First evidence of Pentoxylales in Antarctica

First evidence of Pentoxylales in Antarctica

*S. N. CeÂsari, *{C. A. Parica, *M. B. Remesal

and *F. M. Salani

* CONICET, Departamento de GeologõÂa, Facultad de Ciencias Exactas y Naturales, PabelloÂn II, CiudadUniversitaria, (1428) Buenos Aires, Argentina{ Instituto AntaÂrtico Argentino, Cerrito 1248, (1010) Buenos Aires, Argentina

Revised manuscript accepted 26 February 1998

Plant remains from the Byers Peninsula, South Shetlands Islands, are described. These are leavesreferred to Taeniopteris sp. and female fructi®cations referred to Carnoconites llambiasii CeÂsari sp. nov.They are considered to belong to the Pentoxylales, originally described from India, Australia and NewZealand. This is the ®rst record of this group of gymnosperms from Antarctica. The occurrence con-tributes to further differentiation between ¯oras of this part of Gondwana and Laurasia.

# 1998 Academic Press Limited

KEY WORDS: Cretaceous; Antarctica; Pentoxylales; systematics.

1. Introduction

The Pentoxylales comprises a small group of Jurassic and Early Cretaceous

gymnosperms that has been encountered in a few localities in New Zealand,

Australia and India. According to Bose et al. (1985), these plants were small trees

or shrubs with both long and three kinds of short shoots. Short shoots bear leaves

and reproductive structures. Pentoxylon is the name of these stems, which are

known from structurally preserved specimens. The petiolate leaves are of

Taeniopteris-type, if preserved without structure, with a broad midrib and lateral

veins running at right angles, occasionally dichotomising. Nipaniophyllum is the

generic name used for anatomically preserved Taeniopteris specimens. The pollen

organs are referred to Sahnia, and consist of numerous elongate structures

(microsporophylls or microsporangiophores) arising from a collar-like structure

surrounding the conical shoot apex. Two types of organization were proposed for

the seed-bearing organs called Carnoconites. According to Bose et al. (1985) they

consist of a central branching axis with each peduncle bearing spirally inserted

pedicellate cones. Nevertheless, Harris (1962) and Drinnan & Chambers (1985)

presumed that the pedicellate cones are attached individually to the same shoot.

The Pentoxylales can be probably grouped into two types of plants, one of

Pentoxylon sahnii stems bearing Nipaniophyllum raoi leaves and Carnoconitescompactus cones; and the other of Taeniopteris daintreei leaves, Carnoconitescranwellii cones and Sahnia laxiphora pollen-bearing organs.

The ®nding of Taeniopteris leaves and Carnoconites cones in the Byers Peninsula,

extends the geographical distribution of the Pentoxylales and contributes to our

knowledge of the arrangement of the cones. Some years ago, Drinnan &

Chambers (1985) emphasised that the discovery of the Pentoxylales in other

Gondwanan localities will necessitate a revision of the Mesozoic phytogeography.

Cretaceous Research (1998) 19, 733±743 Article No. cr980128

0195±6671/98/060733 + 11 $30.00/0 # 1998 Academic Press

Hence, the palaeogeographic signi®cance of the new specimens from Antarctica is

discussed in this paper.

The specimens, preserved as imprints, are catalogued BAFCPB and housed in

the Palaeobotanical Collection of the Department of Geology, Buenos Aires

University.

2. Stratigraphic setting

The specimens studied have been recovered by three of us (CAP, MBR and

FMS) from outcrops in the western side of Zigzag Gully, Byers Peninsula

(Figure 1). This locality was named False Cerro Negro by Askin (1983) and

probably corresponds to locality P.-2237 of Crame et al. (1993). Hathway &

Lomas (1998) have presented a new stratigraphic scheme for the Byers Peninsula;

it is modi®ed from Crame et al. (1993). According to this new stratigraphy, the

Byers Group includes ®ve units, the lowest of which is the Anchorage Formation.

