Some Formations Close to the Permo-Triassic Boundary in Tasmania

PAPERS ANn PROCEEDINGS OF' 'l'HI

SOME FORMATIONS CLOSE TO THE PERMO-TRIASSIC BOUNDARY IN TASMANIA

MAXWEL.L. R. BANKS, University oj Tasmania

and

IKRAM H. NAQVI, Department oj Mines, Tasmania, formerly Universi.ty oj Tasmania.

(With three text. figu.l'cS and one plate.)

ABSTRACT boundary elsewhere. Lewis (1940) recognised that The Coal Measures is a unit of carbol1- the occurred coal and feldspathic sandstone in at the not top Cygnet Measure Upper Permian age. It is overlain but was not consistent (1946, pp. 22-3). gradationally or disconi'ormably by the Springs The difficulties are increased by lack of definition Sandstone. the lower member of which, Barnetts and reasonably precise knowledge of the so-called Sandstone, is a thinly-bedded, fine-grained feJd­ basal Triassic units. The Ross Sandstone of spathic to arkosic sandstone, and the upper, the Ronwick (870) and later authors is not properly Mountain Lodge Sandstone. a more thickly-bedded, defined in its type area and may be undefinable medium-grained protoquartzite. The upper part there. The outcrop is certainly not good in the of the Springs Sandstone is probably Otoceratan. Ross area. The Rhyndaston Sandstone (David The Permo-Triassic boundary lies within these 1932) is much better exposed at Rhyndaston than gradational non-marine units and cannot yet be the Ross Sandstone at Ross, but its relationships fixed accurately. The base of the Mountain Lodge to the Permian on the one hand and to younger Member is probably the most convenient boundary units on the other are not yet clear. The Springs to use as the Permo-Triassic boundary in local field Sandstone (Lewis 1940) is moderately well-exposed mapping. in its basal portions and its relationship to under­ lying rocks in the type area determinable but not INTRODUCTION that to younger sedimentary rocks. No accurate section of this formation has been published. Milligan 0849 pp. 17-18) llrst described the coal Recently the basal 'rriassic unit in the Lake St. measures at Southport subsequently correlated with Clair area has been called the Gould Formation the Cygnet Coal Measures. Thureau in 1881 men­ (McLeod et al, 1961, p. 26), the relationships of tioned the coal at Cygnet and his description of which to underlying and overlying units are clear, the section at the coal mines may still be the best. if gradational, but of which no detailed sectIon has Subsequently Johnston (887) descrIbed coal been published. There are, in addition, innate measures of similar age at Adventure Bay on Bruny diffi.culties in defining and correlating units such Island and in 1888 described the section at Cygnet as the Springs, Ross, Rhyndaston and Gould Forma­ and correlated both successions with the New­ tions, because of their depositional environment. castle Coal Measures of New South Wales. Subse­ probably fiood plains with braided channels. quently freshwater beds at this level have been recognised widely in Tasmania and sections In t.his paper, definitions and deta.iled sections measured in several places (eg. Mount La Perouse, of formations near the Permo-Triassic boundary in Hobart, Zeehan, Lake St. Clair, Mole Creek). the Cygnet and Hobart Districts will be offered During 1965, a good succession near the nominal and the problem of the Permo-Triassic boundary type area was established in detail by I. Naqvi and discussed in some detaiL this led to a review of formations adjacent to the Permo-Triassic boundary. CYGNET COAL l\iEASURES Difficulty is experienced in many parts of Tas­ Stra.tigraphy mania in delineating the Permo-Triassic boundary. The nomenclature and synonomy of this formation There is a macrofioral change near the boundary have been dealt with by Smith (1959, pp, 47-8). from the Glossopteris to the - DicTOidium flora but The first Significant report is that of Thureau (1881) pinpointing this change in the field is very difficult who used the term "Mt. Cygnet measures" and because of the poorly fossHiferous character of the recorded a sandstone fioor, 3 ft. of coal, approxi­ formations overlying the Cygnet Coal Measures mately 64 ft. of sandstone, a 5 ft. seam of coal and their correlates. The standard field convention within borders of carbonaceous clay, and a mica­ has been to regard the base of the first thickly ceous, laminated sandstone roof. His map and bedded quartz sandstone above carbonaceous shales section clearly show two seams separated by sand­ (with or without a Glossopteris fiora) of the Cygnet stone but it is not quite certain that the thickness Coal Measures as a convenient place for the Permo­ quoted for the inter-seam sandstone is not Triassic boundary with full recognition that this excessive due to faulting. Twelvetrees (1902) may not correspond exactly to the Permo-Triassic reported a slightly different section at the Mt.

R.S.-3 17 .... A Distl'ibution and Thickness of Cygnet Coal B-Sections in Cygnet and C·-Map of Springs Area. 00 Measures and Barnetts Member in Hobart Area Mt. Wellington Cygnet District o 2.0 Chains

Q 4 Miles L== J Q Metres f·5 " I)" l 10 Ulo ·15 Is: /Pl nnacle Rd)? t<1 (PiUinger .. i o"l Mountain j Pi i:ti Lodge , is: Member >,., F (5 Z to o Medium ot" grained Ul ., Springs t-l ..., ~ t<1 'tJ t

><~ z

tn~ ~ Z 30 Thickness of Bametts Member > 5 Thickness of Cygnet Coal Mine Thickness in Cradac Area after Leaman (pers. comm.) 3 4 5 Section on Mt. Wellington Top FIG. L Barnetts A~ Map of the Cygnet al'ea showing distribution and thickness of Cygnet Coal Measures and Barnetts rvlember. i H. Columnar sections of Cygnet Coal Measures and Springs Sandstone, Cygnet and lvit. Wellington. 500Ft·1 1. Section at Coal Mine Bay. 2. Section at Mt. Cygnet Coal Mine. :)~ 4~ 5, Section on Mt. "'''ellington. t·!\op of D ='===;:::;j C. Sketch map of part of Mt. """"ellington showing positions of sections. CcoL Mine Boy Areo, Cygnetj D. Map of Coal Mine Bay area; Cygnet~ showing position of type sections. MAXWEl.l. R. BANKS AND IKRAM H. NAQVI 19

