressedpreglabellarmedianfurrow, (Figs4,5) small triangular basal lobes, truncated glabellar rear, elongate pygidiai medial node, accessory Material furrows not meeting posteriorly, wide border GST14381-14388 from Loc. 2, GST14390- furrow, and distinct marginal spines. 14401 from Loc. 3. Featuresnotfully inaccordwithPs.{Sulcatag- nostus) are the apparent lack (it is notclearifthe Discussion borderisentireatthispointandthe midposterior Agnostids are common in collections from spinemightbeanexternal featureanyway)ofthe Localities 2 and 3 but they are invariably dis- thirdpygidiai (midposterior)spineandposterior torted to some degree and cannot be identified positionofmarginalspines.Ps. {Sulcatagnostus) with confidence because the parietal structures rugosus Ergaliev, 1980 also lacks the third of the glabella, which are so important to pygidiai marginal spine. As Shergold (1977) taxonomy in thisgroup cannotbeobserved. It is pointed out this form is closely allied to the Ps. not even certain how many species are repre- cyclopyge Group of Pseudagnostus and sented (e.g. are Fig. 5J,I,Lwith blunterglabellar Ergaliev‘s Kazakh species and this Tasmanian anterior and effaced accessory furrows con- form could well belong to Ps. idalis if the mid specific with the restofthe material?). They arc posterior spine is considered subgenerically apparently spectaculate agnostids, with a well- critical. However, the scrobiculate cranidium impressed preglabellar median furrow, wide seems a distinctive feature of these 3 forms deliquiate border furrows, subquadrate giving a basis for distinction. pygidium,accessoryfurrowsinmostspecimens, accessory furrows not reaching border furrow, RhaptagnostusWhitehouse, 1936 marginal spinessituatedposteriorlylevelwithor behind the rear of the acrolobe. These features Type Species are consistentwith the morphology ofPseudag- Agnostus cyclopygeformis Sun, 1924 from the FIG. 3. A-F, Rhaptagnostus mji sp. nov. from Loc. 9. A,B, internal moulds of complete specimens (A is holotype) GSTI4367, 14368, x8 and x5, respectively. C. latex cast of complete specimen GST14369. x9. D,E, internal mouldsofpygidiaGST14370, 14371, x9. F, latex cast ofsmall complete specimen GST14372, xI2. G-I, Lotagnostus tullahensis sp. nov. from Loc. 9. G,H. laiex casts of small complete specimens GST14373, 14374, xl2. T, latex cast of larger articulated holotype GSTI4375. xlO. J,K, Pseudagnostus (Sulcatagnostus) sp. from Loc. 7. J, internal mould of cranidium GST14376. x9. K. internal mould of articulated specimen GST14377, x9. L, Pseudagnostid indel. interna 1 mould ofcranidium and first thoracic segment from Loc. 7, GST14378, x5. M,Neoagnostusciavus Shergold internal mould ofcephalon and first thoracicsegmentfrom Loc.8,GST14379,x8. N,0,RhaptagnostusconvergensPalmer, latexcastand internal mould ofarticulatedspecimen from Loc. 8, GST14380, x7. 464 MEMOIRS OFTHE QUEENSLAND MUSEUM Fig. 4. Pseudagnostus sp. from hoc. 2. A. latex cast ofarliculaled specimen GST14381, x8. B, G, internal mouldandlatexcastofarticulatedspecimenGST14382,x10. C,D,latexcastand internalmouldofarticulated specimen GST14383, xlO. E,F, internal mould and latex cast ofarticulated specimen GST14384, xlO. H, internal mould ofpygidium GST14385, x5. L J, internal moulds of articulated specimens GST14386 and 14387, x6 and x8, respectively. K, L, latex castand internal mould ofarticulated specimen GST14388, x6.