The overlying unit is the marine President Beaches Formation, which in turn is

overlain unconformably by the Chester Cone and Start Hill Formations. The

uppermost part of the Byers Group consists of thick non-marine strata of the

Cerro Negro Formation (Hathway, 1997), which rest unconformably on the

Chester Cone Formation. The fossiliferous sequence bearing Pentoxylales is

included by Hathway (op. cit.) in the Cerro Negro Formation, corresponding to

non-marine volcaniclastic strata, which rests with low angle erosive unconformity

on a clastic marine sequence of gravitational ¯ows with very fragmented remains

of ammonites. Palynomorphs (Askin, 1983) and plant macrofossils (HernaÂndez &

AzcaÂrate, 1971) were recovered from the upper part of the Cerro Negro

Formation to the southwest of Cerro Negro. The rich ¯ora with Pentoxylales

from Zigzag Gully and Cerro Negro, will be described in detail by CeÂsari et al. (in

prep).

Crame et al. (1993) mentioned the presence of an abundant ¯ora at an

exposure to the NW of Sealer Hill (P.-2237, ®g. 2 in Crame et al., 1993) that

includes sphenopsids, ferns such as Coniopteris, Phlebopteris and Phyllopteroides,and Bennettitales (Ptilophyllum). The plant-bearing strata are referred to

``Crystalline tuff '' that rests with apparent unconformity on the Sealer Hill

Figure 1. Location map.

734 S. N. CeÂsari et al.

Member of the Byers Group (Crame et al., 1993) and typically contains silici®ed

trunks. A comparison of the location maps of Crame et al. (1993) and Hathway

(1997) suggests that the P.-2237 and Zigzag Gully fossiliferous localities are

probably identical.

3. Previous records of Pentoxylales

Sahni (1948) originally described the Pentoxyleae as a group of extinct

gymnosperms from Jurassic-Cretaceous sediments in the Rajmahal Hills, India

(Figure 2). This material was later revised by Bose et al. (1985). The ¯oras of the

Rajmahal Hills consist of a number of plant beds associated with lava ¯ows which,

according to Bose et al. (1990), may be divided into two main ¯oras. Cycadophytic

remains dominate the ®rst type of assemblage. The second ¯oral assemblage is best

represented at Nipania (in the Rajmahal Hills) and includes Pentoxylales, conifer

foliage (Brachyphyllum and Elatocladus) and a few cycadophytic remains.

Drinnan & Chambers (1985) ascribed Taeniopteris daintreei to the Order

Pentoxylales based on the association of these leaves with female cones of

Figure 2. Palaeogeographic reconstruction of Gondwana in the Early Cretaceous, showing the fos-siliferous localities bearing pentoxylalean remains; based on Grunow (1993).

First evidence of Pentoxylales in Antarctica 735

Carnoconites cranwellii and pollen organs of Sahnia laxiphora from the Koonwarra

¯ora of Victoria, Australia (Figure 2). According to Drinnan & Chambers (1986)

the vegetation represented by this ¯ora was dominated by a forest of Ginkgo and

several conifers, with an understorey of pentoxylaleans, ferns, sphenophytes and

bryophytes in a cool, possibly montane environment.

White (1981) illustrated leaves of Taeniopteris spatulata McClelland, from the

Talbragar Fish Beds of New South Wales, apparently connected to Pentoxylales

cones. She proposed a new taxon, Pentoxylon australica, for leaves of T. spatulata-type. Female cones of Carnoconites and male microsporangial groups were also

found in association.

Sedimentary strata at Port Waikato, New Zealand (Figure 2), deposited during

the Middle Tithonian (Late Jurassic), contain specimens of T. daintreei and C.cranwellii (Harris, 1962). This succession in other localities also contains the

coniferous fossils Elatocladus and Araucarites, the ferns Cladophlebis and

Coniopteris, and the Bennettitales Ptilophyllum and Pterophyllum (McQueen, 1956).

4. Systematic descriptions

Genus Taeniopteris Brongniart 1828Taeniopteris sp.

Figure 3a-e

Description. Fragments of leaves simple, linear, petiolate, and variable in size,

ranging from 5 to 13 mm wide (exceptionally as narrow as 1.8 mm). Petiole

2 mm wide, continuing into the leaf lamina as a prominent and striated midvein

(0.6-12 mm). Secondary veins parallel, arising from the midvein at 60�-90�, most

dichotomising once near the midrib, or twice before reaching the margin.

Secondary vein density: 30-34 per 10 mm of lamina measured midway between

the midvein and leaf margin. Leaf margins entire, tapering gradually to an acute

base and terminating distally with an acute apex.

Studied specimens. BAFCPB 16087, 16090, 16091, 16099, 16100, 16111, 16118.