Cygnet Mines, a small 2 inch seam, another seam aceous section is six metres thick and is overlaIn 1 ft. thick 12 feet above the first and the main disconformably by conglomerate with a quartz rich seam 3 ft. thick, 25 ft. above the second. The matrix. main and two minor seams of which only one was No macrofossils occur in this section and un­ Reeous siltstone, the lower siltstone resting on dark fortunately specimens submiUed to Dr. G. Play­ sandstone and the upper with a roof of grey hard. ford. University of Queensland, for palynological glistening, quartzose sandstone with a conglomerate study yielded no recognizable spores. in places. Reid (in Hills et al. 1922) reported two main and two minor seams of which only one \vas A thin section (33626) of carbonaceous siltstone being worked. At Berry's, north-west of the main from the middle of the formation revealed a modal working (fig. 1 A) a sandstone roof caps a seam grainsize of the clastic particles between 0.03 12" -14" thick resting on 12 inches of black shale. and 0.045 mm. with variation from very fine silt This shale rests on 8 feet of shale and 20 feet to very fine sand. Aggregates up to about 0.4 mm. of sandstone which in turn overlies mUdstone. Thus long of clay grade carbonaceous matter comprise it would appear that in the nominal type area the about 30% of the rock, the only other clay grade coal measures are at least 6.8 m. thick and may be material being a very minor amount of interstitial over 12 m. thick on Thureau's figures. Recent clay mineral. Clear unstrained quartz is the investigations by the authors show that detailed commonest constituent of the rock, about 55%, information in the type area is not available because but a little quartz with globular inclusions in lines of thick vegetation and deep soil cover. The also occurs. Plagioclase (oligoclase) is about 1 % Cygnet Coal Measures appear to overlie Ferntree of the rock and there is a surprising amount of Mudstone, to be between 12 and 18 m. thick and biotite (about 12%) and some muscovite (about to be overlain by quartz sandstone and thinly­ 3 % ) . The quartz and plagioclase grains are pre­ bedded (beds about 1 cm. thick) micaceous sand­ dominantly angular and very few are as rounded stones and siltstones. as subrounded. Their sphericity varies from equi­ dimensional to very elongate and many have The best section now available is that in shore lengths exceeding three times their breadth. The platforms at Coal Mine Bay and immediately south quartz and feldspar grains lie in a felted matrix of

Rock Types TABLE I The Cygnet Coal Measures and its correlates 1 2 :1 4 consist of sandstone, predominantly feldspathic but AcanthotriZetes tereteangulutu8 x siliceous in many places, carbonaceous and mica­ A napiculatisporites ericianus x x :x ceous siltstone and thin seams of coal. Conglomer­ Anapiculatisporites cf. ericianus x atic beds are rare but occur at Mount La Perouse, Zeehan and Launceston at the present limits of Bascanisporites undosus x outcrop of this formation and at Hugel Creek near Cycadopites vetus x x :x Lake St. Clair. The sandstones are commonly Deltoidospora dirccta x x x x carbonaceous and in many places are cross-bedded Granulatisporites and ripple-marked. No study has yet been made rnicronodosus :x x of current directions. The sandstones predominate Granulatisporitcs trisinus x x x x in most sections especially on Bruny Island, at M arsupipollenites triradiatus x x x Mount La Perouse, Zeehan, the Du Cane Range, PUaspora plurigenus x x x Mount Inglis, Beaconsfield and LHydale, again Proiohaploxypinus aJnplus x x x areas close to the present limits of outcrop. Although insufficient sand-shale ratios are available Protohaploxypinus sewardi x for accurate depiction of variations there is a PllnctatispOTites gretensis x suggestion that the silt content of the formation Retllsotriletes diversijorrnis x x is higher at Poatina, Mole Creek, the Pelion Range, Schopfi,pollenites sinuosus x some areas neal' Lake St. Clair, Maydena, South­ port, Cygnet and Hobart than around the margins Striatoabietites m nltistria tus x x x x of present outcrop. This in turn suggests deriva­ Striato1Jodocarpites fusns x x tion from higher source areas not far outside the Thyrnospora cf. harnatus x present limits of outcrop in the north-east, north, 1 .. - Cradoc ( Welling Ck.) north-west, west and south. 2 -= Mount l'elion. ., Mole Creek. The presence of feldspar as a common constituent 4 --::. \Ve::;tern BiufL of the sandstone is unusual. The non-marine sand­ Taxononlic comhinaUons after Hart (HI65), stones of the I,ower Permain Mel'sey Group do contain feldspar but not in significant amounts and 'The spores listed indicate correlation of the coal t,he prominent sandstones of the Lower Triassic are measures at Cradoc, Mount Mole Creek and quartz-rich sandstones. No detailed Western Bluff with one another with the New- studies have been made and the type of castle Coal Measures of New South Wales wbich not determined. are Kazanian or Tartarian. The Jad;:ey Shale at ·Western Bluff in particular is probably Coal occurs at Coal Hill near Lake St. Clair. to the of the Newcastle Coal and Tier {probably a single Balme reported that the coal measures at seam each two or more seams occur Farm, Hobart and Malanna were of on the Pelion Mole Creek, Bruny Island and The Kazanian or Tartarian age is the being at Cygnet (1.5 m., of the occurrence of the same 1881) . in the Condren Member of Western Australia, Member with goniatites or Lower Kungurian age (Glenister and Furnish 1961, P. 688) and over­ Flora lain by the Hardman Membil' with Aulosteges ingens, and which occur Johnston ( recorded Gangamopteris in the Basleo of Timor are similar to spathulata., G. In.<:w),,,t,,,··.,", orowniana val'. spaeies in the Kazanian of Russia (ColemEm 1957. rem ains from p. 139 i thus Kazanian age. The Vertebraria spores from the Coal Measures lllay how- beyond the limits of the Condren arc clearly pre-Scythian. be referred to determinations from require revision, Prom Scmthport Johnston (l Stratigraphic It!'!ati.onships reported Veriebraria australis. McLeod et al. \1961. 2.5) Gangamopteris and The stratigraphic relationships of the Cygnet. and Banks and Coai Measures as here used are summarised in fig. recorded .these fossils, 2, In north-eastern a.nd eastern Tasmania. the from the Malanna formation is missing. A disconformity is found at area, near Zeehan. Glossopteris and Vertebral'ia the base of the formation in northern and western occur in the Mount La Perouse area. Tasmania whereas in the cent.ral and southern parts of the basin, where the formation is thickest. Spores from the Cygnet Coal Measures and their the basal contact is conform.able. This, with the correlates have been reported by Dulhunty and smaller grainsize in the central parts of the basin Dulhunty (1949), Balme and Hennelly 0955, 1956a, suggests. lower stream velocities, and probably 1956b), Balme (in Jennings 1963) and Balme grades) 111 those areas, indicating perhaps a 10w­ (1964), and their distribution is summarised below lying, fiattish central area bounded by low rather as Table I distant hills. ' 22 SOME FORMATIONS CLOSE TO' THE PERMO-TRIASSIC BOUNDARY IN TASMANIA

-----~------.~; -., --.-~--.--.-.

CYGNET COAL MEASURES Distribution Thickness and Stratigraphie Relationship

TASMANIA

.,'

3aMolecr,ek ~. , lY,.

Zeehon '1- AS­ 'd

Explon otion

D Disconformity } GradatIon Upper Contact of Cygnet

\ ... 30 Flor&ntlne 33 Thickness in metres \ g Valley d DiscOnfOrm!t y } Lower Contact of Cygnet 9 Gradotion 4'~ Maydena R/pf Triassic resting or; Ferntree Mudstone R/Dev.Triassic resting On Devonion

: -SCALE -

! • ___ -f e.o KMS.

FIG. 2.-Map of Tasmania showing distribution, thicknesses and stratigraphic relationships of the Cygnet Coal Measures.

The upper contact shows less systematic variation isopach maps possible in view of the likely effects in its character although again gradation occurs of disconformities and until detailed correlations in most of the thicker sections and disconformity between sections become feasible. A zone of thicker in the thinner. Variation in thickness of the for­ Cygnet Formation appears to extend from south mation and type of upper contact in the area of Zeehan through the Du Cane and Pelion Ranges around Cygnet suggests considerable differential to Mole Creek, Poatina and Beaconsfield but this erosion and possibly uplift before deposition of the has an area within it at Westmoreland Creek, just overlying Barnetts Sandstone (see fig. 1 A), east of Mole Creek, where Triassic rocks rest directly on Ferntree Mudstone (Jennings 1963), The thickness of the formation varies up to Another zone of thicker sediments appears to extend about 105 m. but measured sections are exposed from Mount La Perouse through Southport to too sporadically to make construction of useful Bruny Island and Hobart. SANDSTONE SilTSTONE ~ Thick- bedded ---Siliceous ~ Thin - bedded Carbonaceous rz=J Shaly, fissile D Cross - bedded ~ Micaceous C.TJ Micaceous MUDSTONE I]IJ Quortz r::-:C:J PLANTS ~ Arkose LundblGdi~ [IJ Glos.sopteri~ ~ Feldspothic ~1 yertebraria CONGLO~~ERATE Quartz - pebble Ripple Marks Clay - peLlet Cool ~ Mud - pellet ~> ~ i:'l t" t" ?:l ttl :> Z :>:: w. > Z t:I

H :>:: ~ ~ :z: Metres ;..z 50 [ § 40 f~: a b c d e f 9 h k m MT. WELLINGTON ADVENTURE COAL MINE COAL BORE ~~ORTH SIDE MISERY RA. I'll. MANFRED PELION RA. MIINGLIS BADGER RIVER f~ARAI\OOPA CK. POATINA LILY DALE SPRINGS ARE A BAY BAY SOUTHPORT of LA PEROUS E FLOK V,\LLEY LAKE ST. CLAiR ARE/\ eiAlANNA MOLE CK< 8anks II. Alwar after MinRes.7 after Naqvi Twelvetrees Glenister&: Banks after Corbett Gulline Me Leod etaL Gee Benks & Ahmad Jennings Me.Kellar Longman 1965 i9S; 1963 1962 1963 1957 1966

:FIG. 3.-Columnar sections of the formations close to the Permo-Triassic boundal'Y in Tasmania.

~ eo.> 24 SOME FORMATIONS CLOSE TO THE PERMO-TRIASSIC BOUNDARY IN TASMANIA

SPRINGS SANDSTONE "C" and" E" (jig. 1 C). The main rock type is The name was first used, casually, by Lewis a well-sorted, medium to coarse-grained quartz (1940) and explained more fully by Lewis (1946, sandstone, crotis-bedded and with beds generally pp. 23, 35) by whom the thickness on Mount Wel­ more than 75 cm. thick but with some units more lington was given as 800 feet, 200 feet more than thinly bedded. the average elsewhere. Elsewhere in that publica­ Current ripple marks occur on some bedding tion (Ibid. p. 46) the thickness was given as 500 planes and suggest currents from the northerly feet, or even, it would appear, 1,000 feet (p. 67). quarter. The cross-bedding suggests currents from It was described as massive sandstone. Lewis did northerly, westerly or south-westerly areas. The not define the term very stringently and it is not cross-bedding commonly shows current drag effects clear if he considered the Springs Sandstone as only at the top of the set. The predominant mineral is the cliff-forming sandstone or as all the rock quartz, often with sparkling surfaces which are between the Ferntree Mudstone and the dolerite crystal faces resulting from authigenic outgrowth, contact on the Pinnacle Road. with very minor amounts of muscovite, graphite A section was measured from the top of the and feldspar «2%). Clay-pellet and quartz pebble Ferntree Mudstone on the Pinnacle Road to the conglomerates also occur. A few grains of pink dolerite contact (fig. 1 C) by M. Anand Alwar garnet were seen by hand-lens in the basal unit. and one of the authors (M.R.B.) some years The section above the Springs Sandstone, as ago and is recorded here as figs. 1 B, 3, 4, 5, and here used, on the Pinnacle Road reveals sandstones the lower part of it as fig. 3 A, and in more detail and some siltstones, the sandstones finer-grained in an Appendix. The cliff-forming, "massive" and more thinly-bedded on the whole than those in sandstones below the Springs Hotel (A-E on fig. the Mountain Lodge Member. Some beds with 1 C) form a distinctive mappable unit above which 10%-20% feldspar occur and a green micaceous the section is very discontinuous. mineral, probably a chlorite, occurs on several horizons. Cross-bedding is common and ripple­ It is proposed therejore to dejine the Springs marking present. Dips of cross-bedding appear to Sandstone as that unit oj quartz sandstone with be too variable to allow assessment of current minor siltstone and clay-pellet conglomerate over­ directions without statistical analysis. lying the Cygnet Coal Measures and underlying Fossils have not been found in any of the units a unit oj sandstone and siltstone, probably equiva­ in this section. The Mountain Lodge Member is lent to the Knocklojty Sandstone and Siltstone, in like the lowest sandstone units in the Knocklofty track, road and cliff sections below the Springs Formation a few miles to the east and these sand­ Hotel. It is about 92.5 m. thick in its type section. stones are overlain by fossiliferous siltstones of (511.45 E. 717.05 N. to 511.52 E. 716.53 N.l. No Scythian age. The most likely age for the Springs jossils have been recorded in it jrom its type Sandstone is Scythian but there is no reliable, direct section but it is probably Lower Triassic. evidence for this. In its type area it may be considered as consisting In the Coal Mine Bay area near Cygnet, the of two members, a lower one about 55 m. thick Cygnet Coal Measures are overlain by a sandstone of sandstone, especially sandstone with beds of the formation, divisible into a lower member called the order of 10 cm. to a few tens of centimetres thick Barnetts Sandstone and an upper member, litho­ and fissile sandstone, siltstone and rare clay-pellet logically similar to the Sphinx Rock Member of conglomera,te, the upper one about 37.5 m. :thick the Hobart area. of coarsely cross-bedded cliff-forming sandstone The Barnetts Sandstone may be dejined as the in beds mostly of the order of one metre thick. lower member oj the Springs Sandstone discon­ The lower member may be correlated with the jormably overlying Cygnet Coal Measures at Coal Barnetts Member, well-exposed in cliff and shore­ Mine Bay. It consists oj 27 m. oj jeldspathic sand­ platform sections south of Coal Mine Bay, Cygnet. stone, and is named jor Barnetts Trig. Point (498.20, Unit G (see Appendix) may conveniently be 678.00) near Coal Mine Bay. It is probably Triassic taken as the base of the upper member in the track but may be Upper Permian. section" A "-" C" and unit j in the road section The section of the formation (see fig. 1 Band "D "-" C ". The lower member consists pre­ Leaman & Naqvi 1967) commences with a basal dominantly of well-sorted medium to fine-grained conglomeratic unit (pl!l!te I, f. 1) about one metre sandstones with minor developments of coarse­ thick, conSisting of granule and pebble conglomer­ grained sandstone near the base,and some carbona­ ate, feldspathic sandstone and thin beds 0-2 mm.) ceous siltstone with plant fragments. The sand­ of coal and carbonaceous siltstone. The granules stones consist predominantly of quartz with minor and pebbles are composed of quartz. carbonaceous quantities «10%) of muscovite, graphite and siltstone, a mUdstone (probably Ferntree Mud­ feldspar. They are cross-bedded and have beds a stone), and rarely of schist. Most of the larger few centimetres to about 75 cm. thick and some are fragments are 3-4 cm. long but some of the angular fissile. The carbonaceous siltstone contains plant fragments of carbonaceous matter lying across the fragments (probably sphenopsid), sandy beds a few bedding reach a length of 10 cm. The matrix miIIimetresthick and some clay pellets. Outcrop of consists of angular quartz fragments with a minor this member is somewhat subdued in contrast to clay cement. Quartz comprises about 60% of the the strong development of cliffs by the upper rock. Kaolinised feldspar in the matrix comprises member. about 10% of the rock. Muscovite and graphite The upper member, which may be called the are common, and rutile, limonite and distinctive Mountain Lodge Member, outcrops well in the patches of pink garnet are present. The conglomer­ cliffs beneath the Springs Hotel and at Sphinx ate is disconformable on the Cygnet Coal Measures. Rock. It may be considered to include units G The conglomerate is cross-bedded, the current to Y in the road, track and cliff sections between having come from 50 0 • MAXWELL R BANKS AND IKHAM H. N AQVI 25