Comparisons. A comparable species, Taeniopteris daintreei McCoy, was ®rst

described and ®gured from the Lower Cretaceous of Victoria by McCoy in 1874.

Later, Chapman (1908, 1909) considered this species to be equivalent to

Taeniopteris spatulata (=Nipaniophyllum raoi Sahni) from India, a view supported

by cuticular evidence (Vishnu-Mittre, 1957), because both species have

haplocheilic stomata. Drinnan & Chambers (1985) preferred the form genus

Taeniopteris Brongniart to Nipaniophyllum Sahni for the Victorian specimens,

because the diagnosis for the latter describes in detail vascular anatomy unknown

in the Australian leaves. These authors proposed a size range of between 100 mm

long by 10 mm wide and 300 mm by 30 mm, a vein density of 15-30 per 10 mm

of lamina, and haplocheilic stomata surrounded by 5-8 papillae on the lower leaf

surface. They pointed out that the narrower leaves tend to have secondary veins,

which are more sparsely arranged and are not strictly perpendicular to the

midvein. This combination of characters has not been observed in Antarctic

specimens, where the smaller and narrower leaves have a high density of

veins.The variation in leaf size and shape noted by Vishnu-Mittre (1957) for

Nipaniophyllum raoi is similar to, and within the range reported by, Drinnan &

Chambers (1985) for T. daintreei. However, according to Bose et al. (1985), N.raoi is 4-6 mm wide and has a sharp or obtuse pointed apex where the median

736 S. N. CeÂsari et al.

vein emerges as a minute mucro. Nipaniophyllum hobsonii Bose et al. 1985 is a

wider leaf (10-14 mm) with a contracted and obtusely pointed apex. According to

Bose & Banerji (1981) Taeniopteris spatulata McClelland includes impressions of

leaves up to 1.6 cm in width (usually 0.5-0.8 cm), with an obtuse apex and lateral

veins emerging at 60�-90� at a density ranging from 18-40, but usually 20-30 per

cm. The obtuse apex distinguishes this species from the species of Taeniopteris

described above. The specimens that White (1981) reported from the Talbragar

Fish Beds in Australia are similar in width (0.25-1 cm), and have lateral veins

emerging at right angles at a density of 25-30 per cm; however, it differs from T.

sp. in having an acuminate or bluntly-rounded apex.

HernaÂndez & AzcaÂrate (1971) described specimens referred to Taeniopteris sp.

that are closely similar to those under discussion here. These authors established

a comparison, based on the size, with specimens from the Baquero Formation

described by Archangelsky (1965). Leaves from Santa Cruz have a sharply

pointed apex, a maximum width of 1 cm, and 20-30 lateral veins per cm that are

forked once near the central vein. Baldoni & Ramos (1981) also described small

leaves of Taeniopteris up to 7 mm in maximum width and with a density of 20-25

Figure 3. Taeniopteris sp. a, Specimen BAFCPB 16099, x 3. b, Specimen BAFCPB 16100, showingacute apex, x 3. c, Specimen BAFCPB 16111, detail of venation, x 4. d, Detail of venation,BAFCPB 16090, x 4. e, Specimen BAFCPB 16091 showing petiolate base, x 2.

First evidence of Pentoxylales in Antarctica 737

veins per cm. We believe that all of these latter specimens could be identical to

the species from the Byers Peninsula.

Genus Carnoconites Srivastava 1944

Carnoconites llambiasii CeÂsari sp.nov.

Figure 4a-f

Diagnosis. Seed bearing organs arranged along a main axis 0.7 mm wide. Apical

cone is directly inserted on the main axis; the others are apparently spirally

inserted on very short secondary axes. Cones are oval in longitudinal outline,

compact, about 6 mm long and 5 mm wide; in nearly median longitudinal section

they show an axis 0.6 mm thick. Seeds apparently sessile, about 1 mm long, 30-

35 per cone, densely arranged in a spiral, nearly ovate in longitudinal section,

with rounded outline.

Holotype. BAFCPB 16118, Figure 4a-f.

Type locality. Western side of Zigzag Gully, Byers Peninsula, Livingston Island,

South Shetland Islands.

Derivatio nominis. After Mr Eduardo LlambõÂas, in recognition of his invaluable

assistance during the ®eld season.