Above the conglomerate is a bed of coal and zircon in a matrix of ferruginous material (7%), carbonaceous shale nine cm. thick and at least kaolin (3%) and felted masses of sericite 00%). three metres long (plate 1, f. 2, 3). In places just The clastic particles are mainly equidimensional below the coal is a bed five cm. thick of inter­ although a few quartz grains and some muscovite bedded sandstone and carbonaceous siltstone like laths appeared to have length about four times that immediately beneath the basal conglomerate. breadth. The grains are angular and sub angular The sandstone and siltstone unit passes west into but a few are subrounded. The framework is a unit up to 30 em. thick of interbedded coal, intact. The modal grain size is fine sand with very carbonaceous siltstone and sandstone (plate 1, f. 3). little coarse sand, much medium sand and very fine A fine to medium-grained, fairly well-sorted sand­ sand and some coarse silt. A few lenses of clay stone overlies the coal and, where coal is absent, pellet conglomerate occur in this sandstone. The the basal conglomerate. This is cross-bedded, the cross-bedding indicates currents from the north­ individual laminae being a few millimetres to a west. few centimetres thick, and in the top part ferru­ A reddish, fine-grained, well-sorted, fissile, quartz ginous concretions form up to 2% of the rock. sandstone (33621) overlies the feldspathic sand­ The sandstone (33G24 in the University of Tas­ stone. This rock is predominantly quartz (80%) mania Geology Department collection) consists of with very little muscovite (2%) and plagioclase angular and a few sub angular grains of quartz (2%) which is probably olil-{oclase. The clastic (about 50 %), kaolinised plagioclase (about 30 (70) grains are predominantly in the fine sand grade, a which is oligoclase where it is determinable, some few only in the medium sand grade, some in the micro cline (1 % ) , muscovite (5 %) and a few quartz­ fine sand and little, if any, in the coarse silt ite fragments in a limonitic (10%) and sericitic The framework is intact, made mainly of (5%) cement. The clastiC grains are equi­ angular, equidimensional quartz grains with a few dimensional to elongate (length up to five times quartz and muscovite grains about three times as breadth) and are fine sand (modally) but range long as broad. Many of the quartz grains have from medium silt grade up to coarse sand grade. sutured contacts but others are separated by a The quartz is predominantly clear and unstrained felted sericitic matrix composing ten to fifteen but some with undulose extinction is present. A per cent of the rock. few sand grade grains of zircon and rutile occur. The framework is closed. Nlany quartz grains have The final unit is a thin, transitional bed of sutured contacts against one another whereas the well-sorted, finely laminated boundaries of others appear to be sericitised. The which consists of quartz ferruginous cement seems to have been emplaced (sericitised, 10 muscovite before of the sericite in most places (2 o/c) and an argillaceous but in a few sericite appears to be earlier.- The This member differs significantly from other rock rock 1s an arkose. units in the abundance of the This sandstone is overlain by the thinly-bedded, a few em. thick) feldspathic of small, spheroidal (83623) with pieces of carbonaceous cretions. The abundance of siltstone like that of the Cygnet Formation form­ Mine Bay helps to differentiate the over- ing about 8% of the rock.. At the top of this unit member. The pellets are discoidal to ellip­ is a bed 17 em. thick of conglomerate of which up structures up to eight cm. long and although to 75% comprises this siltstone (plate 1, f. 4). the majority are siltstone, some arc of feldspathic A thin section of the sandstone showed the follow- sandstone, suggesting contemporaneous erosion. composition: quartz, mainly unstrained, some Graphite from this member was examined by X-ray undulose extinction, some with lines of globu­ diITraction methods and its identity established. lar inclusions, 60%' rock fragments, quartzite, silt- The quartz sandstone conformably and tran- stone, micaceous 5%' feldspar, mainly sitionally the Barnetts Member at Coal plagioclase Eot certainty, some Mine f. 8) is a.t least 27 m. thick in carbonaceous this the Bay area. It is medium- 7 % ; senclt,(' grained. 'Nell-sorted, cross .. bedded with beds of the rutile. tourmaline. zircon in small order of one metre or more in trlickness. Limonitic is closed and the rock concretions and clay pellets are present but not and a few Rub·· so common as in the Barnetts Member The rock. are neaT'ly equi- (33619) consists of angular and subangulal' quartz rJil':'.er:sional up to four times particles ), muscovite (5 %) and fragments of the brea.dth~ is nne sl:lnd and siltstone, siltstone, and some fine- particles coarse silt to medium grai:>:led sandstone (10%) with common particles of Some bet.ween quartz or rutile and tourmaline with a minor cement quartz and feldspar are but of sericite laths (about 5 of the ferruginous or sericH.lc separates others. quartz particles showed almost The next unIt is a pale red to purple, fine-grained, contacts with other particles. of the quartz well-sorted, strongly cross-bedded (plate 1, f. 5) is equidimensional but a few particles up to four feldspathic sandstone (336221. Examination of a times as long as broad occur. The quartz grains thin section showed that the rock consists of quartz are predominantly of clear, unstrained quartz but (about 65 %); fresh and sericitised feldspar (oligo­ some strained quartz, some quartz with rutile clase where determinable, 5 .. 7%); muscovite (5-7%), needles and some with euhedral zircon crystals fragments of very fine-grained quartzite, siltstone and lines of inclusions occur. A few grains may and clay (3 o/c), graphite (1 %) and a few grains ot' show signs of regrowth. The tourmaline in this 26 SOME FORMATIONS CLOSE TO THE PERMO-TRIASSIC BOUNDARY IN TASMANIA as in other thin sections from the Barnetts Member tains two spores of stratigraphic significance also is green. The zircons are predominantly euhedra present in the Ross Sandstone. Thus at Cygnet or broken euhedra. Limonite staining is present and Hobart the Permo-Triassic boundary lies within but minor. The grains are mainly in the medium the range of the Cygnet and Springs Formations. sand grade with some coarse, some fine and some very fine sand particles. Tourmaline, ilmenite, melanite and some zircon and magnetite were THE PERMO-TRIASSIC BOUNDARY IN separated as heavy minerals and comprise in all TASMANIA about 0.2% of the rock. This member corresponds Evidence has been presented earlier that the in character and position to the Mountain Lodge Cygnet Coal Measures are Upper Permian, contain Member in the Hobart section. the Dulh.untyispora Microfiora, and at Western Bluff Rocks apparently similar in character to the at least may be correlated with the upper part Barnetts Member and overlain by thickly-bedded, of the Newcastle Coal Measures of New South usually cliff-forming quartz sandstones have been Wales. A siltstone 85 m. above the base of the reported by Mather (1955, pp. 96-7) from the Huon­ Ross Sandstone at Poatina contains Lundblaaispora ville area, and Hale 0953, p. 122) from Brooks brevicuJa and Densoisporites play!ordi (Playford Bay where they are 45 m. thick and overlain by 1965) and a siltstone at Crisp and Gunn's Quarry. cliff-forming quartz sandstone between 30 and 45 m. Knocklofty. in the Knocklofty Formation also con­ thick; from the Police Point area (Ibid. pp. 122-3) tains these spores (G. Playford pers. comm.l. The and from the eastern shore of Port Esperance presence of these spores suggests correlation with ([bid. p. 124). An outcrop on the Esperance River the Collaroy Claystone of the Narrabeen Group of south-west of Geeveston (Ford, 1954, p. 155) may New South Wales (Evans 1966) and with the be a correlate of the Barnebts Member. Further Kock.atea Shale of Western Australia (Playford afield the top 40 m. of the section beneath the 1965) which is Otoceratan (basal Scythian). The thicldy-bedded sandstones on Mount La Perouse two lower spore assemblages (Tn la, TR Ib) reported may correspond to the Barnetts Member and 15 m. from New South Wales and Queensland (Evans of feldspathic sandstone overlies carbonaceous silt­ 1966) have not yet been found in Tasmania. In stone near Maydena and may be a correlate of the New South Wales the thickness of sediment separat­ Barnetts Member CJago 1966). ing the Collaroy ClaYstone from the Newcastle Neit.her the Barnetts Member nor the overlying Coal Measures is nearly 400 m. (Hanlon et aL 1953) member at Coal Mine Bay has yielded fossils. considerably more than that between the Lundblad­ ispora horizon and the Jackey Shale. At this The Barnetts Member in its type area has some stage it is impossible to fix with any precision the unusual features. The second unit (33624) is an boundary In Tasmania corresponding to the base arkose with a high proportion of kaolinised plagio­ of the Narrabeen Group in New South Wales, and, clase and very few rock fragments. Its composition of course, to divide the Tasmanian succession accur­ and texture suggest relatively short transportation ately in terms of a world standard Permo-Triassic by water from a terrane of granitic and meta­ boundary (top of the Cyclolobus Zone). morphic rocks. Higher units contain progressively less feldspar. Although in each case the modal For local convenience in field mapping this grade is fine sand, the estimated mode decreases boundary has been taken at the base of the first from about 0.2 mm. (33624, 33623) to 0.15 mm. thickly-bedded sandstone above the highest car­ (33622) and 0.12 mm. (33621), and the range in bonaceous siltstone. This may well correspond to grainsizes decreases in general from six grades the base of the Sphinx Rock Member as both on (33624) to four (33623, 33621), and an impression Mount Wellington and at Cygnet the Barnetts of better sorting and longer transport emerges from Member contains carbonaceous siltstone, is this preliminary study. No change in terrane relatively thinly-bedded and has a subdued out­ during deposition of this member is evident. The crop whereas the Sphinx Rock Member outcrops cementing material in all sections studied is seri­ boldly and the obvious place for a field mapping citic and ferruginous. boundary is at the base of this member. The entry of cross-bedded, feldspathic sandstones is The upper member at Coal Mine Bay shows a probably not a suitable horizon for distinguishing marked increase in modal grainsize (about 0.35 the Cygnet from the Springs Formation as the mnl.) and in the size of the coarsest fragment Cygnet contains feldspathic sandstones in many present 0.1 mm.), with a continued decrease in places although at others, e.g., Malanna, it does the feldspar content. The current velocity of the not seem to have Significant quantities of feldspar. depositing streams apparently increased but feld­ The base of the Barnetts Member is a convenient spar was no longer available in the source area due boundary as, in some places at least, there is a to deep weathering of quartzo-feldspathic rocks or disconformity and a thin conglomerate at this leveL to the earlier source of feldspar having become However. in the type areas and elsewhere the con­ covered. ditions of deposition did not change significantly at this horizon; coal and feldspathic sandstone Correlations occur both above and below the contact. In addi­ Suggested correlations, based on lithological tion, disconformities, even those overlain by thin similarity and stratigraphic pOSition are shown conglomerates, within fresh-water successions may in fig. 3. The most significant point emerging from well be of very local extent and short duration. these correlations is that if correlation with the Angular unconformities between the Ferntree Mud­ Ross Sandstone at Poatina is correct, the Springs stone and Springs-type sandstone have been Sandstone is Scythian and may be roughly corre­ reported by Woolley (1959) and Alwar (1960) but lated with the Knocklofty Formation which con- the evidence is not incontrovertible. Attempts to MAXWELL R. BANKS AND IKRAM 1L NAQVI 27