Remarks. A Taeniopteris leaf 1.8 mm wide, with a thick midvein and laterals at

about 90� to the margin, is in close association on the same bedding plane,

adjacent to the fertile organ.

Although the exact arrangement of the cones is uncertain, it is assumed from

their close disposition (exactly as a spiral) that short pedicels attached them to the

same axis or peduncle. The possibility that the specimen represents a clump of

pedicellate cones that are not attached to each other seems less likely.

Comparisons. Three species of Carnoconites have been described previously: C.

compactus Srivastava, C. cranwellii Harris and C. rajmahalensis Bose et al. Sahni

(1948) described the cones of C. compactus as being borne in pairs on a

bifurcating stalk, but Vishnu-Mittre (1953) interpreted pedunculate cones arising

singly in spiral succession from a simple axis. Nearly all the specimens of C.

compactus that have been ®gured are 10-11 mm wide, but one cited by Vishnu-

Mittre is 7 mm wide. In transverse section, the organ is elliptical with an acute

apex.

Harris (1962) illustrated up to 12 pedunculate cones arising in a group from

near the apex of the main fertile axis of C. cranwellii from New Zealand, and

described the seeds as rhombic in section. This description correlates closely with

that of the Australian, Koonwarra specimens (Drinnan & Chambers, 1985), the

cones of which are 5 mm wide, and the seeds rhombic or hexagonal in transverse

section; the arrangement of seeds on the axis is either in spiral succession or in

alternating whorls. The specimens from both New Zealand and Australia differ

from C. llambiasii in having long naked axes and seeds that are rhombic in cross-

section.

Bose et al. (1984) examined one imprint described by Feistmantel (1877, pl.

39, ®gs 5, 5a, 5b) from the Rajmahal Hills and referred by Wieland (1911) to

Williamsonia (?) rajmahalensis. They considered it to be equivalent to petri®ed

specimens of Carnoconites laxum, and proposed the new combination C.

rajmahalensis (Wieland) Bose et al. to accommodate it. The cones of this species

738 S. N. CeÂsari et al.

Figure 4. Carnoconites llambiasii CeÂsari sp. nov. Holotype, BAFCPB 16118. a, Detail of the distalcones, x 8. b, Detail of associated leaf with a distinctive midrib and a high density of lateralveins, x 6. c, Nearly longitudinal section of one cone, x 10. d, Longitudinal section of one coneshowing the central axis, x 10. e, Main axis bearing an apical cone and two short stalked cones, x5. f, General view showing seven cones and the Taeniopteris leaf , x 3.

First evidence of Pentoxylales in Antarctica 739

are elongate, up to 5 mm wide and 25 mm long, and the seeds in transverse

section are broadly oval to subspherical. Both C. compactus and C. rajmahalensisare interpreted as pedunculate cones bearing seeds in a more or less dense spiral

arrangement without associated interseminal scales or bracts. The arrangement of

the cones is similar to that in C. llambiasii (Bose et al., 1985, ®gs 36, 40).

Nevertheless the cones are longer than those of C. llambiasii and are apparently

inserted on longer axes.

White (1981) described as Carnoconites sp. fructi®cations associated with

Taeniopteris spatulata from the Talbragar Fish Beds (Australia). These cones are

bigger and contain fewer seeds than Carnoconites llambiasii.We prefer for the moment to separate the Antarctic species according to the

small size of the cones, apparent helicoidal arrangement around a main stem

through short secondary axes, and the rounded outline of the seeds in

longitudinal and transverse section.

5. Age

According to Drinnan & Chambers (1985), the age range of the Pentoxylales

group is uncertain. The Rajmahal Series was considered to be Middle to Late

Jurassic by early Indian palaeobotanists, but ammonites seemed to suggest a

Neocomian age (Arkell, 1956). Palynological evidence also indicated a

Neocomian and possibly post-Barremian age (Playford & Cornelius, 1967;

Srivastava, 1983), and analysis of K-Ar ratios of lava traps suggested an Albian

age (McDougall & McElhinny, 1970). An Early Cretaceous age was assigned

based on subsurface correlation in the region (Das-Gupta, 1974).