determine the age of this boundary at Cygnet and BURNS, K. r.., 1964.-Explanatory Report. One-Mile Geological Map Series-Devon port, Dep. Min. Tasm. Southport by palynological studies failed because COLEMAN, p~ J.~ 1957,-Permian Productacea of Western Aus~

of poor preservation of spores. Disappearance or traHa. Bull. Bur. Min. Resour. AU8to J 40. virtual disappearance of feldspar from the section CORBETT, K. D., 1964.---The Geology of the Florentine Valley. does not appear to be a good criterion for separat­ Unfl}. of Tasm. Thesis. Unpnb. DAVID, T. W. Bot 1932.-E~xplanatory Notes to Accompany iii. ing Barnetts and Mountain Lodge Members as this New Geological Map of the Commonwealth of Australia. seems to have occurred at different levels in differ­ Coun. Sci. Ind. Reso, Melbourne. DULHUNTY. J. A.~ and DULHUNl'Y. R., 1949.-Notes on Micl~O~ ent places. In fact, some of the sandstones above SPOrf-"frypes in Tasmanian Pern1ian Coals. Proc. Linn. the Mountain Lodge Member on Mount Wellington 80c. N.S.lV., 74. 3-4, 132-139. appear to contain considerable quantities of feld­ EVANS, P. ll., 1966.-·-Mesozoic Stratlgraphic Palynology in Australia. A "1,{,8t. Oil (llas. ~l.~ 12. ti. 58~63. spar. The appearance and disappearance of Ij"oRD, H ..J .• 195().--The GEology of the Upper HUOllwArve River feldspar is probably related to uncovering and Area. Pap. P'ruc. Roy. Soc. Twnn .• 90, 147-156. covering of different source areas at different times. GEE, H. D.t 1964.---Permian Hocks of the Mount IngHs Area. At Cygnet and elsewhere the contact between Tech. Rep. Dept. Min. Ta8rn" 8, 57-63. Barnetts and Mountain Lodge Members appears GLP.NISTER, B_ F. W. M .• to be gradational and the main objective criteria Ammonoids Palaeont. p D. 1959.~~Geology of thH Beueonsfield Djp;trict. inclnd~ for disting'uishing them are bedding thickness, Andersong C:reek Ultl'abasic Conlplex. Rec~ Q. grainsize and outcrop behaviour. HI. GULLINE~ A. B., Finally, delineating formations in a persistent Geological Map fluvial environment is difficult just because of HALE, E. A., 1953.-The Geology of the environment and no difficult if the fluvial Pr(J(', Roy. Soc. Tasrn .• 1952, environment changes slowly with time as may be HANLON) P. N.~ OSBORNFJ. G. D.~ and R.AGGATT, N ar.l'aheen Group: Its subdivisions suggested for the interval from Cygnet to Mountain bE'tvleen the South Coast and the Lodge. Distri~tf'i< J. Proc, Roy. Soc. ]v".S. ~V., Thus no a priori case can be established for pre­ HART. P 1965.~ .. ~-·The SysternaticB and Distribution of Perraian Witwatersrand Univ. Press. felTing the base of the Barnetts to the top of the A. M., NYE, P. B., KElT), H. Barnetts as a system boundary and neither may 1922.-.rrhe Coal Resources correspond to the top of the CycloZobus Zone. A Tasm.~ 7. floral occurred at about the time of 1966.-·-Geology of the Maydena Area. Uni". of deposition the Barnetts Member and establish- Thesis. Unpub. I, E., 1963.~-f~xplanatory ReporL One-1vIHe GeologIca] ment of the position of this change in difl'erent Series-11iddleRex. Del}. Min. Tas1n. sections may eventuallY provide a better time JOHNSTON. R., M.. on the GeOlogy of Bruny Island. marker than is now available. As is widely realized Pap. Proe. 1886, 18-26. by Australian palacontologists this change itself -~-~-~------• .lk88.-SYNt(Yfnatic Account of the Geolof}V of may not correspond to the system boundary else­ ~ratlmania. Govt. Pl'lnte1'1 Hobart. ' ----~~.--~--, 1893.--Further Contributions where. Flora of Tasmania, Part 1. Pap. Proc. Roy. (1893), 171-72 CONCLUSIONS D. E .• 1\J6f).~Geo]ogy of the Cl~adoc Area. Uni-v. of Thea'is. l.lnpub. The Cygnet Coal Measures of Upper Permian age -.------.--...-.---- and NAQvr, r. H., 1967.- Geology are followed disconformably or gradationally by the physics of the Cygnet District. Bull. GeoL SU1'v. Springs Sandstone, the basal member of which, in press. Barnetts Sandstone, is thinly-bedded, fine-grained LEWI8. A. N.. 1940.--Record of GlO:~Bopteri8 from Cygne-to and in places arkosic and the upper member, Pap. Pmc. Roy. Soc. Ta8m. (1939), 95-96. -... ------., 1946.-The Geology of the Hobart District. Mountain Lodge Member, a more thickly-bedded, Roy. Soc. Tasrn .• Hobart. coarser-grained protoquartzite. The Springs Sand­ LONGMAN. M., :t966.·~-Explanatory Report. One-Mile Geologica! stone has few identified fossils but by correlation 1vTap Series---Ijaunceston. ncp. 1VIin. '1'aB'rn~ . with the Ross Sandstone Is probably basal Triassic McKELLAR, .f. B. A., lU57.·-··G<~ology of the Western (otocera.tan). The most convenient horizon to the Great Lake, Tasmania. Rec. Q. Viet. 1JI1u8., .MCLEOD, 'V. N., ,TACK. R. H., and THREADER, V. M., take as the Permo-Triassic boundary in Tasmania. Explanatory Report. One-Mile Geological Map is probably the base of the Mountain Lodge Sand­ Du Cane. Dep. M·in. l'usm. stone but this is debatable and certainly may not MATHER, R. P., 1955.-GeoJogy of the Huon District. Pap. correspond temporally to the Permo-Triassic Proc. Roy, So(:, TasTn., 89, 191~202. boundary elsewhere. MILLIGAN, .To, 1849.--·Reports on the Coal Basins Diements Land, Pap. Proc. Roy. Soc. Tasm.~ 1, 13 REFERENCES. NAQVI. L H.. 19G6.-The G,'ology of the Cygnet Peninsula. Univ, of Tns'Yn" Thesis. Unpub. ALWAl<. M. A., 1960.---Geology and Structure of the Middle .PLAYFORD, G., 19G5.-Plant MicrofosRils from Derwent VaHey. Pap. Proc Roy, Soc. TaSl1'i.~ 94, 13-24. ments near Poatina, 'rasmania. J. Geol. BALM;]. B. E., 1964.-The Palynological Record of Australian Pre­ 2. 173-210. Tertiary Floras. Ancient Pacific Floras, Uni1J. of Hawaii Press, 49-80. HEAD, D. ·K. 18130.-Geology of the Claremont-Old 13e,,"h Area. ------and HENN1<~LLY. J'. P. P., 1955.-Bisaccate Sporo~ Un'i-v. of Tawm., Thesis. Unpub. mol' phs from Australian Permian Coals. Aust. J. Bot.. SMITH, E. M .• 19~9.-·-Tasmania. Lexiqu.e Stratig'ral)hiq'Ue lnter­ 3, 1, 89-HH. 't1,ational. VI. 5d. Inter. Geol. eongr. -----.------. 1956a.-Monolete, Mono­ ~.'HVRE:AU, G.) 1881.--Report on thf' Coal Mines in the vieinity colpate and Alete Sporomorphs from Australian Permian of Gardw;'r's and Randalls Bays. 1.'asrn. Leg. Count! Sediments. Aust. J. Bot., 4, 1, 54-67. Pa.p., 91. . ------.-----.----., 1956b.-Trilete Sporo­ morphs from Australian Permian Sediments. Aust. J, TWELVETREES, W. R .• 1902.-Report. on Gold and Coal at Port Bot., 4. 1, 240-260. Cygnet. Tasm. Secy. for Mines Rept., 1901-2. BANKS, M. R. and AHMAD, N., 1962.-The Permian System in -.-.... ------~, 1911.--The Tasmanite Shale Fiel<1s of Western Tasmania. Pap. Proc. Roy. Soc. Tasm., 96, 1-18. the M.,,-sey District. Bull. Geol. Surv. Ta.,m., 11. BONWICK, J., 1870.-A Sketch of the Geology of Tasmania; in WOOLLEY. D. R., 1959.-The Geology of the New Norfolk-Black The Daily Life of the Tasmanian.. London. 267-293. Hills District. Pap. Proe. Roy. Soc. Tasm., 93, 97-110. 28 SOME J"Ol~MATIONS CLOSE TO THE PERMO-TRIASSIC BOUNDARY IN TASMANIA

ApPENDIX I Unit Section from Dolerite Contact on Pinnacle Road K t 0.61 m. SS, beds order of few cm. thick. to Cygnet Coal Measures J, 2.74 m. med to fine-gnd SS, qz, muse, feld TOP-dolerite contact, " F" on fig 1 C « 5 %) , graph uncommon, biot or chlor, beds 7,5 ern. to m. Unit thiek; 10 YR 4/2. F 1.20 m. fiss., fine-gnd S3 and coarse sltst, I, 0,75 m, med.gnd SS, musc, feid 10 Y R 6/6. sets 2.5-5 em. thick, l!"12 2.73 m. e.b. SS, cross-bedded sets approx 5 cm. 130" at 12°; 10 YR 616. thiek; beds of order or one m. thiclc H, 0.'15 m. clay-pellet conglom. D, 1.36 m. very fine-gnd SS, cross-bedded sets abt G , 0,'/5 m. med-fine gnd SS, qz, muse, graph, feid 10 cm. thick; beds 10 to 30 em. thick «5%); beds 0.5-10 cm. thick. 10 approx. YR 6/6. C, 0.91 m. bed of qz ss with some feld; base ]<', 4.42 m. med-gnd SS, qz, feld, muse, graph, orOS8- bedded. pelL up to 5 em. long 5 Y 5/2; B:l 4.12 m. fine-gnd qz ss with some graph; fiss, up to 2.5 m. thick. c.b. Second road to chalet above Mountain A2 1.06 m. ss with beds of order of 30 em. thick. Lodge. Gap of 17.22 m. including first road to Z, 1.21 m. coarse-gnd s1tst, beds 7:5 to 25 em. chalet above Mountain Lodge. thick; non-fiss; 5 Y 6/4. " E" on fig~ 1. Gap of 50.14 m. E, 0.45 m. med-gnd S8, qz, feld «5'/'». muse y, 1.21 m. c.b. ss; sets 2.5 cm. thick; cross­ (rare), graph, green biot, or ch10r; bedding dips 20'r at 15°. contains large rounded frags up to X, 2.12 m. thick bedded (beds about 0.75-1 m. 5 ern. long -which look like F'ern­ thick) ss. tree Mudstone and garnetiferous Wi 1.06 m. e.h. containing qz, phyllite; bedding up to 2.5 em. muse; sets ern. thick; thiek; c.b., 10 YR 8/2. 170 0 5 B G 6/2. D,S.50 m. S8, beds up to 1 m. thiclz. Vi 1.06 m, c.b. ss with discoidal ferruginous C, (J,07 m. 6 beds, eaeh em. tbick concretions to 2.5 cm, long; of clay sets 0.5 to em. thinner to 1 cm. near base; thicker near 1 mm.) U, 0.91 m. thick bedded 53, (some bands up feld these gns are more angular Gap of 16.14 m. in others), muse, rare graph, 1\ 2.'14 m. med-gnd S8; qz, musc; feld un­ beds of conglorn lenticular, 10 common; beds :~o ern, to 1.25 m. 4/2. near 2.5 to 5 em. thick near top; 10 6/2 or 5 BG 5/2. H, 2.40 m. Gap of 50.75 m. A, 0.60 m, feld (5-10%), , beds 45-15 em. 8, 0.61 m. med··gnd 55 with qz, muse, feid ). e.b.; bedding to 20 em. z 1.20 m. fiss, fine-gnd ss with muse, feld. graph, in beds 0.5 to cm. thick; Gap of 3.02 m. ripple marked, 10 Y interbedded Rt 2.25 m. friable med to coarse-gnd ss. with eoarsG-gnd Y 5/2, with Q, 0.60 m. med to coarse-gnd feld (abt ? animal tracks. some fresh), 22-40 cm. Y 14.94 m. c.b.: S8 beds 30 cm. and more current drag structure Pi 0.35 m. coarse to med-gnd sltst, qz, muse, forming; some small graph, very flss. cretions. 0, 3.50 m, ) , muse, }:, 0.30 m. ss, beds 0.5 to 2.5 ern. t.hick. up 1.25 m. W 0.02, pell and qz conglom; pell to 0.05 m. ern. long; changes along Gap of 108 m. to N. Nt 0.76 m. fine-med-gnd feld rare. single bed, 7/4. 'J 6.'70 m. S8, beds abt 1 m. thick; cliff-forming. "lvI, 2.29 m. c.b. S8; basal part with sets 0.5 to U 3.94 m c.b. ISS, sets 20 em. thick; c.b. dips 1 ern. thick, overlain by S8 with 25° at 11". eroded current ripples, elay pellets T 1.50 m. 10 R 6/2 SS, few day pellets; top with and current drag; c.b. dips 310 0 current ripples from north; beds 0 at 16 ; overiain by e.b. ss sets 7.5 about 0.9 cm. thick. to 10 cm. thick. S 0.05 m. fine-gnd fiss ss; beds appro x 1 cm. 1., 0.46 m. one bed ss. thick; qz, musc, graph. MAXWELL It. BANKS AND IKRAM H. NAQVr 29