Pentoxylean specimens from the Koonwarra ¯ora were determined to be

Valanginian-Aptian based on palynological analysis (Dettmann, 1963) and ®ssion

track dating (Gleadow & Duddy, 1980). The Talbragar Fish Beds are considered

to be late Early Jurassic (Hind & Helby, 1969; Loughnan & Evans, 1978). The

¯ora from Waikato Heads is latest Jurassic (Harris, 1962). Drinnan & Chambers

(1985) considered the Taeniopteris and Carnoconites taxa from this locality to be

conspeci®c with those that persist through to the Albian in south-eastern

Australia.

An Early Cretaceous age is suggested for the strata bearing Pentoxylales in

Antarctica. Palynological analysis of the President Beaches and Chester Cone

Formations has yielded an Early Berriasian-Valanginian assemblage (Duane,

1994), indicating a maximum age for the plant-bearing unit. In addition, the

palyno¯oras of the Byers Group suggest phytogeographic af®nities with the

western side of Australia and Patagonia, the palynological species of continental

origin showing, in particular, a strong relationship with assemblages from the

Baquero Formation of Santa Cruz (Duane, 1994). According to HernaÂndez &

AzcaÂrate (1971) and Askin (1983), plant macrofossils and palynomorphs from the

upper part of the Cerro Negro Formation, southwest of Cerro Negro, suggest a

Barremian age for the unit. Recent results from a basal tuff have yielded 40Ar/39Ar

ages of 120.3 � 2.2 Ma on plagioclase and 119.4 � 0.6 and 119.1 � 0.8 Ma on

biotite and plagioclase from an ignimbrite clast in the basal conglomerates.

Plagioclase from the upper part of the formation has given a 40Ar/39Ar age of

119 � 3 Ma. According to Hathway (1997), these dates indicate an earliest

Aptian age for the Cerro Negro Formation.

740 S. N. CeÂsari et al.

6. Discussion

Barale et al. (1995) suggested the existence of an emerged landmass since Late

Triassic times, at least in this sector of Antarctica. The recognition of a Triassic

¯ora at Williams Point in Livingston Island supports this suggestion.

Nevertheless, Rees & Smellie (1989) proposed an early Late Cretaceous age for

all plant-bearing outcrops at Williams Point. This proposal was rejected by Barale

et al. (1995) in the course of discussing the presence of angiosperms.

At the beginning of the Cretaceous, Australia, Antarctica, New Zealand, and a

large part of India were joined as part of the Gondwana landmass. Dettmann

(1992) suggested that during the Early Cretaceous it comprised two broad

¯oristic regions. In the northernmost areas (i.e., South America and Africa), the

cheirolepidiacean conifers were important in the vegetation whereas in southern

Gondwana (Antarctica, Australasia, southernmost South America and India) the

vegetation was characterised by a variety of podocarp/araucarian forests. In the

northern part of this sector (India, Patagonia, the Antarctic Peninsula and

northern Australia), araucarians were more important in the canopy than the

podocarps.

Our record of the Pentoxylales from the Byers Peninsula indicates a greater

similarity between ¯oras of India (Rajmahal Series), Australia (Victorian,

Talbragar) and New Zealand than elsewhere during the Jurassic-Early Cretaceous

(Figure 2). According to the palaeogeographic reconstruction, two of the regions

(Victoria and Port Waikato) occupied a similar palaeolatitudinal position, and

their interrelationship seems close. However, the Indian plate and South Shetland

Islands, occupied a slightly more northerly position.

When Harris (1962) reported fertile pentoxylalean organs from New Zealand,

he pointed out the potential importance of this group during the Jurassic and

Cretaceous for differentiating Gondwanan ¯oras from those of Laurasia where the

group was not identi®ed. The Antarctic occurrence contributes to further

differentiation of the ¯oras and enhances the importance of the Pentoxylales as a

characteristic group of this southernmost region of Gondwana during Jurassic-

Early Cretaceous times.

Acknowledgements

We are very grateful to Dr S. Archangelsky for reading an early version of the

manuscript and to Dr D. J. Cantrill and Prof. D. J. Batten for their valuable

suggestions as reviewers.

References

Archangelsky, S. 1965. Dos nuevas localidades con plantas foÂsiles del Baqueroense (CretaÂcicoinferior). Revista del Museo de La Plata (Nueva Serie) 4, 247±257.

Arkell, W. J. 1956. Jurassic geology of the world, 806 pp. (Oliver & Boyd, London).Askin, R. A. 1983. Tithonian (uppermost Jurassic)-Barremian (Lower Cretaceous) spores, pollen and

microplankton from the South Shetland Islands, Antarctica. In Antarctic earth science (eds Oliver,R. L., James, P. R. & Jago, J. B.), pp. 295±297 (Australian Academy Sciences, Canberra).

Baldoni, A. M. & Ramos, V. A. 1981. Nuevas localidades con plantas foÂsiles CretaÂcicas en la Cordil-lera PatagoÂnica (provincia de Santa Cruz, RepuÂblica Argentina). Actas del VIII Congreso GeoloÂgicoArgentino 4, 743±759.

Barale, G., Philippe, M., Torres, T. & Thevenard, F. 1995. Reappraisal of the Triassic ¯ora from Wil-liams Point, Livingston Island (South Shetland Islands, Antarctica): systematical, biostratigraphi-

First evidence of Pentoxylales in Antarctica 741

cal and paleogeographical implications. Serie Cientõ®ca del Instituto AntaÂrtico Chileno (INACH) 45,9±38.

Bose, M. N. & Banerji, K. 1981. Cycadophytic leaves from Jurassic-Lower Cretaceous rocks of India.The Palaeobotanist 28±29, 218±300.

Bose, M. N., Pal, P. K. & Harris, T. M. 1984. Carnoconites rajmahalensis (Wieland) comb. nov. fromthe Jurassic of Rajmahal Hills, India. The Palaeobotanist 32, 368±369.

Bose, M. N., Pal, P. K. & Harris, T. M. 1985. The Pentoxylon plant. Philosophical Transactions of theRoyal Society of London B 310, 77±108.

Bose, M. N., Taylor, E. & Taylor, T. 1990. Gondwana ¯oras of India and Antarctica. A survey andappraisal. In Antarctic paleobiology. Its role in the reconstruction of Gondwana (eds Taylor, T. & Tay-lor, E.), pp. 119±148 (Springer-Verlag, Berlin).

Brongniart, A. 1828. Prodrome d'une histoire des veÂgeÂtaux fossiles. Dictionnaire des Sciences Naturelles57, 16±212.

Chapman, F. 1908. Report of Jurassic plants. Records of the Geological Survey of Victoria 2, 212±220.Chapman, F. 1909. Jurassic plant remains from Gippsland. Records of the Geological Survey of Victoria

3, 103±111.Crame, J. A., Pirrie, D., Crampton, J. S. & Duane, A. M. 1993. Stratigraphy and regional signi®cance

of the Upper Jurassic-Lower Cretaceous Byers Group, Livingston Island, Antarctica. Journal of theGeological Society, London 150, 1075±1087.

Das-Gupta, S. K. 1974. Reclassi®cation of the Mesozoic-Tertiary stratigraphy of the Jaisalmer Basin,Rajasthan. Proceedings of the Indian Science Congress, Abstracts 12, 135.

Dettmann, M. E. 1963. Upper Mesozoic micro¯oras from south-eastern Australia. Proceedings of theRoyal Society of Victoria 77, 1±40.

Dettmann, M. E. 1992. Structure and ¯oristics of Cretaceous vegetation of southern Gondwana:implications for angiosperm biogeography. The Palaeobotanist 41, 224±233.

Drinnan, A. N. & Chambers, T. C. 1985. A reassessment of Taeniopteris daintreei from the VictorianEarly Cretaceous: a member of the Pentoxylales and a signi®cant Gondwanaland plant. AustralianJournal of Botany 33, 89±100.

Drinnan, A. N. & Chambers, T. C. 1986. Flora of the Lower Cretaceous Koonwarra fossil beds (Kor-umburra Group), South Gippsland, Victoria. Memoirs of the Association of Australasian Palaeontolo-gists 3, 1±77.

Duane, A. M. 1994. Preliminary palynological investigation of the Byers Group (Late Jurassic-EarlyCretaceous) Livingston Islands, Antarctic Peninsula. Review of Palaeobotany and Palynology 84,113±120.

Feistmantel, O. 1877. Jurassic (Liassic) ¯ora of the Rajmahal Group, in the Rajmahal Hills. Memoirs ofthe Geological Survey of India, Palaeontologia Indica Series 2 1, 53±162.