Unit Unit

R 1.60 nl. lned-gnd SSI qz, muse (rare) 1 feld a 2.59 m. med-fine gnd sitst, qz, muse, feId, (rare) clay pell (rare), c.b. beds 0.5 rare plant frags, non-fiss; irregular to 1 cm. up to 60 em.; 10 YR 5/4. blocky fracture. . Gap of 0.4 m. HDH Q 0.50 m. one bed ss. " C H_~' i\" p 015 m. SS, qz, musc, graph; beds <0.5 cm. J 0.90 m. SS, beds more than 30 em. thiclL thick. Road (approx 2.45 m. gap). Gap of 1.2 m. I 1.25 m. coarse-gndwi.th 1.80 m. graph; bedding feld'C<5%, muse , beds 10 cm. or morc 1.00 m. fine to feld H 0,50 m. fine-gnd qz, clay pell cong-lom. wiLlI dip G 1.50 m. med to feld ( Gap of 1 m. muse , (rare), M 4.60 m. med-gnd S8, qz, garnet (very , beds 10 ern. beds 15-60 em, or more thick; some conglom bands forming, 10 YR 6/B. 10 YH 6/6. L 0.20 m' friable white ss. Gap of 10.36 m. K 0.20 m. e.b. fi.ss ss, fine-gnd, qz, graph, muse, F 10.67 m. fine··gnd SS, qz, muse, beds less than one set, 1 em. thick, very fiss near base, less so near top 5 Y 5/2. Third step on track above road at "c (fig. 1 Cl. E l.50 m, med to coarse-gnd, carb sUst with qz, muse, feld, plant fmgs ('? sphen­ He "_" D H opsidJ, some beds fine-gnd ss and L80 m. SS, qz, musc, graph, elay pell, beds clay pell eonglom, somewhat fiss, 7.5 to 15 cm. thick. N4-N5. k 0.15 m. c.b. SS, qz, graph, muse. D 6AO m. very micaceous, fine-gnd SS, beds 1.70 m. med- coarse-gnd SS, beds 0.8-1 m. less 1 cm, thick, 5 Y 5/2. thick; with qz, graph, clay pell C 1.20 m. ss, beds 1-1.5 mm. thick, 5 Y 7/2. near top; near base sets 7.5 to B 1.70 m. ss, beds 1-2.5 cm. thick; otherwise 30 em., c,b. with current dmg; 10 as A. YR 6/6; ss. A 14.60 m. med. to coarse-gnd ss, qz, feld LSO m. fine-gnu S8, musc, qz, beds 0.5-2.5 <10%), muse, graph, 10 YR 8/4. em.

0.30 m, c.b. ss, sets 1-2.5 cm. thick. Base of Springs Sandstone on track from H G " g 0045 m. one set c.b. ss. to "A" on Pinnacle Road (Pillinger Drive). f 0.30 m. fine-gnd ss, qz, muse, graph, beds Thicknesses by abney levelling or by measure­ 1.0-2.5 em. ment on outcrop by steel tape. Gap of 0.75 m. Abbreviutions.-biot--biotite; c,b.-cross-bedded 0.60 m. med-fine gnd SS, qz, musc. graph, or cross-bedding; chlor-chlorite; conglom.-con­ 10 YR 6/4. glomerate; .feld-feldspar; ferrug-ferruginous; Gap of 0.30 m. fiss-fissile; jrags·---fragments; gnd--grained; graph -graphite; m.ed-medium; musc--muscovite; pell d 0.15 m. c,b. ss; sets 1-2 em. thick. --pellet; qz-~quartz; rnderl-rounded--sltst-silt- c 2.25 m. ss .. beds 75 cm, thick. stone; ss-sandstone. b {l.BO m, SS. Colour symbols quoted are from Rock Colour Gap of 3.5 m. Chart, Goddard et al. 30 SOME FORMATIONS CLOSE TO THE PERMO-TRIASSIC BOUNDARY IN TASMANIA

EXPLANATION OF PLATES.

1. Basal conglomerate with cross-bedding, Barnetts Member in type section; note coal lense immediately above it in right-hand side of photograph. 2. Coal lense between basal conglomerate and second unit of Barnetts Member. 3. Second unit of Barnetts Member in type section shown basal conglomerate at lower left passing up and right in interbedded feldspathic sandstone, carbonaceous siltstone and coal; second unit at top of photograph. 4. Clay pellet conglomerate, Barnetts Member, Coal Mine Bay. 5. Cross-bedding and ferruginous concretions, Barnetts Member, Coal Mine Bay. 6. Thinly-bedded sandstone showing diffusion rings' and weathering pattern, Barnetts Member, Coal Mine Bay. 7. Thinly-bedded sandstone, Barnetts Member, Coal Mine Bay. 8. Contact between Barnetts Member below and thickly-bedded, cross-bedded sandstone above as seen in cliff section at Coal Mine Bay 493.3, 677.42. Hammer lying immediately below contact.

[Photos by A. Goede for I. H. Naqvi. PAPli;H8 ANn PHOCF1EDINGH (W THE ROYAL ;SOCiETY 01" TASMANiA, Vm.L!:rfE: ·!01.

1 2

3 4

5 6

7 8