Gleadow, A. J. W. & Duddy, I. R. 1980. Early Cretaceous volcanism and the early breakup history ofsoutheastern Australia: evidence from ®ssion track dating of volcaniclastic sediments. In Proceed-ings of the 5th International Gondwana Symposium (eds Cresswell, M. M. & Vella, P.), pp. 295±300(A. A. Balkema, Rotherdam).

Grunow, A. M. 1993. New paleomagnetic data from the Antarctic Peninsula and their tectonic impli-cations. Journal of Geophysical Research 98, 13 815±13833.

Harris, T. M. 1962. The occurrence of the fructi®cation Carnoconites in New Zealand. Transactions ofthe Royal Society of New Zealand (Geology) 1, 17±27.

Hathway, B. 1997. Nonmarine sedimentation in an Early Cretaceous extensional continental-marginarc, Byers Peninsula, Livingston Island, South Shetland Islands. Journal of Sedimentary Research67, 686±697.

Hathway, B. & Lomas, S. 1998. The uppermost Jurassic-Lower Cretaceous Byers Group, South Shet-land Islands, Antarctica: revised stratigraphy and regional correlations. Cretaceous Research 19, 43±67.

HernaÂndez, P. & AzcaÂrate, V. 1971. Estudio paleobotaÂnico preliminar sobre restos de una tafo¯ora dela PenõÂnsula Byers (Cerro Negro), Isla Livingston, Islas Shetlands del Sur, AntaÂrctica. Serie CientõÂ-®ca del Instituto AntaÂrtico Chileno 2, 15±50.

Hind, M. C. & Helby, R. J. 1969. The Great Artesian Basin in New South Wales. In The geology ofNew South Wales (ed. Packham, G. H.), Journal of the Geological Society of Australia 16, 481±497.

Loughnan, F. C. & Evans, P. R. 1978. The Permian and Mesozoic of the Merriwa-Binnaway-Balli-more area, New South Wales. Journal and Proceedings of the Royal Society of New South Wales 111,107±119.

McCoy, F. 1874. Prodromus of the palaeontology of Victoria, 270 pp. (Geological Survey of Victoria,Melbourne).

McDougall, I. & McElhinny, M. W. 1970. The Rajmahal traps of India K-Ar ages and palaeomagnet-ism. Earth and Planetary Sciences Letters 9, 371±378.

McQueen, D. R. 1956. Leaves of Middle and Upper Cretaceous pteridophytes and cycads from NewZealand. Transactions of the Royal Society of New Zealand 83, 673±685.

742 S. N. CeÂsari et al.

Playford, G. & Cornelius, K. D. 1967. Palynological and lithostratigraphic features of the RazorbackBeds, Mount Morgan District, Queensland. University of Queensland, Department of Geology, Papers6, 81±94.

Rees, P. M. & Smellie, J. L. 1989. Cretaceous angiosperms from an allegedly Triassic ¯ora at WilliamsPoint, Livingston Island, South Shetlands. Antarctic Science 1, 239±248.

Sahni, B. 1948. The Pentoxyleae. A new group of Jurassic gymnosperms from the Rajmahal Hills ofIndia. Botanical Gazette 110, 47±80.

Srivastava, B. P. 1944. Silici®ed plant remains from the Rajmahal Hills. Palaeobotany in India 5. Pro-ceedings of the National Academy of Sciences of India 14, 73±76.

Srivastava, S. K. 1983. Cretaceous phytogeoprovinces and paleogeography of the Indian plate basedon palynological data. In Proceedings of the Symposium on Cretaceous of India: palaeoecology, palaeo-geography and time boundaries, pp. 141±157 (Indian Association of Palynostratigraphers, Luck-now).

Vishnu-Mittre, 1953. A male ¯ower of the Pentoxyleae, with remarks on the female cones of thegroup. The Palaeobotanist 2, 78±84.

Vishnu-Mittre, 1957. Studies in the fossil ¯ora of Nipania (Rajmahal Series) India. Pentoxyleae. ThePalaeobotanist 6, 31±45.

White, M. E. 1981. Revision of the Talbragar Fish Bed ¯ora (Jurassic) of New South Wales. Records ofthe Australian Museum 33, 695±721.

Wieland, G. R. 1911. On the Willamsonia tribe. American Journal of Science, Series 4 32, 433±466.

First evidence of Pentoxylales in Antarctica 743